From 8769177658d3559c4323200a719dd456d2f2675a Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Sat, 18 Mar 2023 17:24:27 +0100 Subject: [PATCH 001/186] ARM: vfp: Record VFP bounces as perf emulation faults VFP 'bouncing' occurs when the VFP unit cannot complete the execution of a VFP instruction, either because it is not implemented at all, or because the values of the arguments are out of range for the hardware implementation, and the software needs to step in to complete the operation. To give some insight in how much certain programs rely on this bouncing, record the emulation of a VFP instruction in perf's emulation-faults counter. This can be used like so perf stat -e emulation-faults ./testfloat -all2 and the output will be something like Performance counter stats for './testfloat -all2': 259,277 emulation-faults:u 6.846432176 seconds time elapsed Reviewed-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/vfp/vfpmodule.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 349dcb944a937..08d5dfcf70796 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -313,6 +314,7 @@ static u32 vfp_emulate_instruction(u32 inst, u32 fpscr, struct pt_regs *regs) * emulate it. */ } + perf_sw_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, regs, regs->ARM_pc); return exceptions & ~VFP_NAN_FLAG; } From 4a0548c6681cd25b8d76e897e01bfb62ce93916d Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Mon, 20 Mar 2023 11:01:16 +0100 Subject: [PATCH 002/186] ARM: vfp: Remove workaround for Feroceon CPUs Feroceon CPUs have a non-standard implementation of VFP which reports synchronous VFP exceptions using the async VFP flag. This requires a workaround which is difficult to reconcile with other implementations, making it tricky to support both versions in a single image. Since this is a v5 CPU, it is not supported by armhf and so the likelihood that anybody is using this with recent distros/kernels and rely on the VFP at the same time is extremely low. So let's just disable VFP support on these cores, so we can remove the workaround. This will help future development to support v5 and v6 CPUs with a single kernel image. Reviewed-by: Linus Walleij Acked-by: Nicolas Pitre Acked-by: Arnd Bergmann Signed-off-by: Ard Biesheuvel --- arch/arm/mm/proc-feroceon.S | 4 ++++ arch/arm/vfp/vfphw.S | 4 ---- arch/arm/vfp/vfpmodule.c | 8 +++++--- 3 files changed, 9 insertions(+), 7 deletions(-) diff --git a/arch/arm/mm/proc-feroceon.S b/arch/arm/mm/proc-feroceon.S index 61ce82aca6f0d..072ff9b451f84 100644 --- a/arch/arm/mm/proc-feroceon.S +++ b/arch/arm/mm/proc-feroceon.S @@ -56,6 +56,10 @@ ENTRY(cpu_feroceon_proc_init) movne r2, r2, lsr #2 @ turned into # of sets sub r2, r2, #(1 << 5) stmia r1, {r2, r3} +#ifdef CONFIG_VFP + mov r1, #1 @ disable quirky VFP + str_l r1, VFP_arch_feroceon, r2 +#endif ret lr /* diff --git a/arch/arm/vfp/vfphw.S b/arch/arm/vfp/vfphw.S index a4610d0f32152..0aeb60ac3b537 100644 --- a/arch/arm/vfp/vfphw.S +++ b/arch/arm/vfp/vfphw.S @@ -110,7 +110,6 @@ ENTRY(vfp_support_entry) beq vfp_reload_hw @ then the hw state needs reloading VFPFSTMIA r4, r5 @ save the working registers VFPFMRX r5, FPSCR @ current status -#ifndef CONFIG_CPU_FEROCEON tst r1, #FPEXC_EX @ is there additional state to save? beq 1f VFPFMRX r6, FPINST @ FPINST (only if FPEXC.EX is set) @@ -118,7 +117,6 @@ ENTRY(vfp_support_entry) beq 1f VFPFMRX r8, FPINST2 @ FPINST2 if needed (and present) 1: -#endif stmia r4, {r1, r5, r6, r8} @ save FPEXC, FPSCR, FPINST, FPINST2 vfp_reload_hw: @@ -153,7 +151,6 @@ vfp_reload_hw: VFPFLDMIA r10, r5 @ reload the working registers while @ FPEXC is in a safe state ldmia r10, {r1, r5, r6, r8} @ load FPEXC, FPSCR, FPINST, FPINST2 -#ifndef CONFIG_CPU_FEROCEON tst r1, #FPEXC_EX @ is there additional state to restore? beq 1f VFPFMXR FPINST, r6 @ restore FPINST (only if FPEXC.EX is set) @@ -161,7 +158,6 @@ vfp_reload_hw: beq 1f VFPFMXR FPINST2, r8 @ FPINST2 if needed (and present) 1: -#endif VFPFMXR FPSCR, r5 @ restore status @ The context stored in the VFP hardware is up to date with this thread diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 08d5dfcf70796..95628e57807b1 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -42,7 +42,11 @@ static bool have_vfp __ro_after_init; * Used in startup: set to non-zero if VFP checks fail * After startup, holds VFP architecture */ -static unsigned int __initdata VFP_arch; +static unsigned int VFP_arch; + +#ifdef CONFIG_CPU_FEROCEON +extern unsigned int VFP_arch_feroceon __alias(VFP_arch); +#endif /* * The pointer to the vfpstate structure of the thread which currently @@ -357,14 +361,12 @@ void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) } if (fpexc & FPEXC_EX) { -#ifndef CONFIG_CPU_FEROCEON /* * Asynchronous exception. The instruction is read from FPINST * and the interrupted instruction has to be restarted. */ trigger = fmrx(FPINST); regs->ARM_pc -= 4; -#endif } else if (!(fpexc & FPEXC_DEX)) { /* * Illegal combination of bits. It can be caused by an From 4708fb041346fa9cc6745dafb8c248a3e2f1075b Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Thu, 16 Mar 2023 00:51:42 +0100 Subject: [PATCH 003/186] ARM: vfp: Reimplement VFP exception entry in C code En/disabling softirqs from asm code turned out to be trickier than expected, so vfp_support_entry now returns by tail calling __local_enable_bh_ip() and passing the same arguments that a C call to local_bh_enable() would pass. However, this is slightly hacky, as we don't want to carry our own implementation of local_bh_enable(). So let's bite the bullet, and get rid of the asm logic in vfp_support_entry that reasons about whether or not to save and/or reload the VFP state, and about whether or not an FP exception is pending, and only keep the VFP loading logic as a function that is callable from C. Replicate the removed logic in vfp_entry(), and use the exact same reasoning as in the asm code. To emphasize the correspondence, retain some of the asm comments in the C version as well. Signed-off-by: Ard Biesheuvel Acked-by: Linus Walleij --- arch/arm/vfp/entry.S | 12 +-- arch/arm/vfp/vfp.h | 1 + arch/arm/vfp/vfphw.S | 204 +++------------------------------------ arch/arm/vfp/vfpmodule.c | 123 +++++++++++++++++++---- 4 files changed, 124 insertions(+), 216 deletions(-) diff --git a/arch/arm/vfp/entry.S b/arch/arm/vfp/entry.S index 62206ef250371..547c94c62cd3a 100644 --- a/arch/arm/vfp/entry.S +++ b/arch/arm/vfp/entry.S @@ -22,10 +22,10 @@ @ IRQs enabled. @ ENTRY(do_vfp) - mov r1, r10 - str lr, [sp, #-8]! - add r3, sp, #4 - str r9, [r3] - bl vfp_entry - ldr pc, [sp], #8 + mov r1, r0 @ pass trigger opcode via R1 + mov r0, sp @ pass struct pt_regs via R0 + bl vfp_support_entry @ dispatch the VFP exception + cmp r0, #0 @ handled successfully? + reteq r9 @ then use R9 as return address + ret lr @ pass to undef handler ENDPROC(do_vfp) diff --git a/arch/arm/vfp/vfp.h b/arch/arm/vfp/vfp.h index 5cd6d50532717..e43a630f8a164 100644 --- a/arch/arm/vfp/vfp.h +++ b/arch/arm/vfp/vfp.h @@ -375,3 +375,4 @@ struct op { }; asmlinkage void vfp_save_state(void *location, u32 fpexc); +asmlinkage u32 vfp_load_state(const void *location); diff --git a/arch/arm/vfp/vfphw.S b/arch/arm/vfp/vfphw.S index 0aeb60ac3b537..d5a03f3c10c50 100644 --- a/arch/arm/vfp/vfphw.S +++ b/arch/arm/vfp/vfphw.S @@ -4,12 +4,6 @@ * * Copyright (C) 2004 ARM Limited. * Written by Deep Blue Solutions Limited. - * - * This code is called from the kernel's undefined instruction trap. - * r1 holds the thread_info pointer - * r3 holds the return address for successful handling. - * lr holds the return address for unrecognised instructions. - * sp points to a struct pt_regs (as defined in include/asm/proc/ptrace.h) */ #include #include @@ -19,20 +13,6 @@ #include #include - .macro DBGSTR, str -#ifdef DEBUG - stmfd sp!, {r0-r3, ip, lr} - ldr r0, =1f - bl _printk - ldmfd sp!, {r0-r3, ip, lr} - - .pushsection .rodata, "a" -1: .ascii KERN_DEBUG "VFP: \str\n" - .byte 0 - .previous -#endif - .endm - .macro DBGSTR1, str, arg #ifdef DEBUG stmfd sp!, {r0-r3, ip, lr} @@ -48,177 +28,25 @@ #endif .endm - .macro DBGSTR3, str, arg1, arg2, arg3 -#ifdef DEBUG - stmfd sp!, {r0-r3, ip, lr} - mov r3, \arg3 - mov r2, \arg2 - mov r1, \arg1 - ldr r0, =1f - bl _printk - ldmfd sp!, {r0-r3, ip, lr} - - .pushsection .rodata, "a" -1: .ascii KERN_DEBUG "VFP: \str\n" - .byte 0 - .previous -#endif - .endm - - -@ VFP hardware support entry point. -@ -@ r0 = instruction opcode (32-bit ARM or two 16-bit Thumb) -@ r1 = thread_info pointer -@ r2 = PC value to resume execution after successful emulation -@ r3 = normal "successful" return address -@ lr = unrecognised instruction return address -@ IRQs enabled. -ENTRY(vfp_support_entry) - ldr r11, [r1, #TI_CPU] @ CPU number - add r10, r1, #TI_VFPSTATE @ r10 = workspace - - DBGSTR3 "instr %08x pc %08x state %p", r0, r2, r10 - - .fpu vfpv2 - VFPFMRX r1, FPEXC @ Is the VFP enabled? - DBGSTR1 "fpexc %08x", r1 - tst r1, #FPEXC_EN - bne look_for_VFP_exceptions @ VFP is already enabled - - DBGSTR1 "enable %x", r10 - ldr r9, vfp_current_hw_state_address - orr r1, r1, #FPEXC_EN @ user FPEXC has the enable bit set - ldr r4, [r9, r11, lsl #2] @ vfp_current_hw_state pointer - bic r5, r1, #FPEXC_EX @ make sure exceptions are disabled - cmp r4, r10 @ this thread owns the hw context? -#ifndef CONFIG_SMP - @ For UP, checking that this thread owns the hw context is - @ sufficient to determine that the hardware state is valid. - beq vfp_hw_state_valid - - @ On UP, we lazily save the VFP context. As a different - @ thread wants ownership of the VFP hardware, save the old - @ state if there was a previous (valid) owner. - - VFPFMXR FPEXC, r5 @ enable VFP, disable any pending - @ exceptions, so we can get at the - @ rest of it - - DBGSTR1 "save old state %p", r4 - cmp r4, #0 @ if the vfp_current_hw_state is NULL - beq vfp_reload_hw @ then the hw state needs reloading - VFPFSTMIA r4, r5 @ save the working registers - VFPFMRX r5, FPSCR @ current status - tst r1, #FPEXC_EX @ is there additional state to save? - beq 1f - VFPFMRX r6, FPINST @ FPINST (only if FPEXC.EX is set) - tst r1, #FPEXC_FP2V @ is there an FPINST2 to read? - beq 1f - VFPFMRX r8, FPINST2 @ FPINST2 if needed (and present) -1: - stmia r4, {r1, r5, r6, r8} @ save FPEXC, FPSCR, FPINST, FPINST2 -vfp_reload_hw: - -#else - @ For SMP, if this thread does not own the hw context, then we - @ need to reload it. No need to save the old state as on SMP, - @ we always save the state when we switch away from a thread. - bne vfp_reload_hw - - @ This thread has ownership of the current hardware context. - @ However, it may have been migrated to another CPU, in which - @ case the saved state is newer than the hardware context. - @ Check this by looking at the CPU number which the state was - @ last loaded onto. - ldr ip, [r10, #VFP_CPU] - teq ip, r11 - beq vfp_hw_state_valid - -vfp_reload_hw: - @ We're loading this threads state into the VFP hardware. Update - @ the CPU number which contains the most up to date VFP context. - str r11, [r10, #VFP_CPU] - - VFPFMXR FPEXC, r5 @ enable VFP, disable any pending - @ exceptions, so we can get at the - @ rest of it -#endif - - DBGSTR1 "load state %p", r10 - str r10, [r9, r11, lsl #2] @ update the vfp_current_hw_state pointer +ENTRY(vfp_load_state) + @ Load the current VFP state + @ r0 - load location + @ returns FPEXC + DBGSTR1 "load VFP state %p", r0 @ Load the saved state back into the VFP - VFPFLDMIA r10, r5 @ reload the working registers while + VFPFLDMIA r0, r1 @ reload the working registers while @ FPEXC is in a safe state - ldmia r10, {r1, r5, r6, r8} @ load FPEXC, FPSCR, FPINST, FPINST2 - tst r1, #FPEXC_EX @ is there additional state to restore? + ldmia r0, {r0-r3} @ load FPEXC, FPSCR, FPINST, FPINST2 + tst r0, #FPEXC_EX @ is there additional state to restore? beq 1f - VFPFMXR FPINST, r6 @ restore FPINST (only if FPEXC.EX is set) - tst r1, #FPEXC_FP2V @ is there an FPINST2 to write? + VFPFMXR FPINST, r2 @ restore FPINST (only if FPEXC.EX is set) + tst r0, #FPEXC_FP2V @ is there an FPINST2 to write? beq 1f - VFPFMXR FPINST2, r8 @ FPINST2 if needed (and present) + VFPFMXR FPINST2, r3 @ FPINST2 if needed (and present) 1: - VFPFMXR FPSCR, r5 @ restore status - -@ The context stored in the VFP hardware is up to date with this thread -vfp_hw_state_valid: - tst r1, #FPEXC_EX - bne process_exception @ might as well handle the pending - @ exception before retrying branch - @ out before setting an FPEXC that - @ stops us reading stuff - VFPFMXR FPEXC, r1 @ Restore FPEXC last - mov sp, r3 @ we think we have handled things - pop {lr} - sub r2, r2, #4 @ Retry current instruction - if Thumb - str r2, [sp, #S_PC] @ mode it's two 16-bit instructions, - @ else it's one 32-bit instruction, so - @ always subtract 4 from the following - @ instruction address. - -local_bh_enable_and_ret: - adr r0, . - mov r1, #SOFTIRQ_DISABLE_OFFSET - b __local_bh_enable_ip @ tail call - -look_for_VFP_exceptions: - @ Check for synchronous or asynchronous exception - tst r1, #FPEXC_EX | FPEXC_DEX - bne process_exception - @ On some implementations of the VFP subarch 1, setting FPSCR.IXE - @ causes all the CDP instructions to be bounced synchronously without - @ setting the FPEXC.EX bit - VFPFMRX r5, FPSCR - tst r5, #FPSCR_IXE - bne process_exception - - tst r5, #FPSCR_LENGTH_MASK - beq skip - orr r1, r1, #FPEXC_DEX - b process_exception -skip: - - @ Fall into hand on to next handler - appropriate coproc instr - @ not recognised by VFP - - DBGSTR "not VFP" - b local_bh_enable_and_ret - -process_exception: - DBGSTR "bounce" - mov sp, r3 @ setup for a return to the user code. - pop {lr} - mov r2, sp @ nothing stacked - regdump is at TOS - - @ Now call the C code to package up the bounce to the support code - @ r0 holds the trigger instruction - @ r1 holds the FPEXC value - @ r2 pointer to register dump - b VFP_bounce @ we have handled this - the support - @ code will raise an exception if - @ required. If not, the user code will - @ retry the faulted instruction -ENDPROC(vfp_support_entry) + VFPFMXR FPSCR, r1 @ restore status + ret lr +ENDPROC(vfp_load_state) ENTRY(vfp_save_state) @ Save the current VFP state @@ -238,10 +66,6 @@ ENTRY(vfp_save_state) ret lr ENDPROC(vfp_save_state) - .align -vfp_current_hw_state_address: - .word vfp_current_hw_state - .macro tbl_branch, base, tmp, shift #ifdef CONFIG_THUMB2_KERNEL adr \tmp, 1f diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 95628e57807b1..7572cb5b28a2e 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -30,11 +30,6 @@ #include "vfpinstr.h" #include "vfp.h" -/* - * Our undef handlers (in entry.S) - */ -asmlinkage void vfp_support_entry(u32, void *, u32, u32); - static bool have_vfp __ro_after_init; /* @@ -325,7 +320,7 @@ static u32 vfp_emulate_instruction(u32 inst, u32 fpscr, struct pt_regs *regs) /* * Package up a bounce condition. */ -void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) +static void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) { u32 fpscr, orig_fpscr, fpsid, exceptions; @@ -374,7 +369,7 @@ void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) * on VFP subarch 1. */ vfp_raise_exceptions(VFP_EXCEPTION_ERROR, trigger, fpscr, regs); - goto exit; + return; } /* @@ -405,7 +400,7 @@ void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) * the FPEXC.FP2V bit is valid only if FPEXC.EX is 1. */ if ((fpexc & (FPEXC_EX | FPEXC_FP2V)) != (FPEXC_EX | FPEXC_FP2V)) - goto exit; + return; /* * The barrier() here prevents fpinst2 being read @@ -418,8 +413,6 @@ void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) exceptions = vfp_emulate_instruction(trigger, orig_fpscr, regs); if (exceptions) vfp_raise_exceptions(exceptions, trigger, orig_fpscr, regs); - exit: - local_bh_enable(); } static void vfp_enable(void *unused) @@ -649,22 +642,112 @@ static int vfp_starting_cpu(unsigned int unused) } /* - * Entered with: + * vfp_support_entry - Handle VFP exception from user mode * - * r0 = instruction opcode (32-bit ARM or two 16-bit Thumb) - * r1 = thread_info pointer - * r2 = PC value to resume execution after successful emulation - * r3 = normal "successful" return address - * lr = unrecognised instruction return address + * @regs: pt_regs structure holding the register state at exception entry + * @trigger: The opcode of the instruction that triggered the exception + * + * Returns 0 if the exception was handled, or an error code otherwise. */ -asmlinkage void vfp_entry(u32 trigger, struct thread_info *ti, u32 resume_pc, - u32 resume_return_address) +asmlinkage int vfp_support_entry(struct pt_regs *regs, u32 trigger) { + struct thread_info *ti = current_thread_info(); + u32 fpexc; + if (unlikely(!have_vfp)) - return; + return -ENODEV; local_bh_disable(); - vfp_support_entry(trigger, ti, resume_pc, resume_return_address); + fpexc = fmrx(FPEXC); + + /* + * If the VFP unit was not enabled yet, we have to check whether the + * VFP state in the CPU's registers is the most recent VFP state + * associated with the process. On UP systems, we don't save the VFP + * state eagerly on a context switch, so we may need to save the + * VFP state to memory first, as it may belong to another process. + */ + if (!(fpexc & FPEXC_EN)) { + /* + * Enable the VFP unit but mask the FP exception flag for the + * time being, so we can access all the registers. + */ + fpexc |= FPEXC_EN; + fmxr(FPEXC, fpexc & ~FPEXC_EX); + + /* + * Check whether or not the VFP state in the CPU's registers is + * the most recent VFP state associated with this task. On SMP, + * migration may result in multiple CPUs holding VFP states + * that belong to the same task, but only the most recent one + * is valid. + */ + if (!vfp_state_in_hw(ti->cpu, ti)) { + if (!IS_ENABLED(CONFIG_SMP) && + vfp_current_hw_state[ti->cpu] != NULL) { + /* + * This CPU is currently holding the most + * recent VFP state associated with another + * task, and we must save that to memory first. + */ + vfp_save_state(vfp_current_hw_state[ti->cpu], + fpexc); + } + + /* + * We can now proceed with loading the task's VFP state + * from memory into the CPU registers. + */ + fpexc = vfp_load_state(&ti->vfpstate); + vfp_current_hw_state[ti->cpu] = &ti->vfpstate; +#ifdef CONFIG_SMP + /* + * Record that this CPU is now the one holding the most + * recent VFP state of the task. + */ + ti->vfpstate.hard.cpu = ti->cpu; +#endif + } + + if (fpexc & FPEXC_EX) + /* + * Might as well handle the pending exception before + * retrying branch out before setting an FPEXC that + * stops us reading stuff. + */ + goto bounce; + + /* + * No FP exception is pending: just enable the VFP and + * replay the instruction that trapped. + */ + fmxr(FPEXC, fpexc); + regs->ARM_pc -= 4; + } else { + /* Check for synchronous or asynchronous exceptions */ + if (!(fpexc & (FPEXC_EX | FPEXC_DEX))) { + u32 fpscr = fmrx(FPSCR); + + /* + * On some implementations of the VFP subarch 1, + * setting FPSCR.IXE causes all the CDP instructions to + * be bounced synchronously without setting the + * FPEXC.EX bit + */ + if (!(fpscr & FPSCR_IXE)) { + if (!(fpscr & FPSCR_LENGTH_MASK)) { + pr_debug("not VFP\n"); + local_bh_enable(); + return -ENOEXEC; + } + fpexc |= FPEXC_DEX; + } + } +bounce: VFP_bounce(trigger, fpexc, regs); + } + + local_bh_enable(); + return 0; } #ifdef CONFIG_KERNEL_MODE_NEON From 6ee1e6772e1e19436f573672de5ff8aab7163be6 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Sun, 19 Mar 2023 23:55:14 +0100 Subject: [PATCH 004/186] ARM: kernel: Get rid of thread_info::used_cp[] array We keep track of which coprocessor triggered a fault in the used_cp[] array in thread_info, but this data is never used anywhere. So let's remove it. Linus did some digging and found out that the last user of this field was removed in commit bb1a773d5b6b ("kill unused dump_fpu() instances"). Reviewed-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/include/asm/thread_info.h | 1 - arch/arm/kernel/asm-offsets.c | 1 - arch/arm/kernel/entry-armv.S | 6 ------ arch/arm/kernel/process.c | 1 - arch/arm/kernel/ptrace.c | 2 -- 5 files changed, 11 deletions(-) diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index 7f092cb55a417..85c5f1e02ebf8 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -66,7 +66,6 @@ struct thread_info { __u32 cpu_domain; /* cpu domain */ struct cpu_context_save cpu_context; /* cpu context */ __u32 abi_syscall; /* ABI type and syscall nr */ - __u8 used_cp[16]; /* thread used copro */ unsigned long tp_value[2]; /* TLS registers */ union fp_state fpstate __attribute__((aligned(8))); union vfp_state vfpstate; diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c index 38121c59cbc26..f9c7111c1d65f 100644 --- a/arch/arm/kernel/asm-offsets.c +++ b/arch/arm/kernel/asm-offsets.c @@ -47,7 +47,6 @@ int main(void) DEFINE(TI_CPU_DOMAIN, offsetof(struct thread_info, cpu_domain)); DEFINE(TI_CPU_SAVE, offsetof(struct thread_info, cpu_context)); DEFINE(TI_ABI_SYSCALL, offsetof(struct thread_info, abi_syscall)); - DEFINE(TI_USED_CP, offsetof(struct thread_info, used_cp)); DEFINE(TI_TP_VALUE, offsetof(struct thread_info, tp_value)); DEFINE(TI_FPSTATE, offsetof(struct thread_info, fpstate)); #ifdef CONFIG_VFP diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index c39303e5c2347..ba47f6aac5ff8 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -591,9 +591,6 @@ call_fpe: and r8, r0, r5 cmp r8, r7 @ NEON instruction? bne 2b - mov r7, #1 - strb r7, [r10, #TI_USED_CP + 10] @ mark CP#10 as used - strb r7, [r10, #TI_USED_CP + 11] @ mark CP#11 as used b do_vfp @ let VFP handler handle this 1: #endif @@ -601,9 +598,6 @@ call_fpe: tstne r0, #0x04000000 @ bit 26 set on both ARM and Thumb-2 reteq lr and r8, r0, #0x00000f00 @ mask out CP number - mov r7, #1 - add r6, r10, r8, lsr #8 @ add used_cp[] array offset first - strb r7, [r6, #TI_USED_CP] @ set appropriate used_cp[] #ifdef CONFIG_IWMMXT @ Test if we need to give access to iWMMXt coprocessors ldr r5, [r10, #TI_FLAGS] diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index 0e8ff85890ade..e16ed102960cb 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -222,7 +222,6 @@ void flush_thread(void) flush_ptrace_hw_breakpoint(tsk); - memset(thread->used_cp, 0, sizeof(thread->used_cp)); memset(&tsk->thread.debug, 0, sizeof(struct debug_info)); memset(&thread->fpstate, 0, sizeof(union fp_state)); diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 2d8e2516906b6..2b945b9bd3662 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -584,8 +584,6 @@ static int fpa_set(struct task_struct *target, { struct thread_info *thread = task_thread_info(target); - thread->used_cp[1] = thread->used_cp[2] = 1; - return user_regset_copyin(&pos, &count, &kbuf, &ubuf, &thread->fpstate, 0, sizeof(struct user_fp)); From cdd87465adfd75e4ebd11507575533c6bf7a5525 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Sun, 19 Mar 2023 00:28:35 +0100 Subject: [PATCH 005/186] ARM: vfp: Use undef hook for handling VFP exceptions Now that the VFP support code has been reimplemented as a C function that takes a struct pt_regs pointer and an opcode, we can use the existing undef_hook framework to deal with undef exceptions triggered by VFP instructions instead of having special handling in assembler. Reviewed-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/kernel/entry-armv.S | 53 ----------------- arch/arm/vfp/Makefile | 2 +- arch/arm/vfp/entry.S | 31 ---------- arch/arm/vfp/vfpmodule.c | 109 +++++++++++++++++------------------ 4 files changed, 54 insertions(+), 141 deletions(-) delete mode 100644 arch/arm/vfp/entry.S diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index ba47f6aac5ff8..0e40b2566f598 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -557,13 +557,6 @@ ENDPROC(__und_usr) * co-processor instructions. However, we have to watch out * for the ARM6/ARM7 SWI bug. * - * NEON is a special case that has to be handled here. Not all - * NEON instructions are co-processor instructions, so we have - * to make a special case of checking for them. Plus, there's - * five groups of them, so we have a table of mask/opcode pairs - * to check against, and if any match then we branch off into the - * NEON handler code. - * * Emulators may wish to make use of the following registers: * r0 = instruction opcode (32-bit ARM or two 16-bit Thumb) * r2 = PC value to resume execution after successful emulation @@ -575,25 +568,8 @@ ENDPROC(__und_usr) @ @ Fall-through from Thumb-2 __und_usr @ -#ifdef CONFIG_NEON - get_thread_info r10 @ get current thread - adr r6, .LCneon_thumb_opcodes - b 2f -#endif call_fpe: get_thread_info r10 @ get current thread -#ifdef CONFIG_NEON - adr r6, .LCneon_arm_opcodes -2: ldr r5, [r6], #4 @ mask value - ldr r7, [r6], #4 @ opcode bits matching in mask - cmp r5, #0 @ end mask? - beq 1f - and r8, r0, r5 - cmp r8, r7 @ NEON instruction? - bne 2b - b do_vfp @ let VFP handler handle this -1: -#endif tst r0, #0x08000000 @ only CDP/CPRT/LDC/STC have bit 27 tstne r0, #0x04000000 @ bit 26 set on both ARM and Thumb-2 reteq lr @@ -620,42 +596,13 @@ call_fpe: ret.w lr @ CP#7 ret.w lr @ CP#8 ret.w lr @ CP#9 -#ifdef CONFIG_VFP - W(b) do_vfp @ CP#10 (VFP) - W(b) do_vfp @ CP#11 (VFP) -#else ret.w lr @ CP#10 (VFP) ret.w lr @ CP#11 (VFP) -#endif ret.w lr @ CP#12 ret.w lr @ CP#13 ret.w lr @ CP#14 (Debug) ret.w lr @ CP#15 (Control) -#ifdef CONFIG_NEON - .align 6 - -.LCneon_arm_opcodes: - .word 0xfe000000 @ mask - .word 0xf2000000 @ opcode - - .word 0xff100000 @ mask - .word 0xf4000000 @ opcode - - .word 0x00000000 @ mask - .word 0x00000000 @ opcode - -.LCneon_thumb_opcodes: - .word 0xef000000 @ mask - .word 0xef000000 @ opcode - - .word 0xff100000 @ mask - .word 0xf9000000 @ opcode - - .word 0x00000000 @ mask - .word 0x00000000 @ opcode -#endif - do_fpe: add r10, r10, #TI_FPSTATE @ r10 = workspace ldr_va pc, fp_enter, tmp=r4 @ Call FP module USR entry point diff --git a/arch/arm/vfp/Makefile b/arch/arm/vfp/Makefile index 749901a72d6dc..dfd64bc2b2fbd 100644 --- a/arch/arm/vfp/Makefile +++ b/arch/arm/vfp/Makefile @@ -8,4 +8,4 @@ # ccflags-y := -DDEBUG # asflags-y := -DDEBUG -obj-y += vfpmodule.o entry.o vfphw.o vfpsingle.o vfpdouble.o +obj-y += vfpmodule.o vfphw.o vfpsingle.o vfpdouble.o diff --git a/arch/arm/vfp/entry.S b/arch/arm/vfp/entry.S deleted file mode 100644 index 547c94c62cd3a..0000000000000 --- a/arch/arm/vfp/entry.S +++ /dev/null @@ -1,31 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* - * linux/arch/arm/vfp/entry.S - * - * Copyright (C) 2004 ARM Limited. - * Written by Deep Blue Solutions Limited. - */ -#include -#include -#include -#include -#include -#include - -@ VFP entry point. -@ -@ r0 = instruction opcode (32-bit ARM or two 16-bit Thumb) -@ r2 = PC value to resume execution after successful emulation -@ r9 = normal "successful" return address -@ r10 = this threads thread_info structure -@ lr = unrecognised instruction return address -@ IRQs enabled. -@ -ENTRY(do_vfp) - mov r1, r0 @ pass trigger opcode via R1 - mov r0, sp @ pass struct pt_regs via R0 - bl vfp_support_entry @ dispatch the VFP exception - cmp r0, #0 @ handled successfully? - reteq r9 @ then use R9 as return address - ret lr @ pass to undef handler -ENDPROC(do_vfp) diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 7572cb5b28a2e..58a9442add24b 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -641,15 +641,37 @@ static int vfp_starting_cpu(unsigned int unused) return 0; } +static int vfp_kmode_exception(struct pt_regs *regs, unsigned int instr) +{ + /* + * If we reach this point, a floating point exception has been raised + * while running in kernel mode. If the NEON/VFP unit was enabled at the + * time, it means a VFP instruction has been issued that requires + * software assistance to complete, something which is not currently + * supported in kernel mode. + * If the NEON/VFP unit was disabled, and the location pointed to below + * is properly preceded by a call to kernel_neon_begin(), something has + * caused the task to be scheduled out and back in again. In this case, + * rebuilding and running with CONFIG_DEBUG_ATOMIC_SLEEP enabled should + * be helpful in localizing the problem. + */ + if (fmrx(FPEXC) & FPEXC_EN) + pr_crit("BUG: unsupported FP instruction in kernel mode\n"); + else + pr_crit("BUG: FP instruction issued in kernel mode with FP unit disabled\n"); + pr_crit("FPEXC == 0x%08x\n", fmrx(FPEXC)); + return 1; +} + /* - * vfp_support_entry - Handle VFP exception from user mode + * vfp_support_entry - Handle VFP exception * * @regs: pt_regs structure holding the register state at exception entry * @trigger: The opcode of the instruction that triggered the exception * * Returns 0 if the exception was handled, or an error code otherwise. */ -asmlinkage int vfp_support_entry(struct pt_regs *regs, u32 trigger) +static int vfp_support_entry(struct pt_regs *regs, u32 trigger) { struct thread_info *ti = current_thread_info(); u32 fpexc; @@ -657,6 +679,9 @@ asmlinkage int vfp_support_entry(struct pt_regs *regs, u32 trigger) if (unlikely(!have_vfp)) return -ENODEV; + if (!user_mode(regs)) + return vfp_kmode_exception(regs, trigger); + local_bh_disable(); fpexc = fmrx(FPEXC); @@ -722,7 +747,6 @@ asmlinkage int vfp_support_entry(struct pt_regs *regs, u32 trigger) * replay the instruction that trapped. */ fmxr(FPEXC, fpexc); - regs->ARM_pc -= 4; } else { /* Check for synchronous or asynchronous exceptions */ if (!(fpexc & (FPEXC_EX | FPEXC_DEX))) { @@ -743,78 +767,47 @@ asmlinkage int vfp_support_entry(struct pt_regs *regs, u32 trigger) fpexc |= FPEXC_DEX; } } -bounce: VFP_bounce(trigger, fpexc, regs); +bounce: regs->ARM_pc += 4; + VFP_bounce(trigger, fpexc, regs); } local_bh_enable(); return 0; } -#ifdef CONFIG_KERNEL_MODE_NEON - -static int vfp_kmode_exception(struct pt_regs *regs, unsigned int instr) -{ - /* - * If we reach this point, a floating point exception has been raised - * while running in kernel mode. If the NEON/VFP unit was enabled at the - * time, it means a VFP instruction has been issued that requires - * software assistance to complete, something which is not currently - * supported in kernel mode. - * If the NEON/VFP unit was disabled, and the location pointed to below - * is properly preceded by a call to kernel_neon_begin(), something has - * caused the task to be scheduled out and back in again. In this case, - * rebuilding and running with CONFIG_DEBUG_ATOMIC_SLEEP enabled should - * be helpful in localizing the problem. - */ - if (fmrx(FPEXC) & FPEXC_EN) - pr_crit("BUG: unsupported FP instruction in kernel mode\n"); - else - pr_crit("BUG: FP instruction issued in kernel mode with FP unit disabled\n"); - pr_crit("FPEXC == 0x%08x\n", fmrx(FPEXC)); - return 1; -} - -static struct undef_hook vfp_kmode_exception_hook[] = {{ +static struct undef_hook neon_support_hook[] = {{ .instr_mask = 0xfe000000, .instr_val = 0xf2000000, - .cpsr_mask = MODE_MASK | PSR_T_BIT, - .cpsr_val = SVC_MODE, - .fn = vfp_kmode_exception, + .cpsr_mask = PSR_T_BIT, + .cpsr_val = 0, + .fn = vfp_support_entry, }, { .instr_mask = 0xff100000, .instr_val = 0xf4000000, - .cpsr_mask = MODE_MASK | PSR_T_BIT, - .cpsr_val = SVC_MODE, - .fn = vfp_kmode_exception, + .cpsr_mask = PSR_T_BIT, + .cpsr_val = 0, + .fn = vfp_support_entry, }, { .instr_mask = 0xef000000, .instr_val = 0xef000000, - .cpsr_mask = MODE_MASK | PSR_T_BIT, - .cpsr_val = SVC_MODE | PSR_T_BIT, - .fn = vfp_kmode_exception, + .cpsr_mask = PSR_T_BIT, + .cpsr_val = PSR_T_BIT, + .fn = vfp_support_entry, }, { .instr_mask = 0xff100000, .instr_val = 0xf9000000, - .cpsr_mask = MODE_MASK | PSR_T_BIT, - .cpsr_val = SVC_MODE | PSR_T_BIT, - .fn = vfp_kmode_exception, -}, { - .instr_mask = 0x0c000e00, - .instr_val = 0x0c000a00, - .cpsr_mask = MODE_MASK, - .cpsr_val = SVC_MODE, - .fn = vfp_kmode_exception, + .cpsr_mask = PSR_T_BIT, + .cpsr_val = PSR_T_BIT, + .fn = vfp_support_entry, }}; -static int __init vfp_kmode_exception_hook_init(void) -{ - int i; +static struct undef_hook vfp_support_hook = { + .instr_mask = 0x0c000e00, + .instr_val = 0x0c000a00, + .fn = vfp_support_entry, +}; - for (i = 0; i < ARRAY_SIZE(vfp_kmode_exception_hook); i++) - register_undef_hook(&vfp_kmode_exception_hook[i]); - return 0; -} -subsys_initcall(vfp_kmode_exception_hook_init); +#ifdef CONFIG_KERNEL_MODE_NEON /* * Kernel-side NEON support functions @@ -919,8 +912,11 @@ static int __init vfp_init(void) * for NEON if the hardware has the MVFR registers. */ if (IS_ENABLED(CONFIG_NEON) && - (fmrx(MVFR1) & 0x000fff00) == 0x00011100) + (fmrx(MVFR1) & 0x000fff00) == 0x00011100) { elf_hwcap |= HWCAP_NEON; + for (int i = 0; i < ARRAY_SIZE(neon_support_hook); i++) + register_undef_hook(&neon_support_hook[i]); + } if (IS_ENABLED(CONFIG_VFPv3)) { u32 mvfr0 = fmrx(MVFR0); @@ -989,6 +985,7 @@ static int __init vfp_init(void) have_vfp = true; + register_undef_hook(&vfp_support_hook); thread_register_notifier(&vfp_notifier_block); vfp_pm_init(); From 8bcba70cb5c2204a011e06278a1fbfb1213e1df1 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Sun, 19 Mar 2023 15:18:25 +0100 Subject: [PATCH 006/186] ARM: entry: Disregard Thumb undef exception in coproc dispatch Now that the only remaining coprocessor instructions being handled via the dispatch in entry-armv.S are ones that only exist in a ARM (A32) encoding, we can simplify the handling of Thumb undef exceptions, and send them straight to the undefined instruction handlers in C code. This also means we can drop the code that partially decodes the instruction to decide whether it is a 16-bit or 32-bit Thumb instruction: this is all taken care of by the undef hook. Acked-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/kernel/entry-armv.S | 121 ++++++----------------------------- 1 file changed, 18 insertions(+), 103 deletions(-) diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 0e40b2566f598..aff6cfe587456 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -446,106 +446,32 @@ ENDPROC(__irq_usr) __und_usr: usr_entry uaccess=0 - mov r2, r4 - mov r3, r5 - - @ r2 = regs->ARM_pc, which is either 2 or 4 bytes ahead of the - @ faulting instruction depending on Thumb mode. - @ r3 = regs->ARM_cpsr - @ - @ The emulation code returns using r9 if it has emulated the - @ instruction, or the more conventional lr if we are to treat - @ this as a real undefined instruction - @ - badr r9, ret_from_exception - @ IRQs must be enabled before attempting to read the instruction from @ user space since that could cause a page/translation fault if the @ page table was modified by another CPU. enable_irq - tst r3, #PSR_T_BIT @ Thumb mode? - bne __und_usr_thumb - sub r4, r2, #4 @ ARM instr at LR - 4 -1: ldrt r0, [r4] - ARM_BE8(rev r0, r0) @ little endian instruction - + tst r5, #PSR_T_BIT @ Thumb mode? + mov r1, #2 @ set insn size to 2 for Thumb + bne 0f @ handle as Thumb undef exception + adr r9, ret_from_exception + bl call_fpe @ returns via R9 on success + mov r1, #4 @ set insn size to 4 for ARM +0: mov r0, sp uaccess_disable ip - - @ r0 = 32-bit ARM instruction which caused the exception - @ r2 = PC value for the following instruction (:= regs->ARM_pc) - @ r4 = PC value for the faulting instruction - @ lr = 32-bit undefined instruction function - badr lr, __und_usr_fault_32 - b call_fpe - -__und_usr_thumb: - @ Thumb instruction - sub r4, r2, #2 @ First half of thumb instr at LR - 2 -#if CONFIG_ARM_THUMB && __LINUX_ARM_ARCH__ >= 6 && CONFIG_CPU_V7 -/* - * Thumb-2 instruction handling. Note that because pre-v6 and >= v6 platforms - * can never be supported in a single kernel, this code is not applicable at - * all when __LINUX_ARM_ARCH__ < 6. This allows simplifying assumptions to be - * made about .arch directives. - */ -#if __LINUX_ARM_ARCH__ < 7 -/* If the target CPU may not be Thumb-2-capable, a run-time check is needed: */ - ldr_va r5, cpu_architecture - cmp r5, #CPU_ARCH_ARMv7 - blo __und_usr_fault_16 @ 16bit undefined instruction -/* - * The following code won't get run unless the running CPU really is v7, so - * coding round the lack of ldrht on older arches is pointless. Temporarily - * override the assembler target arch with the minimum required instead: - */ - .arch armv6t2 -#endif -2: ldrht r5, [r4] -ARM_BE8(rev16 r5, r5) @ little endian instruction - cmp r5, #0xe800 @ 32bit instruction if xx != 0 - blo __und_usr_fault_16_pan @ 16bit undefined instruction -3: ldrht r0, [r2] -ARM_BE8(rev16 r0, r0) @ little endian instruction - uaccess_disable ip - add r2, r2, #2 @ r2 is PC + 2, make it PC + 4 - str r2, [sp, #S_PC] @ it's a 2x16bit instr, update - orr r0, r0, r5, lsl #16 - badr lr, __und_usr_fault_32 - @ r0 = the two 16-bit Thumb instructions which caused the exception - @ r2 = PC value for the following Thumb instruction (:= regs->ARM_pc) - @ r4 = PC value for the first 16-bit Thumb instruction - @ lr = 32bit undefined instruction function - -#if __LINUX_ARM_ARCH__ < 7 -/* If the target arch was overridden, change it back: */ -#ifdef CONFIG_CPU_32v6K - .arch armv6k -#else - .arch armv6 -#endif -#endif /* __LINUX_ARM_ARCH__ < 7 */ -#else /* !(CONFIG_ARM_THUMB && __LINUX_ARM_ARCH__ >= 6 && CONFIG_CPU_V7) */ - b __und_usr_fault_16 -#endif + bl __und_fault + b ret_from_exception UNWIND(.fnend) ENDPROC(__und_usr) /* - * The out of line fixup for the ldrt instructions above. + * The out of line fixup for the ldrt instruction below. */ .pushsection .text.fixup, "ax" .align 2 4: str r4, [sp, #S_PC] @ retry current instruction ret r9 .popsection - .pushsection __ex_table,"a" - .long 1b, 4b -#if CONFIG_ARM_THUMB && __LINUX_ARM_ARCH__ >= 6 && CONFIG_CPU_V7 - .long 2b, 4b - .long 3b, 4b -#endif - .popsection /* * Check whether the instruction is a co-processor instruction. @@ -558,20 +484,22 @@ ENDPROC(__und_usr) * for the ARM6/ARM7 SWI bug. * * Emulators may wish to make use of the following registers: - * r0 = instruction opcode (32-bit ARM or two 16-bit Thumb) - * r2 = PC value to resume execution after successful emulation + * r4 = PC value to resume execution after successful emulation * r9 = normal "successful" return address * r10 = this threads thread_info structure * lr = unrecognised instruction return address * IRQs enabled, FIQs enabled. */ - @ - @ Fall-through from Thumb-2 __und_usr - @ call_fpe: + mov r2, r4 + sub r4, r4, #4 @ ARM instruction at user PC - 4 +USERL( 4b, ldrt r0, [r4]) @ load opcode from user space +ARM_BE8(rev r0, r0) @ little endian instruction + + uaccess_disable ip + get_thread_info r10 @ get current thread tst r0, #0x08000000 @ only CDP/CPRT/LDC/STC have bit 27 - tstne r0, #0x04000000 @ bit 26 set on both ARM and Thumb-2 reteq lr and r8, r0, #0x00000f00 @ mask out CP number #ifdef CONFIG_IWMMXT @@ -626,19 +554,6 @@ ENTRY(no_fp) ret lr ENDPROC(no_fp) -__und_usr_fault_32: - mov r1, #4 - b 1f -__und_usr_fault_16_pan: - uaccess_disable ip -__und_usr_fault_16: - mov r1, #2 -1: mov r0, sp - badr lr, ret_from_exception - b __und_fault -ENDPROC(__und_usr_fault_32) -ENDPROC(__und_usr_fault_16) - .align 5 __pabt_usr: usr_entry From 303d6da167dcbc3dd89adf3ca4e36c369950ed01 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Mon, 20 Mar 2023 00:07:20 +0100 Subject: [PATCH 007/186] ARM: iwmmxt: Use undef hook to enable coprocessor for task Define a undef hook to deal with undef exceptions triggered by iwmmxt instructions that were issued with the coprocessor disabled. This removes the dependency on the coprocessor dispatch code in entry-armv.S, which will be made NWFPE-only in a subsequent patch. Reviewed-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/include/asm/thread_info.h | 16 ++++++++++++++++ arch/arm/kernel/entry-armv.S | 1 + arch/arm/kernel/iwmmxt.S | 18 ++++++++++++++---- arch/arm/kernel/pj4-cp0.c | 1 + arch/arm/kernel/xscale-cp0.c | 1 + 5 files changed, 33 insertions(+), 4 deletions(-) diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index 85c5f1e02ebf8..943ffcf069d29 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -40,6 +40,7 @@ struct task_struct; DECLARE_PER_CPU(struct task_struct *, __entry_task); #include +#include struct cpu_context_save { __u32 r4; @@ -104,6 +105,21 @@ extern void iwmmxt_task_restore(struct thread_info *, void *); extern void iwmmxt_task_release(struct thread_info *); extern void iwmmxt_task_switch(struct thread_info *); +extern int iwmmxt_undef_handler(struct pt_regs *, u32); + +static inline void register_iwmmxt_undef_handler(void) +{ + static struct undef_hook iwmmxt_undef_hook = { + .instr_mask = 0x0c000e00, + .instr_val = 0x0c000000, + .cpsr_mask = MODE_MASK | PSR_T_BIT, + .cpsr_val = USR_MODE, + .fn = iwmmxt_undef_handler, + }; + + register_undef_hook(&iwmmxt_undef_hook); +} + extern void vfp_sync_hwstate(struct thread_info *); extern void vfp_flush_hwstate(struct thread_info *); diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index aff6cfe587456..822b2c83bf083 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -507,6 +507,7 @@ ARM_BE8(rev r0, r0) @ little endian instruction ldr r5, [r10, #TI_FLAGS] rsbs r7, r8, #(1 << 8) @ CP 0 or 1 only movscs r7, r5, lsr #(TIF_USING_IWMMXT + 1) + movcs r0, sp @ pass struct pt_regs bcs iwmmxt_task_enable #endif ARM( add pc, pc, r8, lsr #6 ) diff --git a/arch/arm/kernel/iwmmxt.S b/arch/arm/kernel/iwmmxt.S index d2b4ac06e4ed8..a0218c4867b9b 100644 --- a/arch/arm/kernel/iwmmxt.S +++ b/arch/arm/kernel/iwmmxt.S @@ -58,9 +58,19 @@ .text .arm +ENTRY(iwmmxt_undef_handler) + push {r9, r10, lr} + get_thread_info r10 + mov r9, pc + b iwmmxt_task_enable + mov r0, #0 + pop {r9, r10, pc} +ENDPROC(iwmmxt_undef_handler) + /* * Lazy switching of Concan coprocessor context * + * r0 = struct pt_regs pointer * r10 = struct thread_info pointer * r9 = ret_from_exception * lr = undefined instr exit @@ -84,12 +94,12 @@ ENTRY(iwmmxt_task_enable) PJ4(mcr p15, 0, r2, c1, c0, 2) ldr r3, =concan_owner - add r0, r10, #TI_IWMMXT_STATE @ get task Concan save area - ldr r2, [sp, #60] @ current task pc value + ldr r2, [r0, #S_PC] @ current task pc value ldr r1, [r3] @ get current Concan owner - str r0, [r3] @ this task now owns Concan regs sub r2, r2, #4 @ adjust pc back - str r2, [sp, #60] + str r2, [r0, #S_PC] + add r0, r10, #TI_IWMMXT_STATE @ get task Concan save area + str r0, [r3] @ this task now owns Concan regs mrc p15, 0, r2, c2, c0, 0 mov r2, r2 @ cpwait diff --git a/arch/arm/kernel/pj4-cp0.c b/arch/arm/kernel/pj4-cp0.c index 1d1fb22f44f37..4bca8098c4ff5 100644 --- a/arch/arm/kernel/pj4-cp0.c +++ b/arch/arm/kernel/pj4-cp0.c @@ -126,6 +126,7 @@ static int __init pj4_cp0_init(void) pr_info("PJ4 iWMMXt v%d coprocessor enabled.\n", vers); elf_hwcap |= HWCAP_IWMMXT; thread_register_notifier(&iwmmxt_notifier_block); + register_iwmmxt_undef_handler(); #endif return 0; diff --git a/arch/arm/kernel/xscale-cp0.c b/arch/arm/kernel/xscale-cp0.c index ed4f6e77616da..00d00d3aae972 100644 --- a/arch/arm/kernel/xscale-cp0.c +++ b/arch/arm/kernel/xscale-cp0.c @@ -166,6 +166,7 @@ static int __init xscale_cp0_init(void) pr_info("XScale iWMMXt coprocessor detected.\n"); elf_hwcap |= HWCAP_IWMMXT; thread_register_notifier(&iwmmxt_notifier_block); + register_iwmmxt_undef_handler(); #endif } else { pr_info("XScale DSP coprocessor detected.\n"); From 47ba5f39eab3c2a9a1ba878159a6050f2bbfc0e2 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Mon, 20 Mar 2023 00:25:18 +0100 Subject: [PATCH 008/186] ARM: entry: Make asm coproc dispatch code NWFPE only Now that we can dispatch all VFP and iWMMXT related undef exceptions using undef hooks implemented in C code, we no longer need the asm entry code that takes care of this unless we are using FPE, so we can move it into the FPE entry code. As this means it is ARM only, we can remove the Thumb2 specific decorations as well. It also means the non-standard, asm-only calling convention where returning via LR means failure and returning via R9 means success is now only used on legacy platforms that lack any kind of function return prediction, avoiding the associated performance impact. Reviewed-by: Linus Walleij Signed-off-by: Ard Biesheuvel --- arch/arm/kernel/entry-armv.S | 93 +----------------------------------- arch/arm/nwfpe/entry.S | 77 +++++++++++++++++++++++++++++ 2 files changed, 79 insertions(+), 91 deletions(-) diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 822b2c83bf083..682e92664b07f 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -454,8 +454,10 @@ __und_usr: tst r5, #PSR_T_BIT @ Thumb mode? mov r1, #2 @ set insn size to 2 for Thumb bne 0f @ handle as Thumb undef exception +#ifdef CONFIG_FPE_NWFPE adr r9, ret_from_exception bl call_fpe @ returns via R9 on success +#endif mov r1, #4 @ set insn size to 4 for ARM 0: mov r0, sp uaccess_disable ip @@ -464,97 +466,6 @@ __und_usr: UNWIND(.fnend) ENDPROC(__und_usr) -/* - * The out of line fixup for the ldrt instruction below. - */ - .pushsection .text.fixup, "ax" - .align 2 -4: str r4, [sp, #S_PC] @ retry current instruction - ret r9 - .popsection - -/* - * Check whether the instruction is a co-processor instruction. - * If yes, we need to call the relevant co-processor handler. - * - * Note that we don't do a full check here for the co-processor - * instructions; all instructions with bit 27 set are well - * defined. The only instructions that should fault are the - * co-processor instructions. However, we have to watch out - * for the ARM6/ARM7 SWI bug. - * - * Emulators may wish to make use of the following registers: - * r4 = PC value to resume execution after successful emulation - * r9 = normal "successful" return address - * r10 = this threads thread_info structure - * lr = unrecognised instruction return address - * IRQs enabled, FIQs enabled. - */ -call_fpe: - mov r2, r4 - sub r4, r4, #4 @ ARM instruction at user PC - 4 -USERL( 4b, ldrt r0, [r4]) @ load opcode from user space -ARM_BE8(rev r0, r0) @ little endian instruction - - uaccess_disable ip - - get_thread_info r10 @ get current thread - tst r0, #0x08000000 @ only CDP/CPRT/LDC/STC have bit 27 - reteq lr - and r8, r0, #0x00000f00 @ mask out CP number -#ifdef CONFIG_IWMMXT - @ Test if we need to give access to iWMMXt coprocessors - ldr r5, [r10, #TI_FLAGS] - rsbs r7, r8, #(1 << 8) @ CP 0 or 1 only - movscs r7, r5, lsr #(TIF_USING_IWMMXT + 1) - movcs r0, sp @ pass struct pt_regs - bcs iwmmxt_task_enable -#endif - ARM( add pc, pc, r8, lsr #6 ) - THUMB( lsr r8, r8, #6 ) - THUMB( add pc, r8 ) - nop - - ret.w lr @ CP#0 - W(b) do_fpe @ CP#1 (FPE) - W(b) do_fpe @ CP#2 (FPE) - ret.w lr @ CP#3 - ret.w lr @ CP#4 - ret.w lr @ CP#5 - ret.w lr @ CP#6 - ret.w lr @ CP#7 - ret.w lr @ CP#8 - ret.w lr @ CP#9 - ret.w lr @ CP#10 (VFP) - ret.w lr @ CP#11 (VFP) - ret.w lr @ CP#12 - ret.w lr @ CP#13 - ret.w lr @ CP#14 (Debug) - ret.w lr @ CP#15 (Control) - -do_fpe: - add r10, r10, #TI_FPSTATE @ r10 = workspace - ldr_va pc, fp_enter, tmp=r4 @ Call FP module USR entry point - -/* - * The FP module is called with these registers set: - * r0 = instruction - * r2 = PC+4 - * r9 = normal "successful" return address - * r10 = FP workspace - * lr = unrecognised FP instruction return address - */ - - .pushsection .data - .align 2 -ENTRY(fp_enter) - .word no_fp - .popsection - -ENTRY(no_fp) - ret lr -ENDPROC(no_fp) - .align 5 __pabt_usr: usr_entry diff --git a/arch/arm/nwfpe/entry.S b/arch/arm/nwfpe/entry.S index d8f9915566e15..354d297a193bb 100644 --- a/arch/arm/nwfpe/entry.S +++ b/arch/arm/nwfpe/entry.S @@ -7,6 +7,7 @@ Direct questions, comments to Scott Bambrough */ +#include #include #include @@ -104,6 +105,7 @@ next: @ plain LDR instruction. Weird, but it seems harmless. .pushsection .text.fixup,"ax" .align 2 +.Lrep: str r4, [sp, #S_PC] @ retry current instruction .Lfix: ret r9 @ let the user eat segfaults .popsection @@ -111,3 +113,78 @@ next: .align 3 .long .Lx1, .Lfix .popsection + + @ + @ Check whether the instruction is a co-processor instruction. + @ If yes, we need to call the relevant co-processor handler. + @ Only FPE instructions are dispatched here, everything else + @ is handled by undef hooks. + @ + @ Emulators may wish to make use of the following registers: + @ r4 = PC value to resume execution after successful emulation + @ r9 = normal "successful" return address + @ lr = unrecognised instruction return address + @ IRQs enabled, FIQs enabled. + @ +ENTRY(call_fpe) + mov r2, r4 + sub r4, r4, #4 @ ARM instruction at user PC - 4 +USERL( .Lrep, ldrt r0, [r4]) @ load opcode from user space +ARM_BE8(rev r0, r0) @ little endian instruction + + uaccess_disable ip + + get_thread_info r10 @ get current thread + tst r0, #0x08000000 @ only CDP/CPRT/LDC/STC have bit 27 + reteq lr + and r8, r0, #0x00000f00 @ mask out CP number +#ifdef CONFIG_IWMMXT + @ Test if we need to give access to iWMMXt coprocessors + ldr r5, [r10, #TI_FLAGS] + rsbs r7, r8, #(1 << 8) @ CP 0 or 1 only + movscs r7, r5, lsr #(TIF_USING_IWMMXT + 1) + movcs r0, sp @ pass struct pt_regs + bcs iwmmxt_task_enable +#endif + add pc, pc, r8, lsr #6 + nop + + ret lr @ CP#0 + b do_fpe @ CP#1 (FPE) + b do_fpe @ CP#2 (FPE) + ret lr @ CP#3 + ret lr @ CP#4 + ret lr @ CP#5 + ret lr @ CP#6 + ret lr @ CP#7 + ret lr @ CP#8 + ret lr @ CP#9 + ret lr @ CP#10 (VFP) + ret lr @ CP#11 (VFP) + ret lr @ CP#12 + ret lr @ CP#13 + ret lr @ CP#14 (Debug) + ret lr @ CP#15 (Control) + +do_fpe: + add r10, r10, #TI_FPSTATE @ r10 = workspace + ldr_va pc, fp_enter, tmp=r4 @ Call FP module USR entry point + + @ + @ The FP module is called with these registers set: + @ r0 = instruction + @ r2 = PC+4 + @ r9 = normal "successful" return address + @ r10 = FP workspace + @ lr = unrecognised FP instruction return address + @ + + .pushsection .data + .align 2 +ENTRY(fp_enter) + .word no_fp + .popsection + +no_fp: + ret lr +ENDPROC(no_fp) From cade5397e5461295f3cb87880534b6a07cafa427 Mon Sep 17 00:00:00 2001 From: Andrew Kanner Date: Sat, 1 Jul 2023 17:05:42 +0300 Subject: [PATCH 009/186] fs/jfs: prevent double-free in dbUnmount() after failed jfs_remount() Syzkaller reported the following issue: ================================================================== BUG: KASAN: double-free in slab_free mm/slub.c:3787 [inline] BUG: KASAN: double-free in __kmem_cache_free+0x71/0x110 mm/slub.c:3800 Free of addr ffff888086408000 by task syz-executor.4/12750 [...] Call Trace: [...] kasan_report_invalid_free+0xac/0xd0 mm/kasan/report.c:482 ____kasan_slab_free+0xfb/0x120 kasan_slab_free include/linux/kasan.h:177 [inline] slab_free_hook mm/slub.c:1781 [inline] slab_free_freelist_hook+0x12e/0x1a0 mm/slub.c:1807 slab_free mm/slub.c:3787 [inline] __kmem_cache_free+0x71/0x110 mm/slub.c:3800 dbUnmount+0xf4/0x110 fs/jfs/jfs_dmap.c:264 jfs_umount+0x248/0x3b0 fs/jfs/jfs_umount.c:87 jfs_put_super+0x86/0x190 fs/jfs/super.c:194 generic_shutdown_super+0x130/0x310 fs/super.c:492 kill_block_super+0x79/0xd0 fs/super.c:1386 deactivate_locked_super+0xa7/0xf0 fs/super.c:332 cleanup_mnt+0x494/0x520 fs/namespace.c:1291 task_work_run+0x243/0x300 kernel/task_work.c:179 resume_user_mode_work include/linux/resume_user_mode.h:49 [inline] exit_to_user_mode_loop+0x124/0x150 kernel/entry/common.c:171 exit_to_user_mode_prepare+0xb2/0x140 kernel/entry/common.c:203 __syscall_exit_to_user_mode_work kernel/entry/common.c:285 [inline] syscall_exit_to_user_mode+0x26/0x60 kernel/entry/common.c:296 do_syscall_64+0x49/0xb0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd [...] Allocated by task 13352: kasan_save_stack mm/kasan/common.c:45 [inline] kasan_set_track+0x3d/0x60 mm/kasan/common.c:52 ____kasan_kmalloc mm/kasan/common.c:371 [inline] __kasan_kmalloc+0x97/0xb0 mm/kasan/common.c:380 kmalloc include/linux/slab.h:580 [inline] dbMount+0x54/0x980 fs/jfs/jfs_dmap.c:164 jfs_mount+0x1dd/0x830 fs/jfs/jfs_mount.c:121 jfs_fill_super+0x590/0xc50 fs/jfs/super.c:556 mount_bdev+0x26c/0x3a0 fs/super.c:1359 legacy_get_tree+0xea/0x180 fs/fs_context.c:610 vfs_get_tree+0x88/0x270 fs/super.c:1489 do_new_mount+0x289/0xad0 fs/namespace.c:3145 do_mount fs/namespace.c:3488 [inline] __do_sys_mount fs/namespace.c:3697 [inline] __se_sys_mount+0x2d3/0x3c0 fs/namespace.c:3674 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd Freed by task 13352: kasan_save_stack mm/kasan/common.c:45 [inline] kasan_set_track+0x3d/0x60 mm/kasan/common.c:52 kasan_save_free_info+0x27/0x40 mm/kasan/generic.c:518 ____kasan_slab_free+0xd6/0x120 mm/kasan/common.c:236 kasan_slab_free include/linux/kasan.h:177 [inline] slab_free_hook mm/slub.c:1781 [inline] slab_free_freelist_hook+0x12e/0x1a0 mm/slub.c:1807 slab_free mm/slub.c:3787 [inline] __kmem_cache_free+0x71/0x110 mm/slub.c:3800 dbUnmount+0xf4/0x110 fs/jfs/jfs_dmap.c:264 jfs_mount_rw+0x545/0x740 fs/jfs/jfs_mount.c:247 jfs_remount+0x3db/0x710 fs/jfs/super.c:454 reconfigure_super+0x3bc/0x7b0 fs/super.c:935 vfs_fsconfig_locked fs/fsopen.c:254 [inline] __do_sys_fsconfig fs/fsopen.c:439 [inline] __se_sys_fsconfig+0xad5/0x1060 fs/fsopen.c:314 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd [...] JFS_SBI(ipbmap->i_sb)->bmap wasn't set to NULL after kfree() in dbUnmount(). Syzkaller uses faultinject to reproduce this KASAN double-free warning. The issue is triggered if either diMount() or dbMount() fail in jfs_remount(), since diUnmount() or dbUnmount() already happened in such a case - they will do double-free on next execution: jfs_umount or jfs_remount. Tested on both upstream and jfs-next by syzkaller. Reported-and-tested-by: syzbot+6a93efb725385bc4b2e9@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/000000000000471f2d05f1ce8bad@google.com/T/ Link: https://syzkaller.appspot.com/bug?extid=6a93efb725385bc4b2e9 Signed-off-by: Andrew Kanner Signed-off-by: Dave Kleikamp --- fs/jfs/jfs_dmap.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c index a14a0f18a4c40..88afd108c2dd2 100644 --- a/fs/jfs/jfs_dmap.c +++ b/fs/jfs/jfs_dmap.c @@ -269,6 +269,7 @@ int dbUnmount(struct inode *ipbmap, int mounterror) /* free the memory for the in-memory bmap. */ kfree(bmp); + JFS_SBI(ipbmap->i_sb)->bmap = NULL; return (0); } From c67235d08b2ea9673a2e7b80823f762e45942f5e Mon Sep 17 00:00:00 2001 From: Immad Mir Date: Wed, 28 Jun 2023 13:14:21 +0530 Subject: [PATCH 010/186] FS: JFS: (trivial) Fix grammatical error in extAlloc There is a grammatical error in one the commnents in extAlloc function. Signed-off-by: Immad Mir Signed-off-by: Dave Kleikamp --- fs/jfs/jfs_extent.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c index ae99a7e232eeb..adaa9ad50d4c8 100644 --- a/fs/jfs/jfs_extent.c +++ b/fs/jfs/jfs_extent.c @@ -166,7 +166,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, bool abnr) /* * COMMIT_SyncList flags an anonymous tlock on page that is on * sync list. - * We need to commit the inode to get the page written disk. + * We need to commit the inode to get the page written to the disk. */ if (test_and_clear_cflag(COMMIT_Synclist,ip)) jfs_commit_inode(ip, 0); From 6e2bda2c192d0244b5a78b787ef20aa10cb319b7 Mon Sep 17 00:00:00 2001 From: Liu Shixin via Jfs-discussion Date: Thu, 1 Dec 2022 20:46:28 +0800 Subject: [PATCH 011/186] jfs: fix invalid free of JFS_IP(ipimap)->i_imap in diUnmount syzbot found an invalid-free in diUnmount: BUG: KASAN: double-free in slab_free mm/slub.c:3661 [inline] BUG: KASAN: double-free in __kmem_cache_free+0x71/0x110 mm/slub.c:3674 Free of addr ffff88806f410000 by task syz-executor131/3632 CPU: 0 PID: 3632 Comm: syz-executor131 Not tainted 6.1.0-rc7-syzkaller-00012-gca57f02295f1 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022 Call Trace: __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0x1b1/0x28e lib/dump_stack.c:106 print_address_description+0x74/0x340 mm/kasan/report.c:284 print_report+0x107/0x1f0 mm/kasan/report.c:395 kasan_report_invalid_free+0xac/0xd0 mm/kasan/report.c:460 ____kasan_slab_free+0xfb/0x120 kasan_slab_free include/linux/kasan.h:177 [inline] slab_free_hook mm/slub.c:1724 [inline] slab_free_freelist_hook+0x12e/0x1a0 mm/slub.c:1750 slab_free mm/slub.c:3661 [inline] __kmem_cache_free+0x71/0x110 mm/slub.c:3674 diUnmount+0xef/0x100 fs/jfs/jfs_imap.c:195 jfs_umount+0x108/0x370 fs/jfs/jfs_umount.c:63 jfs_put_super+0x86/0x190 fs/jfs/super.c:194 generic_shutdown_super+0x130/0x310 fs/super.c:492 kill_block_super+0x79/0xd0 fs/super.c:1428 deactivate_locked_super+0xa7/0xf0 fs/super.c:332 cleanup_mnt+0x494/0x520 fs/namespace.c:1186 task_work_run+0x243/0x300 kernel/task_work.c:179 exit_task_work include/linux/task_work.h:38 [inline] do_exit+0x664/0x2070 kernel/exit.c:820 do_group_exit+0x1fd/0x2b0 kernel/exit.c:950 __do_sys_exit_group kernel/exit.c:961 [inline] __se_sys_exit_group kernel/exit.c:959 [inline] __x64_sys_exit_group+0x3b/0x40 kernel/exit.c:959 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd [...] JFS_IP(ipimap)->i_imap is not setting to NULL after free in diUnmount. If jfs_remount() free JFS_IP(ipimap)->i_imap but then failed at diMount(). JFS_IP(ipimap)->i_imap will be freed once again. Fix this problem by setting JFS_IP(ipimap)->i_imap to NULL after free. Reported-by: syzbot+90a11e6b1e810785c6ff@syzkaller.appspotmail.com Signed-off-by: Liu Shixin Signed-off-by: Dave Kleikamp --- fs/jfs/jfs_imap.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/jfs/jfs_imap.c b/fs/jfs/jfs_imap.c index 390cbfce391fc..6fb28572cb2c6 100644 --- a/fs/jfs/jfs_imap.c +++ b/fs/jfs/jfs_imap.c @@ -193,6 +193,7 @@ int diUnmount(struct inode *ipimap, int mounterror) * free in-memory control structure */ kfree(imap); + JFS_IP(ipimap)->i_imap = NULL; return (0); } From 99c58d6480d937dbdc2b4acfdea1bcf7ab113e5e Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 20 Jul 2023 08:22:39 -0400 Subject: [PATCH 012/186] fs: dlm: remove twice newline This patch removes a newline which log_print() already adds, also removes wrapped string that causes a checkpatch warning. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 70a4752ed913a..a34f605d8505b 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -240,8 +240,8 @@ static int dlm_plock_callback(struct plock_op *op) rv = notify(fl, 0); if (rv) { /* XXX: We need to cancel the fs lock here: */ - log_print("dlm_plock_callback: lock granted after lock request " - "failed; dangling lock!\n"); + log_print("%s: lock granted after lock request failed; dangling lock!", + __func__); goto out; } From 568f915655b3b4c40032104e4d0014e5e2c474b9 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 20 Jul 2023 08:22:40 -0400 Subject: [PATCH 013/186] fs: dlm: allow to F_SETLKW getting interrupted This patch implements dlm plock F_SETLKW interruption feature. If a blocking posix lock request got interrupted in user space by a signal a cancellation request for a non granted lock request to the user space lock manager will be send. The user lock manager answers either with zero or a negative errno code. A errno of -ENOENT signals that there is currently no blocking lock request waiting to being granted. In case of -ENOENT it was probably to late to request a cancellation and the pending lock got granted. In any error case we will wait until the lock is being granted as cancellation failed, this causes also that in case of an older user lock manager returning -EINVAL we will wait as cancellation is not supported which should be fine. If a user requires this feature the user should update dlm user space to support lock request cancellation. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 56 ++++++++++++++++++++++------------ include/uapi/linux/dlm_plock.h | 1 + 2 files changed, 37 insertions(+), 20 deletions(-) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index a34f605d8505b..a8ffa07609139 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -74,30 +74,26 @@ static void send_op(struct plock_op *op) wake_up(&send_wq); } -/* If a process was killed while waiting for the only plock on a file, - locks_remove_posix will not see any lock on the file so it won't - send an unlock-close to us to pass on to userspace to clean up the - abandoned waiter. So, we have to insert the unlock-close when the - lock call is interrupted. */ - -static void do_unlock_close(const struct dlm_plock_info *info) +static int do_lock_cancel(const struct dlm_plock_info *orig_info) { struct plock_op *op; + int rv; op = kzalloc(sizeof(*op), GFP_NOFS); if (!op) - return; + return -ENOMEM; + + op->info = *orig_info; + op->info.optype = DLM_PLOCK_OP_CANCEL; + op->info.wait = 0; - op->info.optype = DLM_PLOCK_OP_UNLOCK; - op->info.pid = info->pid; - op->info.fsid = info->fsid; - op->info.number = info->number; - op->info.start = 0; - op->info.end = OFFSET_MAX; - op->info.owner = info->owner; - - op->info.flags |= DLM_PLOCK_FL_CLOSE; send_op(op); + wait_event(recv_wq, (op->done != 0)); + + rv = op->info.rv; + + dlm_release_plock_op(op); + return rv; } int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, @@ -156,7 +152,7 @@ int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, send_op(op); if (op->info.wait) { - rv = wait_event_killable(recv_wq, (op->done != 0)); + rv = wait_event_interruptible(recv_wq, (op->done != 0)); if (rv == -ERESTARTSYS) { spin_lock(&ops_lock); /* recheck under ops_lock if we got a done != 0, @@ -166,17 +162,37 @@ int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, spin_unlock(&ops_lock); goto do_lock_wait; } - list_del(&op->list); spin_unlock(&ops_lock); + rv = do_lock_cancel(&op->info); + switch (rv) { + case 0: + /* waiter was deleted in user space, answer will never come + * remove original request. The original request must be + * on recv_list because the answer of do_lock_cancel() + * synchronized it. + */ + spin_lock(&ops_lock); + list_del(&op->list); + spin_unlock(&ops_lock); + rv = -EINTR; + break; + case -ENOENT: + /* cancellation wasn't successful but op should be done */ + fallthrough; + default: + /* internal error doing cancel we need to wait */ + goto wait; + } + log_debug(ls, "%s: wait interrupted %x %llx pid %d", __func__, ls->ls_global_id, (unsigned long long)number, op->info.pid); - do_unlock_close(&op->info); dlm_release_plock_op(op); goto out; } } else { +wait: wait_event(recv_wq, (op->done != 0)); } diff --git a/include/uapi/linux/dlm_plock.h b/include/uapi/linux/dlm_plock.h index 63b6c1fd91690..eb66afcac40ed 100644 --- a/include/uapi/linux/dlm_plock.h +++ b/include/uapi/linux/dlm_plock.h @@ -22,6 +22,7 @@ enum { DLM_PLOCK_OP_LOCK = 1, DLM_PLOCK_OP_UNLOCK, DLM_PLOCK_OP_GET, + DLM_PLOCK_OP_CANCEL, }; #define DLM_PLOCK_FL_CLOSE 1 From dc52cd2eff4ac924a795efcef27f8fd58a5260bb Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 20 Jul 2023 08:22:41 -0400 Subject: [PATCH 014/186] fs: dlm: fix F_CANCELLK to cancel pending request This patch fixes the current handling of F_CANCELLK by not just doing a unlock as we need to try to cancel a lock at first. A unlock makes sense on a non-blocking lock request but if it's a blocking lock request we need to cancel the request until it's not granted yet. This patch is fixing this behaviour by first try to cancel a lock request and if it's failed it's unlocking the lock which seems to be granted. Note: currently the nfs locking handling was disabled by commit 40595cdc93ed ("nfs: block notification on fs with its own ->lock"). However DLM was never being updated regarding to this change. Future patches will try to fix lockd lock requests for DLM. This patch is currently assuming the upstream DLM lockd handling is correct. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 103 +++++++++++++++++++++++++++++++++----- fs/gfs2/file.c | 9 ++-- fs/ocfs2/stack_user.c | 13 ++--- include/linux/dlm_plock.h | 2 + 4 files changed, 98 insertions(+), 29 deletions(-) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index a8ffa07609139..943d9f8e55645 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -42,6 +42,27 @@ static inline void set_version(struct dlm_plock_info *info) info->version[2] = DLM_PLOCK_VERSION_PATCH; } +static struct plock_op *plock_lookup_waiter(const struct dlm_plock_info *info) +{ + struct plock_op *op = NULL, *iter; + + list_for_each_entry(iter, &recv_list, list) { + if (iter->info.fsid == info->fsid && + iter->info.number == info->number && + iter->info.owner == info->owner && + iter->info.pid == info->pid && + iter->info.start == info->start && + iter->info.end == info->end && + iter->info.ex == info->ex && + iter->info.wait) { + op = iter; + break; + } + } + + return op; +} + static int check_version(struct dlm_plock_info *info) { if ((DLM_PLOCK_VERSION_MAJOR != info->version[0]) || @@ -334,6 +355,74 @@ int dlm_posix_unlock(dlm_lockspace_t *lockspace, u64 number, struct file *file, } EXPORT_SYMBOL_GPL(dlm_posix_unlock); +/* + * NOTE: This implementation can only handle async lock requests as nfs + * do it. It cannot handle cancellation of a pending lock request sitting + * in wait_event(), but for now only nfs is the only user local kernel + * user. + */ +int dlm_posix_cancel(dlm_lockspace_t *lockspace, u64 number, struct file *file, + struct file_lock *fl) +{ + struct dlm_plock_info info; + struct plock_op *op; + struct dlm_ls *ls; + int rv; + + /* this only works for async request for now and nfs is the only + * kernel user right now. + */ + if (WARN_ON_ONCE(!fl->fl_lmops || !fl->fl_lmops->lm_grant)) + return -EOPNOTSUPP; + + ls = dlm_find_lockspace_local(lockspace); + if (!ls) + return -EINVAL; + + memset(&info, 0, sizeof(info)); + info.pid = fl->fl_pid; + info.ex = (fl->fl_type == F_WRLCK); + info.fsid = ls->ls_global_id; + dlm_put_lockspace(ls); + info.number = number; + info.start = fl->fl_start; + info.end = fl->fl_end; + info.owner = (__u64)fl->fl_pid; + + rv = do_lock_cancel(&info); + switch (rv) { + case 0: + spin_lock(&ops_lock); + /* lock request to cancel must be on recv_list because + * do_lock_cancel() synchronizes it. + */ + op = plock_lookup_waiter(&info); + if (WARN_ON_ONCE(!op)) { + rv = -ENOLCK; + break; + } + + list_del(&op->list); + spin_unlock(&ops_lock); + WARN_ON(op->info.optype != DLM_PLOCK_OP_LOCK); + op->data->callback(op->data->fl, -EINTR); + dlm_release_plock_op(op); + rv = -EINTR; + break; + case -ENOENT: + /* if cancel wasn't successful we probably were to late + * or it was a non-blocking lock request, so just unlock it. + */ + rv = dlm_posix_unlock(lockspace, number, file, fl); + break; + default: + break; + } + + return rv; +} +EXPORT_SYMBOL_GPL(dlm_posix_cancel); + int dlm_posix_get(dlm_lockspace_t *lockspace, u64 number, struct file *file, struct file_lock *fl) { @@ -457,19 +546,7 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, */ spin_lock(&ops_lock); if (info.wait) { - list_for_each_entry(iter, &recv_list, list) { - if (iter->info.fsid == info.fsid && - iter->info.number == info.number && - iter->info.owner == info.owner && - iter->info.pid == info.pid && - iter->info.start == info.start && - iter->info.end == info.end && - iter->info.ex == info.ex && - iter->info.wait) { - op = iter; - break; - } - } + op = plock_lookup_waiter(&info); } else { list_for_each_entry(iter, &recv_list, list) { if (!iter->info.wait) { diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 1bf3c4453516f..386eceb2f5746 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -1436,17 +1436,14 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl) if (!(fl->fl_flags & FL_POSIX)) return -ENOLCK; - if (cmd == F_CANCELLK) { - /* Hack: */ - cmd = F_SETLK; - fl->fl_type = F_UNLCK; - } if (unlikely(gfs2_withdrawn(sdp))) { if (fl->fl_type == F_UNLCK) locks_lock_file_wait(file, fl); return -EIO; } - if (IS_GETLK(cmd)) + if (cmd == F_CANCELLK) + return dlm_posix_cancel(ls->ls_dlm, ip->i_no_addr, file, fl); + else if (IS_GETLK(cmd)) return dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl); else if (fl->fl_type == F_UNLCK) return dlm_posix_unlock(ls->ls_dlm, ip->i_no_addr, file, fl); diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c index 05d4414d0c33f..9b76ee66aeb2f 100644 --- a/fs/ocfs2/stack_user.c +++ b/fs/ocfs2/stack_user.c @@ -738,18 +738,11 @@ static int user_plock(struct ocfs2_cluster_connection *conn, * * Internally, fs/dlm will pass these to a misc device, which * a userspace daemon will read and write to. - * - * For now, cancel requests (which happen internally only), - * are turned into unlocks. Most of this function taken from - * gfs2_lock. */ - if (cmd == F_CANCELLK) { - cmd = F_SETLK; - fl->fl_type = F_UNLCK; - } - - if (IS_GETLK(cmd)) + if (cmd == F_CANCELLK) + return dlm_posix_cancel(conn->cc_lockspace, ino, file, fl); + else if (IS_GETLK(cmd)) return dlm_posix_get(conn->cc_lockspace, ino, file, fl); else if (fl->fl_type == F_UNLCK) return dlm_posix_unlock(conn->cc_lockspace, ino, file, fl); diff --git a/include/linux/dlm_plock.h b/include/linux/dlm_plock.h index e6d76e8715a6c..15fc856d198cf 100644 --- a/include/linux/dlm_plock.h +++ b/include/linux/dlm_plock.h @@ -11,6 +11,8 @@ int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, int cmd, struct file_lock *fl); int dlm_posix_unlock(dlm_lockspace_t *lockspace, u64 number, struct file *file, struct file_lock *fl); +int dlm_posix_cancel(dlm_lockspace_t *lockspace, u64 number, struct file *file, + struct file_lock *fl); int dlm_posix_get(dlm_lockspace_t *lockspace, u64 number, struct file *file, struct file_lock *fl); #endif From 43bbddc067883d94de7a43d5756a295439fbe37d Mon Sep 17 00:00:00 2001 From: Baokun Li Date: Mon, 24 Jul 2023 20:10:57 +0800 Subject: [PATCH 015/186] ext4: add two helper functions extent_logical_end() and pa_logical_end() When we use lstart + len to calculate the end of free extent or prealloc space, it may exceed the maximum value of 4294967295(0xffffffff) supported by ext4_lblk_t and cause overflow, which may lead to various problems. Therefore, we add two helper functions, extent_logical_end() and pa_logical_end(), to limit the type of end to loff_t, and also convert lstart to loff_t for calculation to avoid overflow. Signed-off-by: Baokun Li Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230724121059.11834-2-libaokun1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 9 +++------ fs/ext4/mballoc.h | 14 ++++++++++++++ 2 files changed, 17 insertions(+), 6 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 21b903fe546e8..4cb13b3e41b3f 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4432,7 +4432,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, /* first, let's learn actual file size * given current request is allocated */ - size = ac->ac_o_ex.fe_logical + EXT4_C2B(sbi, ac->ac_o_ex.fe_len); + size = extent_logical_end(sbi, &ac->ac_o_ex); size = size << bsbits; if (size < i_size_read(ac->ac_inode)) size = i_size_read(ac->ac_inode); @@ -4766,7 +4766,6 @@ ext4_mb_use_preallocated(struct ext4_allocation_context *ac) struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); struct ext4_locality_group *lg; struct ext4_prealloc_space *tmp_pa = NULL, *cpa = NULL; - loff_t tmp_pa_end; struct rb_node *iter; ext4_fsblk_t goal_block; @@ -4862,9 +4861,7 @@ ext4_mb_use_preallocated(struct ext4_allocation_context *ac) * pa can possibly satisfy the request hence check if it overlaps * original logical start and stop searching if it doesn't. */ - tmp_pa_end = (loff_t)tmp_pa->pa_lstart + EXT4_C2B(sbi, tmp_pa->pa_len); - - if (ac->ac_o_ex.fe_logical >= tmp_pa_end) { + if (ac->ac_o_ex.fe_logical >= pa_logical_end(sbi, tmp_pa)) { spin_unlock(&tmp_pa->pa_lock); goto try_group_pa; } @@ -5769,7 +5766,7 @@ static void ext4_mb_group_or_file(struct ext4_allocation_context *ac) group_pa_eligible = sbi->s_mb_group_prealloc > 0; inode_pa_eligible = true; - size = ac->ac_o_ex.fe_logical + EXT4_C2B(sbi, ac->ac_o_ex.fe_len); + size = extent_logical_end(sbi, &ac->ac_o_ex); isize = (i_size_read(ac->ac_inode) + ac->ac_sb->s_blocksize - 1) >> bsbits; diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index df6b5e7c22741..d7aeb5da7d867 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -233,6 +233,20 @@ static inline ext4_fsblk_t ext4_grp_offs_to_block(struct super_block *sb, (fex->fe_start << EXT4_SB(sb)->s_cluster_bits); } +static inline loff_t extent_logical_end(struct ext4_sb_info *sbi, + struct ext4_free_extent *fex) +{ + /* Use loff_t to avoid end exceeding ext4_lblk_t max. */ + return (loff_t)fex->fe_logical + EXT4_C2B(sbi, fex->fe_len); +} + +static inline loff_t pa_logical_end(struct ext4_sb_info *sbi, + struct ext4_prealloc_space *pa) +{ + /* Use loff_t to avoid end exceeding ext4_lblk_t max. */ + return (loff_t)pa->pa_lstart + EXT4_C2B(sbi, pa->pa_len); +} + typedef int (*ext4_mballoc_query_range_fn)( struct super_block *sb, ext4_group_t agno, From bc056e7163ac7db945366de219745cf94f32a3e6 Mon Sep 17 00:00:00 2001 From: Baokun Li Date: Mon, 24 Jul 2023 20:10:58 +0800 Subject: [PATCH 016/186] ext4: fix BUG in ext4_mb_new_inode_pa() due to overflow When we calculate the end position of ext4_free_extent, this position may be exactly where ext4_lblk_t (i.e. uint) overflows. For example, if ac_g_ex.fe_logical is 4294965248 and ac_orig_goal_len is 2048, then the computed end is 0x100000000, which is 0. If ac->ac_o_ex.fe_logical is not the first case of adjusting the best extent, that is, new_bex_end > 0, the following BUG_ON will be triggered: ========================================================= kernel BUG at fs/ext4/mballoc.c:5116! invalid opcode: 0000 [#1] PREEMPT SMP PTI CPU: 3 PID: 673 Comm: xfs_io Tainted: G E 6.5.0-rc1+ #279 RIP: 0010:ext4_mb_new_inode_pa+0xc5/0x430 Call Trace: ext4_mb_use_best_found+0x203/0x2f0 ext4_mb_try_best_found+0x163/0x240 ext4_mb_regular_allocator+0x158/0x1550 ext4_mb_new_blocks+0x86a/0xe10 ext4_ext_map_blocks+0xb0c/0x13a0 ext4_map_blocks+0x2cd/0x8f0 ext4_iomap_begin+0x27b/0x400 iomap_iter+0x222/0x3d0 __iomap_dio_rw+0x243/0xcb0 iomap_dio_rw+0x16/0x80 ========================================================= A simple reproducer demonstrating the problem: mkfs.ext4 -F /dev/sda -b 4096 100M mount /dev/sda /tmp/test fallocate -l1M /tmp/test/tmp fallocate -l10M /tmp/test/file fallocate -i -o 1M -l16777203M /tmp/test/file fsstress -d /tmp/test -l 0 -n 100000 -p 8 & sleep 10 && killall -9 fsstress rm -f /tmp/test/tmp xfs_io -c "open -ad /tmp/test/file" -c "pwrite -S 0xff 0 8192" We simply refactor the logic for adjusting the best extent by adding a temporary ext4_free_extent ex and use extent_logical_end() to avoid overflow, which also simplifies the code. Cc: stable@kernel.org # 6.4 Fixes: 93cdf49f6eca ("ext4: Fix best extent lstart adjustment logic in ext4_mb_new_inode_pa()") Signed-off-by: Baokun Li Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230724121059.11834-3-libaokun1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 31 ++++++++++++++----------------- 1 file changed, 14 insertions(+), 17 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 4cb13b3e41b3f..86bce870dc5a3 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -5177,8 +5177,11 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) pa = ac->ac_pa; if (ac->ac_b_ex.fe_len < ac->ac_orig_goal_len) { - int new_bex_start; - int new_bex_end; + struct ext4_free_extent ex = { + .fe_logical = ac->ac_g_ex.fe_logical, + .fe_len = ac->ac_orig_goal_len, + }; + loff_t orig_goal_end = extent_logical_end(sbi, &ex); /* we can't allocate as much as normalizer wants. * so, found space must get proper lstart @@ -5197,29 +5200,23 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) * still cover original start * 3. Else, keep the best ex at start of original request. */ - new_bex_end = ac->ac_g_ex.fe_logical + - EXT4_C2B(sbi, ac->ac_orig_goal_len); - new_bex_start = new_bex_end - EXT4_C2B(sbi, ac->ac_b_ex.fe_len); - if (ac->ac_o_ex.fe_logical >= new_bex_start) - goto adjust_bex; + ex.fe_len = ac->ac_b_ex.fe_len; - new_bex_start = ac->ac_g_ex.fe_logical; - new_bex_end = - new_bex_start + EXT4_C2B(sbi, ac->ac_b_ex.fe_len); - if (ac->ac_o_ex.fe_logical < new_bex_end) + ex.fe_logical = orig_goal_end - EXT4_C2B(sbi, ex.fe_len); + if (ac->ac_o_ex.fe_logical >= ex.fe_logical) goto adjust_bex; - new_bex_start = ac->ac_o_ex.fe_logical; - new_bex_end = - new_bex_start + EXT4_C2B(sbi, ac->ac_b_ex.fe_len); + ex.fe_logical = ac->ac_g_ex.fe_logical; + if (ac->ac_o_ex.fe_logical < extent_logical_end(sbi, &ex)) + goto adjust_bex; + ex.fe_logical = ac->ac_o_ex.fe_logical; adjust_bex: - ac->ac_b_ex.fe_logical = new_bex_start; + ac->ac_b_ex.fe_logical = ex.fe_logical; BUG_ON(ac->ac_o_ex.fe_logical < ac->ac_b_ex.fe_logical); BUG_ON(ac->ac_o_ex.fe_len > ac->ac_b_ex.fe_len); - BUG_ON(new_bex_end > (ac->ac_g_ex.fe_logical + - EXT4_C2B(sbi, ac->ac_orig_goal_len))); + BUG_ON(extent_logical_end(sbi, &ex) > orig_goal_end); } pa->pa_lstart = ac->ac_b_ex.fe_logical; From bedc5d34632c21b5adb8ca7143d4c1f794507e4c Mon Sep 17 00:00:00 2001 From: Baokun Li Date: Mon, 24 Jul 2023 20:10:59 +0800 Subject: [PATCH 017/186] ext4: avoid overlapping preallocations due to overflow Let's say we want to allocate 2 blocks starting from 4294966386, after predicting the file size, start is aligned to 4294965248, len is changed to 2048, then end = start + size = 0x100000000. Since end is of type ext4_lblk_t, i.e. uint, end is truncated to 0. This causes (pa->pa_lstart >= end) to always hold when checking if the current extent to be allocated crosses already preallocated blocks, so the resulting ac_g_ex may cross already preallocated blocks. Hence we convert the end type to loff_t and use pa_logical_end() to avoid overflow. Signed-off-by: Baokun Li Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230724121059.11834-4-libaokun1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 86bce870dc5a3..78a4a24e2f578 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4222,12 +4222,13 @@ ext4_mb_pa_rb_next_iter(ext4_lblk_t new_start, ext4_lblk_t cur_start, struct rb_ static inline void ext4_mb_pa_assert_overlap(struct ext4_allocation_context *ac, - ext4_lblk_t start, ext4_lblk_t end) + ext4_lblk_t start, loff_t end) { struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); struct ext4_prealloc_space *tmp_pa; - ext4_lblk_t tmp_pa_start, tmp_pa_end; + ext4_lblk_t tmp_pa_start; + loff_t tmp_pa_end; struct rb_node *iter; read_lock(&ei->i_prealloc_lock); @@ -4236,7 +4237,7 @@ ext4_mb_pa_assert_overlap(struct ext4_allocation_context *ac, tmp_pa = rb_entry(iter, struct ext4_prealloc_space, pa_node.inode_node); tmp_pa_start = tmp_pa->pa_lstart; - tmp_pa_end = tmp_pa->pa_lstart + EXT4_C2B(sbi, tmp_pa->pa_len); + tmp_pa_end = pa_logical_end(sbi, tmp_pa); spin_lock(&tmp_pa->pa_lock); if (tmp_pa->pa_deleted == 0) @@ -4258,14 +4259,14 @@ ext4_mb_pa_assert_overlap(struct ext4_allocation_context *ac, */ static inline void ext4_mb_pa_adjust_overlap(struct ext4_allocation_context *ac, - ext4_lblk_t *start, ext4_lblk_t *end) + ext4_lblk_t *start, loff_t *end) { struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); struct ext4_prealloc_space *tmp_pa = NULL, *left_pa = NULL, *right_pa = NULL; struct rb_node *iter; - ext4_lblk_t new_start, new_end; - ext4_lblk_t tmp_pa_start, tmp_pa_end, left_pa_end = -1, right_pa_start = -1; + ext4_lblk_t new_start, tmp_pa_start, right_pa_start = -1; + loff_t new_end, tmp_pa_end, left_pa_end = -1; new_start = *start; new_end = *end; @@ -4284,7 +4285,7 @@ ext4_mb_pa_adjust_overlap(struct ext4_allocation_context *ac, tmp_pa = rb_entry(iter, struct ext4_prealloc_space, pa_node.inode_node); tmp_pa_start = tmp_pa->pa_lstart; - tmp_pa_end = tmp_pa->pa_lstart + EXT4_C2B(sbi, tmp_pa->pa_len); + tmp_pa_end = pa_logical_end(sbi, tmp_pa); /* PA must not overlap original request */ spin_lock(&tmp_pa->pa_lock); @@ -4364,8 +4365,7 @@ ext4_mb_pa_adjust_overlap(struct ext4_allocation_context *ac, } if (left_pa) { - left_pa_end = - left_pa->pa_lstart + EXT4_C2B(sbi, left_pa->pa_len); + left_pa_end = pa_logical_end(sbi, left_pa); BUG_ON(left_pa_end > ac->ac_o_ex.fe_logical); } @@ -4404,8 +4404,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); struct ext4_super_block *es = sbi->s_es; int bsbits, max; - ext4_lblk_t end; - loff_t size, start_off; + loff_t size, start_off, end; loff_t orig_size __maybe_unused; ext4_lblk_t start; From 1d40165047456923fa4343d519353d9440cd68df Mon Sep 17 00:00:00 2001 From: Guoqing Cai Date: Thu, 13 Apr 2023 17:57:39 +0800 Subject: [PATCH 018/186] fs: jbd2: fix an incorrect warn log In jbd2_journal_load(), when journal_reset fails, it prints an incorrect warn log. Fix this by changing the goto statement to return statement. Also, return actual error code from jbd2_journal_recover() and journal_reset(). Signed-off-by: Guoqing Cai Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230413095740.2222066-1-u202112087@hust.edu.cn Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index fbce16fedaa4a..5c223032f77aa 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2089,8 +2089,11 @@ int jbd2_journal_load(journal_t *journal) /* Let the recovery code check whether it needs to recover any * data from the journal. */ - if (jbd2_journal_recover(journal)) - goto recovery_error; + err = jbd2_journal_recover(journal); + if (err) { + pr_warn("JBD2: journal recovery failed\n"); + return err; + } if (journal->j_failed_commit) { printk(KERN_ERR "JBD2: journal transaction %u on %s " @@ -2107,15 +2110,14 @@ int jbd2_journal_load(journal_t *journal) /* OK, we've finished with the dynamic journal bits: * reinitialise the dynamic contents of the superblock in memory * and reset them on disk. */ - if (journal_reset(journal)) - goto recovery_error; + err = journal_reset(journal); + if (err) { + pr_warn("JBD2: journal reset failed\n"); + return err; + } journal->j_flags |= JBD2_LOADED; return 0; - -recovery_error: - printk(KERN_WARNING "JBD2: recovery failed\n"); - return -EIO; } /** From 98175720c9ed3bac857b0364321517cc2d695a3f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:47 +0200 Subject: [PATCH 019/186] ext4: remove pointless sb_rdonly() checks from freezing code ext4_freeze() and ext4_unfreeze() checks for sb_rdonly(). However this check is pointless as VFS already checks for read-only filesystem before calling filesystem specific methods. Remove the pointless checks. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-1-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index c94ebf704616e..ffc4fb73f133c 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -6347,12 +6347,7 @@ static int ext4_sync_fs(struct super_block *sb, int wait) static int ext4_freeze(struct super_block *sb) { int error = 0; - journal_t *journal; - - if (sb_rdonly(sb)) - return 0; - - journal = EXT4_SB(sb)->s_journal; + journal_t *journal = EXT4_SB(sb)->s_journal; if (journal) { /* Now we set up the journal barrier. */ @@ -6386,7 +6381,7 @@ static int ext4_freeze(struct super_block *sb) */ static int ext4_unfreeze(struct super_block *sb) { - if (sb_rdonly(sb) || ext4_forced_shutdown(EXT4_SB(sb))) + if (ext4_forced_shutdown(EXT4_SB(sb))) return 0; if (EXT4_SB(sb)->s_journal) { From d5d020b3294b69eaf3b8985e7a37ba237849c390 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:48 +0200 Subject: [PATCH 020/186] ext4: use sb_rdonly() helper for checking read-only flag sb_rdonly() helper instead of directly checking sb->s_flags. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-2-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index ffc4fb73f133c..19514f98e2fe8 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -6084,7 +6084,7 @@ static void ext4_update_super(struct super_block *sb) * the clock is set in the future, and this will cause e2fsck * to complain and force a full file system check. */ - if (!(sb->s_flags & SB_RDONLY)) + if (!sb_rdonly(sb)) ext4_update_tstamp(es, s_wtime); es->s_kbytes_written = cpu_to_le64(sbi->s_kbytes_written + @@ -6675,7 +6675,7 @@ static int __ext4_remount(struct fs_context *fc, struct super_block *sb) * If there was a failing r/w to ro transition, we may need to * re-enable quota */ - if ((sb->s_flags & SB_RDONLY) && !(old_sb_flags & SB_RDONLY) && + if (sb_rdonly(sb) && !(old_sb_flags & SB_RDONLY) && sb_any_quota_suspended(sb)) dquot_resume(sb, -1); sb->s_flags = old_sb_flags; From eb8ab4443aec5ffe923a471b337568a8158cd32b Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:49 +0200 Subject: [PATCH 021/186] ext4: make ext4_forced_shutdown() take struct super_block Currently ext4_forced_shutdown() takes struct ext4_sb_info but most callers need to get it from struct super_block anyway. So just pass in struct super_block to save all callers from some boilerplate code. No functional changes. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-3-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 4 ++-- fs/ext4/ext4_jbd2.c | 2 +- fs/ext4/file.c | 13 ++++++------- fs/ext4/fsync.c | 2 +- fs/ext4/ialloc.c | 2 +- fs/ext4/inline.c | 2 +- fs/ext4/inode.c | 24 ++++++++++++------------ fs/ext4/ioctl.c | 2 +- fs/ext4/namei.c | 8 ++++---- fs/ext4/page-io.c | 2 +- fs/ext4/super.c | 14 +++++++------- fs/ext4/xattr.c | 2 +- 12 files changed, 38 insertions(+), 39 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0a2d55faa095e..feb38c9fe1294 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2222,9 +2222,9 @@ extern int ext4_feature_set_ok(struct super_block *sb, int readonly); #define EXT4_FLAGS_SHUTDOWN 1 #define EXT4_FLAGS_BDEV_IS_DAX 2 -static inline int ext4_forced_shutdown(struct ext4_sb_info *sbi) +static inline int ext4_forced_shutdown(struct super_block *sb) { - return test_bit(EXT4_FLAGS_SHUTDOWN, &sbi->s_ext4_flags); + return test_bit(EXT4_FLAGS_SHUTDOWN, &EXT4_SB(sb)->s_ext4_flags); } /* diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 77f318ec8abb7..b72a22a57d20c 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -67,7 +67,7 @@ static int ext4_journal_check_start(struct super_block *sb) might_sleep(); - if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) + if (unlikely(ext4_forced_shutdown(sb))) return -EIO; if (sb_rdonly(sb)) diff --git a/fs/ext4/file.c b/fs/ext4/file.c index c457c8517f0fd..2071b1e4322c5 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -131,7 +131,7 @@ static ssize_t ext4_file_read_iter(struct kiocb *iocb, struct iov_iter *to) { struct inode *inode = file_inode(iocb->ki_filp); - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; if (!iov_iter_count(to)) @@ -153,7 +153,7 @@ static ssize_t ext4_file_splice_read(struct file *in, loff_t *ppos, { struct inode *inode = file_inode(in); - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; return filemap_splice_read(in, ppos, pipe, len, flags); } @@ -709,7 +709,7 @@ ext4_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct inode *inode = file_inode(iocb->ki_filp); - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; #ifdef CONFIG_FS_DAX @@ -807,10 +807,9 @@ static const struct vm_operations_struct ext4_file_vm_ops = { static int ext4_file_mmap(struct file *file, struct vm_area_struct *vma) { struct inode *inode = file->f_mapping->host; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - struct dax_device *dax_dev = sbi->s_daxdev; + struct dax_device *dax_dev = EXT4_SB(inode->i_sb)->s_daxdev; - if (unlikely(ext4_forced_shutdown(sbi))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; /* @@ -886,7 +885,7 @@ static int ext4_file_open(struct inode *inode, struct file *filp) { int ret; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; ret = ext4_sample_last_mounted(inode->i_sb, filp->f_path.mnt); diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index 0c56f3a011a1f..bffc1d0994f54 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -133,7 +133,7 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync) struct inode *inode = file->f_mapping->host; struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (unlikely(ext4_forced_shutdown(sbi))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; ASSERT(ext4_journal_current_handle() == NULL); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 754f961cd9fdf..060630c0b0ca0 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -950,7 +950,7 @@ struct inode *__ext4_new_inode(struct mnt_idmap *idmap, sb = dir->i_sb; sbi = EXT4_SB(sb); - if (unlikely(ext4_forced_shutdown(sbi))) + if (unlikely(ext4_forced_shutdown(sb))) return ERR_PTR(-EIO); ngroups = ext4_get_groups_count(sb); diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index a4b7e4bc32d40..3623dfcc8fc7b 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -228,7 +228,7 @@ static void ext4_write_inline_data(struct inode *inode, struct ext4_iloc *iloc, struct ext4_inode *raw_inode; int cp_len = 0; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return; BUG_ON(!EXT4_I(inode)->i_inline_off); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 43775a6ca5054..c6fa59e57f1ea 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1114,7 +1114,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, pgoff_t index; unsigned from, to; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; trace_ext4_write_begin(inode, pos, len); @@ -2213,7 +2213,7 @@ static int mpage_map_and_submit_extent(handle_t *handle, if (err < 0) { struct super_block *sb = inode->i_sb; - if (ext4_forced_shutdown(EXT4_SB(sb)) || + if (ext4_forced_shutdown(sb) || ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) goto invalidate_dirty_pages; /* @@ -2540,7 +2540,7 @@ static int ext4_do_writepages(struct mpage_da_data *mpd) * *never* be called, so if that ever happens, we would want * the stack trace. */ - if (unlikely(ext4_forced_shutdown(EXT4_SB(mapping->host->i_sb)) || + if (unlikely(ext4_forced_shutdown(mapping->host->i_sb) || ext4_test_mount_flag(inode->i_sb, EXT4_MF_FS_ABORTED))) { ret = -EROFS; goto out_writepages; @@ -2759,7 +2759,7 @@ static int ext4_writepages(struct address_space *mapping, int ret; int alloc_ctx; - if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) + if (unlikely(ext4_forced_shutdown(sb))) return -EIO; alloc_ctx = ext4_writepages_down_read(sb); @@ -2798,16 +2798,16 @@ static int ext4_dax_writepages(struct address_space *mapping, int ret; long nr_to_write = wbc->nr_to_write; struct inode *inode = mapping->host; - struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); int alloc_ctx; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); - ret = dax_writeback_mapping_range(mapping, sbi->s_daxdev, wbc); + ret = dax_writeback_mapping_range(mapping, + EXT4_SB(inode->i_sb)->s_daxdev, wbc); trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); ext4_writepages_up_read(inode->i_sb, alloc_ctx); @@ -2857,7 +2857,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping, pgoff_t index; struct inode *inode = mapping->host; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; index = pos >> PAGE_SHIFT; @@ -5135,7 +5135,7 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) sb_rdonly(inode->i_sb)) return 0; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; if (EXT4_SB(inode->i_sb)->s_journal) { @@ -5255,7 +5255,7 @@ int ext4_setattr(struct mnt_idmap *idmap, struct dentry *dentry, const unsigned int ia_valid = attr->ia_valid; bool inc_ivers = true; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; if (unlikely(IS_IMMUTABLE(inode))) @@ -5676,7 +5676,7 @@ int ext4_mark_iloc_dirty(handle_t *handle, { int err = 0; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) { + if (unlikely(ext4_forced_shutdown(inode->i_sb))) { put_bh(iloc->bh); return -EIO; } @@ -5702,7 +5702,7 @@ ext4_reserve_inode_write(handle_t *handle, struct inode *inode, { int err; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; err = ext4_get_inode_loc(inode, iloc); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 331859511f80f..0d3aef1118cbc 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -801,7 +801,7 @@ int ext4_force_shutdown(struct super_block *sb, u32 flags) if (flags > EXT4_GOING_FLAGS_NOLOGFLUSH) return -EINVAL; - if (ext4_forced_shutdown(sbi)) + if (ext4_forced_shutdown(sb)) return 0; ext4_msg(sb, KERN_ALERT, "shut down requested (%d)", flags); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 0caf6c730ce34..6298cfaaa0bde 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -3142,7 +3142,7 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) struct ext4_dir_entry_2 *de; handle_t *handle = NULL; - if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb)))) + if (unlikely(ext4_forced_shutdown(dir->i_sb))) return -EIO; /* Initialize quotas before so that eventual writes go in @@ -3301,7 +3301,7 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) { int retval; - if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb)))) + if (unlikely(ext4_forced_shutdown(dir->i_sb))) return -EIO; trace_ext4_unlink_enter(dir, dentry); @@ -3369,7 +3369,7 @@ static int ext4_symlink(struct mnt_idmap *idmap, struct inode *dir, struct fscrypt_str disk_link; int retries = 0; - if (unlikely(ext4_forced_shutdown(EXT4_SB(dir->i_sb)))) + if (unlikely(ext4_forced_shutdown(dir->i_sb))) return -EIO; err = fscrypt_prepare_symlink(dir, symname, len, dir->i_sb->s_blocksize, @@ -4189,7 +4189,7 @@ static int ext4_rename2(struct mnt_idmap *idmap, { int err; - if (unlikely(ext4_forced_shutdown(EXT4_SB(old_dir->i_sb)))) + if (unlikely(ext4_forced_shutdown(old_dir->i_sb))) return -EIO; if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT)) diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 3621f29ec6712..dfdd7e5cf0389 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -184,7 +184,7 @@ static int ext4_end_io_end(ext4_io_end_t *io_end) io_end->handle = NULL; /* Following call will use up the handle */ ret = ext4_convert_unwritten_io_end_vec(handle, io_end); - if (ret < 0 && !ext4_forced_shutdown(EXT4_SB(inode->i_sb))) { + if (ret < 0 && !ext4_forced_shutdown(inode->i_sb)) { ext4_msg(inode->i_sb, KERN_EMERG, "failed to convert unwritten extents to written " "extents -- potential data loss! " diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 19514f98e2fe8..0038233eafa81 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -758,7 +758,7 @@ void __ext4_error(struct super_block *sb, const char *function, struct va_format vaf; va_list args; - if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) + if (unlikely(ext4_forced_shutdown(sb))) return; trace_ext4_error(sb, function, line); @@ -783,7 +783,7 @@ void __ext4_error_inode(struct inode *inode, const char *function, va_list args; struct va_format vaf; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return; trace_ext4_error(inode->i_sb, function, line); @@ -818,7 +818,7 @@ void __ext4_error_file(struct file *file, const char *function, struct inode *inode = file_inode(file); char pathname[80], *path; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return; trace_ext4_error(inode->i_sb, function, line); @@ -898,7 +898,7 @@ void __ext4_std_error(struct super_block *sb, const char *function, char nbuf[16]; const char *errstr; - if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) + if (unlikely(ext4_forced_shutdown(sb))) return; /* Special case: if the error is EROFS, and we're not already @@ -992,7 +992,7 @@ __acquires(bitlock) struct va_format vaf; va_list args; - if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) + if (unlikely(ext4_forced_shutdown(sb))) return; trace_ext4_error(sb, function, line); @@ -6298,7 +6298,7 @@ static int ext4_sync_fs(struct super_block *sb, int wait) bool needs_barrier = false; struct ext4_sb_info *sbi = EXT4_SB(sb); - if (unlikely(ext4_forced_shutdown(sbi))) + if (unlikely(ext4_forced_shutdown(sb))) return 0; trace_ext4_sync_fs(sb, wait); @@ -6381,7 +6381,7 @@ static int ext4_freeze(struct super_block *sb) */ static int ext4_unfreeze(struct super_block *sb) { - if (ext4_forced_shutdown(EXT4_SB(sb))) + if (ext4_forced_shutdown(sb)) return 0; if (EXT4_SB(sb)->s_journal) { diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 05151d61b00b3..7cc502c06246e 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -701,7 +701,7 @@ ext4_xattr_get(struct inode *inode, int name_index, const char *name, { int error; - if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) + if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; if (strlen(name) > 255) From 22b8d707b07e6e06f50fe1d9ca8756e1f894eb0d Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:50 +0200 Subject: [PATCH 022/186] ext4: make 'abort' mount option handling standard 'abort' mount option is the only mount option that has special handling and sets a bit in sbi->s_mount_flags. There is not strong reason for that so just simplify the code and make 'abort' set a bit in sbi->s_mount_opt2 as any other mount option. This simplifies the code and will allow us to drop EXT4_MF_FS_ABORTED completely in the following patch. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-4-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 1 + fs/ext4/super.c | 16 ++-------------- 2 files changed, 3 insertions(+), 14 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index feb38c9fe1294..907829007f3fc 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1235,6 +1235,7 @@ struct ext4_inode_info { #define EXT4_MOUNT2_MB_OPTIMIZE_SCAN 0x00000080 /* Optimize group * scanning in mballoc */ +#define EXT4_MOUNT2_ABORT 0x00000100 /* Abort filesystem */ #define clear_opt(sb, opt) EXT4_SB(sb)->s_mount_opt &= \ ~EXT4_MOUNT_##opt diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 0038233eafa81..f84142907cd51 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1897,6 +1897,7 @@ static const struct mount_opts { {Opt_fc_debug_force, EXT4_MOUNT2_JOURNAL_FAST_COMMIT, MOPT_SET | MOPT_2 | MOPT_EXT4_ONLY}, #endif + {Opt_abort, EXT4_MOUNT2_ABORT, MOPT_SET | MOPT_2}, {Opt_err, 0, 0} }; @@ -1965,8 +1966,6 @@ struct ext4_fs_context { unsigned int mask_s_mount_opt; unsigned int vals_s_mount_opt2; unsigned int mask_s_mount_opt2; - unsigned long vals_s_mount_flags; - unsigned long mask_s_mount_flags; unsigned int opt_flags; /* MOPT flags */ unsigned int spec; u32 s_max_batch_time; @@ -2117,12 +2116,6 @@ EXT4_SET_CTX(mount_opt2); EXT4_CLEAR_CTX(mount_opt2); EXT4_TEST_CTX(mount_opt2); -static inline void ctx_set_mount_flag(struct ext4_fs_context *ctx, int bit) -{ - set_bit(bit, &ctx->mask_s_mount_flags); - set_bit(bit, &ctx->vals_s_mount_flags); -} - static int ext4_parse_param(struct fs_context *fc, struct fs_parameter *param) { struct ext4_fs_context *ctx = fc->fs_private; @@ -2186,9 +2179,6 @@ static int ext4_parse_param(struct fs_context *fc, struct fs_parameter *param) ext4_msg(NULL, KERN_WARNING, "Ignoring removed %s option", param->key); return 0; - case Opt_abort: - ctx_set_mount_flag(ctx, EXT4_MF_FS_ABORTED); - return 0; case Opt_inlinecrypt: #ifdef CONFIG_FS_ENCRYPTION_INLINE_CRYPT ctx_set_flags(ctx, SB_INLINECRYPT); @@ -2842,8 +2832,6 @@ static void ext4_apply_options(struct fs_context *fc, struct super_block *sb) sbi->s_mount_opt |= ctx->vals_s_mount_opt; sbi->s_mount_opt2 &= ~ctx->mask_s_mount_opt2; sbi->s_mount_opt2 |= ctx->vals_s_mount_opt2; - sbi->s_mount_flags &= ~ctx->mask_s_mount_flags; - sbi->s_mount_flags |= ctx->vals_s_mount_flags; sb->s_flags &= ~ctx->mask_s_flags; sb->s_flags |= ctx->vals_s_flags; @@ -6497,7 +6485,7 @@ static int __ext4_remount(struct fs_context *fc, struct super_block *sb) goto restore_opts; } - if (ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) + if (test_opt2(sb, ABORT)) ext4_abort(sb, ESHUTDOWN, "Abort forced by user"); sb->s_flags = (sb->s_flags & ~SB_POSIXACL) | From 95257987a6387f02970eda707e55a06cce734e18 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:51 +0200 Subject: [PATCH 023/186] ext4: drop EXT4_MF_FS_ABORTED flag EXT4_MF_FS_ABORTED flag has practically the same intent as EXT4_FLAGS_SHUTDOWN flag. The shutdown flag is checked in many more places than the aborted flag which is mostly the historical artifact where we were relying on SB_RDONLY checks instead of the aborted flag checks. There are only three places - ext4_sync_file(), __ext4_remount(), and mballoc debug code - which check aborted flag and not shutdown flag and this is arguably a bug. Avoid these inconsistencies by removing EXT4_MF_FS_ABORTED flag and using EXT4_FLAGS_SHUTDOWN everywhere. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-5-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 1 - fs/ext4/fsync.c | 7 +++---- fs/ext4/inode.c | 8 +++----- fs/ext4/mballoc.c | 4 ++-- fs/ext4/super.c | 4 ++-- 5 files changed, 10 insertions(+), 14 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 907829007f3fc..89d76d50af4ca 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1799,7 +1799,6 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino) */ enum { EXT4_MF_MNTDIR_SAMPLED, - EXT4_MF_FS_ABORTED, /* Fatal error detected */ EXT4_MF_FC_INELIGIBLE /* Fast commit ineligible */ }; diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index bffc1d0994f54..b40d3b29f7e5c 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -131,7 +131,6 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync) int ret = 0, err; bool needs_barrier = false; struct inode *inode = file->f_mapping->host; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); if (unlikely(ext4_forced_shutdown(inode->i_sb))) return -EIO; @@ -141,14 +140,14 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync) trace_ext4_sync_file_enter(file, datasync); if (sb_rdonly(inode->i_sb)) { - /* Make sure that we read updated s_mount_flags value */ + /* Make sure that we read updated s_ext4_flags value */ smp_rmb(); - if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FS_ABORTED)) + if (ext4_forced_shutdown(inode->i_sb)) ret = -EROFS; goto out; } - if (!sbi->s_journal) { + if (!EXT4_SB(inode->i_sb)->s_journal) { ret = ext4_fsync_nojournal(file, start, end, datasync, &needs_barrier); if (needs_barrier) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index c6fa59e57f1ea..100c3ec6da6c6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2213,8 +2213,7 @@ static int mpage_map_and_submit_extent(handle_t *handle, if (err < 0) { struct super_block *sb = inode->i_sb; - if (ext4_forced_shutdown(sb) || - ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) + if (ext4_forced_shutdown(sb)) goto invalidate_dirty_pages; /* * Let the uper layers retry transient errors. @@ -2534,14 +2533,13 @@ static int ext4_do_writepages(struct mpage_da_data *mpd) * If the filesystem has aborted, it is read-only, so return * right away instead of dumping stack traces later on that * will obscure the real source of the problem. We test - * EXT4_MF_FS_ABORTED instead of sb->s_flag's SB_RDONLY because + * fs shutdown state instead of sb->s_flag's SB_RDONLY because * the latter could be true if the filesystem is mounted * read-only, and in that case, ext4_writepages should * *never* be called, so if that ever happens, we would want * the stack trace. */ - if (unlikely(ext4_forced_shutdown(mapping->host->i_sb) || - ext4_test_mount_flag(inode->i_sb, EXT4_MF_FS_ABORTED))) { + if (unlikely(ext4_forced_shutdown(mapping->host->i_sb))) { ret = -EROFS; goto out_writepages; } diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 78a4a24e2f578..1dc63e329e64b 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -5664,7 +5664,7 @@ static inline void ext4_mb_show_pa(struct super_block *sb) { ext4_group_t i, ngroups; - if (ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) + if (ext4_forced_shutdown(sb)) return; ngroups = ext4_get_groups_count(sb); @@ -5698,7 +5698,7 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) { struct super_block *sb = ac->ac_sb; - if (ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) + if (ext4_forced_shutdown(sb)) return; mb_debug(sb, "Can't allocate:" diff --git a/fs/ext4/super.c b/fs/ext4/super.c index f84142907cd51..20a8e64da4ac7 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -657,7 +657,7 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, WARN_ON_ONCE(1); if (!continue_fs && !sb_rdonly(sb)) { - ext4_set_mount_flag(sb, EXT4_MF_FS_ABORTED); + set_bit(EXT4_FLAGS_SHUTDOWN, &EXT4_SB(sb)->s_ext4_flags); if (journal) jbd2_journal_abort(journal, -EIO); } @@ -6502,7 +6502,7 @@ static int __ext4_remount(struct fs_context *fc, struct super_block *sb) flush_work(&sbi->s_error_work); if ((bool)(fc->sb_flags & SB_RDONLY) != sb_rdonly(sb)) { - if (ext4_test_mount_flag(sb, EXT4_MF_FS_ABORTED)) { + if (ext4_forced_shutdown(sb)) { err = -EROFS; goto restore_opts; } From e0e985f3f8941438a66ab8abb94cb011b9fb39a7 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:52 +0200 Subject: [PATCH 024/186] ext4: avoid starting transaction on read-only fs in ext4_quota_off() When the filesystem gets first remounted read-only and then unmounted, ext4_quota_off() will try to start a transaction (and fail) on read-only filesystem to cleanup inode flags for legacy quota files. Just bail before trying to start a transaction instead since that is going to issue a warning. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-6-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 20a8e64da4ac7..a9a7c38c74423 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -7072,6 +7072,13 @@ static int ext4_quota_off(struct super_block *sb, int type) err = dquot_quota_off(sb, type); if (err || ext4_has_feature_quota(sb)) goto out_put; + /* + * When the filesystem was remounted read-only first, we cannot cleanup + * inode flags here. Bad luck but people should be using QUOTA feature + * these days anyway. + */ + if (sb_rdonly(sb)) + goto out_put; inode_lock(inode); /* From e7fc2b31e04c46c9e2098bba710c9951c6b968af Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:53 +0200 Subject: [PATCH 025/186] ext4: warn on read-only filesystem in ext4_journal_check_start() Now that filesystem abort marks the filesystem as shutdown, we shouldn't be ever hitting the sb_rdonly() check in ext4_journal_check_start(). Since this is a suitable place for catching all sorts of programming errors, convert the check to WARN_ON instead of dropping it. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-7-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/ext4_jbd2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index b72a22a57d20c..ca0eaf2147b0e 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -70,8 +70,9 @@ static int ext4_journal_check_start(struct super_block *sb) if (unlikely(ext4_forced_shutdown(sb))) return -EIO; - if (sb_rdonly(sb)) + if (WARN_ON_ONCE(sb_rdonly(sb))) return -EROFS; + WARN_ON(sb->s_writers.frozen == SB_FREEZE_COMPLETE); journal = EXT4_SB(sb)->s_journal; /* From ffb6844e28ef6b9d76bee378774d7afbc3db6da9 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:54 +0200 Subject: [PATCH 026/186] ext4: drop read-only check in ext4_init_inode_table() We better should not be initializing inode tables on read-only filesystem. The following transaction start will warn us and make the function bail anyway so drop the pointless check. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-8-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/ialloc.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 060630c0b0ca0..e0698f54e17ae 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -1523,12 +1523,6 @@ int ext4_init_inode_table(struct super_block *sb, ext4_group_t group, int num, ret = 0, used_blks = 0; unsigned long used_inos = 0; - /* This should not happen, but just to be sure check this */ - if (sb_rdonly(sb)) { - ret = 1; - goto out; - } - gdp = ext4_get_group_desc(sb, group, &group_desc_bh); if (!gdp || !grp) goto out; From f1128084b40e520bea8bb32b3ff4d03745ab7e64 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:55 +0200 Subject: [PATCH 027/186] ext4: drop read-only check in ext4_write_inode() We should not have dirty inodes on read-only filesystem. Also silently bailing without writing anything would be a problem when we enable quotas during remount while the filesystem is read-only. So drop the read-only check. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-9-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/inode.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 100c3ec6da6c6..1b9003840bc16 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5129,8 +5129,7 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) { int err; - if (WARN_ON_ONCE(current->flags & PF_MEMALLOC) || - sb_rdonly(inode->i_sb)) + if (WARN_ON_ONCE(current->flags & PF_MEMALLOC)) return 0; if (unlikely(ext4_forced_shutdown(inode->i_sb))) From 889860e452d7436ca72018b8a03cbd89c38d6384 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:56 +0200 Subject: [PATCH 028/186] ext4: drop read-only check from ext4_force_commit() JBD2 code will quickly return without doing anything when there's nothing to commit so there's no point in the read-only check in ext4_force_commit(). Just drop it. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-10-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a9a7c38c74423..4613264344b07 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -6270,13 +6270,7 @@ static int ext4_clear_journal_err(struct super_block *sb, */ int ext4_force_commit(struct super_block *sb) { - journal_t *journal; - - if (sb_rdonly(sb)) - return 0; - - journal = EXT4_SB(sb)->s_journal; - return ext4_journal_force_commit(journal); + return ext4_journal_force_commit(EXT4_SB(sb)->s_journal); } static int ext4_sync_fs(struct super_block *sb, int wait) From 1e1566b9c85fbd6150657ea17f50fd42b9166d31 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 16 Jun 2023 18:50:57 +0200 Subject: [PATCH 029/186] ext4: replace read-only check for shutdown check in mmp code The multi-mount protection kthread checks for read-only filesystem and aborts in that case. The remount code actually handles stopping of the kthread on remount so the only purpose of the check is in case of emergency remount read-only. Replace the check for read-only filesystem with a check for shutdown filesystem as running MMP on such is risky anyway and it makes ordering of things during remount simpler. Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230616165109.21695-11-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/mmp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c index 0aaf38ffcb6ec..bd946d0c71b70 100644 --- a/fs/ext4/mmp.c +++ b/fs/ext4/mmp.c @@ -162,7 +162,7 @@ static int kmmpd(void *data) memcpy(mmp->mmp_nodename, init_utsname()->nodename, sizeof(mmp->mmp_nodename)); - while (!kthread_should_stop() && !sb_rdonly(sb)) { + while (!kthread_should_stop() && !ext4_forced_shutdown(sb)) { if (!ext4_has_feature_mmp(sb)) { ext4_warning(sb, "kmmpd being stopped since MMP feature" " has been disabled."); From 304749c0d5e216479ea4d553ad04ba1390d5c707 Mon Sep 17 00:00:00 2001 From: Ojaswin Mujoo Date: Fri, 30 Jun 2023 14:29:27 +0530 Subject: [PATCH 030/186] ext4: replace CR_FAST macro with inline function for readability Replace CR_FAST with ext4_mb_cr_expensive() inline function for better readability. This function returns true if the criteria is one of the expensive/slower ones where lots of disk IO/prefetching is acceptable. No functional changes are intended in this patch. Signed-off-by: Ojaswin Mujoo Reviewed-by: Jan Kara Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230630085927.140137-1-ojaswin@linux.ibm.com Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 7 ++++--- fs/ext4/mballoc.c | 13 +++++++++---- 2 files changed, 13 insertions(+), 7 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 89d76d50af4ca..532b70f613e95 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -176,9 +176,6 @@ enum criteria { EXT4_MB_NUM_CRS }; -/* criteria below which we use fast block scanning and avoid unnecessary IO */ -#define CR_FAST CR_GOAL_LEN_SLOW - /* * Flags used in mballoc's allocation_context flags field. * @@ -2924,6 +2921,10 @@ extern int ext4_trim_fs(struct super_block *, struct fstrim_range *); extern void ext4_process_freed_data(struct super_block *sb, tid_t commit_tid); extern void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block, int len, int state); +static inline bool ext4_mb_cr_expensive(enum criteria cr) +{ + return cr >= CR_GOAL_LEN_SLOW; +} /* inode.c */ void ext4_inode_csum_set(struct inode *inode, struct ext4_inode *raw, diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 1dc63e329e64b..a2d791953da53 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2450,7 +2450,7 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, break; } - if (ac->ac_criteria < CR_FAST) { + if (!ext4_mb_cr_expensive(ac->ac_criteria)) { /* * In CR_GOAL_LEN_FAST and CR_BEST_AVAIL_LEN, we are * sure that this group will have a large enough @@ -2634,7 +2634,12 @@ static int ext4_mb_good_group_nolock(struct ext4_allocation_context *ac, free = grp->bb_free; if (free == 0) goto out; - if (cr <= CR_FAST && free < ac->ac_g_ex.fe_len) + /* + * In all criterias except CR_ANY_FREE we try to avoid groups that + * can't possibly satisfy the full goal request due to insufficient + * free blocks. + */ + if (cr < CR_ANY_FREE && free < ac->ac_g_ex.fe_len) goto out; if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) goto out; @@ -2658,7 +2663,7 @@ static int ext4_mb_good_group_nolock(struct ext4_allocation_context *ac, * sure we locate metadata blocks in the first block group in * the flex_bg if possible. */ - if (cr < CR_FAST && + if (!ext4_mb_cr_expensive(cr) && (!sbi->s_log_groups_per_flex || ((group & ((1 << sbi->s_log_groups_per_flex) - 1)) != 0)) && !(ext4_has_group_desc_csum(sb) && @@ -2852,7 +2857,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) * spend a lot of time loading imperfect groups */ if ((prefetch_grp == group) && - (cr >= CR_FAST || + (ext4_mb_cr_expensive(cr) || prefetch_ios < sbi->s_mb_prefetch_limit)) { nr = sbi->s_mb_prefetch; if (ext4_has_feature_flex_bg(sb)) { From a9ce5993a0f5c0887c8a1b4ffa3b8046fbcfdc93 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:31:55 +0800 Subject: [PATCH 031/186] ext4: correct grp validation in ext4_mb_good_group Group corruption check will access memory of grp and will trigger kernel crash if grp is NULL. So do NULL check before corruption check. Fixes: 5354b2af3406 ("ext4: allow ext4_get_group_info() to fail") Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-2-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index a2d791953da53..e07d2a4fbcd10 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2553,7 +2553,7 @@ static bool ext4_mb_good_group(struct ext4_allocation_context *ac, BUG_ON(cr < CR_POWER2_ALIGNED || cr >= EXT4_MB_NUM_CRS); - if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(grp) || !grp)) + if (unlikely(!grp || EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) return false; free = grp->bb_free; From 60c672b7f2d1e5dd1774f2399b355c9314e709f8 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:31:56 +0800 Subject: [PATCH 032/186] ext4: avoid potential data overflow in next_linear_group ngroups is ext4_group_t (unsigned int) while next_linear_group treat it in int. If ngroups is bigger than max number described by int, it will be treat as a negative number. Then "return group + 1 >= ngroups ? 0 : group + 1;" may keep returning 0. Switch int to ext4_group_t in next_linear_group to fix the overflow. Fixes: 196e402adf2e ("ext4: improve cr 0 / cr 1 group scanning") Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-3-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index e07d2a4fbcd10..bf041932c5997 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1080,8 +1080,9 @@ static inline int should_optimize_scan(struct ext4_allocation_context *ac) * Return next linear group for allocation. If linear traversal should not be * performed, this function just returns the same group */ -static int -next_linear_group(struct ext4_allocation_context *ac, int group, int ngroups) +static ext4_group_t +next_linear_group(struct ext4_allocation_context *ac, ext4_group_t group, + ext4_group_t ngroups) { if (!should_optimize_scan(ac)) goto inc_and_return; From 919eb90cec4049cecf4a9f996afb0f14e3864fca Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:31:57 +0800 Subject: [PATCH 033/186] ext4: return found group directly in ext4_mb_choose_next_group_p2_aligned Return good group when it's found in loop to remove unnecessary NULL initialization of grp and futher check if good group is found after loop. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-4-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 21 +++++++-------------- 1 file changed, 7 insertions(+), 14 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index bf041932c5997..880b9731edaa8 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -874,7 +874,7 @@ static void ext4_mb_choose_next_group_p2_aligned(struct ext4_allocation_context enum criteria *new_cr, ext4_group_t *group, ext4_group_t ngroups) { struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); - struct ext4_group_info *iter, *grp; + struct ext4_group_info *iter; int i; if (ac->ac_status == AC_STATUS_FOUND) @@ -883,7 +883,6 @@ static void ext4_mb_choose_next_group_p2_aligned(struct ext4_allocation_context if (unlikely(sbi->s_mb_stats && ac->ac_flags & EXT4_MB_CR_POWER2_ALIGNED_OPTIMIZED)) atomic_inc(&sbi->s_bal_p2_aligned_bad_suggestions); - grp = NULL; for (i = ac->ac_2order; i < MB_NUM_ORDERS(ac->ac_sb); i++) { if (list_empty(&sbi->s_mb_largest_free_orders[i])) continue; @@ -892,28 +891,22 @@ static void ext4_mb_choose_next_group_p2_aligned(struct ext4_allocation_context read_unlock(&sbi->s_mb_largest_free_orders_locks[i]); continue; } - grp = NULL; list_for_each_entry(iter, &sbi->s_mb_largest_free_orders[i], bb_largest_free_order_node) { if (sbi->s_mb_stats) atomic64_inc(&sbi->s_bal_cX_groups_considered[CR_POWER2_ALIGNED]); if (likely(ext4_mb_good_group(ac, iter->bb_group, CR_POWER2_ALIGNED))) { - grp = iter; - break; + *group = iter->bb_group; + ac->ac_flags |= EXT4_MB_CR_POWER2_ALIGNED_OPTIMIZED; + read_unlock(&sbi->s_mb_largest_free_orders_locks[i]); + return; } } read_unlock(&sbi->s_mb_largest_free_orders_locks[i]); - if (grp) - break; } - if (!grp) { - /* Increment cr and search again */ - *new_cr = CR_GOAL_LEN_FAST; - } else { - *group = grp->bb_group; - ac->ac_flags |= EXT4_MB_CR_POWER2_ALIGNED_OPTIMIZED; - } + /* Increment cr and search again if no group is found */ + *new_cr = CR_GOAL_LEN_FAST; } /* From bb60caa2db6697c20a0842b5b3c192aa1800da1a Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:31:58 +0800 Subject: [PATCH 034/186] ext4: use is_power_of_2 helper in ext4_mb_regular_allocator Use intuitive is_power_of_2 helper in ext4_mb_regular_allocator. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-5-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 880b9731edaa8..3dd2609ea1333 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2799,10 +2799,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) * requests upto maximum buddy size we have constructed. */ if (i >= sbi->s_mb_order2_reqs && i <= MB_NUM_ORDERS(sb)) { - /* - * This should tell if fe_len is exactly power of 2 - */ - if ((ac->ac_g_ex.fe_len & (~(1 << (i - 1)))) == 0) + if (is_power_of_2(ac->ac_g_ex.fe_len)) ac->ac_2order = array_index_nospec(i - 1, MB_NUM_ORDERS(sb)); } From ad635507b5b22d59457b6db6d8a0e4ddf7ad2b4c Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:31:59 +0800 Subject: [PATCH 035/186] ext4: remove unnecessary return for void function The return at end of void function is unnecessary, just remove it. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-6-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 3dd2609ea1333..154ae1fca10c1 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4976,7 +4976,6 @@ static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, mb_set_bits(bitmap, entry->efd_start_cluster, entry->efd_count); n = rb_next(n); } - return; } /* @@ -5727,12 +5726,10 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) #else static inline void ext4_mb_show_pa(struct super_block *sb) { - return; } static inline void ext4_mb_show_ac(struct ext4_allocation_context *ac) { ext4_mb_show_pa(ac->ac_sb); - return; } #endif @@ -5973,12 +5970,9 @@ static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) spin_unlock(&lg->lg_prealloc_lock); /* Now trim the list to be not more than 8 elements */ - if (lg_prealloc_count > 8) { + if (lg_prealloc_count > 8) ext4_mb_discard_lg_preallocations(sb, lg, order, lg_prealloc_count); - return; - } - return ; } /* @@ -6632,7 +6626,6 @@ static void ext4_mb_clear_bb(handle_t *handle, struct inode *inode, error_return: brelse(bitmap_bh); ext4_std_error(sb, err); - return; } /** @@ -6735,7 +6728,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, } ext4_mb_clear_bb(handle, inode, block, count, flags); - return; } /** From de8bf0e5ee7482585450357c6d4eddec8efc5cb7 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:32:00 +0800 Subject: [PATCH 036/186] ext4: replace the traditional ternary conditional operator with with max()/min() Replace the traditional ternary conditional operator with with max()/min() Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-7-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 154ae1fca10c1..f0deb5f2f81de 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -6917,8 +6917,7 @@ __releases(ext4_group_lock_ptr(sb, e4b->bd_group)) void *bitmap; bitmap = e4b->bd_bitmap; - start = (e4b->bd_info->bb_first_free > start) ? - e4b->bd_info->bb_first_free : start; + start = max(e4b->bd_info->bb_first_free, start); count = 0; free_count = 0; @@ -7135,8 +7134,7 @@ ext4_mballoc_query_range( ext4_lock_group(sb, group); - start = (e4b.bd_info->bb_first_free > start) ? - e4b.bd_info->bb_first_free : start; + start = max(e4b.bd_info->bb_first_free, start); if (end >= EXT4_CLUSTERS_PER_GROUP(sb)) end = EXT4_CLUSTERS_PER_GROUP(sb) - 1; From f6c72fef1272e65eff8d5ecef8c744686f6b7745 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:32:01 +0800 Subject: [PATCH 037/186] ext4: remove unused ext4_{set}/{clear}_bit_atomic Remove ext4_set_bit_atomic and ext4_clear_bit_atomic which are defined but not used. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-8-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 532b70f613e95..fb4d914ea8883 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1250,10 +1250,8 @@ struct ext4_inode_info { #define ext4_test_and_set_bit __test_and_set_bit_le #define ext4_set_bit __set_bit_le -#define ext4_set_bit_atomic ext2_set_bit_atomic #define ext4_test_and_clear_bit __test_and_clear_bit_le #define ext4_clear_bit __clear_bit_le -#define ext4_clear_bit_atomic ext2_clear_bit_atomic #define ext4_test_bit test_bit_le #define ext4_find_next_zero_bit find_next_zero_bit_le #define ext4_find_next_bit find_next_bit_le From b50675a4a6a69110c0c2baadebd2075d3b31b25c Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:32:02 +0800 Subject: [PATCH 038/186] ext4: return found group directly in ext4_mb_choose_next_group_goal_fast Return good group when it's found in loop to remove futher check if good group is found after loop. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-9-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index f0deb5f2f81de..3b1f90dfb1192 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -959,16 +959,14 @@ static void ext4_mb_choose_next_group_goal_fast(struct ext4_allocation_context * for (i = mb_avg_fragment_size_order(ac->ac_sb, ac->ac_g_ex.fe_len); i < MB_NUM_ORDERS(ac->ac_sb); i++) { grp = ext4_mb_find_good_group_avg_frag_lists(ac, i); - if (grp) - break; + if (grp) { + *group = grp->bb_group; + ac->ac_flags |= EXT4_MB_CR_GOAL_LEN_FAST_OPTIMIZED; + return; + } } - if (grp) { - *group = grp->bb_group; - ac->ac_flags |= EXT4_MB_CR_GOAL_LEN_FAST_OPTIMIZED; - } else { - *new_cr = CR_BEST_AVAIL_LEN; - } + *new_cr = CR_BEST_AVAIL_LEN; } /* From bcb123ac9b9887478da4185b55dfbf1a72550848 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:32:03 +0800 Subject: [PATCH 039/186] ext4: return found group directly in ext4_mb_choose_next_group_best_avail Return good group when it's found in loop to remove futher check if good group is found after loop. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-10-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 3b1f90dfb1192..f9189f7566fba 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1042,18 +1042,16 @@ static void ext4_mb_choose_next_group_best_avail(struct ext4_allocation_context ac->ac_g_ex.fe_len); grp = ext4_mb_find_good_group_avg_frag_lists(ac, frag_order); - if (grp) - break; + if (grp) { + *group = grp->bb_group; + ac->ac_flags |= EXT4_MB_CR_BEST_AVAIL_LEN_OPTIMIZED; + return; + } } - if (grp) { - *group = grp->bb_group; - ac->ac_flags |= EXT4_MB_CR_BEST_AVAIL_LEN_OPTIMIZED; - } else { - /* Reset goal length to original goal length before falling into CR_GOAL_LEN_SLOW */ - ac->ac_g_ex.fe_len = ac->ac_orig_goal_len; - *new_cr = CR_GOAL_LEN_SLOW; - } + /* Reset goal length to original goal length before falling into CR_GOAL_LEN_SLOW */ + ac->ac_g_ex.fe_len = ac->ac_orig_goal_len; + *new_cr = CR_GOAL_LEN_SLOW; } static inline int should_optimize_scan(struct ext4_allocation_context *ac) From 4eea9fbed950f240bf6e627e1c784b8d54c54988 Mon Sep 17 00:00:00 2001 From: Kemeng Shi Date: Tue, 1 Aug 2023 22:32:04 +0800 Subject: [PATCH 040/186] ext4: correct some stale comment of criteria We named criteria with CR_XXX, correct stale comment to criteria with raw number. Signed-off-by: Kemeng Shi Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/20230801143204.2284343-11-shikemeng@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index f9189f7566fba..b89b5f0816e71 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2782,8 +2782,8 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) /* * ac->ac_2order is set only if the fe_len is a power of 2 - * if ac->ac_2order is set we also set criteria to 0 so that we - * try exact allocation using buddy. + * if ac->ac_2order is set we also set criteria to CR_POWER2_ALIGNED + * so that we try exact allocation using buddy. */ i = fls(ac->ac_g_ex.fe_len); ac->ac_2order = 0; @@ -2840,8 +2840,8 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) /* * Batch reads of the block allocation bitmaps * to get multiple READs in flight; limit - * prefetching at cr=0/1, otherwise mballoc can - * spend a lot of time loading imperfect groups + * prefetching at inexpensive CR, otherwise mballoc + * can spend a lot of time loading imperfect groups */ if ((prefetch_grp == group) && (ext4_mb_cr_expensive(cr) || From 373ac521799d9e97061515aca6ec6621789036bb Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 14 Jul 2023 10:55:26 +0800 Subject: [PATCH 041/186] jbd2: fix checkpoint cleanup performance regression journal_clean_one_cp_list() has been merged into journal_shrink_one_cp_list(), but do chekpoint buffer cleanup from the committing process is just a best effort, it should stop scan once it meet a busy buffer, or else it will cause a lot of invalid buffer scan and checks. We catch a performance regression when doing fs_mark tests below. Test cmd: ./fs_mark -d scratch -s 1024 -n 10000 -t 1 -D 100 -N 100 Before merging checkpoint buffer cleanup: FSUse% Count Size Files/sec App Overhead 95 10000 1024 8304.9 49033 After merging checkpoint buffer cleanup: FSUse% Count Size Files/sec App Overhead 95 10000 1024 7649.0 50012 FSUse% Count Size Files/sec App Overhead 95 10000 1024 2107.1 50871 After merging checkpoint buffer cleanup, the total loop count in journal_shrink_one_cp_list() could be up to 6,261,600+ (50,000+ ~ 100,000+ in general), most of them are invalid. This patch fix it through passing 'shrink_type' into journal_shrink_one_cp_list() and add a new 'SHRINK_BUSY_STOP' to indicate it should stop once meet a busy buffer. After fix, the loop count descending back to 10,000+. After this fix: FSUse% Count Size Files/sec App Overhead 95 10000 1024 8558.4 49109 Cc: stable@kernel.org Fixes: b98dba273a0e ("jbd2: remove journal_clean_one_cp_list()") Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230714025528.564988-2-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/checkpoint.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 9ec91017a7f3c..936c6d758a651 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -349,6 +349,8 @@ int jbd2_cleanup_journal_tail(journal_t *journal) /* Checkpoint list management */ +enum shrink_type {SHRINK_DESTROY, SHRINK_BUSY_STOP, SHRINK_BUSY_SKIP}; + /* * journal_shrink_one_cp_list * @@ -360,7 +362,8 @@ int jbd2_cleanup_journal_tail(journal_t *journal) * Called with j_list_lock held. */ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh, - bool destroy, bool *released) + enum shrink_type type, + bool *released) { struct journal_head *last_jh; struct journal_head *next_jh = jh; @@ -376,12 +379,15 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh, jh = next_jh; next_jh = jh->b_cpnext; - if (destroy) { + if (type == SHRINK_DESTROY) { ret = __jbd2_journal_remove_checkpoint(jh); } else { ret = jbd2_journal_try_remove_checkpoint(jh); - if (ret < 0) - continue; + if (ret < 0) { + if (type == SHRINK_BUSY_SKIP) + continue; + break; + } } nr_freed++; @@ -445,7 +451,7 @@ unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal, tid = transaction->t_tid; freed = journal_shrink_one_cp_list(transaction->t_checkpoint_list, - false, &released); + SHRINK_BUSY_SKIP, &released); nr_freed += freed; (*nr_to_scan) -= min(*nr_to_scan, freed); if (*nr_to_scan == 0) @@ -485,19 +491,21 @@ unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal, void __jbd2_journal_clean_checkpoint_list(journal_t *journal, bool destroy) { transaction_t *transaction, *last_transaction, *next_transaction; + enum shrink_type type; bool released; transaction = journal->j_checkpoint_transactions; if (!transaction) return; + type = destroy ? SHRINK_DESTROY : SHRINK_BUSY_STOP; last_transaction = transaction->t_cpprev; next_transaction = transaction; do { transaction = next_transaction; next_transaction = transaction->t_cpnext; journal_shrink_one_cp_list(transaction->t_checkpoint_list, - destroy, &released); + type, &released); /* * This function only frees up some memory if possible so we * dont have an obligation to finish processing. Bail out if From 590a809ff743e7bd890ba5fb36bc38e20a36de53 Mon Sep 17 00:00:00 2001 From: Zhihao Cheng Date: Fri, 14 Jul 2023 10:55:27 +0800 Subject: [PATCH 042/186] jbd2: check 'jh->b_transaction' before removing it from checkpoint MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Following process will corrupt ext4 image: Step 1: jbd2_journal_commit_transaction __jbd2_journal_insert_checkpoint(jh, commit_transaction) // Put jh into trans1->t_checkpoint_list journal->j_checkpoint_transactions = commit_transaction // Put trans1 into journal->j_checkpoint_transactions Step 2: do_get_write_access test_clear_buffer_dirty(bh) // clear buffer dirty，set jbd dirty __jbd2_journal_file_buffer(jh, transaction) // jh belongs to trans2 Step 3: drop_cache journal_shrink_one_cp_list jbd2_journal_try_remove_checkpoint if (!trylock_buffer(bh)) // lock bh, true if (buffer_dirty(bh)) // buffer is not dirty __jbd2_journal_remove_checkpoint(jh) // remove jh from trans1->t_checkpoint_list Step 4: jbd2_log_do_checkpoint trans1 = journal->j_checkpoint_transactions // jh is not in trans1->t_checkpoint_list jbd2_cleanup_journal_tail(journal) // trans1 is done Step 5: Power cut, trans2 is not committed, jh is lost in next mounting. Fix it by checking 'jh->b_transaction' before remove it from checkpoint. Cc: stable@kernel.org Fixes: 46f881b5b175 ("jbd2: fix a race when checking checkpoint buffer busy") Signed-off-by: Zhihao Cheng Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230714025528.564988-3-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/checkpoint.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 936c6d758a651..f033ac807013c 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -639,6 +639,8 @@ int jbd2_journal_try_remove_checkpoint(struct journal_head *jh) { struct buffer_head *bh = jh2bh(jh); + if (jh->b_transaction) + return -EBUSY; if (!trylock_buffer(bh)) return -EBUSY; if (buffer_dirty(bh)) { From 5f02a30eac5cc1c081cbdb42d19fd0ded00b0618 Mon Sep 17 00:00:00 2001 From: Yang Li Date: Fri, 14 Jul 2023 10:55:28 +0800 Subject: [PATCH 043/186] jbd2: remove unused function '__cp_buffer_busy' The code calling function '__cp_buffer_busy' has been removed, so the function should also be removed. silence the warning: fs/jbd2/checkpoint.c:48:20: warning: unused function '__cp_buffer_busy' Reported-by: Abaci Robot Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=5518 Signed-off-by: Yang Li Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230714025528.564988-4-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/checkpoint.c | 12 ------------ 1 file changed, 12 deletions(-) diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index f033ac807013c..118699fff2f90 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -40,18 +40,6 @@ static inline void __buffer_unlink(struct journal_head *jh) } } -/* - * Check a checkpoint buffer could be release or not. - * - * Requires j_list_lock - */ -static inline bool __cp_buffer_busy(struct journal_head *jh) -{ - struct buffer_head *bh = jh2bh(jh); - - return (jh->b_transaction || buffer_locked(bh) || buffer_dirty(bh)); -} - /* * __jbd2_log_wait_for_space: wait until there is space in the journal. * From 7ca4b085f430f3774c3838b3da569ceccd6a0177 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lu=C3=ADs=20Henriques?= Date: Thu, 3 Aug 2023 10:17:13 +0100 Subject: [PATCH 044/186] ext4: fix memory leaks in ext4_fname_{setup_filename,prepare_lookup} MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If the filename casefolding fails, we'll be leaking memory from the fscrypt_name struct, namely from the 'crypto_buf.name' member. Make sure we free it in the error path on both ext4_fname_setup_filename() and ext4_fname_prepare_lookup() functions. Cc: stable@kernel.org Fixes: 1ae98e295fa2 ("ext4: optimize match for casefolded encrypted dirs") Signed-off-by: Luís Henriques Reviewed-by: Eric Biggers Link: https://lore.kernel.org/r/20230803091713.13239-1-lhenriques@suse.de Signed-off-by: Theodore Ts'o --- fs/ext4/crypto.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/ext4/crypto.c b/fs/ext4/crypto.c index e20ac0654b3f2..453d4da5de520 100644 --- a/fs/ext4/crypto.c +++ b/fs/ext4/crypto.c @@ -33,6 +33,8 @@ int ext4_fname_setup_filename(struct inode *dir, const struct qstr *iname, #if IS_ENABLED(CONFIG_UNICODE) err = ext4_fname_setup_ci_filename(dir, iname, fname); + if (err) + ext4_fname_free_filename(fname); #endif return err; } @@ -51,6 +53,8 @@ int ext4_fname_prepare_lookup(struct inode *dir, struct dentry *dentry, #if IS_ENABLED(CONFIG_UNICODE) err = ext4_fname_setup_ci_filename(dir, &dentry->d_name, fname); + if (err) + ext4_fname_free_filename(fname); #endif return err; } From e15e117bbbe18258a5ad506bbf6c58ff129c9576 Mon Sep 17 00:00:00 2001 From: Wang Jianjian Date: Wed, 2 Aug 2023 22:45:34 +0800 Subject: [PATCH 045/186] jbd2: remove unused t_handle_lock Since commit f7f497cb7024 ("jbd2: kill t_handle_lock transaction spinlock"), this lock has been no use. Fixes: f7f497cb7024 ("jbd2: kill t_handle_lock transaction spinlock") Signed-off-by: Wang Jianjian Reviewed-by: Ritesh Harjani (IBM) Link: https://lore.kernel.org/r/tencent_8477CBE568348A1862C64E393D587B342008@qq.com Signed-off-by: Theodore Ts'o --- include/linux/jbd2.h | 5 ----- 1 file changed, 5 deletions(-) diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index 44c298aa58d44..52772c826c868 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -630,11 +630,6 @@ struct transaction_s */ struct list_head t_inode_list; - /* - * Protects info related to handles - */ - spinlock_t t_handle_lock; - /* * Longest time some handle had to wait for running transaction */ From 772c9f691dcf3a487f29ddb90a5a15c78d7328e1 Mon Sep 17 00:00:00 2001 From: Ritesh Harjani Date: Sun, 16 Jul 2023 19:33:34 +0530 Subject: [PATCH 046/186] ext4: don't use CR_BEST_AVAIL_LEN for non-regular files Using CR_BEST_AVAIL_LEN only make sense for regular files, as for non-regular files we never normalize the allocation request length i.e. goal len is same as original length (ac_g_ex.fe_len == ac_o_ex.fe_len). Hence there is no scope of trimming the goal length to make it satisfy original request len. Thus this patch avoids using CR_BEST_AVAIL_LEN criteria for non-regular files request. Cc: stable@kernel.org Fixes: 33122aa930f1 ("ext4: Add allocation criteria 1.5 (CR1_5)") Reported-by: Eric Whitney Signed-off-by: Ritesh Harjani (IBM) Tested-by: Eric Whitney Link: https://lore.kernel.org/r/2a694c748ff8b8c4b416995a24f06f07b55047a8.1689516047.git.ritesh.list@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index b89b5f0816e71..3d5b0b71d7f5b 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -966,7 +966,18 @@ static void ext4_mb_choose_next_group_goal_fast(struct ext4_allocation_context * } } - *new_cr = CR_BEST_AVAIL_LEN; + /* + * CR_BEST_AVAIL_LEN works based on the concept that we have + * a larger normalized goal len request which can be trimmed to + * a smaller goal len such that it can still satisfy original + * request len. However, allocation request for non-regular + * files never gets normalized. + * See function ext4_mb_normalize_request() (EXT4_MB_HINT_DATA). + */ + if (ac->ac_flags & EXT4_MB_HINT_DATA) + *new_cr = CR_BEST_AVAIL_LEN; + else + *new_cr = CR_GOAL_LEN_SLOW; } /* From e717f2e8e4896f4c59a865b11d5cb957b0bfb0e1 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:38 -0400 Subject: [PATCH 047/186] fs: dlm: add missing spin_unlock This patch fixes commit dc52cd2eff4a ("fs: dlm: fix F_CANCELLK to cancel pending request") that we don't unlock the ops_lock in a rate case when a waiter cannot be found. This case can only happen when cancellation of plock operation was successful but no kernel waiter was being found. Fixes: dc52cd2eff4a ("fs: dlm: fix F_CANCELLK to cancel pending request") Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 943d9f8e55645..44b3aab5b709b 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -398,6 +398,7 @@ int dlm_posix_cancel(dlm_lockspace_t *lockspace, u64 number, struct file *file, */ op = plock_lookup_waiter(&info); if (WARN_ON_ONCE(!op)) { + spin_unlock(&ops_lock); rv = -ENOLCK; break; } From 4b056db81c5dd79d786b44c371f6e0b4371735c3 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:39 -0400 Subject: [PATCH 048/186] fs: dlm: remove unused processed_nodes The variable processed_nodes is not being used by commit 1696c75f1864 ("fs: dlm: add send ack threshold and append acks to msgs"). This patch removes the leftover of this commit. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/lowcomms.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index 9f14ea9f63224..f7bc22e74db27 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -863,7 +863,6 @@ struct dlm_processed_nodes { static void process_dlm_messages(struct work_struct *work) { struct processqueue_entry *pentry; - LIST_HEAD(processed_nodes); spin_lock(&processqueue_lock); pentry = list_first_entry_or_null(&processqueue, From 541adb0d4d10b4daf15f4b6b73c5d6b855d23eb5 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:40 -0400 Subject: [PATCH 049/186] fs: dlm: debugfs for queued callbacks It was useful to debug an issue with the callback queue to check if any callbacks in any lkb are for some reason not processed by the callback workqueue. The mentioned issue was fixed by commit a034c1370ded ("fs: dlm: fix DLM_IFL_CB_PENDING gets overwritten"). If there are similar issue that looks like a ast callback was not processed, we can confirm now that it is not sitting to be processed by the callback workqueue anymore. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/debug_fs.c | 101 +++++++++++++++++++++++++++++++++++++++++- fs/dlm/dlm_internal.h | 1 + 2 files changed, 101 insertions(+), 1 deletion(-) diff --git a/fs/dlm/debug_fs.c b/fs/dlm/debug_fs.c index a1aca41c49d06..5aabcb6f0f157 100644 --- a/fs/dlm/debug_fs.c +++ b/fs/dlm/debug_fs.c @@ -18,6 +18,7 @@ #include "dlm_internal.h" #include "midcomms.h" #include "lock.h" +#include "ast.h" #define DLM_DEBUG_BUF_LEN 4096 static char debug_buf[DLM_DEBUG_BUF_LEN]; @@ -365,6 +366,52 @@ static void print_format4(struct dlm_rsb *r, struct seq_file *s) unlock_rsb(r); } +static void print_format5_lock(struct seq_file *s, struct dlm_lkb *lkb) +{ + struct dlm_callback *cb; + + /* lkb_id lkb_flags mode flags sb_status sb_flags */ + + spin_lock(&lkb->lkb_cb_lock); + list_for_each_entry(cb, &lkb->lkb_callbacks, list) { + seq_printf(s, "%x %x %d %x %d %x\n", + lkb->lkb_id, + dlm_iflags_val(lkb), + cb->mode, + cb->flags, + cb->sb_status, + cb->sb_flags); + } + spin_unlock(&lkb->lkb_cb_lock); +} + +static void print_format5(struct dlm_rsb *r, struct seq_file *s) +{ + struct dlm_lkb *lkb; + + lock_rsb(r); + + list_for_each_entry(lkb, &r->res_grantqueue, lkb_statequeue) { + print_format5_lock(s, lkb); + if (seq_has_overflowed(s)) + goto out; + } + + list_for_each_entry(lkb, &r->res_convertqueue, lkb_statequeue) { + print_format5_lock(s, lkb); + if (seq_has_overflowed(s)) + goto out; + } + + list_for_each_entry(lkb, &r->res_waitqueue, lkb_statequeue) { + print_format5_lock(s, lkb); + if (seq_has_overflowed(s)) + goto out; + } + out: + unlock_rsb(r); +} + struct rsbtbl_iter { struct dlm_rsb *rsb; unsigned bucket; @@ -408,6 +455,13 @@ static int table_seq_show(struct seq_file *seq, void *iter_ptr) } print_format4(ri->rsb, seq); break; + case 5: + if (ri->header) { + seq_puts(seq, "lkb_id lkb_flags mode flags sb_status sb_flags\n"); + ri->header = 0; + } + print_format5(ri->rsb, seq); + break; } return 0; @@ -417,6 +471,7 @@ static const struct seq_operations format1_seq_ops; static const struct seq_operations format2_seq_ops; static const struct seq_operations format3_seq_ops; static const struct seq_operations format4_seq_ops; +static const struct seq_operations format5_seq_ops; static void *table_seq_start(struct seq_file *seq, loff_t *pos) { @@ -448,6 +503,8 @@ static void *table_seq_start(struct seq_file *seq, loff_t *pos) ri->format = 3; if (seq->op == &format4_seq_ops) ri->format = 4; + if (seq->op == &format5_seq_ops) + ri->format = 5; tree = toss ? &ls->ls_rsbtbl[bucket].toss : &ls->ls_rsbtbl[bucket].keep; @@ -602,10 +659,18 @@ static const struct seq_operations format4_seq_ops = { .show = table_seq_show, }; +static const struct seq_operations format5_seq_ops = { + .start = table_seq_start, + .next = table_seq_next, + .stop = table_seq_stop, + .show = table_seq_show, +}; + static const struct file_operations format1_fops; static const struct file_operations format2_fops; static const struct file_operations format3_fops; static const struct file_operations format4_fops; +static const struct file_operations format5_fops; static int table_open1(struct inode *inode, struct file *file) { @@ -683,7 +748,21 @@ static int table_open4(struct inode *inode, struct file *file) struct seq_file *seq; int ret; - ret = seq_open(file, &format4_seq_ops); + ret = seq_open(file, &format5_seq_ops); + if (ret) + return ret; + + seq = file->private_data; + seq->private = inode->i_private; /* the dlm_ls */ + return 0; +} + +static int table_open5(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + int ret; + + ret = seq_open(file, &format5_seq_ops); if (ret) return ret; @@ -725,6 +804,14 @@ static const struct file_operations format4_fops = { .release = seq_release }; +static const struct file_operations format5_fops = { + .owner = THIS_MODULE, + .open = table_open5, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release +}; + /* * dump lkb's on the ls_waiters list */ @@ -793,6 +880,7 @@ void dlm_delete_debug_file(struct dlm_ls *ls) debugfs_remove(ls->ls_debug_locks_dentry); debugfs_remove(ls->ls_debug_all_dentry); debugfs_remove(ls->ls_debug_toss_dentry); + debugfs_remove(ls->ls_debug_queued_asts_dentry); } static int dlm_state_show(struct seq_file *file, void *offset) @@ -936,6 +1024,17 @@ void dlm_create_debug_file(struct dlm_ls *ls) dlm_root, ls, &waiters_fops); + + /* format 5 */ + + memset(name, 0, sizeof(name)); + snprintf(name, DLM_LOCKSPACE_LEN + 8, "%s_queued_asts", ls->ls_name); + + ls->ls_debug_queued_asts_dentry = debugfs_create_file(name, + 0644, + dlm_root, + ls, + &format5_fops); } void __init dlm_register_debugfs(void) diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index c8156770205e6..dfc444dad3298 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -598,6 +598,7 @@ struct dlm_ls { struct dentry *ls_debug_locks_dentry; /* debugfs */ struct dentry *ls_debug_all_dentry; /* debugfs */ struct dentry *ls_debug_toss_dentry; /* debugfs */ + struct dentry *ls_debug_queued_asts_dentry; /* debugfs */ wait_queue_head_t ls_uevent_wait; /* user part of join/leave */ int ls_uevent_result; From 67b5da9a40fc984d25bda90a918e490e8c2555b7 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:41 -0400 Subject: [PATCH 050/186] fs: dlm: check on plock ops when exit dlm To be sure we don't have any issues that there are leftover plock ops in either send_list or recv_list we simple check if either one of the list are empty when we exit the dlm subsystem. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 44b3aab5b709b..5c2cc8d940efc 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -628,5 +628,7 @@ int dlm_plock_init(void) void dlm_plock_exit(void) { misc_deregister(&plock_dev_misc); + WARN_ON(!list_empty(&send_list)); + WARN_ON(!list_empty(&recv_list)); } From 8c95006d55726eeebf3b863335accfba50d4bc8f Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:42 -0400 Subject: [PATCH 051/186] fs: dlm: add plock dev tracepoints I currently debug nfs plock handling and introduce those two tracepoints for getting more information about what is happening there if the user space reads plock operations from kernel and writing the result back. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 6 +++++ include/trace/events/dlm.h | 51 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 57 insertions(+) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 5c2cc8d940efc..00e1d802a81cb 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -11,6 +11,8 @@ #include #include +#include + #include "dlm_internal.h" #include "lockspace.h" @@ -509,6 +511,8 @@ static ssize_t dev_read(struct file *file, char __user *u, size_t count, if (!op) return -EAGAIN; + trace_dlm_plock_read(&info); + /* there is no need to get a reply from userspace for unlocks that were generated by the vfs cleaning up for a close (the process did not make an unlock call). */ @@ -536,6 +540,8 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, if (copy_from_user(&info, u, sizeof(info))) return -EFAULT; + trace_dlm_plock_write(&info); + if (check_version(&info)) return -EINVAL; diff --git a/include/trace/events/dlm.h b/include/trace/events/dlm.h index 2b09574e12430..c1a146f9fc911 100644 --- a/include/trace/events/dlm.h +++ b/include/trace/events/dlm.h @@ -7,6 +7,7 @@ #include #include +#include #include #include "../../../fs/dlm/dlm_internal.h" @@ -585,6 +586,56 @@ TRACE_EVENT(dlm_recv_message, ); +DECLARE_EVENT_CLASS(dlm_plock_template, + + TP_PROTO(const struct dlm_plock_info *info), + + TP_ARGS(info), + + TP_STRUCT__entry( + __field(uint8_t, optype) + __field(uint8_t, ex) + __field(uint8_t, wait) + __field(uint8_t, flags) + __field(uint32_t, pid) + __field(int32_t, nodeid) + __field(int32_t, rv) + __field(uint32_t, fsid) + __field(uint64_t, number) + __field(uint64_t, start) + __field(uint64_t, end) + __field(uint64_t, owner) + ), + + TP_fast_assign( + __entry->optype = info->optype; + __entry->ex = info->ex; + __entry->wait = info->wait; + __entry->flags = info->flags; + __entry->pid = info->pid; + __entry->nodeid = info->nodeid; + __entry->rv = info->rv; + __entry->fsid = info->fsid; + __entry->number = info->number; + __entry->start = info->start; + __entry->end = info->end; + __entry->owner = info->owner; + ), + + TP_printk("fsid=%u number=%llx owner=%llx optype=%d ex=%d wait=%d flags=%x pid=%u nodeid=%d rv=%d start=%llx end=%llx", + __entry->fsid, __entry->number, __entry->owner, + __entry->optype, __entry->ex, __entry->wait, + __entry->flags, __entry->pid, __entry->nodeid, + __entry->rv, __entry->start, __entry->end) + +); + +DEFINE_EVENT(dlm_plock_template, dlm_plock_read, + TP_PROTO(const struct dlm_plock_info *info), TP_ARGS(info)); + +DEFINE_EVENT(dlm_plock_template, dlm_plock_write, + TP_PROTO(const struct dlm_plock_info *info), TP_ARGS(info)); + TRACE_EVENT(dlm_send, TP_PROTO(int nodeid, int ret), From c84c47333abbbfd83212fcfe2867be4a47e82056 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:43 -0400 Subject: [PATCH 052/186] fs: dlm: remove clear_members_cb This patch is just a small cleanup to directly call remove_remote_member() instead of going over clear_members_cb() which just calls remove_remote_member(). Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/member.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/fs/dlm/member.c b/fs/dlm/member.c index 77d202e4a02a4..f303ea8bd256e 100644 --- a/fs/dlm/member.c +++ b/fs/dlm/member.c @@ -393,14 +393,9 @@ static void remove_remote_member(int nodeid) dlm_midcomms_remove_member(nodeid); } -static void clear_members_cb(int nodeid) -{ - remove_remote_member(nodeid); -} - void dlm_clear_members(struct dlm_ls *ls) { - clear_memb_list(&ls->ls_nodes, clear_members_cb); + clear_memb_list(&ls->ls_nodes, remove_remote_member); ls->ls_num_nodes = 0; } From 643f5cfa610f475c7465e4158b2b1fdd170fac10 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:44 -0400 Subject: [PATCH 053/186] fs: dlm: cleanup lock order This patch cleanups the lock order to hold at first the close_lock and then held the nodes_srcu read lock. Probably it will never be a problem as nodes_srcu is only a read lock preventing the node pointer getting freed. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/midcomms.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/dlm/midcomms.c b/fs/dlm/midcomms.c index e1a0df67b5669..8ebffbfdc00ae 100644 --- a/fs/dlm/midcomms.c +++ b/fs/dlm/midcomms.c @@ -1489,12 +1489,12 @@ int dlm_midcomms_close(int nodeid) synchronize_srcu(&nodes_srcu); - idx = srcu_read_lock(&nodes_srcu); mutex_lock(&close_lock); + idx = srcu_read_lock(&nodes_srcu); node = nodeid2node(nodeid, 0); if (!node) { - mutex_unlock(&close_lock); srcu_read_unlock(&nodes_srcu, idx); + mutex_unlock(&close_lock); return dlm_lowcomms_close(nodeid); } From c4f4e135c27b503d325d414819831909023b113d Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:45 -0400 Subject: [PATCH 054/186] fs: dlm: get recovery sequence number as parameter This patch removes a read of the ls->ls_recover_seq uint64_t number in _create_rcom(). If the ls->ls_recover_seq is readed the ls_recover_lock need to held. However this number was always readed before when any rcom message is received and it's not necessary to read it again from a per lockspace variable to use it for the replying message. This patch will pass the sequence number as parameter so another read of ls->ls_recover_seq and holding the ls->ls_recover_lock is not required. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/dir.c | 4 +-- fs/dlm/dir.h | 2 +- fs/dlm/lock.c | 5 ++-- fs/dlm/lock.h | 3 ++- fs/dlm/member.c | 6 ++--- fs/dlm/rcom.c | 68 ++++++++++++++++++++++++++--------------------- fs/dlm/rcom.h | 10 ++++--- fs/dlm/recover.c | 58 +++++++++++++++++++++------------------- fs/dlm/recover.h | 12 ++++----- fs/dlm/recoverd.c | 16 +++++------ 10 files changed, 99 insertions(+), 85 deletions(-) diff --git a/fs/dlm/dir.c b/fs/dlm/dir.c index fb1981654bb24..3bf5bf7a37b4e 100644 --- a/fs/dlm/dir.c +++ b/fs/dlm/dir.c @@ -58,7 +58,7 @@ void dlm_recover_dir_nodeid(struct dlm_ls *ls) up_read(&ls->ls_root_sem); } -int dlm_recover_directory(struct dlm_ls *ls) +int dlm_recover_directory(struct dlm_ls *ls, uint64_t seq) { struct dlm_member *memb; char *b, *last_name = NULL; @@ -90,7 +90,7 @@ int dlm_recover_directory(struct dlm_ls *ls) } error = dlm_rcom_names(ls, memb->nodeid, - last_name, last_len); + last_name, last_len, seq); if (error) goto out_free; diff --git a/fs/dlm/dir.h b/fs/dlm/dir.h index 03844d086be23..0635582da003e 100644 --- a/fs/dlm/dir.h +++ b/fs/dlm/dir.h @@ -15,7 +15,7 @@ int dlm_dir_nodeid(struct dlm_rsb *rsb); int dlm_hash2nodeid(struct dlm_ls *ls, uint32_t hash); void dlm_recover_dir_nodeid(struct dlm_ls *ls); -int dlm_recover_directory(struct dlm_ls *ls); +int dlm_recover_directory(struct dlm_ls *ls, uint64_t seq); void dlm_copy_master_names(struct dlm_ls *ls, char *inbuf, int inlen, char *outbuf, int outlen, int nodeid); diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index f511a9d7d416e..b489da38e685f 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -5464,7 +5464,8 @@ int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc) } /* needs at least dlm_rcom + rcom_lock */ -int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc) +int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc, + uint64_t seq) { struct rcom_lock *rl = (struct rcom_lock *) rc->rc_buf; struct dlm_rsb *r; @@ -5509,7 +5510,7 @@ int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc) lkid, le32_to_cpu(rc->rc_header.h_nodeid), remid, result); - dlm_send_rcom_lock(r, lkb); + dlm_send_rcom_lock(r, lkb, seq); goto out; case -EEXIST: case 0: diff --git a/fs/dlm/lock.h b/fs/dlm/lock.h index aa5ad44d902bf..222e682523b90 100644 --- a/fs/dlm/lock.h +++ b/fs/dlm/lock.h @@ -37,7 +37,8 @@ void dlm_recover_grant(struct dlm_ls *ls); int dlm_recover_waiters_post(struct dlm_ls *ls); void dlm_recover_waiters_pre(struct dlm_ls *ls); int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc); -int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc); +int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc, + uint64_t seq); int dlm_user_request(struct dlm_ls *ls, struct dlm_user_args *ua, int mode, uint32_t flags, void *name, unsigned int namelen); diff --git a/fs/dlm/member.c b/fs/dlm/member.c index f303ea8bd256e..19f3cd96f3c0d 100644 --- a/fs/dlm/member.c +++ b/fs/dlm/member.c @@ -449,7 +449,7 @@ static void make_member_array(struct dlm_ls *ls) /* send a status request to all members just to establish comms connections */ -static int ping_members(struct dlm_ls *ls) +static int ping_members(struct dlm_ls *ls, uint64_t seq) { struct dlm_member *memb; int error = 0; @@ -459,7 +459,7 @@ static int ping_members(struct dlm_ls *ls) error = -EINTR; break; } - error = dlm_rcom_status(ls, memb->nodeid, 0); + error = dlm_rcom_status(ls, memb->nodeid, 0, seq); if (error) break; } @@ -607,7 +607,7 @@ int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out) make_member_array(ls); *neg_out = neg; - error = ping_members(ls); + error = ping_members(ls, rv->seq); log_rinfo(ls, "dlm_recover_members %d nodes", ls->ls_num_nodes); return error; } diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c index f4afdf892f785..efe45e68287f4 100644 --- a/fs/dlm/rcom.c +++ b/fs/dlm/rcom.c @@ -28,7 +28,8 @@ static int rcom_response(struct dlm_ls *ls) } static void _create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len, - struct dlm_rcom **rc_ret, char *mb, int mb_len) + struct dlm_rcom **rc_ret, char *mb, int mb_len, + uint64_t seq) { struct dlm_rcom *rc; @@ -41,16 +42,14 @@ static void _create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len, rc->rc_header.h_cmd = DLM_RCOM; rc->rc_type = cpu_to_le32(type); - - spin_lock(&ls->ls_recover_lock); - rc->rc_seq = cpu_to_le64(ls->ls_recover_seq); - spin_unlock(&ls->ls_recover_lock); + rc->rc_seq = cpu_to_le64(seq); *rc_ret = rc; } static int create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len, - struct dlm_rcom **rc_ret, struct dlm_mhandle **mh_ret) + struct dlm_rcom **rc_ret, struct dlm_mhandle **mh_ret, + uint64_t seq) { int mb_len = sizeof(struct dlm_rcom) + len; struct dlm_mhandle *mh; @@ -63,14 +62,14 @@ static int create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len, return -ENOBUFS; } - _create_rcom(ls, to_nodeid, type, len, rc_ret, mb, mb_len); + _create_rcom(ls, to_nodeid, type, len, rc_ret, mb, mb_len, seq); *mh_ret = mh; return 0; } static int create_rcom_stateless(struct dlm_ls *ls, int to_nodeid, int type, int len, struct dlm_rcom **rc_ret, - struct dlm_msg **msg_ret) + struct dlm_msg **msg_ret, uint64_t seq) { int mb_len = sizeof(struct dlm_rcom) + len; struct dlm_msg *msg; @@ -84,7 +83,7 @@ static int create_rcom_stateless(struct dlm_ls *ls, int to_nodeid, int type, return -ENOBUFS; } - _create_rcom(ls, to_nodeid, type, len, rc_ret, mb, mb_len); + _create_rcom(ls, to_nodeid, type, len, rc_ret, mb, mb_len, seq); *msg_ret = msg; return 0; } @@ -170,7 +169,8 @@ static void disallow_sync_reply(struct dlm_ls *ls) * node's rcom_config. */ -int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags) +int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags, + uint64_t seq) { struct dlm_rcom *rc; struct dlm_msg *msg; @@ -186,7 +186,8 @@ int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags) retry: error = create_rcom_stateless(ls, nodeid, DLM_RCOM_STATUS, - sizeof(struct rcom_status), &rc, &msg); + sizeof(struct rcom_status), &rc, &msg, + seq); if (error) goto out; @@ -220,7 +221,8 @@ int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags) return error; } -static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in, + uint64_t seq) { struct dlm_rcom *rc; struct rcom_status *rs; @@ -251,7 +253,7 @@ static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in) do_create: error = create_rcom_stateless(ls, nodeid, DLM_RCOM_STATUS_REPLY, - len, &rc, &msg); + len, &rc, &msg, seq); if (error) return; @@ -302,7 +304,8 @@ static void receive_sync_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in) spin_unlock(&ls->ls_rcom_spin); } -int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len) +int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, + int last_len, uint64_t seq) { struct dlm_rcom *rc; struct dlm_msg *msg; @@ -312,7 +315,7 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len) retry: error = create_rcom_stateless(ls, nodeid, DLM_RCOM_NAMES, last_len, - &rc, &msg); + &rc, &msg, seq); if (error) goto out; memcpy(rc->rc_buf, last_name, last_len); @@ -330,7 +333,8 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len) return error; } -static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in, + uint64_t seq) { struct dlm_rcom *rc; int error, inlen, outlen, nodeid; @@ -342,7 +346,7 @@ static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in) outlen = DLM_MAX_APP_BUFSIZE - sizeof(struct dlm_rcom); error = create_rcom_stateless(ls, nodeid, DLM_RCOM_NAMES_REPLY, outlen, - &rc, &msg); + &rc, &msg, seq); if (error) return; rc->rc_id = rc_in->rc_id; @@ -353,7 +357,7 @@ static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in) send_rcom_stateless(msg, rc); } -int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid) +int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid, uint64_t seq) { struct dlm_rcom *rc; struct dlm_mhandle *mh; @@ -361,7 +365,7 @@ int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid) int error; error = create_rcom(ls, dir_nodeid, DLM_RCOM_LOOKUP, r->res_length, - &rc, &mh); + &rc, &mh, seq); if (error) goto out; memcpy(rc->rc_buf, r->res_name, r->res_length); @@ -372,7 +376,8 @@ int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid) return error; } -static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in, + uint64_t seq) { struct dlm_rcom *rc; struct dlm_mhandle *mh; @@ -387,7 +392,8 @@ static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in) return; } - error = create_rcom(ls, nodeid, DLM_RCOM_LOOKUP_REPLY, 0, &rc, &mh); + error = create_rcom(ls, nodeid, DLM_RCOM_LOOKUP_REPLY, 0, &rc, &mh, + seq); if (error) return; @@ -437,7 +443,7 @@ static void pack_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, memcpy(rl->rl_lvb, lkb->lkb_lvbptr, r->res_ls->ls_lvblen); } -int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb) +int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, uint64_t seq) { struct dlm_ls *ls = r->res_ls; struct dlm_rcom *rc; @@ -448,7 +454,8 @@ int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb) if (lkb->lkb_lvbptr) len += ls->ls_lvblen; - error = create_rcom(ls, r->res_nodeid, DLM_RCOM_LOCK, len, &rc, &mh); + error = create_rcom(ls, r->res_nodeid, DLM_RCOM_LOCK, len, &rc, &mh, + seq); if (error) goto out; @@ -462,7 +469,8 @@ int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb) } /* needs at least dlm_rcom + rcom_lock */ -static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in, + uint64_t seq) { struct dlm_rcom *rc; struct dlm_mhandle *mh; @@ -471,7 +479,7 @@ static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in) dlm_recover_master_copy(ls, rc_in); error = create_rcom(ls, nodeid, DLM_RCOM_LOCK_REPLY, - sizeof(struct rcom_lock), &rc, &mh); + sizeof(struct rcom_lock), &rc, &mh, seq); if (error) return; @@ -620,21 +628,21 @@ void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid) switch (rc->rc_type) { case cpu_to_le32(DLM_RCOM_STATUS): - receive_rcom_status(ls, rc); + receive_rcom_status(ls, rc, seq); break; case cpu_to_le32(DLM_RCOM_NAMES): - receive_rcom_names(ls, rc); + receive_rcom_names(ls, rc, seq); break; case cpu_to_le32(DLM_RCOM_LOOKUP): - receive_rcom_lookup(ls, rc); + receive_rcom_lookup(ls, rc, seq); break; case cpu_to_le32(DLM_RCOM_LOCK): if (le16_to_cpu(rc->rc_header.h_length) < lock_size) goto Eshort; - receive_rcom_lock(ls, rc); + receive_rcom_lock(ls, rc, seq); break; case cpu_to_le32(DLM_RCOM_STATUS_REPLY): @@ -652,7 +660,7 @@ void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid) case cpu_to_le32(DLM_RCOM_LOCK_REPLY): if (le16_to_cpu(rc->rc_header.h_length) < lock_size) goto Eshort; - dlm_recover_process_copy(ls, rc); + dlm_recover_process_copy(ls, rc, seq); break; default: diff --git a/fs/dlm/rcom.h b/fs/dlm/rcom.h index 454d3c4814abe..9dd06d43ddb46 100644 --- a/fs/dlm/rcom.h +++ b/fs/dlm/rcom.h @@ -12,10 +12,12 @@ #ifndef __RCOM_DOT_H__ #define __RCOM_DOT_H__ -int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags); -int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name,int last_len); -int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid); -int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb); +int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags, + uint64_t seq); +int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, + int last_len, uint64_t seq); +int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid, uint64_t seq); +int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, uint64_t seq); void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid); int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in); diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index 29d71a5018d43..ddb6b3312cc1c 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -93,7 +93,7 @@ void dlm_set_recover_status(struct dlm_ls *ls, uint32_t status) } static int wait_status_all(struct dlm_ls *ls, uint32_t wait_status, - int save_slots) + int save_slots, uint64_t seq) { struct dlm_rcom *rc = ls->ls_recover_buf; struct dlm_member *memb; @@ -107,7 +107,7 @@ static int wait_status_all(struct dlm_ls *ls, uint32_t wait_status, goto out; } - error = dlm_rcom_status(ls, memb->nodeid, 0); + error = dlm_rcom_status(ls, memb->nodeid, 0, seq); if (error) goto out; @@ -126,7 +126,7 @@ static int wait_status_all(struct dlm_ls *ls, uint32_t wait_status, } static int wait_status_low(struct dlm_ls *ls, uint32_t wait_status, - uint32_t status_flags) + uint32_t status_flags, uint64_t seq) { struct dlm_rcom *rc = ls->ls_recover_buf; int error = 0, delay = 0, nodeid = ls->ls_low_nodeid; @@ -137,7 +137,7 @@ static int wait_status_low(struct dlm_ls *ls, uint32_t wait_status, goto out; } - error = dlm_rcom_status(ls, nodeid, status_flags); + error = dlm_rcom_status(ls, nodeid, status_flags, seq); if (error) break; @@ -151,22 +151,22 @@ static int wait_status_low(struct dlm_ls *ls, uint32_t wait_status, return error; } -static int wait_status(struct dlm_ls *ls, uint32_t status) +static int wait_status(struct dlm_ls *ls, uint32_t status, uint64_t seq) { uint32_t status_all = status << 1; int error; if (ls->ls_low_nodeid == dlm_our_nodeid()) { - error = wait_status_all(ls, status, 0); + error = wait_status_all(ls, status, 0, seq); if (!error) dlm_set_recover_status(ls, status_all); } else - error = wait_status_low(ls, status_all, 0); + error = wait_status_low(ls, status_all, 0, seq); return error; } -int dlm_recover_members_wait(struct dlm_ls *ls) +int dlm_recover_members_wait(struct dlm_ls *ls, uint64_t seq) { struct dlm_member *memb; struct dlm_slot *slots; @@ -180,7 +180,7 @@ int dlm_recover_members_wait(struct dlm_ls *ls) } if (ls->ls_low_nodeid == dlm_our_nodeid()) { - error = wait_status_all(ls, DLM_RS_NODES, 1); + error = wait_status_all(ls, DLM_RS_NODES, 1, seq); if (error) goto out; @@ -199,7 +199,8 @@ int dlm_recover_members_wait(struct dlm_ls *ls) dlm_set_recover_status(ls, DLM_RS_NODES_ALL); } } else { - error = wait_status_low(ls, DLM_RS_NODES_ALL, DLM_RSF_NEED_SLOTS); + error = wait_status_low(ls, DLM_RS_NODES_ALL, + DLM_RSF_NEED_SLOTS, seq); if (error) goto out; @@ -209,19 +210,19 @@ int dlm_recover_members_wait(struct dlm_ls *ls) return error; } -int dlm_recover_directory_wait(struct dlm_ls *ls) +int dlm_recover_directory_wait(struct dlm_ls *ls, uint64_t seq) { - return wait_status(ls, DLM_RS_DIR); + return wait_status(ls, DLM_RS_DIR, seq); } -int dlm_recover_locks_wait(struct dlm_ls *ls) +int dlm_recover_locks_wait(struct dlm_ls *ls, uint64_t seq) { - return wait_status(ls, DLM_RS_LOCKS); + return wait_status(ls, DLM_RS_LOCKS, seq); } -int dlm_recover_done_wait(struct dlm_ls *ls) +int dlm_recover_done_wait(struct dlm_ls *ls, uint64_t seq) { - return wait_status(ls, DLM_RS_DONE); + return wait_status(ls, DLM_RS_DONE, seq); } /* @@ -441,7 +442,7 @@ static void set_new_master(struct dlm_rsb *r) * equals our_nodeid below). */ -static int recover_master(struct dlm_rsb *r, unsigned int *count) +static int recover_master(struct dlm_rsb *r, unsigned int *count, uint64_t seq) { struct dlm_ls *ls = r->res_ls; int our_nodeid, dir_nodeid; @@ -472,7 +473,7 @@ static int recover_master(struct dlm_rsb *r, unsigned int *count) error = 0; } else { recover_idr_add(r); - error = dlm_send_rcom_lookup(r, dir_nodeid); + error = dlm_send_rcom_lookup(r, dir_nodeid, seq); } (*count)++; @@ -520,7 +521,7 @@ static int recover_master_static(struct dlm_rsb *r, unsigned int *count) * the correct dir node. */ -int dlm_recover_masters(struct dlm_ls *ls) +int dlm_recover_masters(struct dlm_ls *ls, uint64_t seq) { struct dlm_rsb *r; unsigned int total = 0; @@ -542,7 +543,7 @@ int dlm_recover_masters(struct dlm_ls *ls) if (nodir) error = recover_master_static(r, &count); else - error = recover_master(r, &count); + error = recover_master(r, &count, seq); unlock_rsb(r); cond_resched(); total++; @@ -614,13 +615,14 @@ int dlm_recover_master_reply(struct dlm_ls *ls, struct dlm_rcom *rc) * an equal number of replies then recovery for the rsb is done */ -static int recover_locks_queue(struct dlm_rsb *r, struct list_head *head) +static int recover_locks_queue(struct dlm_rsb *r, struct list_head *head, + uint64_t seq) { struct dlm_lkb *lkb; int error = 0; list_for_each_entry(lkb, head, lkb_statequeue) { - error = dlm_send_rcom_lock(r, lkb); + error = dlm_send_rcom_lock(r, lkb, seq); if (error) break; r->res_recover_locks_count++; @@ -629,7 +631,7 @@ static int recover_locks_queue(struct dlm_rsb *r, struct list_head *head) return error; } -static int recover_locks(struct dlm_rsb *r) +static int recover_locks(struct dlm_rsb *r, uint64_t seq) { int error = 0; @@ -637,13 +639,13 @@ static int recover_locks(struct dlm_rsb *r) DLM_ASSERT(!r->res_recover_locks_count, dlm_dump_rsb(r);); - error = recover_locks_queue(r, &r->res_grantqueue); + error = recover_locks_queue(r, &r->res_grantqueue, seq); if (error) goto out; - error = recover_locks_queue(r, &r->res_convertqueue); + error = recover_locks_queue(r, &r->res_convertqueue, seq); if (error) goto out; - error = recover_locks_queue(r, &r->res_waitqueue); + error = recover_locks_queue(r, &r->res_waitqueue, seq); if (error) goto out; @@ -656,7 +658,7 @@ static int recover_locks(struct dlm_rsb *r) return error; } -int dlm_recover_locks(struct dlm_ls *ls) +int dlm_recover_locks(struct dlm_ls *ls, uint64_t seq) { struct dlm_rsb *r; int error, count = 0; @@ -677,7 +679,7 @@ int dlm_recover_locks(struct dlm_ls *ls) goto out; } - error = recover_locks(r); + error = recover_locks(r, seq); if (error) { up_read(&ls->ls_root_sem); goto out; diff --git a/fs/dlm/recover.h b/fs/dlm/recover.h index 235e0d25cd48c..c5ce2ef139342 100644 --- a/fs/dlm/recover.h +++ b/fs/dlm/recover.h @@ -15,13 +15,13 @@ int dlm_wait_function(struct dlm_ls *ls, int (*testfn) (struct dlm_ls *ls)); uint32_t dlm_recover_status(struct dlm_ls *ls); void dlm_set_recover_status(struct dlm_ls *ls, uint32_t status); -int dlm_recover_members_wait(struct dlm_ls *ls); -int dlm_recover_directory_wait(struct dlm_ls *ls); -int dlm_recover_locks_wait(struct dlm_ls *ls); -int dlm_recover_done_wait(struct dlm_ls *ls); -int dlm_recover_masters(struct dlm_ls *ls); +int dlm_recover_members_wait(struct dlm_ls *ls, uint64_t seq); +int dlm_recover_directory_wait(struct dlm_ls *ls, uint64_t seq); +int dlm_recover_locks_wait(struct dlm_ls *ls, uint64_t seq); +int dlm_recover_done_wait(struct dlm_ls *ls, uint64_t seq); +int dlm_recover_masters(struct dlm_ls *ls, uint64_t seq); int dlm_recover_master_reply(struct dlm_ls *ls, struct dlm_rcom *rc); -int dlm_recover_locks(struct dlm_ls *ls); +int dlm_recover_locks(struct dlm_ls *ls, uint64_t seq); void dlm_recovered_lock(struct dlm_rsb *r); int dlm_create_root_list(struct dlm_ls *ls); void dlm_release_root_list(struct dlm_ls *ls); diff --git a/fs/dlm/recoverd.c b/fs/dlm/recoverd.c index 19da816cfb09d..4d17491dea2fe 100644 --- a/fs/dlm/recoverd.c +++ b/fs/dlm/recoverd.c @@ -90,7 +90,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) dlm_set_recover_status(ls, DLM_RS_NODES); - error = dlm_recover_members_wait(ls); + error = dlm_recover_members_wait(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_members_wait error %d", error); goto fail; @@ -103,7 +103,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) * nodes their master rsb names that hash to us. */ - error = dlm_recover_directory(ls); + error = dlm_recover_directory(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_directory error %d", error); goto fail; @@ -111,7 +111,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) dlm_set_recover_status(ls, DLM_RS_DIR); - error = dlm_recover_directory_wait(ls); + error = dlm_recover_directory_wait(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_directory_wait error %d", error); goto fail; @@ -145,7 +145,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) * departed nodes. */ - error = dlm_recover_masters(ls); + error = dlm_recover_masters(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_masters error %d", error); goto fail; @@ -155,7 +155,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) * Send our locks on remastered rsb's to the new masters. */ - error = dlm_recover_locks(ls); + error = dlm_recover_locks(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_locks error %d", error); goto fail; @@ -163,7 +163,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) dlm_set_recover_status(ls, DLM_RS_LOCKS); - error = dlm_recover_locks_wait(ls); + error = dlm_recover_locks_wait(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_locks_wait error %d", error); goto fail; @@ -187,7 +187,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) */ dlm_set_recover_status(ls, DLM_RS_LOCKS); - error = dlm_recover_locks_wait(ls); + error = dlm_recover_locks_wait(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_locks_wait error %d", error); goto fail; @@ -206,7 +206,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv) dlm_set_recover_status(ls, DLM_RS_DONE); - error = dlm_recover_done_wait(ls); + error = dlm_recover_done_wait(ls, rv->seq); if (error) { log_rinfo(ls, "dlm_recover_done_wait error %d", error); goto fail; From 561c67d8a10142250baa2a2a4e8b5d95e9c97df9 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:46 -0400 Subject: [PATCH 055/186] fs: dlm: drop rxbuf manipulation in dlm_copy_master_names This patch removes the manipulation of the receive buffer in case of an error and be sure the buffer is null terminated before an error messagea is printed out. Instead of manipulate the receive buffer we tell inside the format string the maximum length the string buffer is being read. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/dir.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/fs/dlm/dir.c b/fs/dlm/dir.c index 3bf5bf7a37b4e..768cf8d43b2b4 100644 --- a/fs/dlm/dir.c +++ b/fs/dlm/dir.c @@ -245,9 +245,8 @@ void dlm_copy_master_names(struct dlm_ls *ls, char *inbuf, int inlen, if (inlen > 1) { r = find_rsb_root(ls, inbuf, inlen); if (!r) { - inbuf[inlen - 1] = '\0'; - log_error(ls, "copy_master_names from %d start %d %s", - nodeid, inlen, inbuf); + log_error(ls, "copy_master_names from %d start %d %.*s", + nodeid, inlen, inlen, inbuf); goto out; } list = r->res_root_list.next; From b9d2f6ada0083bad46f37a1238fea718b575e0fa Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:47 -0400 Subject: [PATCH 056/186] fs: dlm: drop rxbuf manipulation in dlm_recover_master_copy Currently dlm_recover_master_copy() manipulates the receive buffer of an rcom lock message and modifies it on the fly so a later memcpy() to a new rcom message with the same message has those new values. This patch avoids manipulating the received rcom message by store the values for the new rcom message in paremter assigned with call by reference. Later when dlm_send_rcom_lock() constructs a new message and memcpy() the receive buffer those values will be set on the new constructed message. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/lock.c | 10 +++++++--- fs/dlm/lock.h | 3 ++- fs/dlm/rcom.c | 12 ++++++++---- 3 files changed, 17 insertions(+), 8 deletions(-) diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index b489da38e685f..1cf666c7401db 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -5384,7 +5384,8 @@ static int receive_rcom_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, back the rcom_lock struct we got but with the remid field filled in. */ /* needs at least dlm_rcom + rcom_lock */ -int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc) +int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc, + __le32 *rl_remid, __le32 *rl_result) { struct rcom_lock *rl = (struct rcom_lock *) rc->rc_buf; struct dlm_rsb *r; @@ -5393,6 +5394,9 @@ int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc) int from_nodeid = le32_to_cpu(rc->rc_header.h_nodeid); int error; + /* init rl_remid with rcom lock rl_remid */ + *rl_remid = rl->rl_remid; + if (rl->rl_parent_lkid) { error = -EOPNOTSUPP; goto out; @@ -5448,7 +5452,7 @@ int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc) out_remid: /* this is the new value returned to the lock holder for saving in its process-copy lkb */ - rl->rl_remid = cpu_to_le32(lkb->lkb_id); + *rl_remid = cpu_to_le32(lkb->lkb_id); lkb->lkb_recover_seq = ls->ls_recover_seq; @@ -5459,7 +5463,7 @@ int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc) if (error && error != -EEXIST) log_rinfo(ls, "dlm_recover_master_copy remote %d %x error %d", from_nodeid, remid, error); - rl->rl_result = cpu_to_le32(error); + *rl_result = cpu_to_le32(error); return error; } diff --git a/fs/dlm/lock.h b/fs/dlm/lock.h index 222e682523b90..cd67ccfbbf9b4 100644 --- a/fs/dlm/lock.h +++ b/fs/dlm/lock.h @@ -36,7 +36,8 @@ void dlm_purge_mstcpy_locks(struct dlm_rsb *r); void dlm_recover_grant(struct dlm_ls *ls); int dlm_recover_waiters_post(struct dlm_ls *ls); void dlm_recover_waiters_pre(struct dlm_ls *ls); -int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc); +int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc, + __le32 *rl_remid, __le32 *rl_result); int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc, uint64_t seq); diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c index efe45e68287f4..0946431e370a0 100644 --- a/fs/dlm/rcom.c +++ b/fs/dlm/rcom.c @@ -472,21 +472,25 @@ int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, uint64_t seq) static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in, uint64_t seq) { + __le32 rl_remid, rl_result; + struct rcom_lock *rl; struct dlm_rcom *rc; struct dlm_mhandle *mh; int error, nodeid = le32_to_cpu(rc_in->rc_header.h_nodeid); - dlm_recover_master_copy(ls, rc_in); + dlm_recover_master_copy(ls, rc_in, &rl_remid, &rl_result); error = create_rcom(ls, nodeid, DLM_RCOM_LOCK_REPLY, sizeof(struct rcom_lock), &rc, &mh, seq); if (error) return; - /* We send back the same rcom_lock struct we received, but - dlm_recover_master_copy() has filled in rl_remid and rl_result */ - memcpy(rc->rc_buf, rc_in->rc_buf, sizeof(struct rcom_lock)); + rl = (struct rcom_lock *)rc->rc_buf; + /* set rl_remid and rl_result from dlm_recover_master_copy() */ + rl->rl_remid = rl_remid; + rl->rl_result = rl_result; + rc->rc_id = rc_in->rc_id; rc->rc_seq_reply = rc_in->rc_seq; From 1151935182b40bbe398905850f6f7f4fbb262e06 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:48 -0400 Subject: [PATCH 057/186] fs: dlm: constify receive buffer The dlm receive buffer should be never manipulated as DLM is the last instance of parsing layer. This patch constify the whole receive buffer so we are sure it never gets manipulated when it's being parsed. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/dir.c | 5 +- fs/dlm/dir.h | 4 +- fs/dlm/lock.c | 109 ++++++++++++++++++++++-------------------- fs/dlm/lock.h | 14 +++--- fs/dlm/member.c | 2 +- fs/dlm/member.h | 2 +- fs/dlm/midcomms.c | 16 ++++--- fs/dlm/rcom.c | 20 ++++---- fs/dlm/rcom.h | 5 +- fs/dlm/recover.c | 2 +- fs/dlm/recover.h | 2 +- fs/dlm/requestqueue.c | 3 +- fs/dlm/requestqueue.h | 3 +- 13 files changed, 101 insertions(+), 86 deletions(-) diff --git a/fs/dlm/dir.c b/fs/dlm/dir.c index 768cf8d43b2b4..f6acba4310a7b 100644 --- a/fs/dlm/dir.c +++ b/fs/dlm/dir.c @@ -196,7 +196,8 @@ int dlm_recover_directory(struct dlm_ls *ls, uint64_t seq) return error; } -static struct dlm_rsb *find_rsb_root(struct dlm_ls *ls, char *name, int len) +static struct dlm_rsb *find_rsb_root(struct dlm_ls *ls, const char *name, + int len) { struct dlm_rsb *r; uint32_t hash, bucket; @@ -232,7 +233,7 @@ static struct dlm_rsb *find_rsb_root(struct dlm_ls *ls, char *name, int len) for rsb's we're master of and whose directory node matches the requesting node. inbuf is the rsb name last sent, inlen is the name's length */ -void dlm_copy_master_names(struct dlm_ls *ls, char *inbuf, int inlen, +void dlm_copy_master_names(struct dlm_ls *ls, const char *inbuf, int inlen, char *outbuf, int outlen, int nodeid) { struct list_head *list; diff --git a/fs/dlm/dir.h b/fs/dlm/dir.h index 0635582da003e..39ecb69d7ef36 100644 --- a/fs/dlm/dir.h +++ b/fs/dlm/dir.h @@ -16,8 +16,8 @@ int dlm_dir_nodeid(struct dlm_rsb *rsb); int dlm_hash2nodeid(struct dlm_ls *ls, uint32_t hash); void dlm_recover_dir_nodeid(struct dlm_ls *ls); int dlm_recover_directory(struct dlm_ls *ls, uint64_t seq); -void dlm_copy_master_names(struct dlm_ls *ls, char *inbuf, int inlen, - char *outbuf, int outlen, int nodeid); +void dlm_copy_master_names(struct dlm_ls *ls, const char *inbuf, int inlen, + char *outbuf, int outlen, int nodeid); #endif /* __DIR_DOT_H__ */ diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 1cf666c7401db..652c51fbbf768 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -86,8 +86,8 @@ static int send_remove(struct dlm_rsb *r); static int _request_lock(struct dlm_rsb *r, struct dlm_lkb *lkb); static int _cancel_lock(struct dlm_rsb *r, struct dlm_lkb *lkb); static void __receive_convert_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, - struct dlm_message *ms, bool local); -static int receive_extralen(struct dlm_message *ms); + const struct dlm_message *ms, bool local); +static int receive_extralen(const struct dlm_message *ms); static void do_purge(struct dlm_ls *ls, int nodeid, int pid); static void toss_rsb(struct kref *kref); @@ -984,8 +984,8 @@ static void __dlm_master_lookup(struct dlm_ls *ls, struct dlm_rsb *r, int our_no * . dlm_master_lookup RECOVER_MASTER (fix_master 1, from_master 0) */ -int dlm_master_lookup(struct dlm_ls *ls, int from_nodeid, char *name, int len, - unsigned int flags, int *r_nodeid, int *result) +int dlm_master_lookup(struct dlm_ls *ls, int from_nodeid, const char *name, + int len, unsigned int flags, int *r_nodeid, int *result) { struct dlm_rsb *r = NULL; uint32_t hash, b; @@ -1106,7 +1106,7 @@ static void dlm_dump_rsb_hash(struct dlm_ls *ls, uint32_t hash) } } -void dlm_dump_rsb_name(struct dlm_ls *ls, char *name, int len) +void dlm_dump_rsb_name(struct dlm_ls *ls, const char *name, int len) { struct dlm_rsb *r = NULL; uint32_t hash, b; @@ -1459,7 +1459,7 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) set RESEND and dlm_recover_waiters_post() */ static int _remove_from_waiters(struct dlm_lkb *lkb, int mstype, - struct dlm_message *ms) + const struct dlm_message *ms) { struct dlm_ls *ls = lkb->lkb_resource->res_ls; int overlap_done = 0; @@ -1557,8 +1557,8 @@ static int remove_from_waiters(struct dlm_lkb *lkb, int mstype) /* Handles situations where we might be processing a "fake" or "local" reply in which we can't try to take waiters_mutex again. */ -static int remove_from_waiters_ms(struct dlm_lkb *lkb, struct dlm_message *ms, - bool local) +static int remove_from_waiters_ms(struct dlm_lkb *lkb, + const struct dlm_message *ms, bool local) { struct dlm_ls *ls = lkb->lkb_resource->res_ls; int error; @@ -1800,7 +1800,7 @@ static void set_lvb_unlock(struct dlm_rsb *r, struct dlm_lkb *lkb) /* lkb is process copy (pc) */ static void set_lvb_lock_pc(struct dlm_rsb *r, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { int b; @@ -1907,7 +1907,7 @@ static void grant_lock(struct dlm_rsb *r, struct dlm_lkb *lkb) } static void grant_lock_pc(struct dlm_rsb *r, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { set_lvb_lock_pc(r, lkb, ms); _grant_lock(r, lkb); @@ -1945,7 +1945,7 @@ static void munge_demoted(struct dlm_lkb *lkb) lkb->lkb_grmode = DLM_LOCK_NL; } -static void munge_altmode(struct dlm_lkb *lkb, struct dlm_message *ms) +static void munge_altmode(struct dlm_lkb *lkb, const struct dlm_message *ms) { if (ms->m_type != cpu_to_le32(DLM_MSG_REQUEST_REPLY) && ms->m_type != cpu_to_le32(DLM_MSG_GRANT)) { @@ -3641,8 +3641,9 @@ static int send_cancel_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, int rv) return send_common_reply(r, lkb, DLM_MSG_CANCEL_REPLY, rv); } -static int send_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms_in, - int ret_nodeid, int rv) +static int send_lookup_reply(struct dlm_ls *ls, + const struct dlm_message *ms_in, int ret_nodeid, + int rv) { struct dlm_rsb *r = &ls->ls_local_rsb; struct dlm_message *ms; @@ -3667,14 +3668,15 @@ static int send_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms_in, of message, unlike the send side where we can safely send everything about the lkb for any type of message */ -static void receive_flags(struct dlm_lkb *lkb, struct dlm_message *ms) +static void receive_flags(struct dlm_lkb *lkb, const struct dlm_message *ms) { lkb->lkb_exflags = le32_to_cpu(ms->m_exflags); dlm_set_sbflags_val(lkb, le32_to_cpu(ms->m_sbflags)); dlm_set_dflags_val(lkb, le32_to_cpu(ms->m_flags)); } -static void receive_flags_reply(struct dlm_lkb *lkb, struct dlm_message *ms, +static void receive_flags_reply(struct dlm_lkb *lkb, + const struct dlm_message *ms, bool local) { if (local) @@ -3684,14 +3686,14 @@ static void receive_flags_reply(struct dlm_lkb *lkb, struct dlm_message *ms, dlm_set_dflags_val(lkb, le32_to_cpu(ms->m_flags)); } -static int receive_extralen(struct dlm_message *ms) +static int receive_extralen(const struct dlm_message *ms) { return (le16_to_cpu(ms->m_header.h_length) - sizeof(struct dlm_message)); } static int receive_lvb(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { int len; @@ -3719,7 +3721,7 @@ static void fake_astfn(void *astparam) } static int receive_request_args(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { lkb->lkb_nodeid = le32_to_cpu(ms->m_header.h_nodeid); lkb->lkb_ownpid = le32_to_cpu(ms->m_pid); @@ -3741,7 +3743,7 @@ static int receive_request_args(struct dlm_ls *ls, struct dlm_lkb *lkb, } static int receive_convert_args(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { if (lkb->lkb_status != DLM_LKSTS_GRANTED) return -EBUSY; @@ -3756,7 +3758,7 @@ static int receive_convert_args(struct dlm_ls *ls, struct dlm_lkb *lkb, } static int receive_unlock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_message *ms) + const struct dlm_message *ms) { if (receive_lvb(ls, lkb, ms)) return -ENOMEM; @@ -3766,7 +3768,7 @@ static int receive_unlock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, /* We fill in the local-lkb fields with the info that send_xxxx_reply() uses to send a reply and that the remote end uses to process the reply. */ -static void setup_local_lkb(struct dlm_ls *ls, struct dlm_message *ms) +static void setup_local_lkb(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb = &ls->ls_local_lkb; lkb->lkb_nodeid = le32_to_cpu(ms->m_header.h_nodeid); @@ -3776,7 +3778,7 @@ static void setup_local_lkb(struct dlm_ls *ls, struct dlm_message *ms) /* This is called after the rsb is locked so that we can safely inspect fields in the lkb. */ -static int validate_message(struct dlm_lkb *lkb, struct dlm_message *ms) +static int validate_message(struct dlm_lkb *lkb, const struct dlm_message *ms) { int from = le32_to_cpu(ms->m_header.h_nodeid); int error = 0; @@ -3828,7 +3830,7 @@ static int validate_message(struct dlm_lkb *lkb, struct dlm_message *ms) return error; } -static int receive_request(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_request(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -3907,7 +3909,7 @@ static int receive_request(struct dlm_ls *ls, struct dlm_message *ms) return error; } -static int receive_convert(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_convert(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -3963,7 +3965,7 @@ static int receive_convert(struct dlm_ls *ls, struct dlm_message *ms) return error; } -static int receive_unlock(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_unlock(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4015,7 +4017,7 @@ static int receive_unlock(struct dlm_ls *ls, struct dlm_message *ms) return error; } -static int receive_cancel(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_cancel(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4051,7 +4053,7 @@ static int receive_cancel(struct dlm_ls *ls, struct dlm_message *ms) return error; } -static int receive_grant(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_grant(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4082,7 +4084,7 @@ static int receive_grant(struct dlm_ls *ls, struct dlm_message *ms) return 0; } -static int receive_bast(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_bast(struct dlm_ls *ls, const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4110,7 +4112,7 @@ static int receive_bast(struct dlm_ls *ls, struct dlm_message *ms) return 0; } -static void receive_lookup(struct dlm_ls *ls, struct dlm_message *ms) +static void receive_lookup(struct dlm_ls *ls, const struct dlm_message *ms) { int len, error, ret_nodeid, from_nodeid, our_nodeid; @@ -4130,7 +4132,7 @@ static void receive_lookup(struct dlm_ls *ls, struct dlm_message *ms) send_lookup_reply(ls, ms, ret_nodeid, error); } -static void receive_remove(struct dlm_ls *ls, struct dlm_message *ms) +static void receive_remove(struct dlm_ls *ls, const struct dlm_message *ms) { char name[DLM_RESNAME_MAXLEN+1]; struct dlm_rsb *r; @@ -4218,12 +4220,13 @@ static void receive_remove(struct dlm_ls *ls, struct dlm_message *ms) } } -static void receive_purge(struct dlm_ls *ls, struct dlm_message *ms) +static void receive_purge(struct dlm_ls *ls, const struct dlm_message *ms) { do_purge(ls, le32_to_cpu(ms->m_nodeid), le32_to_cpu(ms->m_pid)); } -static int receive_request_reply(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_request_reply(struct dlm_ls *ls, + const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4345,7 +4348,7 @@ static int receive_request_reply(struct dlm_ls *ls, struct dlm_message *ms) } static void __receive_convert_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, - struct dlm_message *ms, bool local) + const struct dlm_message *ms, bool local) { /* this is the value returned from do_convert() on the master */ switch (from_dlm_errno(le32_to_cpu(ms->m_result))) { @@ -4388,8 +4391,8 @@ static void __receive_convert_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, } } -static void _receive_convert_reply(struct dlm_lkb *lkb, struct dlm_message *ms, - bool local) +static void _receive_convert_reply(struct dlm_lkb *lkb, + const struct dlm_message *ms, bool local) { struct dlm_rsb *r = lkb->lkb_resource; int error; @@ -4412,7 +4415,8 @@ static void _receive_convert_reply(struct dlm_lkb *lkb, struct dlm_message *ms, put_rsb(r); } -static int receive_convert_reply(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_convert_reply(struct dlm_ls *ls, + const struct dlm_message *ms) { struct dlm_lkb *lkb; int error; @@ -4426,8 +4430,8 @@ static int receive_convert_reply(struct dlm_ls *ls, struct dlm_message *ms) return 0; } -static void _receive_unlock_reply(struct dlm_lkb *lkb, struct dlm_message *ms, - bool local) +static void _receive_unlock_reply(struct dlm_lkb *lkb, + const struct dlm_message *ms, bool local) { struct dlm_rsb *r = lkb->lkb_resource; int error; @@ -4463,7 +4467,8 @@ static void _receive_unlock_reply(struct dlm_lkb *lkb, struct dlm_message *ms, put_rsb(r); } -static int receive_unlock_reply(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_unlock_reply(struct dlm_ls *ls, + const struct dlm_message *ms) { struct dlm_lkb *lkb; int error; @@ -4477,8 +4482,8 @@ static int receive_unlock_reply(struct dlm_ls *ls, struct dlm_message *ms) return 0; } -static void _receive_cancel_reply(struct dlm_lkb *lkb, struct dlm_message *ms, - bool local) +static void _receive_cancel_reply(struct dlm_lkb *lkb, + const struct dlm_message *ms, bool local) { struct dlm_rsb *r = lkb->lkb_resource; int error; @@ -4515,7 +4520,8 @@ static void _receive_cancel_reply(struct dlm_lkb *lkb, struct dlm_message *ms, put_rsb(r); } -static int receive_cancel_reply(struct dlm_ls *ls, struct dlm_message *ms) +static int receive_cancel_reply(struct dlm_ls *ls, + const struct dlm_message *ms) { struct dlm_lkb *lkb; int error; @@ -4529,7 +4535,8 @@ static int receive_cancel_reply(struct dlm_ls *ls, struct dlm_message *ms) return 0; } -static void receive_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms) +static void receive_lookup_reply(struct dlm_ls *ls, + const struct dlm_message *ms) { struct dlm_lkb *lkb; struct dlm_rsb *r; @@ -4608,7 +4615,7 @@ static void receive_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms) dlm_put_lkb(lkb); } -static void _receive_message(struct dlm_ls *ls, struct dlm_message *ms, +static void _receive_message(struct dlm_ls *ls, const struct dlm_message *ms, uint32_t saved_seq) { int error = 0, noent = 0; @@ -4744,7 +4751,7 @@ static void _receive_message(struct dlm_ls *ls, struct dlm_message *ms, requestqueue, to processing all the saved messages, to processing new messages as they arrive. */ -static void dlm_receive_message(struct dlm_ls *ls, struct dlm_message *ms, +static void dlm_receive_message(struct dlm_ls *ls, const struct dlm_message *ms, int nodeid) { if (dlm_locking_stopped(ls)) { @@ -4767,7 +4774,7 @@ static void dlm_receive_message(struct dlm_ls *ls, struct dlm_message *ms, /* This is called by dlm_recoverd to process messages that were saved on the requestqueue. */ -void dlm_receive_message_saved(struct dlm_ls *ls, struct dlm_message *ms, +void dlm_receive_message_saved(struct dlm_ls *ls, const struct dlm_message *ms, uint32_t saved_seq) { _receive_message(ls, ms, saved_seq); @@ -4778,9 +4785,9 @@ void dlm_receive_message_saved(struct dlm_ls *ls, struct dlm_message *ms, standard locking activity) or an RCOM (recovery message sent as part of lockspace recovery). */ -void dlm_receive_buffer(union dlm_packet *p, int nodeid) +void dlm_receive_buffer(const union dlm_packet *p, int nodeid) { - struct dlm_header *hd = &p->header; + const struct dlm_header *hd = &p->header; struct dlm_ls *ls; int type = 0; @@ -5334,7 +5341,7 @@ static struct dlm_lkb *search_remid(struct dlm_rsb *r, int nodeid, /* needs at least dlm_rcom + rcom_lock */ static int receive_rcom_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_rsb *r, struct dlm_rcom *rc) + struct dlm_rsb *r, const struct dlm_rcom *rc) { struct rcom_lock *rl = (struct rcom_lock *) rc->rc_buf; @@ -5384,7 +5391,7 @@ static int receive_rcom_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, back the rcom_lock struct we got but with the remid field filled in. */ /* needs at least dlm_rcom + rcom_lock */ -int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc, +int dlm_recover_master_copy(struct dlm_ls *ls, const struct dlm_rcom *rc, __le32 *rl_remid, __le32 *rl_result) { struct rcom_lock *rl = (struct rcom_lock *) rc->rc_buf; @@ -5468,7 +5475,7 @@ int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc, } /* needs at least dlm_rcom + rcom_lock */ -int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc, +int dlm_recover_process_copy(struct dlm_ls *ls, const struct dlm_rcom *rc, uint64_t seq) { struct rcom_lock *rl = (struct rcom_lock *) rc->rc_buf; diff --git a/fs/dlm/lock.h b/fs/dlm/lock.h index cd67ccfbbf9b4..b54e2cbbe6e27 100644 --- a/fs/dlm/lock.h +++ b/fs/dlm/lock.h @@ -12,11 +12,11 @@ #define __LOCK_DOT_H__ void dlm_dump_rsb(struct dlm_rsb *r); -void dlm_dump_rsb_name(struct dlm_ls *ls, char *name, int len); +void dlm_dump_rsb_name(struct dlm_ls *ls, const char *name, int len); void dlm_print_lkb(struct dlm_lkb *lkb); -void dlm_receive_message_saved(struct dlm_ls *ls, struct dlm_message *ms, +void dlm_receive_message_saved(struct dlm_ls *ls, const struct dlm_message *ms, uint32_t saved_seq); -void dlm_receive_buffer(union dlm_packet *p, int nodeid); +void dlm_receive_buffer(const union dlm_packet *p, int nodeid); int dlm_modes_compat(int mode1, int mode2); void dlm_put_rsb(struct dlm_rsb *r); void dlm_hold_rsb(struct dlm_rsb *r); @@ -25,8 +25,8 @@ void dlm_scan_rsbs(struct dlm_ls *ls); int dlm_lock_recovery_try(struct dlm_ls *ls); void dlm_unlock_recovery(struct dlm_ls *ls); -int dlm_master_lookup(struct dlm_ls *ls, int nodeid, char *name, int len, - unsigned int flags, int *r_nodeid, int *result); +int dlm_master_lookup(struct dlm_ls *ls, int from_nodeid, const char *name, + int len, unsigned int flags, int *r_nodeid, int *result); int dlm_search_rsb_tree(struct rb_root *tree, const void *name, int len, struct dlm_rsb **r_ret); @@ -36,9 +36,9 @@ void dlm_purge_mstcpy_locks(struct dlm_rsb *r); void dlm_recover_grant(struct dlm_ls *ls); int dlm_recover_waiters_post(struct dlm_ls *ls); void dlm_recover_waiters_pre(struct dlm_ls *ls); -int dlm_recover_master_copy(struct dlm_ls *ls, struct dlm_rcom *rc, +int dlm_recover_master_copy(struct dlm_ls *ls, const struct dlm_rcom *rc, __le32 *rl_remid, __le32 *rl_result); -int dlm_recover_process_copy(struct dlm_ls *ls, struct dlm_rcom *rc, +int dlm_recover_process_copy(struct dlm_ls *ls, const struct dlm_rcom *rc, uint64_t seq); int dlm_user_request(struct dlm_ls *ls, struct dlm_user_args *ua, int mode, diff --git a/fs/dlm/member.c b/fs/dlm/member.c index 19f3cd96f3c0d..be7909ead71b4 100644 --- a/fs/dlm/member.c +++ b/fs/dlm/member.c @@ -18,7 +18,7 @@ #include "midcomms.h" #include "lowcomms.h" -int dlm_slots_version(struct dlm_header *h) +int dlm_slots_version(const struct dlm_header *h) { if ((le32_to_cpu(h->h_version) & 0x0000FFFF) < DLM_HEADER_SLOTS) return 0; diff --git a/fs/dlm/member.h b/fs/dlm/member.h index 433b2fac9f4a7..f61cfde463140 100644 --- a/fs/dlm/member.h +++ b/fs/dlm/member.h @@ -18,7 +18,7 @@ void dlm_clear_members_gone(struct dlm_ls *ls); int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv,int *neg_out); int dlm_is_removed(struct dlm_ls *ls, int nodeid); int dlm_is_member(struct dlm_ls *ls, int nodeid); -int dlm_slots_version(struct dlm_header *h); +int dlm_slots_version(const struct dlm_header *h); void dlm_slot_save(struct dlm_ls *ls, struct dlm_rcom *rc, struct dlm_member *memb); void dlm_slots_copy_out(struct dlm_ls *ls, struct dlm_rcom *rc); diff --git a/fs/dlm/midcomms.c b/fs/dlm/midcomms.c index 8ebffbfdc00ae..c125496e03bf1 100644 --- a/fs/dlm/midcomms.c +++ b/fs/dlm/midcomms.c @@ -499,7 +499,8 @@ static void dlm_pas_fin_ack_rcv(struct midcomms_node *node) spin_unlock(&node->state_lock); } -static void dlm_receive_buffer_3_2_trace(uint32_t seq, union dlm_packet *p) +static void dlm_receive_buffer_3_2_trace(uint32_t seq, + const union dlm_packet *p) { switch (p->header.h_cmd) { case DLM_MSG: @@ -513,7 +514,7 @@ static void dlm_receive_buffer_3_2_trace(uint32_t seq, union dlm_packet *p) } } -static void dlm_midcomms_receive_buffer(union dlm_packet *p, +static void dlm_midcomms_receive_buffer(const union dlm_packet *p, struct midcomms_node *node, uint32_t seq) { @@ -708,7 +709,8 @@ static int dlm_midcomms_version_check_3_2(struct midcomms_node *node) return 0; } -static int dlm_opts_check_msglen(union dlm_packet *p, uint16_t msglen, int nodeid) +static int dlm_opts_check_msglen(const union dlm_packet *p, uint16_t msglen, + int nodeid) { int len = msglen; @@ -757,7 +759,7 @@ static int dlm_opts_check_msglen(union dlm_packet *p, uint16_t msglen, int nodei return 0; } -static void dlm_midcomms_receive_buffer_3_2(union dlm_packet *p, int nodeid) +static void dlm_midcomms_receive_buffer_3_2(const union dlm_packet *p, int nodeid) { uint16_t msglen = le16_to_cpu(p->header.h_length); struct midcomms_node *node; @@ -878,7 +880,7 @@ static int dlm_midcomms_version_check_3_1(struct midcomms_node *node) return 0; } -static void dlm_midcomms_receive_buffer_3_1(union dlm_packet *p, int nodeid) +static void dlm_midcomms_receive_buffer_3_1(const union dlm_packet *p, int nodeid) { uint16_t msglen = le16_to_cpu(p->header.h_length); struct midcomms_node *node; @@ -977,10 +979,10 @@ int dlm_process_incoming_buffer(int nodeid, unsigned char *buf, int len) switch (hd->h_version) { case cpu_to_le32(DLM_VERSION_3_1): - dlm_midcomms_receive_buffer_3_1((union dlm_packet *)ptr, nodeid); + dlm_midcomms_receive_buffer_3_1((const union dlm_packet *)ptr, nodeid); break; case cpu_to_le32(DLM_VERSION_3_2): - dlm_midcomms_receive_buffer_3_2((union dlm_packet *)ptr, nodeid); + dlm_midcomms_receive_buffer_3_2((const union dlm_packet *)ptr, nodeid); break; default: log_print("received invalid version header: %u from node %d, will skip this message", diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c index 0946431e370a0..6ab029149a1d7 100644 --- a/fs/dlm/rcom.c +++ b/fs/dlm/rcom.c @@ -221,7 +221,8 @@ int dlm_rcom_status(struct dlm_ls *ls, int nodeid, uint32_t status_flags, return error; } -static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in, +static void receive_rcom_status(struct dlm_ls *ls, + const struct dlm_rcom *rc_in, uint64_t seq) { struct dlm_rcom *rc; @@ -283,7 +284,7 @@ static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in, send_rcom_stateless(msg, rc); } -static void receive_sync_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_sync_reply(struct dlm_ls *ls, const struct dlm_rcom *rc_in) { spin_lock(&ls->ls_rcom_spin); if (!test_bit(LSFL_RCOM_WAIT, &ls->ls_flags) || @@ -333,7 +334,7 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, return error; } -static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in, +static void receive_rcom_names(struct dlm_ls *ls, const struct dlm_rcom *rc_in, uint64_t seq) { struct dlm_rcom *rc; @@ -376,8 +377,8 @@ int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid, uint64_t seq) return error; } -static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in, - uint64_t seq) +static void receive_rcom_lookup(struct dlm_ls *ls, + const struct dlm_rcom *rc_in, uint64_t seq) { struct dlm_rcom *rc; struct dlm_mhandle *mh; @@ -408,7 +409,8 @@ static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in, send_rcom(mh, rc); } -static void receive_rcom_lookup_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in) +static void receive_rcom_lookup_reply(struct dlm_ls *ls, + const struct dlm_rcom *rc_in) { dlm_recover_master_reply(ls, rc_in); } @@ -469,7 +471,7 @@ int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, uint64_t seq) } /* needs at least dlm_rcom + rcom_lock */ -static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in, +static void receive_rcom_lock(struct dlm_ls *ls, const struct dlm_rcom *rc_in, uint64_t seq) { __le32 rl_remid, rl_result; @@ -500,7 +502,7 @@ static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in, /* If the lockspace doesn't exist then still send a status message back; it's possible that it just doesn't have its global_id yet. */ -int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in) +int dlm_send_ls_not_ready(int nodeid, const struct dlm_rcom *rc_in) { struct dlm_rcom *rc; struct rcom_config *rf; @@ -578,7 +580,7 @@ int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in) /* Called by dlm_recv; corresponds to dlm_receive_message() but special recovery-only comms are sent through here. */ -void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid) +void dlm_receive_rcom(struct dlm_ls *ls, const struct dlm_rcom *rc, int nodeid) { int lock_size = sizeof(struct dlm_rcom) + sizeof(struct rcom_lock); int stop, reply = 0, names = 0, lookup = 0, lock = 0; diff --git a/fs/dlm/rcom.h b/fs/dlm/rcom.h index 9dd06d43ddb46..765926ae0020d 100644 --- a/fs/dlm/rcom.h +++ b/fs/dlm/rcom.h @@ -18,8 +18,9 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len, uint64_t seq); int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid, uint64_t seq); int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, uint64_t seq); -void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid); -int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in); +void dlm_receive_rcom(struct dlm_ls *ls, const struct dlm_rcom *rc, + int nodeid); +int dlm_send_ls_not_ready(int nodeid, const struct dlm_rcom *rc_in); #endif diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index ddb6b3312cc1c..53917c0aa3c07 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -564,7 +564,7 @@ int dlm_recover_masters(struct dlm_ls *ls, uint64_t seq) return error; } -int dlm_recover_master_reply(struct dlm_ls *ls, struct dlm_rcom *rc) +int dlm_recover_master_reply(struct dlm_ls *ls, const struct dlm_rcom *rc) { struct dlm_rsb *r; int ret_nodeid, new_master; diff --git a/fs/dlm/recover.h b/fs/dlm/recover.h index c5ce2ef139342..dbc51013ecadb 100644 --- a/fs/dlm/recover.h +++ b/fs/dlm/recover.h @@ -20,7 +20,7 @@ int dlm_recover_directory_wait(struct dlm_ls *ls, uint64_t seq); int dlm_recover_locks_wait(struct dlm_ls *ls, uint64_t seq); int dlm_recover_done_wait(struct dlm_ls *ls, uint64_t seq); int dlm_recover_masters(struct dlm_ls *ls, uint64_t seq); -int dlm_recover_master_reply(struct dlm_ls *ls, struct dlm_rcom *rc); +int dlm_recover_master_reply(struct dlm_ls *ls, const struct dlm_rcom *rc); int dlm_recover_locks(struct dlm_ls *ls, uint64_t seq); void dlm_recovered_lock(struct dlm_rsb *r); int dlm_create_root_list(struct dlm_ls *ls); diff --git a/fs/dlm/requestqueue.c b/fs/dlm/requestqueue.c index 8be2893ad15bb..892d6ca21e742 100644 --- a/fs/dlm/requestqueue.c +++ b/fs/dlm/requestqueue.c @@ -30,7 +30,8 @@ struct rq_entry { * lockspace is enabled on some while still suspended on others. */ -void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_message *ms) +void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, + const struct dlm_message *ms) { struct rq_entry *e; int length = le16_to_cpu(ms->m_header.h_length) - diff --git a/fs/dlm/requestqueue.h b/fs/dlm/requestqueue.h index 4e403469a8458..42bfe23ceabea 100644 --- a/fs/dlm/requestqueue.h +++ b/fs/dlm/requestqueue.h @@ -11,7 +11,8 @@ #ifndef __REQUESTQUEUE_DOT_H__ #define __REQUESTQUEUE_DOT_H__ -void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_message *ms); +void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, + const struct dlm_message *ms); int dlm_process_requestqueue(struct dlm_ls *ls); void dlm_wait_requestqueue(struct dlm_ls *ls); void dlm_purge_requestqueue(struct dlm_ls *ls); From 63e711b081609748d631fc3a08b14ba01c8e4f16 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:49 -0400 Subject: [PATCH 058/186] fs: dlm: create midcomms nodes when configure This patch puts the life of a midcomms node the same as a lowcomms connection. The lowcomms connection lifetime was changed by commit 6f0b0b5d7ae7 ("fs: dlm: remove dlm_node_addrs lookup list"). In the future the midcomms node instances can be merged with lowcomms connection structure as the lifetime is the same and states can be controlled over values or flags. Before midcomms nodes were generated during version detection. This is not necessary anymore when the nodes are created when the cluster manager configures DLM via configfs. When a midcomms node is created over configfs it well set DLM_VERSION_NOT_SET as version. This indicates that the version of the midcomms node is still unknown and need to be probed via certain rcom messages. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/config.c | 2 +- fs/dlm/midcomms.c | 286 +++++++++++++++++----------------------------- fs/dlm/midcomms.h | 1 + 3 files changed, 110 insertions(+), 179 deletions(-) diff --git a/fs/dlm/config.c b/fs/dlm/config.c index 2beceff024e39..e55e0a2cd2e86 100644 --- a/fs/dlm/config.c +++ b/fs/dlm/config.c @@ -664,7 +664,7 @@ static ssize_t comm_addr_store(struct config_item *item, const char *buf, memcpy(addr, buf, len); - rv = dlm_lowcomms_addr(cm->nodeid, addr, len); + rv = dlm_midcomms_addr(cm->nodeid, addr, len); if (rv) { kfree(addr); return rv; diff --git a/fs/dlm/midcomms.c b/fs/dlm/midcomms.c index c125496e03bf1..f641b36a36db0 100644 --- a/fs/dlm/midcomms.c +++ b/fs/dlm/midcomms.c @@ -330,18 +330,23 @@ static void midcomms_node_reset(struct midcomms_node *node) wake_up(&node->shutdown_wait); } -static struct midcomms_node *nodeid2node(int nodeid, gfp_t alloc) +static struct midcomms_node *nodeid2node(int nodeid) { - struct midcomms_node *node, *tmp; - int r = nodeid_hash(nodeid); + return __find_node(nodeid, nodeid_hash(nodeid)); +} + +int dlm_midcomms_addr(int nodeid, struct sockaddr_storage *addr, int len) +{ + int ret, r = nodeid_hash(nodeid); + struct midcomms_node *node; - node = __find_node(nodeid, r); - if (node || !alloc) - return node; + ret = dlm_lowcomms_addr(nodeid, addr, len); + if (ret) + return ret; - node = kmalloc(sizeof(*node), alloc); + node = kmalloc(sizeof(*node), GFP_NOFS); if (!node) - return NULL; + return -ENOMEM; node->nodeid = nodeid; spin_lock_init(&node->state_lock); @@ -353,21 +358,11 @@ static struct midcomms_node *nodeid2node(int nodeid, gfp_t alloc) midcomms_node_reset(node); spin_lock(&nodes_lock); - /* check again if there was somebody else - * earlier here to add the node - */ - tmp = __find_node(nodeid, r); - if (tmp) { - spin_unlock(&nodes_lock); - kfree(node); - return tmp; - } - hlist_add_head_rcu(&node->hlist, &node_hash[r]); spin_unlock(&nodes_lock); node->debugfs = dlm_create_debug_comms_file(nodeid, node); - return node; + return 0; } static int dlm_send_ack(int nodeid, uint32_t seq) @@ -603,112 +598,6 @@ static void dlm_midcomms_receive_buffer(const union dlm_packet *p, } } -static struct midcomms_node * -dlm_midcomms_recv_node_lookup(int nodeid, const union dlm_packet *p, - uint16_t msglen, int (*cb)(struct midcomms_node *node)) -{ - struct midcomms_node *node = NULL; - gfp_t allocation = 0; - int ret; - - switch (p->header.h_cmd) { - case DLM_RCOM: - if (msglen < sizeof(struct dlm_rcom)) { - log_print("rcom msg too small: %u, will skip this message from node %d", - msglen, nodeid); - return NULL; - } - - switch (p->rcom.rc_type) { - case cpu_to_le32(DLM_RCOM_NAMES): - fallthrough; - case cpu_to_le32(DLM_RCOM_NAMES_REPLY): - fallthrough; - case cpu_to_le32(DLM_RCOM_STATUS): - fallthrough; - case cpu_to_le32(DLM_RCOM_STATUS_REPLY): - node = nodeid2node(nodeid, 0); - if (node) { - spin_lock(&node->state_lock); - if (node->state != DLM_ESTABLISHED) - pr_debug("receive begin RCOM msg from node %d with state %s\n", - node->nodeid, dlm_state_str(node->state)); - - switch (node->state) { - case DLM_CLOSED: - node->state = DLM_ESTABLISHED; - pr_debug("switch node %d to state %s\n", - node->nodeid, dlm_state_str(node->state)); - break; - case DLM_ESTABLISHED: - break; - default: - spin_unlock(&node->state_lock); - return NULL; - } - spin_unlock(&node->state_lock); - } - - allocation = GFP_NOFS; - break; - default: - break; - } - - break; - default: - break; - } - - node = nodeid2node(nodeid, allocation); - if (!node) { - switch (p->header.h_cmd) { - case DLM_OPTS: - if (msglen < sizeof(struct dlm_opts)) { - log_print("opts msg too small: %u, will skip this message from node %d", - msglen, nodeid); - return NULL; - } - - log_print_ratelimited("received dlm opts message nextcmd %d from node %d in an invalid sequence", - p->opts.o_nextcmd, nodeid); - break; - default: - log_print_ratelimited("received dlm message cmd %d from node %d in an invalid sequence", - p->header.h_cmd, nodeid); - break; - } - - return NULL; - } - - ret = cb(node); - if (ret < 0) - return NULL; - - return node; -} - -static int dlm_midcomms_version_check_3_2(struct midcomms_node *node) -{ - switch (node->version) { - case DLM_VERSION_NOT_SET: - node->version = DLM_VERSION_3_2; - wake_up(&node->shutdown_wait); - log_print("version 0x%08x for node %d detected", DLM_VERSION_3_2, - node->nodeid); - break; - case DLM_VERSION_3_2: - break; - default: - log_print_ratelimited("version mismatch detected, assumed 0x%08x but node %d has 0x%08x", - DLM_VERSION_3_2, node->nodeid, node->version); - return -1; - } - - return 0; -} - static int dlm_opts_check_msglen(const union dlm_packet *p, uint16_t msglen, int nodeid) { @@ -767,10 +656,37 @@ static void dlm_midcomms_receive_buffer_3_2(const union dlm_packet *p, int nodei int ret, idx; idx = srcu_read_lock(&nodes_srcu); - node = dlm_midcomms_recv_node_lookup(nodeid, p, msglen, - dlm_midcomms_version_check_3_2); - if (!node) + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) + goto out; + + switch (node->version) { + case DLM_VERSION_NOT_SET: + node->version = DLM_VERSION_3_2; + wake_up(&node->shutdown_wait); + log_print("version 0x%08x for node %d detected", DLM_VERSION_3_2, + node->nodeid); + + spin_lock(&node->state_lock); + switch (node->state) { + case DLM_CLOSED: + node->state = DLM_ESTABLISHED; + pr_debug("switch node %d to state %s\n", + node->nodeid, dlm_state_str(node->state)); + break; + default: + break; + } + spin_unlock(&node->state_lock); + + break; + case DLM_VERSION_3_2: + break; + default: + log_print_ratelimited("version mismatch detected, assumed 0x%08x but node %d has 0x%08x", + DLM_VERSION_3_2, node->nodeid, node->version); goto out; + } switch (p->header.h_cmd) { case DLM_RCOM: @@ -860,8 +776,19 @@ static void dlm_midcomms_receive_buffer_3_2(const union dlm_packet *p, int nodei srcu_read_unlock(&nodes_srcu, idx); } -static int dlm_midcomms_version_check_3_1(struct midcomms_node *node) +static void dlm_midcomms_receive_buffer_3_1(const union dlm_packet *p, int nodeid) { + uint16_t msglen = le16_to_cpu(p->header.h_length); + struct midcomms_node *node; + int idx; + + idx = srcu_read_lock(&nodes_srcu); + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) { + srcu_read_unlock(&nodes_srcu, idx); + return; + } + switch (node->version) { case DLM_VERSION_NOT_SET: node->version = DLM_VERSION_3_1; @@ -874,22 +801,6 @@ static int dlm_midcomms_version_check_3_1(struct midcomms_node *node) default: log_print_ratelimited("version mismatch detected, assumed 0x%08x but node %d has 0x%08x", DLM_VERSION_3_1, node->nodeid, node->version); - return -1; - } - - return 0; -} - -static void dlm_midcomms_receive_buffer_3_1(const union dlm_packet *p, int nodeid) -{ - uint16_t msglen = le16_to_cpu(p->header.h_length); - struct midcomms_node *node; - int idx; - - idx = srcu_read_lock(&nodes_srcu); - node = dlm_midcomms_recv_node_lookup(nodeid, p, msglen, - dlm_midcomms_version_check_3_1); - if (!node) { srcu_read_unlock(&nodes_srcu, idx); return; } @@ -1005,8 +916,8 @@ void dlm_midcomms_unack_msg_resend(int nodeid) int idx, ret; idx = srcu_read_lock(&nodes_srcu); - node = nodeid2node(nodeid, 0); - if (!node) { + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) { srcu_read_unlock(&nodes_srcu, idx); return; } @@ -1092,11 +1003,9 @@ struct dlm_mhandle *dlm_midcomms_get_mhandle(int nodeid, int len, int idx; idx = srcu_read_lock(&nodes_srcu); - node = nodeid2node(nodeid, 0); - if (!node) { - WARN_ON_ONCE(1); + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) goto err; - } /* this is a bug, however we going on and hope it will be resolved */ WARN_ON_ONCE(test_bit(DLM_NODE_FLAG_STOP_TX, &node->flags)); @@ -1237,8 +1146,34 @@ void dlm_midcomms_init(void) dlm_lowcomms_init(); } +static void midcomms_node_release(struct rcu_head *rcu) +{ + struct midcomms_node *node = container_of(rcu, struct midcomms_node, rcu); + + WARN_ON_ONCE(atomic_read(&node->send_queue_cnt)); + dlm_send_queue_flush(node); + kfree(node); +} + void dlm_midcomms_exit(void) { + struct midcomms_node *node; + int i, idx; + + idx = srcu_read_lock(&nodes_srcu); + for (i = 0; i < CONN_HASH_SIZE; i++) { + hlist_for_each_entry_rcu(node, &node_hash[i], hlist) { + dlm_delete_debug_comms_file(node->debugfs); + + spin_lock(&nodes_lock); + hlist_del_rcu(&node->hlist); + spin_unlock(&nodes_lock); + + call_srcu(&nodes_srcu, &node->rcu, midcomms_node_release); + } + } + srcu_read_unlock(&nodes_srcu, idx); + dlm_lowcomms_exit(); } @@ -1279,8 +1214,8 @@ void dlm_midcomms_add_member(int nodeid) int idx; idx = srcu_read_lock(&nodes_srcu); - node = nodeid2node(nodeid, GFP_NOFS); - if (!node) { + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) { srcu_read_unlock(&nodes_srcu, idx); return; } @@ -1324,8 +1259,8 @@ void dlm_midcomms_remove_member(int nodeid) int idx; idx = srcu_read_lock(&nodes_srcu); - node = nodeid2node(nodeid, 0); - if (!node) { + node = nodeid2node(nodeid); + if (WARN_ON_ONCE(!node)) { srcu_read_unlock(&nodes_srcu, idx); return; } @@ -1369,15 +1304,6 @@ void dlm_midcomms_remove_member(int nodeid) srcu_read_unlock(&nodes_srcu, idx); } -static void midcomms_node_release(struct rcu_head *rcu) -{ - struct midcomms_node *node = container_of(rcu, struct midcomms_node, rcu); - - WARN_ON_ONCE(atomic_read(&node->send_queue_cnt)); - dlm_send_queue_flush(node); - kfree(node); -} - void dlm_midcomms_version_wait(void) { struct midcomms_node *node; @@ -1440,7 +1366,7 @@ static void midcomms_shutdown(struct midcomms_node *node) node->state == DLM_CLOSED || test_bit(DLM_NODE_FLAG_CLOSE, &node->flags), DLM_SHUTDOWN_TIMEOUT); - if (!ret || test_bit(DLM_NODE_FLAG_CLOSE, &node->flags)) + if (!ret) pr_debug("active shutdown timed out for node %d with state %s\n", node->nodeid, dlm_state_str(node->state)); else @@ -1458,14 +1384,6 @@ void dlm_midcomms_shutdown(void) for (i = 0; i < CONN_HASH_SIZE; i++) { hlist_for_each_entry_rcu(node, &node_hash[i], hlist) { midcomms_shutdown(node); - - dlm_delete_debug_comms_file(node->debugfs); - - spin_lock(&nodes_lock); - hlist_del_rcu(&node->hlist); - spin_unlock(&nodes_lock); - - call_srcu(&nodes_srcu, &node->rcu, midcomms_node_release); } } srcu_read_unlock(&nodes_srcu, idx); @@ -1481,7 +1399,7 @@ int dlm_midcomms_close(int nodeid) idx = srcu_read_lock(&nodes_srcu); /* Abort pending close/remove operation */ - node = nodeid2node(nodeid, 0); + node = nodeid2node(nodeid); if (node) { /* let shutdown waiters leave */ set_bit(DLM_NODE_FLAG_CLOSE, &node->flags); @@ -1493,7 +1411,7 @@ int dlm_midcomms_close(int nodeid) mutex_lock(&close_lock); idx = srcu_read_lock(&nodes_srcu); - node = nodeid2node(nodeid, 0); + node = nodeid2node(nodeid); if (!node) { srcu_read_unlock(&nodes_srcu, idx); mutex_unlock(&close_lock); @@ -1501,10 +1419,22 @@ int dlm_midcomms_close(int nodeid) } ret = dlm_lowcomms_close(nodeid); - spin_lock(&node->state_lock); - midcomms_node_reset(node); - spin_unlock(&node->state_lock); + dlm_delete_debug_comms_file(node->debugfs); + + spin_lock(&nodes_lock); + hlist_del_rcu(&node->hlist); + spin_unlock(&nodes_lock); srcu_read_unlock(&nodes_srcu, idx); + + /* wait that all readers left until flush send queue */ + synchronize_srcu(&nodes_srcu); + + /* drop all pending dlm messages, this is fine as + * this function get called when the node is fenced + */ + dlm_send_queue_flush(node); + + call_srcu(&nodes_srcu, &node->rcu, midcomms_node_release); mutex_unlock(&close_lock); return ret; diff --git a/fs/dlm/midcomms.h b/fs/dlm/midcomms.h index 9f8c9605013d0..e7246fb3ef577 100644 --- a/fs/dlm/midcomms.h +++ b/fs/dlm/midcomms.h @@ -20,6 +20,7 @@ struct dlm_mhandle *dlm_midcomms_get_mhandle(int nodeid, int len, gfp_t allocation, char **ppc); void dlm_midcomms_commit_mhandle(struct dlm_mhandle *mh, const void *name, int namelen); +int dlm_midcomms_addr(int nodeid, struct sockaddr_storage *addr, int len); void dlm_midcomms_version_wait(void); int dlm_midcomms_close(int nodeid); int dlm_midcomms_start(void); From a3d85fcf268ea40c024e864b219b72516236d15b Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Tue, 1 Aug 2023 14:09:50 -0400 Subject: [PATCH 059/186] fs: dlm: don't use RCOM_NAMES for version detection Currently RCOM_STATUS and RCOM_NAMES inclusive their replies are being used to determine the DLM version. The RCOM_NAMES messages are triggered in DLM recovery when calling dlm_recover_directory() only. At this time the DLM version need to be determined. I ran some tests and did not expirenced some issues. When the DLM version detection was developed probably I run once in a case of RCOM_NAMES and the version was not detected yet. However it seems to be not necessary. For backwards compatibility we still need to accept RCOM_NAMES messages which are not protected regarding the DLM message reliability layer aka stateless message. This patch changes that RCOM_NAMES we are sending out after this patch are not stateless anymore. Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/rcom.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c index 6ab029149a1d7..3b734aed26b54 100644 --- a/fs/dlm/rcom.c +++ b/fs/dlm/rcom.c @@ -308,15 +308,15 @@ static void receive_sync_reply(struct dlm_ls *ls, const struct dlm_rcom *rc_in) int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len, uint64_t seq) { + struct dlm_mhandle *mh; struct dlm_rcom *rc; - struct dlm_msg *msg; int error = 0; ls->ls_recover_nodeid = nodeid; retry: - error = create_rcom_stateless(ls, nodeid, DLM_RCOM_NAMES, last_len, - &rc, &msg, seq); + error = create_rcom(ls, nodeid, DLM_RCOM_NAMES, last_len, + &rc, &mh, seq); if (error) goto out; memcpy(rc->rc_buf, last_name, last_len); @@ -324,7 +324,7 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, allow_sync_reply(ls, &rc->rc_id); memset(ls->ls_recover_buf, 0, DLM_MAX_SOCKET_BUFSIZE); - send_rcom_stateless(msg, rc); + send_rcom(mh, rc); error = dlm_wait_function(ls, &rcom_response); disallow_sync_reply(ls); @@ -337,17 +337,17 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, static void receive_rcom_names(struct dlm_ls *ls, const struct dlm_rcom *rc_in, uint64_t seq) { + struct dlm_mhandle *mh; struct dlm_rcom *rc; int error, inlen, outlen, nodeid; - struct dlm_msg *msg; nodeid = le32_to_cpu(rc_in->rc_header.h_nodeid); inlen = le16_to_cpu(rc_in->rc_header.h_length) - sizeof(struct dlm_rcom); outlen = DLM_MAX_APP_BUFSIZE - sizeof(struct dlm_rcom); - error = create_rcom_stateless(ls, nodeid, DLM_RCOM_NAMES_REPLY, outlen, - &rc, &msg, seq); + error = create_rcom(ls, nodeid, DLM_RCOM_NAMES_REPLY, outlen, + &rc, &mh, seq); if (error) return; rc->rc_id = rc_in->rc_id; @@ -355,7 +355,7 @@ static void receive_rcom_names(struct dlm_ls *ls, const struct dlm_rcom *rc_in, dlm_copy_master_names(ls, rc_in->rc_buf, inlen, rc->rc_buf, outlen, nodeid); - send_rcom_stateless(msg, rc); + send_rcom(mh, rc); } int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid, uint64_t seq) From e6b51532d5273eeefba84106daea3d392c602837 Mon Sep 17 00:00:00 2001 From: Tomislav Novak Date: Tue, 20 Jun 2023 18:54:11 +0100 Subject: [PATCH 060/186] ARM: 9316/1: hw_breakpoint: fix single-stepping when using bpf_overflow_handler Arm platforms use is_default_overflow_handler() to determine if the hw_breakpoint code should single-step over the breakpoint trigger or let the custom handler deal with it. Since bpf_overflow_handler() currently isn't recognized as a default handler, attaching a BPF program to a PERF_TYPE_BREAKPOINT event causes it to keep firing (the instruction triggering the data abort exception is never skipped). For example: # bpftrace -e 'watchpoint:0x10000:4:w { print("hit") }' -c ./test Attaching 1 probe... hit hit [...] ^C (./test performs a single 4-byte store to 0x10000) This patch replaces the check with uses_default_overflow_handler(), which accounts for the bpf_overflow_handler() case by also testing if one of the perf_event_output functions gets invoked indirectly, via orig_default_handler. Link: https://lore.kernel.org/linux-arm-kernel/20220923203644.2731604-1-tnovak@fb.com/ Signed-off-by: Tomislav Novak Tested-by: Samuel Gosselin # arm64 Reviewed-by: Catalin Marinas Acked-by: Alexei Starovoitov Signed-off-by: Russell King (Oracle) --- arch/arm/kernel/hw_breakpoint.c | 8 ++++---- arch/arm64/kernel/hw_breakpoint.c | 4 ++-- include/linux/perf_event.h | 22 +++++++++++++++++++--- 3 files changed, 25 insertions(+), 9 deletions(-) diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index 054e9199f30db..dc0fb7a813715 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -626,7 +626,7 @@ int hw_breakpoint_arch_parse(struct perf_event *bp, hw->address &= ~alignment_mask; hw->ctrl.len <<= offset; - if (is_default_overflow_handler(bp)) { + if (uses_default_overflow_handler(bp)) { /* * Mismatch breakpoints are required for single-stepping * breakpoints. @@ -798,7 +798,7 @@ static void watchpoint_handler(unsigned long addr, unsigned int fsr, * Otherwise, insert a temporary mismatch breakpoint so that * we can single-step over the watchpoint trigger. */ - if (!is_default_overflow_handler(wp)) + if (!uses_default_overflow_handler(wp)) continue; step: enable_single_step(wp, instruction_pointer(regs)); @@ -811,7 +811,7 @@ static void watchpoint_handler(unsigned long addr, unsigned int fsr, info->trigger = addr; pr_debug("watchpoint fired: address = 0x%x\n", info->trigger); perf_bp_event(wp, regs); - if (is_default_overflow_handler(wp)) + if (uses_default_overflow_handler(wp)) enable_single_step(wp, instruction_pointer(regs)); } @@ -886,7 +886,7 @@ static void breakpoint_handler(unsigned long unknown, struct pt_regs *regs) info->trigger = addr; pr_debug("breakpoint fired: address = 0x%x\n", addr); perf_bp_event(bp, regs); - if (is_default_overflow_handler(bp)) + if (uses_default_overflow_handler(bp)) enable_single_step(bp, addr); goto unlock; } diff --git a/arch/arm64/kernel/hw_breakpoint.c b/arch/arm64/kernel/hw_breakpoint.c index db2a1861bb978..35225632d70ad 100644 --- a/arch/arm64/kernel/hw_breakpoint.c +++ b/arch/arm64/kernel/hw_breakpoint.c @@ -654,7 +654,7 @@ static int breakpoint_handler(unsigned long unused, unsigned long esr, perf_bp_event(bp, regs); /* Do we need to handle the stepping? */ - if (is_default_overflow_handler(bp)) + if (uses_default_overflow_handler(bp)) step = 1; unlock: rcu_read_unlock(); @@ -733,7 +733,7 @@ static u64 get_distance_from_watchpoint(unsigned long addr, u64 val, static int watchpoint_report(struct perf_event *wp, unsigned long addr, struct pt_regs *regs) { - int step = is_default_overflow_handler(wp); + int step = uses_default_overflow_handler(wp); struct arch_hw_breakpoint *info = counter_arch_bp(wp); info->trigger = addr; diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 2166a69e3bf2e..e657916c9509c 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1316,15 +1316,31 @@ extern int perf_event_output(struct perf_event *event, struct pt_regs *regs); static inline bool -is_default_overflow_handler(struct perf_event *event) +__is_default_overflow_handler(perf_overflow_handler_t overflow_handler) { - if (likely(event->overflow_handler == perf_event_output_forward)) + if (likely(overflow_handler == perf_event_output_forward)) return true; - if (unlikely(event->overflow_handler == perf_event_output_backward)) + if (unlikely(overflow_handler == perf_event_output_backward)) return true; return false; } +#define is_default_overflow_handler(event) \ + __is_default_overflow_handler((event)->overflow_handler) + +#ifdef CONFIG_BPF_SYSCALL +static inline bool uses_default_overflow_handler(struct perf_event *event) +{ + if (likely(is_default_overflow_handler(event))) + return true; + + return __is_default_overflow_handler(event->orig_overflow_handler); +} +#else +#define uses_default_overflow_handler(event) \ + is_default_overflow_handler(event) +#endif + extern void perf_event_header__init_id(struct perf_event_header *header, struct perf_sample_data *data, From 8922ba71c969d2a0c01a94372a71477d879470de Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A5rten=20Lindahl?= Date: Tue, 8 Aug 2023 09:37:32 +0100 Subject: [PATCH 061/186] ARM: 9317/1: kexec: Make smp stop calls asynchronous MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If a panic is triggered by a hrtimer interrupt all online cpus will be notified and set offline. But as highlighted by commit 19dbdcb8039c ("smp: Warn on function calls from softirq context") this call should not be made synchronous with disabled interrupts: softdog: Initiating panic Kernel panic - not syncing: Software Watchdog Timer expired WARNING: CPU: 1 PID: 0 at kernel/smp.c:753 smp_call_function_many_cond unwind_backtrace: show_stack dump_stack_lvl __warn warn_slowpath_fmt smp_call_function_many_cond smp_call_function crash_smp_send_stop.part.0 machine_crash_shutdown __crash_kexec panic softdog_fire __hrtimer_run_queues hrtimer_interrupt Make the smp call for machine_crash_nonpanic_core() asynchronous. Signed-off-by: Mårten Lindahl Signed-off-by: Russell King (Oracle) --- arch/arm/kernel/machine_kexec.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/arch/arm/kernel/machine_kexec.c b/arch/arm/kernel/machine_kexec.c index 46364b699cc30..5d07cf9e0044d 100644 --- a/arch/arm/kernel/machine_kexec.c +++ b/arch/arm/kernel/machine_kexec.c @@ -94,16 +94,28 @@ static void machine_crash_nonpanic_core(void *unused) } } +static DEFINE_PER_CPU(call_single_data_t, cpu_stop_csd) = + CSD_INIT(machine_crash_nonpanic_core, NULL); + void crash_smp_send_stop(void) { static int cpus_stopped; unsigned long msecs; + call_single_data_t *csd; + int cpu, this_cpu = raw_smp_processor_id(); if (cpus_stopped) return; atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1); - smp_call_function(machine_crash_nonpanic_core, NULL, false); + for_each_online_cpu(cpu) { + if (cpu == this_cpu) + continue; + + csd = &per_cpu(cpu_stop_csd, cpu); + smp_call_function_single_async(cpu, csd); + } + msecs = 1000; /* Wait at most a second for the other cpus to stop */ while ((atomic_read(&waiting_for_crash_ipi) > 0) && msecs) { mdelay(1); From b0a6da43a510fdfff23fcda12f90fba37cab1c05 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 9 Aug 2023 00:44:06 +0100 Subject: [PATCH 062/186] ARM: 9318/1: locomo: move kernel-doc to prevent warnings Move the kernel-doc comments for locomo_probe() to immediately before that function instead of before __locomo_probe() to prevent kernel-doc warnings. Adjust the documented function parameters and make the return values compatible with ReST by adding bullets to them. Add more possible return values to the list. Prevents these warnings: arch/arm/common/locomo.c:368: warning: Function parameter or member 'me' not described in '__locomo_probe' arch/arm/common/locomo.c:368: warning: Function parameter or member 'mem' not described in '__locomo_probe' arch/arm/common/locomo.c:368: warning: Function parameter or member 'irq' not described in '__locomo_probe' arch/arm/common/locomo.c:368: warning: expecting prototype for locomo_probe(). Prototype was for __locomo_probe() instead Link: lore.kernel.org/r/202308050851.zAvHe6y7-lkp@intel.com Fixes: 5eb6e280432d ("ARM: 9289/1: Allow pre-ARMv5 builds with ld.lld 16.0.0 and newer") Signed-off-by: Randy Dunlap Reported-by: kernel test robot Cc: Nick Desaulniers Cc: Arnd Bergmann Cc: Nathan Chancellor Cc: linux-arm-kernel@lists.infradead.org Cc: patches@armlinux.org.uk Signed-off-by: Russell King (Oracle) --- arch/arm/common/locomo.c | 28 +++++++++++++++------------- 1 file changed, 15 insertions(+), 13 deletions(-) diff --git a/arch/arm/common/locomo.c b/arch/arm/common/locomo.c index 309b747834684..70480dd9e96db 100644 --- a/arch/arm/common/locomo.c +++ b/arch/arm/common/locomo.c @@ -350,19 +350,6 @@ static int locomo_resume(struct platform_device *dev) } #endif - -/** - * locomo_probe - probe for a single LoCoMo chip. - * @phys_addr: physical address of device. - * - * Probe for a LoCoMo chip. This must be called - * before any other locomo-specific code. - * - * Returns: - * %-ENODEV device not found. - * %-EBUSY physical address already marked in-use. - * %0 successful. - */ static int __locomo_probe(struct device *me, struct resource *mem, int irq) { @@ -479,6 +466,21 @@ static void __locomo_remove(struct locomo *lchip) kfree(lchip); } +/** + * locomo_probe - probe for a single LoCoMo chip. + * @dev: platform device + * + * Probe for a LoCoMo chip. This must be called + * before any other locomo-specific code. + * + * Returns: + * * %-EINVAL - device's IORESOURCE_MEM not found + * * %-ENXIO - could not allocate an IRQ for the device + * * %-ENODEV - device not found. + * * %-EBUSY - physical address already marked in-use. + * * %-ENOMEM - could not allocate or iomap memory. + * * %0 - successful. + */ static int locomo_probe(struct platform_device *dev) { struct resource *mem; From 133789d4a458c761f60ef71cc5a6ede5a617ed7e Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Mon, 14 Aug 2023 23:22:51 +0100 Subject: [PATCH 063/186] Revert part of ae1f8d793a19 ("ARM: 9304/1: add prototype for function called only from asm") The build bot reports: >> arch/arm/vfp/vfpmodule.c:324:13: error: static declaration of 'VFP_bounce' follows non-static declaration static void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs) ^ arch/arm/include/asm/vfp.h:105:6: note: previous declaration is here void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs); ^ 1 error generated. This is due to a merge conflict between commit ae1f8d793a19 ("ARM: 9304/1: add prototype for function called only from asm") and Ard's commit 4708fb041346 ("ARM: vfp: Reimplement VFP exception entry in C code"). Fix this by removing Arnd's change. No need to backport. Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202308150547.m54XHV12-lkp@intel.com/ Signed-off-by: Russell King (Oracle) --- arch/arm/include/asm/vfp.h | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arm/include/asm/vfp.h b/arch/arm/include/asm/vfp.h index 5b57b8768bacc..157ea34261586 100644 --- a/arch/arm/include/asm/vfp.h +++ b/arch/arm/include/asm/vfp.h @@ -102,7 +102,6 @@ #ifndef __ASSEMBLY__ void vfp_disable(void); -void VFP_bounce(u32 trigger, u32 fpexc, struct pt_regs *regs); #endif #endif /* __ASM_VFP_H */ From 29a511e49f33426c8d24700db4842234a84678b2 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:35:59 +0800 Subject: [PATCH 064/186] jbd2: move load_superblock() dependent functions Move load_superblock() declaration and the functions it calls before journal_init_common(). This is a preparation for moving a call to load_superblock() from jbd2_journal_load() and jbd2_journal_wipe() to journal_init_common(). No functional changes. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-2-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 337 +++++++++++++++++++++++----------------------- 1 file changed, 168 insertions(+), 169 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 5c223032f77aa..c3f9689096184 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1336,6 +1336,174 @@ static unsigned long jbd2_journal_shrink_count(struct shrinker *shrink, return count; } +/* + * If the journal init or create aborts, we need to mark the journal + * superblock as being NULL to prevent the journal destroy from writing + * back a bogus superblock. + */ +static void journal_fail_superblock(journal_t *journal) +{ + struct buffer_head *bh = journal->j_sb_buffer; + brelse(bh); + journal->j_sb_buffer = NULL; +} + +/* + * Read the superblock for a given journal, performing initial + * validation of the format. + */ +static int journal_get_superblock(journal_t *journal) +{ + struct buffer_head *bh; + journal_superblock_t *sb; + int err; + + bh = journal->j_sb_buffer; + + J_ASSERT(bh != NULL); + if (buffer_verified(bh)) + return 0; + + err = bh_read(bh, 0); + if (err < 0) { + printk(KERN_ERR + "JBD2: IO error reading journal superblock\n"); + goto out; + } + + sb = journal->j_superblock; + + err = -EINVAL; + + if (sb->s_header.h_magic != cpu_to_be32(JBD2_MAGIC_NUMBER) || + sb->s_blocksize != cpu_to_be32(journal->j_blocksize)) { + printk(KERN_WARNING "JBD2: no valid journal superblock found\n"); + goto out; + } + + if (be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V1 && + be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V2) { + printk(KERN_WARNING "JBD2: unrecognised superblock format ID\n"); + goto out; + } + + if (be32_to_cpu(sb->s_maxlen) > journal->j_total_len) { + printk(KERN_WARNING "JBD2: journal file too short\n"); + goto out; + } + + if (be32_to_cpu(sb->s_first) == 0 || + be32_to_cpu(sb->s_first) >= journal->j_total_len) { + printk(KERN_WARNING + "JBD2: Invalid start block of journal: %u\n", + be32_to_cpu(sb->s_first)); + goto out; + } + + if (jbd2_has_feature_csum2(journal) && + jbd2_has_feature_csum3(journal)) { + /* Can't have checksum v2 and v3 at the same time! */ + printk(KERN_ERR "JBD2: Can't enable checksumming v2 and v3 " + "at the same time!\n"); + goto out; + } + + if (jbd2_journal_has_csum_v2or3_feature(journal) && + jbd2_has_feature_checksum(journal)) { + /* Can't have checksum v1 and v2 on at the same time! */ + printk(KERN_ERR "JBD2: Can't enable checksumming v1 and v2/3 " + "at the same time!\n"); + goto out; + } + + if (!jbd2_verify_csum_type(journal, sb)) { + printk(KERN_ERR "JBD2: Unknown checksum type\n"); + goto out; + } + + /* Load the checksum driver */ + if (jbd2_journal_has_csum_v2or3_feature(journal)) { + journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); + if (IS_ERR(journal->j_chksum_driver)) { + printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); + err = PTR_ERR(journal->j_chksum_driver); + journal->j_chksum_driver = NULL; + goto out; + } + /* Check superblock checksum */ + if (sb->s_checksum != jbd2_superblock_csum(journal, sb)) { + printk(KERN_ERR "JBD2: journal checksum error\n"); + err = -EFSBADCRC; + goto out; + } + } + set_buffer_verified(bh); + return 0; + +out: + journal_fail_superblock(journal); + return err; +} + +static int journal_revoke_records_per_block(journal_t *journal) +{ + int record_size; + int space = journal->j_blocksize - sizeof(jbd2_journal_revoke_header_t); + + if (jbd2_has_feature_64bit(journal)) + record_size = 8; + else + record_size = 4; + + if (jbd2_journal_has_csum_v2or3(journal)) + space -= sizeof(struct jbd2_journal_block_tail); + return space / record_size; +} + +/* + * Load the on-disk journal superblock and read the key fields into the + * journal_t. + */ +static int load_superblock(journal_t *journal) +{ + int err; + journal_superblock_t *sb; + int num_fc_blocks; + + err = journal_get_superblock(journal); + if (err) + return err; + + sb = journal->j_superblock; + + journal->j_tail_sequence = be32_to_cpu(sb->s_sequence); + journal->j_tail = be32_to_cpu(sb->s_start); + journal->j_first = be32_to_cpu(sb->s_first); + journal->j_errno = be32_to_cpu(sb->s_errno); + journal->j_last = be32_to_cpu(sb->s_maxlen); + + if (be32_to_cpu(sb->s_maxlen) < journal->j_total_len) + journal->j_total_len = be32_to_cpu(sb->s_maxlen); + /* Precompute checksum seed for all metadata */ + if (jbd2_journal_has_csum_v2or3(journal)) + journal->j_csum_seed = jbd2_chksum(journal, ~0, sb->s_uuid, + sizeof(sb->s_uuid)); + journal->j_revoke_records_per_block = + journal_revoke_records_per_block(journal); + + if (jbd2_has_feature_fast_commit(journal)) { + journal->j_fc_last = be32_to_cpu(sb->s_maxlen); + num_fc_blocks = jbd2_journal_get_num_fc_blks(sb); + if (journal->j_last - num_fc_blocks >= JBD2_MIN_JOURNAL_BLOCKS) + journal->j_last = journal->j_fc_last - num_fc_blocks; + journal->j_fc_first = journal->j_last + 1; + journal->j_fc_off = 0; + } + + return 0; +} + + /* * Management for journal control blocks: functions to create and * destroy journal_t structures, and to initialise and read existing @@ -1521,18 +1689,6 @@ journal_t *jbd2_journal_init_inode(struct inode *inode) return journal; } -/* - * If the journal init or create aborts, we need to mark the journal - * superblock as being NULL to prevent the journal destroy from writing - * back a bogus superblock. - */ -static void journal_fail_superblock(journal_t *journal) -{ - struct buffer_head *bh = journal->j_sb_buffer; - brelse(bh); - journal->j_sb_buffer = NULL; -} - /* * Given a journal_t structure, initialise the various fields for * startup of a new journaling session. We use this both when creating @@ -1889,163 +2045,6 @@ void jbd2_journal_update_sb_errno(journal_t *journal) } EXPORT_SYMBOL(jbd2_journal_update_sb_errno); -static int journal_revoke_records_per_block(journal_t *journal) -{ - int record_size; - int space = journal->j_blocksize - sizeof(jbd2_journal_revoke_header_t); - - if (jbd2_has_feature_64bit(journal)) - record_size = 8; - else - record_size = 4; - - if (jbd2_journal_has_csum_v2or3(journal)) - space -= sizeof(struct jbd2_journal_block_tail); - return space / record_size; -} - -/* - * Read the superblock for a given journal, performing initial - * validation of the format. - */ -static int journal_get_superblock(journal_t *journal) -{ - struct buffer_head *bh; - journal_superblock_t *sb; - int err; - - bh = journal->j_sb_buffer; - - J_ASSERT(bh != NULL); - if (buffer_verified(bh)) - return 0; - - err = bh_read(bh, 0); - if (err < 0) { - printk(KERN_ERR - "JBD2: IO error reading journal superblock\n"); - goto out; - } - - sb = journal->j_superblock; - - err = -EINVAL; - - if (sb->s_header.h_magic != cpu_to_be32(JBD2_MAGIC_NUMBER) || - sb->s_blocksize != cpu_to_be32(journal->j_blocksize)) { - printk(KERN_WARNING "JBD2: no valid journal superblock found\n"); - goto out; - } - - if (be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V1 && - be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V2) { - printk(KERN_WARNING "JBD2: unrecognised superblock format ID\n"); - goto out; - } - - if (be32_to_cpu(sb->s_maxlen) > journal->j_total_len) { - printk(KERN_WARNING "JBD2: journal file too short\n"); - goto out; - } - - if (be32_to_cpu(sb->s_first) == 0 || - be32_to_cpu(sb->s_first) >= journal->j_total_len) { - printk(KERN_WARNING - "JBD2: Invalid start block of journal: %u\n", - be32_to_cpu(sb->s_first)); - goto out; - } - - if (jbd2_has_feature_csum2(journal) && - jbd2_has_feature_csum3(journal)) { - /* Can't have checksum v2 and v3 at the same time! */ - printk(KERN_ERR "JBD2: Can't enable checksumming v2 and v3 " - "at the same time!\n"); - goto out; - } - - if (jbd2_journal_has_csum_v2or3_feature(journal) && - jbd2_has_feature_checksum(journal)) { - /* Can't have checksum v1 and v2 on at the same time! */ - printk(KERN_ERR "JBD2: Can't enable checksumming v1 and v2/3 " - "at the same time!\n"); - goto out; - } - - if (!jbd2_verify_csum_type(journal, sb)) { - printk(KERN_ERR "JBD2: Unknown checksum type\n"); - goto out; - } - - /* Load the checksum driver */ - if (jbd2_journal_has_csum_v2or3_feature(journal)) { - journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); - if (IS_ERR(journal->j_chksum_driver)) { - printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); - err = PTR_ERR(journal->j_chksum_driver); - journal->j_chksum_driver = NULL; - goto out; - } - /* Check superblock checksum */ - if (sb->s_checksum != jbd2_superblock_csum(journal, sb)) { - printk(KERN_ERR "JBD2: journal checksum error\n"); - err = -EFSBADCRC; - goto out; - } - } - set_buffer_verified(bh); - return 0; - -out: - journal_fail_superblock(journal); - return err; -} - -/* - * Load the on-disk journal superblock and read the key fields into the - * journal_t. - */ - -static int load_superblock(journal_t *journal) -{ - int err; - journal_superblock_t *sb; - int num_fc_blocks; - - err = journal_get_superblock(journal); - if (err) - return err; - - sb = journal->j_superblock; - - journal->j_tail_sequence = be32_to_cpu(sb->s_sequence); - journal->j_tail = be32_to_cpu(sb->s_start); - journal->j_first = be32_to_cpu(sb->s_first); - journal->j_errno = be32_to_cpu(sb->s_errno); - journal->j_last = be32_to_cpu(sb->s_maxlen); - - if (be32_to_cpu(sb->s_maxlen) < journal->j_total_len) - journal->j_total_len = be32_to_cpu(sb->s_maxlen); - /* Precompute checksum seed for all metadata */ - if (jbd2_journal_has_csum_v2or3(journal)) - journal->j_csum_seed = jbd2_chksum(journal, ~0, sb->s_uuid, - sizeof(sb->s_uuid)); - journal->j_revoke_records_per_block = - journal_revoke_records_per_block(journal); - - if (jbd2_has_feature_fast_commit(journal)) { - journal->j_fc_last = be32_to_cpu(sb->s_maxlen); - num_fc_blocks = jbd2_journal_get_num_fc_blks(sb); - if (journal->j_last - num_fc_blocks >= JBD2_MIN_JOURNAL_BLOCKS) - journal->j_last = journal->j_fc_last - num_fc_blocks; - journal->j_fc_first = journal->j_last + 1; - journal->j_fc_off = 0; - } - - return 0; -} - - /** * jbd2_journal_load() - Read journal from disk. * @journal: Journal to act on. From c30713084ba5b6fa343129613ec349ea91f0c458 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:00 +0800 Subject: [PATCH 065/186] jbd2: move load_superblock() into journal_init_common() Move the call to load_superblock() from jbd2_journal_load() and jbd2_journal_wipe() early into journal_init_common(), the journal superblock gets read and the in-memory journal_t structure gets initialised after calling jbd2_journal_init_{dev,inode}, it's safe to do following initialization according to it. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-3-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index c3f9689096184..98b43a9dcabe1 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1582,6 +1582,10 @@ static journal_t *journal_init_common(struct block_device *bdev, journal->j_sb_buffer = bh; journal->j_superblock = (journal_superblock_t *)bh->b_data; + err = load_superblock(journal); + if (err) + goto err_cleanup; + journal->j_shrink_transaction = NULL; journal->j_shrinker.scan_objects = jbd2_journal_shrink_scan; journal->j_shrinker.count_objects = jbd2_journal_shrink_count; @@ -2056,13 +2060,7 @@ EXPORT_SYMBOL(jbd2_journal_update_sb_errno); int jbd2_journal_load(journal_t *journal) { int err; - journal_superblock_t *sb; - - err = load_superblock(journal); - if (err) - return err; - - sb = journal->j_superblock; + journal_superblock_t *sb = journal->j_superblock; /* * If this is a V2 superblock, then we have to check the @@ -2523,10 +2521,6 @@ int jbd2_journal_wipe(journal_t *journal, int write) J_ASSERT (!(journal->j_flags & JBD2_LOADED)); - err = load_superblock(journal); - if (err) - return err; - if (!journal->j_tail) goto no_recovery; From 9600f3e5cfd0360b10c271149032c77917baedc5 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:01 +0800 Subject: [PATCH 066/186] jbd2: don't load superblock in jbd2_journal_check_used_features() Since load_superblock() has been moved to journal_init_common(), the in-memory superblock structure is initialized and contains valid data once the file system has a journal_t object, so it's safe to access it, let's drop the call to journal_get_superblock() from jbd2_journal_check_used_features() and also drop the setting/clearing of the veirfy bit of the superblock buffer. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-4-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 98b43a9dcabe1..95499b3184aaa 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1361,8 +1361,6 @@ static int journal_get_superblock(journal_t *journal) bh = journal->j_sb_buffer; J_ASSERT(bh != NULL); - if (buffer_verified(bh)) - return 0; err = bh_read(bh, 0); if (err < 0) { @@ -1437,7 +1435,6 @@ static int journal_get_superblock(journal_t *journal) goto out; } } - set_buffer_verified(bh); return 0; out: @@ -2226,8 +2223,6 @@ int jbd2_journal_check_used_features(journal_t *journal, unsigned long compat, if (!compat && !ro && !incompat) return 1; - if (journal_get_superblock(journal)) - return 0; if (!jbd2_format_support_feature(journal)) return 0; From e4adf8b837087b5bb57fff6827e10ec877a50f64 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:02 +0800 Subject: [PATCH 067/186] jbd2: checking valid features early in journal_get_superblock() journal_get_superblock() is used to check validity of the jounal supberblock, so move the features checks from jbd2_journal_load() to journal_get_superblock(). Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-5-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 95499b3184aaa..4d4494b42b39f 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1398,6 +1398,21 @@ static int journal_get_superblock(journal_t *journal) goto out; } + /* + * If this is a V2 superblock, then we have to check the + * features flags on it. + */ + if (!jbd2_format_support_feature(journal)) + return 0; + + if ((sb->s_feature_ro_compat & + ~cpu_to_be32(JBD2_KNOWN_ROCOMPAT_FEATURES)) || + (sb->s_feature_incompat & + ~cpu_to_be32(JBD2_KNOWN_INCOMPAT_FEATURES))) { + printk(KERN_WARNING "JBD2: Unrecognised features on journal\n"); + goto out; + } + if (jbd2_has_feature_csum2(journal) && jbd2_has_feature_csum3(journal)) { /* Can't have checksum v2 and v3 at the same time! */ @@ -2059,21 +2074,6 @@ int jbd2_journal_load(journal_t *journal) int err; journal_superblock_t *sb = journal->j_superblock; - /* - * If this is a V2 superblock, then we have to check the - * features flags on it. - */ - if (jbd2_format_support_feature(journal)) { - if ((sb->s_feature_ro_compat & - ~cpu_to_be32(JBD2_KNOWN_ROCOMPAT_FEATURES)) || - (sb->s_feature_incompat & - ~cpu_to_be32(JBD2_KNOWN_INCOMPAT_FEATURES))) { - printk(KERN_WARNING - "JBD2: Unrecognised features on journal\n"); - return -EINVAL; - } - } - /* * Create a slab for this blocksize */ From 18dad509e7bd3189ac1e7f7904faf1561a908871 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:03 +0800 Subject: [PATCH 068/186] jbd2: open code jbd2_verify_csum_type() helper jbd2_verify_csum_type() helper check checksum type in the superblock for v2 or v3 checksum feature, it always return true if these features are not enabled, and it has only one user, so open code it is more clear. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-6-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 18 +++++------------- 1 file changed, 5 insertions(+), 13 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 4d4494b42b39f..5d4744203e391 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -115,14 +115,6 @@ void __jbd2_debug(int level, const char *file, const char *func, #endif /* Checksumming functions */ -static int jbd2_verify_csum_type(journal_t *j, journal_superblock_t *sb) -{ - if (!jbd2_journal_has_csum_v2or3_feature(j)) - return 1; - - return sb->s_checksum_type == JBD2_CRC32C_CHKSUM; -} - static __be32 jbd2_superblock_csum(journal_t *j, journal_superblock_t *sb) { __u32 csum; @@ -1429,13 +1421,13 @@ static int journal_get_superblock(journal_t *journal) goto out; } - if (!jbd2_verify_csum_type(journal, sb)) { - printk(KERN_ERR "JBD2: Unknown checksum type\n"); - goto out; - } - /* Load the checksum driver */ if (jbd2_journal_has_csum_v2or3_feature(journal)) { + if (sb->s_checksum_type != JBD2_CRC32C_CHKSUM) { + printk(KERN_ERR "JBD2: Unknown checksum type\n"); + goto out; + } + journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); if (IS_ERR(journal->j_chksum_driver)) { printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); From 054d9c8fef14d476f1a9c6434de86813c5990052 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:04 +0800 Subject: [PATCH 069/186] jbd2: cleanup load_superblock() Rename load_superblock() to journal_load_superblock(), move getting and reading superblock from journal_init_common() and journal_get_superblock() to this function, and also rename journal_get_superblock() to journal_check_superblock(), make it a pure check helper to check superblock validity from disk. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-7-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 85 +++++++++++++++++++---------------------------- 1 file changed, 35 insertions(+), 50 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 5d4744203e391..89f9eb35323da 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1341,45 +1341,29 @@ static void journal_fail_superblock(journal_t *journal) } /* - * Read the superblock for a given journal, performing initial + * Check the superblock for a given journal, performing initial * validation of the format. */ -static int journal_get_superblock(journal_t *journal) +static int journal_check_superblock(journal_t *journal) { - struct buffer_head *bh; - journal_superblock_t *sb; - int err; - - bh = journal->j_sb_buffer; - - J_ASSERT(bh != NULL); - - err = bh_read(bh, 0); - if (err < 0) { - printk(KERN_ERR - "JBD2: IO error reading journal superblock\n"); - goto out; - } - - sb = journal->j_superblock; - - err = -EINVAL; + journal_superblock_t *sb = journal->j_superblock; + int err = -EINVAL; if (sb->s_header.h_magic != cpu_to_be32(JBD2_MAGIC_NUMBER) || sb->s_blocksize != cpu_to_be32(journal->j_blocksize)) { printk(KERN_WARNING "JBD2: no valid journal superblock found\n"); - goto out; + return err; } if (be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V1 && be32_to_cpu(sb->s_header.h_blocktype) != JBD2_SUPERBLOCK_V2) { printk(KERN_WARNING "JBD2: unrecognised superblock format ID\n"); - goto out; + return err; } if (be32_to_cpu(sb->s_maxlen) > journal->j_total_len) { printk(KERN_WARNING "JBD2: journal file too short\n"); - goto out; + return err; } if (be32_to_cpu(sb->s_first) == 0 || @@ -1387,7 +1371,7 @@ static int journal_get_superblock(journal_t *journal) printk(KERN_WARNING "JBD2: Invalid start block of journal: %u\n", be32_to_cpu(sb->s_first)); - goto out; + return err; } /* @@ -1402,7 +1386,7 @@ static int journal_get_superblock(journal_t *journal) (sb->s_feature_incompat & ~cpu_to_be32(JBD2_KNOWN_INCOMPAT_FEATURES))) { printk(KERN_WARNING "JBD2: Unrecognised features on journal\n"); - goto out; + return err; } if (jbd2_has_feature_csum2(journal) && @@ -1410,7 +1394,7 @@ static int journal_get_superblock(journal_t *journal) /* Can't have checksum v2 and v3 at the same time! */ printk(KERN_ERR "JBD2: Can't enable checksumming v2 and v3 " "at the same time!\n"); - goto out; + return err; } if (jbd2_journal_has_csum_v2or3_feature(journal) && @@ -1418,14 +1402,14 @@ static int journal_get_superblock(journal_t *journal) /* Can't have checksum v1 and v2 on at the same time! */ printk(KERN_ERR "JBD2: Can't enable checksumming v1 and v2/3 " "at the same time!\n"); - goto out; + return err; } /* Load the checksum driver */ if (jbd2_journal_has_csum_v2or3_feature(journal)) { if (sb->s_checksum_type != JBD2_CRC32C_CHKSUM) { printk(KERN_ERR "JBD2: Unknown checksum type\n"); - goto out; + return err; } journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0); @@ -1433,20 +1417,17 @@ static int journal_get_superblock(journal_t *journal) printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n"); err = PTR_ERR(journal->j_chksum_driver); journal->j_chksum_driver = NULL; - goto out; + return err; } /* Check superblock checksum */ if (sb->s_checksum != jbd2_superblock_csum(journal, sb)) { printk(KERN_ERR "JBD2: journal checksum error\n"); err = -EFSBADCRC; - goto out; + return err; } } - return 0; -out: - journal_fail_superblock(journal); - return err; + return 0; } static int journal_revoke_records_per_block(journal_t *journal) @@ -1468,17 +1449,31 @@ static int journal_revoke_records_per_block(journal_t *journal) * Load the on-disk journal superblock and read the key fields into the * journal_t. */ -static int load_superblock(journal_t *journal) +static int journal_load_superblock(journal_t *journal) { int err; + struct buffer_head *bh; journal_superblock_t *sb; int num_fc_blocks; - err = journal_get_superblock(journal); - if (err) - return err; + bh = getblk_unmovable(journal->j_dev, journal->j_blk_offset, + journal->j_blocksize); + if (bh) + err = bh_read(bh, 0); + if (!bh || err < 0) { + pr_err("%s: Cannot read journal superblock\n", __func__); + brelse(bh); + return -EIO; + } - sb = journal->j_superblock; + journal->j_sb_buffer = bh; + sb = (journal_superblock_t *)bh->b_data; + journal->j_superblock = sb; + err = journal_check_superblock(journal); + if (err) { + journal_fail_superblock(journal); + return err; + } journal->j_tail_sequence = be32_to_cpu(sb->s_sequence); journal->j_tail = be32_to_cpu(sb->s_start); @@ -1524,7 +1519,6 @@ static journal_t *journal_init_common(struct block_device *bdev, static struct lock_class_key jbd2_trans_commit_key; journal_t *journal; int err; - struct buffer_head *bh; int n; journal = kzalloc(sizeof(*journal), GFP_KERNEL); @@ -1577,16 +1571,7 @@ static journal_t *journal_init_common(struct block_device *bdev, if (!journal->j_wbuf) goto err_cleanup; - bh = getblk_unmovable(journal->j_dev, start, journal->j_blocksize); - if (!bh) { - pr_err("%s: Cannot get buffer for journal superblock\n", - __func__); - goto err_cleanup; - } - journal->j_sb_buffer = bh; - journal->j_superblock = (journal_superblock_t *)bh->b_data; - - err = load_superblock(journal); + err = journal_load_superblock(journal); if (err) goto err_cleanup; From 0dbc759ae9971568af24def1b01d5b1aa87bd546 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:05 +0800 Subject: [PATCH 070/186] jbd2: add fast_commit space check If JBD2_FEATURE_INCOMPAT_FAST_COMMIT bit is set, it means the journal have fast commit records need to recover, so the fast commit size should not be too large, and the leftover normal journal size should never less than JBD2_MIN_JOURNAL_BLOCKS. If it happens, the journal->j_last is likely to be wrong and will probably lead to incorrect journal recovery. So add a check into the journal_check_superblock(), and drop the pointless check when initializing the fastcommit parameters. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-8-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 89f9eb35323da..ef9d75cca3620 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1347,6 +1347,7 @@ static void journal_fail_superblock(journal_t *journal) static int journal_check_superblock(journal_t *journal) { journal_superblock_t *sb = journal->j_superblock; + int num_fc_blks; int err = -EINVAL; if (sb->s_header.h_magic != cpu_to_be32(JBD2_MAGIC_NUMBER) || @@ -1389,6 +1390,15 @@ static int journal_check_superblock(journal_t *journal) return err; } + num_fc_blks = jbd2_has_feature_fast_commit(journal) ? + jbd2_journal_get_num_fc_blks(sb) : 0; + if (be32_to_cpu(sb->s_maxlen) < JBD2_MIN_JOURNAL_BLOCKS || + be32_to_cpu(sb->s_maxlen) - JBD2_MIN_JOURNAL_BLOCKS < num_fc_blks) { + printk(KERN_ERR "JBD2: journal file too short %u,%d\n", + be32_to_cpu(sb->s_maxlen), num_fc_blks); + return err; + } + if (jbd2_has_feature_csum2(journal) && jbd2_has_feature_csum3(journal)) { /* Can't have checksum v2 and v3 at the same time! */ @@ -1454,7 +1464,6 @@ static int journal_load_superblock(journal_t *journal) int err; struct buffer_head *bh; journal_superblock_t *sb; - int num_fc_blocks; bh = getblk_unmovable(journal->j_dev, journal->j_blk_offset, journal->j_blocksize); @@ -1492,9 +1501,8 @@ static int journal_load_superblock(journal_t *journal) if (jbd2_has_feature_fast_commit(journal)) { journal->j_fc_last = be32_to_cpu(sb->s_maxlen); - num_fc_blocks = jbd2_journal_get_num_fc_blks(sb); - if (journal->j_last - num_fc_blocks >= JBD2_MIN_JOURNAL_BLOCKS) - journal->j_last = journal->j_fc_last - num_fc_blocks; + journal->j_last = journal->j_fc_last - + jbd2_journal_get_num_fc_blks(sb); journal->j_fc_first = journal->j_last + 1; journal->j_fc_off = 0; } From 49887e47a5262cc7b87d547de57a21a072c6ea5e Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:06 +0800 Subject: [PATCH 071/186] jbd2: cleanup journal_init_common() Adjust the initialization sequence and error handle of journal_t, moving load superblock to the begin, and classify others initialization. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-9-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 45 ++++++++++++++++++++++++--------------------- 1 file changed, 24 insertions(+), 21 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index ef9d75cca3620..04b67844118c9 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1533,6 +1533,16 @@ static journal_t *journal_init_common(struct block_device *bdev, if (!journal) return NULL; + journal->j_blocksize = blocksize; + journal->j_dev = bdev; + journal->j_fs_dev = fs_dev; + journal->j_blk_offset = start; + journal->j_total_len = len; + + err = journal_load_superblock(journal); + if (err) + goto err_cleanup; + init_waitqueue_head(&journal->j_wait_transaction_locked); init_waitqueue_head(&journal->j_wait_done_commit); init_waitqueue_head(&journal->j_wait_commit); @@ -1544,12 +1554,15 @@ static journal_t *journal_init_common(struct block_device *bdev, mutex_init(&journal->j_checkpoint_mutex); spin_lock_init(&journal->j_revoke_lock); spin_lock_init(&journal->j_list_lock); + spin_lock_init(&journal->j_history_lock); rwlock_init(&journal->j_state_lock); journal->j_commit_interval = (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE); journal->j_min_batch_time = 0; journal->j_max_batch_time = 15000; /* 15ms */ atomic_set(&journal->j_reserved_credits, 0); + lockdep_init_map(&journal->j_trans_commit_map, "jbd2_handle", + &jbd2_trans_commit_key, 0); /* The journal is marked for error until we succeed with recovery! */ journal->j_flags = JBD2_ABORT; @@ -1559,18 +1572,10 @@ static journal_t *journal_init_common(struct block_device *bdev, if (err) goto err_cleanup; - spin_lock_init(&journal->j_history_lock); - - lockdep_init_map(&journal->j_trans_commit_map, "jbd2_handle", - &jbd2_trans_commit_key, 0); - - /* journal descriptor can store up to n blocks -bzzz */ - journal->j_blocksize = blocksize; - journal->j_dev = bdev; - journal->j_fs_dev = fs_dev; - journal->j_blk_offset = start; - journal->j_total_len = len; - /* We need enough buffers to write out full descriptor block. */ + /* + * journal descriptor can store up to n blocks, we need enough + * buffers to write out full descriptor block. + */ n = journal->j_blocksize / jbd2_min_tag_size(); journal->j_wbufsize = n; journal->j_fc_wbuf = NULL; @@ -1579,7 +1584,8 @@ static journal_t *journal_init_common(struct block_device *bdev, if (!journal->j_wbuf) goto err_cleanup; - err = journal_load_superblock(journal); + err = percpu_counter_init(&journal->j_checkpoint_jh_count, 0, + GFP_KERNEL); if (err) goto err_cleanup; @@ -1588,21 +1594,18 @@ static journal_t *journal_init_common(struct block_device *bdev, journal->j_shrinker.count_objects = jbd2_journal_shrink_count; journal->j_shrinker.seeks = DEFAULT_SEEKS; journal->j_shrinker.batch = journal->j_max_transaction_buffers; - - if (percpu_counter_init(&journal->j_checkpoint_jh_count, 0, GFP_KERNEL)) + err = register_shrinker(&journal->j_shrinker, "jbd2-journal:(%u:%u)", + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev)); + if (err) goto err_cleanup; - if (register_shrinker(&journal->j_shrinker, "jbd2-journal:(%u:%u)", - MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev))) { - percpu_counter_destroy(&journal->j_checkpoint_jh_count); - goto err_cleanup; - } return journal; err_cleanup: - brelse(journal->j_sb_buffer); + percpu_counter_destroy(&journal->j_checkpoint_jh_count); kfree(journal->j_wbuf); jbd2_journal_destroy_revoke(journal); + journal_fail_superblock(journal); kfree(journal); return NULL; } From d9a45496019a73c240bd22912ae18a04b8496364 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:07 +0800 Subject: [PATCH 072/186] jbd2: drop useless error tag in jbd2_journal_wipe() no_recovery is redundant, just drop it. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-10-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 04b67844118c9..6482fcca3dc67 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2500,12 +2500,12 @@ int jbd2_journal_flush(journal_t *journal, unsigned int flags) int jbd2_journal_wipe(journal_t *journal, int write) { - int err = 0; + int err; J_ASSERT (!(journal->j_flags & JBD2_LOADED)); if (!journal->j_tail) - goto no_recovery; + return 0; printk(KERN_WARNING "JBD2: %s recovery information on journal\n", write ? "Clearing" : "Ignoring"); @@ -2518,7 +2518,6 @@ int jbd2_journal_wipe(journal_t *journal, int write) mutex_unlock(&journal->j_checkpoint_mutex); } - no_recovery: return err; } From 8e6cf5fbb7b47d337998faab3fcacdceaa547ead Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:08 +0800 Subject: [PATCH 073/186] jbd2: jbd2_journal_init_{dev,inode} return proper error return value Current jbd2_journal_init_{dev,inode} return NULL if some error happens, make them to pass out proper error return value. [ Fix from Yang Yingliang folded in. ] Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-11-yi.zhang@huaweicloud.com Link: https://lore.kernel.org/r/20230822030018.644419-1-yangyingliang@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 4 ++-- fs/jbd2/journal.c | 19 +++++++++---------- fs/ocfs2/journal.c | 8 ++++---- 3 files changed, 15 insertions(+), 16 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 4613264344b07..279e37c3b2758 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5815,7 +5815,7 @@ static journal_t *ext4_get_journal(struct super_block *sb, return NULL; journal = jbd2_journal_init_inode(journal_inode); - if (!journal) { + if (IS_ERR(journal)) { ext4_msg(sb, KERN_ERR, "Could not load journal inode"); iput(journal_inode); return NULL; @@ -5894,7 +5894,7 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, journal = jbd2_journal_init_dev(bdev, sb->s_bdev, start, len, blocksize); - if (!journal) { + if (IS_ERR(journal)) { ext4_msg(sb, KERN_ERR, "failed to create device journal"); goto out_bdev; } diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 6482fcca3dc67..15e33c26c6cd7 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1531,7 +1531,7 @@ static journal_t *journal_init_common(struct block_device *bdev, journal = kzalloc(sizeof(*journal), GFP_KERNEL); if (!journal) - return NULL; + return ERR_PTR(-ENOMEM); journal->j_blocksize = blocksize; journal->j_dev = bdev; @@ -1576,6 +1576,7 @@ static journal_t *journal_init_common(struct block_device *bdev, * journal descriptor can store up to n blocks, we need enough * buffers to write out full descriptor block. */ + err = -ENOMEM; n = journal->j_blocksize / jbd2_min_tag_size(); journal->j_wbufsize = n; journal->j_fc_wbuf = NULL; @@ -1607,7 +1608,7 @@ static journal_t *journal_init_common(struct block_device *bdev, jbd2_journal_destroy_revoke(journal); journal_fail_superblock(journal); kfree(journal); - return NULL; + return ERR_PTR(err); } /* jbd2_journal_init_dev and jbd2_journal_init_inode: @@ -1640,8 +1641,8 @@ journal_t *jbd2_journal_init_dev(struct block_device *bdev, journal_t *journal; journal = journal_init_common(bdev, fs_dev, start, len, blocksize); - if (!journal) - return NULL; + if (IS_ERR(journal)) + return ERR_CAST(journal); snprintf(journal->j_devname, sizeof(journal->j_devname), "%pg", journal->j_dev); @@ -1667,11 +1668,9 @@ journal_t *jbd2_journal_init_inode(struct inode *inode) blocknr = 0; err = bmap(inode, &blocknr); - if (err || !blocknr) { - pr_err("%s: Cannot locate journal superblock\n", - __func__); - return NULL; + pr_err("%s: Cannot locate journal superblock\n", __func__); + return err ? ERR_PTR(err) : ERR_PTR(-EINVAL); } jbd2_debug(1, "JBD2: inode %s/%ld, size %lld, bits %d, blksize %ld\n", @@ -1681,8 +1680,8 @@ journal_t *jbd2_journal_init_inode(struct inode *inode) journal = journal_init_common(inode->i_sb->s_bdev, inode->i_sb->s_bdev, blocknr, inode->i_size >> inode->i_sb->s_blocksize_bits, inode->i_sb->s_blocksize); - if (!journal) - return NULL; + if (IS_ERR(journal)) + return ERR_CAST(journal); journal->j_inode = inode; snprintf(journal->j_devname, sizeof(journal->j_devname), diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 25d8072ccfce4..1d2960e8ce745 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -911,9 +911,9 @@ int ocfs2_journal_init(struct ocfs2_super *osb, int *dirty) /* call the kernels journal init function now */ j_journal = jbd2_journal_init_inode(inode); - if (j_journal == NULL) { + if (IS_ERR(j_journal)) { mlog(ML_ERROR, "Linux journal layer error\n"); - status = -EINVAL; + status = PTR_ERR(j_journal); goto done; } @@ -1687,9 +1687,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb, } journal = jbd2_journal_init_inode(inode); - if (journal == NULL) { + if (IS_ERR(journal)) { mlog(ML_ERROR, "Linux journal layer error\n"); - status = -EIO; + status = PTR_ERR(journal); goto done; } From bc74e6a38d16d745a9bc28a7e343494019066492 Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:09 +0800 Subject: [PATCH 074/186] ext4: cleanup ext4_get_dev_journal() and ext4_get_journal() Factor out a new helper form ext4_get_dev_journal() to get external journal bdev and check validation of this device, drop ext4_blkdev_get() helper, and also remove duplicate check of journal feature. It makes ext4_get_dev_journal() more clear than before. Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-12-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 109 ++++++++++++++++++++++-------------------------- 1 file changed, 49 insertions(+), 60 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 279e37c3b2758..0eee238b290e0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1105,26 +1105,6 @@ static const struct blk_holder_ops ext4_holder_ops = { .mark_dead = ext4_bdev_mark_dead, }; -/* - * Open the external journal device - */ -static struct block_device *ext4_blkdev_get(dev_t dev, struct super_block *sb) -{ - struct block_device *bdev; - - bdev = blkdev_get_by_dev(dev, BLK_OPEN_READ | BLK_OPEN_WRITE, sb, - &ext4_holder_ops); - if (IS_ERR(bdev)) - goto fail; - return bdev; - -fail: - ext4_msg(sb, KERN_ERR, - "failed to open journal device unknown-block(%u,%u) %ld", - MAJOR(dev), MINOR(dev), PTR_ERR(bdev)); - return NULL; -} - /* * Release the journal device */ @@ -5768,14 +5748,14 @@ static struct inode *ext4_get_journal_inode(struct super_block *sb, ext4_msg(sb, KERN_ERR, "journal inode is deleted"); return NULL; } - - ext4_debug("Journal inode found at %p: %lld bytes\n", - journal_inode, journal_inode->i_size); if (!S_ISREG(journal_inode->i_mode) || IS_ENCRYPTED(journal_inode)) { ext4_msg(sb, KERN_ERR, "invalid journal inode"); iput(journal_inode); return NULL; } + + ext4_debug("Journal inode found at %p: %lld bytes\n", + journal_inode, journal_inode->i_size); return journal_inode; } @@ -5807,9 +5787,6 @@ static journal_t *ext4_get_journal(struct super_block *sb, struct inode *journal_inode; journal_t *journal; - if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) - return NULL; - journal_inode = ext4_get_journal_inode(sb, journal_inum); if (!journal_inode) return NULL; @@ -5826,25 +5803,25 @@ static journal_t *ext4_get_journal(struct super_block *sb, return journal; } -static journal_t *ext4_get_dev_journal(struct super_block *sb, - dev_t j_dev) +static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, + dev_t j_dev, ext4_fsblk_t *j_start, + ext4_fsblk_t *j_len) { struct buffer_head *bh; - journal_t *journal; - ext4_fsblk_t start; - ext4_fsblk_t len; + struct block_device *bdev; int hblock, blocksize; ext4_fsblk_t sb_block; unsigned long offset; struct ext4_super_block *es; - struct block_device *bdev; - if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) - return NULL; - - bdev = ext4_blkdev_get(j_dev, sb); - if (bdev == NULL) + bdev = blkdev_get_by_dev(j_dev, BLK_OPEN_READ | BLK_OPEN_WRITE, sb, + &ext4_holder_ops); + if (IS_ERR(bdev)) { + ext4_msg(sb, KERN_ERR, + "failed to open journal device unknown-block(%u,%u) %ld", + MAJOR(j_dev), MINOR(j_dev), PTR_ERR(bdev)); return NULL; + } blocksize = sb->s_blocksize; hblock = bdev_logical_block_size(bdev); @@ -5857,7 +5834,8 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, sb_block = EXT4_MIN_BLOCK_SIZE / blocksize; offset = EXT4_MIN_BLOCK_SIZE % blocksize; set_blocksize(bdev, blocksize); - if (!(bh = __bread(bdev, sb_block, blocksize))) { + bh = __bread(bdev, sb_block, blocksize); + if (!bh) { ext4_msg(sb, KERN_ERR, "couldn't read superblock of " "external journal"); goto out_bdev; @@ -5867,56 +5845,67 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, if ((le16_to_cpu(es->s_magic) != EXT4_SUPER_MAGIC) || !(le32_to_cpu(es->s_feature_incompat) & EXT4_FEATURE_INCOMPAT_JOURNAL_DEV)) { - ext4_msg(sb, KERN_ERR, "external journal has " - "bad superblock"); - brelse(bh); - goto out_bdev; + ext4_msg(sb, KERN_ERR, "external journal has bad superblock"); + goto out_bh; } if ((le32_to_cpu(es->s_feature_ro_compat) & EXT4_FEATURE_RO_COMPAT_METADATA_CSUM) && es->s_checksum != ext4_superblock_csum(sb, es)) { - ext4_msg(sb, KERN_ERR, "external journal has " - "corrupt superblock"); - brelse(bh); - goto out_bdev; + ext4_msg(sb, KERN_ERR, "external journal has corrupt superblock"); + goto out_bh; } if (memcmp(EXT4_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) { ext4_msg(sb, KERN_ERR, "journal UUID does not match"); - brelse(bh); - goto out_bdev; + goto out_bh; } - len = ext4_blocks_count(es); - start = sb_block + 1; - brelse(bh); /* we're done with the superblock */ + *j_start = sb_block + 1; + *j_len = ext4_blocks_count(es); + brelse(bh); + return bdev; + +out_bh: + brelse(bh); +out_bdev: + blkdev_put(bdev, sb); + return NULL; +} + +static journal_t *ext4_get_dev_journal(struct super_block *sb, + dev_t j_dev) +{ + journal_t *journal; + ext4_fsblk_t j_start; + ext4_fsblk_t j_len; + struct block_device *journal_bdev; + + journal_bdev = ext4_get_journal_blkdev(sb, j_dev, &j_start, &j_len); + if (!journal_bdev) + return NULL; - journal = jbd2_journal_init_dev(bdev, sb->s_bdev, - start, len, blocksize); + journal = jbd2_journal_init_dev(journal_bdev, sb->s_bdev, j_start, + j_len, sb->s_blocksize); if (IS_ERR(journal)) { ext4_msg(sb, KERN_ERR, "failed to create device journal"); goto out_bdev; } - journal->j_private = sb; - if (ext4_read_bh_lock(journal->j_sb_buffer, REQ_META | REQ_PRIO, true)) { - ext4_msg(sb, KERN_ERR, "I/O error on journal device"); - goto out_journal; - } if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) { ext4_msg(sb, KERN_ERR, "External journal has more than one " "user (unsupported) - %d", be32_to_cpu(journal->j_superblock->s_nr_users)); goto out_journal; } - EXT4_SB(sb)->s_journal_bdev = bdev; + journal->j_private = sb; + EXT4_SB(sb)->s_journal_bdev = journal_bdev; ext4_init_journal_params(sb, journal); return journal; out_journal: jbd2_journal_destroy(journal); out_bdev: - blkdev_put(bdev, sb); + blkdev_put(journal_bdev, sb); return NULL; } From bb05a617f06b7a882e19c4f475b8e37f14d9ceac Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Wed, 24 May 2023 17:27:08 -0400 Subject: [PATCH 075/186] NFSv4.2: Fix READ_PLUS smatch warnings Smatch reports: fs/nfs/nfs42xdr.c:1131 decode_read_plus() warn: missing error code? 'status' Which Dan suggests to fix by doing a hardcoded "return 0" from the "if (segments == 0)" check. Additionally, smatch reports that the "status = -EIO" assignment is not used. This patch addresses both these issues. Reported-by: kernel test robot Reported-by: Dan Carpenter Closes: https://lore.kernel.org/r/202305222209.6l5VM2lL-lkp@intel.com/ Fixes: d3b00a802c845 ("NFS: Replace the READ_PLUS decoding code") Signed-off-by: Anna Schumaker --- fs/nfs/nfs42xdr.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 95234208dc9ee..d0919c5bf61c7 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -1056,13 +1056,12 @@ static int decode_read_plus(struct xdr_stream *xdr, struct nfs_pgio_res *res) res->eof = be32_to_cpup(p++); segments = be32_to_cpup(p++); if (segments == 0) - return status; + return 0; segs = kmalloc_array(segments, sizeof(*segs), GFP_KERNEL); if (!segs) return -ENOMEM; - status = -EIO; for (i = 0; i < segments; i++) { status = decode_read_plus_segment(xdr, &segs[i]); if (status < 0) From 8d18f6c5bb864d97a730f471c56cdecf313efe64 Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Wed, 31 May 2023 17:02:54 -0400 Subject: [PATCH 076/186] NFSv4.2: Fix READ_PLUS size calculations I bump the decode_read_plus_maxsz to account for hole segments, but I need to subtract out this increase when calling rpc_prepare_reply_pages() so the common case of single data segment replies can be directly placed into the xdr pages without needing to be shifted around. Reported-by: Chuck Lever Fixes: d3b00a802c845 ("NFS: Replace the READ_PLUS decoding code") Signed-off-by: Anna Schumaker --- fs/nfs/nfs42xdr.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index d0919c5bf61c7..78193f04d8928 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -54,10 +54,16 @@ (1 /* data_content4 */ + \ 2 /* data_info4.di_offset */ + \ 1 /* data_info4.di_length */) +#define NFS42_READ_PLUS_HOLE_SEGMENT_SIZE \ + (1 /* data_content4 */ + \ + 2 /* data_info4.di_offset */ + \ + 2 /* data_info4.di_length */) +#define READ_PLUS_SEGMENT_SIZE_DIFF (NFS42_READ_PLUS_HOLE_SEGMENT_SIZE - \ + NFS42_READ_PLUS_DATA_SEGMENT_SIZE) #define decode_read_plus_maxsz (op_decode_hdr_maxsz + \ 1 /* rpr_eof */ + \ 1 /* rpr_contents count */ + \ - NFS42_READ_PLUS_DATA_SEGMENT_SIZE) + NFS42_READ_PLUS_HOLE_SEGMENT_SIZE) #define encode_seek_maxsz (op_encode_hdr_maxsz + \ encode_stateid_maxsz + \ 2 /* offset */ + \ @@ -617,8 +623,8 @@ static void nfs4_xdr_enc_read_plus(struct rpc_rqst *req, encode_putfh(xdr, args->fh, &hdr); encode_read_plus(xdr, args, &hdr); - rpc_prepare_reply_pages(req, args->pages, args->pgbase, - args->count, hdr.replen); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, args->count, + hdr.replen - READ_PLUS_SEGMENT_SIZE_DIFF); encode_nops(&hdr); } From 303a78052091c81e9003915c521fdca1c7e117af Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Fri, 9 Jun 2023 15:26:25 -0400 Subject: [PATCH 077/186] NFSv4.2: Rework scratch handling for READ_PLUS (again) I found that the read code might send multiple requests using the same nfs_pgio_header, but nfs4_proc_read_setup() is only called once. This is how we ended up occasionally double-freeing the scratch buffer, but also means we set a NULL pointer but non-zero length to the xdr scratch buffer. This results in an oops the first time decoding needs to copy something to scratch, which frequently happens when decoding READ_PLUS hole segments. I fix this by moving scratch handling into the pageio read code. I provide a function to allocate scratch space for decoding read replies, and free the scratch buffer when the nfs_pgio_header is freed. Fixes: fbd2a05f29a9 (NFSv4.2: Rework scratch handling for READ_PLUS) Signed-off-by: Anna Schumaker --- fs/nfs/internal.h | 1 + fs/nfs/nfs42.h | 1 + fs/nfs/nfs42xdr.c | 2 +- fs/nfs/nfs4proc.c | 13 +------------ fs/nfs/read.c | 10 ++++++++++ 5 files changed, 14 insertions(+), 13 deletions(-) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 913c09806c7f5..41abea340ad84 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -493,6 +493,7 @@ extern const struct nfs_pgio_completion_ops nfs_async_read_completion_ops; extern void nfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode, bool force_mds, const struct nfs_pgio_completion_ops *compl_ops); +extern bool nfs_read_alloc_scratch(struct nfs_pgio_header *hdr, size_t size); extern int nfs_read_add_folio(struct nfs_pageio_descriptor *pgio, struct nfs_open_context *ctx, struct folio *folio); diff --git a/fs/nfs/nfs42.h b/fs/nfs/nfs42.h index 0fe5aacbcfdf1..b59876b01a1e3 100644 --- a/fs/nfs/nfs42.h +++ b/fs/nfs/nfs42.h @@ -13,6 +13,7 @@ * more? Need to consider not to pre-alloc too much for a compound. */ #define PNFS_LAYOUTSTATS_MAXDEV (4) +#define READ_PLUS_SCRATCH_SIZE (16) /* nfs4.2proc.c */ #ifdef CONFIG_NFS_V4_2 diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 78193f04d8928..9e3ae53e22058 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -1433,7 +1433,7 @@ static int nfs4_xdr_dec_read_plus(struct rpc_rqst *rqstp, struct compound_hdr hdr; int status; - xdr_set_scratch_buffer(xdr, res->scratch, sizeof(res->scratch)); + xdr_set_scratch_buffer(xdr, res->scratch, READ_PLUS_SCRATCH_SIZE); status = decode_compound_hdr(xdr, &hdr); if (status) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 832fa226b8f26..3c24c3c99e8ac 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -5438,18 +5438,8 @@ static bool nfs4_read_plus_not_supported(struct rpc_task *task, return false; } -static inline void nfs4_read_plus_scratch_free(struct nfs_pgio_header *hdr) -{ - if (hdr->res.scratch) { - kfree(hdr->res.scratch); - hdr->res.scratch = NULL; - } -} - static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr) { - nfs4_read_plus_scratch_free(hdr); - if (!nfs4_sequence_done(task, &hdr->res.seq_res)) return -EAGAIN; if (nfs4_read_stateid_changed(task, &hdr->args)) @@ -5469,8 +5459,7 @@ static bool nfs42_read_plus_support(struct nfs_pgio_header *hdr, /* Note: We don't use READ_PLUS with pNFS yet */ if (nfs_server_capable(hdr->inode, NFS_CAP_READ_PLUS) && !hdr->ds_clp) { msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ_PLUS]; - hdr->res.scratch = kmalloc(32, GFP_KERNEL); - return hdr->res.scratch != NULL; + return nfs_read_alloc_scratch(hdr, READ_PLUS_SCRATCH_SIZE); } return false; } diff --git a/fs/nfs/read.c b/fs/nfs/read.c index f71eeee67e201..7dc21a48e3e7b 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -47,6 +47,8 @@ static struct nfs_pgio_header *nfs_readhdr_alloc(void) static void nfs_readhdr_free(struct nfs_pgio_header *rhdr) { + if (rhdr->res.scratch != NULL) + kfree(rhdr->res.scratch); kmem_cache_free(nfs_rdata_cachep, rhdr); } @@ -108,6 +110,14 @@ void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio) } EXPORT_SYMBOL_GPL(nfs_pageio_reset_read_mds); +bool nfs_read_alloc_scratch(struct nfs_pgio_header *hdr, size_t size) +{ + WARN_ON(hdr->res.scratch != NULL); + hdr->res.scratch = kmalloc(size, GFP_KERNEL); + return hdr->res.scratch != NULL; +} +EXPORT_SYMBOL_GPL(nfs_read_alloc_scratch); + static void nfs_readpage_release(struct nfs_page *req, int error) { struct folio *folio = nfs_page_to_folio(req); From 61182c796d74f54ba66d17bac6f516183ec09af2 Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Fri, 23 Jun 2023 11:43:14 -0400 Subject: [PATCH 078/186] SUNRPC: kmap() the xdr pages during decode If the pages are in HIGHMEM then we need to make sure they're mapped before trying to read data off of them, otherwise we could end up with a NULL pointer dereference. The downside to this is that we need an extra cleanup step at the end of decode to kunmap() the last page. I introduced an xdr_finish_decode() function to do this. Right now this function only calls the unmap_current_page() function, but other generic cleanup steps could be added in the future if we come across anything else. Reported-by: Krzysztof Kozlowski Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xdr.h | 2 ++ net/sunrpc/clnt.c | 1 + net/sunrpc/svc.c | 2 ++ net/sunrpc/xdr.c | 27 ++++++++++++++++++++++++++- 4 files changed, 31 insertions(+), 1 deletion(-) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index f89ec4b5ea169..adc844db1ea52 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -224,6 +224,7 @@ struct xdr_stream { struct kvec *iov; /* pointer to the current kvec */ struct kvec scratch; /* Scratch buffer */ struct page **page_ptr; /* pointer to the current page */ + void *page_kaddr; /* kmapped address of the current page */ unsigned int nwords; /* Remaining decode buffer length */ struct rpc_rqst *rqst; /* For debugging */ @@ -255,6 +256,7 @@ extern void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p, struct rpc_rqst *rqst); extern void xdr_init_decode_pages(struct xdr_stream *xdr, struct xdr_buf *buf, struct page **pages, unsigned int len); +extern void xdr_finish_decode(struct xdr_stream *xdr); extern __be32 *xdr_inline_decode(struct xdr_stream *xdr, size_t nbytes); extern unsigned int xdr_read_pages(struct xdr_stream *xdr, unsigned int len); extern void xdr_enter_page(struct xdr_stream *xdr, unsigned int len); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d7c697af3762f..ca2c6efe19c95 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -2602,6 +2602,7 @@ call_decode(struct rpc_task *task) case 0: task->tk_action = rpc_exit_task; task->tk_status = rpcauth_unwrap_resp(task, &xdr); + xdr_finish_decode(&xdr); return; case -EAGAIN: task->tk_status = 0; diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 587811a002c98..a864414ce8119 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1370,6 +1370,8 @@ svc_process_common(struct svc_rqst *rqstp) rc = process.dispatch(rqstp); if (procp->pc_release) procp->pc_release(rqstp); + xdr_finish_decode(xdr); + if (!rc) goto dropit; if (rqstp->rq_auth_stat != rpc_auth_ok) diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 2a22e78af116e..f5011344dfe78 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -1288,6 +1288,14 @@ static unsigned int xdr_set_tail_base(struct xdr_stream *xdr, return xdr_set_iov(xdr, buf->tail, base, len); } +static void xdr_stream_unmap_current_page(struct xdr_stream *xdr) +{ + if (xdr->page_kaddr) { + kunmap_local(xdr->page_kaddr); + xdr->page_kaddr = NULL; + } +} + static unsigned int xdr_set_page_base(struct xdr_stream *xdr, unsigned int base, unsigned int len) { @@ -1305,12 +1313,18 @@ static unsigned int xdr_set_page_base(struct xdr_stream *xdr, if (len > maxlen) len = maxlen; + xdr_stream_unmap_current_page(xdr); xdr_stream_page_set_pos(xdr, base); base += xdr->buf->page_base; pgnr = base >> PAGE_SHIFT; xdr->page_ptr = &xdr->buf->pages[pgnr]; - kaddr = page_address(*xdr->page_ptr); + + if (PageHighMem(*xdr->page_ptr)) { + xdr->page_kaddr = kmap_local_page(*xdr->page_ptr); + kaddr = xdr->page_kaddr; + } else + kaddr = page_address(*xdr->page_ptr); pgoff = base & ~PAGE_MASK; xdr->p = (__be32*)(kaddr + pgoff); @@ -1364,6 +1378,7 @@ void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p, struct rpc_rqst *rqst) { xdr->buf = buf; + xdr->page_kaddr = NULL; xdr_reset_scratch_buffer(xdr); xdr->nwords = XDR_QUADLEN(buf->len); if (xdr_set_iov(xdr, buf->head, 0, buf->len) == 0 && @@ -1396,6 +1411,16 @@ void xdr_init_decode_pages(struct xdr_stream *xdr, struct xdr_buf *buf, } EXPORT_SYMBOL_GPL(xdr_init_decode_pages); +/** + * xdr_finish_decode - Clean up the xdr_stream after decoding data. + * @xdr: pointer to xdr_stream struct + */ +void xdr_finish_decode(struct xdr_stream *xdr) +{ + xdr_stream_unmap_current_page(xdr); +} +EXPORT_SYMBOL(xdr_finish_decode); + static __be32 * __xdr_inline_decode(struct xdr_stream *xdr, size_t nbytes) { unsigned int nwords = XDR_QUADLEN(nbytes); From 9cf2744d249144fc0fe17667b56da78216678378 Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Mon, 17 Jul 2023 16:33:14 -0400 Subject: [PATCH 079/186] NFS: Enable the READ_PLUS operation by default Now that the remaining issues have been worked out, including some unexpected 32 bit issues, we can safely enable the feature by default. Signed-off-by: Anna Schumaker --- fs/nfs/Kconfig | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index b6fc169be1b16..7df2503cef6c3 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -209,8 +209,6 @@ config NFS_DISABLE_UDP_SUPPORT config NFS_V4_2_READ_PLUS bool "NFS: Enable support for the NFSv4.2 READ_PLUS operation" depends on NFS_V4_2 - default n + default y help - This is intended for developers only. The READ_PLUS operation has - been shown to have issues under specific conditions and should not - be used in production. + Choose Y here to enable use of the NFS v4.2 READ_PLUS operation. From f9597ba8872a8f79f97b712ca098ffec841a374c Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Sat, 29 Jul 2023 20:31:52 +0800 Subject: [PATCH 080/186] xprtrdma: Remove unused function declaration rpcrdma_bc_post_recv() rpcrdma_bc_post_recv() is never implemented since introduction in commit f531a5dbc451 ("xprtrdma: Pre-allocate backward rpc_rqst and send/receive buffers"). Signed-off-by: Yue Haibing Reviewed-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/xprtrdma/xprt_rdma.h | 1 - 1 file changed, 1 deletion(-) diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h index 5e5ff6784ef5f..da409450dfc05 100644 --- a/net/sunrpc/xprtrdma/xprt_rdma.h +++ b/net/sunrpc/xprtrdma/xprt_rdma.h @@ -593,7 +593,6 @@ void xprt_rdma_cleanup(void); int xprt_rdma_bc_setup(struct rpc_xprt *, unsigned int); size_t xprt_rdma_bc_maxpayload(struct rpc_xprt *); unsigned int xprt_rdma_bc_max_slots(struct rpc_xprt *); -int rpcrdma_bc_post_recv(struct rpcrdma_xprt *, unsigned int); void rpcrdma_bc_receive_call(struct rpcrdma_xprt *, struct rpcrdma_rep *); int xprt_rdma_bc_send_reply(struct rpc_rqst *rqst); void xprt_rdma_bc_free_rqst(struct rpc_rqst *); From e87cf8a28e7592bd19064e8181324ae26bc02932 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 30 Jun 2023 12:46:53 +0300 Subject: [PATCH 081/186] SUNRPC: clean up integer overflow check This integer overflow check works as intended but Clang and GCC and warn about it when compiling with W=1. include/linux/sunrpc/xdr.h:539:17: error: comparison is always false due to limited range of data type [-Werror=type-limits] Use size_mul() to prevent the integer overflow. It silences the warning and it's cleaner as well. Reported-by: Dmitry Antipov Closes: https://lore.kernel.org/all/20230601143332.255312-1-dmantipov@yandex.ru/ Signed-off-by: Dan Carpenter Acked-by: Jeff Layton Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xdr.h | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index adc844db1ea52..68915180a29cb 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -777,9 +777,7 @@ xdr_stream_decode_uint32_array(struct xdr_stream *xdr, if (unlikely(xdr_stream_decode_u32(xdr, &len) < 0)) return -EBADMSG; - if (len > SIZE_MAX / sizeof(*p)) - return -EBADMSG; - p = xdr_inline_decode(xdr, len * sizeof(*p)); + p = xdr_inline_decode(xdr, size_mul(len, sizeof(*p))); if (unlikely(!p)) return -EBADMSG; if (array == NULL) From 08be82ba0cffdfa15ce2e2c312cb704823971862 Mon Sep 17 00:00:00 2001 From: GUO Zihua Date: Fri, 18 Aug 2023 11:31:02 +0800 Subject: [PATCH 082/186] NFS: Move common includes outside ifdef module.h, clnt.h, addr.h and dns_resolve.h is always included whether CONFIG_NFS_USE_KERNEL_DNS is set or not and their order does not seems to matter. Move them outside the ifdef to simplify code and avoid checkincludes message. Signed-off-by: GUO Zihua Signed-off-by: Anna Schumaker --- fs/nfs/dns_resolve.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index 6603b5cee029c..714975e5c0dbd 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -7,14 +7,16 @@ * Resolves DNS hostnames into valid ip addresses */ -#ifdef CONFIG_NFS_USE_KERNEL_DNS - #include #include #include -#include + #include "dns_resolve.h" +#ifdef CONFIG_NFS_USE_KERNEL_DNS + +#include + ssize_t nfs_dns_resolve_name(struct net *net, char *name, size_t namelen, struct sockaddr_storage *ss, size_t salen) { @@ -35,7 +37,6 @@ ssize_t nfs_dns_resolve_name(struct net *net, char *name, size_t namelen, #else -#include #include #include #include @@ -43,15 +44,12 @@ ssize_t nfs_dns_resolve_name(struct net *net, char *name, size_t namelen, #include #include #include -#include -#include #include #include #include #include #include "nfs4_fs.h" -#include "dns_resolve.h" #include "cache_lib.h" #include "netns.h" From 96562c45af5c31b89a197af28f79bfa838fb8391 Mon Sep 17 00:00:00 2001 From: Fedor Pchelkin Date: Thu, 20 Jul 2023 18:37:51 +0300 Subject: [PATCH 083/186] NFSv4/pnfs: minor fix for cleanup path in nfs4_get_device_info It is an almost improbable error case but when page allocating loop in nfs4_get_device_info() fails then we should only free the already allocated pages, as __free_page() can't deal with NULL arguments. Found by Linux Verification Center (linuxtesting.org). Cc: stable@vger.kernel.org Signed-off-by: Fedor Pchelkin Reviewed-by: Benjamin Coddington Signed-off-by: Anna Schumaker --- fs/nfs/pnfs_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c index ddbbf4fcda867..178001c90156f 100644 --- a/fs/nfs/pnfs_dev.c +++ b/fs/nfs/pnfs_dev.c @@ -154,7 +154,7 @@ nfs4_get_device_info(struct nfs_server *server, set_bit(NFS_DEVICEID_NOCACHE, &d->flags); out_free_pages: - for (i = 0; i < max_pages; i++) + while (--i >= 0) __free_page(pages[i]); kfree(pages); out_free_pdev: From a841c9cb9b04b05525f0928633e84e95921ab298 Mon Sep 17 00:00:00 2001 From: "huzhi001@208suo.com" Date: Wed, 19 Jul 2023 19:00:38 +0800 Subject: [PATCH 084/186] filemap: Fix errors in file.c The following checkpatch errors are removed: ERROR: "foo * bar" should be "foo *bar" "foo * bar" should be "foo *bar" Signed-off-by: ZhiHu Signed-off-by: Anna Schumaker --- fs/nfs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 79b1b3fcd3fcf..3f9768810427d 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -200,7 +200,7 @@ nfs_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe EXPORT_SYMBOL_GPL(nfs_file_splice_read); int -nfs_file_mmap(struct file * file, struct vm_area_struct * vma) +nfs_file_mmap(struct file *file, struct vm_area_struct *vma) { struct inode *inode = file_inode(file); int status; From 08b45fcb2d4675f6182fe0edc0d8b1fe604051fa Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 24 Jul 2023 11:08:46 +0300 Subject: [PATCH 085/186] nfs/blocklayout: Use the passed in gfp flags This allocation should use the passed in GFP_ flags instead of GFP_KERNEL. One places where this matters is in filelayout_pg_init_write() which uses GFP_NOFS as the allocation flags. Fixes: 5c83746a0cf2 ("pnfs/blocklayout: in-kernel GETDEVICEINFO XDR parsing") Signed-off-by: Dan Carpenter Reviewed-by: Christoph Hellwig Signed-off-by: Anna Schumaker --- fs/nfs/blocklayout/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/nfs/blocklayout/dev.c b/fs/nfs/blocklayout/dev.c index 70f5563a8e81c..65cbb5607a5fc 100644 --- a/fs/nfs/blocklayout/dev.c +++ b/fs/nfs/blocklayout/dev.c @@ -404,7 +404,7 @@ bl_parse_concat(struct nfs_server *server, struct pnfs_block_dev *d, int ret, i; d->children = kcalloc(v->concat.volumes_count, - sizeof(struct pnfs_block_dev), GFP_KERNEL); + sizeof(struct pnfs_block_dev), gfp_mask); if (!d->children) return -ENOMEM; @@ -433,7 +433,7 @@ bl_parse_stripe(struct nfs_server *server, struct pnfs_block_dev *d, int ret, i; d->children = kcalloc(v->stripe.volumes_count, - sizeof(struct pnfs_block_dev), GFP_KERNEL); + sizeof(struct pnfs_block_dev), gfp_mask); if (!d->children) return -ENOMEM; From 14e7316a3c73cf45cef4422549f3585fc1c53521 Mon Sep 17 00:00:00 2001 From: Kinglong Mee Date: Fri, 21 Jul 2023 09:23:00 +0800 Subject: [PATCH 086/186] nfs: fix redundant readdir request after get eof When a directory contains 17 files (except . and ..), nfs client sends a redundant readdir request after get eof. A simple reproduce, At NFS server, create a directory with 17 files under exported directory. # mkdir test # cd test # for i in {0..16} ; do touch $i; done At NFS client, no matter mounting through nfsv3 or nfsv4, does ls (or ll) at the created test directory. A tshark output likes following (for nfsv4), # tshark -i eth0 tcp port 2049 -Tfields -e ip.src -e ip.dst -e nfs -e nfs.cookie4 srcip dstip SEQUENCE, PUTFH, READDIR 0 dstip srcip SEQUENCE PUTFH READDIR 909539109313539306,2108391201987888856,2305312124304486544,2566335452463141496,2978225129081509984,4263037479923412583,4304697173036510679,4666703455469210097,4759208201298769007,4776701232145978803,5338408478512081262,5949498658935544804,5971526429894832903,6294060338267709855,6528840566229532529,8600463293536422524,9223372036854775807 srcip dstip srcip dstip SEQUENCE, PUTFH, READDIR 9223372036854775807 dstip srcip SEQUENCE PUTFH READDIR The READDIR with cookie 9223372036854775807(0x7FFFFFFFFFFFFFFF) is redundant. Reviewed-by: Benjamin Coddington Signed-off-by: Kinglong Mee Signed-off-by: Anna Schumaker --- fs/nfs/dir.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 8f3112e71a6a6..e6a51fd94fea8 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1089,6 +1089,17 @@ static void nfs_do_filldir(struct nfs_readdir_descriptor *desc, for (i = desc->cache_entry_index; i < array->size; i++) { struct nfs_cache_array_entry *ent; + /* + * nfs_readdir_handle_cache_misses return force clear at + * (cache_misses > NFS_READDIR_CACHE_MISS_THRESHOLD) for + * readdir heuristic, NFS_READDIR_CACHE_MISS_THRESHOLD + 1 + * entries need be emitted here. + */ + if (first_emit && i > NFS_READDIR_CACHE_MISS_THRESHOLD + 2) { + desc->eob = true; + break; + } + ent = &array->array[i]; if (!dir_emit(desc->ctx, ent->name, ent->name_len, nfs_compat_user_ino64(ent->ino), ent->d_type)) { @@ -1107,10 +1118,6 @@ static void nfs_do_filldir(struct nfs_readdir_descriptor *desc, desc->ctx->pos = desc->dir_cookie; else desc->ctx->pos++; - if (first_emit && i > NFS_READDIR_CACHE_MISS_THRESHOLD + 1) { - desc->eob = true; - break; - } } if (array->folio_is_eof) desc->eof = !desc->eob; From 88975a55969e11f26fe3846bf4fbf8e7dc8cbbd4 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:22:14 -0400 Subject: [PATCH 087/186] NFS: Fix a potential data corruption We must ensure that the subrequests are joined back into the head before we can retransmit a request. If the head was not on the commit lists, because the server wrote it synchronously, we still need to add it back to the retransmission list. Add a call that mirrors the effect of nfs_cancel_remove_inode() for O_DIRECT. Fixes: ed5d588fe47f ("NFS: Try to join page groups before an O_DIRECT retransmission") Cc: stable@vger.kernel.org Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- fs/nfs/direct.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index aaffaaa336cc5..47d892a1d363d 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -472,13 +472,31 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter, return result; } +static void nfs_direct_add_page_head(struct list_head *list, + struct nfs_page *req) +{ + struct nfs_page *head = req->wb_head; + + if (!list_empty(&head->wb_list) || !nfs_lock_request(head)) + return; + if (!list_empty(&head->wb_list)) { + nfs_unlock_request(head); + return; + } + list_add(&head->wb_list, list); + kref_get(&head->wb_kref); + kref_get(&head->wb_kref); +} + static void nfs_direct_join_group(struct list_head *list, struct inode *inode) { struct nfs_page *req, *subreq; list_for_each_entry(req, list, wb_list) { - if (req->wb_head != req) + if (req->wb_head != req) { + nfs_direct_add_page_head(&req->wb_list, req); continue; + } subreq = req->wb_this_page; if (subreq == req) continue; From 3a107f07403aa3bf0d604996a30922812e30f8a9 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:32:21 -0400 Subject: [PATCH 088/186] SUNRPC: Set the TCP_SYNCNT to match the socket timeout Set the TCP SYN count so that we abort the connection attempt at around the expected timeout value. Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- net/sunrpc/xprtsock.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 9f010369100a2..47d0b6a8c32e5 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -2230,9 +2230,13 @@ static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt, struct socket *sock) { struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); + struct net *net = sock_net(sock->sk); + unsigned long connect_timeout; + unsigned long syn_retries; unsigned int keepidle; unsigned int keepcnt; unsigned int timeo; + unsigned long t; spin_lock(&xprt->transport_lock); keepidle = DIV_ROUND_UP(xprt->timeout->to_initval, HZ); @@ -2250,6 +2254,16 @@ static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt, /* TCP user timeout (see RFC5482) */ tcp_sock_set_user_timeout(sock->sk, timeo); + + /* Connect timeout */ + connect_timeout = max_t(unsigned long, + DIV_ROUND_UP(xprt->connect_timeout, HZ), 1); + syn_retries = max_t(unsigned long, + READ_ONCE(net->ipv4.sysctl_tcp_syn_retries), 1); + for (t = 0; t <= syn_retries && (1UL << t) < connect_timeout; t++) + ; + if (t <= syn_retries) + tcp_sock_set_syncnt(sock->sk, t - 1); } static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt, From 3e6ff89d2e4b605d7064686e6d8991b6f780df3f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:32:22 -0400 Subject: [PATCH 089/186] SUNRPC: Refactor and simplify connect timeout Instead of requiring the requests to redrive the connection several times, just let the TCP connect code manage it now that we've adjusted the TCP_SYNCNT value. Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- net/sunrpc/xprtsock.c | 34 +++++++++++++++++++++------------- 1 file changed, 21 insertions(+), 13 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 47d0b6a8c32e5..e558f0024fe59 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -2266,6 +2266,25 @@ static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt, tcp_sock_set_syncnt(sock->sk, t - 1); } +static void xs_tcp_do_set_connect_timeout(struct rpc_xprt *xprt, + unsigned long connect_timeout) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); + struct rpc_timeout to; + unsigned long initval; + + memcpy(&to, xprt->timeout, sizeof(to)); + /* Arbitrary lower limit */ + initval = max_t(unsigned long, connect_timeout, XS_TCP_INIT_REEST_TO); + to.to_initval = initval; + to.to_maxval = initval; + to.to_retries = 0; + memcpy(&transport->tcp_timeout, &to, sizeof(transport->tcp_timeout)); + xprt->timeout = &transport->tcp_timeout; + xprt->connect_timeout = connect_timeout; +} + static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt, unsigned long connect_timeout, unsigned long reconnect_timeout) @@ -2277,19 +2296,8 @@ static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt, spin_lock(&xprt->transport_lock); if (reconnect_timeout < xprt->max_reconnect_timeout) xprt->max_reconnect_timeout = reconnect_timeout; - if (connect_timeout < xprt->connect_timeout) { - memcpy(&to, xprt->timeout, sizeof(to)); - initval = DIV_ROUND_UP(connect_timeout, to.to_retries + 1); - /* Arbitrary lower limit */ - if (initval < XS_TCP_INIT_REEST_TO << 1) - initval = XS_TCP_INIT_REEST_TO << 1; - to.to_initval = initval; - to.to_maxval = initval; - memcpy(&transport->tcp_timeout, &to, - sizeof(transport->tcp_timeout)); - xprt->timeout = &transport->tcp_timeout; - xprt->connect_timeout = connect_timeout; - } + if (connect_timeout < xprt->connect_timeout) + xs_tcp_do_set_connect_timeout(xprt, connect_timeout); set_bit(XPRT_SOCK_UPD_TIMEOUT, &transport->sock_state); spin_unlock(&xprt->transport_lock); } From d2ee413884cdbdfcfc1560526615519311a47d33 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:32:23 -0400 Subject: [PATCH 090/186] SUNRPC: Allow specification of TCP client connect timeout at setup When we create a TCP transport, the connect timeout parameters are currently fixed to be 90s. This is problematic in the pNFS flexfiles case, where we may have multiple mirrors, and we would like to fail over quickly to the next mirror if a data server is down. This patch adds the ability to specify the connection parameters at RPC client creation time. Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- include/linux/sunrpc/clnt.h | 2 ++ include/linux/sunrpc/xprt.h | 2 ++ net/sunrpc/clnt.c | 2 ++ net/sunrpc/xprtsock.c | 7 +++++-- 4 files changed, 11 insertions(+), 2 deletions(-) diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 4f41d839face4..af7358277f1c3 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -148,6 +148,8 @@ struct rpc_create_args { const struct cred *cred; unsigned int max_connect; struct xprtsec_parms xprtsec; + unsigned long connect_timeout; + unsigned long reconnect_timeout; }; struct rpc_add_xprt_test { diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index b52411bcfe4e7..4ecc89301eb74 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -351,6 +351,8 @@ struct xprt_create { struct rpc_xprt_switch *bc_xps; unsigned int flags; struct xprtsec_parms xprtsec; + unsigned long connect_timeout; + unsigned long reconnect_timeout; }; struct xprt_class { diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index ca2c6efe19c95..06df08b0ee9e1 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -534,6 +534,8 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args) .servername = args->servername, .bc_xprt = args->bc_xprt, .xprtsec = args->xprtsec, + .connect_timeout = args->connect_timeout, + .reconnect_timeout = args->reconnect_timeout, }; char servername[48]; struct rpc_clnt *clnt; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index e558f0024fe59..6e845e51cbf31 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -2290,8 +2290,6 @@ static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt, unsigned long reconnect_timeout) { struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); - struct rpc_timeout to; - unsigned long initval; spin_lock(&xprt->transport_lock); if (reconnect_timeout < xprt->max_reconnect_timeout) @@ -3350,8 +3348,13 @@ static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args) xprt->timeout = &xs_tcp_default_timeout; xprt->max_reconnect_timeout = xprt->timeout->to_maxval; + if (args->reconnect_timeout) + xprt->max_reconnect_timeout = args->reconnect_timeout; + xprt->connect_timeout = xprt->timeout->to_initval * (xprt->timeout->to_retries + 1); + if (args->connect_timeout) + xs_tcp_do_set_connect_timeout(xprt, args->connect_timeout); INIT_WORK(&transport->recv_worker, xs_stream_data_receive_workfn); INIT_WORK(&transport->error_worker, xs_error_handle); From cd18f24085f012b46b8271640b3c60fb27c0b05f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:32:24 -0400 Subject: [PATCH 091/186] SUNRPC: Don't override connect timeouts in rpc_clnt_add_xprt() If the caller specifies the connect timeouts in the arguments to rpc_clnt_add_xprt(), then we shouldn't override them. Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- net/sunrpc/clnt.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 06df08b0ee9e1..8d75290f1a31d 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -3072,6 +3072,11 @@ int rpc_clnt_add_xprt(struct rpc_clnt *clnt, } xprt->resvport = resvport; xprt->reuseport = reuseport; + + if (xprtargs->connect_timeout) + connect_timeout = xprtargs->connect_timeout; + if (xprtargs->reconnect_timeout) + reconnect_timeout = xprtargs->reconnect_timeout; if (xprt->ops->set_connect_timeout != NULL) xprt->ops->set_connect_timeout(xprt, connect_timeout, From 537935f72eb28a3dd0097386f06402e25e66359a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 19 Aug 2023 17:32:25 -0400 Subject: [PATCH 092/186] NFS/pNFS: Set the connect timeout for the pNFS flexfiles driver Ensure that the connect timeout for the pNFS flexfiles driver is of the same order as the I/O timeout, so that we can fail over quickly when trying to read from a data server that is down. Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- fs/nfs/client.c | 2 ++ fs/nfs/internal.h | 2 ++ fs/nfs/nfs3client.c | 3 +++ fs/nfs/pnfs_nfs.c | 3 +++ 4 files changed, 10 insertions(+) diff --git a/fs/nfs/client.c b/fs/nfs/client.c index e4c5f193ed5e8..44eca51b28085 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -517,6 +517,8 @@ int nfs_create_rpc_client(struct nfs_client *clp, .authflavor = flavor, .cred = cl_init->cred, .xprtsec = cl_init->xprtsec, + .connect_timeout = cl_init->connect_timeout, + .reconnect_timeout = cl_init->reconnect_timeout, }; if (test_bit(NFS_CS_DISCRTRY, &clp->cl_flags)) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 41abea340ad84..9c9cf764f6000 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -82,6 +82,8 @@ struct nfs_client_initdata { const struct rpc_timeout *timeparms; const struct cred *cred; struct xprtsec_parms xprtsec; + unsigned long connect_timeout; + unsigned long reconnect_timeout; }; /* diff --git a/fs/nfs/nfs3client.c b/fs/nfs/nfs3client.c index eff3802c5e035..674c012868b1a 100644 --- a/fs/nfs/nfs3client.c +++ b/fs/nfs/nfs3client.c @@ -86,6 +86,7 @@ struct nfs_client *nfs3_set_ds_client(struct nfs_server *mds_srv, int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans) { struct rpc_timeout ds_timeout; + unsigned long connect_timeout = ds_timeo * (ds_retrans + 1) * HZ / 10; struct nfs_client *mds_clp = mds_srv->nfs_client; struct nfs_client_initdata cl_init = { .addr = ds_addr, @@ -98,6 +99,8 @@ struct nfs_client *nfs3_set_ds_client(struct nfs_server *mds_srv, .timeparms = &ds_timeout, .cred = mds_srv->cred, .xprtsec = mds_clp->cl_xprtsec, + .connect_timeout = connect_timeout, + .reconnect_timeout = connect_timeout, }; struct nfs_client *clp; char buf[INET6_ADDRSTRLEN + 1]; diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c index a0112ad4937aa..a08cfda6fff1f 100644 --- a/fs/nfs/pnfs_nfs.c +++ b/fs/nfs/pnfs_nfs.c @@ -852,6 +852,7 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv, { struct nfs_client *clp = ERR_PTR(-EIO); struct nfs4_pnfs_ds_addr *da; + unsigned long connect_timeout = timeo * (retrans + 1) * HZ / 10; int status = 0; dprintk("--> %s DS %s\n", __func__, ds->ds_remotestr); @@ -870,6 +871,8 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv, .dstaddr = (struct sockaddr *)&da->da_addr, .addrlen = da->da_addrlen, .servername = clp->cl_hostname, + .connect_timeout = connect_timeout, + .reconnect_timeout = connect_timeout, }; if (da->da_transport != clp->cl_proto) From 51d674a5e4889f1c8e223ac131cf218e1631e423 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Thu, 13 Jul 2023 13:02:38 -0400 Subject: [PATCH 093/186] NFSv4.1: use EXCHGID4_FLAG_USE_PNFS_DS for DS server After receiving the location(s) of the DS server(s) in the GETDEVINCEINFO, create the request for the clientid to such server and indicate that the client is connecting to a DS. Signed-off-by: Olga Kornievskaia Signed-off-by: Anna Schumaker --- fs/nfs/nfs4client.c | 3 +++ fs/nfs/nfs4proc.c | 4 ++++ 2 files changed, 7 insertions(+) diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index d9114a754db73..27fb25567ce75 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -232,6 +232,8 @@ struct nfs_client *nfs4_alloc_client(const struct nfs_client_initdata *cl_init) __set_bit(NFS_CS_DISCRTRY, &clp->cl_flags); __set_bit(NFS_CS_NO_RETRANS_TIMEOUT, &clp->cl_flags); + if (test_bit(NFS_CS_DS, &cl_init->init_flags)) + __set_bit(NFS_CS_DS, &clp->cl_flags); /* * Set up the connection to the server before we add add to the * global list. @@ -1007,6 +1009,7 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv, if (mds_srv->flags & NFS_MOUNT_NORESVPORT) __set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags); + __set_bit(NFS_CS_DS, &cl_init.init_flags); /* * Set an authflavor equual to the MDS value. Use the MDS nfs_client * cl_ipaddr so as to use the same EXCHANGE_ID co_ownerid as the MDS diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 3c24c3c99e8ac..3bc6bfdf7b814 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -8787,6 +8787,8 @@ nfs4_run_exchange_id(struct nfs_client *clp, const struct cred *cred, #ifdef CONFIG_NFS_V4_1_MIGRATION calldata->args.flags |= EXCHGID4_FLAG_SUPP_MOVED_MIGR; #endif + if (test_bit(NFS_CS_DS, &clp->cl_flags)) + calldata->args.flags |= EXCHGID4_FLAG_USE_PNFS_DS; msg.rpc_argp = &calldata->args; msg.rpc_resp = &calldata->res; task_setup_data.callback_data = calldata; @@ -8864,6 +8866,8 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, const struct cred *cre /* Save the EXCHANGE_ID verifier session trunk tests */ memcpy(clp->cl_confirm.data, argp->verifier.data, sizeof(clp->cl_confirm.data)); + if (resp->flags & EXCHGID4_FLAG_USE_PNFS_DS) + set_bit(NFS_CS_DS, &clp->cl_flags); out: trace_nfs4_exchange_id(clp, status); rpc_put_task(task); From 7c53e847ff5e97f033fdd31f71949807633d506b Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 24 Aug 2023 16:51:42 -0400 Subject: [PATCH 094/186] dlm: fix plock lookup when using multiple lockspaces All posix lock ops, for all lockspaces (gfs2 file systems) are sent to userspace (dlm_controld) through a single misc device. The dlm_controld daemon reads the ops from the misc device and sends them to other cluster nodes using separate, per-lockspace cluster api communication channels. The ops for a single lockspace are ordered at this level, so that the results are received in the same sequence that the requests were sent. When the results are sent back to the kernel via the misc device, they are again funneled through the single misc device for all lockspaces. When the dlm code in the kernel processes the results from the misc device, these results will be returned in the same sequence that the requests were sent, on a per-lockspace basis. A recent change in this request/reply matching code missed the "per-lockspace" check (fsid comparison) when matching request and reply, so replies could be incorrectly matched to requests from other lockspaces. Cc: stable@vger.kernel.org Reported-by: Barry Marson Fixes: 57e2c2f2d94c ("fs: dlm: fix mismatch of plock results from userspace") Signed-off-by: Alexander Aring Signed-off-by: David Teigland --- fs/dlm/plock.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 00e1d802a81cb..e6b4c1a214466 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -556,7 +556,8 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, op = plock_lookup_waiter(&info); } else { list_for_each_entry(iter, &recv_list, list) { - if (!iter->info.wait) { + if (!iter->info.wait && + iter->info.fsid == info.fsid) { op = iter; break; } @@ -568,8 +569,7 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, if (info.wait) WARN_ON(op->info.optype != DLM_PLOCK_OP_LOCK); else - WARN_ON(op->info.fsid != info.fsid || - op->info.number != info.number || + WARN_ON(op->info.number != info.number || op->info.owner != info.owner || op->info.optype != info.optype); From ee5c807137ce283acebd83297f8855428cdd839a Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Fri, 11 Aug 2023 14:36:10 +0800 Subject: [PATCH 095/186] ext4: ext4_get_{dev}_journal return proper error value ext4_get_journal() and ext4_get_dev_journal() return NULL if they failed to init journal, making them return proper error value instead, also rename them to ext4_open_{inode,dev}_journal(). [ Folded fix to ext4_calculate_overhead() to check for an ERR_PTR instead of NULL. ] Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230811063610.2980059-13-yi.zhang@huaweicloud.com Reported-by: syzbot+b3123e6d9842e526de39@syzkaller.appspotmail.com Link: https://lore.kernel.org/r/20230826011029.2023140-1-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 53 +++++++++++++++++++++++++++++-------------------- 1 file changed, 31 insertions(+), 22 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 0eee238b290e0..6edf7deeb2dbe 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4200,7 +4200,7 @@ int ext4_calculate_overhead(struct super_block *sb) else if (ext4_has_feature_journal(sb) && !sbi->s_journal && j_inum) { /* j_inum for internal journal is non-zero */ j_inode = ext4_get_journal_inode(sb, j_inum); - if (j_inode) { + if (!IS_ERR(j_inode)) { j_blocks = j_inode->i_size >> sb->s_blocksize_bits; overhead += EXT4_NUM_B2C(sbi, j_blocks); iput(j_inode); @@ -5740,18 +5740,18 @@ static struct inode *ext4_get_journal_inode(struct super_block *sb, journal_inode = ext4_iget(sb, journal_inum, EXT4_IGET_SPECIAL); if (IS_ERR(journal_inode)) { ext4_msg(sb, KERN_ERR, "no journal found"); - return NULL; + return ERR_CAST(journal_inode); } if (!journal_inode->i_nlink) { make_bad_inode(journal_inode); iput(journal_inode); ext4_msg(sb, KERN_ERR, "journal inode is deleted"); - return NULL; + return ERR_PTR(-EFSCORRUPTED); } if (!S_ISREG(journal_inode->i_mode) || IS_ENCRYPTED(journal_inode)) { ext4_msg(sb, KERN_ERR, "invalid journal inode"); iput(journal_inode); - return NULL; + return ERR_PTR(-EFSCORRUPTED); } ext4_debug("Journal inode found at %p: %lld bytes\n", @@ -5781,21 +5781,21 @@ static int ext4_journal_bmap(journal_t *journal, sector_t *block) return 0; } -static journal_t *ext4_get_journal(struct super_block *sb, - unsigned int journal_inum) +static journal_t *ext4_open_inode_journal(struct super_block *sb, + unsigned int journal_inum) { struct inode *journal_inode; journal_t *journal; journal_inode = ext4_get_journal_inode(sb, journal_inum); - if (!journal_inode) - return NULL; + if (IS_ERR(journal_inode)) + return ERR_CAST(journal_inode); journal = jbd2_journal_init_inode(journal_inode); if (IS_ERR(journal)) { ext4_msg(sb, KERN_ERR, "Could not load journal inode"); iput(journal_inode); - return NULL; + return ERR_CAST(journal); } journal->j_private = sb; journal->j_bmap = ext4_journal_bmap; @@ -5813,6 +5813,7 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, ext4_fsblk_t sb_block; unsigned long offset; struct ext4_super_block *es; + int errno; bdev = blkdev_get_by_dev(j_dev, BLK_OPEN_READ | BLK_OPEN_WRITE, sb, &ext4_holder_ops); @@ -5820,7 +5821,7 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, ext4_msg(sb, KERN_ERR, "failed to open journal device unknown-block(%u,%u) %ld", MAJOR(j_dev), MINOR(j_dev), PTR_ERR(bdev)); - return NULL; + return ERR_CAST(bdev); } blocksize = sb->s_blocksize; @@ -5828,6 +5829,7 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, if (blocksize < hblock) { ext4_msg(sb, KERN_ERR, "blocksize too small for journal device"); + errno = -EINVAL; goto out_bdev; } @@ -5838,6 +5840,7 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, if (!bh) { ext4_msg(sb, KERN_ERR, "couldn't read superblock of " "external journal"); + errno = -EINVAL; goto out_bdev; } @@ -5846,6 +5849,7 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, !(le32_to_cpu(es->s_feature_incompat) & EXT4_FEATURE_INCOMPAT_JOURNAL_DEV)) { ext4_msg(sb, KERN_ERR, "external journal has bad superblock"); + errno = -EFSCORRUPTED; goto out_bh; } @@ -5853,11 +5857,13 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, EXT4_FEATURE_RO_COMPAT_METADATA_CSUM) && es->s_checksum != ext4_superblock_csum(sb, es)) { ext4_msg(sb, KERN_ERR, "external journal has corrupt superblock"); + errno = -EFSCORRUPTED; goto out_bh; } if (memcmp(EXT4_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) { ext4_msg(sb, KERN_ERR, "journal UUID does not match"); + errno = -EFSCORRUPTED; goto out_bh; } @@ -5870,31 +5876,34 @@ static struct block_device *ext4_get_journal_blkdev(struct super_block *sb, brelse(bh); out_bdev: blkdev_put(bdev, sb); - return NULL; + return ERR_PTR(errno); } -static journal_t *ext4_get_dev_journal(struct super_block *sb, - dev_t j_dev) +static journal_t *ext4_open_dev_journal(struct super_block *sb, + dev_t j_dev) { journal_t *journal; ext4_fsblk_t j_start; ext4_fsblk_t j_len; struct block_device *journal_bdev; + int errno = 0; journal_bdev = ext4_get_journal_blkdev(sb, j_dev, &j_start, &j_len); - if (!journal_bdev) - return NULL; + if (IS_ERR(journal_bdev)) + return ERR_CAST(journal_bdev); journal = jbd2_journal_init_dev(journal_bdev, sb->s_bdev, j_start, j_len, sb->s_blocksize); if (IS_ERR(journal)) { ext4_msg(sb, KERN_ERR, "failed to create device journal"); + errno = PTR_ERR(journal); goto out_bdev; } if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) { ext4_msg(sb, KERN_ERR, "External journal has more than one " "user (unsupported) - %d", be32_to_cpu(journal->j_superblock->s_nr_users)); + errno = -EINVAL; goto out_journal; } journal->j_private = sb; @@ -5906,7 +5915,7 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, jbd2_journal_destroy(journal); out_bdev: blkdev_put(journal_bdev, sb); - return NULL; + return ERR_PTR(errno); } static int ext4_load_journal(struct super_block *sb, @@ -5938,13 +5947,13 @@ static int ext4_load_journal(struct super_block *sb, } if (journal_inum) { - journal = ext4_get_journal(sb, journal_inum); - if (!journal) - return -EINVAL; + journal = ext4_open_inode_journal(sb, journal_inum); + if (IS_ERR(journal)) + return PTR_ERR(journal); } else { - journal = ext4_get_dev_journal(sb, journal_dev); - if (!journal) - return -EINVAL; + journal = ext4_open_dev_journal(sb, journal_dev); + if (IS_ERR(journal)) + return PTR_ERR(journal); } journal_dev_ro = bdev_read_only(journal->j_dev); From 2dfba3bb40ad8536b9fa802364f2d40da31aa88e Mon Sep 17 00:00:00 2001 From: Zhang Yi Date: Mon, 26 Jun 2023 15:33:22 +0800 Subject: [PATCH 096/186] jbd2: correct the end of the journal recovery scan range We got a filesystem inconsistency issue below while running generic/475 I/O failure pressure test with fast_commit feature enabled. Symlink /p3/d3/d1c/d6c/dd6/dce/l101 (inode #132605) is invalid. If fast_commit feature is enabled, a special fast_commit journal area is appended to the end of the normal journal area. The journal->j_last point to the first unused block behind the normal journal area instead of the whole log area, and the journal->j_fc_last point to the first unused block behind the fast_commit journal area. While doing journal recovery, do_one_pass(PASS_SCAN) should first scan the normal journal area and turn around to the first block once it meet journal->j_last, but the wrap() macro misuse the journal->j_fc_last, so the recovering could not read the next magic block (commit block perhaps) and would end early mistakenly and missing tN and every transaction after it in the following example. Finally, it could lead to filesystem inconsistency. | normal journal area | fast commit area | +-------------------------------------------------+------------------+ | tN(rere) | tN+1 |~| tN-x |...| tN-1 | tN(front) | .... | +-------------------------------------------------+------------------+ / / / start journal->j_last journal->j_fc_last This patch fix it by use the correct ending journal->j_last. Fixes: 5b849b5f96b4 ("jbd2: fast commit recovery path") Cc: stable@kernel.org Reported-by: Theodore Ts'o Link: https://lore.kernel.org/linux-ext4/20230613043120.GB1584772@mit.edu/ Signed-off-by: Zhang Yi Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230626073322.3956567-1-yi.zhang@huaweicloud.com Signed-off-by: Theodore Ts'o --- fs/jbd2/recovery.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c index 0184931d47f7d..c269a7d29a465 100644 --- a/fs/jbd2/recovery.c +++ b/fs/jbd2/recovery.c @@ -230,12 +230,8 @@ static int count_tags(journal_t *journal, struct buffer_head *bh) /* Make sure we wrap around the log correctly! */ #define wrap(journal, var) \ do { \ - unsigned long _wrap_last = \ - jbd2_has_feature_fast_commit(journal) ? \ - (journal)->j_fc_last : (journal)->j_last; \ - \ - if (var >= _wrap_last) \ - var -= (_wrap_last - (journal)->j_first); \ + if (var >= (journal)->j_last) \ + var -= ((journal)->j_last - (journal)->j_first); \ } while (0) static int fc_do_one_pass(journal_t *journal, @@ -524,9 +520,7 @@ static int do_one_pass(journal_t *journal, break; jbd2_debug(2, "Scanning for sequence ID %u at %lu/%lu\n", - next_commit_ID, next_log_block, - jbd2_has_feature_fast_commit(journal) ? - journal->j_fc_last : journal->j_last); + next_commit_ID, next_log_block, journal->j_last); /* Skip over each chunk of the transaction looking * either the next descriptor block or the final commit From 1524773425ae8113b0b782886366e68656b34e53 Mon Sep 17 00:00:00 2001 From: Zhihao Cheng Date: Wed, 28 Jun 2023 21:20:11 +0800 Subject: [PATCH 097/186] ext4: fix unttached inode after power cut with orphan file feature enabled Running generic/475(filesystem consistent tests after power cut) could easily trigger unattached inode error while doing fsck: Unattached zero-length inode 39405. Clear? no Unattached inode 39405 Connect to /lost+found? no Above inconsistence is caused by following process: P1 P2 ext4_create inode = ext4_new_inode_start_handle // itable records nlink=1 ext4_add_nondir err = ext4_add_entry // ENOSPC ext4_append ext4_bread ext4_getblk ext4_map_blocks // returns ENOSPC drop_nlink(inode) // won't be updated into disk inode ext4_orphan_add(handle, inode) ext4_orphan_file_add ext4_journal_stop(handle) jbd2_journal_commit_transaction // commit success >> power cut << ext4_fill_super ext4_load_and_init_journal // itable records nlink=1 ext4_orphan_cleanup ext4_process_orphan if (inode->i_nlink) // true, inode won't be deleted Then, allocated inode will be reserved on disk and corresponds to no dentries, so e2fsck reports 'unattached inode' problem. The problem won't happen if orphan file feature is disabled, because ext4_orphan_add() will update disk inode in orphan list mode. There are several places not updating disk inode while putting inode into orphan area, such as ext4_add_nondir(), ext4_symlink() and whiteout in ext4_rename(). Fix it by updating inode into disk in all error branches of these places. Link: https://bugzilla.kernel.org/show_bug.cgi?id=217605 Fixes: 02f310fcf47f ("ext4: Speedup ext4 orphan inode handling") Signed-off-by: Zhihao Cheng Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230628132011.650383-1-chengzhihao1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/namei.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 6298cfaaa0bde..9a13431656cd8 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2799,6 +2799,7 @@ static int ext4_add_nondir(handle_t *handle, return err; } drop_nlink(inode); + ext4_mark_inode_dirty(handle, inode); ext4_orphan_add(handle, inode); unlock_new_inode(inode); return err; @@ -3436,6 +3437,7 @@ static int ext4_symlink(struct mnt_idmap *idmap, struct inode *dir, err_drop_inode: clear_nlink(inode); + ext4_mark_inode_dirty(handle, inode); ext4_orphan_add(handle, inode); unlock_new_inode(inode); if (handle) @@ -4021,6 +4023,7 @@ static int ext4_rename(struct mnt_idmap *idmap, struct inode *old_dir, ext4_resetent(handle, &old, old.inode->i_ino, old_file_type); drop_nlink(whiteout); + ext4_mark_inode_dirty(handle, whiteout); ext4_orphan_add(handle, whiteout); } unlock_new_inode(whiteout); From 89cadf6e22a958014d09c901caf0cd2105780dbe Mon Sep 17 00:00:00 2001 From: Lu Hongfei Date: Fri, 7 Jul 2023 18:55:16 +0800 Subject: [PATCH 098/186] ext4: change the type of blocksize in ext4_mb_init_cache() The return value type of i_blocksize() is 'unsigned int', so the type of blocksize has been modified from 'int' to 'unsigned int' to ensure data type consistency. Signed-off-by: Lu Hongfei Link: https://lore.kernel.org/r/20230707105516.9156-1-luhongfei@vivo.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 3d5b0b71d7f5b..96068d687d9da 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1256,7 +1256,7 @@ void ext4_mb_generate_buddy(struct super_block *sb, static int ext4_mb_init_cache(struct page *page, char *incore, gfp_t gfp) { ext4_group_t ngroups; - int blocksize; + unsigned int blocksize; int blocks_per_page; int groups_per_page; int err = 0; From 79ebf48c44b5ba05a98af23f8830883daf36f4d3 Mon Sep 17 00:00:00 2001 From: Lu Hongfei Date: Fri, 7 Jul 2023 19:59:07 +0800 Subject: [PATCH 099/186] ext4: use sbi instead of EXT4_SB(sb) in ext4_mb_new_blocks_simple() Signed-off-by: Lu Hongfei Link: https://lore.kernel.org/r/20230707115907.26637-1-luhongfei@vivo.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 96068d687d9da..a807e8bf86643 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -6092,7 +6092,7 @@ ext4_mb_new_blocks_simple(struct ext4_allocation_request *ar, int *errp) ext4_grpblk_t max = EXT4_CLUSTERS_PER_GROUP(sb); ext4_grpblk_t i = 0; ext4_fsblk_t goal, block; - struct ext4_super_block *es = EXT4_SB(sb)->s_es; + struct ext4_super_block *es = sbi->s_es; goal = ar->goal; if (goal < le32_to_cpu(es->s_first_data_block) || From a50bda147421e24c1a5d47ddcc0675360b7cb3ac Mon Sep 17 00:00:00 2001 From: Su Hui Date: Tue, 25 Jul 2023 12:33:11 +0800 Subject: [PATCH 100/186] ext4: mballoc: avoid garbage value from err clang's static analysis warning: fs/ext4/mballoc.c line 4178, column 6, Branch condition evaluates to a garbage value. err is uninitialized and will be judged when 'len <= 0' or it first enters the loop while the condition "!ext4_sb_block_valid()" is true. Although this can't make problems now, it's better to correct it. Signed-off-by: Su Hui Reviewed-by: Nick Desaulniers Link: https://lore.kernel.org/r/20230725043310.1227621-1-suhui@nfschina.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index a807e8bf86643..1e4c667812a9d 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4087,7 +4087,7 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block, struct ext4_sb_info *sbi = EXT4_SB(sb); ext4_group_t group; ext4_grpblk_t blkoff; - int i, err; + int i, err = 0; int already; unsigned int clen, clen_changed, thisgrp_len; From b6c7d6dc8aebc04cefd342d6cccd24932be37d12 Mon Sep 17 00:00:00 2001 From: Cai Xinchen Date: Wed, 2 Aug 2023 03:00:25 +0000 Subject: [PATCH 101/186] ext4: remove unused function declaration These functions do not have its function implementation. So those function declaration is useless. Remove these Signed-off-by: Cai Xinchen Link: https://lore.kernel.org/r/20230802030025.173148-1-caixinchen1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 6 ------ 1 file changed, 6 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index fb4d914ea8883..ae458cde55d13 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2697,7 +2697,6 @@ extern ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode, extern int ext4_claim_free_clusters(struct ext4_sb_info *sbi, s64 nclusters, unsigned int flags); extern ext4_fsblk_t ext4_count_free_clusters(struct super_block *); -extern void ext4_check_blocks_bitmap(struct super_block *); extern struct ext4_group_desc * ext4_get_group_desc(struct super_block * sb, ext4_group_t block_group, struct buffer_head ** bh); @@ -2853,7 +2852,6 @@ extern void ext4_free_inode(handle_t *, struct inode *); extern struct inode * ext4_orphan_get(struct super_block *, unsigned long); extern unsigned long ext4_count_free_inodes(struct super_block *); extern unsigned long ext4_count_dirs(struct super_block *); -extern void ext4_check_inodes_bitmap(struct super_block *); extern void ext4_mark_bitmap_end(int start_bit, int end_bit, char *bitmap); extern int ext4_init_inode_table(struct super_block *sb, ext4_group_t group, int barrier); @@ -2896,7 +2894,6 @@ extern int ext4_mb_init(struct super_block *); extern int ext4_mb_release(struct super_block *); extern ext4_fsblk_t ext4_mb_new_blocks(handle_t *, struct ext4_allocation_request *, int *); -extern int ext4_mb_reserve_blocks(struct super_block *, int); extern void ext4_discard_preallocations(struct inode *, unsigned int); extern int __init ext4_init_mballoc(void); extern void ext4_exit_mballoc(void); @@ -2976,7 +2973,6 @@ extern void ext4_evict_inode(struct inode *); extern void ext4_clear_inode(struct inode *); extern int ext4_file_getattr(struct mnt_idmap *, const struct path *, struct kstat *, u32, unsigned int); -extern int ext4_sync_inode(handle_t *, struct inode *); extern void ext4_dirty_inode(struct inode *, int); extern int ext4_change_inode_journal_flag(struct inode *, int); extern int ext4_get_inode_loc(struct inode *, struct ext4_iloc *); @@ -3524,8 +3520,6 @@ extern loff_t ext4_llseek(struct file *file, loff_t offset, int origin); /* inline.c */ extern int ext4_get_max_inline_size(struct inode *inode); extern int ext4_find_inline_data_nolock(struct inode *inode); -extern int ext4_init_inline_data(handle_t *handle, struct inode *inode, - unsigned int len); extern int ext4_destroy_inline_data(handle_t *handle, struct inode *inode); int ext4_readpage_inline(struct inode *inode, struct folio *folio); From 68228da51c9a436872a4ef4b5a7692e29f7e5bc7 Mon Sep 17 00:00:00 2001 From: Wang Jianjian Date: Thu, 3 Aug 2023 00:28:39 +0800 Subject: [PATCH 102/186] ext4: add correct group descriptors and reserved GDT blocks to system zone When setup_system_zone, flex_bg is not initialized so it is always 1. Use a new helper function, ext4_num_base_meta_blocks() which does not depend on sbi->s_log_groups_per_flex being initialized. [ Squashed two patches in the Link URL's below together into a single commit, which is simpler to review/understand. Also fix checkpatch warnings. --TYT ] Cc: stable@kernel.org Signed-off-by: Wang Jianjian Link: https://lore.kernel.org/r/tencent_21AF0D446A9916ED5C51492CC6C9A0A77B05@qq.com Link: https://lore.kernel.org/r/tencent_D744D1450CC169AEA77FCF0A64719909ED05@qq.com Signed-off-by: Theodore Ts'o --- fs/ext4/balloc.c | 15 +++++++++++---- fs/ext4/block_validity.c | 8 ++++---- fs/ext4/ext4.h | 2 ++ 3 files changed, 17 insertions(+), 8 deletions(-) diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 1f72f977c6dba..79b20d6ae39ec 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -913,11 +913,11 @@ unsigned long ext4_bg_num_gdb(struct super_block *sb, ext4_group_t group) } /* - * This function returns the number of file system metadata clusters at + * This function returns the number of file system metadata blocks at * the beginning of a block group, including the reserved gdt blocks. */ -static unsigned ext4_num_base_meta_clusters(struct super_block *sb, - ext4_group_t block_group) +unsigned int ext4_num_base_meta_blocks(struct super_block *sb, + ext4_group_t block_group) { struct ext4_sb_info *sbi = EXT4_SB(sb); unsigned num; @@ -935,8 +935,15 @@ static unsigned ext4_num_base_meta_clusters(struct super_block *sb, } else { /* For META_BG_BLOCK_GROUPS */ num += ext4_bg_num_gdb_meta(sb, block_group); } - return EXT4_NUM_B2C(sbi, num); + return num; } + +static unsigned int ext4_num_base_meta_clusters(struct super_block *sb, + ext4_group_t block_group) +{ + return EXT4_NUM_B2C(EXT4_SB(sb), ext4_num_base_meta_blocks(sb, block_group)); +} + /** * ext4_inode_to_goal_block - return a hint for block allocation * @inode: inode for block allocation diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 5504f72bbbbe7..6fe3c941b5651 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -215,7 +215,6 @@ int ext4_setup_system_zone(struct super_block *sb) struct ext4_system_blocks *system_blks; struct ext4_group_desc *gdp; ext4_group_t i; - int flex_size = ext4_flex_bg_size(sbi); int ret; system_blks = kzalloc(sizeof(*system_blks), GFP_KERNEL); @@ -223,12 +222,13 @@ int ext4_setup_system_zone(struct super_block *sb) return -ENOMEM; for (i=0; i < ngroups; i++) { + unsigned int meta_blks = ext4_num_base_meta_blocks(sb, i); + cond_resched(); - if (ext4_bg_has_super(sb, i) && - ((i < 5) || ((i % flex_size) == 0))) { + if (meta_blks != 0) { ret = add_system_zone(system_blks, ext4_group_first_block_no(sb, i), - ext4_bg_num_gdb(sb, i) + 1, 0); + meta_blks, 0); if (ret) goto err; } diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index ae458cde55d13..2c2c3191bf41d 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -3079,6 +3079,8 @@ extern const char *ext4_decode_error(struct super_block *sb, int errno, extern void ext4_mark_group_bitmap_corrupted(struct super_block *sb, ext4_group_t block_group, unsigned int flags); +extern unsigned int ext4_num_base_meta_blocks(struct super_block *sb, + ext4_group_t block_group); extern __printf(7, 8) void __ext4_error(struct super_block *, const char *, unsigned int, bool, From 194505b55dd7899da114a4d47825204eefc0fff5 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Thu, 10 Aug 2023 12:55:59 -0400 Subject: [PATCH 103/186] ext4: drop dio overwrite only flag and associated warning The commit referenced below opened up concurrent unaligned dio under shared locking for pure overwrites. In doing so, it enabled use of the IOMAP_DIO_OVERWRITE_ONLY flag and added a warning on unexpected -EAGAIN returns as an extra precaution, since ext4 does not retry writes in such cases. The flag itself is advisory in this case since ext4 checks for unaligned I/Os and uses appropriate locking up front, rather than on a retry in response to -EAGAIN. As it turns out, the warning check is susceptible to false positives because there are scenarios where -EAGAIN can be expected from lower layers without necessarily having IOCB_NOWAIT set on the iocb. For example, one instance of the warning has been seen where io_uring sets IOCB_HIPRI, which in turn results in REQ_POLLED|REQ_NOWAIT on the bio. This results in -EAGAIN if the block layer is unable to allocate a request, etc. [Note that there is an outstanding patch to untangle REQ_POLLED and REQ_NOWAIT such that the latter relies on IOCB_NOWAIT, which would also address this instance of the warning.] Another instance of the warning has been reproduced by syzbot. A dio write is interrupted down in __get_user_pages_locked() waiting on the mm lock and returns -EAGAIN up the stack. If the iomap dio iteration layer has made no progress on the write to this point, -EAGAIN returns up to the filesystem and triggers the warning. This use of the overwrite flag in ext4 is precautionary and half-baked. I.e., ext4 doesn't actually implement overwrite checking in the iomap callbacks when the flag is set, so the only extra verification it provides are i_size checks in the generic iomap dio layer. Combined with the tendency for false positives, the added verification is not worth the extra trouble. Remove the flag, associated warning, and update the comments to document when concurrent unaligned dio writes are allowed and why said flag is not used. Cc: stable@kernel.org Reported-by: syzbot+5050ad0fb47527b1808a@syzkaller.appspotmail.com Reported-by: Pengfei Xu Fixes: 310ee0902b8d ("ext4: allow concurrent unaligned dio overwrites") Signed-off-by: Brian Foster Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230810165559.946222-1-bfoster@redhat.com Signed-off-by: Theodore Ts'o --- fs/ext4/file.c | 25 ++++++++++--------------- 1 file changed, 10 insertions(+), 15 deletions(-) diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 2071b1e4322c5..e99cc17b6bd27 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -476,6 +476,11 @@ static ssize_t ext4_dio_write_checks(struct kiocb *iocb, struct iov_iter *from, * required to change security info in file_modified(), for extending * I/O, any form of non-overwrite I/O, and unaligned I/O to unwritten * extents (as partial block zeroing may be required). + * + * Note that unaligned writes are allowed under shared lock so long as + * they are pure overwrites. Otherwise, concurrent unaligned writes risk + * data corruption due to partial block zeroing in the dio layer, and so + * the I/O must occur exclusively. */ if (*ilock_shared && ((!IS_NOSEC(inode) || *extend || !overwrite || @@ -492,21 +497,12 @@ static ssize_t ext4_dio_write_checks(struct kiocb *iocb, struct iov_iter *from, /* * Now that locking is settled, determine dio flags and exclusivity - * requirements. Unaligned writes are allowed under shared lock so long - * as they are pure overwrites. Set the iomap overwrite only flag as an - * added precaution in this case. Even though this is unnecessary, we - * can detect and warn on unexpected -EAGAIN if an unsafe unaligned - * write is ever submitted. - * - * Otherwise, concurrent unaligned writes risk data corruption due to - * partial block zeroing in the dio layer, and so the I/O must occur - * exclusively. The inode lock is already held exclusive if the write is - * non-overwrite or extending, so drain all outstanding dio and set the - * force wait dio flag. + * requirements. We don't use DIO_OVERWRITE_ONLY because we enforce + * behavior already. The inode lock is already held exclusive if the + * write is non-overwrite or extending, so drain all outstanding dio and + * set the force wait dio flag. */ - if (*ilock_shared && unaligned_io) { - *dio_flags = IOMAP_DIO_OVERWRITE_ONLY; - } else if (!*ilock_shared && (unaligned_io || *extend)) { + if (!*ilock_shared && (unaligned_io || *extend)) { if (iocb->ki_flags & IOCB_NOWAIT) { ret = -EAGAIN; goto out; @@ -608,7 +604,6 @@ static ssize_t ext4_dio_write_iter(struct kiocb *iocb, struct iov_iter *from) iomap_ops = &ext4_iomap_overwrite_ops; ret = iomap_dio_rw(iocb, from, iomap_ops, &ext4_dio_write_ops, dio_flags, NULL, 0); - WARN_ON_ONCE(ret == -EAGAIN && !(iocb->ki_flags & IOCB_NOWAIT)); if (ret == -ENOTBLK) ret = 0; From ff0722de896eb278fca193888d22278c28f2782c Mon Sep 17 00:00:00 2001 From: Vitaliy Kuznetsov Date: Thu, 10 Aug 2023 18:38:52 +0400 Subject: [PATCH 104/186] ext4: add periodic superblock update check This patch introduces a mechanism to periodically check and update the superblock within the ext4 file system. The main purpose of this patch is to keep the disk superblock up to date. The update will be performed if more than one hour has passed since the last update, and if more than 16MB of data have been written to disk. This check and update is performed within the ext4_journal_commit_callback function, ensuring that the superblock is written while the disk is active, rather than based on a timer that may trigger during disk idle periods. Discussion https://www.spinics.net/lists/linux-ext4/msg85865.html Signed-off-by: Vitaliy Kuznetsov Link: https://lore.kernel.org/r/20230810143852.40228-1-vk.en.mail@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 62 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 61 insertions(+), 1 deletion(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 6edf7deeb2dbe..bf0cfdffa9d07 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -433,6 +433,57 @@ static time64_t __ext4_get_tstamp(__le32 *lo, __u8 *hi) #define ext4_get_tstamp(es, tstamp) \ __ext4_get_tstamp(&(es)->tstamp, &(es)->tstamp ## _hi) +#define EXT4_SB_REFRESH_INTERVAL_SEC (3600) /* seconds (1 hour) */ +#define EXT4_SB_REFRESH_INTERVAL_KB (16384) /* kilobytes (16MB) */ + +/* + * The ext4_maybe_update_superblock() function checks and updates the + * superblock if needed. + * + * This function is designed to update the on-disk superblock only under + * certain conditions to prevent excessive disk writes and unnecessary + * waking of the disk from sleep. The superblock will be updated if: + * 1. More than an hour has passed since the last superblock update, and + * 2. More than 16MB have been written since the last superblock update. + * + * @sb: The superblock + */ +static void ext4_maybe_update_superblock(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + journal_t *journal = sbi->s_journal; + time64_t now; + __u64 last_update; + __u64 lifetime_write_kbytes; + __u64 diff_size; + + if (sb_rdonly(sb) || !(sb->s_flags & SB_ACTIVE) || + !journal || (journal->j_flags & JBD2_UNMOUNT)) + return; + + now = ktime_get_real_seconds(); + last_update = ext4_get_tstamp(es, s_wtime); + + if (likely(now - last_update < EXT4_SB_REFRESH_INTERVAL_SEC)) + return; + + lifetime_write_kbytes = sbi->s_kbytes_written + + ((part_stat_read(sb->s_bdev, sectors[STAT_WRITE]) - + sbi->s_sectors_written_start) >> 1); + + /* Get the number of kilobytes not written to disk to account + * for statistics and compare with a multiple of 16 MB. This + * is used to determine when the next superblock commit should + * occur (i.e. not more often than once per 16MB if there was + * less written in an hour). + */ + diff_size = lifetime_write_kbytes - le64_to_cpu(es->s_kbytes_written); + + if (diff_size > EXT4_SB_REFRESH_INTERVAL_KB) + schedule_work(&EXT4_SB(sb)->s_error_work); +} + /* * The del_gendisk() function uninitializes the disk-specific data * structures, including the bdi structure, without telling anyone @@ -459,6 +510,7 @@ static void ext4_journal_commit_callback(journal_t *journal, transaction_t *txn) BUG_ON(txn->t_state == T_FINISHED); ext4_process_freed_data(sb, txn->t_tid); + ext4_maybe_update_superblock(sb); spin_lock(&sbi->s_md_lock); while (!list_empty(&txn->t_private_list)) { @@ -715,6 +767,7 @@ static void flush_stashed_error_work(struct work_struct *work) */ if (!sb_rdonly(sbi->s_sb) && journal) { struct buffer_head *sbh = sbi->s_sbh; + bool call_notify_err; handle = jbd2_journal_start(journal, 1); if (IS_ERR(handle)) goto write_directly; @@ -722,6 +775,10 @@ static void flush_stashed_error_work(struct work_struct *work) jbd2_journal_stop(handle); goto write_directly; } + + if (sbi->s_add_error_count > 0) + call_notify_err = true; + ext4_update_super(sbi->s_sb); if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) { ext4_msg(sbi->s_sb, KERN_ERR, "previous I/O error to " @@ -735,7 +792,10 @@ static void flush_stashed_error_work(struct work_struct *work) goto write_directly; } jbd2_journal_stop(handle); - ext4_notify_error_sysfs(sbi); + + if (call_notify_err) + ext4_notify_error_sysfs(sbi); + return; } write_directly: From bb15cea20f211e110150e528fca806f38d5789e0 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Tue, 22 Aug 2023 23:43:38 -0400 Subject: [PATCH 105/186] ext4: rename s_error_work to s_sb_upd_work The most common use that s_error_work will get scheduled is now the periodic update of the superblock. So rename it to s_sb_upd_work. Also rename the function flush_stashed_error_work() to update_super_work(). Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 9 ++++++--- fs/ext4/super.c | 32 ++++++++++++++++---------------- 2 files changed, 22 insertions(+), 19 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 2c2c3191bf41d..84618c46f2390 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1698,10 +1698,13 @@ struct ext4_sb_info { const char *s_last_error_func; time64_t s_last_error_time; /* - * If we are in a context where we cannot update error information in - * the on-disk superblock, we queue this work to do it. + * If we are in a context where we cannot update the on-disk + * superblock, we queue the work here. This is used to update + * the error information in the superblock, and for periodic + * updates of the superblock called from the commit callback + * function. */ - struct work_struct s_error_work; + struct work_struct s_sb_upd_work; /* Ext4 fast commit sub transaction ID */ atomic_t s_fc_subtid; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index bf0cfdffa9d07..91f20afa1d71c 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -481,7 +481,7 @@ static void ext4_maybe_update_superblock(struct super_block *sb) diff_size = lifetime_write_kbytes - le64_to_cpu(es->s_kbytes_written); if (diff_size > EXT4_SB_REFRESH_INTERVAL_KB) - schedule_work(&EXT4_SB(sb)->s_error_work); + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); } /* @@ -723,7 +723,7 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, * defer superblock flushing to a workqueue. */ if (continue_fs && journal) - schedule_work(&EXT4_SB(sb)->s_error_work); + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); else ext4_commit_super(sb); } @@ -750,10 +750,10 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, sb->s_flags |= SB_RDONLY; } -static void flush_stashed_error_work(struct work_struct *work) +static void update_super_work(struct work_struct *work) { struct ext4_sb_info *sbi = container_of(work, struct ext4_sb_info, - s_error_work); + s_sb_upd_work); journal_t *journal = sbi->s_journal; handle_t *handle; @@ -1078,7 +1078,7 @@ __acquires(bitlock) if (!bdev_read_only(sb->s_bdev)) { save_error_info(sb, EFSCORRUPTED, ino, block, function, line); - schedule_work(&EXT4_SB(sb)->s_error_work); + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); } return; } @@ -1318,10 +1318,10 @@ static void ext4_put_super(struct super_block *sb) * Unregister sysfs before destroying jbd2 journal. * Since we could still access attr_journal_task attribute via sysfs * path which could have sbi->s_journal->j_task as NULL - * Unregister sysfs before flush sbi->s_error_work. + * Unregister sysfs before flush sbi->s_sb_upd_work. * Since user may read /proc/fs/ext4/xx/mb_groups during umount, If * read metadata verify failed then will queue error work. - * flush_stashed_error_work will call start_this_handle may trigger + * update_super_work will call start_this_handle may trigger * BUG_ON. */ ext4_unregister_sysfs(sb); @@ -1333,7 +1333,7 @@ static void ext4_put_super(struct super_block *sb) ext4_unregister_li_request(sb); ext4_quotas_off(sb, EXT4_MAXQUOTAS); - flush_work(&sbi->s_error_work); + flush_work(&sbi->s_sb_upd_work); destroy_workqueue(sbi->rsv_conversion_wq); ext4_release_orphan_info(sb); @@ -4998,8 +4998,8 @@ static int ext4_load_and_init_journal(struct super_block *sb, return 0; out: - /* flush s_error_work before journal destroy. */ - flush_work(&sbi->s_error_work); + /* flush s_sb_upd_work before destroying the journal. */ + flush_work(&sbi->s_sb_upd_work); jbd2_journal_destroy(sbi->s_journal); sbi->s_journal = NULL; return -EINVAL; @@ -5322,7 +5322,7 @@ static int __ext4_fill_super(struct fs_context *fc, struct super_block *sb) timer_setup(&sbi->s_err_report, print_daily_error_info, 0); spin_lock_init(&sbi->s_error_lock); - INIT_WORK(&sbi->s_error_work, flush_stashed_error_work); + INIT_WORK(&sbi->s_sb_upd_work, update_super_work); err = ext4_group_desc_init(sb, es, logical_sb_block, &first_not_zeroed); if (err) @@ -5666,16 +5666,16 @@ failed_mount9: __maybe_unused sbi->s_ea_block_cache = NULL; if (sbi->s_journal) { - /* flush s_error_work before journal destroy. */ - flush_work(&sbi->s_error_work); + /* flush s_sb_upd_work before journal destroy. */ + flush_work(&sbi->s_sb_upd_work); jbd2_journal_destroy(sbi->s_journal); sbi->s_journal = NULL; } failed_mount3a: ext4_es_unregister_shrinker(sbi); failed_mount3: - /* flush s_error_work before sbi destroy */ - flush_work(&sbi->s_error_work); + /* flush s_sb_upd_work before sbi destroy */ + flush_work(&sbi->s_sb_upd_work); del_timer_sync(&sbi->s_err_report); ext4_stop_mmpd(sbi); ext4_group_desc_free(sbi); @@ -6551,7 +6551,7 @@ static int __ext4_remount(struct fs_context *fc, struct super_block *sb) } /* Flush outstanding errors before changing fs state */ - flush_work(&sbi->s_error_work); + flush_work(&sbi->s_sb_upd_work); if ((bool)(fc->sb_flags & SB_RDONLY) != sb_rdonly(sb)) { if (ext4_forced_shutdown(sb)) { From 03de20bed203b0819225d4de98353c1f8755a1dd Mon Sep 17 00:00:00 2001 From: Liu Song Date: Thu, 10 Aug 2023 23:43:33 +0800 Subject: [PATCH 106/186] ext4: do not mark inode dirty every time when appending using delalloc In the delalloc append write scenario, if inode's i_size is extended due to buffer write, there are delalloc writes pending in the range up to i_size, and no need to touch i_disksize since writeback will push i_disksize up to i_size eventually. Offers significant performance improvement in high-frequency append write scenarios. I conducted tests in my 32-core environment by launching 32 concurrent threads to append write to the same file. Each write operation had a length of 1024 bytes and was repeated 100000 times. Without using this patch, the test was completed in 7705 ms. However, with this patch, the test was completed in 5066 ms, resulting in a performance improvement of 34%. Moreover, in test scenarios of Kafka version 2.6.2, using packet size of 2K, with this patch resulted in a 10% performance improvement. Signed-off-by: Liu Song Suggested-by: Jan Kara Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230810154333.84921-1-liusong@linux.alibaba.com Signed-off-by: Theodore Ts'o --- fs/ext4/inode.c | 88 ++++++++++++++++++++++++++++++++++--------------- 1 file changed, 62 insertions(+), 26 deletions(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 1b9003840bc16..c5d8f8933c8c6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2935,14 +2935,73 @@ static int ext4_da_should_update_i_disksize(struct folio *folio, return 1; } +static int ext4_da_do_write_end(struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page) +{ + struct inode *inode = mapping->host; + loff_t old_size = inode->i_size; + bool disksize_changed = false; + loff_t new_i_size; + + /* + * block_write_end() will mark the inode as dirty with I_DIRTY_PAGES + * flag, which all that's needed to trigger page writeback. + */ + copied = block_write_end(NULL, mapping, pos, len, copied, page, NULL); + new_i_size = pos + copied; + + /* + * It's important to update i_size while still holding page lock, + * because page writeout could otherwise come in and zero beyond + * i_size. + * + * Since we are holding inode lock, we are sure i_disksize <= + * i_size. We also know that if i_disksize < i_size, there are + * delalloc writes pending in the range up to i_size. If the end of + * the current write is <= i_size, there's no need to touch + * i_disksize since writeback will push i_disksize up to i_size + * eventually. If the end of the current write is > i_size and + * inside an allocated block which ext4_da_should_update_i_disksize() + * checked, we need to update i_disksize here as certain + * ext4_writepages() paths not allocating blocks and update i_disksize. + */ + if (new_i_size > inode->i_size) { + unsigned long end; + + i_size_write(inode, new_i_size); + end = (new_i_size - 1) & (PAGE_SIZE - 1); + if (copied && ext4_da_should_update_i_disksize(page_folio(page), end)) { + ext4_update_i_disksize(inode, new_i_size); + disksize_changed = true; + } + } + + unlock_page(page); + put_page(page); + + if (old_size < pos) + pagecache_isize_extended(inode, old_size, pos); + + if (disksize_changed) { + handle_t *handle; + + handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); + if (IS_ERR(handle)) + return PTR_ERR(handle); + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + } + + return copied; +} + static int ext4_da_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { struct inode *inode = mapping->host; - loff_t new_i_size; - unsigned long start, end; int write_mode = (int)(unsigned long)fsdata; struct folio *folio = page_folio(page); @@ -2961,30 +3020,7 @@ static int ext4_da_write_end(struct file *file, if (unlikely(copied < len) && !PageUptodate(page)) copied = 0; - start = pos & (PAGE_SIZE - 1); - end = start + copied - 1; - - /* - * Since we are holding inode lock, we are sure i_disksize <= - * i_size. We also know that if i_disksize < i_size, there are - * delalloc writes pending in the range upto i_size. If the end of - * the current write is <= i_size, there's no need to touch - * i_disksize since writeback will push i_disksize upto i_size - * eventually. If the end of the current write is > i_size and - * inside an allocated block (ext4_da_should_update_i_disksize() - * check), we need to update i_disksize here as certain - * ext4_writepages() paths not allocating blocks update i_disksize. - * - * Note that we defer inode dirtying to generic_write_end() / - * ext4_da_write_inline_data_end(). - */ - new_i_size = pos + copied; - if (copied && new_i_size > inode->i_size && - ext4_da_should_update_i_disksize(folio, end)) - ext4_update_i_disksize(inode, new_i_size); - - return generic_write_end(file, mapping, pos, len, copied, &folio->page, - fsdata); + return ext4_da_do_write_end(mapping, pos, len, copied, &folio->page); } /* From 0f6bc57971c63f7352c3564d19a5dc707fe8332a Mon Sep 17 00:00:00 2001 From: Ruan Jinjie Date: Sat, 12 Aug 2023 15:18:39 +0800 Subject: [PATCH 107/186] ext4: use LIST_HEAD() to initialize the list_head in mballoc.c Use LIST_HEAD() to initialize the list_head instead of open-coding it. Signed-off-by: Ruan Jinjie Link: https://lore.kernel.org/r/20230812071839.3481909-1-ruanjinjie@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 18 +++++------------- 1 file changed, 5 insertions(+), 13 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 1e4c667812a9d..c91db9f57524c 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -3504,11 +3504,10 @@ static void ext4_discard_work(struct work_struct *work) struct super_block *sb = sbi->s_sb; struct ext4_free_data *fd, *nfd; struct ext4_buddy e4b; - struct list_head discard_list; + LIST_HEAD(discard_list); ext4_group_t grp, load_grp; int err = 0; - INIT_LIST_HEAD(&discard_list); spin_lock(&sbi->s_md_lock); list_splice_init(&sbi->s_discard_list, &discard_list); spin_unlock(&sbi->s_md_lock); @@ -3882,12 +3881,10 @@ void ext4_process_freed_data(struct super_block *sb, tid_t commit_tid) { struct ext4_sb_info *sbi = EXT4_SB(sb); struct ext4_free_data *entry, *tmp; - struct list_head freed_data_list; + LIST_HEAD(freed_data_list); struct list_head *cut_pos = NULL; bool wake; - INIT_LIST_HEAD(&freed_data_list); - spin_lock(&sbi->s_md_lock); list_for_each_entry(entry, &sbi->s_freed_data_list, efd_list) { if (entry->efd_tid != commit_tid) @@ -5414,7 +5411,7 @@ ext4_mb_discard_group_preallocations(struct super_block *sb, struct ext4_group_info *grp = ext4_get_group_info(sb, group); struct buffer_head *bitmap_bh = NULL; struct ext4_prealloc_space *pa, *tmp; - struct list_head list; + LIST_HEAD(list); struct ext4_buddy e4b; struct ext4_inode_info *ei; int err; @@ -5443,7 +5440,6 @@ ext4_mb_discard_group_preallocations(struct super_block *sb, goto out_dbg; } - INIT_LIST_HEAD(&list); ext4_lock_group(sb, group); list_for_each_entry_safe(pa, tmp, &grp->bb_prealloc_list, pa_group_list) { @@ -5524,7 +5520,7 @@ void ext4_discard_preallocations(struct inode *inode, unsigned int needed) struct buffer_head *bitmap_bh = NULL; struct ext4_prealloc_space *pa, *tmp; ext4_group_t group = 0; - struct list_head list; + LIST_HEAD(list); struct ext4_buddy e4b; struct rb_node *iter; int err; @@ -5541,8 +5537,6 @@ void ext4_discard_preallocations(struct inode *inode, unsigned int needed) trace_ext4_discard_preallocations(inode, atomic_read(&ei->i_prealloc_active), needed); - INIT_LIST_HEAD(&list); - if (needed == 0) needed = UINT_MAX; @@ -5858,13 +5852,11 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb, { ext4_group_t group = 0; struct ext4_buddy e4b; - struct list_head discard_list; + LIST_HEAD(discard_list); struct ext4_prealloc_space *pa, *tmp; mb_debug(sb, "discard locality group preallocation\n"); - INIT_LIST_HEAD(&discard_list); - spin_lock(&lg->lg_prealloc_lock); list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[order], pa_node.lg_list, From 8216776ccff6fcd40e3fdaa109aa4150ebe760b3 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Mon, 14 Aug 2023 11:29:01 -0700 Subject: [PATCH 108/186] ext4: reject casefold inode flag without casefold feature It is invalid for the casefold inode flag to be set without the casefold superblock feature flag also being set. e2fsck already considers this case to be invalid and handles it by offering to clear the casefold flag on the inode. __ext4_iget() also already considered this to be invalid, sort of, but it only got so far as logging an error message; it didn't actually reject the inode. Make it reject the inode so that other code doesn't have to handle this case. This matches what f2fs does. Note: we could check 's_encoding != NULL' instead of ext4_has_feature_casefold(). This would make the check robust against the casefold feature being enabled by userspace writing to the page cache of the mounted block device. However, it's unsolvable in general for filesystems to be robust against concurrent writes to the page cache of the mounted block device. Though this very particular scenario involving the casefold feature is solvable, we should not pretend that we can support this model, so let's just check the casefold feature. tune2fs already forbids enabling casefold on a mounted filesystem. Signed-off-by: Eric Biggers Link: https://lore.kernel.org/r/20230814182903.37267-2-ebiggers@kernel.org Signed-off-by: Theodore Ts'o --- fs/ext4/inode.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index c5d8f8933c8c6..6c490f05e2baf 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4974,9 +4974,12 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino, "iget: bogus i_mode (%o)", inode->i_mode); goto bad_inode; } - if (IS_CASEFOLDED(inode) && !ext4_has_feature_casefold(inode->i_sb)) + if (IS_CASEFOLDED(inode) && !ext4_has_feature_casefold(inode->i_sb)) { ext4_error_inode(inode, function, line, 0, "casefold flag without casefold feature"); + ret = -EFSCORRUPTED; + goto bad_inode; + } if ((err_str = check_igot_inode(inode, flags)) != NULL) { ext4_error_inode(inode, function, line, 0, err_str); ret = -EFSCORRUPTED; From b81427939590450172716093dafdda8ef52e020f Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Mon, 14 Aug 2023 11:29:02 -0700 Subject: [PATCH 109/186] ext4: remove redundant checks of s_encoding Now that ext4 does not allow inodes with the casefold flag to be instantiated when unsupported, it's unnecessary to repeatedly check for support later on during random filesystem operations. Signed-off-by: Eric Biggers Link: https://lore.kernel.org/r/20230814182903.37267-3-ebiggers@kernel.org Signed-off-by: Theodore Ts'o --- fs/ext4/hash.c | 2 +- fs/ext4/namei.c | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/ext4/hash.c b/fs/ext4/hash.c index 46c3423ddfa17..deabe29da7fbc 100644 --- a/fs/ext4/hash.c +++ b/fs/ext4/hash.c @@ -300,7 +300,7 @@ int ext4fs_dirhash(const struct inode *dir, const char *name, int len, unsigned char *buff; struct qstr qstr = {.name = name, .len = len }; - if (len && IS_CASEFOLDED(dir) && um && + if (len && IS_CASEFOLDED(dir) && (!IS_ENCRYPTED(dir) || fscrypt_has_encryption_key(dir))) { buff = kzalloc(sizeof(char) * PATH_MAX, GFP_KERNEL); if (!buff) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 9a13431656cd8..c0f0b4e2413b5 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1445,7 +1445,7 @@ int ext4_fname_setup_ci_filename(struct inode *dir, const struct qstr *iname, struct dx_hash_info *hinfo = &name->hinfo; int len; - if (!IS_CASEFOLDED(dir) || !dir->i_sb->s_encoding || + if (!IS_CASEFOLDED(dir) || (IS_ENCRYPTED(dir) && !fscrypt_has_encryption_key(dir))) { cf_name->name = NULL; return 0; @@ -1496,7 +1496,7 @@ static bool ext4_match(struct inode *parent, #endif #if IS_ENABLED(CONFIG_UNICODE) - if (parent->i_sb->s_encoding && IS_CASEFOLDED(parent) && + if (IS_CASEFOLDED(parent) && (!IS_ENCRYPTED(parent) || fscrypt_has_encryption_key(parent))) { if (fname->cf_name.name) { struct qstr cf = {.name = fname->cf_name.name, @@ -2393,7 +2393,7 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, #if IS_ENABLED(CONFIG_UNICODE) if (sb_has_strict_encoding(sb) && IS_CASEFOLDED(dir) && - sb->s_encoding && utf8_validate(sb->s_encoding, &dentry->d_name)) + utf8_validate(sb->s_encoding, &dentry->d_name)) return -EINVAL; #endif From af494af38580a35b92f921639a60630a2307bcc2 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Mon, 14 Aug 2023 11:29:03 -0700 Subject: [PATCH 110/186] libfs: remove redundant checks of s_encoding Now that neither ext4 nor f2fs allows inodes with the casefold flag to be instantiated when unsupported, it's unnecessary to repeatedly check for support later on during random filesystem operations. Signed-off-by: Eric Biggers Link: https://lore.kernel.org/r/20230814182903.37267-4-ebiggers@kernel.org Signed-off-by: Theodore Ts'o --- fs/libfs.c | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/fs/libfs.c b/fs/libfs.c index 5b851315eeed0..5197ea8c66d35 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -1381,16 +1381,6 @@ bool is_empty_dir_inode(struct inode *inode) } #if IS_ENABLED(CONFIG_UNICODE) -/* - * Determine if the name of a dentry should be casefolded. - * - * Return: if names will need casefolding - */ -static bool needs_casefold(const struct inode *dir) -{ - return IS_CASEFOLDED(dir) && dir->i_sb->s_encoding; -} - /** * generic_ci_d_compare - generic d_compare implementation for casefolding filesystems * @dentry: dentry whose name we are checking against @@ -1411,7 +1401,7 @@ static int generic_ci_d_compare(const struct dentry *dentry, unsigned int len, char strbuf[DNAME_INLINE_LEN]; int ret; - if (!dir || !needs_casefold(dir)) + if (!dir || !IS_CASEFOLDED(dir)) goto fallback; /* * If the dentry name is stored in-line, then it may be concurrently @@ -1453,7 +1443,7 @@ static int generic_ci_d_hash(const struct dentry *dentry, struct qstr *str) const struct unicode_map *um = sb->s_encoding; int ret = 0; - if (!dir || !needs_casefold(dir)) + if (!dir || !IS_CASEFOLDED(dir)) return 0; ret = utf8_casefold_hash(um, dentry, str); From 768d612f79822d30a1e7d132a4d4b05337ce42ec Mon Sep 17 00:00:00 2001 From: Baokun Li Date: Tue, 15 Aug 2023 15:08:08 +0800 Subject: [PATCH 111/186] ext4: fix slab-use-after-free in ext4_es_insert_extent() Yikebaer reported an issue: ================================================================== BUG: KASAN: slab-use-after-free in ext4_es_insert_extent+0xc68/0xcb0 fs/ext4/extents_status.c:894 Read of size 4 at addr ffff888112ecc1a4 by task syz-executor/8438 CPU: 1 PID: 8438 Comm: syz-executor Not tainted 6.5.0-rc5 #1 Call Trace: [...] kasan_report+0xba/0xf0 mm/kasan/report.c:588 ext4_es_insert_extent+0xc68/0xcb0 fs/ext4/extents_status.c:894 ext4_map_blocks+0x92a/0x16f0 fs/ext4/inode.c:680 ext4_alloc_file_blocks.isra.0+0x2df/0xb70 fs/ext4/extents.c:4462 ext4_zero_range fs/ext4/extents.c:4622 [inline] ext4_fallocate+0x251c/0x3ce0 fs/ext4/extents.c:4721 [...] Allocated by task 8438: [...] kmem_cache_zalloc include/linux/slab.h:693 [inline] __es_alloc_extent fs/ext4/extents_status.c:469 [inline] ext4_es_insert_extent+0x672/0xcb0 fs/ext4/extents_status.c:873 ext4_map_blocks+0x92a/0x16f0 fs/ext4/inode.c:680 ext4_alloc_file_blocks.isra.0+0x2df/0xb70 fs/ext4/extents.c:4462 ext4_zero_range fs/ext4/extents.c:4622 [inline] ext4_fallocate+0x251c/0x3ce0 fs/ext4/extents.c:4721 [...] Freed by task 8438: [...] kmem_cache_free+0xec/0x490 mm/slub.c:3823 ext4_es_try_to_merge_right fs/ext4/extents_status.c:593 [inline] __es_insert_extent+0x9f4/0x1440 fs/ext4/extents_status.c:802 ext4_es_insert_extent+0x2ca/0xcb0 fs/ext4/extents_status.c:882 ext4_map_blocks+0x92a/0x16f0 fs/ext4/inode.c:680 ext4_alloc_file_blocks.isra.0+0x2df/0xb70 fs/ext4/extents.c:4462 ext4_zero_range fs/ext4/extents.c:4622 [inline] ext4_fallocate+0x251c/0x3ce0 fs/ext4/extents.c:4721 [...] ================================================================== The flow of issue triggering is as follows: 1. remove es raw es es removed es1 |-------------------| -> |----|.......|------| 2. insert es es insert es1 merge with es es1 merge with es and free es1 |----|.......|------| -> |------------|------| -> |-------------------| es merges with newes, then merges with es1, frees es1, then determines if es1->es_len is 0 and triggers a UAF. The code flow is as follows: ext4_es_insert_extent es1 = __es_alloc_extent(true); es2 = __es_alloc_extent(true); __es_remove_extent(inode, lblk, end, NULL, es1) __es_insert_extent(inode, &newes, es1) ---> insert es1 to es tree __es_insert_extent(inode, &newes, es2) ext4_es_try_to_merge_right ext4_es_free_extent(inode, es1) ---> es1 is freed if (es1 && !es1->es_len) // Trigger UAF by determining if es1 is used. We determine whether es1 or es2 is used immediately after calling __es_remove_extent() or __es_insert_extent() to avoid triggering a UAF if es1 or es2 is freed. Reported-by: Yikebaer Aizezi Closes: https://lore.kernel.org/lkml/CALcu4raD4h9coiyEBL4Bm0zjDwxC2CyPiTwsP3zFuhot6y9Beg@mail.gmail.com Fixes: 2a69c450083d ("ext4: using nofail preallocation in ext4_es_insert_extent()") Cc: stable@kernel.org Signed-off-by: Baokun Li Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20230815070808.3377171-1-libaokun1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/extents_status.c | 44 +++++++++++++++++++++++++++------------- 1 file changed, 30 insertions(+), 14 deletions(-) diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 9b5b8951afb44..6f7de14c0fa86 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -878,23 +878,29 @@ void ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, err1 = __es_remove_extent(inode, lblk, end, NULL, es1); if (err1 != 0) goto error; + /* Free preallocated extent if it didn't get used. */ + if (es1) { + if (!es1->es_len) + __es_free_extent(es1); + es1 = NULL; + } err2 = __es_insert_extent(inode, &newes, es2); if (err2 == -ENOMEM && !ext4_es_must_keep(&newes)) err2 = 0; if (err2 != 0) goto error; + /* Free preallocated extent if it didn't get used. */ + if (es2) { + if (!es2->es_len) + __es_free_extent(es2); + es2 = NULL; + } if (sbi->s_cluster_ratio > 1 && test_opt(inode->i_sb, DELALLOC) && (status & EXTENT_STATUS_WRITTEN || status & EXTENT_STATUS_UNWRITTEN)) __revise_pending(inode, lblk, len); - - /* es is pre-allocated but not used, free it. */ - if (es1 && !es1->es_len) - __es_free_extent(es1); - if (es2 && !es2->es_len) - __es_free_extent(es2); error: write_unlock(&EXT4_I(inode)->i_es_lock); if (err1 || err2) @@ -1491,8 +1497,12 @@ void ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, */ write_lock(&EXT4_I(inode)->i_es_lock); err = __es_remove_extent(inode, lblk, end, &reserved, es); - if (es && !es->es_len) - __es_free_extent(es); + /* Free preallocated extent if it didn't get used. */ + if (es) { + if (!es->es_len) + __es_free_extent(es); + es = NULL; + } write_unlock(&EXT4_I(inode)->i_es_lock); if (err) goto retry; @@ -2047,19 +2057,25 @@ void ext4_es_insert_delayed_block(struct inode *inode, ext4_lblk_t lblk, err1 = __es_remove_extent(inode, lblk, lblk, NULL, es1); if (err1 != 0) goto error; + /* Free preallocated extent if it didn't get used. */ + if (es1) { + if (!es1->es_len) + __es_free_extent(es1); + es1 = NULL; + } err2 = __es_insert_extent(inode, &newes, es2); if (err2 != 0) goto error; + /* Free preallocated extent if it didn't get used. */ + if (es2) { + if (!es2->es_len) + __es_free_extent(es2); + es2 = NULL; + } if (allocated) __insert_pending(inode, lblk); - - /* es is pre-allocated but not used, free it. */ - if (es1 && !es1->es_len) - __es_free_extent(es1); - if (es2 && !es2->es_len) - __es_free_extent(es2); error: write_unlock(&EXT4_I(inode)->i_es_lock); if (err1 || err2) From 87098a0d9e42cba2ec49b56dfbf1e4944a6e7bb6 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Fri, 18 Aug 2023 16:02:10 +0100 Subject: [PATCH 112/186] jfs: remove redundant initialization to pointer ip The pointer ip is being initialized with a value that is never read, it is being re-assigned later on. The assignment is redundant and can be removed. Cleans up clang scan warning: fs/jfs/namei.c:886:16: warning: Value stored to 'ip' during its initialization is never read [deadcode.DeadStores] Signed-off-by: Colin Ian King Signed-off-by: Dave Kleikamp --- fs/jfs/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 9b030297aa64c..a61366146cbfe 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -883,7 +883,7 @@ static int jfs_symlink(struct mnt_idmap *idmap, struct inode *dip, struct component_name dname; u32 ssize; /* source pathname size */ struct btstack btstack; - struct inode *ip = d_inode(dentry); + struct inode *ip; s64 xlen = 0; int bmask = 0, xsize; s64 xaddr; From 0225e10972fa809728b8d4c1bd2772b3ec3fdb57 Mon Sep 17 00:00:00 2001 From: Alexei Filippov Date: Sat, 19 Aug 2023 20:32:16 +0300 Subject: [PATCH 113/186] jfs: validate max amount of blocks before allocation. The lack of checking bmp->db_max_freebud in extBalloc() can lead to shift out of bounds, so this patch prevents undefined behavior, because bmp->db_max_freebud == -1 only if there is no free space. Signed-off-by: Aleksei Filippov Signed-off-by: Dave Kleikamp Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-and-tested-by: syzbot+5f088f29593e6b4c8db8@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?id=01abadbd6ae6a08b1f1987aa61554c6b3ac19ff2 --- fs/jfs/jfs_extent.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c index adaa9ad50d4c8..63d21822d309b 100644 --- a/fs/jfs/jfs_extent.c +++ b/fs/jfs/jfs_extent.c @@ -311,6 +311,11 @@ extBalloc(struct inode *ip, s64 hint, s64 * nblocks, s64 * blkno) * blocks in the map. in that case, we'll start off with the * maximum free. */ + + /* give up if no space left */ + if (bmp->db_maxfreebud == -1) + return -ENOSPC; + max = (s64) 1 << bmp->db_maxfreebud; if (*nblocks >= max && *nblocks > nbperpage) nb = nblks = (max > nbperpage) ? max : nbperpage; From 084ba46fc41c21ba827fd92e61f78def7a6e52ea Mon Sep 17 00:00:00 2001 From: Yang Yingliang Date: Tue, 25 Jul 2023 20:31:47 +0800 Subject: [PATCH 114/186] ksmbd: switch to use kmemdup_nul() helper Use kmemdup_nul() helper instead of open-coding to simplify the code. Acked-by: Namjae Jeon Signed-off-by: Yang Yingliang Signed-off-by: Steve French --- fs/smb/server/asn1.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/smb/server/asn1.c b/fs/smb/server/asn1.c index cc6384f796759..4a4b2b03ff33d 100644 --- a/fs/smb/server/asn1.c +++ b/fs/smb/server/asn1.c @@ -214,12 +214,10 @@ static int ksmbd_neg_token_alloc(void *context, size_t hdrlen, { struct ksmbd_conn *conn = context; - conn->mechToken = kmalloc(vlen + 1, GFP_KERNEL); + conn->mechToken = kmemdup_nul(value, vlen, GFP_KERNEL); if (!conn->mechToken) return -ENOMEM; - memcpy(conn->mechToken, value, vlen); - conn->mechToken[vlen] = '\0'; return 0; } From e2b76ab8b5c9327ab2dae6da05d0752eb2f4771d Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Tue, 29 Aug 2023 23:39:31 +0900 Subject: [PATCH 115/186] ksmbd: add support for read compound MacOS sends a compound request including read to the server (e.g. open-read-close). So far, ksmbd has not handled read as a compound request. For compatibility between ksmbd and an OS that supports SMB, This patch provides compound support for read requests. Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/auth.c | 11 +- fs/smb/server/connection.c | 55 +--- fs/smb/server/connection.h | 2 +- fs/smb/server/ksmbd_work.c | 91 +++++- fs/smb/server/ksmbd_work.h | 34 ++- fs/smb/server/oplock.c | 17 +- fs/smb/server/server.c | 8 +- fs/smb/server/smb2pdu.c | 507 ++++++++++++++------------------- fs/smb/server/smb_common.c | 13 +- fs/smb/server/transport_rdma.c | 4 +- fs/smb/server/vfs.c | 4 +- fs/smb/server/vfs.h | 4 +- 12 files changed, 380 insertions(+), 370 deletions(-) diff --git a/fs/smb/server/auth.c b/fs/smb/server/auth.c index 5e5e120edcc22..af7b2cdba1262 100644 --- a/fs/smb/server/auth.c +++ b/fs/smb/server/auth.c @@ -1029,11 +1029,15 @@ static struct scatterlist *ksmbd_init_sg(struct kvec *iov, unsigned int nvec, { struct scatterlist *sg; unsigned int assoc_data_len = sizeof(struct smb2_transform_hdr) - 20; - int i, nr_entries[3] = {0}, total_entries = 0, sg_idx = 0; + int i, *nr_entries, total_entries = 0, sg_idx = 0; if (!nvec) return NULL; + nr_entries = kcalloc(nvec, sizeof(int), GFP_KERNEL); + if (!nr_entries) + return NULL; + for (i = 0; i < nvec - 1; i++) { unsigned long kaddr = (unsigned long)iov[i + 1].iov_base; @@ -1051,8 +1055,10 @@ static struct scatterlist *ksmbd_init_sg(struct kvec *iov, unsigned int nvec, total_entries += 2; sg = kmalloc_array(total_entries, sizeof(struct scatterlist), GFP_KERNEL); - if (!sg) + if (!sg) { + kfree(nr_entries); return NULL; + } sg_init_table(sg, total_entries); smb2_sg_set_buf(&sg[sg_idx++], iov[0].iov_base + 24, assoc_data_len); @@ -1086,6 +1092,7 @@ static struct scatterlist *ksmbd_init_sg(struct kvec *iov, unsigned int nvec, } } smb2_sg_set_buf(&sg[sg_idx], sign, SMB2_SIGNATURE_SIZE); + kfree(nr_entries); return sg; } diff --git a/fs/smb/server/connection.c b/fs/smb/server/connection.c index 2a717d158f02e..0d990c2f33cda 100644 --- a/fs/smb/server/connection.c +++ b/fs/smb/server/connection.c @@ -123,28 +123,22 @@ void ksmbd_conn_enqueue_request(struct ksmbd_work *work) } } -int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) +void ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) { struct ksmbd_conn *conn = work->conn; - int ret = 1; if (list_empty(&work->request_entry) && list_empty(&work->async_request_entry)) - return 0; + return; - if (!work->multiRsp) - atomic_dec(&conn->req_running); - if (!work->multiRsp) { - spin_lock(&conn->request_lock); - list_del_init(&work->request_entry); - spin_unlock(&conn->request_lock); - if (work->asynchronous) - release_async_work(work); - ret = 0; - } + atomic_dec(&conn->req_running); + spin_lock(&conn->request_lock); + list_del_init(&work->request_entry); + spin_unlock(&conn->request_lock); + if (work->asynchronous) + release_async_work(work); wake_up_all(&conn->req_running_q); - return ret; } void ksmbd_conn_lock(struct ksmbd_conn *conn) @@ -193,41 +187,22 @@ void ksmbd_conn_wait_idle(struct ksmbd_conn *conn, u64 sess_id) int ksmbd_conn_write(struct ksmbd_work *work) { struct ksmbd_conn *conn = work->conn; - size_t len = 0; int sent; - struct kvec iov[3]; - int iov_idx = 0; if (!work->response_buf) { pr_err("NULL response header\n"); return -EINVAL; } - if (work->tr_buf) { - iov[iov_idx] = (struct kvec) { work->tr_buf, - sizeof(struct smb2_transform_hdr) + 4 }; - len += iov[iov_idx++].iov_len; - } - - if (work->aux_payload_sz) { - iov[iov_idx] = (struct kvec) { work->response_buf, work->resp_hdr_sz }; - len += iov[iov_idx++].iov_len; - iov[iov_idx] = (struct kvec) { work->aux_payload_buf, work->aux_payload_sz }; - len += iov[iov_idx++].iov_len; - } else { - if (work->tr_buf) - iov[iov_idx].iov_len = work->resp_hdr_sz; - else - iov[iov_idx].iov_len = get_rfc1002_len(work->response_buf) + 4; - iov[iov_idx].iov_base = work->response_buf; - len += iov[iov_idx++].iov_len; - } + if (work->send_no_response) + return 0; ksmbd_conn_lock(conn); - sent = conn->transport->ops->writev(conn->transport, &iov[0], - iov_idx, len, - work->need_invalidate_rkey, - work->remote_key); + sent = conn->transport->ops->writev(conn->transport, work->iov, + work->iov_cnt, + get_rfc1002_len(work->iov[0].iov_base) + 4, + work->need_invalidate_rkey, + work->remote_key); ksmbd_conn_unlock(conn); if (sent < 0) { diff --git a/fs/smb/server/connection.h b/fs/smb/server/connection.h index ad8dfaa48ffb3..ab2583f030ceb 100644 --- a/fs/smb/server/connection.h +++ b/fs/smb/server/connection.h @@ -158,7 +158,7 @@ int ksmbd_conn_rdma_write(struct ksmbd_conn *conn, struct smb2_buffer_desc_v1 *desc, unsigned int desc_len); void ksmbd_conn_enqueue_request(struct ksmbd_work *work); -int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work); +void ksmbd_conn_try_dequeue_request(struct ksmbd_work *work); void ksmbd_conn_init_server_callbacks(struct ksmbd_conn_ops *ops); int ksmbd_conn_handler_loop(void *p); int ksmbd_conn_transport_init(void); diff --git a/fs/smb/server/ksmbd_work.c b/fs/smb/server/ksmbd_work.c index 14b9caebf7a4f..f49c2e01ea9fc 100644 --- a/fs/smb/server/ksmbd_work.c +++ b/fs/smb/server/ksmbd_work.c @@ -27,18 +27,35 @@ struct ksmbd_work *ksmbd_alloc_work_struct(void) INIT_LIST_HEAD(&work->async_request_entry); INIT_LIST_HEAD(&work->fp_entry); INIT_LIST_HEAD(&work->interim_entry); + INIT_LIST_HEAD(&work->aux_read_list); + work->iov_alloc_cnt = 4; + work->iov = kcalloc(work->iov_alloc_cnt, sizeof(struct kvec), + GFP_KERNEL); + if (!work->iov) { + kmem_cache_free(work_cache, work); + work = NULL; + } } return work; } void ksmbd_free_work_struct(struct ksmbd_work *work) { + struct aux_read *ar, *tmp; + WARN_ON(work->saved_cred != NULL); kvfree(work->response_buf); - kvfree(work->aux_payload_buf); + + list_for_each_entry_safe(ar, tmp, &work->aux_read_list, entry) { + kvfree(ar->buf); + list_del(&ar->entry); + kfree(ar); + } + kfree(work->tr_buf); kvfree(work->request_buf); + kfree(work->iov); if (work->async_id) ksmbd_release_id(&work->conn->async_ida, work->async_id); kmem_cache_free(work_cache, work); @@ -77,3 +94,75 @@ bool ksmbd_queue_work(struct ksmbd_work *work) { return queue_work(ksmbd_wq, &work->work); } + +static int ksmbd_realloc_iov_pin(struct ksmbd_work *work, void *ib, + unsigned int ib_len) +{ + + if (work->iov_alloc_cnt <= work->iov_cnt) { + struct kvec *new; + + work->iov_alloc_cnt += 4; + new = krealloc(work->iov, + sizeof(struct kvec) * work->iov_alloc_cnt, + GFP_KERNEL | __GFP_ZERO); + if (!new) + return -ENOMEM; + work->iov = new; + } + + work->iov[++work->iov_idx].iov_base = ib; + work->iov[work->iov_idx].iov_len = ib_len; + work->iov_cnt++; + + return 0; +} + +static int __ksmbd_iov_pin_rsp(struct ksmbd_work *work, void *ib, int len, + void *aux_buf, unsigned int aux_size) +{ + /* Plus rfc_length size on first iov */ + if (!work->iov_idx) { + work->iov[work->iov_idx].iov_base = work->response_buf; + *(__be32 *)work->iov[0].iov_base = 0; + work->iov[work->iov_idx].iov_len = 4; + work->iov_cnt++; + } + + ksmbd_realloc_iov_pin(work, ib, len); + inc_rfc1001_len(work->iov[0].iov_base, len); + + if (aux_size) { + struct aux_read *ar; + + ksmbd_realloc_iov_pin(work, aux_buf, aux_size); + inc_rfc1001_len(work->iov[0].iov_base, aux_size); + + ar = kmalloc(sizeof(struct aux_read), GFP_KERNEL); + if (!ar) + return -ENOMEM; + + ar->buf = aux_buf; + list_add(&ar->entry, &work->aux_read_list); + } + + return 0; +} + +int ksmbd_iov_pin_rsp(struct ksmbd_work *work, void *ib, int len) +{ + return __ksmbd_iov_pin_rsp(work, ib, len, NULL, 0); +} + +int ksmbd_iov_pin_rsp_read(struct ksmbd_work *work, void *ib, int len, + void *aux_buf, unsigned int aux_size) +{ + return __ksmbd_iov_pin_rsp(work, ib, len, aux_buf, aux_size); +} + +void ksmbd_iov_reset(struct ksmbd_work *work) +{ + work->iov_idx = 0; + work->iov_cnt = 0; + *(__be32 *)work->iov[0].iov_base = 0; +} diff --git a/fs/smb/server/ksmbd_work.h b/fs/smb/server/ksmbd_work.h index f8ae6144c0aea..255157eb26dc4 100644 --- a/fs/smb/server/ksmbd_work.h +++ b/fs/smb/server/ksmbd_work.h @@ -19,6 +19,11 @@ enum { KSMBD_WORK_CLOSED, }; +struct aux_read { + void *buf; + struct list_head entry; +}; + /* one of these for every pending CIFS request at the connection */ struct ksmbd_work { /* Server corresponding to this mid */ @@ -31,13 +36,19 @@ struct ksmbd_work { /* Response buffer */ void *response_buf; - /* Read data buffer */ - void *aux_payload_buf; + struct list_head aux_read_list; + + struct kvec *iov; + int iov_alloc_cnt; + int iov_cnt; + int iov_idx; /* Next cmd hdr in compound req buf*/ int next_smb2_rcv_hdr_off; /* Next cmd hdr in compound rsp buf*/ int next_smb2_rsp_hdr_off; + /* Current cmd hdr in compound rsp buf*/ + int curr_smb2_rsp_hdr_off; /* * Current Local FID assigned compound response if SMB2 CREATE @@ -53,16 +64,11 @@ struct ksmbd_work { unsigned int credits_granted; /* response smb header size */ - unsigned int resp_hdr_sz; unsigned int response_sz; - /* Read data count */ - unsigned int aux_payload_sz; void *tr_buf; unsigned char state; - /* Multiple responses for one request e.g. SMB ECHO */ - bool multiRsp:1; /* No response for cancelled request */ bool send_no_response:1; /* Request is encrypted */ @@ -95,6 +101,15 @@ static inline void *ksmbd_resp_buf_next(struct ksmbd_work *work) return work->response_buf + work->next_smb2_rsp_hdr_off + 4; } +/** + * ksmbd_resp_buf_curr - Get current buffer on compound response. + * @work: smb work containing response buffer + */ +static inline void *ksmbd_resp_buf_curr(struct ksmbd_work *work) +{ + return work->response_buf + work->curr_smb2_rsp_hdr_off + 4; +} + /** * ksmbd_req_buf_next - Get next buffer on compound request. * @work: smb work containing response buffer @@ -113,5 +128,8 @@ int ksmbd_work_pool_init(void); int ksmbd_workqueue_init(void); void ksmbd_workqueue_destroy(void); bool ksmbd_queue_work(struct ksmbd_work *work); - +int ksmbd_iov_pin_rsp_read(struct ksmbd_work *work, void *ib, int len, + void *aux_buf, unsigned int aux_size); +int ksmbd_iov_pin_rsp(struct ksmbd_work *work, void *ib, int len); +void ksmbd_iov_reset(struct ksmbd_work *work); #endif /* __KSMBD_WORK_H__ */ diff --git a/fs/smb/server/oplock.c b/fs/smb/server/oplock.c index 844b303baf293..c42b2cff61464 100644 --- a/fs/smb/server/oplock.c +++ b/fs/smb/server/oplock.c @@ -639,7 +639,6 @@ static void __smb2_oplock_break_noti(struct work_struct *wk) { struct smb2_oplock_break *rsp = NULL; struct ksmbd_work *work = container_of(wk, struct ksmbd_work, work); - struct ksmbd_conn *conn = work->conn; struct oplock_break_info *br_info = work->request_buf; struct smb2_hdr *rsp_hdr; struct ksmbd_file *fp; @@ -656,8 +655,6 @@ static void __smb2_oplock_break_noti(struct work_struct *wk) rsp_hdr = smb2_get_msg(work->response_buf); memset(rsp_hdr, 0, sizeof(struct smb2_hdr) + 2); - *(__be32 *)work->response_buf = - cpu_to_be32(conn->vals->header_size); rsp_hdr->ProtocolId = SMB2_PROTO_NUMBER; rsp_hdr->StructureSize = SMB2_HEADER_STRUCTURE_SIZE; rsp_hdr->CreditRequest = cpu_to_le16(0); @@ -684,13 +681,15 @@ static void __smb2_oplock_break_noti(struct work_struct *wk) rsp->PersistentFid = fp->persistent_id; rsp->VolatileFid = fp->volatile_id; - inc_rfc1001_len(work->response_buf, 24); + ksmbd_fd_put(work, fp); + if (ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_oplock_break))) + goto out; ksmbd_debug(OPLOCK, "sending oplock break v_id %llu p_id = %llu lock level = %d\n", rsp->VolatileFid, rsp->PersistentFid, rsp->OplockLevel); - ksmbd_fd_put(work, fp); ksmbd_conn_write(work); out: @@ -751,7 +750,6 @@ static void __smb2_lease_break_noti(struct work_struct *wk) struct smb2_lease_break *rsp = NULL; struct ksmbd_work *work = container_of(wk, struct ksmbd_work, work); struct lease_break_info *br_info = work->request_buf; - struct ksmbd_conn *conn = work->conn; struct smb2_hdr *rsp_hdr; if (allocate_oplock_break_buf(work)) { @@ -761,8 +759,6 @@ static void __smb2_lease_break_noti(struct work_struct *wk) rsp_hdr = smb2_get_msg(work->response_buf); memset(rsp_hdr, 0, sizeof(struct smb2_hdr) + 2); - *(__be32 *)work->response_buf = - cpu_to_be32(conn->vals->header_size); rsp_hdr->ProtocolId = SMB2_PROTO_NUMBER; rsp_hdr->StructureSize = SMB2_HEADER_STRUCTURE_SIZE; rsp_hdr->CreditRequest = cpu_to_le16(0); @@ -791,7 +787,9 @@ static void __smb2_lease_break_noti(struct work_struct *wk) rsp->AccessMaskHint = 0; rsp->ShareMaskHint = 0; - inc_rfc1001_len(work->response_buf, 44); + if (ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_lease_break))) + goto out; ksmbd_conn_write(work); @@ -845,6 +843,7 @@ static int smb2_lease_break_noti(struct oplock_info *opinfo) setup_async_work(in_work, NULL, NULL); smb2_send_interim_resp(in_work, STATUS_PENDING); list_del(&in_work->interim_entry); + ksmbd_iov_reset(in_work); } INIT_WORK(&work->work, __smb2_lease_break_noti); ksmbd_queue_work(work); diff --git a/fs/smb/server/server.c b/fs/smb/server/server.c index 9df121bdf3492..801cd0929209c 100644 --- a/fs/smb/server/server.c +++ b/fs/smb/server/server.c @@ -163,6 +163,7 @@ static void __handle_ksmbd_work(struct ksmbd_work *work, { u16 command = 0; int rc; + bool is_chained = false; if (conn->ops->allocate_rsp_buf(work)) return; @@ -229,14 +230,13 @@ static void __handle_ksmbd_work(struct ksmbd_work *work, } } + is_chained = is_chained_smb2_message(work); + if (work->sess && (work->sess->sign || smb3_11_final_sess_setup_resp(work) || conn->ops->is_sign_req(work, command))) conn->ops->set_sign_rsp(work); - } while (is_chained_smb2_message(work)); - - if (work->send_no_response) - return; + } while (is_chained == true); send: smb3_preauth_hash_rsp(work); diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 7cc1b0c47d0a2..14354e2787d93 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -145,12 +145,18 @@ void smb2_set_err_rsp(struct ksmbd_work *work) err_rsp = smb2_get_msg(work->response_buf); if (err_rsp->hdr.Status != STATUS_STOPPED_ON_SYMLINK) { + int err; + err_rsp->StructureSize = SMB2_ERROR_STRUCTURE_SIZE2_LE; err_rsp->ErrorContextCount = 0; err_rsp->Reserved = 0; err_rsp->ByteCount = 0; err_rsp->ErrorData[0] = 0; - inc_rfc1001_len(work->response_buf, SMB2_ERROR_STRUCTURE_SIZE2); + err = ksmbd_iov_pin_rsp(work, (void *)err_rsp, + work->conn->vals->header_size + + SMB2_ERROR_STRUCTURE_SIZE2); + if (err) + work->send_no_response = 1; } } @@ -245,9 +251,7 @@ int init_smb2_neg_rsp(struct ksmbd_work *work) struct smb2_hdr *rsp_hdr; struct smb2_negotiate_rsp *rsp; struct ksmbd_conn *conn = work->conn; - - *(__be32 *)work->response_buf = - cpu_to_be32(conn->vals->header_size); + int err; rsp_hdr = smb2_get_msg(work->response_buf); memset(rsp_hdr, 0, sizeof(struct smb2_hdr) + 2); @@ -286,12 +290,13 @@ int init_smb2_neg_rsp(struct ksmbd_work *work) rsp->SecurityBufferLength = cpu_to_le16(AUTH_GSS_LENGTH); ksmbd_copy_gss_neg_header((char *)(&rsp->hdr) + le16_to_cpu(rsp->SecurityBufferOffset)); - inc_rfc1001_len(work->response_buf, - sizeof(struct smb2_negotiate_rsp) - - sizeof(struct smb2_hdr) + AUTH_GSS_LENGTH); rsp->SecurityMode = SMB2_NEGOTIATE_SIGNING_ENABLED_LE; if (server_conf.signing == KSMBD_CONFIG_OPT_MANDATORY) rsp->SecurityMode |= SMB2_NEGOTIATE_SIGNING_REQUIRED_LE; + err = ksmbd_iov_pin_rsp(work, rsp, + sizeof(struct smb2_negotiate_rsp) + AUTH_GSS_LENGTH); + if (err) + return err; conn->use_spnego = true; ksmbd_conn_set_need_negotiate(conn); @@ -390,11 +395,12 @@ static void init_chained_smb2_rsp(struct ksmbd_work *work) next_hdr_offset = le32_to_cpu(req->NextCommand); new_len = ALIGN(len, 8); - inc_rfc1001_len(work->response_buf, - sizeof(struct smb2_hdr) + new_len - len); + work->iov[work->iov_idx].iov_len += (new_len - len); + inc_rfc1001_len(work->response_buf, new_len - len); rsp->NextCommand = cpu_to_le32(new_len); work->next_smb2_rcv_hdr_off += next_hdr_offset; + work->curr_smb2_rsp_hdr_off = work->next_smb2_rsp_hdr_off; work->next_smb2_rsp_hdr_off += new_len; ksmbd_debug(SMB, "Compound req new_len = %d rcv off = %d rsp off = %d\n", @@ -470,10 +476,10 @@ bool is_chained_smb2_message(struct ksmbd_work *work) len = len - get_rfc1002_len(work->response_buf); if (len) { ksmbd_debug(SMB, "padding len %u\n", len); + work->iov[work->iov_idx].iov_len += len; inc_rfc1001_len(work->response_buf, len); - if (work->aux_payload_sz) - work->aux_payload_sz += len; } + work->curr_smb2_rsp_hdr_off = work->next_smb2_rsp_hdr_off; } return false; } @@ -488,11 +494,8 @@ int init_smb2_rsp_hdr(struct ksmbd_work *work) { struct smb2_hdr *rsp_hdr = smb2_get_msg(work->response_buf); struct smb2_hdr *rcv_hdr = smb2_get_msg(work->request_buf); - struct ksmbd_conn *conn = work->conn; memset(rsp_hdr, 0, sizeof(struct smb2_hdr) + 2); - *(__be32 *)work->response_buf = - cpu_to_be32(conn->vals->header_size); rsp_hdr->ProtocolId = rcv_hdr->ProtocolId; rsp_hdr->StructureSize = SMB2_HEADER_STRUCTURE_SIZE; rsp_hdr->Command = rcv_hdr->Command; @@ -657,7 +660,7 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) struct ksmbd_conn *conn = work->conn; int id; - rsp_hdr = smb2_get_msg(work->response_buf); + rsp_hdr = ksmbd_resp_buf_next(work); rsp_hdr->Flags |= SMB2_FLAGS_ASYNC_COMMAND; id = ksmbd_acquire_async_msg_id(&conn->async_ida); @@ -707,14 +710,12 @@ void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; - rsp_hdr = smb2_get_msg(work->response_buf); + rsp_hdr = ksmbd_resp_buf_next(work); smb2_set_err_rsp(work); rsp_hdr->Status = status; - work->multiRsp = 1; ksmbd_conn_write(work); rsp_hdr->Status = 0; - work->multiRsp = 0; } static __le32 smb2_get_reparse_tag_special_file(umode_t mode) @@ -821,9 +822,8 @@ static void build_posix_ctxt(struct smb2_posix_neg_context *pneg_ctxt) pneg_ctxt->Name[15] = 0x7C; } -static void assemble_neg_contexts(struct ksmbd_conn *conn, - struct smb2_negotiate_rsp *rsp, - void *smb2_buf_len) +static unsigned int assemble_neg_contexts(struct ksmbd_conn *conn, + struct smb2_negotiate_rsp *rsp) { char * const pneg_ctxt = (char *)rsp + le32_to_cpu(rsp->NegotiateContextOffset); @@ -834,7 +834,6 @@ static void assemble_neg_contexts(struct ksmbd_conn *conn, "assemble SMB2_PREAUTH_INTEGRITY_CAPABILITIES context\n"); build_preauth_ctxt((struct smb2_preauth_neg_context *)pneg_ctxt, conn->preauth_info->Preauth_HashId); - inc_rfc1001_len(smb2_buf_len, AUTH_GSS_PADDING); ctxt_size = sizeof(struct smb2_preauth_neg_context); if (conn->cipher_type) { @@ -874,7 +873,7 @@ static void assemble_neg_contexts(struct ksmbd_conn *conn, } rsp->NegotiateContextCount = cpu_to_le16(neg_ctxt_cnt); - inc_rfc1001_len(smb2_buf_len, ctxt_size); + return ctxt_size + AUTH_GSS_PADDING; } static __le32 decode_preauth_ctxt(struct ksmbd_conn *conn, @@ -1090,7 +1089,7 @@ int smb2_handle_negotiate(struct ksmbd_work *work) struct smb2_negotiate_req *req = smb2_get_msg(work->request_buf); struct smb2_negotiate_rsp *rsp = smb2_get_msg(work->response_buf); int rc = 0; - unsigned int smb2_buf_len, smb2_neg_size; + unsigned int smb2_buf_len, smb2_neg_size, neg_ctxt_len = 0; __le32 status; ksmbd_debug(SMB, "Received negotiate request\n"); @@ -1183,7 +1182,7 @@ int smb2_handle_negotiate(struct ksmbd_work *work) conn->preauth_info->Preauth_HashValue); rsp->NegotiateContextOffset = cpu_to_le32(OFFSET_OF_NEG_CONTEXT); - assemble_neg_contexts(conn, rsp, work->response_buf); + neg_ctxt_len = assemble_neg_contexts(conn, rsp); break; case SMB302_PROT_ID: init_smb3_02_server(conn); @@ -1233,8 +1232,7 @@ int smb2_handle_negotiate(struct ksmbd_work *work) rsp->SecurityBufferLength = cpu_to_le16(AUTH_GSS_LENGTH); ksmbd_copy_gss_neg_header((char *)(&rsp->hdr) + le16_to_cpu(rsp->SecurityBufferOffset)); - inc_rfc1001_len(work->response_buf, sizeof(struct smb2_negotiate_rsp) - - sizeof(struct smb2_hdr) + AUTH_GSS_LENGTH); + rsp->SecurityMode = SMB2_NEGOTIATE_SIGNING_ENABLED_LE; conn->use_spnego = true; @@ -1252,9 +1250,15 @@ int smb2_handle_negotiate(struct ksmbd_work *work) ksmbd_conn_set_need_negotiate(conn); err_out: + if (rc) + rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; + + if (!rc) + rc = ksmbd_iov_pin_rsp(work, rsp, + sizeof(struct smb2_negotiate_rsp) + + AUTH_GSS_LENGTH + neg_ctxt_len); if (rc < 0) smb2_set_err_rsp(work); - return rc; } @@ -1454,7 +1458,6 @@ static int ntlm_authenticate(struct ksmbd_work *work, memcpy((char *)&rsp->hdr.ProtocolId + sz, spnego_blob, spnego_blob_len); rsp->SecurityBufferLength = cpu_to_le16(spnego_blob_len); kfree(spnego_blob); - inc_rfc1001_len(work->response_buf, spnego_blob_len - 1); } user = session_user(conn, req); @@ -1600,7 +1603,6 @@ static int krb5_authenticate(struct ksmbd_work *work, return -EINVAL; } rsp->SecurityBufferLength = cpu_to_le16(out_len); - inc_rfc1001_len(work->response_buf, out_len - 1); if ((conn->sign || server_conf.enforced_signing) || (req->SecurityMode & SMB2_NEGOTIATE_SIGNING_REQUIRED)) @@ -1672,7 +1674,6 @@ int smb2_sess_setup(struct ksmbd_work *work) rsp->SessionFlags = 0; rsp->SecurityBufferOffset = cpu_to_le16(72); rsp->SecurityBufferLength = 0; - inc_rfc1001_len(work->response_buf, 9); ksmbd_conn_lock(conn); if (!req->hdr.SessionId) { @@ -1808,13 +1809,6 @@ int smb2_sess_setup(struct ksmbd_work *work) goto out_err; rsp->hdr.Status = STATUS_MORE_PROCESSING_REQUIRED; - /* - * Note: here total size -1 is done as an - * adjustment for 0 size blob - */ - inc_rfc1001_len(work->response_buf, - le16_to_cpu(rsp->SecurityBufferLength) - 1); - } else if (negblob->MessageType == NtLmAuthenticate) { rc = ntlm_authenticate(work, req, rsp); if (rc) @@ -1899,6 +1893,17 @@ int smb2_sess_setup(struct ksmbd_work *work) ksmbd_conn_set_need_negotiate(conn); } } + } else { + unsigned int iov_len; + + if (rsp->SecurityBufferLength) + iov_len = offsetof(struct smb2_sess_setup_rsp, Buffer) + + le16_to_cpu(rsp->SecurityBufferLength); + else + iov_len = sizeof(struct smb2_sess_setup_rsp); + rc = ksmbd_iov_pin_rsp(work, rsp, iov_len); + if (rc) + rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; } ksmbd_conn_unlock(conn); @@ -1977,13 +1982,16 @@ int smb2_tree_connect(struct ksmbd_work *work) status.tree_conn->posix_extensions = true; rsp->StructureSize = cpu_to_le16(16); - inc_rfc1001_len(work->response_buf, 16); out_err1: rsp->Capabilities = 0; rsp->Reserved = 0; /* default manual caching */ rsp->ShareFlags = SMB2_SHAREFLAG_MANUAL_CACHING; + rc = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_tree_connect_rsp)); + if (rc) + status.ret = KSMBD_TREE_CONN_STATUS_NOMEM; + if (!IS_ERR(treename)) kfree(treename); if (!IS_ERR(name)) @@ -2096,20 +2104,27 @@ int smb2_tree_disconnect(struct ksmbd_work *work) struct smb2_tree_disconnect_req *req; struct ksmbd_session *sess = work->sess; struct ksmbd_tree_connect *tcon = work->tcon; + int err; WORK_BUFFERS(work, req, rsp); - rsp->StructureSize = cpu_to_le16(4); - inc_rfc1001_len(work->response_buf, 4); - ksmbd_debug(SMB, "request\n"); + rsp->StructureSize = cpu_to_le16(4); + err = ksmbd_iov_pin_rsp(work, rsp, + sizeof(struct smb2_tree_disconnect_rsp)); + if (err) { + rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; + smb2_set_err_rsp(work); + return err; + } + if (!tcon || test_and_set_bit(TREE_CONN_EXPIRE, &tcon->status)) { ksmbd_debug(SMB, "Invalid tid %d\n", req->hdr.Id.SyncId.TreeId); rsp->hdr.Status = STATUS_NETWORK_NAME_DELETED; smb2_set_err_rsp(work); - return 0; + return -ENOENT; } ksmbd_close_tree_conn_fds(work); @@ -2131,15 +2146,21 @@ int smb2_session_logoff(struct ksmbd_work *work) struct smb2_logoff_rsp *rsp; struct ksmbd_session *sess; u64 sess_id; + int err; WORK_BUFFERS(work, req, rsp); + ksmbd_debug(SMB, "request\n"); + sess_id = le64_to_cpu(req->hdr.SessionId); rsp->StructureSize = cpu_to_le16(4); - inc_rfc1001_len(work->response_buf, 4); - - ksmbd_debug(SMB, "request\n"); + err = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_logoff_rsp)); + if (err) { + rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; + smb2_set_err_rsp(work); + return err; + } ksmbd_all_conn_set_status(sess_id, KSMBD_SESS_NEED_RECONNECT); ksmbd_close_session_fds(work); @@ -2154,7 +2175,7 @@ int smb2_session_logoff(struct ksmbd_work *work) ksmbd_debug(SMB, "Invalid tid %d\n", req->hdr.Id.SyncId.TreeId); rsp->hdr.Status = STATUS_NETWORK_NAME_DELETED; smb2_set_err_rsp(work); - return 0; + return -ENOENT; } ksmbd_destroy_file_table(&sess->file_table); @@ -2215,7 +2236,10 @@ static noinline int create_smb2_pipe(struct ksmbd_work *work) rsp->CreateContextsOffset = 0; rsp->CreateContextsLength = 0; - inc_rfc1001_len(work->response_buf, 88); /* StructureSize - 1*/ + err = ksmbd_iov_pin_rsp(work, rsp, offsetof(struct smb2_create_rsp, Buffer)); + if (err) + goto out; + kfree(name); return 0; @@ -2597,6 +2621,7 @@ int smb2_open(struct ksmbd_work *work) u64 time; umode_t posix_mode = 0; __le32 daccess, maximal_access = 0; + int iov_len = 0; WORK_BUFFERS(work, req, rsp); @@ -3248,7 +3273,7 @@ int smb2_open(struct ksmbd_work *work) rsp->CreateContextsOffset = 0; rsp->CreateContextsLength = 0; - inc_rfc1001_len(work->response_buf, 88); /* StructureSize - 1*/ + iov_len = offsetof(struct smb2_create_rsp, Buffer); /* If lease is request send lease context response */ if (opinfo && opinfo->is_lease) { @@ -3263,8 +3288,7 @@ int smb2_open(struct ksmbd_work *work) create_lease_buf(rsp->Buffer, opinfo->o_lease); le32_add_cpu(&rsp->CreateContextsLength, conn->vals->create_lease_size); - inc_rfc1001_len(work->response_buf, - conn->vals->create_lease_size); + iov_len += conn->vals->create_lease_size; next_ptr = &lease_ccontext->Next; next_off = conn->vals->create_lease_size; } @@ -3284,8 +3308,7 @@ int smb2_open(struct ksmbd_work *work) le32_to_cpu(maximal_access)); le32_add_cpu(&rsp->CreateContextsLength, conn->vals->create_mxac_size); - inc_rfc1001_len(work->response_buf, - conn->vals->create_mxac_size); + iov_len += conn->vals->create_mxac_size; if (next_ptr) *next_ptr = cpu_to_le32(next_off); next_ptr = &mxac_ccontext->Next; @@ -3303,8 +3326,7 @@ int smb2_open(struct ksmbd_work *work) stat.ino, tcon->id); le32_add_cpu(&rsp->CreateContextsLength, conn->vals->create_disk_id_size); - inc_rfc1001_len(work->response_buf, - conn->vals->create_disk_id_size); + iov_len += conn->vals->create_disk_id_size; if (next_ptr) *next_ptr = cpu_to_le32(next_off); next_ptr = &disk_id_ccontext->Next; @@ -3318,8 +3340,7 @@ int smb2_open(struct ksmbd_work *work) fp); le32_add_cpu(&rsp->CreateContextsLength, conn->vals->create_posix_size); - inc_rfc1001_len(work->response_buf, - conn->vals->create_posix_size); + iov_len += conn->vals->create_posix_size; if (next_ptr) *next_ptr = cpu_to_le32(next_off); } @@ -3337,7 +3358,8 @@ int smb2_open(struct ksmbd_work *work) } ksmbd_revert_fsids(work); err_out1: - + if (!rc) + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, iov_len); if (rc) { if (rc == -EINVAL) rsp->hdr.Status = STATUS_INVALID_PARAMETER; @@ -4063,7 +4085,10 @@ int smb2_query_dir(struct ksmbd_work *work) rsp->OutputBufferOffset = cpu_to_le16(0); rsp->OutputBufferLength = cpu_to_le32(0); rsp->Buffer[0] = 0; - inc_rfc1001_len(work->response_buf, 9); + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_query_directory_rsp)); + if (rc) + goto err_out; } else { no_buf_len: ((struct file_directory_info *) @@ -4075,7 +4100,11 @@ int smb2_query_dir(struct ksmbd_work *work) rsp->StructureSize = cpu_to_le16(9); rsp->OutputBufferOffset = cpu_to_le16(72); rsp->OutputBufferLength = cpu_to_le32(d_info.data_count); - inc_rfc1001_len(work->response_buf, 8 + d_info.data_count); + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_query_directory_rsp, Buffer) + + d_info.data_count); + if (rc) + goto err_out; } kfree(srch_ptr); @@ -4122,21 +4151,13 @@ int smb2_query_dir(struct ksmbd_work *work) */ static int buffer_check_err(int reqOutputBufferLength, struct smb2_query_info_rsp *rsp, - void *rsp_org, int infoclass_size) + void *rsp_org) { if (reqOutputBufferLength < le32_to_cpu(rsp->OutputBufferLength)) { - if (reqOutputBufferLength < infoclass_size) { - pr_err("Invalid Buffer Size Requested\n"); - rsp->hdr.Status = STATUS_INFO_LENGTH_MISMATCH; - *(__be32 *)rsp_org = cpu_to_be32(sizeof(struct smb2_hdr)); - return -EINVAL; - } - - ksmbd_debug(SMB, "Buffer Overflow\n"); - rsp->hdr.Status = STATUS_BUFFER_OVERFLOW; - *(__be32 *)rsp_org = cpu_to_be32(sizeof(struct smb2_hdr) + - reqOutputBufferLength); - rsp->OutputBufferLength = cpu_to_le32(reqOutputBufferLength); + pr_err("Invalid Buffer Size Requested\n"); + rsp->hdr.Status = STATUS_INFO_LENGTH_MISMATCH; + *(__be32 *)rsp_org = cpu_to_be32(sizeof(struct smb2_hdr)); + return -EINVAL; } return 0; } @@ -4155,7 +4176,6 @@ static void get_standard_info_pipe(struct smb2_query_info_rsp *rsp, sinfo->Directory = 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_standard_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_standard_info)); } static void get_internal_info_pipe(struct smb2_query_info_rsp *rsp, u64 num, @@ -4169,7 +4189,6 @@ static void get_internal_info_pipe(struct smb2_query_info_rsp *rsp, u64 num, file_info->IndexNumber = cpu_to_le64(num | (1ULL << 63)); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_internal_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_internal_info)); } static int smb2_get_info_file_pipe(struct ksmbd_session *sess, @@ -4195,14 +4214,12 @@ static int smb2_get_info_file_pipe(struct ksmbd_session *sess, case FILE_STANDARD_INFORMATION: get_standard_info_pipe(rsp, rsp_org); rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), - rsp, rsp_org, - FILE_STANDARD_INFORMATION_SIZE); + rsp, rsp_org); break; case FILE_INTERNAL_INFORMATION: get_internal_info_pipe(rsp, id, rsp_org); rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), - rsp, rsp_org, - FILE_INTERNAL_INFORMATION_SIZE); + rsp, rsp_org); break; default: ksmbd_debug(SMB, "smb2_info_file_pipe for %u not supported\n", @@ -4370,7 +4387,6 @@ static int smb2_get_ea(struct ksmbd_work *work, struct ksmbd_file *fp, if (rsp_data_cnt == 0) rsp->hdr.Status = STATUS_NO_EAS_ON_FILE; rsp->OutputBufferLength = cpu_to_le32(rsp_data_cnt); - inc_rfc1001_len(rsp_org, rsp_data_cnt); out: kvfree(xattr_list); return rc; @@ -4385,7 +4401,6 @@ static void get_file_access_info(struct smb2_query_info_rsp *rsp, file_info->AccessFlags = fp->daccess; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_access_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_access_info)); } static int get_file_basic_info(struct smb2_query_info_rsp *rsp, @@ -4415,7 +4430,6 @@ static int get_file_basic_info(struct smb2_query_info_rsp *rsp, basic_info->Pad1 = 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_basic_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_basic_info)); return 0; } @@ -4440,8 +4454,6 @@ static void get_file_standard_info(struct smb2_query_info_rsp *rsp, sinfo->Directory = S_ISDIR(stat.mode) ? 1 : 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_standard_info)); - inc_rfc1001_len(rsp_org, - sizeof(struct smb2_file_standard_info)); } static void get_file_alignment_info(struct smb2_query_info_rsp *rsp, @@ -4453,8 +4465,6 @@ static void get_file_alignment_info(struct smb2_query_info_rsp *rsp, file_info->AlignmentRequirement = 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_alignment_info)); - inc_rfc1001_len(rsp_org, - sizeof(struct smb2_file_alignment_info)); } static int get_file_all_info(struct ksmbd_work *work, @@ -4518,7 +4528,6 @@ static int get_file_all_info(struct ksmbd_work *work, rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_all_info) + conv_len - 1); kfree(filename); - inc_rfc1001_len(rsp_org, le32_to_cpu(rsp->OutputBufferLength)); return 0; } @@ -4541,7 +4550,6 @@ static void get_file_alternate_info(struct ksmbd_work *work, file_info->FileNameLength = cpu_to_le32(conv_len); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_alt_name_info) + conv_len); - inc_rfc1001_len(rsp_org, le32_to_cpu(rsp->OutputBufferLength)); } static void get_file_stream_info(struct ksmbd_work *work, @@ -4641,7 +4649,6 @@ static void get_file_stream_info(struct ksmbd_work *work, kvfree(xattr_list); rsp->OutputBufferLength = cpu_to_le32(nbytes); - inc_rfc1001_len(rsp_org, nbytes); } static void get_file_internal_info(struct smb2_query_info_rsp *rsp, @@ -4656,7 +4663,6 @@ static void get_file_internal_info(struct smb2_query_info_rsp *rsp, file_info->IndexNumber = cpu_to_le64(stat.ino); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_internal_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_internal_info)); } static int get_file_network_open_info(struct smb2_query_info_rsp *rsp, @@ -4692,7 +4698,6 @@ static int get_file_network_open_info(struct smb2_query_info_rsp *rsp, file_info->Reserved = cpu_to_le32(0); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_ntwrk_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_ntwrk_info)); return 0; } @@ -4704,7 +4709,6 @@ static void get_file_ea_info(struct smb2_query_info_rsp *rsp, void *rsp_org) file_info->EASize = 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_ea_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_ea_info)); } static void get_file_position_info(struct smb2_query_info_rsp *rsp, @@ -4716,7 +4720,6 @@ static void get_file_position_info(struct smb2_query_info_rsp *rsp, file_info->CurrentByteOffset = cpu_to_le64(fp->filp->f_pos); rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_pos_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_pos_info)); } static void get_file_mode_info(struct smb2_query_info_rsp *rsp, @@ -4728,7 +4731,6 @@ static void get_file_mode_info(struct smb2_query_info_rsp *rsp, file_info->Mode = fp->coption & FILE_MODE_INFO_MASK; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_mode_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_mode_info)); } static void get_file_compression_info(struct smb2_query_info_rsp *rsp, @@ -4750,7 +4752,6 @@ static void get_file_compression_info(struct smb2_query_info_rsp *rsp, rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_comp_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_comp_info)); } static int get_file_attribute_tag_info(struct smb2_query_info_rsp *rsp, @@ -4769,11 +4770,10 @@ static int get_file_attribute_tag_info(struct smb2_query_info_rsp *rsp, file_info->ReparseTag = 0; rsp->OutputBufferLength = cpu_to_le32(sizeof(struct smb2_file_attr_tag_info)); - inc_rfc1001_len(rsp_org, sizeof(struct smb2_file_attr_tag_info)); return 0; } -static int find_file_posix_info(struct smb2_query_info_rsp *rsp, +static void find_file_posix_info(struct smb2_query_info_rsp *rsp, struct ksmbd_file *fp, void *rsp_org) { struct smb311_posix_qinfo *file_info; @@ -4811,8 +4811,6 @@ static int find_file_posix_info(struct smb2_query_info_rsp *rsp, SIDUNIX_GROUP, (struct smb_sid *)&file_info->Sids[16]); rsp->OutputBufferLength = cpu_to_le32(out_buf_len); - inc_rfc1001_len(rsp_org, out_buf_len); - return out_buf_len; } static int smb2_get_info_file(struct ksmbd_work *work, @@ -4822,7 +4820,6 @@ static int smb2_get_info_file(struct ksmbd_work *work, struct ksmbd_file *fp; int fileinfoclass = 0; int rc = 0; - int file_infoclass_size; unsigned int id = KSMBD_NO_FID, pid = KSMBD_NO_FID; if (test_share_config_flag(work->tcon->share_conf, @@ -4855,85 +4852,69 @@ static int smb2_get_info_file(struct ksmbd_work *work, switch (fileinfoclass) { case FILE_ACCESS_INFORMATION: get_file_access_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_ACCESS_INFORMATION_SIZE; break; case FILE_BASIC_INFORMATION: rc = get_file_basic_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_BASIC_INFORMATION_SIZE; break; case FILE_STANDARD_INFORMATION: get_file_standard_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_STANDARD_INFORMATION_SIZE; break; case FILE_ALIGNMENT_INFORMATION: get_file_alignment_info(rsp, work->response_buf); - file_infoclass_size = FILE_ALIGNMENT_INFORMATION_SIZE; break; case FILE_ALL_INFORMATION: rc = get_file_all_info(work, rsp, fp, work->response_buf); - file_infoclass_size = FILE_ALL_INFORMATION_SIZE; break; case FILE_ALTERNATE_NAME_INFORMATION: get_file_alternate_info(work, rsp, fp, work->response_buf); - file_infoclass_size = FILE_ALTERNATE_NAME_INFORMATION_SIZE; break; case FILE_STREAM_INFORMATION: get_file_stream_info(work, rsp, fp, work->response_buf); - file_infoclass_size = FILE_STREAM_INFORMATION_SIZE; break; case FILE_INTERNAL_INFORMATION: get_file_internal_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_INTERNAL_INFORMATION_SIZE; break; case FILE_NETWORK_OPEN_INFORMATION: rc = get_file_network_open_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_NETWORK_OPEN_INFORMATION_SIZE; break; case FILE_EA_INFORMATION: get_file_ea_info(rsp, work->response_buf); - file_infoclass_size = FILE_EA_INFORMATION_SIZE; break; case FILE_FULL_EA_INFORMATION: rc = smb2_get_ea(work, fp, req, rsp, work->response_buf); - file_infoclass_size = FILE_FULL_EA_INFORMATION_SIZE; break; case FILE_POSITION_INFORMATION: get_file_position_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_POSITION_INFORMATION_SIZE; break; case FILE_MODE_INFORMATION: get_file_mode_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_MODE_INFORMATION_SIZE; break; case FILE_COMPRESSION_INFORMATION: get_file_compression_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_COMPRESSION_INFORMATION_SIZE; break; case FILE_ATTRIBUTE_TAG_INFORMATION: rc = get_file_attribute_tag_info(rsp, fp, work->response_buf); - file_infoclass_size = FILE_ATTRIBUTE_TAG_INFORMATION_SIZE; break; case SMB_FIND_FILE_POSIX_INFO: if (!work->tcon->posix_extensions) { pr_err("client doesn't negotiate with SMB3.1.1 POSIX Extensions\n"); rc = -EOPNOTSUPP; } else { - file_infoclass_size = find_file_posix_info(rsp, fp, - work->response_buf); + find_file_posix_info(rsp, fp, work->response_buf); } break; default: @@ -4943,8 +4924,7 @@ static int smb2_get_info_file(struct ksmbd_work *work, } if (!rc) rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), - rsp, work->response_buf, - file_infoclass_size); + rsp, work->response_buf); ksmbd_fd_put(work, fp); return rc; } @@ -4960,7 +4940,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, struct kstatfs stfs; struct path path; int rc = 0, len; - int fs_infoclass_size = 0; if (!share->path) return -EIO; @@ -4990,8 +4969,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->DeviceType = cpu_to_le32(stfs.f_type); info->DeviceCharacteristics = cpu_to_le32(0x00000020); rsp->OutputBufferLength = cpu_to_le32(8); - inc_rfc1001_len(work->response_buf, 8); - fs_infoclass_size = FS_DEVICE_INFORMATION_SIZE; break; } case FS_ATTRIBUTE_INFORMATION: @@ -5020,8 +4997,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->FileSystemNameLen = cpu_to_le32(len); sz = sizeof(struct filesystem_attribute_info) - 2 + len; rsp->OutputBufferLength = cpu_to_le32(sz); - inc_rfc1001_len(work->response_buf, sz); - fs_infoclass_size = FS_ATTRIBUTE_INFORMATION_SIZE; break; } case FS_VOLUME_INFORMATION: @@ -5048,8 +5023,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->Reserved = 0; sz = sizeof(struct filesystem_vol_info) - 2 + len; rsp->OutputBufferLength = cpu_to_le32(sz); - inc_rfc1001_len(work->response_buf, sz); - fs_infoclass_size = FS_VOLUME_INFORMATION_SIZE; break; } case FS_SIZE_INFORMATION: @@ -5062,8 +5035,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->SectorsPerAllocationUnit = cpu_to_le32(1); info->BytesPerSector = cpu_to_le32(stfs.f_bsize); rsp->OutputBufferLength = cpu_to_le32(24); - inc_rfc1001_len(work->response_buf, 24); - fs_infoclass_size = FS_SIZE_INFORMATION_SIZE; break; } case FS_FULL_SIZE_INFORMATION: @@ -5079,8 +5050,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->SectorsPerAllocationUnit = cpu_to_le32(1); info->BytesPerSector = cpu_to_le32(stfs.f_bsize); rsp->OutputBufferLength = cpu_to_le32(32); - inc_rfc1001_len(work->response_buf, 32); - fs_infoclass_size = FS_FULL_SIZE_INFORMATION_SIZE; break; } case FS_OBJECT_ID_INFORMATION: @@ -5100,8 +5069,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->extended_info.rel_date = 0; memcpy(info->extended_info.version_string, "1.1.0", strlen("1.1.0")); rsp->OutputBufferLength = cpu_to_le32(64); - inc_rfc1001_len(work->response_buf, 64); - fs_infoclass_size = FS_OBJECT_ID_INFORMATION_SIZE; break; } case FS_SECTOR_SIZE_INFORMATION: @@ -5123,8 +5090,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->ByteOffsetForSectorAlignment = 0; info->ByteOffsetForPartitionAlignment = 0; rsp->OutputBufferLength = cpu_to_le32(28); - inc_rfc1001_len(work->response_buf, 28); - fs_infoclass_size = FS_SECTOR_SIZE_INFORMATION_SIZE; break; } case FS_CONTROL_INFORMATION: @@ -5145,8 +5110,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->DefaultQuotaLimit = cpu_to_le64(SMB2_NO_FID); info->Padding = 0; rsp->OutputBufferLength = cpu_to_le32(48); - inc_rfc1001_len(work->response_buf, 48); - fs_infoclass_size = FS_CONTROL_INFORMATION_SIZE; break; } case FS_POSIX_INFORMATION: @@ -5166,8 +5129,6 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, info->TotalFileNodes = cpu_to_le64(stfs.f_files); info->FreeFileNodes = cpu_to_le64(stfs.f_ffree); rsp->OutputBufferLength = cpu_to_le32(56); - inc_rfc1001_len(work->response_buf, 56); - fs_infoclass_size = FS_POSIX_INFORMATION_SIZE; } break; } @@ -5176,8 +5137,7 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, return -EOPNOTSUPP; } rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), - rsp, work->response_buf, - fs_infoclass_size); + rsp, work->response_buf); path_put(&path); return rc; } @@ -5211,7 +5171,6 @@ static int smb2_get_info_sec(struct ksmbd_work *work, secdesclen = sizeof(struct smb_ntsd); rsp->OutputBufferLength = cpu_to_le32(secdesclen); - inc_rfc1001_len(work->response_buf, secdesclen); return 0; } @@ -5256,7 +5215,6 @@ static int smb2_get_info_sec(struct ksmbd_work *work, return rc; rsp->OutputBufferLength = cpu_to_le32(secdesclen); - inc_rfc1001_len(work->response_buf, secdesclen); return 0; } @@ -5295,6 +5253,14 @@ int smb2_query_info(struct ksmbd_work *work) rc = -EOPNOTSUPP; } + if (!rc) { + rsp->StructureSize = cpu_to_le16(9); + rsp->OutputBufferOffset = cpu_to_le16(72); + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_query_info_rsp, Buffer) + + le32_to_cpu(rsp->OutputBufferLength)); + } + if (rc < 0) { if (rc == -EACCES) rsp->hdr.Status = STATUS_ACCESS_DENIED; @@ -5302,6 +5268,8 @@ int smb2_query_info(struct ksmbd_work *work) rsp->hdr.Status = STATUS_FILE_CLOSED; else if (rc == -EIO) rsp->hdr.Status = STATUS_UNEXPECTED_IO_ERROR; + else if (rc == -ENOMEM) + rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; else if (rc == -EOPNOTSUPP || rsp->hdr.Status == 0) rsp->hdr.Status = STATUS_INVALID_INFO_CLASS; smb2_set_err_rsp(work); @@ -5310,9 +5278,6 @@ int smb2_query_info(struct ksmbd_work *work) rc); return rc; } - rsp->StructureSize = cpu_to_le16(9); - rsp->OutputBufferOffset = cpu_to_le16(72); - inc_rfc1001_len(work->response_buf, 8); return 0; } @@ -5343,8 +5308,9 @@ static noinline int smb2_close_pipe(struct ksmbd_work *work) rsp->AllocationSize = 0; rsp->EndOfFile = 0; rsp->Attributes = 0; - inc_rfc1001_len(work->response_buf, 60); - return 0; + + return ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_close_rsp)); } /** @@ -5449,15 +5415,17 @@ int smb2_close(struct ksmbd_work *work) err = ksmbd_close_fd(work, volatile_id); out: + if (!err) + err = ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_close_rsp)); + if (err) { if (rsp->hdr.Status == 0) rsp->hdr.Status = STATUS_FILE_CLOSED; smb2_set_err_rsp(work); - } else { - inc_rfc1001_len(work->response_buf, 60); } - return 0; + return err; } /** @@ -5475,8 +5443,7 @@ int smb2_echo(struct ksmbd_work *work) rsp->StructureSize = cpu_to_le16(4); rsp->Reserved = 0; - inc_rfc1001_len(work->response_buf, 4); - return 0; + return ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_echo_rsp)); } static int smb2_rename(struct ksmbd_work *work, @@ -6068,7 +6035,10 @@ int smb2_set_info(struct ksmbd_work *work) goto err_out; rsp->StructureSize = cpu_to_le16(2); - inc_rfc1001_len(work->response_buf, 2); + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + sizeof(struct smb2_set_info_rsp)); + if (rc) + goto err_out; ksmbd_fd_put(work, fp); return 0; @@ -6115,28 +6085,36 @@ static noinline int smb2_read_pipe(struct ksmbd_work *work) id = req->VolatileFileId; - inc_rfc1001_len(work->response_buf, 16); rpc_resp = ksmbd_rpc_read(work->sess, id); if (rpc_resp) { + void *aux_payload_buf; + if (rpc_resp->flags != KSMBD_RPC_OK) { err = -EINVAL; goto out; } - work->aux_payload_buf = + aux_payload_buf = kvmalloc(rpc_resp->payload_sz, GFP_KERNEL); - if (!work->aux_payload_buf) { + if (!aux_payload_buf) { err = -ENOMEM; goto out; } - memcpy(work->aux_payload_buf, rpc_resp->payload, - rpc_resp->payload_sz); + memcpy(aux_payload_buf, rpc_resp->payload, rpc_resp->payload_sz); nbytes = rpc_resp->payload_sz; - work->resp_hdr_sz = get_rfc1002_len(work->response_buf) + 4; - work->aux_payload_sz = nbytes; kvfree(rpc_resp); + err = ksmbd_iov_pin_rsp_read(work, (void *)rsp, + offsetof(struct smb2_read_rsp, Buffer), + aux_payload_buf, nbytes); + if (err) + goto out; + } else { + err = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_read_rsp, Buffer)); + if (err) + goto out; } rsp->StructureSize = cpu_to_le16(17); @@ -6145,7 +6123,6 @@ static noinline int smb2_read_pipe(struct ksmbd_work *work) rsp->DataLength = cpu_to_le32(nbytes); rsp->DataRemaining = 0; rsp->Flags = 0; - inc_rfc1001_len(work->response_buf, nbytes); return 0; out: @@ -6219,13 +6196,8 @@ int smb2_read(struct ksmbd_work *work) int err = 0; bool is_rdma_channel = false; unsigned int max_read_size = conn->vals->max_read_size; - - WORK_BUFFERS(work, req, rsp); - if (work->next_smb2_rcv_hdr_off) { - work->send_no_response = 1; - err = -EOPNOTSUPP; - goto out; - } + unsigned int id = KSMBD_NO_FID, pid = KSMBD_NO_FID; + void *aux_payload_buf; if (test_share_config_flag(work->tcon->share_conf, KSMBD_SHARE_FLAG_PIPE)) { @@ -6233,6 +6205,25 @@ int smb2_read(struct ksmbd_work *work) return smb2_read_pipe(work); } + if (work->next_smb2_rcv_hdr_off) { + req = ksmbd_req_buf_next(work); + rsp = ksmbd_resp_buf_next(work); + if (!has_file_id(req->VolatileFileId)) { + ksmbd_debug(SMB, "Compound request set FID = %llu\n", + work->compound_fid); + id = work->compound_fid; + pid = work->compound_pfid; + } + } else { + req = smb2_get_msg(work->request_buf); + rsp = smb2_get_msg(work->response_buf); + } + + if (!has_file_id(id)) { + id = req->VolatileFileId; + pid = req->PersistentFileId; + } + if (req->Channel == SMB2_CHANNEL_RDMA_V1_INVALIDATE || req->Channel == SMB2_CHANNEL_RDMA_V1) { is_rdma_channel = true; @@ -6255,7 +6246,7 @@ int smb2_read(struct ksmbd_work *work) goto out; } - fp = ksmbd_lookup_fd_slow(work, req->VolatileFileId, req->PersistentFileId); + fp = ksmbd_lookup_fd_slow(work, id, pid); if (!fp) { err = -ENOENT; goto out; @@ -6281,21 +6272,20 @@ int smb2_read(struct ksmbd_work *work) ksmbd_debug(SMB, "filename %pD, offset %lld, len %zu\n", fp->filp, offset, length); - work->aux_payload_buf = kvzalloc(length, GFP_KERNEL); - if (!work->aux_payload_buf) { + aux_payload_buf = kvzalloc(length, GFP_KERNEL); + if (!aux_payload_buf) { err = -ENOMEM; goto out; } - nbytes = ksmbd_vfs_read(work, fp, length, &offset); + nbytes = ksmbd_vfs_read(work, fp, length, &offset, aux_payload_buf); if (nbytes < 0) { err = nbytes; goto out; } if ((nbytes == 0 && length != 0) || nbytes < mincount) { - kvfree(work->aux_payload_buf); - work->aux_payload_buf = NULL; + kvfree(aux_payload_buf); rsp->hdr.Status = STATUS_END_OF_FILE; smb2_set_err_rsp(work); ksmbd_fd_put(work, fp); @@ -6308,10 +6298,9 @@ int smb2_read(struct ksmbd_work *work) if (is_rdma_channel == true) { /* write data to the client using rdma channel */ remain_bytes = smb2_read_rdma_channel(work, req, - work->aux_payload_buf, + aux_payload_buf, nbytes); - kvfree(work->aux_payload_buf); - work->aux_payload_buf = NULL; + kvfree(aux_payload_buf); nbytes = 0; if (remain_bytes < 0) { @@ -6326,10 +6315,11 @@ int smb2_read(struct ksmbd_work *work) rsp->DataLength = cpu_to_le32(nbytes); rsp->DataRemaining = cpu_to_le32(remain_bytes); rsp->Flags = 0; - inc_rfc1001_len(work->response_buf, 16); - work->resp_hdr_sz = get_rfc1002_len(work->response_buf) + 4; - work->aux_payload_sz = nbytes; - inc_rfc1001_len(work->response_buf, nbytes); + err = ksmbd_iov_pin_rsp_read(work, (void *)rsp, + offsetof(struct smb2_read_rsp, Buffer), + aux_payload_buf, nbytes); + if (err) + goto out; ksmbd_fd_put(work, fp); return 0; @@ -6412,8 +6402,8 @@ static noinline int smb2_write_pipe(struct ksmbd_work *work) rsp->DataLength = cpu_to_le32(length); rsp->DataRemaining = 0; rsp->Reserved2 = 0; - inc_rfc1001_len(work->response_buf, 16); - return 0; + err = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_write_rsp, Buffer)); out: if (err) { rsp->hdr.Status = STATUS_INVALID_HANDLE; @@ -6569,7 +6559,9 @@ int smb2_write(struct ksmbd_work *work) rsp->DataLength = cpu_to_le32(nbytes); rsp->DataRemaining = 0; rsp->Reserved2 = 0; - inc_rfc1001_len(work->response_buf, 16); + err = ksmbd_iov_pin_rsp(work, rsp, offsetof(struct smb2_write_rsp, Buffer)); + if (err) + goto out; ksmbd_fd_put(work, fp); return 0; @@ -6616,15 +6608,11 @@ int smb2_flush(struct ksmbd_work *work) rsp->StructureSize = cpu_to_le16(4); rsp->Reserved = 0; - inc_rfc1001_len(work->response_buf, 4); - return 0; + return ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_flush_rsp)); out: - if (err) { - rsp->hdr.Status = STATUS_INVALID_HANDLE; - smb2_set_err_rsp(work); - } - + rsp->hdr.Status = STATUS_INVALID_HANDLE; + smb2_set_err_rsp(work); return err; } @@ -7061,6 +7049,8 @@ int smb2_lock(struct ksmbd_work *work) list_del(&work->fp_entry); spin_unlock(&fp->f_lock); + ksmbd_iov_reset(work); + if (work->state != KSMBD_WORK_ACTIVE) { list_del(&smb_lock->llist); spin_lock(&work->conn->llist_lock); @@ -7079,7 +7069,6 @@ int smb2_lock(struct ksmbd_work *work) } init_smb2_rsp_hdr(work); - smb2_set_err_rsp(work); rsp->hdr.Status = STATUS_RANGE_NOT_LOCKED; kfree(smb_lock); @@ -7114,7 +7103,10 @@ int smb2_lock(struct ksmbd_work *work) ksmbd_debug(SMB, "successful in taking lock\n"); rsp->hdr.Status = STATUS_SUCCESS; rsp->Reserved = 0; - inc_rfc1001_len(work->response_buf, 4); + err = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_lock_rsp)); + if (err) + goto out; + ksmbd_fd_put(work, fp); return 0; @@ -7910,9 +7902,9 @@ int smb2_ioctl(struct ksmbd_work *work) rsp->Reserved = cpu_to_le16(0); rsp->Flags = cpu_to_le32(0); rsp->Reserved2 = cpu_to_le32(0); - inc_rfc1001_len(work->response_buf, 48 + nbytes); - - return 0; + ret = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_ioctl_rsp) + nbytes); + if (!ret) + return ret; out: if (ret == -EACCES) @@ -8047,8 +8039,9 @@ static void smb20_oplock_break_ack(struct ksmbd_work *work) rsp->Reserved2 = 0; rsp->VolatileFid = volatile_id; rsp->PersistentFid = persistent_id; - inc_rfc1001_len(work->response_buf, 24); - return; + ret = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_oplock_break)); + if (!ret) + return; err_out: opinfo->op_state = OPLOCK_STATE_NONE; @@ -8198,8 +8191,9 @@ static void smb21_lease_break_ack(struct ksmbd_work *work) memcpy(rsp->LeaseKey, req->LeaseKey, 16); rsp->LeaseState = lease_state; rsp->LeaseDuration = 0; - inc_rfc1001_len(work->response_buf, 36); - return; + ret = ksmbd_iov_pin_rsp(work, rsp, sizeof(struct smb2_lease_ack)); + if (!ret) + return; err_out: opinfo->op_state = OPLOCK_STATE_NONE; @@ -8337,43 +8331,19 @@ int smb2_check_sign_req(struct ksmbd_work *work) void smb2_set_sign_rsp(struct ksmbd_work *work) { struct smb2_hdr *hdr; - struct smb2_hdr *req_hdr; char signature[SMB2_HMACSHA256_SIZE]; - struct kvec iov[2]; - size_t len; + struct kvec *iov; int n_vec = 1; - hdr = smb2_get_msg(work->response_buf); - if (work->next_smb2_rsp_hdr_off) - hdr = ksmbd_resp_buf_next(work); - - req_hdr = ksmbd_req_buf_next(work); - - if (!work->next_smb2_rsp_hdr_off) { - len = get_rfc1002_len(work->response_buf); - if (req_hdr->NextCommand) - len = ALIGN(len, 8); - } else { - len = get_rfc1002_len(work->response_buf) - - work->next_smb2_rsp_hdr_off; - len = ALIGN(len, 8); - } - - if (req_hdr->NextCommand) - hdr->NextCommand = cpu_to_le32(len); - + hdr = ksmbd_resp_buf_curr(work); hdr->Flags |= SMB2_FLAGS_SIGNED; memset(hdr->Signature, 0, SMB2_SIGNATURE_SIZE); - iov[0].iov_base = (char *)&hdr->ProtocolId; - iov[0].iov_len = len; - - if (work->aux_payload_sz) { - iov[0].iov_len -= work->aux_payload_sz; - - iov[1].iov_base = work->aux_payload_buf; - iov[1].iov_len = work->aux_payload_sz; + if (hdr->Command == SMB2_READ) { + iov = &work->iov[work->iov_idx - 1]; n_vec++; + } else { + iov = &work->iov[work->iov_idx]; } if (!ksmbd_sign_smb2_pdu(work->conn, work->sess->sess_key, iov, n_vec, @@ -8449,29 +8419,14 @@ int smb3_check_sign_req(struct ksmbd_work *work) void smb3_set_sign_rsp(struct ksmbd_work *work) { struct ksmbd_conn *conn = work->conn; - struct smb2_hdr *req_hdr, *hdr; + struct smb2_hdr *hdr; struct channel *chann; char signature[SMB2_CMACAES_SIZE]; - struct kvec iov[2]; + struct kvec *iov; int n_vec = 1; - size_t len; char *signing_key; - hdr = smb2_get_msg(work->response_buf); - if (work->next_smb2_rsp_hdr_off) - hdr = ksmbd_resp_buf_next(work); - - req_hdr = ksmbd_req_buf_next(work); - - if (!work->next_smb2_rsp_hdr_off) { - len = get_rfc1002_len(work->response_buf); - if (req_hdr->NextCommand) - len = ALIGN(len, 8); - } else { - len = get_rfc1002_len(work->response_buf) - - work->next_smb2_rsp_hdr_off; - len = ALIGN(len, 8); - } + hdr = ksmbd_resp_buf_curr(work); if (conn->binding == false && le16_to_cpu(hdr->Command) == SMB2_SESSION_SETUP_HE) { @@ -8487,21 +8442,18 @@ void smb3_set_sign_rsp(struct ksmbd_work *work) if (!signing_key) return; - if (req_hdr->NextCommand) - hdr->NextCommand = cpu_to_le32(len); - hdr->Flags |= SMB2_FLAGS_SIGNED; memset(hdr->Signature, 0, SMB2_SIGNATURE_SIZE); - iov[0].iov_base = (char *)&hdr->ProtocolId; - iov[0].iov_len = len; - if (work->aux_payload_sz) { - iov[0].iov_len -= work->aux_payload_sz; - iov[1].iov_base = work->aux_payload_buf; - iov[1].iov_len = work->aux_payload_sz; + + if (hdr->Command == SMB2_READ) { + iov = &work->iov[work->iov_idx - 1]; n_vec++; + } else { + iov = &work->iov[work->iov_idx]; } - if (!ksmbd_sign_smb3_pdu(conn, signing_key, iov, n_vec, signature)) + if (!ksmbd_sign_smb3_pdu(conn, signing_key, iov, n_vec, + signature)) memcpy(hdr->Signature, signature, SMB2_SIGNATURE_SIZE); } @@ -8568,45 +8520,22 @@ static void fill_transform_hdr(void *tr_buf, char *old_buf, __le16 cipher_type) int smb3_encrypt_resp(struct ksmbd_work *work) { - char *buf = work->response_buf; - struct kvec iov[3]; + struct kvec *iov = work->iov; int rc = -ENOMEM; - int buf_size = 0, rq_nvec = 2 + (work->aux_payload_sz ? 1 : 0); + void *tr_buf; - if (ARRAY_SIZE(iov) < rq_nvec) - return -ENOMEM; - - work->tr_buf = kzalloc(sizeof(struct smb2_transform_hdr) + 4, GFP_KERNEL); - if (!work->tr_buf) + tr_buf = kzalloc(sizeof(struct smb2_transform_hdr) + 4, GFP_KERNEL); + if (!tr_buf) return rc; /* fill transform header */ - fill_transform_hdr(work->tr_buf, buf, work->conn->cipher_type); + fill_transform_hdr(tr_buf, work->response_buf, work->conn->cipher_type); - iov[0].iov_base = work->tr_buf; + iov[0].iov_base = tr_buf; iov[0].iov_len = sizeof(struct smb2_transform_hdr) + 4; - buf_size += iov[0].iov_len - 4; - - iov[1].iov_base = buf + 4; - iov[1].iov_len = get_rfc1002_len(buf); - if (work->aux_payload_sz) { - iov[1].iov_len = work->resp_hdr_sz - 4; - - iov[2].iov_base = work->aux_payload_buf; - iov[2].iov_len = work->aux_payload_sz; - buf_size += iov[2].iov_len; - } - buf_size += iov[1].iov_len; - work->resp_hdr_sz = iov[1].iov_len; + work->tr_buf = tr_buf; - rc = ksmbd_crypt_message(work, iov, rq_nvec, 1); - if (rc) - return rc; - - memmove(buf, iov[1].iov_base, iov[1].iov_len); - *(__be32 *)work->tr_buf = cpu_to_be32(buf_size); - - return rc; + return ksmbd_crypt_message(work, iov, work->iov_idx + 1, 1); } bool smb3_is_transform_hdr(void *buf) diff --git a/fs/smb/server/smb_common.c b/fs/smb/server/smb_common.c index c2b75d8988528..e6ba1e9b8589a 100644 --- a/fs/smb/server/smb_common.c +++ b/fs/smb/server/smb_common.c @@ -319,12 +319,6 @@ static int init_smb1_rsp_hdr(struct ksmbd_work *work) struct smb_hdr *rsp_hdr = (struct smb_hdr *)work->response_buf; struct smb_hdr *rcv_hdr = (struct smb_hdr *)work->request_buf; - /* - * Remove 4 byte direct TCP header. - */ - *(__be32 *)work->response_buf = - cpu_to_be32(sizeof(struct smb_hdr) - 4); - rsp_hdr->Command = SMB_COM_NEGOTIATE; *(__le32 *)rsp_hdr->Protocol = SMB1_PROTO_NUMBER; rsp_hdr->Flags = SMBFLG_RESPONSE; @@ -560,10 +554,11 @@ static int smb_handle_negotiate(struct ksmbd_work *work) ksmbd_debug(SMB, "Unsupported SMB1 protocol\n"); - /* Add 2 byte bcc and 2 byte DialectIndex. */ - inc_rfc1001_len(work->response_buf, 4); - neg_rsp->hdr.Status.CifsError = STATUS_SUCCESS; + if (ksmbd_iov_pin_rsp(work, (void *)neg_rsp, + sizeof(struct smb_negotiate_rsp) - 4)) + return -ENOMEM; + neg_rsp->hdr.Status.CifsError = STATUS_SUCCESS; neg_rsp->hdr.WordCount = 1; neg_rsp->DialectIndex = cpu_to_le16(work->conn->dialect); neg_rsp->ByteCount = 0; diff --git a/fs/smb/server/transport_rdma.c b/fs/smb/server/transport_rdma.c index c06efc020bd95..7f222787c52c2 100644 --- a/fs/smb/server/transport_rdma.c +++ b/fs/smb/server/transport_rdma.c @@ -1241,14 +1241,12 @@ static int smb_direct_writev(struct ksmbd_transport *t, //FIXME: skip RFC1002 header.. buflen -= 4; - iov[0].iov_base += 4; - iov[0].iov_len -= 4; remaining_data_length = buflen; ksmbd_debug(RDMA, "Sending smb (RDMA): smb_len=%u\n", buflen); smb_direct_send_ctx_init(st, &send_ctx, need_invalidate, remote_key); - start = i = 0; + start = i = 1; buflen = 0; while (true) { buflen += iov[i].iov_len; diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c index 3d5d652153a5b..cf9007dc2c710 100644 --- a/fs/smb/server/vfs.c +++ b/fs/smb/server/vfs.c @@ -367,15 +367,15 @@ static int check_lock_range(struct file *filp, loff_t start, loff_t end, * @fid: file id of open file * @count: read byte count * @pos: file pos + * @rbuf: read data buffer * * Return: number of read bytes on success, otherwise error */ int ksmbd_vfs_read(struct ksmbd_work *work, struct ksmbd_file *fp, size_t count, - loff_t *pos) + loff_t *pos, char *rbuf) { struct file *filp = fp->filp; ssize_t nbytes = 0; - char *rbuf = work->aux_payload_buf; struct inode *inode = file_inode(filp); if (S_ISDIR(inode->i_mode)) diff --git a/fs/smb/server/vfs.h b/fs/smb/server/vfs.h index 72f9fb4b48d13..00968081856e3 100644 --- a/fs/smb/server/vfs.h +++ b/fs/smb/server/vfs.h @@ -76,8 +76,8 @@ void ksmbd_vfs_query_maximal_access(struct mnt_idmap *idmap, struct dentry *dentry, __le32 *daccess); int ksmbd_vfs_create(struct ksmbd_work *work, const char *name, umode_t mode); int ksmbd_vfs_mkdir(struct ksmbd_work *work, const char *name, umode_t mode); -int ksmbd_vfs_read(struct ksmbd_work *work, struct ksmbd_file *fp, - size_t count, loff_t *pos); +int ksmbd_vfs_read(struct ksmbd_work *work, struct ksmbd_file *fp, size_t count, + loff_t *pos, char *rbuf); int ksmbd_vfs_write(struct ksmbd_work *work, struct ksmbd_file *fp, char *buf, size_t count, loff_t *pos, bool sync, ssize_t *written); From 041bba4414cda37d00063952c9bff9c3d5812a19 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 19 Aug 2023 20:26:17 +0900 Subject: [PATCH 116/186] ksmbd: fix wrong interim response on compound If smb2_lock or smb2_open request is compound, ksmbd could send wrong interim response to client. ksmbd allocate new interim buffer instead of using resonse buffer to support compound request. Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/ksmbd_work.c | 10 ++++++---- fs/smb/server/ksmbd_work.h | 2 +- fs/smb/server/oplock.c | 14 ++------------ fs/smb/server/smb2pdu.c | 26 +++++++++++++++++--------- 4 files changed, 26 insertions(+), 26 deletions(-) diff --git a/fs/smb/server/ksmbd_work.c b/fs/smb/server/ksmbd_work.c index f49c2e01ea9fc..51def3ca74c01 100644 --- a/fs/smb/server/ksmbd_work.c +++ b/fs/smb/server/ksmbd_work.c @@ -160,9 +160,11 @@ int ksmbd_iov_pin_rsp_read(struct ksmbd_work *work, void *ib, int len, return __ksmbd_iov_pin_rsp(work, ib, len, aux_buf, aux_size); } -void ksmbd_iov_reset(struct ksmbd_work *work) +int allocate_interim_rsp_buf(struct ksmbd_work *work) { - work->iov_idx = 0; - work->iov_cnt = 0; - *(__be32 *)work->iov[0].iov_base = 0; + work->response_buf = kzalloc(MAX_CIFS_SMALL_BUFFER_SIZE, GFP_KERNEL); + if (!work->response_buf) + return -ENOMEM; + work->response_sz = MAX_CIFS_SMALL_BUFFER_SIZE; + return 0; } diff --git a/fs/smb/server/ksmbd_work.h b/fs/smb/server/ksmbd_work.h index 255157eb26dc4..8ca2c813246e6 100644 --- a/fs/smb/server/ksmbd_work.h +++ b/fs/smb/server/ksmbd_work.h @@ -131,5 +131,5 @@ bool ksmbd_queue_work(struct ksmbd_work *work); int ksmbd_iov_pin_rsp_read(struct ksmbd_work *work, void *ib, int len, void *aux_buf, unsigned int aux_size); int ksmbd_iov_pin_rsp(struct ksmbd_work *work, void *ib, int len); -void ksmbd_iov_reset(struct ksmbd_work *work); +int allocate_interim_rsp_buf(struct ksmbd_work *work); #endif /* __KSMBD_WORK_H__ */ diff --git a/fs/smb/server/oplock.c b/fs/smb/server/oplock.c index c42b2cff61464..6bc8a1e481712 100644 --- a/fs/smb/server/oplock.c +++ b/fs/smb/server/oplock.c @@ -616,15 +616,6 @@ static int oplock_break_pending(struct oplock_info *opinfo, int req_op_level) return 0; } -static inline int allocate_oplock_break_buf(struct ksmbd_work *work) -{ - work->response_buf = kzalloc(MAX_CIFS_SMALL_BUFFER_SIZE, GFP_KERNEL); - if (!work->response_buf) - return -ENOMEM; - work->response_sz = MAX_CIFS_SMALL_BUFFER_SIZE; - return 0; -} - /** * __smb2_oplock_break_noti() - send smb2 oplock break cmd from conn * to client @@ -647,7 +638,7 @@ static void __smb2_oplock_break_noti(struct work_struct *wk) if (!fp) goto out; - if (allocate_oplock_break_buf(work)) { + if (allocate_interim_rsp_buf(work)) { pr_err("smb2_allocate_rsp_buf failed! "); ksmbd_fd_put(work, fp); goto out; @@ -752,7 +743,7 @@ static void __smb2_lease_break_noti(struct work_struct *wk) struct lease_break_info *br_info = work->request_buf; struct smb2_hdr *rsp_hdr; - if (allocate_oplock_break_buf(work)) { + if (allocate_interim_rsp_buf(work)) { ksmbd_debug(OPLOCK, "smb2_allocate_rsp_buf failed! "); goto out; } @@ -843,7 +834,6 @@ static int smb2_lease_break_noti(struct oplock_info *opinfo) setup_async_work(in_work, NULL, NULL); smb2_send_interim_resp(in_work, STATUS_PENDING); list_del(&in_work->interim_entry); - ksmbd_iov_reset(in_work); } INIT_WORK(&work->work, __smb2_lease_break_noti); ksmbd_queue_work(work); diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 14354e2787d93..95520bf8aa167 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -153,8 +153,8 @@ void smb2_set_err_rsp(struct ksmbd_work *work) err_rsp->ByteCount = 0; err_rsp->ErrorData[0] = 0; err = ksmbd_iov_pin_rsp(work, (void *)err_rsp, - work->conn->vals->header_size + - SMB2_ERROR_STRUCTURE_SIZE2); + __SMB2_HEADER_STRUCTURE_SIZE + + SMB2_ERROR_STRUCTURE_SIZE2); if (err) work->send_no_response = 1; } @@ -709,13 +709,24 @@ void release_async_work(struct ksmbd_work *work) void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; + struct ksmbd_work *in_work = ksmbd_alloc_work_struct(); - rsp_hdr = ksmbd_resp_buf_next(work); - smb2_set_err_rsp(work); + if (allocate_interim_rsp_buf(in_work)) { + pr_err("smb_allocate_rsp_buf failed!\n"); + ksmbd_free_work_struct(in_work); + return; + } + + in_work->conn = work->conn; + memcpy(smb2_get_msg(in_work->response_buf), ksmbd_resp_buf_next(work), + __SMB2_HEADER_STRUCTURE_SIZE); + + rsp_hdr = smb2_get_msg(in_work->response_buf); + smb2_set_err_rsp(in_work); rsp_hdr->Status = status; - ksmbd_conn_write(work); - rsp_hdr->Status = 0; + ksmbd_conn_write(in_work); + ksmbd_free_work_struct(in_work); } static __le32 smb2_get_reparse_tag_special_file(umode_t mode) @@ -7049,8 +7060,6 @@ int smb2_lock(struct ksmbd_work *work) list_del(&work->fp_entry); spin_unlock(&fp->f_lock); - ksmbd_iov_reset(work); - if (work->state != KSMBD_WORK_ACTIVE) { list_del(&smb_lock->llist); spin_lock(&work->conn->llist_lock); @@ -7068,7 +7077,6 @@ int smb2_lock(struct ksmbd_work *work) goto out; } - init_smb2_rsp_hdr(work); rsp->hdr.Status = STATUS_RANGE_NOT_LOCKED; kfree(smb_lock); From 65656f5242e500dcfeffa6a0a1519eae14724f86 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Atte=20Heikkil=C3=A4?= Date: Thu, 10 Aug 2023 23:01:32 +0300 Subject: [PATCH 117/186] ksmbd: fix `force create mode' and `force directory mode' MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit `force create mode' and `force directory mode' should be bitwise ORed with the perms after `create mask' and `directory mask' have been applied, respectively. Signed-off-by: Atte Heikkilä Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/mgmt/share_config.h | 29 +++++++++++------------------ 1 file changed, 11 insertions(+), 18 deletions(-) diff --git a/fs/smb/server/mgmt/share_config.h b/fs/smb/server/mgmt/share_config.h index 3fd3382939421..5f591751b9236 100644 --- a/fs/smb/server/mgmt/share_config.h +++ b/fs/smb/server/mgmt/share_config.h @@ -34,29 +34,22 @@ struct ksmbd_share_config { #define KSMBD_SHARE_INVALID_UID ((__u16)-1) #define KSMBD_SHARE_INVALID_GID ((__u16)-1) -static inline int share_config_create_mode(struct ksmbd_share_config *share, - umode_t posix_mode) +static inline umode_t +share_config_create_mode(struct ksmbd_share_config *share, + umode_t posix_mode) { - if (!share->force_create_mode) { - if (!posix_mode) - return share->create_mask; - else - return posix_mode & share->create_mask; - } - return share->force_create_mode & share->create_mask; + umode_t mode = (posix_mode ?: (umode_t)-1) & share->create_mask; + + return mode | share->force_create_mode; } -static inline int share_config_directory_mode(struct ksmbd_share_config *share, - umode_t posix_mode) +static inline umode_t +share_config_directory_mode(struct ksmbd_share_config *share, + umode_t posix_mode) { - if (!share->force_directory_mode) { - if (!posix_mode) - return share->directory_mask; - else - return posix_mode & share->directory_mask; - } + umode_t mode = (posix_mode ?: (umode_t)-1) & share->directory_mask; - return share->force_directory_mode & share->directory_mask; + return mode | share->force_directory_mode; } static inline int test_share_config_flag(struct ksmbd_share_config *share, From e628bf939aafb61fbc56e9bdac8795cea5127e25 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Tue, 29 Aug 2023 23:40:37 +0900 Subject: [PATCH 118/186] ksmbd: reduce descriptor size if remaining bytes is less than request size Create 3 kinds of files to reproduce this problem. dd if=/dev/urandom of=127k.bin bs=1024 count=127 dd if=/dev/urandom of=128k.bin bs=1024 count=128 dd if=/dev/urandom of=129k.bin bs=1024 count=129 When copying files from ksmbd share to windows or cifs.ko, The following error message happen from windows client. "The file '129k.bin' is too large for the destination filesystem." We can see the error logs from ksmbd debug prints [48394.611537] ksmbd: RDMA r/w request 0x0: token 0x669d, length 0x20000 [48394.612054] ksmbd: smb_direct: RDMA write, len 0x20000, needed credits 0x1 [48394.612572] ksmbd: filename 129k.bin, offset 131072, len 131072 [48394.614189] ksmbd: nbytes 1024, offset 132096 mincount 0 [48394.614585] ksmbd: Failed to process 8 [-22] And we can reproduce it with cifs.ko, e.g. dd if=129k.bin of=/dev/null bs=128KB count=2 This problem is that ksmbd rdma return error if remaining bytes is less than Length of Buffer Descriptor V1 Structure. smb_direct_rdma_xmit() ... if (desc_buf_len == 0 || total_length > buf_len || total_length > t->max_rdma_rw_size) return -EINVAL; This patch reduce descriptor size with remaining bytes and remove the check for total_length and buf_len. Cc: stable@vger.kernel.org Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/transport_rdma.c | 25 ++++++++++++++++++------- 1 file changed, 18 insertions(+), 7 deletions(-) diff --git a/fs/smb/server/transport_rdma.c b/fs/smb/server/transport_rdma.c index 7f222787c52c2..3b269e1f523a1 100644 --- a/fs/smb/server/transport_rdma.c +++ b/fs/smb/server/transport_rdma.c @@ -1364,24 +1364,35 @@ static int smb_direct_rdma_xmit(struct smb_direct_transport *t, LIST_HEAD(msg_list); char *desc_buf; int credits_needed; - unsigned int desc_buf_len; - size_t total_length = 0; + unsigned int desc_buf_len, desc_num = 0; if (t->status != SMB_DIRECT_CS_CONNECTED) return -ENOTCONN; + if (buf_len > t->max_rdma_rw_size) + return -EINVAL; + /* calculate needed credits */ credits_needed = 0; desc_buf = buf; for (i = 0; i < desc_len / sizeof(*desc); i++) { + if (!buf_len) + break; + desc_buf_len = le32_to_cpu(desc[i].length); + if (!desc_buf_len) + return -EINVAL; + + if (desc_buf_len > buf_len) { + desc_buf_len = buf_len; + desc[i].length = cpu_to_le32(desc_buf_len); + buf_len = 0; + } credits_needed += calc_rw_credits(t, desc_buf, desc_buf_len); desc_buf += desc_buf_len; - total_length += desc_buf_len; - if (desc_buf_len == 0 || total_length > buf_len || - total_length > t->max_rdma_rw_size) - return -EINVAL; + buf_len -= desc_buf_len; + desc_num++; } ksmbd_debug(RDMA, "RDMA %s, len %#x, needed credits %#x\n", @@ -1393,7 +1404,7 @@ static int smb_direct_rdma_xmit(struct smb_direct_transport *t, /* build rdma_rw_ctx for each descriptor */ desc_buf = buf; - for (i = 0; i < desc_len / sizeof(*desc); i++) { + for (i = 0; i < desc_num; i++) { msg = kzalloc(offsetof(struct smb_direct_rdma_rw_msg, sg_list) + sizeof(struct scatterlist) * SG_CHUNK_SIZE, GFP_KERNEL); if (!msg) { From bf26f1b4e0918f017775edfeacf6d867204b680b Mon Sep 17 00:00:00 2001 From: Yang Li Date: Mon, 21 Aug 2023 16:29:03 +0800 Subject: [PATCH 119/186] ksmbd: Fix one kernel-doc comment Fix one kernel-doc comment to silence the warning: fs/smb/server/smb2pdu.c:4160: warning: Excess function parameter 'infoclass_size' description in 'buffer_check_err' Signed-off-by: Yang Li Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/smb2pdu.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 95520bf8aa167..2d4b8efaf19fc 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -4156,7 +4156,6 @@ int smb2_query_dir(struct ksmbd_work *work) * @reqOutputBufferLength: max buffer length expected in command response * @rsp: query info response buffer contains output buffer length * @rsp_org: base response buffer pointer in case of chained response - * @infoclass_size: query info class response buffer size * * Return: 0 on success, otherwise error */ From 17d5b135bb720832364e8f55f6a887a3c7ec8fdb Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Fri, 25 Aug 2023 23:39:40 +0900 Subject: [PATCH 120/186] ksmbd: fix wrong DataOffset validation of create context If ->DataOffset of create context is 0, DataBuffer size is not correctly validated. This patch change wrong validation code and consider tag length in request. Cc: stable@vger.kernel.org Reported-by: zdi-disclosures@trendmicro.com # ZDI-CAN-21824 Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/oplock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/smb/server/oplock.c b/fs/smb/server/oplock.c index 6bc8a1e481712..9bc0103720f57 100644 --- a/fs/smb/server/oplock.c +++ b/fs/smb/server/oplock.c @@ -1481,7 +1481,7 @@ struct create_context *smb2_find_context_vals(void *open_req, const char *tag, i name_len < 4 || name_off + name_len > cc_len || (value_off & 0x7) != 0 || - (value_off && (value_off < name_off + name_len)) || + (value_len && value_off < name_off + (name_len < 8 ? 8 : name_len)) || ((u64)value_off + value_len > cc_len)) return ERR_PTR(-EINVAL); From 4b081ce0d830b684fdf967abc3696d1261387254 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Fri, 25 Aug 2023 23:40:31 +0900 Subject: [PATCH 121/186] ksmbd: fix slub overflow in ksmbd_decode_ntlmssp_auth_blob() If authblob->SessionKey.Length is bigger than session key size(CIFS_KEY_SIZE), slub overflow can happen in key exchange codes. cifs_arc4_crypt copy to session key array from SessionKey from client. Cc: stable@vger.kernel.org Reported-by: zdi-disclosures@trendmicro.com # ZDI-CAN-21940 Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/auth.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/smb/server/auth.c b/fs/smb/server/auth.c index af7b2cdba1262..229a6527870d0 100644 --- a/fs/smb/server/auth.c +++ b/fs/smb/server/auth.c @@ -355,6 +355,9 @@ int ksmbd_decode_ntlmssp_auth_blob(struct authenticate_message *authblob, if (blob_len < (u64)sess_key_off + sess_key_len) return -EINVAL; + if (sess_key_len > CIFS_KEY_SIZE) + return -EINVAL; + ctx_arc4 = kmalloc(sizeof(*ctx_arc4), GFP_KERNEL); if (!ctx_arc4) return -ENOMEM; From 0ba5439d9afa2722e7728df56f272c89987540a4 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Fri, 25 Aug 2023 23:41:58 +0900 Subject: [PATCH 122/186] ksmbd: replace one-element array with flex-array member in struct smb2_ea_info UBSAN complains about out-of-bounds array indexes on 1-element arrays in struct smb2_ea_info. UBSAN: array-index-out-of-bounds in fs/smb/server/smb2pdu.c:4335:15 index 1 is out of range for type 'char [1]' CPU: 1 PID: 354 Comm: kworker/1:4 Not tainted 6.5.0-rc4 #1 Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 07/22/2020 Workqueue: ksmbd-io handle_ksmbd_work [ksmbd] Call Trace: __dump_stack linux/lib/dump_stack.c:88 dump_stack_lvl+0x48/0x70 linux/lib/dump_stack.c:106 dump_stack+0x10/0x20 linux/lib/dump_stack.c:113 ubsan_epilogue linux/lib/ubsan.c:217 __ubsan_handle_out_of_bounds+0xc6/0x110 linux/lib/ubsan.c:348 smb2_get_ea linux/fs/smb/server/smb2pdu.c:4335 smb2_get_info_file linux/fs/smb/server/smb2pdu.c:4900 smb2_query_info+0x63ae/0x6b20 linux/fs/smb/server/smb2pdu.c:5275 __process_request linux/fs/smb/server/server.c:145 __handle_ksmbd_work linux/fs/smb/server/server.c:213 handle_ksmbd_work+0x348/0x10b0 linux/fs/smb/server/server.c:266 process_one_work+0x85a/0x1500 linux/kernel/workqueue.c:2597 worker_thread+0xf3/0x13a0 linux/kernel/workqueue.c:2748 kthread+0x2b7/0x390 linux/kernel/kthread.c:389 ret_from_fork+0x44/0x90 linux/arch/x86/kernel/process.c:145 ret_from_fork_asm+0x1b/0x30 linux/arch/x86/entry/entry_64.S:304 Cc: stable@vger.kernel.org Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/smb2pdu.c | 2 +- fs/smb/server/smb2pdu.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 2d4b8efaf19fc..d12d995f52d7e 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -4335,7 +4335,7 @@ static int smb2_get_ea(struct ksmbd_work *work, struct ksmbd_file *fp, if (!strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN)) name_len -= XATTR_USER_PREFIX_LEN; - ptr = (char *)(&eainfo->name + name_len + 1); + ptr = eainfo->name + name_len + 1; buf_free_len -= (offsetof(struct smb2_ea_info, name) + name_len + 1); /* bailout if xattr can't fit in buf_free_len */ diff --git a/fs/smb/server/smb2pdu.h b/fs/smb/server/smb2pdu.h index 2767c08a534a3..d12cfd3b09278 100644 --- a/fs/smb/server/smb2pdu.h +++ b/fs/smb/server/smb2pdu.h @@ -361,7 +361,7 @@ struct smb2_ea_info { __u8 Flags; __u8 EaNameLength; __le16 EaValueLength; - char name[1]; + char name[]; /* optionally followed by value */ } __packed; /* level 15 Query */ From 0e2378eaa2b3a663726cf740d4aaa8a801e2cb31 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Tue, 29 Aug 2023 19:37:40 +0900 Subject: [PATCH 123/186] ksmbd: add missing calling smb2_set_err_rsp() on error If some error happen on smb2_sess_setup(), Need to call smb2_set_err_rsp() to set error response. This patch add missing calling smb2_set_err_rsp() on error. Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/smb2pdu.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index d12d995f52d7e..3dbde9fb775f9 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -1904,6 +1904,7 @@ int smb2_sess_setup(struct ksmbd_work *work) ksmbd_conn_set_need_negotiate(conn); } } + smb2_set_err_rsp(work); } else { unsigned int iov_len; From f16ff1cafbf1e65cc706af912df90bcc15d39a6c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 28 Aug 2023 09:23:00 -0400 Subject: [PATCH 124/186] SUNRPC: Fix the recent bv_offset fix Jeff confirmed his original fix addressed his pynfs test failure, but this same bug also impacted qemu: accessing qcow2 virtual disks using direct I/O was failing. Jeff's fix missed that you have to shorten the bio_vec element by the same amount as you increased the page offset. Reported-by: Maxim Levitsky Fixes: c96e2a695e00 ("sunrpc: set the bv_offset of first bvec in svc_tcp_sendmsg") Tested-by: Maxim Levitsky Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/svcsock.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 2eb8df44f894d..589020ed909dc 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1244,8 +1244,10 @@ static int svc_tcp_sendmsg(struct socket *sock, struct xdr_buf *xdr, if (ret != head->iov_len) goto out; - if (xdr_buf_pagecount(xdr)) + if (xdr_buf_pagecount(xdr)) { xdr->bvec[0].bv_offset = offset_in_page(xdr->page_base); + xdr->bvec[0].bv_len -= offset_in_page(xdr->page_base); + } msg.msg_flags = MSG_SPLICE_PAGES; iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, xdr->bvec, From d67cd907cf8ae2cd42e4f3859ad4de4c16d0c2a3 Mon Sep 17 00:00:00 2001 From: Dai Ngo Date: Thu, 29 Jun 2023 18:52:37 -0700 Subject: [PATCH 125/186] locks: allow support for write delegation Remove the check for F_WRLCK in generic_add_lease to allow file_lock to be used for write delegation. First consumer is NFSD. Signed-off-by: Dai Ngo Signed-off-by: Chuck Lever --- fs/locks.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/fs/locks.c b/fs/locks.c index df8b26a425248..08fb0b4fd4f8f 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -1729,13 +1729,6 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr if (is_deleg && !inode_trylock(inode)) return -EAGAIN; - if (is_deleg && arg == F_WRLCK) { - /* Write delegations are not currently supported: */ - inode_unlock(inode); - WARN_ON_ONCE(1); - return -EINVAL; - } - percpu_down_read(&file_rwsem); spin_lock(&ctx->flc_lock); time_out_leases(inode, &dispose); From fd19ca36fd782b84f71b86525b91a905cda913a4 Mon Sep 17 00:00:00 2001 From: Dai Ngo Date: Thu, 29 Jun 2023 18:52:39 -0700 Subject: [PATCH 126/186] NFSD: handle GETATTR conflict with write delegation If the GETATTR request on a file that has write delegation in effect and the request attributes include the change info and size attribute then the write delegation is recalled. If the delegation is returned within 30ms then the GETATTR is serviced as normal otherwise the NFS4ERR_DELAY error is returned for the GETATTR. Add counter for write delegation recall due to conflict GETATTR. This is used to evaluate the need to implement CB_GETATTR to adoid recalling the delegation with conflit GETATTR. Signed-off-by: Dai Ngo Signed-off-by: Chuck Lever --- fs/nfsd/nfs4state.c | 65 +++++++++++++++++++++++++++++++++++++++++++++ fs/nfsd/nfs4xdr.c | 5 ++++ fs/nfsd/state.h | 3 +++ fs/nfsd/stats.c | 2 ++ fs/nfsd/stats.h | 7 +++++ 5 files changed, 82 insertions(+) diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index daf305daa7516..b56ea72d43501 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -8341,3 +8341,68 @@ nfsd4_get_writestateid(struct nfsd4_compound_state *cstate, { get_stateid(cstate, &u->write.wr_stateid); } + +/** + * nfsd4_deleg_getattr_conflict - Recall if GETATTR causes conflict + * @rqstp: RPC transaction context + * @inode: file to be checked for a conflict + * + * This function is called when there is a conflict between a write + * delegation and a change/size GETATTR from another client. The server + * must either use the CB_GETATTR to get the current values of the + * attributes from the client that holds the delegation or recall the + * delegation before replying to the GETATTR. See RFC 8881 section + * 18.7.4. + * + * The current implementation does not support CB_GETATTR yet. However + * this can avoid recalling the delegation could be added in follow up + * work. + * + * Returns 0 if there is no conflict; otherwise an nfs_stat + * code is returned. + */ +__be32 +nfsd4_deleg_getattr_conflict(struct svc_rqst *rqstp, struct inode *inode) +{ + __be32 status; + struct file_lock_context *ctx; + struct file_lock *fl; + struct nfs4_delegation *dp; + + ctx = locks_inode_context(inode); + if (!ctx) + return 0; + spin_lock(&ctx->flc_lock); + list_for_each_entry(fl, &ctx->flc_lease, fl_list) { + if (fl->fl_flags == FL_LAYOUT) + continue; + if (fl->fl_lmops != &nfsd_lease_mng_ops) { + /* + * non-nfs lease, if it's a lease with F_RDLCK then + * we are done; there isn't any write delegation + * on this inode + */ + if (fl->fl_type == F_RDLCK) + break; + goto break_lease; + } + if (fl->fl_type == F_WRLCK) { + dp = fl->fl_owner; + if (dp->dl_recall.cb_clp == *(rqstp->rq_lease_breaker)) { + spin_unlock(&ctx->flc_lock); + return 0; + } +break_lease: + spin_unlock(&ctx->flc_lock); + nfsd_stats_wdeleg_getattr_inc(); + status = nfserrno(nfsd_open_break_lease(inode, NFSD_MAY_READ)); + if (status != nfserr_jukebox || + !nfsd_wait_for_delegreturn(rqstp, inode)) + return status; + return 0; + } + break; + } + spin_unlock(&ctx->flc_lock); + return 0; +} diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index b30dca7de8cc0..95aafe52fa0c6 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2984,6 +2984,11 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp, if (status) goto out; } + if (bmval0 & (FATTR4_WORD0_CHANGE | FATTR4_WORD0_SIZE)) { + status = nfsd4_deleg_getattr_conflict(rqstp, d_inode(dentry)); + if (status) + goto out; + } err = vfs_getattr(&path, &stat, STATX_BASIC_STATS | STATX_BTIME | STATX_CHANGE_COOKIE, diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index d49d3060ed4f7..cbddcf484dbac 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -732,4 +732,7 @@ static inline bool try_to_expire_client(struct nfs4_client *clp) cmpxchg(&clp->cl_state, NFSD4_COURTESY, NFSD4_EXPIRABLE); return clp->cl_state == NFSD4_EXPIRABLE; } + +extern __be32 nfsd4_deleg_getattr_conflict(struct svc_rqst *rqstp, + struct inode *inode); #endif /* NFSD4_STATE_H */ diff --git a/fs/nfsd/stats.c b/fs/nfsd/stats.c index 777e24e5da33b..63797635e1c32 100644 --- a/fs/nfsd/stats.c +++ b/fs/nfsd/stats.c @@ -65,6 +65,8 @@ static int nfsd_show(struct seq_file *seq, void *v) seq_printf(seq, " %lld", percpu_counter_sum_positive(&nfsdstats.counter[NFSD_STATS_NFS4_OP(i)])); } + seq_printf(seq, "\nwdeleg_getattr %lld", + percpu_counter_sum_positive(&nfsdstats.counter[NFSD_STATS_WDELEG_GETATTR])); seq_putc(seq, '\n'); #endif diff --git a/fs/nfsd/stats.h b/fs/nfsd/stats.h index 9b43dc3d99913..cf5524e7ca062 100644 --- a/fs/nfsd/stats.h +++ b/fs/nfsd/stats.h @@ -22,6 +22,7 @@ enum { NFSD_STATS_FIRST_NFS4_OP, /* count of individual nfsv4 operations */ NFSD_STATS_LAST_NFS4_OP = NFSD_STATS_FIRST_NFS4_OP + LAST_NFS4_OP, #define NFSD_STATS_NFS4_OP(op) (NFSD_STATS_FIRST_NFS4_OP + (op)) + NFSD_STATS_WDELEG_GETATTR, /* count of getattr conflict with wdeleg */ #endif NFSD_STATS_COUNTERS_NUM }; @@ -93,4 +94,10 @@ static inline void nfsd_stats_drc_mem_usage_sub(struct nfsd_net *nn, s64 amount) percpu_counter_sub(&nn->counter[NFSD_NET_DRC_MEM_USAGE], amount); } +#ifdef CONFIG_NFSD_V4 +static inline void nfsd_stats_wdeleg_getattr_inc(void) +{ + percpu_counter_inc(&nfsdstats.counter[NFSD_STATS_WDELEG_GETATTR]); +} +#endif #endif /* _NFSD_STATS_H */ From 50bce06f0e7993aeaa03d39a8f8979b31e5862bd Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 11:33:09 -0400 Subject: [PATCH 127/186] NFSD: Report zero space limit for write delegations Replace the -1 (no limit) with a zero (no reserved space). This prevents certain non-determinant client behavior, such as silly-renaming a file when the only open reference is a write delegation. Such a rename can leave unexpected .nfs files in a directory that is otherwise supposed to be empty. Note that other server implementations that support write delegation also set this field to zero. Suggested-by: Dai Ngo Signed-off-by: Chuck Lever --- fs/nfsd/nfs4xdr.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 95aafe52fa0c6..d4de39404cde8 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3978,17 +3978,20 @@ nfsd4_encode_open(struct nfsd4_compoundres *resp, __be32 nfserr, nfserr = nfsd4_encode_stateid(xdr, &open->op_delegate_stateid); if (nfserr) return nfserr; - p = xdr_reserve_space(xdr, 32); + + p = xdr_reserve_space(xdr, XDR_UNIT * 8); if (!p) return nfserr_resource; *p++ = cpu_to_be32(open->op_recall); /* + * Always flush on close + * * TODO: space_limit's in delegations */ *p++ = cpu_to_be32(NFS4_LIMIT_SIZE); - *p++ = cpu_to_be32(~(u32)0); - *p++ = cpu_to_be32(~(u32)0); + *p++ = xdr_zero; + *p++ = xdr_zero; /* * TODO: ACE's in delegations From 1d3dd1d56ce8322fb5b2a143ec9ff38c703bfeda Mon Sep 17 00:00:00 2001 From: Dai Ngo Date: Thu, 29 Jun 2023 18:52:40 -0700 Subject: [PATCH 128/186] NFSD: Enable write delegation support This patch grants write delegations for OPEN with NFS4_SHARE_ACCESS_WRITE if there is no conflict with other OPENs. Write delegation conflicts with another OPEN, REMOVE, RENAME and SETATTR are handled the same as read delegation using notify_change, try_break_deleg. The NFSv4.0 protocol does not enable a server to determine that a conflicting GETATTR originated from the client holding the delegation versus coming from some other client. With NFSv4.1 and later, the SEQUENCE operation that begins each COMPOUND contains a client ID, so delegation recall can be safely squelched in this case. With NFSv4.0, however, the server must recall or send a CB_GETATTR (per RFC 7530 Section 16.7.5) even when the GETATTR originates from the client holding that delegation. An NFSv4.0 client can trigger a pathological situation if it always sends a DELEGRETURN preceded by a conflicting GETATTR in the same COMPOUND. COMPOUND execution will always stop at the GETATTR and the DELEGRETURN will never get executed. The server eventually revokes the delegation, which can result in loss of open or lock state. Tracepoint added to track whether read or write delegation is granted. Signed-off-by: Dai Ngo Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs4state.c | 97 +++++++++++++++++++++++++++++++++++---------- fs/nfsd/trace.h | 1 + 2 files changed, 78 insertions(+), 20 deletions(-) diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b56ea72d43501..8534693eb6a49 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -649,6 +649,18 @@ find_readable_file(struct nfs4_file *f) return ret; } +static struct nfsd_file * +find_rw_file(struct nfs4_file *f) +{ + struct nfsd_file *ret; + + spin_lock(&f->fi_lock); + ret = nfsd_file_get(f->fi_fds[O_RDWR]); + spin_unlock(&f->fi_lock); + + return ret; +} + struct nfsd_file * find_any_file(struct nfs4_file *f) { @@ -1144,7 +1156,7 @@ static void block_delegations(struct knfsd_fh *fh) static struct nfs4_delegation * alloc_init_deleg(struct nfs4_client *clp, struct nfs4_file *fp, - struct nfs4_clnt_odstate *odstate) + struct nfs4_clnt_odstate *odstate, u32 dl_type) { struct nfs4_delegation *dp; long n; @@ -1170,7 +1182,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_file *fp, INIT_LIST_HEAD(&dp->dl_recall_lru); dp->dl_clnt_odstate = odstate; get_clnt_odstate(odstate); - dp->dl_type = NFS4_OPEN_DELEGATE_READ; + dp->dl_type = dl_type; dp->dl_retries = 1; dp->dl_recalled = false; nfsd4_init_cb(&dp->dl_recall, dp->dl_stid.sc_client, @@ -5449,8 +5461,9 @@ nfs4_set_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, struct nfs4_file *fp = stp->st_stid.sc_file; struct nfs4_clnt_odstate *odstate = stp->st_clnt_odstate; struct nfs4_delegation *dp; - struct nfsd_file *nf; + struct nfsd_file *nf = NULL; struct file_lock *fl; + u32 dl_type; /* * The fi_had_conflict and nfs_get_existing_delegation checks @@ -5460,15 +5473,35 @@ nfs4_set_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, if (fp->fi_had_conflict) return ERR_PTR(-EAGAIN); - nf = find_readable_file(fp); - if (!nf) { - /* - * We probably could attempt another open and get a read - * delegation, but for now, don't bother until the - * client actually sends us one. - */ - return ERR_PTR(-EAGAIN); + /* + * Try for a write delegation first. RFC8881 section 10.4 says: + * + * "An OPEN_DELEGATE_WRITE delegation allows the client to handle, + * on its own, all opens." + * + * Furthermore the client can use a write delegation for most READ + * operations as well, so we require a O_RDWR file here. + * + * Offer a write delegation in the case of a BOTH open, and ensure + * we get the O_RDWR descriptor. + */ + if ((open->op_share_access & NFS4_SHARE_ACCESS_BOTH) == NFS4_SHARE_ACCESS_BOTH) { + nf = find_rw_file(fp); + dl_type = NFS4_OPEN_DELEGATE_WRITE; } + + /* + * If the file is being opened O_RDONLY or we couldn't get a O_RDWR + * file for some reason, then try for a read delegation instead. + */ + if (!nf && (open->op_share_access & NFS4_SHARE_ACCESS_READ)) { + nf = find_readable_file(fp); + dl_type = NFS4_OPEN_DELEGATE_READ; + } + + if (!nf) + return ERR_PTR(-EAGAIN); + spin_lock(&state_lock); spin_lock(&fp->fi_lock); if (nfs4_delegation_exists(clp, fp)) @@ -5491,11 +5524,11 @@ nfs4_set_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, return ERR_PTR(status); status = -ENOMEM; - dp = alloc_init_deleg(clp, fp, odstate); + dp = alloc_init_deleg(clp, fp, odstate, dl_type); if (!dp) goto out_delegees; - fl = nfs4_alloc_init_lease(dp, NFS4_OPEN_DELEGATE_READ); + fl = nfs4_alloc_init_lease(dp, dl_type); if (!fl) goto out_clnt_odstate; @@ -5568,10 +5601,28 @@ static void nfsd4_open_deleg_none_ext(struct nfsd4_open *open, int status) } /* - * Attempt to hand out a delegation. + * The Linux NFS server does not offer write delegations to NFSv4.0 + * clients in order to avoid conflicts between write delegations and + * GETATTRs requesting CHANGE or SIZE attributes. + * + * With NFSv4.1 and later minorversions, the SEQUENCE operation that + * begins each COMPOUND contains a client ID. Delegation recall can + * be avoided when the server recognizes the client sending a + * GETATTR also holds write delegation it conflicts with. + * + * However, the NFSv4.0 protocol does not enable a server to + * determine that a GETATTR originated from the client holding the + * conflicting delegation versus coming from some other client. Per + * RFC 7530 Section 16.7.5, the server must recall or send a + * CB_GETATTR even when the GETATTR originates from the client that + * holds the conflicting delegation. * - * Note we don't support write delegations, and won't until the vfs has - * proper support for them. + * An NFSv4.0 client can trigger a pathological situation if it + * always sends a DELEGRETURN preceded by a conflicting GETATTR in + * the same COMPOUND. COMPOUND execution will always stop at the + * GETATTR and the DELEGRETURN will never get executed. The server + * eventually revokes the delegation, which can result in loss of + * open or lock state. */ static void nfs4_open_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, @@ -5590,8 +5641,6 @@ nfs4_open_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, case NFS4_OPEN_CLAIM_PREVIOUS: if (!cb_up) open->op_recall = 1; - if (open->op_delegate_type != NFS4_OPEN_DELEGATE_READ) - goto out_no_deleg; break; case NFS4_OPEN_CLAIM_NULL: parent = currentfh; @@ -5606,6 +5655,9 @@ nfs4_open_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, goto out_no_deleg; if (!cb_up || !(oo->oo_flags & NFS4_OO_CONFIRMED)) goto out_no_deleg; + if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE && + !clp->cl_minorversion) + goto out_no_deleg; break; default: goto out_no_deleg; @@ -5616,8 +5668,13 @@ nfs4_open_delegation(struct nfsd4_open *open, struct nfs4_ol_stateid *stp, memcpy(&open->op_delegate_stateid, &dp->dl_stid.sc_stateid, sizeof(dp->dl_stid.sc_stateid)); - trace_nfsd_deleg_read(&dp->dl_stid.sc_stateid); - open->op_delegate_type = NFS4_OPEN_DELEGATE_READ; + if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE) { + open->op_delegate_type = NFS4_OPEN_DELEGATE_WRITE; + trace_nfsd_deleg_write(&dp->dl_stid.sc_stateid); + } else { + open->op_delegate_type = NFS4_OPEN_DELEGATE_READ; + trace_nfsd_deleg_read(&dp->dl_stid.sc_stateid); + } nfs4_put_stid(&dp->dl_stid); return; out_no_deleg: diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h index 2af74983f1461..693fe6d465aa6 100644 --- a/fs/nfsd/trace.h +++ b/fs/nfsd/trace.h @@ -607,6 +607,7 @@ DEFINE_STATEID_EVENT(layout_recall_release); DEFINE_STATEID_EVENT(open); DEFINE_STATEID_EVENT(deleg_read); +DEFINE_STATEID_EVENT(deleg_write); DEFINE_STATEID_EVENT(deleg_return); DEFINE_STATEID_EVENT(deleg_recall); From 788849b64dff397c7875ea0f68564ff57d1a7515 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:50:39 -0400 Subject: [PATCH 129/186] SUNRPC: Remove RPCSEC_GSS_KRB5_ENCTYPES_DES Make it impossible to enable support for the DES or DES3 Kerberos encryption types in SunRPC. These enctypes were deprecated by RFCs 6649 and 8429 because they are known to be insecure. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/.kunitconfig | 1 - net/sunrpc/Kconfig | 28 ---------------------------- 2 files changed, 29 deletions(-) diff --git a/net/sunrpc/.kunitconfig b/net/sunrpc/.kunitconfig index a55a00fa649ba..eb02b906c2959 100644 --- a/net/sunrpc/.kunitconfig +++ b/net/sunrpc/.kunitconfig @@ -23,7 +23,6 @@ CONFIG_NFS_FS=y CONFIG_SUNRPC=y CONFIG_SUNRPC_GSS=y CONFIG_RPCSEC_GSS_KRB5=y -CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_DES=y CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1=y CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_CAMELLIA=y CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA2=y diff --git a/net/sunrpc/Kconfig b/net/sunrpc/Kconfig index 4afc5fd71d44f..68c95cfd8afa0 100644 --- a/net/sunrpc/Kconfig +++ b/net/sunrpc/Kconfig @@ -34,38 +34,10 @@ config RPCSEC_GSS_KRB5 If unsure, say Y. -config RPCSEC_GSS_KRB5_SIMPLIFIED - bool - depends on RPCSEC_GSS_KRB5 - config RPCSEC_GSS_KRB5_CRYPTOSYSTEM bool depends on RPCSEC_GSS_KRB5 -config RPCSEC_GSS_KRB5_ENCTYPES_DES - bool "Enable Kerberos enctypes based on DES (deprecated)" - depends on RPCSEC_GSS_KRB5 - depends on CRYPTO_CBC && CRYPTO_CTS && CRYPTO_ECB - depends on CRYPTO_HMAC && CRYPTO_MD5 && CRYPTO_SHA1 - depends on CRYPTO_DES - default n - select RPCSEC_GSS_KRB5_SIMPLIFIED - help - Choose Y to enable the use of deprecated Kerberos 5 - encryption types that utilize Data Encryption Standard - (DES) based ciphers. These include des-cbc-md5, - des-cbc-crc, and des-cbc-md4, which were deprecated by - RFC 6649, and des3-cbc-sha1, which was deprecated by RFC - 8429. - - These encryption types are known to be insecure, therefore - the default setting of this option is N. Support for these - encryption types is available only for compatibility with - legacy NFS client and server implementations. - - Removal of support is planned for a subsequent kernel - release. - config RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1 bool "Enable Kerberos enctypes based on AES and SHA-1" depends on RPCSEC_GSS_KRB5 From 649879561d64b05e2b4b0c6068c1a53eccc5214e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:50:46 -0400 Subject: [PATCH 130/186] SUNRPC: Remove Kunit tests for the DES3 encryption type The DES3 encryption type is no longer implemented. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_test.c | 196 ---------------------------- 1 file changed, 196 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_test.c b/net/sunrpc/auth_gss/gss_krb5_test.c index 95ca783795c5e..85625e3f3814e 100644 --- a/net/sunrpc/auth_gss/gss_krb5_test.c +++ b/net/sunrpc/auth_gss/gss_krb5_test.c @@ -320,208 +320,12 @@ static void rfc3961_nfold_case(struct kunit *test) "result mismatch"); } -/* - * RFC 3961 Appendix A.3. DES3 DR and DK - * - * These tests show the derived-random and derived-key values for the - * des3-hmac-sha1-kd encryption scheme, using the DR and DK functions - * defined in section 6.3.1. The input keys were randomly generated; - * the usage values are from this specification. - * - * This test material is copyright (C) The Internet Society (2005). - */ - -DEFINE_HEX_XDR_NETOBJ(des3_dk_usage_155, - 0x00, 0x00, 0x00, 0x01, 0x55 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_usage_1aa, - 0x00, 0x00, 0x00, 0x01, 0xaa -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_usage_kerberos, - 0x6b, 0x65, 0x72, 0x62, 0x65, 0x72, 0x6f, 0x73 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test1_base_key, - 0xdc, 0xe0, 0x6b, 0x1f, 0x64, 0xc8, 0x57, 0xa1, - 0x1c, 0x3d, 0xb5, 0x7c, 0x51, 0x89, 0x9b, 0x2c, - 0xc1, 0x79, 0x10, 0x08, 0xce, 0x97, 0x3b, 0x92 -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test1_derived_key, - 0x92, 0x51, 0x79, 0xd0, 0x45, 0x91, 0xa7, 0x9b, - 0x5d, 0x31, 0x92, 0xc4, 0xa7, 0xe9, 0xc2, 0x89, - 0xb0, 0x49, 0xc7, 0x1f, 0x6e, 0xe6, 0x04, 0xcd -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test2_base_key, - 0x5e, 0x13, 0xd3, 0x1c, 0x70, 0xef, 0x76, 0x57, - 0x46, 0x57, 0x85, 0x31, 0xcb, 0x51, 0xc1, 0x5b, - 0xf1, 0x1c, 0xa8, 0x2c, 0x97, 0xce, 0xe9, 0xf2 -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test2_derived_key, - 0x9e, 0x58, 0xe5, 0xa1, 0x46, 0xd9, 0x94, 0x2a, - 0x10, 0x1c, 0x46, 0x98, 0x45, 0xd6, 0x7a, 0x20, - 0xe3, 0xc4, 0x25, 0x9e, 0xd9, 0x13, 0xf2, 0x07 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test3_base_key, - 0x98, 0xe6, 0xfd, 0x8a, 0x04, 0xa4, 0xb6, 0x85, - 0x9b, 0x75, 0xa1, 0x76, 0x54, 0x0b, 0x97, 0x52, - 0xba, 0xd3, 0xec, 0xd6, 0x10, 0xa2, 0x52, 0xbc -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test3_derived_key, - 0x13, 0xfe, 0xf8, 0x0d, 0x76, 0x3e, 0x94, 0xec, - 0x6d, 0x13, 0xfd, 0x2c, 0xa1, 0xd0, 0x85, 0x07, - 0x02, 0x49, 0xda, 0xd3, 0x98, 0x08, 0xea, 0xbf -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test4_base_key, - 0x62, 0x2a, 0xec, 0x25, 0xa2, 0xfe, 0x2c, 0xad, - 0x70, 0x94, 0x68, 0x0b, 0x7c, 0x64, 0x94, 0x02, - 0x80, 0x08, 0x4c, 0x1a, 0x7c, 0xec, 0x92, 0xb5 -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test4_derived_key, - 0xf8, 0xdf, 0xbf, 0x04, 0xb0, 0x97, 0xe6, 0xd9, - 0xdc, 0x07, 0x02, 0x68, 0x6b, 0xcb, 0x34, 0x89, - 0xd9, 0x1f, 0xd9, 0xa4, 0x51, 0x6b, 0x70, 0x3e -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test5_base_key, - 0xd3, 0xf8, 0x29, 0x8c, 0xcb, 0x16, 0x64, 0x38, - 0xdc, 0xb9, 0xb9, 0x3e, 0xe5, 0xa7, 0x62, 0x92, - 0x86, 0xa4, 0x91, 0xf8, 0x38, 0xf8, 0x02, 0xfb -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test5_derived_key, - 0x23, 0x70, 0xda, 0x57, 0x5d, 0x2a, 0x3d, 0xa8, - 0x64, 0xce, 0xbf, 0xdc, 0x52, 0x04, 0xd5, 0x6d, - 0xf7, 0x79, 0xa7, 0xdf, 0x43, 0xd9, 0xda, 0x43 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test6_base_key, - 0xc1, 0x08, 0x16, 0x49, 0xad, 0xa7, 0x43, 0x62, - 0xe6, 0xa1, 0x45, 0x9d, 0x01, 0xdf, 0xd3, 0x0d, - 0x67, 0xc2, 0x23, 0x4c, 0x94, 0x07, 0x04, 0xda -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test6_derived_key, - 0x34, 0x80, 0x57, 0xec, 0x98, 0xfd, 0xc4, 0x80, - 0x16, 0x16, 0x1c, 0x2a, 0x4c, 0x7a, 0x94, 0x3e, - 0x92, 0xae, 0x49, 0x2c, 0x98, 0x91, 0x75, 0xf7 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test7_base_key, - 0x5d, 0x15, 0x4a, 0xf2, 0x38, 0xf4, 0x67, 0x13, - 0x15, 0x57, 0x19, 0xd5, 0x5e, 0x2f, 0x1f, 0x79, - 0x0d, 0xd6, 0x61, 0xf2, 0x79, 0xa7, 0x91, 0x7c -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test7_derived_key, - 0xa8, 0x80, 0x8a, 0xc2, 0x67, 0xda, 0xda, 0x3d, - 0xcb, 0xe9, 0xa7, 0xc8, 0x46, 0x26, 0xfb, 0xc7, - 0x61, 0xc2, 0x94, 0xb0, 0x13, 0x15, 0xe5, 0xc1 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test8_base_key, - 0x79, 0x85, 0x62, 0xe0, 0x49, 0x85, 0x2f, 0x57, - 0xdc, 0x8c, 0x34, 0x3b, 0xa1, 0x7f, 0x2c, 0xa1, - 0xd9, 0x73, 0x94, 0xef, 0xc8, 0xad, 0xc4, 0x43 -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test8_derived_key, - 0xc8, 0x13, 0xf8, 0x8a, 0x3b, 0xe3, 0xb3, 0x34, - 0xf7, 0x54, 0x25, 0xce, 0x91, 0x75, 0xfb, 0xe3, - 0xc8, 0x49, 0x3b, 0x89, 0xc8, 0x70, 0x3b, 0x49 -); - -DEFINE_HEX_XDR_NETOBJ(des3_dk_test9_base_key, - 0x26, 0xdc, 0xe3, 0x34, 0xb5, 0x45, 0x29, 0x2f, - 0x2f, 0xea, 0xb9, 0xa8, 0x70, 0x1a, 0x89, 0xa4, - 0xb9, 0x9e, 0xb9, 0x94, 0x2c, 0xec, 0xd0, 0x16 -); -DEFINE_HEX_XDR_NETOBJ(des3_dk_test9_derived_key, - 0xf4, 0x8f, 0xfd, 0x6e, 0x83, 0xf8, 0x3e, 0x73, - 0x54, 0xe6, 0x94, 0xfd, 0x25, 0x2c, 0xf8, 0x3b, - 0xfe, 0x58, 0xf7, 0xd5, 0xba, 0x37, 0xec, 0x5d -); - -static const struct gss_krb5_test_param rfc3961_kdf_test_params[] = { - { - .desc = "des3-hmac-sha1 key derivation case 1", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test1_base_key, - .usage = &des3_dk_usage_155, - .expected_result = &des3_dk_test1_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 2", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test2_base_key, - .usage = &des3_dk_usage_1aa, - .expected_result = &des3_dk_test2_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 3", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test3_base_key, - .usage = &des3_dk_usage_155, - .expected_result = &des3_dk_test3_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 4", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test4_base_key, - .usage = &des3_dk_usage_1aa, - .expected_result = &des3_dk_test4_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 5", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test5_base_key, - .usage = &des3_dk_usage_kerberos, - .expected_result = &des3_dk_test5_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 6", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test6_base_key, - .usage = &des3_dk_usage_155, - .expected_result = &des3_dk_test6_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 7", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test7_base_key, - .usage = &des3_dk_usage_1aa, - .expected_result = &des3_dk_test7_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 8", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test8_base_key, - .usage = &des3_dk_usage_155, - .expected_result = &des3_dk_test8_derived_key, - }, - { - .desc = "des3-hmac-sha1 key derivation case 9", - .enctype = ENCTYPE_DES3_CBC_RAW, - .base_key = &des3_dk_test9_base_key, - .usage = &des3_dk_usage_1aa, - .expected_result = &des3_dk_test9_derived_key, - }, -}; - -/* Creates the function rfc3961_kdf_gen_params */ -KUNIT_ARRAY_PARAM(rfc3961_kdf, rfc3961_kdf_test_params, gss_krb5_get_desc); - static struct kunit_case rfc3961_test_cases[] = { { .name = "RFC 3961 n-fold", .run_case = rfc3961_nfold_case, .generate_params = rfc3961_nfold_gen_params, }, - { - .name = "RFC 3961 key derivation", - .run_case = kdf_case, - .generate_params = rfc3961_kdf_gen_params, - }, {} }; From 2024b89d90ecb11f214e05acc7fa7929121bebbf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:50:52 -0400 Subject: [PATCH 131/186] SUNRPC: Remove DES and DES3 enctypes from the supported enctypes list These enctypes can no longer be enabled via CONFIG. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_mech.c | 52 ----------------------------- 1 file changed, 52 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 20e21d08badb3..39160a8ca3b6a 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -39,52 +39,6 @@ static int gss_krb5_import_ctx_v2(struct krb5_ctx *ctx, gfp_t gfp_mask); #endif static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { -#if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_DES) - /* - * DES (All DES enctypes are mapped to the same gss functionality) - */ - { - .etype = ENCTYPE_DES_CBC_RAW, - .ctype = CKSUMTYPE_RSA_MD5, - .name = "des-cbc-crc", - .encrypt_name = "cbc(des)", - .cksum_name = "md5", - .import_ctx = gss_krb5_import_ctx_des, - .get_mic = gss_krb5_get_mic_v1, - .verify_mic = gss_krb5_verify_mic_v1, - .wrap = gss_krb5_wrap_v1, - .unwrap = gss_krb5_unwrap_v1, - .signalg = SGN_ALG_DES_MAC_MD5, - .sealalg = SEAL_ALG_DES, - .keybytes = 7, - .keylength = 8, - .cksumlength = 8, - .keyed_cksum = 0, - }, - /* - * 3DES - */ - { - .etype = ENCTYPE_DES3_CBC_RAW, - .ctype = CKSUMTYPE_HMAC_SHA1_DES3, - .name = "des3-hmac-sha1", - .encrypt_name = "cbc(des3_ede)", - .cksum_name = "hmac(sha1)", - .import_ctx = gss_krb5_import_ctx_v1, - .derive_key = krb5_derive_key_v1, - .get_mic = gss_krb5_get_mic_v1, - .verify_mic = gss_krb5_verify_mic_v1, - .wrap = gss_krb5_wrap_v1, - .unwrap = gss_krb5_unwrap_v1, - .signalg = SGN_ALG_HMAC_SHA1_DES3_KD, - .sealalg = SEAL_ALG_DES3KD, - .keybytes = 21, - .keylength = 24, - .cksumlength = 20, - .keyed_cksum = 1, - }, -#endif - #if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1) /* * AES-128 with SHA-1 (RFC 3962) @@ -283,12 +237,6 @@ static void gss_krb5_prepare_enctype_priority_list(void) #if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1) ENCTYPE_AES256_CTS_HMAC_SHA1_96, ENCTYPE_AES128_CTS_HMAC_SHA1_96, -#endif -#if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_DES) - ENCTYPE_DES3_CBC_SHA1, - ENCTYPE_DES_CBC_MD5, - ENCTYPE_DES_CBC_CRC, - ENCTYPE_DES_CBC_MD4, #endif }; size_t total, i; From ec596aaf9b489e2aefa44697e126d95c6896bc4c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:00 -0400 Subject: [PATCH 132/186] SUNRPC: Remove code behind CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED None of this code can be enabled any more. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_internal.h | 9 - net/sunrpc/auth_gss/gss_krb5_mech.c | 44 ---- net/sunrpc/auth_gss/gss_krb5_seal.c | 69 ------ net/sunrpc/auth_gss/gss_krb5_unseal.c | 77 ------- net/sunrpc/auth_gss/gss_krb5_wrap.c | 287 ------------------------ 5 files changed, 486 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_internal.h b/net/sunrpc/auth_gss/gss_krb5_internal.h index b673e2626acb2..3471a574997ae 100644 --- a/net/sunrpc/auth_gss/gss_krb5_internal.h +++ b/net/sunrpc/auth_gss/gss_krb5_internal.h @@ -85,24 +85,15 @@ struct krb5_ctx { * GSS Kerberos 5 mechanism Per-Message calls. */ -u32 gss_krb5_get_mic_v1(struct krb5_ctx *ctx, struct xdr_buf *text, - struct xdr_netobj *token); u32 gss_krb5_get_mic_v2(struct krb5_ctx *ctx, struct xdr_buf *text, struct xdr_netobj *token); -u32 gss_krb5_verify_mic_v1(struct krb5_ctx *ctx, struct xdr_buf *message_buffer, - struct xdr_netobj *read_token); u32 gss_krb5_verify_mic_v2(struct krb5_ctx *ctx, struct xdr_buf *message_buffer, struct xdr_netobj *read_token); -u32 gss_krb5_wrap_v1(struct krb5_ctx *kctx, int offset, - struct xdr_buf *buf, struct page **pages); u32 gss_krb5_wrap_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf, struct page **pages); -u32 gss_krb5_unwrap_v1(struct krb5_ctx *kctx, int offset, int len, - struct xdr_buf *buf, unsigned int *slack, - unsigned int *align); u32 gss_krb5_unwrap_v2(struct krb5_ctx *kctx, int offset, int len, struct xdr_buf *buf, unsigned int *slack, unsigned int *align); diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 39160a8ca3b6a..890ad877792f4 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -30,10 +30,6 @@ static struct gss_api_mech gss_kerberos_mech; -#if defined(CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED) -static int gss_krb5_import_ctx_des(struct krb5_ctx *ctx, gfp_t gfp_mask); -static int gss_krb5_import_ctx_v1(struct krb5_ctx *ctx, gfp_t gfp_mask); -#endif #if defined(CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM) static int gss_krb5_import_ctx_v2(struct krb5_ctx *ctx, gfp_t gfp_mask); #endif @@ -414,46 +410,6 @@ gss_import_v1_context(const void *p, const void *end, struct krb5_ctx *ctx) return PTR_ERR(p); } -#if defined(CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED) -static int -gss_krb5_import_ctx_des(struct krb5_ctx *ctx, gfp_t gfp_mask) -{ - return -EINVAL; -} - -static int -gss_krb5_import_ctx_v1(struct krb5_ctx *ctx, gfp_t gfp_mask) -{ - struct xdr_netobj keyin, keyout; - - keyin.data = ctx->Ksess; - keyin.len = ctx->gk5e->keylength; - - ctx->seq = gss_krb5_alloc_cipher_v1(ctx, &keyin); - if (ctx->seq == NULL) - goto out_err; - ctx->enc = gss_krb5_alloc_cipher_v1(ctx, &keyin); - if (ctx->enc == NULL) - goto out_free_seq; - - /* derive cksum */ - keyout.data = ctx->cksum; - keyout.len = ctx->gk5e->keylength; - if (krb5_derive_key(ctx, &keyin, &keyout, KG_USAGE_SIGN, - KEY_USAGE_SEED_CHECKSUM, gfp_mask)) - goto out_free_enc; - - return 0; - -out_free_enc: - crypto_free_sync_skcipher(ctx->enc); -out_free_seq: - crypto_free_sync_skcipher(ctx->seq); -out_err: - return -EINVAL; -} -#endif - #if defined(CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM) static struct crypto_sync_skcipher * diff --git a/net/sunrpc/auth_gss/gss_krb5_seal.c b/net/sunrpc/auth_gss/gss_krb5_seal.c index 146aa755f07df..ce540df9bce46 100644 --- a/net/sunrpc/auth_gss/gss_krb5_seal.c +++ b/net/sunrpc/auth_gss/gss_krb5_seal.c @@ -71,75 +71,6 @@ # define RPCDBG_FACILITY RPCDBG_AUTH #endif -#if defined(CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED) - -static void * -setup_token(struct krb5_ctx *ctx, struct xdr_netobj *token) -{ - u16 *ptr; - void *krb5_hdr; - int body_size = GSS_KRB5_TOK_HDR_LEN + ctx->gk5e->cksumlength; - - token->len = g_token_size(&ctx->mech_used, body_size); - - ptr = (u16 *)token->data; - g_make_token_header(&ctx->mech_used, body_size, (unsigned char **)&ptr); - - /* ptr now at start of header described in rfc 1964, section 1.2.1: */ - krb5_hdr = ptr; - *ptr++ = KG_TOK_MIC_MSG; - /* - * signalg is stored as if it were converted from LE to host endian, even - * though it's an opaque pair of bytes according to the RFC. - */ - *ptr++ = (__force u16)cpu_to_le16(ctx->gk5e->signalg); - *ptr++ = SEAL_ALG_NONE; - *ptr = 0xffff; - - return krb5_hdr; -} - -u32 -gss_krb5_get_mic_v1(struct krb5_ctx *ctx, struct xdr_buf *text, - struct xdr_netobj *token) -{ - char cksumdata[GSS_KRB5_MAX_CKSUM_LEN]; - struct xdr_netobj md5cksum = {.len = sizeof(cksumdata), - .data = cksumdata}; - void *ptr; - time64_t now; - u32 seq_send; - u8 *cksumkey; - - dprintk("RPC: %s\n", __func__); - BUG_ON(ctx == NULL); - - now = ktime_get_real_seconds(); - - ptr = setup_token(ctx, token); - - if (ctx->gk5e->keyed_cksum) - cksumkey = ctx->cksum; - else - cksumkey = NULL; - - if (make_checksum(ctx, ptr, 8, text, 0, cksumkey, - KG_USAGE_SIGN, &md5cksum)) - return GSS_S_FAILURE; - - memcpy(ptr + GSS_KRB5_TOK_HDR_LEN, md5cksum.data, md5cksum.len); - - seq_send = atomic_fetch_inc(&ctx->seq_send); - - if (krb5_make_seq_num(ctx, ctx->seq, ctx->initiate ? 0 : 0xff, - seq_send, ptr + GSS_KRB5_TOK_HDR_LEN, ptr + 8)) - return GSS_S_FAILURE; - - return (ctx->endtime < now) ? GSS_S_CONTEXT_EXPIRED : GSS_S_COMPLETE; -} - -#endif - static void * setup_token_v2(struct krb5_ctx *ctx, struct xdr_netobj *token) { diff --git a/net/sunrpc/auth_gss/gss_krb5_unseal.c b/net/sunrpc/auth_gss/gss_krb5_unseal.c index 7d6d4ae4a3c96..4fbc50a0a2c4b 100644 --- a/net/sunrpc/auth_gss/gss_krb5_unseal.c +++ b/net/sunrpc/auth_gss/gss_krb5_unseal.c @@ -69,83 +69,6 @@ # define RPCDBG_FACILITY RPCDBG_AUTH #endif - -#if defined(CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED) -/* read_token is a mic token, and message_buffer is the data that the mic was - * supposedly taken over. */ -u32 -gss_krb5_verify_mic_v1(struct krb5_ctx *ctx, struct xdr_buf *message_buffer, - struct xdr_netobj *read_token) -{ - int signalg; - int sealalg; - char cksumdata[GSS_KRB5_MAX_CKSUM_LEN]; - struct xdr_netobj md5cksum = {.len = sizeof(cksumdata), - .data = cksumdata}; - s32 now; - int direction; - u32 seqnum; - unsigned char *ptr = (unsigned char *)read_token->data; - int bodysize; - u8 *cksumkey; - - dprintk("RPC: krb5_read_token\n"); - - if (g_verify_token_header(&ctx->mech_used, &bodysize, &ptr, - read_token->len)) - return GSS_S_DEFECTIVE_TOKEN; - - if ((ptr[0] != ((KG_TOK_MIC_MSG >> 8) & 0xff)) || - (ptr[1] != (KG_TOK_MIC_MSG & 0xff))) - return GSS_S_DEFECTIVE_TOKEN; - - /* XXX sanity-check bodysize?? */ - - signalg = ptr[2] + (ptr[3] << 8); - if (signalg != ctx->gk5e->signalg) - return GSS_S_DEFECTIVE_TOKEN; - - sealalg = ptr[4] + (ptr[5] << 8); - if (sealalg != SEAL_ALG_NONE) - return GSS_S_DEFECTIVE_TOKEN; - - if ((ptr[6] != 0xff) || (ptr[7] != 0xff)) - return GSS_S_DEFECTIVE_TOKEN; - - if (ctx->gk5e->keyed_cksum) - cksumkey = ctx->cksum; - else - cksumkey = NULL; - - if (make_checksum(ctx, ptr, 8, message_buffer, 0, - cksumkey, KG_USAGE_SIGN, &md5cksum)) - return GSS_S_FAILURE; - - if (memcmp(md5cksum.data, ptr + GSS_KRB5_TOK_HDR_LEN, - ctx->gk5e->cksumlength)) - return GSS_S_BAD_SIG; - - /* it got through unscathed. Make sure the context is unexpired */ - - now = ktime_get_real_seconds(); - - if (now > ctx->endtime) - return GSS_S_CONTEXT_EXPIRED; - - /* do sequencing checks */ - - if (krb5_get_seq_num(ctx, ptr + GSS_KRB5_TOK_HDR_LEN, ptr + 8, - &direction, &seqnum)) - return GSS_S_FAILURE; - - if ((ctx->initiate && direction != 0xff) || - (!ctx->initiate && direction != 0)) - return GSS_S_BAD_SIG; - - return GSS_S_COMPLETE; -} -#endif - u32 gss_krb5_verify_mic_v2(struct krb5_ctx *ctx, struct xdr_buf *message_buffer, struct xdr_netobj *read_token) diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c index 6d6b082380b23..b3e1738ff6bfa 100644 --- a/net/sunrpc/auth_gss/gss_krb5_wrap.c +++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c @@ -40,293 +40,6 @@ # define RPCDBG_FACILITY RPCDBG_AUTH #endif -#if defined(CONFIG_RPCSEC_GSS_KRB5_SIMPLIFIED) - -static inline int -gss_krb5_padding(int blocksize, int length) -{ - return blocksize - (length % blocksize); -} - -static inline void -gss_krb5_add_padding(struct xdr_buf *buf, int offset, int blocksize) -{ - int padding = gss_krb5_padding(blocksize, buf->len - offset); - char *p; - struct kvec *iov; - - if (buf->page_len || buf->tail[0].iov_len) - iov = &buf->tail[0]; - else - iov = &buf->head[0]; - p = iov->iov_base + iov->iov_len; - iov->iov_len += padding; - buf->len += padding; - memset(p, padding, padding); -} - -static inline int -gss_krb5_remove_padding(struct xdr_buf *buf, int blocksize) -{ - u8 *ptr; - u8 pad; - size_t len = buf->len; - - if (len <= buf->head[0].iov_len) { - pad = *(u8 *)(buf->head[0].iov_base + len - 1); - if (pad > buf->head[0].iov_len) - return -EINVAL; - buf->head[0].iov_len -= pad; - goto out; - } else - len -= buf->head[0].iov_len; - if (len <= buf->page_len) { - unsigned int last = (buf->page_base + len - 1) - >>PAGE_SHIFT; - unsigned int offset = (buf->page_base + len - 1) - & (PAGE_SIZE - 1); - ptr = kmap_atomic(buf->pages[last]); - pad = *(ptr + offset); - kunmap_atomic(ptr); - goto out; - } else - len -= buf->page_len; - BUG_ON(len > buf->tail[0].iov_len); - pad = *(u8 *)(buf->tail[0].iov_base + len - 1); -out: - /* XXX: NOTE: we do not adjust the page lengths--they represent - * a range of data in the real filesystem page cache, and we need - * to know that range so the xdr code can properly place read data. - * However adjusting the head length, as we do above, is harmless. - * In the case of a request that fits into a single page, the server - * also uses length and head length together to determine the original - * start of the request to copy the request for deferal; so it's - * easier on the server if we adjust head and tail length in tandem. - * It's not really a problem that we don't fool with the page and - * tail lengths, though--at worst badly formed xdr might lead the - * server to attempt to parse the padding. - * XXX: Document all these weird requirements for gss mechanism - * wrap/unwrap functions. */ - if (pad > blocksize) - return -EINVAL; - if (buf->len > pad) - buf->len -= pad; - else - return -EINVAL; - return 0; -} - -/* Assumptions: the head and tail of inbuf are ours to play with. - * The pages, however, may be real pages in the page cache and we replace - * them with scratch pages from **pages before writing to them. */ -/* XXX: obviously the above should be documentation of wrap interface, - * and shouldn't be in this kerberos-specific file. */ - -/* XXX factor out common code with seal/unseal. */ - -u32 -gss_krb5_wrap_v1(struct krb5_ctx *kctx, int offset, - struct xdr_buf *buf, struct page **pages) -{ - char cksumdata[GSS_KRB5_MAX_CKSUM_LEN]; - struct xdr_netobj md5cksum = {.len = sizeof(cksumdata), - .data = cksumdata}; - int blocksize = 0, plainlen; - unsigned char *ptr, *msg_start; - time64_t now; - int headlen; - struct page **tmp_pages; - u32 seq_send; - u8 *cksumkey; - u32 conflen = crypto_sync_skcipher_blocksize(kctx->enc); - - dprintk("RPC: %s\n", __func__); - - now = ktime_get_real_seconds(); - - blocksize = crypto_sync_skcipher_blocksize(kctx->enc); - gss_krb5_add_padding(buf, offset, blocksize); - BUG_ON((buf->len - offset) % blocksize); - plainlen = conflen + buf->len - offset; - - headlen = g_token_size(&kctx->mech_used, - GSS_KRB5_TOK_HDR_LEN + kctx->gk5e->cksumlength + plainlen) - - (buf->len - offset); - - ptr = buf->head[0].iov_base + offset; - /* shift data to make room for header. */ - xdr_extend_head(buf, offset, headlen); - - /* XXX Would be cleverer to encrypt while copying. */ - BUG_ON((buf->len - offset - headlen) % blocksize); - - g_make_token_header(&kctx->mech_used, - GSS_KRB5_TOK_HDR_LEN + - kctx->gk5e->cksumlength + plainlen, &ptr); - - - /* ptr now at header described in rfc 1964, section 1.2.1: */ - ptr[0] = (unsigned char) ((KG_TOK_WRAP_MSG >> 8) & 0xff); - ptr[1] = (unsigned char) (KG_TOK_WRAP_MSG & 0xff); - - msg_start = ptr + GSS_KRB5_TOK_HDR_LEN + kctx->gk5e->cksumlength; - - /* - * signalg and sealalg are stored as if they were converted from LE - * to host endian, even though they're opaque pairs of bytes according - * to the RFC. - */ - *(__le16 *)(ptr + 2) = cpu_to_le16(kctx->gk5e->signalg); - *(__le16 *)(ptr + 4) = cpu_to_le16(kctx->gk5e->sealalg); - ptr[6] = 0xff; - ptr[7] = 0xff; - - krb5_make_confounder(msg_start, conflen); - - if (kctx->gk5e->keyed_cksum) - cksumkey = kctx->cksum; - else - cksumkey = NULL; - - /* XXXJBF: UGH!: */ - tmp_pages = buf->pages; - buf->pages = pages; - if (make_checksum(kctx, ptr, 8, buf, offset + headlen - conflen, - cksumkey, KG_USAGE_SEAL, &md5cksum)) - return GSS_S_FAILURE; - buf->pages = tmp_pages; - - memcpy(ptr + GSS_KRB5_TOK_HDR_LEN, md5cksum.data, md5cksum.len); - - seq_send = atomic_fetch_inc(&kctx->seq_send); - - /* XXX would probably be more efficient to compute checksum - * and encrypt at the same time: */ - if ((krb5_make_seq_num(kctx, kctx->seq, kctx->initiate ? 0 : 0xff, - seq_send, ptr + GSS_KRB5_TOK_HDR_LEN, ptr + 8))) - return GSS_S_FAILURE; - - if (gss_encrypt_xdr_buf(kctx->enc, buf, - offset + headlen - conflen, pages)) - return GSS_S_FAILURE; - - return (kctx->endtime < now) ? GSS_S_CONTEXT_EXPIRED : GSS_S_COMPLETE; -} - -u32 -gss_krb5_unwrap_v1(struct krb5_ctx *kctx, int offset, int len, - struct xdr_buf *buf, unsigned int *slack, - unsigned int *align) -{ - int signalg; - int sealalg; - char cksumdata[GSS_KRB5_MAX_CKSUM_LEN]; - struct xdr_netobj md5cksum = {.len = sizeof(cksumdata), - .data = cksumdata}; - time64_t now; - int direction; - s32 seqnum; - unsigned char *ptr; - int bodysize; - void *data_start, *orig_start; - int data_len; - int blocksize; - u32 conflen = crypto_sync_skcipher_blocksize(kctx->enc); - int crypt_offset; - u8 *cksumkey; - unsigned int saved_len = buf->len; - - dprintk("RPC: gss_unwrap_kerberos\n"); - - ptr = (u8 *)buf->head[0].iov_base + offset; - if (g_verify_token_header(&kctx->mech_used, &bodysize, &ptr, - len - offset)) - return GSS_S_DEFECTIVE_TOKEN; - - if ((ptr[0] != ((KG_TOK_WRAP_MSG >> 8) & 0xff)) || - (ptr[1] != (KG_TOK_WRAP_MSG & 0xff))) - return GSS_S_DEFECTIVE_TOKEN; - - /* XXX sanity-check bodysize?? */ - - /* get the sign and seal algorithms */ - - signalg = ptr[2] + (ptr[3] << 8); - if (signalg != kctx->gk5e->signalg) - return GSS_S_DEFECTIVE_TOKEN; - - sealalg = ptr[4] + (ptr[5] << 8); - if (sealalg != kctx->gk5e->sealalg) - return GSS_S_DEFECTIVE_TOKEN; - - if ((ptr[6] != 0xff) || (ptr[7] != 0xff)) - return GSS_S_DEFECTIVE_TOKEN; - - /* - * Data starts after token header and checksum. ptr points - * to the beginning of the token header - */ - crypt_offset = ptr + (GSS_KRB5_TOK_HDR_LEN + kctx->gk5e->cksumlength) - - (unsigned char *)buf->head[0].iov_base; - - buf->len = len; - if (gss_decrypt_xdr_buf(kctx->enc, buf, crypt_offset)) - return GSS_S_DEFECTIVE_TOKEN; - - if (kctx->gk5e->keyed_cksum) - cksumkey = kctx->cksum; - else - cksumkey = NULL; - - if (make_checksum(kctx, ptr, 8, buf, crypt_offset, - cksumkey, KG_USAGE_SEAL, &md5cksum)) - return GSS_S_FAILURE; - - if (memcmp(md5cksum.data, ptr + GSS_KRB5_TOK_HDR_LEN, - kctx->gk5e->cksumlength)) - return GSS_S_BAD_SIG; - - /* it got through unscathed. Make sure the context is unexpired */ - - now = ktime_get_real_seconds(); - - if (now > kctx->endtime) - return GSS_S_CONTEXT_EXPIRED; - - /* do sequencing checks */ - - if (krb5_get_seq_num(kctx, ptr + GSS_KRB5_TOK_HDR_LEN, - ptr + 8, &direction, &seqnum)) - return GSS_S_BAD_SIG; - - if ((kctx->initiate && direction != 0xff) || - (!kctx->initiate && direction != 0)) - return GSS_S_BAD_SIG; - - /* Copy the data back to the right position. XXX: Would probably be - * better to copy and encrypt at the same time. */ - - blocksize = crypto_sync_skcipher_blocksize(kctx->enc); - data_start = ptr + (GSS_KRB5_TOK_HDR_LEN + kctx->gk5e->cksumlength) + - conflen; - orig_start = buf->head[0].iov_base + offset; - data_len = (buf->head[0].iov_base + buf->head[0].iov_len) - data_start; - memmove(orig_start, data_start, data_len); - buf->head[0].iov_len -= (data_start - orig_start); - buf->len = len - (data_start - orig_start); - - if (gss_krb5_remove_padding(buf, blocksize)) - return GSS_S_DEFECTIVE_TOKEN; - - /* slack must include room for krb5 padding */ - *slack = XDR_QUADLEN(saved_len - buf->len); - /* The GSS blob always precedes the RPC message payload */ - *align = *slack; - return GSS_S_COMPLETE; -} - -#endif - /* * We can shift data by up to LOCAL_BUF_LEN bytes in a pass. If we need * to do more than that, we shift repeatedly. Kevin Coffman reports From 6964629f4c188af3103a645a672877ee0e9bac91 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:06 -0400 Subject: [PATCH 133/186] SUNRPC: Remove krb5_derive_key_v1() This function is no longer used. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_internal.h | 6 -- net/sunrpc/auth_gss/gss_krb5_keys.c | 84 ------------------------- 2 files changed, 90 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_internal.h b/net/sunrpc/auth_gss/gss_krb5_internal.h index 3471a574997ae..c1aea062c01be 100644 --- a/net/sunrpc/auth_gss/gss_krb5_internal.h +++ b/net/sunrpc/auth_gss/gss_krb5_internal.h @@ -104,12 +104,6 @@ u32 gss_krb5_unwrap_v2(struct krb5_ctx *kctx, int offset, int len, /* Key Derivation Functions */ -int krb5_derive_key_v1(const struct gss_krb5_enctype *gk5e, - const struct xdr_netobj *inkey, - struct xdr_netobj *outkey, - const struct xdr_netobj *label, - gfp_t gfp_mask); - int krb5_derive_key_v2(const struct gss_krb5_enctype *gk5e, const struct xdr_netobj *inkey, struct xdr_netobj *outkey, diff --git a/net/sunrpc/auth_gss/gss_krb5_keys.c b/net/sunrpc/auth_gss/gss_krb5_keys.c index 5347fe1cc93f8..06d8ee0db000f 100644 --- a/net/sunrpc/auth_gss/gss_krb5_keys.c +++ b/net/sunrpc/auth_gss/gss_krb5_keys.c @@ -222,90 +222,6 @@ static int krb5_DK(const struct gss_krb5_enctype *gk5e, return ret; } -#define smask(step) ((1<>step)&smask(step))) -#define parity_char(x) pstep(pstep(pstep((x), 4), 2), 1) - -static void mit_des_fixup_key_parity(u8 key[8]) -{ - int i; - for (i = 0; i < 8; i++) { - key[i] &= 0xfe; - key[i] |= 1^parity_char(key[i]); - } -} - -static int krb5_random_to_key_v1(const struct gss_krb5_enctype *gk5e, - struct xdr_netobj *randombits, - struct xdr_netobj *key) -{ - int i, ret = -EINVAL; - - if (key->len != 24) { - dprintk("%s: key->len is %d\n", __func__, key->len); - goto err_out; - } - if (randombits->len != 21) { - dprintk("%s: randombits->len is %d\n", - __func__, randombits->len); - goto err_out; - } - - /* take the seven bytes, move them around into the top 7 bits of the - 8 key bytes, then compute the parity bits. Do this three times. */ - - for (i = 0; i < 3; i++) { - memcpy(key->data + i*8, randombits->data + i*7, 7); - key->data[i*8+7] = (((key->data[i*8]&1)<<1) | - ((key->data[i*8+1]&1)<<2) | - ((key->data[i*8+2]&1)<<3) | - ((key->data[i*8+3]&1)<<4) | - ((key->data[i*8+4]&1)<<5) | - ((key->data[i*8+5]&1)<<6) | - ((key->data[i*8+6]&1)<<7)); - - mit_des_fixup_key_parity(key->data + i*8); - } - ret = 0; -err_out: - return ret; -} - -/** - * krb5_derive_key_v1 - Derive a subkey for an RFC 3961 enctype - * @gk5e: Kerberos 5 enctype profile - * @inkey: base protocol key - * @outkey: OUT: derived key - * @label: subkey usage label - * @gfp_mask: memory allocation control flags - * - * Caller sets @outkey->len to the desired length of the derived key. - * - * On success, returns 0 and fills in @outkey. A negative errno value - * is returned on failure. - */ -int krb5_derive_key_v1(const struct gss_krb5_enctype *gk5e, - const struct xdr_netobj *inkey, - struct xdr_netobj *outkey, - const struct xdr_netobj *label, - gfp_t gfp_mask) -{ - struct xdr_netobj inblock; - int ret; - - inblock.len = gk5e->keybytes; - inblock.data = kmalloc(inblock.len, gfp_mask); - if (!inblock.data) - return -ENOMEM; - - ret = krb5_DK(gk5e, inkey, inblock.data, label, gfp_mask); - if (!ret) - ret = krb5_random_to_key_v1(gk5e, &inblock, outkey); - - kfree_sensitive(inblock.data); - return ret; -} - /* * This is the identity function, with some sanity checking. */ From da33d635bb4a5093b588dd99c97470ffe3922154 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:13 -0400 Subject: [PATCH 134/186] SUNRPC: Remove gss_import_v1_context() We no longer support importing v1 contexts. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_mech.c | 142 +--------------------------- 1 file changed, 1 insertion(+), 141 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 890ad877792f4..09fff5011d118 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -273,143 +273,6 @@ const struct gss_krb5_enctype *gss_krb5_lookup_enctype(u32 etype) } EXPORT_SYMBOL_IF_KUNIT(gss_krb5_lookup_enctype); -static struct crypto_sync_skcipher * -gss_krb5_alloc_cipher_v1(struct krb5_ctx *ctx, struct xdr_netobj *key) -{ - struct crypto_sync_skcipher *tfm; - - tfm = crypto_alloc_sync_skcipher(ctx->gk5e->encrypt_name, 0, 0); - if (IS_ERR(tfm)) - return NULL; - if (crypto_sync_skcipher_setkey(tfm, key->data, key->len)) { - crypto_free_sync_skcipher(tfm); - return NULL; - } - return tfm; -} - -static inline const void * -get_key(const void *p, const void *end, - struct krb5_ctx *ctx, struct crypto_sync_skcipher **res) -{ - struct crypto_sync_skcipher *tfm; - struct xdr_netobj key; - int alg; - - p = simple_get_bytes(p, end, &alg, sizeof(alg)); - if (IS_ERR(p)) - goto out_err; - switch (alg) { - case ENCTYPE_DES_CBC_CRC: - case ENCTYPE_DES_CBC_MD4: - case ENCTYPE_DES_CBC_MD5: - /* Map all these key types to ENCTYPE_DES_CBC_RAW */ - alg = ENCTYPE_DES_CBC_RAW; - break; - } - if (!gss_krb5_lookup_enctype(alg)) { - pr_warn("gss_krb5: unsupported enctype: %d\n", alg); - goto out_err_inval; - } - - p = simple_get_netobj(p, end, &key); - if (IS_ERR(p)) - goto out_err; - tfm = gss_krb5_alloc_cipher_v1(ctx, &key); - kfree(key.data); - if (!tfm) { - pr_warn("gss_krb5: failed to initialize cipher '%s'\n", - ctx->gk5e->encrypt_name); - goto out_err_inval; - } - *res = tfm; - - return p; - -out_err_inval: - p = ERR_PTR(-EINVAL); -out_err: - return p; -} - -static int -gss_import_v1_context(const void *p, const void *end, struct krb5_ctx *ctx) -{ - u32 seq_send; - int tmp; - u32 time32; - - p = simple_get_bytes(p, end, &ctx->initiate, sizeof(ctx->initiate)); - if (IS_ERR(p)) - goto out_err; - - /* Old format supports only DES! Any other enctype uses new format */ - ctx->enctype = ENCTYPE_DES_CBC_RAW; - - ctx->gk5e = gss_krb5_lookup_enctype(ctx->enctype); - if (ctx->gk5e == NULL) { - p = ERR_PTR(-EINVAL); - goto out_err; - } - - /* The downcall format was designed before we completely understood - * the uses of the context fields; so it includes some stuff we - * just give some minimal sanity-checking, and some we ignore - * completely (like the next twenty bytes): */ - if (unlikely(p + 20 > end || p + 20 < p)) { - p = ERR_PTR(-EFAULT); - goto out_err; - } - p += 20; - p = simple_get_bytes(p, end, &tmp, sizeof(tmp)); - if (IS_ERR(p)) - goto out_err; - if (tmp != SGN_ALG_DES_MAC_MD5) { - p = ERR_PTR(-ENOSYS); - goto out_err; - } - p = simple_get_bytes(p, end, &tmp, sizeof(tmp)); - if (IS_ERR(p)) - goto out_err; - if (tmp != SEAL_ALG_DES) { - p = ERR_PTR(-ENOSYS); - goto out_err; - } - p = simple_get_bytes(p, end, &time32, sizeof(time32)); - if (IS_ERR(p)) - goto out_err; - /* unsigned 32-bit time overflows in year 2106 */ - ctx->endtime = (time64_t)time32; - p = simple_get_bytes(p, end, &seq_send, sizeof(seq_send)); - if (IS_ERR(p)) - goto out_err; - atomic_set(&ctx->seq_send, seq_send); - p = simple_get_netobj(p, end, &ctx->mech_used); - if (IS_ERR(p)) - goto out_err; - p = get_key(p, end, ctx, &ctx->enc); - if (IS_ERR(p)) - goto out_err_free_mech; - p = get_key(p, end, ctx, &ctx->seq); - if (IS_ERR(p)) - goto out_err_free_key1; - if (p != end) { - p = ERR_PTR(-EFAULT); - goto out_err_free_key2; - } - - return 0; - -out_err_free_key2: - crypto_free_sync_skcipher(ctx->seq); -out_err_free_key1: - crypto_free_sync_skcipher(ctx->enc); -out_err_free_mech: - kfree(ctx->mech_used.data); -out_err: - return PTR_ERR(p); -} - #if defined(CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM) static struct crypto_sync_skcipher * @@ -622,10 +485,7 @@ gss_krb5_import_sec_context(const void *p, size_t len, struct gss_ctx *ctx_id, if (ctx == NULL) return -ENOMEM; - if (len == 85) - ret = gss_import_v1_context(p, end, ctx); - else - ret = gss_import_v2_context(p, end, ctx, gfp_mask); + ret = gss_import_v2_context(p, end, ctx, gfp_mask); memzero_explicit(&ctx->Ksess, sizeof(ctx->Ksess)); if (ret) { kfree(ctx); From 6c922ea71170e7d1f9e7a9049289d9edccb7b21e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:19 -0400 Subject: [PATCH 135/186] SUNRPC: Remove CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM This code is now always on, so the ifdef can be removed. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/Kconfig | 7 ------- net/sunrpc/auth_gss/gss_krb5_mech.c | 7 ------- 2 files changed, 14 deletions(-) diff --git a/net/sunrpc/Kconfig b/net/sunrpc/Kconfig index 68c95cfd8afa0..2d8b67dac7b5b 100644 --- a/net/sunrpc/Kconfig +++ b/net/sunrpc/Kconfig @@ -34,10 +34,6 @@ config RPCSEC_GSS_KRB5 If unsure, say Y. -config RPCSEC_GSS_KRB5_CRYPTOSYSTEM - bool - depends on RPCSEC_GSS_KRB5 - config RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1 bool "Enable Kerberos enctypes based on AES and SHA-1" depends on RPCSEC_GSS_KRB5 @@ -45,7 +41,6 @@ config RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1 depends on CRYPTO_HMAC && CRYPTO_SHA1 depends on CRYPTO_AES default y - select RPCSEC_GSS_KRB5_CRYPTOSYSTEM help Choose Y to enable the use of Kerberos 5 encryption types that utilize Advanced Encryption Standard (AES) ciphers and @@ -58,7 +53,6 @@ config RPCSEC_GSS_KRB5_ENCTYPES_CAMELLIA depends on CRYPTO_CBC && CRYPTO_CTS && CRYPTO_CAMELLIA depends on CRYPTO_CMAC default n - select RPCSEC_GSS_KRB5_CRYPTOSYSTEM help Choose Y to enable the use of Kerberos 5 encryption types that utilize Camellia ciphers (RFC 3713) and CMAC digests @@ -72,7 +66,6 @@ config RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA2 depends on CRYPTO_HMAC && CRYPTO_SHA256 && CRYPTO_SHA512 depends on CRYPTO_AES default n - select RPCSEC_GSS_KRB5_CRYPTOSYSTEM help Choose Y to enable the use of Kerberos 5 encryption types that utilize Advanced Encryption Standard (AES) ciphers and diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 09fff5011d118..38a0c93e4b60e 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -29,10 +29,7 @@ #endif static struct gss_api_mech gss_kerberos_mech; - -#if defined(CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM) static int gss_krb5_import_ctx_v2(struct krb5_ctx *ctx, gfp_t gfp_mask); -#endif static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { #if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1) @@ -273,8 +270,6 @@ const struct gss_krb5_enctype *gss_krb5_lookup_enctype(u32 etype) } EXPORT_SYMBOL_IF_KUNIT(gss_krb5_lookup_enctype); -#if defined(CONFIG_RPCSEC_GSS_KRB5_CRYPTOSYSTEM) - static struct crypto_sync_skcipher * gss_krb5_alloc_cipher_v2(const char *cname, const struct xdr_netobj *key) { @@ -403,8 +398,6 @@ gss_krb5_import_ctx_v2(struct krb5_ctx *ctx, gfp_t gfp_mask) goto out; } -#endif - static int gss_import_v2_context(const void *p, const void *end, struct krb5_ctx *ctx, gfp_t gfp_mask) From cfb6b328c47e2e798906383d407d9c02d73a7476 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:26 -0400 Subject: [PATCH 136/186] SUNRPC: Remove the ->import_ctx method All supported encryption types now use the same context import function. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/gss_krb5_internal.h | 1 - net/sunrpc/auth_gss/gss_krb5_mech.c | 12 +----------- 2 files changed, 1 insertion(+), 12 deletions(-) diff --git a/net/sunrpc/auth_gss/gss_krb5_internal.h b/net/sunrpc/auth_gss/gss_krb5_internal.h index c1aea062c01be..9a4b110a6154b 100644 --- a/net/sunrpc/auth_gss/gss_krb5_internal.h +++ b/net/sunrpc/auth_gss/gss_krb5_internal.h @@ -33,7 +33,6 @@ struct gss_krb5_enctype { const u32 Ke_length; /* encryption subkey length, in octets */ const u32 Ki_length; /* integrity subkey length, in octets */ - int (*import_ctx)(struct krb5_ctx *ctx, gfp_t gfp_mask); int (*derive_key)(const struct gss_krb5_enctype *gk5e, const struct xdr_netobj *in, struct xdr_netobj *out, diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 38a0c93e4b60e..e31cfdf7eadcb 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -29,7 +29,6 @@ #endif static struct gss_api_mech gss_kerberos_mech; -static int gss_krb5_import_ctx_v2(struct krb5_ctx *ctx, gfp_t gfp_mask); static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { #if defined(CONFIG_RPCSEC_GSS_KRB5_ENCTYPES_AES_SHA1) @@ -43,7 +42,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .encrypt_name = "cts(cbc(aes))", .aux_cipher = "cbc(aes)", .cksum_name = "hmac(sha1)", - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_derive_key_v2, .encrypt = gss_krb5_aes_encrypt, .decrypt = gss_krb5_aes_decrypt, @@ -73,7 +71,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .encrypt_name = "cts(cbc(aes))", .aux_cipher = "cbc(aes)", .cksum_name = "hmac(sha1)", - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_derive_key_v2, .encrypt = gss_krb5_aes_encrypt, .decrypt = gss_krb5_aes_decrypt, @@ -113,7 +110,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .Ke_length = BITS2OCTETS(128), .Ki_length = BITS2OCTETS(128), - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_kdf_feedback_cmac, .encrypt = gss_krb5_aes_encrypt, .decrypt = gss_krb5_aes_decrypt, @@ -140,7 +136,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .Ke_length = BITS2OCTETS(256), .Ki_length = BITS2OCTETS(256), - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_kdf_feedback_cmac, .encrypt = gss_krb5_aes_encrypt, .decrypt = gss_krb5_aes_decrypt, @@ -170,7 +165,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .Ke_length = BITS2OCTETS(128), .Ki_length = BITS2OCTETS(128), - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_kdf_hmac_sha2, .encrypt = krb5_etm_encrypt, .decrypt = krb5_etm_decrypt, @@ -197,7 +191,6 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .Ke_length = BITS2OCTETS(256), .Ki_length = BITS2OCTETS(192), - .import_ctx = gss_krb5_import_ctx_v2, .derive_key = krb5_kdf_hmac_sha2, .encrypt = krb5_etm_encrypt, .decrypt = krb5_etm_decrypt, @@ -431,9 +424,6 @@ gss_import_v2_context(const void *p, const void *end, struct krb5_ctx *ctx, p = simple_get_bytes(p, end, &ctx->enctype, sizeof(ctx->enctype)); if (IS_ERR(p)) goto out_err; - /* Map ENCTYPE_DES3_CBC_SHA1 to ENCTYPE_DES3_CBC_RAW */ - if (ctx->enctype == ENCTYPE_DES3_CBC_SHA1) - ctx->enctype = ENCTYPE_DES3_CBC_RAW; ctx->gk5e = gss_krb5_lookup_enctype(ctx->enctype); if (ctx->gk5e == NULL) { dprintk("gss_kerberos_mech: unsupported krb5 enctype %u\n", @@ -460,7 +450,7 @@ gss_import_v2_context(const void *p, const void *end, struct krb5_ctx *ctx, } ctx->mech_used.len = gss_kerberos_mech.gm_oid.len; - return ctx->gk5e->import_ctx(ctx, gfp_mask); + return gss_krb5_import_ctx_v2(ctx, gfp_mask); out_err: return PTR_ERR(p); From 2a9893f796a3d5098dd13eae9c2cf8e2f6de5b25 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 29 Jun 2023 13:51:32 -0400 Subject: [PATCH 137/186] SUNRPC: Remove net/sunrpc/auth_gss/gss_krb5_seqnum.c These functions are no longer used. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/auth_gss/Makefile | 2 +- net/sunrpc/auth_gss/gss_krb5_internal.h | 7 -- net/sunrpc/auth_gss/gss_krb5_seqnum.c | 106 ------------------------ 3 files changed, 1 insertion(+), 114 deletions(-) delete mode 100644 net/sunrpc/auth_gss/gss_krb5_seqnum.c diff --git a/net/sunrpc/auth_gss/Makefile b/net/sunrpc/auth_gss/Makefile index 012ae17206894..ad1736d93b763 100644 --- a/net/sunrpc/auth_gss/Makefile +++ b/net/sunrpc/auth_gss/Makefile @@ -12,6 +12,6 @@ auth_rpcgss-y := auth_gss.o gss_generic_token.o \ obj-$(CONFIG_RPCSEC_GSS_KRB5) += rpcsec_gss_krb5.o rpcsec_gss_krb5-y := gss_krb5_mech.o gss_krb5_seal.o gss_krb5_unseal.o \ - gss_krb5_seqnum.o gss_krb5_wrap.o gss_krb5_crypto.o gss_krb5_keys.o + gss_krb5_wrap.o gss_krb5_crypto.o gss_krb5_keys.o obj-$(CONFIG_RPCSEC_GSS_KRB5_KUNIT_TEST) += gss_krb5_test.o diff --git a/net/sunrpc/auth_gss/gss_krb5_internal.h b/net/sunrpc/auth_gss/gss_krb5_internal.h index 9a4b110a6154b..3afd4065bf3d0 100644 --- a/net/sunrpc/auth_gss/gss_krb5_internal.h +++ b/net/sunrpc/auth_gss/gss_krb5_internal.h @@ -153,13 +153,6 @@ static inline int krb5_derive_key(struct krb5_ctx *kctx, return gk5e->derive_key(gk5e, inkey, outkey, &label, gfp_mask); } -s32 krb5_make_seq_num(struct krb5_ctx *kctx, struct crypto_sync_skcipher *key, - int direction, u32 seqnum, unsigned char *cksum, - unsigned char *buf); - -s32 krb5_get_seq_num(struct krb5_ctx *kctx, unsigned char *cksum, - unsigned char *buf, int *direction, u32 *seqnum); - void krb5_make_confounder(u8 *p, int conflen); u32 make_checksum(struct krb5_ctx *kctx, char *header, int hdrlen, diff --git a/net/sunrpc/auth_gss/gss_krb5_seqnum.c b/net/sunrpc/auth_gss/gss_krb5_seqnum.c deleted file mode 100644 index 1babc3474e102..0000000000000 --- a/net/sunrpc/auth_gss/gss_krb5_seqnum.c +++ /dev/null @@ -1,106 +0,0 @@ -/* - * linux/net/sunrpc/gss_krb5_seqnum.c - * - * Adapted from MIT Kerberos 5-1.2.1 lib/gssapi/krb5/util_seqnum.c - * - * Copyright (c) 2000 The Regents of the University of Michigan. - * All rights reserved. - * - * Andy Adamson - */ - -/* - * Copyright 1993 by OpenVision Technologies, Inc. - * - * Permission to use, copy, modify, distribute, and sell this software - * and its documentation for any purpose is hereby granted without fee, - * provided that the above copyright notice appears in all copies and - * that both that copyright notice and this permission notice appear in - * supporting documentation, and that the name of OpenVision not be used - * in advertising or publicity pertaining to distribution of the software - * without specific, written prior permission. OpenVision makes no - * representations about the suitability of this software for any - * purpose. It is provided "as is" without express or implied warranty. - * - * OPENVISION DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, - * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO - * EVENT SHALL OPENVISION BE LIABLE FOR ANY SPECIAL, INDIRECT OR - * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF - * USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR - * OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR - * PERFORMANCE OF THIS SOFTWARE. - */ - -#include -#include -#include - -#include "gss_krb5_internal.h" - -#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) -# define RPCDBG_FACILITY RPCDBG_AUTH -#endif - -s32 -krb5_make_seq_num(struct krb5_ctx *kctx, - struct crypto_sync_skcipher *key, - int direction, - u32 seqnum, - unsigned char *cksum, unsigned char *buf) -{ - unsigned char *plain; - s32 code; - - plain = kmalloc(8, GFP_KERNEL); - if (!plain) - return -ENOMEM; - - plain[0] = (unsigned char) (seqnum & 0xff); - plain[1] = (unsigned char) ((seqnum >> 8) & 0xff); - plain[2] = (unsigned char) ((seqnum >> 16) & 0xff); - plain[3] = (unsigned char) ((seqnum >> 24) & 0xff); - - plain[4] = direction; - plain[5] = direction; - plain[6] = direction; - plain[7] = direction; - - code = krb5_encrypt(key, cksum, plain, buf, 8); - kfree(plain); - return code; -} - -s32 -krb5_get_seq_num(struct krb5_ctx *kctx, - unsigned char *cksum, - unsigned char *buf, - int *direction, u32 *seqnum) -{ - s32 code; - unsigned char *plain; - struct crypto_sync_skcipher *key = kctx->seq; - - dprintk("RPC: krb5_get_seq_num:\n"); - - plain = kmalloc(8, GFP_KERNEL); - if (!plain) - return -ENOMEM; - - if ((code = krb5_decrypt(key, cksum, buf, plain, 8))) - goto out; - - if ((plain[4] != plain[5]) || (plain[4] != plain[6]) || - (plain[4] != plain[7])) { - code = (s32)KG_BAD_SEQ; - goto out; - } - - *direction = plain[4]; - - *seqnum = ((plain[0]) | - (plain[1] << 8) | (plain[2] << 16) | (plain[3] << 24)); - -out: - kfree(plain); - return code; -} From 35308e7f0fc3942edc87d9c6dc78c4a096428957 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:16 -0400 Subject: [PATCH 138/186] NFSD: Refactor nfsd_reply_cache_free_locked() To reduce contention on the bucket locks, we must avoid calling kfree() while each bucket lock is held. Start by refactoring nfsd_reply_cache_free_locked() into a helper that removes an entry from the bucket (and must therefore run under the lock) and a second helper that frees the entry (which does not need to hold the lock). For readability, rename the helpers nfsd_cacherep_. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfscache.c | 27 ++++++++++++++++++++------- 1 file changed, 20 insertions(+), 7 deletions(-) diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index a8eda1c85829e..3fde7f6e4ef87 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -110,21 +110,33 @@ nfsd_reply_cache_alloc(struct svc_rqst *rqstp, __wsum csum, return rp; } +static void nfsd_cacherep_free(struct svc_cacherep *rp) +{ + if (rp->c_type == RC_REPLBUFF) + kfree(rp->c_replvec.iov_base); + kmem_cache_free(drc_slab, rp); +} + static void -nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, - struct nfsd_net *nn) +nfsd_cacherep_unlink_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, + struct svc_cacherep *rp) { - if (rp->c_type == RC_REPLBUFF && rp->c_replvec.iov_base) { + if (rp->c_type == RC_REPLBUFF && rp->c_replvec.iov_base) nfsd_stats_drc_mem_usage_sub(nn, rp->c_replvec.iov_len); - kfree(rp->c_replvec.iov_base); - } if (rp->c_state != RC_UNUSED) { rb_erase(&rp->c_node, &b->rb_head); list_del(&rp->c_lru); atomic_dec(&nn->num_drc_entries); nfsd_stats_drc_mem_usage_sub(nn, sizeof(*rp)); } - kmem_cache_free(drc_slab, rp); +} + +static void +nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, + struct nfsd_net *nn) +{ + nfsd_cacherep_unlink_locked(nn, b, rp); + nfsd_cacherep_free(rp); } static void @@ -132,8 +144,9 @@ nfsd_reply_cache_free(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, struct nfsd_net *nn) { spin_lock(&b->cache_lock); - nfsd_reply_cache_free_locked(b, rp, nn); + nfsd_cacherep_unlink_locked(nn, b, rp); spin_unlock(&b->cache_lock); + nfsd_cacherep_free(rp); } int nfsd_drc_slab_create(void) From ff0d169329768c1102b7b07eebe5a9839aa1c143 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:22 -0400 Subject: [PATCH 139/186] NFSD: Rename nfsd_reply_cache_alloc() For readability, rename to match the other helpers. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfscache.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 3fde7f6e4ef87..02259d280f51c 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -85,8 +85,8 @@ nfsd_hashsize(unsigned int limit) } static struct svc_cacherep * -nfsd_reply_cache_alloc(struct svc_rqst *rqstp, __wsum csum, - struct nfsd_net *nn) +nfsd_cacherep_alloc(struct svc_rqst *rqstp, __wsum csum, + struct nfsd_net *nn) { struct svc_cacherep *rp; @@ -458,7 +458,7 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) * preallocate an entry. */ nn = net_generic(SVC_NET(rqstp), nfsd_net_id); - rp = nfsd_reply_cache_alloc(rqstp, csum, nn); + rp = nfsd_cacherep_alloc(rqstp, csum, nn); if (!rp) goto out; From a9507f6af1450ed26a4a36d979af518f5bb21e5d Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:29 -0400 Subject: [PATCH 140/186] NFSD: Replace nfsd_prune_bucket() Enable nfsd_prune_bucket() to drop the bucket lock while calling kfree(). Use the same pattern that Jeff recently introduced in the NFSD filecache. A few percpu operations are moved outside the lock since they temporarily disable local IRQs which is expensive and does not need to be done while the lock is held. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfscache.c | 78 +++++++++++++++++++++++++++++++++++++--------- fs/nfsd/trace.h | 22 +++++++++++++ 2 files changed, 85 insertions(+), 15 deletions(-) diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 02259d280f51c..787d15b623369 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -117,6 +117,21 @@ static void nfsd_cacherep_free(struct svc_cacherep *rp) kmem_cache_free(drc_slab, rp); } +static unsigned long +nfsd_cacherep_dispose(struct list_head *dispose) +{ + struct svc_cacherep *rp; + unsigned long freed = 0; + + while (!list_empty(dispose)) { + rp = list_first_entry(dispose, struct svc_cacherep, c_lru); + list_del(&rp->c_lru); + nfsd_cacherep_free(rp); + freed++; + } + return freed; +} + static void nfsd_cacherep_unlink_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, struct svc_cacherep *rp) @@ -260,6 +275,41 @@ nfsd_cache_bucket_find(__be32 xid, struct nfsd_net *nn) return &nn->drc_hashtbl[hash]; } +/* + * Remove and return no more than @max expired entries in bucket @b. + * If @max is zero, do not limit the number of removed entries. + */ +static void +nfsd_prune_bucket_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, + unsigned int max, struct list_head *dispose) +{ + unsigned long expiry = jiffies - RC_EXPIRE; + struct svc_cacherep *rp, *tmp; + unsigned int freed = 0; + + lockdep_assert_held(&b->cache_lock); + + /* The bucket LRU is ordered oldest-first. */ + list_for_each_entry_safe(rp, tmp, &b->lru_head, c_lru) { + /* + * Don't free entries attached to calls that are still + * in-progress, but do keep scanning the list. + */ + if (rp->c_state == RC_INPROG) + continue; + + if (atomic_read(&nn->num_drc_entries) <= nn->max_drc_entries && + time_before(expiry, rp->c_timestamp)) + break; + + nfsd_cacherep_unlink_locked(nn, b, rp); + list_add(&rp->c_lru, dispose); + + if (max && ++freed > max) + break; + } +} + static long prune_bucket(struct nfsd_drc_bucket *b, struct nfsd_net *nn, unsigned int max) { @@ -283,11 +333,6 @@ static long prune_bucket(struct nfsd_drc_bucket *b, struct nfsd_net *nn, return freed; } -static long nfsd_prune_bucket(struct nfsd_drc_bucket *b, struct nfsd_net *nn) -{ - return prune_bucket(b, nn, 3); -} - /* * Walk the LRU list and prune off entries that are older than RC_EXPIRE. * Also prune the oldest ones when the total exceeds the max number of entries. @@ -443,6 +488,8 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) __wsum csum; struct nfsd_drc_bucket *b; int type = rqstp->rq_cachetype; + unsigned long freed; + LIST_HEAD(dispose); int rtn = RC_DOIT; rqstp->rq_cacherep = NULL; @@ -467,20 +514,18 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) found = nfsd_cache_insert(b, rp, nn); if (found != rp) goto found_entry; - - nfsd_stats_rc_misses_inc(); rqstp->rq_cacherep = rp; rp->c_state = RC_INPROG; + nfsd_prune_bucket_locked(nn, b, 3, &dispose); + spin_unlock(&b->cache_lock); + freed = nfsd_cacherep_dispose(&dispose); + trace_nfsd_drc_gc(nn, freed); + + nfsd_stats_rc_misses_inc(); atomic_inc(&nn->num_drc_entries); nfsd_stats_drc_mem_usage_add(nn, sizeof(*rp)); - - nfsd_prune_bucket(b, nn); - -out_unlock: - spin_unlock(&b->cache_lock); -out: - return rtn; + goto out; found_entry: /* We found a matching entry which is either in progress or done. */ @@ -518,7 +563,10 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) out_trace: trace_nfsd_drc_found(nn, rqstp, rtn); - goto out_unlock; +out_unlock: + spin_unlock(&b->cache_lock); +out: + return rtn; } /** diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h index 693fe6d465aa6..c48419c0a58a5 100644 --- a/fs/nfsd/trace.h +++ b/fs/nfsd/trace.h @@ -1262,6 +1262,28 @@ TRACE_EVENT(nfsd_drc_mismatch, __entry->ingress) ); +TRACE_EVENT_CONDITION(nfsd_drc_gc, + TP_PROTO( + const struct nfsd_net *nn, + unsigned long freed + ), + TP_ARGS(nn, freed), + TP_CONDITION(freed > 0), + TP_STRUCT__entry( + __field(unsigned long long, boot_time) + __field(unsigned long, freed) + __field(int, total) + ), + TP_fast_assign( + __entry->boot_time = nn->boot_time; + __entry->freed = freed; + __entry->total = atomic_read(&nn->num_drc_entries); + ), + TP_printk("boot_time=%16llx total=%d freed=%lu", + __entry->boot_time, __entry->total, __entry->freed + ) +); + TRACE_EVENT(nfsd_cb_args, TP_PROTO( const struct nfs4_client *clp, From c135e1269f34dfdea4bd94c11060c83a3c0b3c12 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:35 -0400 Subject: [PATCH 141/186] NFSD: Refactor the duplicate reply cache shrinker Avoid holding the bucket lock while freeing cache entries. This change also caps the number of entries that are freed when the shrinker calls to reduce the shrinker's impact on the cache's effectiveness. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfscache.c | 82 ++++++++++++++++++++++------------------------ 1 file changed, 39 insertions(+), 43 deletions(-) diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 787d15b623369..6599a12830dbf 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -310,68 +310,64 @@ nfsd_prune_bucket_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, } } -static long prune_bucket(struct nfsd_drc_bucket *b, struct nfsd_net *nn, - unsigned int max) +/** + * nfsd_reply_cache_count - count_objects method for the DRC shrinker + * @shrink: our registered shrinker context + * @sc: garbage collection parameters + * + * Returns the total number of entries in the duplicate reply cache. To + * keep things simple and quick, this is not the number of expired entries + * in the cache (ie, the number that would be removed by a call to + * nfsd_reply_cache_scan). + */ +static unsigned long +nfsd_reply_cache_count(struct shrinker *shrink, struct shrink_control *sc) { - struct svc_cacherep *rp, *tmp; - long freed = 0; + struct nfsd_net *nn = container_of(shrink, + struct nfsd_net, nfsd_reply_cache_shrinker); - list_for_each_entry_safe(rp, tmp, &b->lru_head, c_lru) { - /* - * Don't free entries attached to calls that are still - * in-progress, but do keep scanning the list. - */ - if (rp->c_state == RC_INPROG) - continue; - if (atomic_read(&nn->num_drc_entries) <= nn->max_drc_entries && - time_before(jiffies, rp->c_timestamp + RC_EXPIRE)) - break; - nfsd_reply_cache_free_locked(b, rp, nn); - if (max && freed++ > max) - break; - } - return freed; + return atomic_read(&nn->num_drc_entries); } -/* - * Walk the LRU list and prune off entries that are older than RC_EXPIRE. - * Also prune the oldest ones when the total exceeds the max number of entries. +/** + * nfsd_reply_cache_scan - scan_objects method for the DRC shrinker + * @shrink: our registered shrinker context + * @sc: garbage collection parameters + * + * Free expired entries on each bucket's LRU list until we've released + * nr_to_scan freed objects. Nothing will be released if the cache + * has not exceeded it's max_drc_entries limit. + * + * Returns the number of entries released by this call. */ -static long -prune_cache_entries(struct nfsd_net *nn) +static unsigned long +nfsd_reply_cache_scan(struct shrinker *shrink, struct shrink_control *sc) { + struct nfsd_net *nn = container_of(shrink, + struct nfsd_net, nfsd_reply_cache_shrinker); + unsigned long freed = 0; + LIST_HEAD(dispose); unsigned int i; - long freed = 0; for (i = 0; i < nn->drc_hashsize; i++) { struct nfsd_drc_bucket *b = &nn->drc_hashtbl[i]; if (list_empty(&b->lru_head)) continue; + spin_lock(&b->cache_lock); - freed += prune_bucket(b, nn, 0); + nfsd_prune_bucket_locked(nn, b, 0, &dispose); spin_unlock(&b->cache_lock); - } - return freed; -} -static unsigned long -nfsd_reply_cache_count(struct shrinker *shrink, struct shrink_control *sc) -{ - struct nfsd_net *nn = container_of(shrink, - struct nfsd_net, nfsd_reply_cache_shrinker); + freed += nfsd_cacherep_dispose(&dispose); + if (freed > sc->nr_to_scan) + break; + } - return atomic_read(&nn->num_drc_entries); + trace_nfsd_drc_gc(nn, freed); + return freed; } -static unsigned long -nfsd_reply_cache_scan(struct shrinker *shrink, struct shrink_control *sc) -{ - struct nfsd_net *nn = container_of(shrink, - struct nfsd_net, nfsd_reply_cache_shrinker); - - return prune_cache_entries(nn); -} /* * Walk an xdr_buf and get a CRC for at most the first RC_CSUMLEN bytes */ From cb18eca4b86768ec79e847795d1043356c9ee3b0 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:41 -0400 Subject: [PATCH 142/186] NFSD: Remove svc_rqst::rq_cacherep Over time I'd like to see NFS-specific fields moved out of struct svc_rqst, which is an RPC layer object. These fields are layering violations. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/cache.h | 6 ++++-- fs/nfsd/nfscache.c | 11 ++++++----- fs/nfsd/nfssvc.c | 10 ++++++---- include/linux/sunrpc/svc.h | 1 - 4 files changed, 16 insertions(+), 12 deletions(-) diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 4c9b87850ab12..27610b0718804 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -84,8 +84,10 @@ int nfsd_net_reply_cache_init(struct nfsd_net *nn); void nfsd_net_reply_cache_destroy(struct nfsd_net *nn); int nfsd_reply_cache_init(struct nfsd_net *); void nfsd_reply_cache_shutdown(struct nfsd_net *); -int nfsd_cache_lookup(struct svc_rqst *); -void nfsd_cache_update(struct svc_rqst *, int, __be32 *); +int nfsd_cache_lookup(struct svc_rqst *rqstp, + struct svc_cacherep **cacherep); +void nfsd_cache_update(struct svc_rqst *rqstp, struct svc_cacherep *rp, + int cachetype, __be32 *statp); int nfsd_reply_cache_stats_show(struct seq_file *m, void *v); #endif /* NFSCACHE_H */ diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 6599a12830dbf..b259fc373ae76 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -465,6 +465,7 @@ nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key, /** * nfsd_cache_lookup - Find an entry in the duplicate reply cache * @rqstp: Incoming Call to find + * @cacherep: OUT: DRC entry for this request * * Try to find an entry matching the current call in the cache. When none * is found, we try to grab the oldest expired entry off the LRU list. If @@ -477,7 +478,7 @@ nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key, * %RC_REPLY: Reply from cache * %RC_DROPIT: Do not process the request further */ -int nfsd_cache_lookup(struct svc_rqst *rqstp) +int nfsd_cache_lookup(struct svc_rqst *rqstp, struct svc_cacherep **cacherep) { struct nfsd_net *nn; struct svc_cacherep *rp, *found; @@ -488,7 +489,6 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) LIST_HEAD(dispose); int rtn = RC_DOIT; - rqstp->rq_cacherep = NULL; if (type == RC_NOCACHE) { nfsd_stats_rc_nocache_inc(); goto out; @@ -510,7 +510,7 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) found = nfsd_cache_insert(b, rp, nn); if (found != rp) goto found_entry; - rqstp->rq_cacherep = rp; + *cacherep = rp; rp->c_state = RC_INPROG; nfsd_prune_bucket_locked(nn, b, 3, &dispose); spin_unlock(&b->cache_lock); @@ -568,6 +568,7 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) /** * nfsd_cache_update - Update an entry in the duplicate reply cache. * @rqstp: svc_rqst with a finished Reply + * @rp: IN: DRC entry for this request * @cachetype: which cache to update * @statp: pointer to Reply's NFS status code, or NULL * @@ -585,10 +586,10 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp) * nfsd failed to encode a reply that otherwise would have been cached. * In this case, nfsd_cache_update is called with statp == NULL. */ -void nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) +void nfsd_cache_update(struct svc_rqst *rqstp, struct svc_cacherep *rp, + int cachetype, __be32 *statp) { struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id); - struct svc_cacherep *rp = rqstp->rq_cacherep; struct kvec *resv = &rqstp->rq_res.head[0], *cachv; struct nfsd_drc_bucket *b; int len; diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 2154fa63c5f2c..f91fb343313de 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -1046,6 +1046,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) { const struct svc_procedure *proc = rqstp->rq_procinfo; __be32 *statp = rqstp->rq_accept_statp; + struct svc_cacherep *rp; /* * Give the xdr decoder a chance to change this if it wants @@ -1056,7 +1057,8 @@ int nfsd_dispatch(struct svc_rqst *rqstp) if (!proc->pc_decode(rqstp, &rqstp->rq_arg_stream)) goto out_decode_err; - switch (nfsd_cache_lookup(rqstp)) { + rp = NULL; + switch (nfsd_cache_lookup(rqstp, &rp)) { case RC_DOIT: break; case RC_REPLY: @@ -1072,7 +1074,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) if (!proc->pc_encode(rqstp, &rqstp->rq_res_stream)) goto out_encode_err; - nfsd_cache_update(rqstp, rqstp->rq_cachetype, statp + 1); + nfsd_cache_update(rqstp, rp, rqstp->rq_cachetype, statp + 1); out_cached_reply: return 1; @@ -1082,13 +1084,13 @@ int nfsd_dispatch(struct svc_rqst *rqstp) return 1; out_update_drop: - nfsd_cache_update(rqstp, RC_NOCACHE, NULL); + nfsd_cache_update(rqstp, rp, RC_NOCACHE, NULL); out_dropit: return 0; out_encode_err: trace_nfsd_cant_encode_err(rqstp); - nfsd_cache_update(rqstp, RC_NOCACHE, NULL); + nfsd_cache_update(rqstp, rp, RC_NOCACHE, NULL); *statp = rpc_system_err; return 1; } diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index f8751118c1221..fe1394cc13716 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -265,7 +265,6 @@ struct svc_rqst { /* Catering to nfsd */ struct auth_domain * rq_client; /* RPC peer info */ struct auth_domain * rq_gssclient; /* "gss/"-style peer info */ - struct svc_cacherep * rq_cacherep; /* cache info */ struct task_struct *rq_task; /* service thread */ struct net *rq_bc_net; /* pointer to backchannel's * net namespace From e7421ce71437ec8e4d69cc6bdf35b6853adc5050 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 9 Jul 2023 11:45:48 -0400 Subject: [PATCH 143/186] NFSD: Rename struct svc_cacherep The svc_ prefix is identified with the SunRPC layer. Although the duplicate reply cache caches RPC replies, it is only for the NFS protocol. Rename the struct to better reflect its purpose. Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/cache.h | 6 +++--- fs/nfsd/nfscache.c | 44 ++++++++++++++++++++++---------------------- fs/nfsd/nfssvc.c | 2 +- fs/nfsd/trace.h | 4 ++-- 4 files changed, 28 insertions(+), 28 deletions(-) diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 27610b0718804..929248c6ca84c 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -19,7 +19,7 @@ * typical sockaddr_storage. This is for space reasons, since sockaddr_storage * is much larger than a sockaddr_in6. */ -struct svc_cacherep { +struct nfsd_cacherep { struct { /* Keep often-read xid, csum in the same cache line: */ __be32 k_xid; @@ -85,8 +85,8 @@ void nfsd_net_reply_cache_destroy(struct nfsd_net *nn); int nfsd_reply_cache_init(struct nfsd_net *); void nfsd_reply_cache_shutdown(struct nfsd_net *); int nfsd_cache_lookup(struct svc_rqst *rqstp, - struct svc_cacherep **cacherep); -void nfsd_cache_update(struct svc_rqst *rqstp, struct svc_cacherep *rp, + struct nfsd_cacherep **cacherep); +void nfsd_cache_update(struct svc_rqst *rqstp, struct nfsd_cacherep *rp, int cachetype, __be32 *statp); int nfsd_reply_cache_stats_show(struct seq_file *m, void *v); diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index b259fc373ae76..80621a7095107 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -84,11 +84,11 @@ nfsd_hashsize(unsigned int limit) return roundup_pow_of_two(limit / TARGET_BUCKET_SIZE); } -static struct svc_cacherep * +static struct nfsd_cacherep * nfsd_cacherep_alloc(struct svc_rqst *rqstp, __wsum csum, struct nfsd_net *nn) { - struct svc_cacherep *rp; + struct nfsd_cacherep *rp; rp = kmem_cache_alloc(drc_slab, GFP_KERNEL); if (rp) { @@ -110,7 +110,7 @@ nfsd_cacherep_alloc(struct svc_rqst *rqstp, __wsum csum, return rp; } -static void nfsd_cacherep_free(struct svc_cacherep *rp) +static void nfsd_cacherep_free(struct nfsd_cacherep *rp) { if (rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); @@ -120,11 +120,11 @@ static void nfsd_cacherep_free(struct svc_cacherep *rp) static unsigned long nfsd_cacherep_dispose(struct list_head *dispose) { - struct svc_cacherep *rp; + struct nfsd_cacherep *rp; unsigned long freed = 0; while (!list_empty(dispose)) { - rp = list_first_entry(dispose, struct svc_cacherep, c_lru); + rp = list_first_entry(dispose, struct nfsd_cacherep, c_lru); list_del(&rp->c_lru); nfsd_cacherep_free(rp); freed++; @@ -134,7 +134,7 @@ nfsd_cacherep_dispose(struct list_head *dispose) static void nfsd_cacherep_unlink_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, - struct svc_cacherep *rp) + struct nfsd_cacherep *rp) { if (rp->c_type == RC_REPLBUFF && rp->c_replvec.iov_base) nfsd_stats_drc_mem_usage_sub(nn, rp->c_replvec.iov_len); @@ -147,7 +147,7 @@ nfsd_cacherep_unlink_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, } static void -nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, +nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct nfsd_cacherep *rp, struct nfsd_net *nn) { nfsd_cacherep_unlink_locked(nn, b, rp); @@ -155,7 +155,7 @@ nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, } static void -nfsd_reply_cache_free(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, +nfsd_reply_cache_free(struct nfsd_drc_bucket *b, struct nfsd_cacherep *rp, struct nfsd_net *nn) { spin_lock(&b->cache_lock); @@ -167,7 +167,7 @@ nfsd_reply_cache_free(struct nfsd_drc_bucket *b, struct svc_cacherep *rp, int nfsd_drc_slab_create(void) { drc_slab = kmem_cache_create("nfsd_drc", - sizeof(struct svc_cacherep), 0, 0, NULL); + sizeof(struct nfsd_cacherep), 0, 0, NULL); return drc_slab ? 0: -ENOMEM; } @@ -236,7 +236,7 @@ int nfsd_reply_cache_init(struct nfsd_net *nn) void nfsd_reply_cache_shutdown(struct nfsd_net *nn) { - struct svc_cacherep *rp; + struct nfsd_cacherep *rp; unsigned int i; unregister_shrinker(&nn->nfsd_reply_cache_shrinker); @@ -244,7 +244,7 @@ void nfsd_reply_cache_shutdown(struct nfsd_net *nn) for (i = 0; i < nn->drc_hashsize; i++) { struct list_head *head = &nn->drc_hashtbl[i].lru_head; while (!list_empty(head)) { - rp = list_first_entry(head, struct svc_cacherep, c_lru); + rp = list_first_entry(head, struct nfsd_cacherep, c_lru); nfsd_reply_cache_free_locked(&nn->drc_hashtbl[i], rp, nn); } @@ -261,7 +261,7 @@ void nfsd_reply_cache_shutdown(struct nfsd_net *nn) * not already scheduled. */ static void -lru_put_end(struct nfsd_drc_bucket *b, struct svc_cacherep *rp) +lru_put_end(struct nfsd_drc_bucket *b, struct nfsd_cacherep *rp) { rp->c_timestamp = jiffies; list_move_tail(&rp->c_lru, &b->lru_head); @@ -284,7 +284,7 @@ nfsd_prune_bucket_locked(struct nfsd_net *nn, struct nfsd_drc_bucket *b, unsigned int max, struct list_head *dispose) { unsigned long expiry = jiffies - RC_EXPIRE; - struct svc_cacherep *rp, *tmp; + struct nfsd_cacherep *rp, *tmp; unsigned int freed = 0; lockdep_assert_held(&b->cache_lock); @@ -402,8 +402,8 @@ nfsd_cache_csum(struct svc_rqst *rqstp) } static int -nfsd_cache_key_cmp(const struct svc_cacherep *key, - const struct svc_cacherep *rp, struct nfsd_net *nn) +nfsd_cache_key_cmp(const struct nfsd_cacherep *key, + const struct nfsd_cacherep *rp, struct nfsd_net *nn) { if (key->c_key.k_xid == rp->c_key.k_xid && key->c_key.k_csum != rp->c_key.k_csum) { @@ -419,11 +419,11 @@ nfsd_cache_key_cmp(const struct svc_cacherep *key, * Must be called with cache_lock held. Returns the found entry or * inserts an empty key on failure. */ -static struct svc_cacherep * -nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key, +static struct nfsd_cacherep * +nfsd_cache_insert(struct nfsd_drc_bucket *b, struct nfsd_cacherep *key, struct nfsd_net *nn) { - struct svc_cacherep *rp, *ret = key; + struct nfsd_cacherep *rp, *ret = key; struct rb_node **p = &b->rb_head.rb_node, *parent = NULL; unsigned int entries = 0; @@ -432,7 +432,7 @@ nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key, while (*p != NULL) { ++entries; parent = *p; - rp = rb_entry(parent, struct svc_cacherep, c_node); + rp = rb_entry(parent, struct nfsd_cacherep, c_node); cmp = nfsd_cache_key_cmp(key, rp, nn); if (cmp < 0) @@ -478,10 +478,10 @@ nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key, * %RC_REPLY: Reply from cache * %RC_DROPIT: Do not process the request further */ -int nfsd_cache_lookup(struct svc_rqst *rqstp, struct svc_cacherep **cacherep) +int nfsd_cache_lookup(struct svc_rqst *rqstp, struct nfsd_cacherep **cacherep) { struct nfsd_net *nn; - struct svc_cacherep *rp, *found; + struct nfsd_cacherep *rp, *found; __wsum csum; struct nfsd_drc_bucket *b; int type = rqstp->rq_cachetype; @@ -586,7 +586,7 @@ int nfsd_cache_lookup(struct svc_rqst *rqstp, struct svc_cacherep **cacherep) * nfsd failed to encode a reply that otherwise would have been cached. * In this case, nfsd_cache_update is called with statp == NULL. */ -void nfsd_cache_update(struct svc_rqst *rqstp, struct svc_cacherep *rp, +void nfsd_cache_update(struct svc_rqst *rqstp, struct nfsd_cacherep *rp, int cachetype, __be32 *statp) { struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id); diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index f91fb343313de..97830e28c140c 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -1046,7 +1046,7 @@ int nfsd_dispatch(struct svc_rqst *rqstp) { const struct svc_procedure *proc = rqstp->rq_procinfo; __be32 *statp = rqstp->rq_accept_statp; - struct svc_cacherep *rp; + struct nfsd_cacherep *rp; /* * Give the xdr decoder a chance to change this if it wants diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h index c48419c0a58a5..8039043488718 100644 --- a/fs/nfsd/trace.h +++ b/fs/nfsd/trace.h @@ -1241,8 +1241,8 @@ TRACE_EVENT(nfsd_drc_found, TRACE_EVENT(nfsd_drc_mismatch, TP_PROTO( const struct nfsd_net *nn, - const struct svc_cacherep *key, - const struct svc_cacherep *rp + const struct nfsd_cacherep *key, + const struct nfsd_cacherep *rp ), TP_ARGS(nn, key, rp), TP_STRUCT__entry( From 5865bafa197a90ecadd086c2874948fa0c474943 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 20 Jul 2023 09:34:53 -0400 Subject: [PATCH 144/186] nfsd: add a MODULE_DESCRIPTION I got this today from modpost: WARNING: modpost: missing MODULE_DESCRIPTION() in fs/nfsd/nfsd.o Add a module description. Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfsctl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 4302ca0ff6ed5..33f80d289d638 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1627,6 +1627,7 @@ static void __exit exit_nfsd(void) } MODULE_AUTHOR("Olaf Kirch "); +MODULE_DESCRIPTION("In-kernel NFS server"); MODULE_LICENSE("GPL"); module_init(init_nfsd) module_exit(exit_nfsd) From a332018a91c419a9a475c41d827544c771986876 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jul 2023 10:29:10 -0400 Subject: [PATCH 145/186] nfsd: handle failure to collect pre/post-op attrs more sanely Collecting pre_op_attrs can fail, in which case it's probably best to fail the whole operation. Change fh_fill_pre_attrs and fh_fill_both_attrs to return __be32, and have the callers check the return code and abort the operation if it's not nfs_ok. Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs3proc.c | 4 +++- fs/nfsd/nfs4proc.c | 14 +++++++------ fs/nfsd/nfsfh.c | 26 ++++++++++++++--------- fs/nfsd/nfsfh.h | 6 +++--- fs/nfsd/vfs.c | 52 +++++++++++++++++++++++++++++++--------------- 5 files changed, 65 insertions(+), 37 deletions(-) diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index fc8d5b7db9f81..268ef57751c48 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -307,7 +307,9 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp, if (!IS_POSIXACL(inode)) iap->ia_mode &= ~current_umask(); - fh_fill_pre_attrs(fhp); + status = fh_fill_pre_attrs(fhp); + if (status != nfs_ok) + goto out; host_err = vfs_create(&nop_mnt_idmap, inode, child, iap->ia_mode, true); if (host_err < 0) { status = nfserrno(host_err); diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 5ae670807449b..cdf58d181c9e6 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -297,12 +297,12 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp, } if (d_really_is_positive(child)) { - status = nfs_ok; - /* NFSv4 protocol requires change attributes even though * no change happened. */ - fh_fill_both_attrs(fhp); + status = fh_fill_both_attrs(fhp); + if (status != nfs_ok) + goto out; switch (open->op_createmode) { case NFS4_CREATE_UNCHECKED: @@ -345,7 +345,9 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp, if (!IS_POSIXACL(inode)) iap->ia_mode &= ~current_umask(); - fh_fill_pre_attrs(fhp); + status = fh_fill_pre_attrs(fhp); + if (status != nfs_ok) + goto out; status = nfsd4_vfs_create(fhp, child, open); if (status != nfs_ok) goto out; @@ -424,11 +426,11 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru } else { status = nfsd_lookup(rqstp, current_fh, open->op_fname, open->op_fnamelen, *resfh); - if (!status) + if (status == nfs_ok) /* NFSv4 protocol requires change attributes even though * no change happened. */ - fh_fill_both_attrs(current_fh); + status = fh_fill_both_attrs(current_fh); } if (status) goto out; diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index c291389a1d71d..355bf0db3235b 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -614,7 +614,7 @@ fh_update(struct svc_fh *fhp) * @fhp: file handle to be updated * */ -void fh_fill_pre_attrs(struct svc_fh *fhp) +__be32 __must_check fh_fill_pre_attrs(struct svc_fh *fhp) { bool v4 = (fhp->fh_maxsize == NFS4_FHSIZE); struct inode *inode; @@ -622,12 +622,12 @@ void fh_fill_pre_attrs(struct svc_fh *fhp) __be32 err; if (fhp->fh_no_wcc || fhp->fh_pre_saved) - return; + return nfs_ok; inode = d_inode(fhp->fh_dentry); err = fh_getattr(fhp, &stat); if (err) - return; + return err; if (v4) fhp->fh_pre_change = nfsd4_change_attribute(&stat, inode); @@ -636,6 +636,7 @@ void fh_fill_pre_attrs(struct svc_fh *fhp) fhp->fh_pre_ctime = stat.ctime; fhp->fh_pre_size = stat.size; fhp->fh_pre_saved = true; + return nfs_ok; } /** @@ -643,26 +644,27 @@ void fh_fill_pre_attrs(struct svc_fh *fhp) * @fhp: file handle to be updated * */ -void fh_fill_post_attrs(struct svc_fh *fhp) +__be32 fh_fill_post_attrs(struct svc_fh *fhp) { bool v4 = (fhp->fh_maxsize == NFS4_FHSIZE); struct inode *inode = d_inode(fhp->fh_dentry); __be32 err; if (fhp->fh_no_wcc) - return; + return nfs_ok; if (fhp->fh_post_saved) printk("nfsd: inode locked twice during operation.\n"); err = fh_getattr(fhp, &fhp->fh_post_attr); if (err) - return; + return err; fhp->fh_post_saved = true; if (v4) fhp->fh_post_change = nfsd4_change_attribute(&fhp->fh_post_attr, inode); + return nfs_ok; } /** @@ -672,16 +674,20 @@ void fh_fill_post_attrs(struct svc_fh *fhp) * This is used when the directory wasn't changed, but wcc attributes * are needed anyway. */ -void fh_fill_both_attrs(struct svc_fh *fhp) +__be32 __must_check fh_fill_both_attrs(struct svc_fh *fhp) { - fh_fill_post_attrs(fhp); - if (!fhp->fh_post_saved) - return; + __be32 err; + + err = fh_fill_post_attrs(fhp); + if (err) + return err; + fhp->fh_pre_change = fhp->fh_post_change; fhp->fh_pre_mtime = fhp->fh_post_attr.mtime; fhp->fh_pre_ctime = fhp->fh_post_attr.ctime; fhp->fh_pre_size = fhp->fh_post_attr.size; fhp->fh_pre_saved = true; + return nfs_ok; } /* diff --git a/fs/nfsd/nfsfh.h b/fs/nfsd/nfsfh.h index 4e0ecf0ae2cf2..40426f899e760 100644 --- a/fs/nfsd/nfsfh.h +++ b/fs/nfsd/nfsfh.h @@ -294,7 +294,7 @@ static inline void fh_clear_pre_post_attrs(struct svc_fh *fhp) } u64 nfsd4_change_attribute(struct kstat *stat, struct inode *inode); -extern void fh_fill_pre_attrs(struct svc_fh *fhp); -extern void fh_fill_post_attrs(struct svc_fh *fhp); -extern void fh_fill_both_attrs(struct svc_fh *fhp); +__be32 __must_check fh_fill_pre_attrs(struct svc_fh *fhp); +__be32 fh_fill_post_attrs(struct svc_fh *fhp); +__be32 __must_check fh_fill_both_attrs(struct svc_fh *fhp); #endif /* _LINUX_NFSD_NFSFH_H */ diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 2c9074ab2315a..c463ef5e08211 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1540,7 +1540,9 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, dput(dchild); if (err) goto out_unlock; - fh_fill_pre_attrs(fhp); + err = fh_fill_pre_attrs(fhp); + if (err != nfs_ok) + goto out_unlock; err = nfsd_create_locked(rqstp, fhp, attrs, type, rdev, resfhp); fh_fill_post_attrs(fhp); out_unlock: @@ -1635,13 +1637,16 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp, inode_unlock(dentry->d_inode); goto out_drop_write; } - fh_fill_pre_attrs(fhp); + err = fh_fill_pre_attrs(fhp); + if (err != nfs_ok) + goto out_unlock; host_err = vfs_symlink(&nop_mnt_idmap, d_inode(dentry), dnew, path); err = nfserrno(host_err); cerr = fh_compose(resfhp, fhp->fh_export, dnew, fhp); if (!err) nfsd_create_setattr(rqstp, fhp, resfhp, attrs); fh_fill_post_attrs(fhp); +out_unlock: inode_unlock(dentry->d_inode); if (!err) err = nfserrno(commit_metadata(fhp)); @@ -1703,7 +1708,9 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, err = nfserr_noent; if (d_really_is_negative(dold)) goto out_dput; - fh_fill_pre_attrs(ffhp); + err = fh_fill_pre_attrs(ffhp); + if (err != nfs_ok) + goto out_dput; host_err = vfs_link(dold, &nop_mnt_idmap, dirp, dnew, NULL); fh_fill_post_attrs(ffhp); inode_unlock(dirp); @@ -1789,8 +1796,12 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen, } trap = lock_rename(tdentry, fdentry); - fh_fill_pre_attrs(ffhp); - fh_fill_pre_attrs(tfhp); + err = fh_fill_pre_attrs(ffhp); + if (err != nfs_ok) + goto out_unlock; + err = fh_fill_pre_attrs(tfhp); + if (err != nfs_ok) + goto out_unlock; odentry = lookup_one_len(fname, fdentry, flen); host_err = PTR_ERR(odentry); @@ -1857,6 +1868,7 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen, fh_fill_post_attrs(ffhp); fh_fill_post_attrs(tfhp); } +out_unlock: unlock_rename(tdentry, fdentry); fh_drop_write(ffhp); @@ -1916,12 +1928,14 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, goto out_unlock; } rinode = d_inode(rdentry); - ihold(rinode); + err = fh_fill_pre_attrs(fhp); + if (err != nfs_ok) + goto out_unlock; + ihold(rinode); if (!type) type = d_inode(rdentry)->i_mode & S_IFMT; - fh_fill_pre_attrs(fhp); if (type != S_IFDIR) { int retries; @@ -2341,16 +2355,18 @@ nfsd_removexattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name) return nfserrno(ret); inode_lock(fhp->fh_dentry->d_inode); - fh_fill_pre_attrs(fhp); - + err = fh_fill_pre_attrs(fhp); + if (err != nfs_ok) + goto out_unlock; ret = __vfs_removexattr_locked(&nop_mnt_idmap, fhp->fh_dentry, name, NULL); - + err = nfsd_xattr_errno(ret); fh_fill_post_attrs(fhp); +out_unlock: inode_unlock(fhp->fh_dentry->d_inode); fh_drop_write(fhp); - return nfsd_xattr_errno(ret); + return err; } __be32 @@ -2368,15 +2384,17 @@ nfsd_setxattr(struct svc_rqst *rqstp, struct svc_fh *fhp, char *name, if (ret) return nfserrno(ret); inode_lock(fhp->fh_dentry->d_inode); - fh_fill_pre_attrs(fhp); - - ret = __vfs_setxattr_locked(&nop_mnt_idmap, fhp->fh_dentry, name, buf, - len, flags, NULL); + err = fh_fill_pre_attrs(fhp); + if (err != nfs_ok) + goto out_unlock; + ret = __vfs_setxattr_locked(&nop_mnt_idmap, fhp->fh_dentry, + name, buf, len, flags, NULL); fh_fill_post_attrs(fhp); + err = nfsd_xattr_errno(ret); +out_unlock: inode_unlock(fhp->fh_dentry->d_inode); fh_drop_write(fhp); - - return nfsd_xattr_errno(ret); + return err; } #endif From 976626073a7502fc91416155bd037a29deee729b Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 21 Jul 2023 10:29:11 -0400 Subject: [PATCH 146/186] nfsd: remove unsafe BUG_ON from set_change_info At one time, nfsd would scrape inode information directly out of struct inode in order to populate the change_info4. At that time, the BUG_ON in set_change_info made some sense, since having it unset meant a coding error. More recently, it calls vfs_getattr to get this information, which can fail. If that fails, fh_pre_saved can end up not being set. While this situation is unfortunate, we don't need to crash the box. Move set_change_info to nfs4proc.c since all of the callers are there. Revise the condition for setting "atomic" to also check for fh_pre_saved. Drop the BUG_ON and just have it zero out both change_attr4s when this occurs. Reported-by: Boyang Xue Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2223560 Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs4proc.c | 32 ++++++++++++++++++++++++++++++++ fs/nfsd/xdr4.h | 11 ----------- 2 files changed, 32 insertions(+), 11 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index cdf58d181c9e6..a9c84339ff659 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -382,6 +382,38 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp, return status; } +/** + * set_change_info - set up the change_info4 for a reply + * @cinfo: pointer to nfsd4_change_info to be populated + * @fhp: pointer to svc_fh to use as source + * + * Many operations in NFSv4 require change_info4 in the reply. This function + * populates that from the info that we (should!) have already collected. In + * the event that we didn't get any pre-attrs, just zero out both. + */ +static void +set_change_info(struct nfsd4_change_info *cinfo, struct svc_fh *fhp) +{ + cinfo->atomic = (u32)(fhp->fh_pre_saved && fhp->fh_post_saved && !fhp->fh_no_atomic_attr); + cinfo->before_change = fhp->fh_pre_change; + cinfo->after_change = fhp->fh_post_change; + + /* + * If fetching the pre-change attributes failed, then we should + * have already failed the whole operation. We could have still + * failed to fetch post-change attributes however. + * + * If we didn't get post-op attrs, just zero-out the after + * field since we don't know what it should be. If the pre_saved + * field isn't set for some reason, throw warning and just copy + * whatever is in the after field. + */ + if (WARN_ON_ONCE(!fhp->fh_pre_saved)) + cinfo->before_change = 0; + if (!fhp->fh_post_saved) + cinfo->after_change = 0; +} + static __be32 do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_open *open, struct svc_fh **resfh) { diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h index 510978e602da6..9d918a79dc166 100644 --- a/fs/nfsd/xdr4.h +++ b/fs/nfsd/xdr4.h @@ -774,17 +774,6 @@ void warn_on_nonidempotent_op(struct nfsd4_op *op); #define NFS4_SVC_XDRSIZE sizeof(struct nfsd4_compoundargs) -static inline void -set_change_info(struct nfsd4_change_info *cinfo, struct svc_fh *fhp) -{ - BUG_ON(!fhp->fh_pre_saved); - cinfo->atomic = (u32)(fhp->fh_post_saved && !fhp->fh_no_atomic_attr); - - cinfo->before_change = fhp->fh_pre_change; - cinfo->after_change = fhp->fh_post_change; -} - - bool nfsd4_mach_creds_match(struct nfs4_client *cl, struct svc_rqst *rqstp); bool nfs4svc_decode_compoundargs(struct svc_rqst *rqstp, struct xdr_stream *xdr); bool nfs4svc_encode_compoundres(struct svc_rqst *rqstp, struct xdr_stream *xdr); From f2b7019d2e3c4f1f55be658659804b337dcfac60 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 24 Jul 2023 10:53:39 -0400 Subject: [PATCH 147/186] nfsd: set missing after_change as before_change + 1 In the event that we can't fetch post_op_attr attributes, we still need to set a value for the after_change. The operation has already happened, so we're not able to return an error at that point, but we do want to ensure that the client knows that its cache should be invalidated. If we weren't able to fetch post-op attrs, then just set the after_change to before_change + 1. The atomic flag should already be clear in this case. Suggested-by: Neil Brown Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index a9c84339ff659..7588fd1859a4f 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -411,7 +411,7 @@ set_change_info(struct nfsd4_change_info *cinfo, struct svc_fh *fhp) if (WARN_ON_ONCE(!fhp->fh_pre_saved)) cinfo->before_change = 0; if (!fhp->fh_post_saved) - cinfo->after_change = 0; + cinfo->after_change = cinfo->before_change + 1; } static __be32 From be2be5f7f4436442d8f6bffbb97a6f438df2896b Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 20 Jul 2023 08:58:04 -0400 Subject: [PATCH 148/186] lockd: nlm_blocked list race fixes This patch fixes races when lockd accesses the global nlm_blocked list. It was mostly safe to access the list because everything was accessed from the lockd kernel thread context but there exist cases like nlmsvc_grant_deferred() that could manipulate the nlm_blocked list and it can be called from any context. Signed-off-by: Alexander Aring Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/lockd/svclock.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index c43ccdf28ed91..28abec5c451d1 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -131,12 +131,14 @@ static void nlmsvc_insert_block(struct nlm_block *block, unsigned long when) static inline void nlmsvc_remove_block(struct nlm_block *block) { + spin_lock(&nlm_blocked_lock); if (!list_empty(&block->b_list)) { - spin_lock(&nlm_blocked_lock); list_del_init(&block->b_list); spin_unlock(&nlm_blocked_lock); nlmsvc_release_block(block); + return; } + spin_unlock(&nlm_blocked_lock); } /* @@ -152,6 +154,7 @@ nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock) file, lock->fl.fl_pid, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end, lock->fl.fl_type); + spin_lock(&nlm_blocked_lock); list_for_each_entry(block, &nlm_blocked, b_list) { fl = &block->b_call->a_args.lock.fl; dprintk("lockd: check f=%p pd=%d %Ld-%Ld ty=%d cookie=%s\n", @@ -161,9 +164,11 @@ nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock) nlmdbg_cookie2a(&block->b_call->a_args.cookie)); if (block->b_file == file && nlm_compare_locks(fl, &lock->fl)) { kref_get(&block->b_count); + spin_unlock(&nlm_blocked_lock); return block; } } + spin_unlock(&nlm_blocked_lock); return NULL; } @@ -185,16 +190,19 @@ nlmsvc_find_block(struct nlm_cookie *cookie) { struct nlm_block *block; + spin_lock(&nlm_blocked_lock); list_for_each_entry(block, &nlm_blocked, b_list) { if (nlm_cookie_match(&block->b_call->a_args.cookie,cookie)) goto found; } + spin_unlock(&nlm_blocked_lock); return NULL; found: dprintk("nlmsvc_find_block(%s): block=%p\n", nlmdbg_cookie2a(cookie), block); kref_get(&block->b_count); + spin_unlock(&nlm_blocked_lock); return block; } @@ -317,6 +325,7 @@ void nlmsvc_traverse_blocks(struct nlm_host *host, restart: mutex_lock(&file->f_mutex); + spin_lock(&nlm_blocked_lock); list_for_each_entry_safe(block, next, &file->f_blocks, b_flist) { if (!match(block->b_host, host)) continue; @@ -325,11 +334,13 @@ void nlmsvc_traverse_blocks(struct nlm_host *host, if (list_empty(&block->b_list)) continue; kref_get(&block->b_count); + spin_unlock(&nlm_blocked_lock); mutex_unlock(&file->f_mutex); nlmsvc_unlink_block(block); nlmsvc_release_block(block); goto restart; } + spin_unlock(&nlm_blocked_lock); mutex_unlock(&file->f_mutex); } From f80774787aa2b719d9c5f2d67a5901b59f219ce7 Mon Sep 17 00:00:00 2001 From: YueHaibing Date: Sat, 22 Jul 2023 11:31:16 +0800 Subject: [PATCH 149/186] sunrpc: Remove unused extern declarations Since commit 49b28684fdba ("nfsd: Remove deprecated nfsctl system call and related code.") these declarations are unused, so can remove it. Signed-off-by: YueHaibing Signed-off-by: Chuck Lever --- include/linux/sunrpc/svcauth.h | 3 --- 1 file changed, 3 deletions(-) diff --git a/include/linux/sunrpc/svcauth.h b/include/linux/sunrpc/svcauth.h index 6d9cc9080aca7..27582d3b538f2 100644 --- a/include/linux/sunrpc/svcauth.h +++ b/include/linux/sunrpc/svcauth.h @@ -157,11 +157,8 @@ extern void svc_auth_unregister(rpc_authflavor_t flavor); extern struct auth_domain *unix_domain_find(char *name); extern void auth_domain_put(struct auth_domain *item); -extern int auth_unix_add_addr(struct net *net, struct in6_addr *addr, struct auth_domain *dom); extern struct auth_domain *auth_domain_lookup(char *name, struct auth_domain *new); extern struct auth_domain *auth_domain_find(char *name); -extern struct auth_domain *auth_unix_lookup(struct net *net, struct in6_addr *addr); -extern int auth_unix_forget_old(struct auth_domain *dom); extern void svcauth_unix_purge(struct net *net); extern void svcauth_unix_info_release(struct svc_xprt *xpt); extern int svcauth_unix_set_client(struct svc_rqst *rqstp); From d424797032c6e24b44037e6c7a2d32fd958300f0 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 24 Jul 2023 08:13:05 -0400 Subject: [PATCH 150/186] nfsd: inherit required unset default acls from effective set A well-formed NFSv4 ACL will always contain OWNER@/GROUP@/EVERYONE@ ACEs, but there is no requirement for inheritable entries for those entities. POSIX ACLs must always have owner/group/other entries, even for a default ACL. nfsd builds the default ACL from inheritable ACEs, but the current code just leaves any unspecified ACEs zeroed out. The result is that adding a default user or group ACE to an inode can leave it with unwanted deny entries. For instance, a newly created directory with no acl will look something like this: # NFSv4 translation by server A::OWNER@:rwaDxtTcCy A::GROUP@:rxtcy A::EVERYONE@:rxtcy # POSIX ACL of underlying file user::rwx group::r-x other::r-x ...if I then add new v4 ACE: nfs4_setfacl -a A:fd:1000:rwx /mnt/local/test ...I end up with a result like this today: user::rwx user:1000:rwx group::r-x mask::rwx other::r-x default:user::--- default:user:1000:rwx default:group::--- default:mask::rwx default:other::--- A::OWNER@:rwaDxtTcCy A::1000:rwaDxtcy A::GROUP@:rxtcy A::EVERYONE@:rxtcy D:fdi:OWNER@:rwaDx A:fdi:OWNER@:tTcCy A:fdi:1000:rwaDxtcy A:fdi:GROUP@:tcy A:fdi:EVERYONE@:tcy ...which is not at all expected. Adding a single inheritable allow ACE should not result in everyone else losing access. The setfacl command solves a silimar issue by copying owner/group/other entries from the effective ACL when none of them are set: "If a Default ACL entry is created, and the Default ACL contains no owner, owning group, or others entry, a copy of the ACL owner, owning group, or others entry is added to the Default ACL. Having nfsd do the same provides a more sane result (with no deny ACEs in the resulting set): user::rwx user:1000:rwx group::r-x mask::rwx other::r-x default:user::rwx default:user:1000:rwx default:group::r-x default:mask::rwx default:other::r-x A::OWNER@:rwaDxtTcCy A::1000:rwaDxtcy A::GROUP@:rxtcy A::EVERYONE@:rxtcy A:fdi:OWNER@:rwaDxtTcCy A:fdi:1000:rwaDxtcy A:fdi:GROUP@:rxtcy A:fdi:EVERYONE@:rxtcy Reported-by: Ondrej Valousek Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2136452 Suggested-by: Andreas Gruenbacher Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs4acl.c | 34 +++++++++++++++++++++++++++++----- 1 file changed, 29 insertions(+), 5 deletions(-) diff --git a/fs/nfsd/nfs4acl.c b/fs/nfsd/nfs4acl.c index 518203821790c..96e786b5e5444 100644 --- a/fs/nfsd/nfs4acl.c +++ b/fs/nfsd/nfs4acl.c @@ -441,7 +441,7 @@ struct posix_ace_state_array { * calculated so far: */ struct posix_acl_state { - int empty; + unsigned char valid; struct posix_ace_state owner; struct posix_ace_state group; struct posix_ace_state other; @@ -457,7 +457,6 @@ init_state(struct posix_acl_state *state, int cnt) int alloc; memset(state, 0, sizeof(struct posix_acl_state)); - state->empty = 1; /* * In the worst case, each individual acl could be for a distinct * named user or group, but we don't know which, so we allocate @@ -500,7 +499,7 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) * and effective cases: when there are no inheritable ACEs, * calls ->set_acl with a NULL ACL structure. */ - if (state->empty && (flags & NFS4_ACL_TYPE_DEFAULT)) + if (!state->valid && (flags & NFS4_ACL_TYPE_DEFAULT)) return NULL; /* @@ -622,11 +621,12 @@ static void process_one_v4_ace(struct posix_acl_state *state, struct nfs4_ace *ace) { u32 mask = ace->access_mask; + short type = ace2type(ace); int i; - state->empty = 0; + state->valid |= type; - switch (ace2type(ace)) { + switch (type) { case ACL_USER_OBJ: if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) { allow_bits(&state->owner, mask); @@ -726,6 +726,30 @@ static int nfs4_acl_nfsv4_to_posix(struct nfs4_acl *acl, if (!(ace->flag & NFS4_ACE_INHERIT_ONLY_ACE)) process_one_v4_ace(&effective_acl_state, ace); } + + /* + * At this point, the default ACL may have zeroed-out entries for owner, + * group and other. That usually results in a non-sensical resulting ACL + * that denies all access except to any ACE that was explicitly added. + * + * The setfacl command solves a similar problem with this logic: + * + * "If a Default ACL entry is created, and the Default ACL contains + * no owner, owning group, or others entry, a copy of the ACL + * owner, owning group, or others entry is added to the Default ACL." + * + * Copy any missing ACEs from the effective set, if any ACEs were + * explicitly set. + */ + if (default_acl_state.valid) { + if (!(default_acl_state.valid & ACL_USER_OBJ)) + default_acl_state.owner = effective_acl_state.owner; + if (!(default_acl_state.valid & ACL_GROUP_OBJ)) + default_acl_state.group = effective_acl_state.group; + if (!(default_acl_state.valid & ACL_OTHER)) + default_acl_state.other = effective_acl_state.other; + } + *pacl = posix_state_to_acl(&effective_acl_state, flags); if (IS_ERR(*pacl)) { ret = PTR_ERR(*pacl); From 2eb2b93581813b74c7174961126f6ec38eadb5a7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 14:31:03 -0400 Subject: [PATCH 151/186] SUNRPC: Convert svc_tcp_sendmsg to use bio_vecs directly Add a helper to convert a whole xdr_buf directly into an array of bio_vecs, then send this array instead of iterating piecemeal over the xdr_buf containing the outbound RPC message. Reviewed-by: David Howells Signed-off-by: Chuck Lever --- include/linux/sunrpc/xdr.h | 2 ++ net/sunrpc/svcsock.c | 64 ++++++++++++-------------------------- net/sunrpc/xdr.c | 50 +++++++++++++++++++++++++++++ 3 files changed, 72 insertions(+), 44 deletions(-) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index f89ec4b5ea169..42f9d7eb9a1a3 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -139,6 +139,8 @@ void xdr_terminate_string(const struct xdr_buf *, const u32); size_t xdr_buf_pagecount(const struct xdr_buf *buf); int xdr_alloc_bvec(struct xdr_buf *buf, gfp_t gfp); void xdr_free_bvec(struct xdr_buf *buf); +unsigned int xdr_buf_to_bvec(struct bio_vec *bvec, unsigned int bvec_size, + const struct xdr_buf *xdr); static inline __be32 *xdr_encode_array(__be32 *p, const void *s, unsigned int len) { diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 589020ed909dc..90b1ab95c223e 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -36,6 +36,8 @@ #include #include #include +#include + #include #include #include @@ -1194,77 +1196,52 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) return 0; /* record not complete */ } -static int svc_tcp_send_kvec(struct socket *sock, const struct kvec *vec, - int flags) -{ - struct msghdr msg = { .msg_flags = MSG_SPLICE_PAGES | flags, }; - - iov_iter_kvec(&msg.msg_iter, ITER_SOURCE, vec, 1, vec->iov_len); - return sock_sendmsg(sock, &msg); -} - /* * MSG_SPLICE_PAGES is used exclusively to reduce the number of * copy operations in this path. Therefore the caller must ensure * that the pages backing @xdr are unchanging. * - * In addition, the logic assumes that * .bv_len is never larger - * than PAGE_SIZE. + * Note that the send is non-blocking. The caller has incremented + * the reference count on each page backing the RPC message, and + * the network layer will "put" these pages when transmission is + * complete. + * + * This is safe for our RPC services because the memory backing + * the head and tail components is never kmalloc'd. These always + * come from pages in the svc_rqst::rq_pages array. */ -static int svc_tcp_sendmsg(struct socket *sock, struct xdr_buf *xdr, +static int svc_tcp_sendmsg(struct svc_sock *svsk, struct svc_rqst *rqstp, rpc_fraghdr marker, unsigned int *sentp) { - const struct kvec *head = xdr->head; - const struct kvec *tail = xdr->tail; struct kvec rm = { .iov_base = &marker, .iov_len = sizeof(marker), }; struct msghdr msg = { - .msg_flags = 0, + .msg_flags = MSG_MORE, }; + unsigned int count; int ret; *sentp = 0; - ret = xdr_alloc_bvec(xdr, GFP_KERNEL); - if (ret < 0) - return ret; - ret = kernel_sendmsg(sock, &msg, &rm, 1, rm.iov_len); + ret = kernel_sendmsg(svsk->sk_sock, &msg, &rm, 1, rm.iov_len); if (ret < 0) return ret; *sentp += ret; if (ret != rm.iov_len) return -EAGAIN; - ret = svc_tcp_send_kvec(sock, head, 0); - if (ret < 0) - return ret; - *sentp += ret; - if (ret != head->iov_len) - goto out; - - if (xdr_buf_pagecount(xdr)) { - xdr->bvec[0].bv_offset = offset_in_page(xdr->page_base); - xdr->bvec[0].bv_len -= offset_in_page(xdr->page_base); - } + count = xdr_buf_to_bvec(rqstp->rq_bvec, ARRAY_SIZE(rqstp->rq_bvec), + &rqstp->rq_res); msg.msg_flags = MSG_SPLICE_PAGES; - iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, xdr->bvec, - xdr_buf_pagecount(xdr), xdr->page_len); - ret = sock_sendmsg(sock, &msg); + iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, rqstp->rq_bvec, + count, rqstp->rq_res.len); + ret = sock_sendmsg(svsk->sk_sock, &msg); if (ret < 0) return ret; *sentp += ret; - - if (tail->iov_len) { - ret = svc_tcp_send_kvec(sock, tail, 0); - if (ret < 0) - return ret; - *sentp += ret; - } - -out: return 0; } @@ -1295,8 +1272,7 @@ static int svc_tcp_sendto(struct svc_rqst *rqstp) if (svc_xprt_is_dead(xprt)) goto out_notconn; tcp_sock_set_cork(svsk->sk_sk, true); - err = svc_tcp_sendmsg(svsk->sk_sock, xdr, marker, &sent); - xdr_free_bvec(xdr); + err = svc_tcp_sendmsg(svsk, rqstp, marker, &sent); trace_svcsock_tcp_send(xprt, err < 0 ? (long)err : sent); if (err < 0 || sent != (xdr->len + sizeof(marker))) goto out_close; diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 2a22e78af116e..358e6de91775e 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -164,6 +164,56 @@ xdr_free_bvec(struct xdr_buf *buf) buf->bvec = NULL; } +/** + * xdr_buf_to_bvec - Copy components of an xdr_buf into a bio_vec array + * @bvec: bio_vec array to populate + * @bvec_size: element count of @bio_vec + * @xdr: xdr_buf to be copied + * + * Returns the number of entries consumed in @bvec. + */ +unsigned int xdr_buf_to_bvec(struct bio_vec *bvec, unsigned int bvec_size, + const struct xdr_buf *xdr) +{ + const struct kvec *head = xdr->head; + const struct kvec *tail = xdr->tail; + unsigned int count = 0; + + if (head->iov_len) { + bvec_set_virt(bvec++, head->iov_base, head->iov_len); + ++count; + } + + if (xdr->page_len) { + unsigned int offset, len, remaining; + struct page **pages = xdr->pages; + + offset = offset_in_page(xdr->page_base); + remaining = xdr->page_len; + while (remaining > 0) { + len = min_t(unsigned int, remaining, + PAGE_SIZE - offset); + bvec_set_page(bvec++, *pages++, len, offset); + remaining -= len; + offset = 0; + if (unlikely(++count > bvec_size)) + goto bvec_overflow; + } + } + + if (tail->iov_len) { + bvec_set_virt(bvec, tail->iov_base, tail->iov_len); + if (unlikely(++count > bvec_size)) + goto bvec_overflow; + } + + return count; + +bvec_overflow: + pr_warn_once("%s: bio_vec array overflow\n", __func__); + return count - 1; +} + /** * xdr_inline_pages - Prepare receive buffer for a large reply * @xdr: xdr_buf into which reply will be placed From e18e157bb5c8c1cd8a9ba25acfdcf4f3035836f4 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 14:31:09 -0400 Subject: [PATCH 152/186] SUNRPC: Send RPC message on TCP with a single sock_sendmsg() call There is now enough infrastructure in place to combine the stream record marker into the biovec array used to send each outgoing RPC message on TCP. The whole message can be more efficiently sent with a single call to sock_sendmsg() using a bio_vec iterator. Note that this also helps with RPC-with-TLS: the TLS implementation can now clearly see where the upper layer message boundaries are. Before, it would send each component of the xdr_buf (record marker, head, page payload, tail) in separate TLS records. Suggested-by: David Howells Reviewed-by: David Howells Signed-off-by: Chuck Lever --- include/linux/sunrpc/svcsock.h | 2 ++ net/sunrpc/svcsock.c | 33 ++++++++++++++++++--------------- 2 files changed, 20 insertions(+), 15 deletions(-) diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index a7116048a4d4b..caf3308f1f07e 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -38,6 +38,8 @@ struct svc_sock { /* Number of queued send requests */ atomic_t sk_sendqlen; + struct page_frag_cache sk_frag_cache; + struct completion sk_handshake_done; struct page * sk_pages[RPCSVC_MAXPAGES]; /* received data */ diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 90b1ab95c223e..d4d816036c040 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1213,31 +1213,30 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) static int svc_tcp_sendmsg(struct svc_sock *svsk, struct svc_rqst *rqstp, rpc_fraghdr marker, unsigned int *sentp) { - struct kvec rm = { - .iov_base = &marker, - .iov_len = sizeof(marker), - }; struct msghdr msg = { - .msg_flags = MSG_MORE, + .msg_flags = MSG_SPLICE_PAGES, }; unsigned int count; + void *buf; int ret; *sentp = 0; - ret = kernel_sendmsg(svsk->sk_sock, &msg, &rm, 1, rm.iov_len); - if (ret < 0) - return ret; - *sentp += ret; - if (ret != rm.iov_len) - return -EAGAIN; + /* The stream record marker is copied into a temporary page + * fragment buffer so that it can be included in rq_bvec. + */ + buf = page_frag_alloc(&svsk->sk_frag_cache, sizeof(marker), + GFP_KERNEL); + if (!buf) + return -ENOMEM; + memcpy(buf, &marker, sizeof(marker)); + bvec_set_virt(rqstp->rq_bvec, buf, sizeof(marker)); - count = xdr_buf_to_bvec(rqstp->rq_bvec, ARRAY_SIZE(rqstp->rq_bvec), - &rqstp->rq_res); + count = xdr_buf_to_bvec(rqstp->rq_bvec + 1, + ARRAY_SIZE(rqstp->rq_bvec) - 1, &rqstp->rq_res); - msg.msg_flags = MSG_SPLICE_PAGES; iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, rqstp->rq_bvec, - count, rqstp->rq_res.len); + 1 + count, sizeof(marker) + rqstp->rq_res.len); ret = sock_sendmsg(svsk->sk_sock, &msg); if (ret < 0) return ret; @@ -1616,6 +1615,7 @@ static void svc_tcp_sock_detach(struct svc_xprt *xprt) static void svc_sock_free(struct svc_xprt *xprt) { struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); + struct page_frag_cache *pfc = &svsk->sk_frag_cache; struct socket *sock = svsk->sk_sock; trace_svcsock_free(svsk, sock); @@ -1625,5 +1625,8 @@ static void svc_sock_free(struct svc_xprt *xprt) sockfd_put(sock); else sock_release(sock); + if (pfc->va) + __page_frag_cache_drain(virt_to_head_page(pfc->va), + pfc->pagecnt_bias); kfree(svsk); } From baabf59c24145612e4a975f459a5024389f13f5d Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 14:31:16 -0400 Subject: [PATCH 153/186] SUNRPC: Convert svc_udp_sendto() to use the per-socket bio_vec array Commit da1661b93bf4 ("SUNRPC: Teach server to use xprt_sock_sendmsg for socket sends") modified svc_udp_sendto() to use xprt_sock_sendmsg() because we originally believed xprt_sock_sendmsg() would be needed for TLS support. That does not actually appear to be the case. In addition, the linkage between the client and server send code has been a bit of a maintenance headache because of the distinct ways that the client and server handle memory allocation. Going forward, eventually the XDR layer will deal with its buffers in the form of bio_vec arrays, so convert this function accordingly. Signed-off-by: Chuck Lever --- net/sunrpc/svcsock.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index d4d816036c040..2ff730335a56c 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -693,9 +693,10 @@ static int svc_udp_sendto(struct svc_rqst *rqstp) .msg_name = &rqstp->rq_addr, .msg_namelen = rqstp->rq_addrlen, .msg_control = cmh, + .msg_flags = MSG_SPLICE_PAGES, .msg_controllen = sizeof(buffer), }; - unsigned int sent; + unsigned int count; int err; svc_udp_release_ctxt(xprt, rqstp->rq_xprt_ctxt); @@ -708,22 +709,23 @@ static int svc_udp_sendto(struct svc_rqst *rqstp) if (svc_xprt_is_dead(xprt)) goto out_notconn; - err = xdr_alloc_bvec(xdr, GFP_KERNEL); - if (err < 0) - goto out_unlock; + count = xdr_buf_to_bvec(rqstp->rq_bvec, + ARRAY_SIZE(rqstp->rq_bvec), xdr); - err = xprt_sock_sendmsg(svsk->sk_sock, &msg, xdr, 0, 0, &sent); + iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, rqstp->rq_bvec, + count, 0); + err = sock_sendmsg(svsk->sk_sock, &msg); if (err == -ECONNREFUSED) { /* ICMP error on earlier request. */ - err = xprt_sock_sendmsg(svsk->sk_sock, &msg, xdr, 0, 0, &sent); + iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, rqstp->rq_bvec, + count, 0); + err = sock_sendmsg(svsk->sk_sock, &msg); } - xdr_free_bvec(xdr); + trace_svcsock_udp_send(xprt, err); -out_unlock: + mutex_unlock(&xprt->xpt_mutex); - if (err < 0) - return err; - return sent; + return err; out_notconn: mutex_unlock(&xprt->xpt_mutex); From 89d2d9fbeadcbdbd6302d3d0cd6bfbe219d85b68 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 14:31:22 -0400 Subject: [PATCH 154/186] SUNRPC: Revert e0a912e8ddba Flamegraph analysis showed that the cork/uncork calls consume nearly a third of the CPU time spent in svc_tcp_sendto(). The other two consumers are mutex lock/unlock and svc_tcp_sendmsg(). Now that svc_tcp_sendto() coalesces RPC messages properly, there is no need to introduce artificial delays to prevent sending partial messages. After applying this change, I measured a 1.2K read IOPS increase for 8KB random I/O (several percent) on 56Gb IP over IB. Reviewed-by: David Howells Signed-off-by: Chuck Lever --- include/linux/sunrpc/svcsock.h | 2 -- net/sunrpc/svcsock.c | 6 ------ 2 files changed, 8 deletions(-) diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index caf3308f1f07e..a7ea54460b1a6 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -35,8 +35,6 @@ struct svc_sock { /* Total length of the data (not including fragment headers) * received so far in the fragments making up this rpc: */ u32 sk_datalen; - /* Number of queued send requests */ - atomic_t sk_sendqlen; struct page_frag_cache sk_frag_cache; diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 2ff730335a56c..8db8227e6c834 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1268,22 +1268,17 @@ static int svc_tcp_sendto(struct svc_rqst *rqstp) svc_tcp_release_ctxt(xprt, rqstp->rq_xprt_ctxt); rqstp->rq_xprt_ctxt = NULL; - atomic_inc(&svsk->sk_sendqlen); mutex_lock(&xprt->xpt_mutex); if (svc_xprt_is_dead(xprt)) goto out_notconn; - tcp_sock_set_cork(svsk->sk_sk, true); err = svc_tcp_sendmsg(svsk, rqstp, marker, &sent); trace_svcsock_tcp_send(xprt, err < 0 ? (long)err : sent); if (err < 0 || sent != (xdr->len + sizeof(marker))) goto out_close; - if (atomic_dec_and_test(&svsk->sk_sendqlen)) - tcp_sock_set_cork(svsk->sk_sk, false); mutex_unlock(&xprt->xpt_mutex); return sent; out_notconn: - atomic_dec(&svsk->sk_sendqlen); mutex_unlock(&xprt->xpt_mutex); return -ENOTCONN; out_close: @@ -1292,7 +1287,6 @@ static int svc_tcp_sendto(struct svc_rqst *rqstp) (err < 0) ? "got error" : "sent", (err < 0) ? err : sent, xdr->len); svc_xprt_deferred_close(xprt); - atomic_dec(&svsk->sk_sendqlen); mutex_unlock(&xprt->xpt_mutex); return -EAGAIN; } From 2b877fc53e975bdc5614c0a7d64047cfdbae3894 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 14:31:29 -0400 Subject: [PATCH 155/186] SUNRPC: Reduce thread wake-up rate when receiving large RPC messages With large NFS WRITE requests on TCP, I measured 5-10 thread wake- ups to receive each request. This is because the socket layer calls ->sk_data_ready() frequently, and each call triggers a thread wake-up. Each recvmsg() seems to pull in less than 100KB. Have the socket layer hold ->sk_data_ready() calls until the full incoming message has arrived to reduce the wake-up rate. Signed-off-by: Chuck Lever --- net/sunrpc/svcsock.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 8db8227e6c834..2864af3abdca4 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1089,6 +1089,9 @@ static void svc_tcp_fragment_received(struct svc_sock *svsk) /* If we have more data, signal svc_xprt_enqueue() to try again */ svsk->sk_tcplen = 0; svsk->sk_marker = xdr_zero; + + smp_wmb(); + tcp_set_rcvlowat(svsk->sk_sk, 1); } /** @@ -1178,10 +1181,17 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) goto err_delete; if (len == want) svc_tcp_fragment_received(svsk); - else + else { + /* Avoid more ->sk_data_ready() calls until the rest + * of the message has arrived. This reduces service + * thread wake-ups on large incoming messages. */ + tcp_set_rcvlowat(svsk->sk_sk, + svc_sock_reclen(svsk) - svsk->sk_tcplen); + trace_svcsock_tcp_recv_short(&svsk->sk_xprt, svc_sock_reclen(svsk), svsk->sk_tcplen - sizeof(rpc_fraghdr)); + } goto err_noclose; error: if (len != -EAGAIN) From 7afdc0c902c06554aef9fa58c2b6de7dc4c1fb45 Mon Sep 17 00:00:00 2001 From: Zhu Wang Date: Mon, 31 Jul 2023 19:23:00 +0800 Subject: [PATCH 156/186] exportfs: remove kernel-doc warnings in exportfs Remove kernel-doc warning in exportfs: fs/exportfs/expfs.c:395: warning: Function parameter or member 'parent' not described in 'exportfs_encode_inode_fh' Signed-off-by: Zhu Wang Signed-off-by: Chuck Lever --- fs/exportfs/expfs.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index d1dbe47c7975f..c20704aa21b3d 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -386,6 +386,7 @@ static int export_encode_fh(struct inode *inode, struct fid *fid, * @inode: the object to encode * @fid: where to store the file handle fragment * @max_len: maximum length to store there + * @parent: parent directory inode, if wanted * @flags: properties of the requested file handle * * Returns an enum fid_type or a negative errno. From de8d38cf44bac43e83bad28357ba84784c412752 Mon Sep 17 00:00:00 2001 From: Su Hui Date: Fri, 4 Aug 2023 09:26:57 +0800 Subject: [PATCH 157/186] fs: lockd: avoid possible wrong NULL parameter clang's static analysis warning: fs/lockd/mon.c: line 293, column 2: Null pointer passed as 2nd argument to memory copy function. Assuming 'hostname' is NULL and calling 'nsm_create_handle()', this will pass NULL as 2nd argument to memory copy function 'memcpy()'. So return NULL if 'hostname' is invalid. Fixes: 77a3ef33e2de ("NSM: More clean up of nsm_get_handle()") Signed-off-by: Su Hui Reviewed-by: Nick Desaulniers Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/lockd/mon.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 1d9488cf05348..87a0f207df0b9 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -276,6 +276,9 @@ static struct nsm_handle *nsm_create_handle(const struct sockaddr *sap, { struct nsm_handle *new; + if (!hostname) + return NULL; + new = kzalloc(sizeof(*new) + hostname_len + 1, GFP_KERNEL); if (unlikely(new == NULL)) return NULL; From 8db14cad28ae8ec3fde0fef18e969782bca204d1 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 18 Jul 2023 16:38:08 +1000 Subject: [PATCH 158/186] lockd: remove SIGKILL handling lockd allows SIGKILL and responds by dropping all locks and restarting the grace period. This functionality has been present since 2.1.32 when lockd was added to Linux. This functionality is undocumented and most likely added as a useful debug aid. When there is a need to drop locks, the better approach is to use /proc/fs/nfsd/unlock_*. This patch removes SIGKILL handling as part of preparation for removing all signal handling from sunrpc service threads. Signed-off-by: NeilBrown Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/lockd/svc.c | 24 ------------------------ 1 file changed, 24 deletions(-) diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 22d3ff3818f5f..614faa5f69cd8 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -45,7 +45,6 @@ #define NLMDBG_FACILITY NLMDBG_SVC #define LOCKD_BUFSIZE (1024 + NLMSVC_XDRSIZE) -#define ALLOWED_SIGS (sigmask(SIGKILL)) static struct svc_program nlmsvc_program; @@ -111,19 +110,6 @@ static void set_grace_period(struct net *net) schedule_delayed_work(&ln->grace_period_end, grace_period); } -static void restart_grace(void) -{ - if (nlmsvc_ops) { - struct net *net = &init_net; - struct lockd_net *ln = net_generic(net, lockd_net_id); - - cancel_delayed_work_sync(&ln->grace_period_end); - locks_end_grace(&ln->lockd_manager); - nlmsvc_invalidate_all(); - set_grace_period(net); - } -} - /* * This is the lockd kernel thread */ @@ -138,9 +124,6 @@ lockd(void *vrqstp) /* try_to_freeze() is called from svc_recv() */ set_freezable(); - /* Allow SIGKILL to tell lockd to drop all of its locks */ - allow_signal(SIGKILL); - dprintk("NFS locking service started (ver " LOCKD_VERSION ").\n"); /* @@ -154,12 +137,6 @@ lockd(void *vrqstp) /* update sv_maxconn if it has changed */ rqstp->rq_server->sv_maxconn = nlm_max_connections; - if (signalled()) { - flush_signals(current); - restart_grace(); - continue; - } - timeout = nlmsvc_retry_blocked(); /* @@ -174,7 +151,6 @@ lockd(void *vrqstp) svc_process(rqstp); } - flush_signals(current); if (nlmsvc_ops) nlmsvc_invalidate_all(); nlm_shutdown_hosts(); From 3903902401451b1cd9d797a8c79769eb26ac7fe5 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 18 Jul 2023 16:38:08 +1000 Subject: [PATCH 159/186] nfsd: don't allow nfsd threads to be signalled. The original implementation of nfsd used signals to stop threads during shutdown. In Linux 2.3.46pre5 nfsd gained the ability to shutdown threads internally it if was asked to run "0" threads. After this user-space transitioned to using "rpc.nfsd 0" to stop nfsd and sending signals to threads was no longer an important part of the API. In commit 3ebdbe5203a8 ("SUNRPC: discard svo_setup and rename svc_set_num_threads_sync()") (v5.17-rc1~75^2~41) we finally removed the use of signals for stopping threads, using kthread_stop() instead. This patch makes the "obvious" next step and removes the ability to signal nfsd threads - or any svc threads. nfsd stops allowing signals and we don't check for their delivery any more. This will allow for some simplification in later patches. A change worth noting is in nfsd4_ssc_setup_dul(). There was previously a signal_pending() check which would only succeed when the thread was being shut down. It should really have tested kthread_should_stop() as well. Now it just does the latter, not the former. Signed-off-by: NeilBrown Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfs/callback.c | 9 +-------- fs/nfsd/nfs4proc.c | 5 ++--- fs/nfsd/nfssvc.c | 12 ------------ net/sunrpc/svc_xprt.c | 16 ++++++---------- 4 files changed, 9 insertions(+), 33 deletions(-) diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 456af7d230cf1..46a0a2d6962e1 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -80,9 +80,6 @@ nfs4_callback_svc(void *vrqstp) set_freezable(); while (!kthread_freezable_should_stop(NULL)) { - - if (signal_pending(current)) - flush_signals(current); /* * Listen for a request on the socket */ @@ -112,11 +109,7 @@ nfs41_callback_svc(void *vrqstp) set_freezable(); while (!kthread_freezable_should_stop(NULL)) { - - if (signal_pending(current)) - flush_signals(current); - - prepare_to_wait(&serv->sv_cb_waitq, &wq, TASK_INTERRUPTIBLE); + prepare_to_wait(&serv->sv_cb_waitq, &wq, TASK_IDLE); spin_lock_bh(&serv->sv_cb_lock); if (!list_empty(&serv->sv_cb_list)) { req = list_first_entry(&serv->sv_cb_list, diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 7588fd1859a4f..5ca748309c262 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1347,12 +1347,11 @@ static __be32 nfsd4_ssc_setup_dul(struct nfsd_net *nn, char *ipaddr, /* found a match */ if (ni->nsui_busy) { /* wait - and try again */ - prepare_to_wait(&nn->nfsd_ssc_waitq, &wait, - TASK_INTERRUPTIBLE); + prepare_to_wait(&nn->nfsd_ssc_waitq, &wait, TASK_IDLE); spin_unlock(&nn->nfsd_ssc_lock); /* allow 20secs for mount/unmount for now - revisit */ - if (signal_pending(current) || + if (kthread_should_stop() || (schedule_timeout(20*HZ) == 0)) { finish_wait(&nn->nfsd_ssc_waitq, &wait); kfree(work); diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 97830e28c140c..439fca1959254 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -965,15 +965,6 @@ nfsd(void *vrqstp) current->fs->umask = 0; - /* - * thread is spawned with all signals set to SIG_IGN, re-enable - * the ones that will bring down the thread - */ - allow_signal(SIGKILL); - allow_signal(SIGHUP); - allow_signal(SIGINT); - allow_signal(SIGQUIT); - atomic_inc(&nfsdstats.th_cnt); set_freezable(); @@ -998,9 +989,6 @@ nfsd(void *vrqstp) validate_process_creds(); } - /* Clear signals before calling svc_exit_thread() */ - flush_signals(current); - atomic_dec(&nfsdstats.th_cnt); out: diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 62c7919ea6106..d4a7fd5aabd14 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -701,8 +701,8 @@ static int svc_alloc_arg(struct svc_rqst *rqstp) /* Made progress, don't sleep yet */ continue; - set_current_state(TASK_INTERRUPTIBLE); - if (signalled() || kthread_should_stop()) { + set_current_state(TASK_IDLE); + if (kthread_should_stop()) { set_current_state(TASK_RUNNING); return -EINTR; } @@ -740,7 +740,7 @@ rqst_should_sleep(struct svc_rqst *rqstp) return false; /* are we shutting down? */ - if (signalled() || kthread_should_stop()) + if (kthread_should_stop()) return false; /* are we freezing? */ @@ -762,11 +762,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) if (rqstp->rq_xprt) goto out_found; - /* - * We have to be able to interrupt this wait - * to bring down the daemons ... - */ - set_current_state(TASK_INTERRUPTIBLE); + set_current_state(TASK_IDLE); smp_mb__before_atomic(); clear_bit(SP_CONGESTED, &pool->sp_flags); clear_bit(RQ_BUSY, &rqstp->rq_flags); @@ -788,7 +784,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) if (!time_left) percpu_counter_inc(&pool->sp_threads_timedout); - if (signalled() || kthread_should_stop()) + if (kthread_should_stop()) return ERR_PTR(-EINTR); return ERR_PTR(-EAGAIN); out_found: @@ -885,7 +881,7 @@ int svc_recv(struct svc_rqst *rqstp, long timeout) try_to_freeze(); cond_resched(); err = -EINTR; - if (signalled() || kthread_should_stop()) + if (kthread_should_stop()) goto out; xprt = svc_get_next_xprt(rqstp, timeout); From 18e4cf915543257eae2925671934937163f5639b Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 31 Jul 2023 16:48:31 +1000 Subject: [PATCH 160/186] nfsd: Simplify code around svc_exit_thread() call in nfsd() Previously a thread could exit asynchronously (due to a signal) so some care was needed to hold nfsd_mutex over the last svc_put() call. Now a thread can only exit when svc_set_num_threads() is called, and this is always called under nfsd_mutex. So no care is needed. Not only is the mutex held when a thread exits now, but the svc refcount is elevated, so the svc_put() in svc_exit_thread() will never be a final put, so the mutex isn't even needed at this point in the code. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- fs/nfsd/nfssvc.c | 23 ----------------------- include/linux/sunrpc/svc.h | 13 ------------- 2 files changed, 36 deletions(-) diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 439fca1959254..56a776cd4e2d5 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -992,31 +992,8 @@ nfsd(void *vrqstp) atomic_dec(&nfsdstats.th_cnt); out: - /* Take an extra ref so that the svc_put in svc_exit_thread() - * doesn't call svc_destroy() - */ - svc_get(nn->nfsd_serv); - /* Release the thread */ svc_exit_thread(rqstp); - - /* We need to drop a ref, but may not drop the last reference - * without holding nfsd_mutex, and we cannot wait for nfsd_mutex as that - * could deadlock with nfsd_shutdown_threads() waiting for us. - * So three options are: - * - drop a non-final reference, - * - get the mutex without waiting - * - sleep briefly andd try the above again - */ - while (!svc_put_not_last(nn->nfsd_serv)) { - if (mutex_trylock(&nfsd_mutex)) { - nfsd_put(net); - mutex_unlock(&nfsd_mutex); - break; - } - msleep(20); - } - return 0; } diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index fe1394cc13716..2230148d9d687 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -120,19 +120,6 @@ static inline void svc_put(struct svc_serv *serv) kref_put(&serv->sv_refcnt, svc_destroy); } -/** - * svc_put_not_last - decrement non-final reference count on SUNRPC serv - * @serv: the svc_serv to have count decremented - * - * Returns: %true is refcount was decremented. - * - * If the refcount is 1, it is not decremented and instead failure is reported. - */ -static inline bool svc_put_not_last(struct svc_serv *serv) -{ - return refcount_dec_not_one(&serv->sv_refcnt.refcount); -} - /* * Maximum payload size supported by a kernel RPC server. * This is use to determine the max number of pages nfsd is From 9f28a971ee9fdf1bf8ce8c88b103f483be610277 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 31 Jul 2023 16:48:32 +1000 Subject: [PATCH 161/186] nfsd: separate nfsd_last_thread() from nfsd_put() Now that the last nfsd thread is stopped by an explicit act of calling svc_set_num_threads() with a count of zero, we only have a limited number of places that can happen, and don't need to call nfsd_last_thread() in nfsd_put() So separate that out and call it at the two places where the number of threads is set to zero. Move the clearing of ->nfsd_serv and the call to svc_xprt_destroy_all() into nfsd_last_thread(), as they are really part of the same action. nfsd_put() is now a thin wrapper around svc_put(), so make it a static inline. nfsd_put() cannot be called after nfsd_last_thread(), so in a couple of places we have to use svc_put() instead. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- fs/nfsd/nfsd.h | 7 ++++++- fs/nfsd/nfssvc.c | 52 ++++++++++++++++++------------------------------ 2 files changed, 25 insertions(+), 34 deletions(-) diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index d88498f8b275c..11c14faa6c67b 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -96,7 +96,12 @@ int nfsd_pool_stats_open(struct inode *, struct file *); int nfsd_pool_stats_release(struct inode *, struct file *); void nfsd_shutdown_threads(struct net *net); -void nfsd_put(struct net *net); +static inline void nfsd_put(struct net *net) +{ + struct nfsd_net *nn = net_generic(net, nfsd_net_id); + + svc_put(nn->nfsd_serv); +} bool i_am_nfsd(void); diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 56a776cd4e2d5..a0b16e3dd91ac 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -542,9 +542,14 @@ static struct notifier_block nfsd_inet6addr_notifier = { /* Only used under nfsd_mutex, so this atomic may be overkill: */ static atomic_t nfsd_notifier_refcount = ATOMIC_INIT(0); -static void nfsd_last_thread(struct svc_serv *serv, struct net *net) +static void nfsd_last_thread(struct net *net) { struct nfsd_net *nn = net_generic(net, nfsd_net_id); + struct svc_serv *serv = nn->nfsd_serv; + + spin_lock(&nfsd_notifier_lock); + nn->nfsd_serv = NULL; + spin_unlock(&nfsd_notifier_lock); /* check if the notifier still has clients */ if (atomic_dec_return(&nfsd_notifier_refcount) == 0) { @@ -554,6 +559,8 @@ static void nfsd_last_thread(struct svc_serv *serv, struct net *net) #endif } + svc_xprt_destroy_all(serv, net); + /* * write_ports can create the server without actually starting * any threads--if we get shut down before any threads are @@ -644,7 +651,8 @@ void nfsd_shutdown_threads(struct net *net) svc_get(serv); /* Kill outstanding nfsd threads */ svc_set_num_threads(serv, NULL, 0); - nfsd_put(net); + nfsd_last_thread(net); + svc_put(serv); mutex_unlock(&nfsd_mutex); } @@ -674,9 +682,6 @@ int nfsd_create_serv(struct net *net) serv->sv_maxconn = nn->max_connections; error = svc_bind(serv, net); if (error < 0) { - /* NOT nfsd_put() as notifiers (see below) haven't - * been set up yet. - */ svc_put(serv); return error; } @@ -719,29 +724,6 @@ int nfsd_get_nrthreads(int n, int *nthreads, struct net *net) return 0; } -/* This is the callback for kref_put() below. - * There is no code here as the first thing to be done is - * call svc_shutdown_net(), but we cannot get the 'net' from - * the kref. So do all the work when kref_put returns true. - */ -static void nfsd_noop(struct kref *ref) -{ -} - -void nfsd_put(struct net *net) -{ - struct nfsd_net *nn = net_generic(net, nfsd_net_id); - - if (kref_put(&nn->nfsd_serv->sv_refcnt, nfsd_noop)) { - svc_xprt_destroy_all(nn->nfsd_serv, net); - nfsd_last_thread(nn->nfsd_serv, net); - svc_destroy(&nn->nfsd_serv->sv_refcnt); - spin_lock(&nfsd_notifier_lock); - nn->nfsd_serv = NULL; - spin_unlock(&nfsd_notifier_lock); - } -} - int nfsd_set_nrthreads(int n, int *nthreads, struct net *net) { int i = 0; @@ -792,7 +774,7 @@ int nfsd_set_nrthreads(int n, int *nthreads, struct net *net) if (err) break; } - nfsd_put(net); + svc_put(nn->nfsd_serv); return err; } @@ -807,6 +789,7 @@ nfsd_svc(int nrservs, struct net *net, const struct cred *cred) int error; bool nfsd_up_before; struct nfsd_net *nn = net_generic(net, nfsd_net_id); + struct svc_serv *serv; mutex_lock(&nfsd_mutex); dprintk("nfsd: creating service\n"); @@ -826,22 +809,25 @@ nfsd_svc(int nrservs, struct net *net, const struct cred *cred) goto out; nfsd_up_before = nn->nfsd_net_up; + serv = nn->nfsd_serv; error = nfsd_startup_net(net, cred); if (error) goto out_put; - error = svc_set_num_threads(nn->nfsd_serv, NULL, nrservs); + error = svc_set_num_threads(serv, NULL, nrservs); if (error) goto out_shutdown; - error = nn->nfsd_serv->sv_nrthreads; + error = serv->sv_nrthreads; + if (error == 0) + nfsd_last_thread(net); out_shutdown: if (error < 0 && !nfsd_up_before) nfsd_shutdown_net(net); out_put: /* Threads now hold service active */ if (xchg(&nn->keep_active, 0)) - nfsd_put(net); - nfsd_put(net); + svc_put(serv); + svc_put(serv); out: mutex_unlock(&nfsd_mutex); return error; From f78116d3bf4fd7a84451e1a2adc35df7a63fbbf4 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 18 Jul 2023 16:38:08 +1000 Subject: [PATCH 162/186] SUNRPC: call svc_process() from svc_recv(). All callers of svc_recv() go on to call svc_process() on success. Simplify callers by having svc_recv() do that for them. This loses one call to validate_process_creds() in nfsd. That was debugging code added 14 years ago. I don't think we need to keep it. Signed-off-by: NeilBrown Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/lockd/svc.c | 5 ----- fs/nfs/callback.c | 1 - fs/nfsd/nfssvc.c | 3 +-- net/sunrpc/svc.c | 1 - net/sunrpc/svc_xprt.c | 3 ++- 5 files changed, 3 insertions(+), 10 deletions(-) diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 614faa5f69cd8..91ef139a7757c 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -132,7 +132,6 @@ lockd(void *vrqstp) */ while (!kthread_should_stop()) { long timeout = MAX_SCHEDULE_TIMEOUT; - RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); /* update sv_maxconn if it has changed */ rqstp->rq_server->sv_maxconn = nlm_max_connections; @@ -146,10 +145,6 @@ lockd(void *vrqstp) err = svc_recv(rqstp, timeout); if (err == -EAGAIN || err == -EINTR) continue; - dprintk("lockd: request from %s\n", - svc_print_addr(rqstp, buf, sizeof(buf))); - - svc_process(rqstp); } if (nlmsvc_ops) nlmsvc_invalidate_all(); diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 46a0a2d6962e1..2d94384bd6a99 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -86,7 +86,6 @@ nfs4_callback_svc(void *vrqstp) err = svc_recv(rqstp, MAX_SCHEDULE_TIMEOUT); if (err == -EAGAIN || err == -EINTR) continue; - svc_process(rqstp); } svc_exit_thread(rqstp); diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index a0b16e3dd91ac..547feb8ad0af7 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -970,8 +970,7 @@ nfsd(void *vrqstp) ; if (err == -EINTR) break; - validate_process_creds(); - svc_process(rqstp); + validate_process_creds(); } diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 587811a002c98..c69896c124a4f 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1516,7 +1516,6 @@ void svc_process(struct svc_rqst *rqstp) out_drop: svc_drop(rqstp); } -EXPORT_SYMBOL_GPL(svc_process); #if defined(CONFIG_SUNRPC_BACKCHANNEL) /* diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index d4a7fd5aabd14..8430b151bd716 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -906,7 +906,8 @@ int svc_recv(struct svc_rqst *rqstp, long timeout) if (serv->sv_stats) serv->sv_stats->netcnt++; rqstp->rq_stime = ktime_get(); - return len; + svc_process(rqstp); + return 0; out_release: rqstp->rq_res.len = 0; svc_xprt_release(rqstp); From 7b719e2bf342a59e88b2b6215b98ca4cf824bc58 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 18 Jul 2023 16:38:08 +1000 Subject: [PATCH 163/186] SUNRPC: change svc_recv() to return void. svc_recv() currently returns a 0 on success or one of two errors: - -EAGAIN means no message was successfully received - -EINTR means the thread has been told to stop Previously nfsd would stop as the result of a signal as well as following kthread_stop(). In that case the difference was useful: EINTR means stop unconditionally. EAGAIN means stop if kthread_should_stop(), continue otherwise. Now threads only exit when kthread_should_stop() so we don't need the distinction. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- fs/lockd/svc.c | 9 +-------- fs/nfs/callback.c | 11 ++--------- fs/nfsd/nfssvc.c | 13 ++----------- include/linux/sunrpc/svcsock.h | 2 +- net/sunrpc/svc_xprt.c | 28 +++++++++++----------------- 5 files changed, 17 insertions(+), 46 deletions(-) diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 91ef139a7757c..cf4ff7d3564c3 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -116,7 +116,6 @@ static void set_grace_period(struct net *net) static int lockd(void *vrqstp) { - int err = 0; struct svc_rqst *rqstp = vrqstp; struct net *net = &init_net; struct lockd_net *ln = net_generic(net, lockd_net_id); @@ -138,13 +137,7 @@ lockd(void *vrqstp) timeout = nlmsvc_retry_blocked(); - /* - * Find a socket with data available and call its - * recvfrom routine. - */ - err = svc_recv(rqstp, timeout); - if (err == -EAGAIN || err == -EINTR) - continue; + svc_recv(rqstp, timeout); } if (nlmsvc_ops) nlmsvc_invalidate_all(); diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 2d94384bd6a99..54155b484f7b5 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -74,19 +74,12 @@ static int nfs4_callback_up_net(struct svc_serv *serv, struct net *net) static int nfs4_callback_svc(void *vrqstp) { - int err; struct svc_rqst *rqstp = vrqstp; set_freezable(); - while (!kthread_freezable_should_stop(NULL)) { - /* - * Listen for a request on the socket - */ - err = svc_recv(rqstp, MAX_SCHEDULE_TIMEOUT); - if (err == -EAGAIN || err == -EINTR) - continue; - } + while (!kthread_freezable_should_stop(NULL)) + svc_recv(rqstp, MAX_SCHEDULE_TIMEOUT); svc_exit_thread(rqstp); return 0; diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 547feb8ad0af7..4c0ab101d90b9 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -939,7 +939,6 @@ nfsd(void *vrqstp) struct svc_xprt *perm_sock = list_entry(rqstp->rq_server->sv_permsocks.next, typeof(struct svc_xprt), xpt_list); struct net *net = perm_sock->xpt_net; struct nfsd_net *nn = net_generic(net, nfsd_net_id); - int err; /* At this point, the thread shares current->fs * with the init process. We need to create files with the @@ -958,19 +957,11 @@ nfsd(void *vrqstp) /* * The main request loop */ - for (;;) { + while (!kthread_should_stop()) { /* Update sv_maxconn if it has changed */ rqstp->rq_server->sv_maxconn = nn->max_connections; - /* - * Find a socket with data available and call its - * recvfrom routine. - */ - while ((err = svc_recv(rqstp, 60*60*HZ)) == -EAGAIN) - ; - if (err == -EINTR) - break; - + svc_recv(rqstp, 60*60*HZ); validate_process_creds(); } diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index a7ea54460b1a6..c0ddd331d82fd 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -57,7 +57,7 @@ static inline u32 svc_sock_final_rec(struct svc_sock *svsk) * Function prototypes. */ void svc_close_net(struct svc_serv *, struct net *); -int svc_recv(struct svc_rqst *, long); +void svc_recv(struct svc_rqst *, long); void svc_send(struct svc_rqst *rqstp); void svc_drop(struct svc_rqst *); void svc_sock_update_bufs(struct svc_serv *serv); diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 8430b151bd716..d7d69143011c5 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -679,7 +679,7 @@ static void svc_check_conn_limits(struct svc_serv *serv) } } -static int svc_alloc_arg(struct svc_rqst *rqstp) +static bool svc_alloc_arg(struct svc_rqst *rqstp) { struct svc_serv *serv = rqstp->rq_server; struct xdr_buf *arg = &rqstp->rq_arg; @@ -704,7 +704,7 @@ static int svc_alloc_arg(struct svc_rqst *rqstp) set_current_state(TASK_IDLE); if (kthread_should_stop()) { set_current_state(TASK_RUNNING); - return -EINTR; + return false; } trace_svc_alloc_arg_err(pages, ret); memalloc_retry_wait(GFP_KERNEL); @@ -723,7 +723,7 @@ static int svc_alloc_arg(struct svc_rqst *rqstp) arg->tail[0].iov_len = 0; rqstp->rq_xid = xdr_zero; - return 0; + return true; } static bool @@ -785,8 +785,8 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) percpu_counter_inc(&pool->sp_threads_timedout); if (kthread_should_stop()) - return ERR_PTR(-EINTR); - return ERR_PTR(-EAGAIN); + return NULL; + return NULL; out_found: /* Normally we will wait up to 5 seconds for any required * cache information to be provided. @@ -868,32 +868,27 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) * organised not to touch any cachelines in the shared svc_serv * structure, only cachelines in the local svc_pool. */ -int svc_recv(struct svc_rqst *rqstp, long timeout) +void svc_recv(struct svc_rqst *rqstp, long timeout) { struct svc_xprt *xprt = NULL; struct svc_serv *serv = rqstp->rq_server; - int len, err; + int len; - err = svc_alloc_arg(rqstp); - if (err) + if (!svc_alloc_arg(rqstp)) goto out; try_to_freeze(); cond_resched(); - err = -EINTR; if (kthread_should_stop()) goto out; xprt = svc_get_next_xprt(rqstp, timeout); - if (IS_ERR(xprt)) { - err = PTR_ERR(xprt); + if (!xprt) goto out; - } len = svc_handle_xprt(rqstp, xprt); /* No data, incomplete (TCP) read, or accept() */ - err = -EAGAIN; if (len <= 0) goto out_release; @@ -907,12 +902,11 @@ int svc_recv(struct svc_rqst *rqstp, long timeout) serv->sv_stats->netcnt++; rqstp->rq_stime = ktime_get(); svc_process(rqstp); - return 0; +out: + return; out_release: rqstp->rq_res.len = 0; svc_xprt_release(rqstp); -out: - return err; } EXPORT_SYMBOL_GPL(svc_recv); From c743b4259c3af2c0637c307f08a062d25fa3c99f Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 18 Jul 2023 16:38:08 +1000 Subject: [PATCH 164/186] SUNRPC: remove timeout arg from svc_recv() Most svc threads have no interest in a timeout. nfsd sets it to 1 hour, but this is a wart of no significance. lockd uses the timeout so that it can call nlmsvc_retry_blocked(). It also sometimes calls svc_wake_up() to ensure this is called. So change lockd to be consistent and always use svc_wake_up() to trigger nlmsvc_retry_blocked() - using a timer instead of a timeout to svc_recv(). And change svc_recv() to not take a timeout arg. This makes the sp_threads_timedout counter always zero. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- fs/lockd/svc.c | 14 +++++++++----- fs/lockd/svclock.c | 5 +++-- fs/nfs/callback.c | 2 +- fs/nfsd/nfssvc.c | 2 +- include/linux/lockd/lockd.h | 4 +++- include/linux/sunrpc/svc.h | 1 - include/linux/sunrpc/svcsock.h | 2 +- net/sunrpc/svc.c | 2 -- net/sunrpc/svc_xprt.c | 34 ++++++++++++++++------------------ 9 files changed, 34 insertions(+), 32 deletions(-) diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index cf4ff7d3564c3..ef3f77a59556c 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -56,6 +56,12 @@ static unsigned int nlmsvc_users; static struct svc_serv *nlmsvc_serv; unsigned long nlmsvc_timeout; +static void nlmsvc_request_retry(struct timer_list *tl) +{ + svc_wake_up(nlmsvc_serv); +} +DEFINE_TIMER(nlmsvc_retry, nlmsvc_request_retry); + unsigned int lockd_net_id; /* @@ -130,14 +136,11 @@ lockd(void *vrqstp) * NFS mount or NFS daemon has gone away. */ while (!kthread_should_stop()) { - long timeout = MAX_SCHEDULE_TIMEOUT; - /* update sv_maxconn if it has changed */ rqstp->rq_server->sv_maxconn = nlm_max_connections; - timeout = nlmsvc_retry_blocked(); - - svc_recv(rqstp, timeout); + nlmsvc_retry_blocked(); + svc_recv(rqstp); } if (nlmsvc_ops) nlmsvc_invalidate_all(); @@ -371,6 +374,7 @@ static void lockd_put(void) #endif svc_set_num_threads(nlmsvc_serv, NULL, 0); + timer_delete_sync(&nlmsvc_retry); nlmsvc_serv = NULL; dprintk("lockd_down: service destroyed\n"); } diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 28abec5c451d1..43aeba9de55cb 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -1019,7 +1019,7 @@ retry_deferred_block(struct nlm_block *block) * picks up locks that can be granted, or grant notifications that must * be retransmitted. */ -unsigned long +void nlmsvc_retry_blocked(void) { unsigned long timeout = MAX_SCHEDULE_TIMEOUT; @@ -1049,5 +1049,6 @@ nlmsvc_retry_blocked(void) } spin_unlock(&nlm_blocked_lock); - return timeout; + if (timeout < MAX_SCHEDULE_TIMEOUT) + mod_timer(&nlmsvc_retry, jiffies + timeout); } diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 54155b484f7b5..39a0ba746267a 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -79,7 +79,7 @@ nfs4_callback_svc(void *vrqstp) set_freezable(); while (!kthread_freezable_should_stop(NULL)) - svc_recv(rqstp, MAX_SCHEDULE_TIMEOUT); + svc_recv(rqstp); svc_exit_thread(rqstp); return 0; diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 4c0ab101d90b9..1582af33e204a 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -961,7 +961,7 @@ nfsd(void *vrqstp) /* Update sv_maxconn if it has changed */ rqstp->rq_server->sv_maxconn = nn->max_connections; - svc_recv(rqstp, 60*60*HZ); + svc_recv(rqstp); validate_process_creds(); } diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index f42594a9efe0d..0f016d69c996e 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -204,6 +204,8 @@ extern unsigned long nlmsvc_timeout; extern bool nsm_use_hostnames; extern u32 nsm_local_state; +extern struct timer_list nlmsvc_retry; + /* * Lockd client functions */ @@ -280,7 +282,7 @@ __be32 nlmsvc_testlock(struct svc_rqst *, struct nlm_file *, struct nlm_host *, struct nlm_lock *, struct nlm_lock *, struct nlm_cookie *); __be32 nlmsvc_cancel_blocked(struct net *net, struct nlm_file *, struct nlm_lock *); -unsigned long nlmsvc_retry_blocked(void); +void nlmsvc_retry_blocked(void); void nlmsvc_traverse_blocks(struct nlm_host *, struct nlm_file *, nlm_host_match_fn_t match); void nlmsvc_grant_reply(struct nlm_cookie *, __be32); diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 2230148d9d687..b206fdde8e97a 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -41,7 +41,6 @@ struct svc_pool { /* statistics on pool operation */ struct percpu_counter sp_sockets_queued; struct percpu_counter sp_threads_woken; - struct percpu_counter sp_threads_timedout; #define SP_TASK_PENDING (0) /* still work to do even if no * xprt is queued. */ diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index c0ddd331d82fd..d4a173c5b3be1 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -57,7 +57,7 @@ static inline u32 svc_sock_final_rec(struct svc_sock *svsk) * Function prototypes. */ void svc_close_net(struct svc_serv *, struct net *); -void svc_recv(struct svc_rqst *, long); +void svc_recv(struct svc_rqst *rqstp); void svc_send(struct svc_rqst *rqstp); void svc_drop(struct svc_rqst *); void svc_sock_update_bufs(struct svc_serv *serv); diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index c69896c124a4f..030f8c759ee60 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -515,7 +515,6 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools, percpu_counter_init(&pool->sp_sockets_queued, 0, GFP_KERNEL); percpu_counter_init(&pool->sp_threads_woken, 0, GFP_KERNEL); - percpu_counter_init(&pool->sp_threads_timedout, 0, GFP_KERNEL); } return serv; @@ -590,7 +589,6 @@ svc_destroy(struct kref *ref) percpu_counter_destroy(&pool->sp_sockets_queued); percpu_counter_destroy(&pool->sp_threads_woken); - percpu_counter_destroy(&pool->sp_threads_timedout); } kfree(serv->sv_pools); kfree(serv); diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index d7d69143011c5..9bdcdd8401b86 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -750,10 +750,9 @@ rqst_should_sleep(struct svc_rqst *rqstp) return true; } -static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) +static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp) { struct svc_pool *pool = rqstp->rq_pool; - long time_left = 0; /* rq_xprt should be clear on entry */ WARN_ON_ONCE(rqstp->rq_xprt); @@ -769,7 +768,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) smp_mb__after_atomic(); if (likely(rqst_should_sleep(rqstp))) - time_left = schedule_timeout(timeout); + schedule(); else __set_current_state(TASK_RUNNING); @@ -781,9 +780,6 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp, long timeout) if (rqstp->rq_xprt) goto out_found; - if (!time_left) - percpu_counter_inc(&pool->sp_threads_timedout); - if (kthread_should_stop()) return NULL; return NULL; @@ -863,12 +859,15 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) return len; } -/* - * Receive the next request on any transport. This code is carefully - * organised not to touch any cachelines in the shared svc_serv - * structure, only cachelines in the local svc_pool. +/** + * svc_recv - Receive and process the next request on any transport + * @rqstp: an idle RPC service thread + * + * This code is carefully organised not to touch any cachelines in + * the shared svc_serv structure, only cachelines in the local + * svc_pool. */ -void svc_recv(struct svc_rqst *rqstp, long timeout) +void svc_recv(struct svc_rqst *rqstp) { struct svc_xprt *xprt = NULL; struct svc_serv *serv = rqstp->rq_server; @@ -882,7 +881,7 @@ void svc_recv(struct svc_rqst *rqstp, long timeout) if (kthread_should_stop()) goto out; - xprt = svc_get_next_xprt(rqstp, timeout); + xprt = svc_get_next_xprt(rqstp); if (!xprt) goto out; @@ -1447,12 +1446,11 @@ static int svc_pool_stats_show(struct seq_file *m, void *p) return 0; } - seq_printf(m, "%u %llu %llu %llu %llu\n", - pool->sp_id, - percpu_counter_sum_positive(&pool->sp_sockets_queued), - percpu_counter_sum_positive(&pool->sp_sockets_queued), - percpu_counter_sum_positive(&pool->sp_threads_woken), - percpu_counter_sum_positive(&pool->sp_threads_timedout)); + seq_printf(m, "%u %llu %llu %llu 0\n", + pool->sp_id, + percpu_counter_sum_positive(&pool->sp_sockets_queued), + percpu_counter_sum_positive(&pool->sp_sockets_queued), + percpu_counter_sum_positive(&pool->sp_threads_woken)); return 0; } From ba4bba6c97d40fa9f2aa25a34f6e9717a468c8f3 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Sat, 29 Jul 2023 14:31:55 -0400 Subject: [PATCH 165/186] SUNRPC: change cache_head.flags bits to enum When a sequence of numbers are needed for internal-use only, an enum is typically best. The sequence will inevitably need to be changed one day, and having an enum means the developer doesn't need to think about renumbering after insertion or deletion. Such patches will be easier to review. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/cache.h | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h index 518bd28f5ab8c..35766963dd145 100644 --- a/include/linux/sunrpc/cache.h +++ b/include/linux/sunrpc/cache.h @@ -56,10 +56,14 @@ struct cache_head { struct kref ref; unsigned long flags; }; -#define CACHE_VALID 0 /* Entry contains valid data */ -#define CACHE_NEGATIVE 1 /* Negative entry - there is no match for the key */ -#define CACHE_PENDING 2 /* An upcall has been sent but no reply received yet*/ -#define CACHE_CLEANED 3 /* Entry has been cleaned from cache */ + +/* cache_head.flags */ +enum { + CACHE_VALID, /* Entry contains valid data */ + CACHE_NEGATIVE, /* Negative entry - there is no match for the key */ + CACHE_PENDING, /* An upcall has been sent but no reply received yet*/ + CACHE_CLEANED, /* Entry has been cleaned from cache */ +}; #define CACHE_NEW_EXPIRY 120 /* keep new things pending confirmation for 120 seconds */ From 3275694adf0f89e1cdcacfee16103be6643c2a0c Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Sat, 29 Jul 2023 14:33:05 -0400 Subject: [PATCH 166/186] SUNRPC: change svc_pool::sp_flags bits to enum When a sequence of numbers are needed for internal-use only, an enum is typically best. The sequence will inevitably need to be changed one day, and having an enum means the developer doesn't need to think about renumbering after insertion or deletion. Such patches will be easier to review. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index b206fdde8e97a..dc526240de5ea 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -42,12 +42,16 @@ struct svc_pool { struct percpu_counter sp_sockets_queued; struct percpu_counter sp_threads_woken; -#define SP_TASK_PENDING (0) /* still work to do even if no - * xprt is queued. */ -#define SP_CONGESTED (1) unsigned long sp_flags; } ____cacheline_aligned_in_smp; +/* bits for sp_flags */ +enum { + SP_TASK_PENDING, /* still work to do even if no xprt is queued */ + SP_CONGESTED, /* all threads are busy, none idle */ +}; + + /* * RPC service. * From a6b4ec39036fd78b95205eef3be121454b7d2973 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Sat, 29 Jul 2023 14:34:12 -0400 Subject: [PATCH 167/186] SUNRPC: change svc_rqst::rq_flags bits to enum When a sequence of numbers are needed for internal-use only, an enum is typically best. The sequence will inevitably need to be changed one day, and having an enum means the developer doesn't need to think about renumbering after insertion or deletion. Such patches will be easier to review. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index dc526240de5ea..9b429253dbd9e 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -222,16 +222,6 @@ struct svc_rqst { u32 rq_proc; /* procedure number */ u32 rq_prot; /* IP protocol */ int rq_cachetype; /* catering to nfsd */ -#define RQ_SECURE (0) /* secure port */ -#define RQ_LOCAL (1) /* local request */ -#define RQ_USEDEFERRAL (2) /* use deferral */ -#define RQ_DROPME (3) /* drop current reply */ -#define RQ_SPLICE_OK (4) /* turned off in gss privacy - * to prevent encrypting page - * cache pages */ -#define RQ_VICTIM (5) /* about to be shut down */ -#define RQ_BUSY (6) /* request is busy */ -#define RQ_DATA (7) /* request has data */ unsigned long rq_flags; /* flags field */ ktime_t rq_qtime; /* enqueue time */ @@ -262,6 +252,19 @@ struct svc_rqst { void ** rq_lease_breaker; /* The v4 client breaking a lease */ }; +/* bits for rq_flags */ +enum { + RQ_SECURE, /* secure port */ + RQ_LOCAL, /* local request */ + RQ_USEDEFERRAL, /* use deferral */ + RQ_DROPME, /* drop current reply */ + RQ_SPLICE_OK, /* turned off in gss privacy to prevent + * encrypting page cache pages */ + RQ_VICTIM, /* about to be shut down */ + RQ_BUSY, /* request is busy */ + RQ_DATA, /* request has data */ +}; + #define SVC_NET(rqst) (rqst->rq_xprt ? rqst->rq_xprt->xpt_net : rqst->rq_bc_net) /* From d75e490f35601aae12c7284d3c22684c65fb8354 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sat, 29 Jul 2023 14:36:10 -0400 Subject: [PATCH 168/186] SUNRPC: change svc_xprt::xpt_flags bits to enum When a sequence of numbers are needed for internal-use only, an enum is typically best. The sequence will inevitably need to be changed one day, and having an enum means the developer doesn't need to think about renumbering after insertion or deletion. Such patches will be easier to review. Suggested-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc_xprt.h | 38 ++++++++++++----------- include/trace/events/sunrpc.h | 53 ++++++++++++++++++++++----------- 2 files changed, 56 insertions(+), 35 deletions(-) diff --git a/include/linux/sunrpc/svc_xprt.h b/include/linux/sunrpc/svc_xprt.h index a6b12631db21c..fa55d12dc7651 100644 --- a/include/linux/sunrpc/svc_xprt.h +++ b/include/linux/sunrpc/svc_xprt.h @@ -56,23 +56,6 @@ struct svc_xprt { struct list_head xpt_list; struct list_head xpt_ready; unsigned long xpt_flags; -#define XPT_BUSY 0 /* enqueued/receiving */ -#define XPT_CONN 1 /* conn pending */ -#define XPT_CLOSE 2 /* dead or dying */ -#define XPT_DATA 3 /* data pending */ -#define XPT_TEMP 4 /* connected transport */ -#define XPT_DEAD 6 /* transport closed */ -#define XPT_CHNGBUF 7 /* need to change snd/rcv buf sizes */ -#define XPT_DEFERRED 8 /* deferred request pending */ -#define XPT_OLD 9 /* used for xprt aging mark+sweep */ -#define XPT_LISTENER 10 /* listening endpoint */ -#define XPT_CACHE_AUTH 11 /* cache auth info */ -#define XPT_LOCAL 12 /* connection from loopback interface */ -#define XPT_KILL_TEMP 13 /* call xpo_kill_temp_xprt before closing */ -#define XPT_CONG_CTRL 14 /* has congestion control */ -#define XPT_HANDSHAKE 15 /* xprt requests a handshake */ -#define XPT_TLS_SESSION 16 /* transport-layer security established */ -#define XPT_PEER_AUTH 17 /* peer has been authenticated */ struct svc_serv *xpt_server; /* service for transport */ atomic_t xpt_reserved; /* space on outq that is rsvd */ @@ -97,6 +80,27 @@ struct svc_xprt { struct rpc_xprt_switch *xpt_bc_xps; /* NFSv4.1 backchannel */ }; +/* flag bits for xpt_flags */ +enum { + XPT_BUSY, /* enqueued/receiving */ + XPT_CONN, /* conn pending */ + XPT_CLOSE, /* dead or dying */ + XPT_DATA, /* data pending */ + XPT_TEMP, /* connected transport */ + XPT_DEAD, /* transport closed */ + XPT_CHNGBUF, /* need to change snd/rcv buf sizes */ + XPT_DEFERRED, /* deferred request pending */ + XPT_OLD, /* used for xprt aging mark+sweep */ + XPT_LISTENER, /* listening endpoint */ + XPT_CACHE_AUTH, /* cache auth info */ + XPT_LOCAL, /* connection from loopback interface */ + XPT_KILL_TEMP, /* call xpo_kill_temp_xprt before closing */ + XPT_CONG_CTRL, /* has congestion control */ + XPT_HANDSHAKE, /* xprt requests a handshake */ + XPT_TLS_SESSION, /* transport-layer security established */ + XPT_PEER_AUTH, /* peer has been authenticated */ +}; + static inline void unregister_xpt_user(struct svc_xprt *xpt, struct svc_xpt_user *u) { spin_lock(&xpt->xpt_lock); diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 43711753616a1..00db9e1fb7118 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -1918,25 +1918,42 @@ TRACE_EVENT(svc_stats_latency, __get_str(procedure), __entry->execute) ); +/* + * from include/linux/sunrpc/svc_xprt.h + */ +#define SVC_XPRT_FLAG_LIST \ + svc_xprt_flag(BUSY) \ + svc_xprt_flag(CONN) \ + svc_xprt_flag(CLOSE) \ + svc_xprt_flag(DATA) \ + svc_xprt_flag(TEMP) \ + svc_xprt_flag(DEAD) \ + svc_xprt_flag(CHNGBUF) \ + svc_xprt_flag(DEFERRED) \ + svc_xprt_flag(OLD) \ + svc_xprt_flag(LISTENER) \ + svc_xprt_flag(CACHE_AUTH) \ + svc_xprt_flag(LOCAL) \ + svc_xprt_flag(KILL_TEMP) \ + svc_xprt_flag(CONG_CTRL) \ + svc_xprt_flag(HANDSHAKE) \ + svc_xprt_flag(TLS_SESSION) \ + svc_xprt_flag_end(PEER_AUTH) + +#undef svc_xprt_flag +#undef svc_xprt_flag_end +#define svc_xprt_flag(x) TRACE_DEFINE_ENUM(XPT_##x); +#define svc_xprt_flag_end(x) TRACE_DEFINE_ENUM(XPT_##x); + +SVC_XPRT_FLAG_LIST + +#undef svc_xprt_flag +#undef svc_xprt_flag_end +#define svc_xprt_flag(x) { BIT(XPT_##x), #x }, +#define svc_xprt_flag_end(x) { BIT(XPT_##x), #x } + #define show_svc_xprt_flags(flags) \ - __print_flags(flags, "|", \ - { BIT(XPT_BUSY), "BUSY" }, \ - { BIT(XPT_CONN), "CONN" }, \ - { BIT(XPT_CLOSE), "CLOSE" }, \ - { BIT(XPT_DATA), "DATA" }, \ - { BIT(XPT_TEMP), "TEMP" }, \ - { BIT(XPT_DEAD), "DEAD" }, \ - { BIT(XPT_CHNGBUF), "CHNGBUF" }, \ - { BIT(XPT_DEFERRED), "DEFERRED" }, \ - { BIT(XPT_OLD), "OLD" }, \ - { BIT(XPT_LISTENER), "LISTENER" }, \ - { BIT(XPT_CACHE_AUTH), "CACHE_AUTH" }, \ - { BIT(XPT_LOCAL), "LOCAL" }, \ - { BIT(XPT_KILL_TEMP), "KILL_TEMP" }, \ - { BIT(XPT_CONG_CTRL), "CONG_CTRL" }, \ - { BIT(XPT_HANDSHAKE), "HANDSHAKE" }, \ - { BIT(XPT_TLS_SESSION), "TLS_SESSION" }, \ - { BIT(XPT_PEER_AUTH), "PEER_AUTH" }) + __print_flags(flags, "|", SVC_XPRT_FLAG_LIST) TRACE_EVENT(svc_xprt_create_err, TP_PROTO( From 78c542f916bccafffef4f3bec9bc60d7cda548f5 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sat, 29 Jul 2023 20:58:54 -0400 Subject: [PATCH 169/186] SUNRPC: Add enum svc_auth_status In addition to the benefits of using an enum rather than a set of macros, we now have a named type that can improve static type checking of function return values. As part of this change, I removed a stale comment from svcauth.h; the return values from current implementations of the auth_ops::release method are all zero/negative errno, not the SVC_OK enum values as the old comment suggested. Suggested-by: NeilBrown Signed-off-by: Chuck Lever --- fs/lockd/svc.c | 2 +- fs/nfs/callback.c | 2 +- include/linux/sunrpc/svc.h | 2 +- include/linux/sunrpc/svcauth.h | 50 +++++++++++++++---------------- include/trace/events/sunrpc.h | 9 ++++-- net/sunrpc/auth_gss/svcauth_gss.c | 7 ++--- net/sunrpc/svc.c | 6 +++- net/sunrpc/svcauth.c | 35 ++++++++++++++++++---- net/sunrpc/svcauth_unix.c | 9 +++--- 9 files changed, 73 insertions(+), 49 deletions(-) diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index ef3f77a59556c..6579948070a48 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -506,7 +506,7 @@ static inline int is_callback(u32 proc) } -static int lockd_authenticate(struct svc_rqst *rqstp) +static enum svc_auth_status lockd_authenticate(struct svc_rqst *rqstp) { rqstp->rq_client = NULL; switch (rqstp->rq_authop->flavour) { diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 39a0ba746267a..466ebf1d41b2b 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -372,7 +372,7 @@ check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp) * All other checking done after NFS decoding where the nfs_client can be * found in nfs4_callback_compound */ -static int nfs_callback_authenticate(struct svc_rqst *rqstp) +static enum svc_auth_status nfs_callback_authenticate(struct svc_rqst *rqstp) { rqstp->rq_auth_stat = rpc_autherr_badcred; diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 9b429253dbd9e..1c491f02efc87 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -336,7 +336,7 @@ struct svc_program { char * pg_name; /* service name */ char * pg_class; /* class name: services sharing authentication */ struct svc_stat * pg_stats; /* rpc statistics */ - int (*pg_authenticate)(struct svc_rqst *); + enum svc_auth_status (*pg_authenticate)(struct svc_rqst *rqstp); __be32 (*pg_init_request)(struct svc_rqst *, const struct svc_program *, struct svc_process_info *); diff --git a/include/linux/sunrpc/svcauth.h b/include/linux/sunrpc/svcauth.h index 27582d3b538f2..6f90203edbf8d 100644 --- a/include/linux/sunrpc/svcauth.h +++ b/include/linux/sunrpc/svcauth.h @@ -83,6 +83,19 @@ struct auth_domain { struct rcu_head rcu_head; }; +enum svc_auth_status { + SVC_GARBAGE = 1, + SVC_SYSERR, + SVC_VALID, + SVC_NEGATIVE, + SVC_OK, + SVC_DROP, + SVC_CLOSE, + SVC_DENIED, + SVC_PENDING, + SVC_COMPLETE, +}; + /* * Each authentication flavour registers an auth_ops * structure. @@ -98,6 +111,8 @@ struct auth_domain { * is (probably) already in place. Certainly space is * reserved for it. * DROP - simply drop the request. It may have been deferred + * CLOSE - like SVC_DROP, but request is definitely lost. + * If there is a tcp connection, it should be closed. * GARBAGE - rpc garbage_args error * SYSERR - rpc system_err error * DENIED - authp holds reason for denial. @@ -111,14 +126,10 @@ struct auth_domain { * * release() is given a request after the procedure has been run. * It should sign/encrypt the results if needed - * It should return: - * OK - the resbuf is ready to be sent - * DROP - the reply should be quitely dropped - * DENIED - authp holds a reason for MSG_DENIED - * SYSERR - rpc system_err * * domain_release() * This call releases a domain. + * * set_client() * Givens a pending request (struct svc_rqst), finds and assigns * an appropriate 'auth_domain' as the client. @@ -127,31 +138,18 @@ struct auth_ops { char * name; struct module *owner; int flavour; - int (*accept)(struct svc_rqst *rq); - int (*release)(struct svc_rqst *rq); - void (*domain_release)(struct auth_domain *); - int (*set_client)(struct svc_rqst *rq); -}; -#define SVC_GARBAGE 1 -#define SVC_SYSERR 2 -#define SVC_VALID 3 -#define SVC_NEGATIVE 4 -#define SVC_OK 5 -#define SVC_DROP 6 -#define SVC_CLOSE 7 /* Like SVC_DROP, but request is definitely - * lost so if there is a tcp connection, it - * should be closed - */ -#define SVC_DENIED 8 -#define SVC_PENDING 9 -#define SVC_COMPLETE 10 + enum svc_auth_status (*accept)(struct svc_rqst *rqstp); + int (*release)(struct svc_rqst *rqstp); + void (*domain_release)(struct auth_domain *dom); + enum svc_auth_status (*set_client)(struct svc_rqst *rqstp); +}; struct svc_xprt; -extern int svc_authenticate(struct svc_rqst *rqstp); +extern enum svc_auth_status svc_authenticate(struct svc_rqst *rqstp); extern int svc_authorise(struct svc_rqst *rqstp); -extern int svc_set_client(struct svc_rqst *rqstp); +extern enum svc_auth_status svc_set_client(struct svc_rqst *rqstp); extern int svc_auth_register(rpc_authflavor_t flavor, struct auth_ops *aops); extern void svc_auth_unregister(rpc_authflavor_t flavor); @@ -161,7 +159,7 @@ extern struct auth_domain *auth_domain_lookup(char *name, struct auth_domain *ne extern struct auth_domain *auth_domain_find(char *name); extern void svcauth_unix_purge(struct net *net); extern void svcauth_unix_info_release(struct svc_xprt *xpt); -extern int svcauth_unix_set_client(struct svc_rqst *rqstp); +extern enum svc_auth_status svcauth_unix_set_client(struct svc_rqst *rqstp); extern int unix_gid_cache_create(struct net *net); extern void unix_gid_cache_destroy(struct net *net); diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 00db9e1fb7118..55716b62ce914 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -1706,7 +1706,7 @@ TRACE_DEFINE_ENUM(SVC_DENIED); TRACE_DEFINE_ENUM(SVC_PENDING); TRACE_DEFINE_ENUM(SVC_COMPLETE); -#define svc_show_status(status) \ +#define show_svc_auth_status(status) \ __print_symbolic(status, \ { SVC_GARBAGE, "SVC_GARBAGE" }, \ { SVC_SYSERR, "SVC_SYSERR" }, \ @@ -1743,7 +1743,10 @@ TRACE_DEFINE_ENUM(SVC_COMPLETE); __entry->xid, __get_sockaddr(server), __get_sockaddr(client) TRACE_EVENT_CONDITION(svc_authenticate, - TP_PROTO(const struct svc_rqst *rqst, int auth_res), + TP_PROTO( + const struct svc_rqst *rqst, + enum svc_auth_status auth_res + ), TP_ARGS(rqst, auth_res), @@ -1766,7 +1769,7 @@ TRACE_EVENT_CONDITION(svc_authenticate, TP_printk(SVC_RQST_ENDPOINT_FORMAT " auth_res=%s auth_stat=%s", SVC_RQST_ENDPOINT_VARARGS, - svc_show_status(__entry->svc_status), + show_svc_auth_status(__entry->svc_status), rpc_show_auth_stat(__entry->auth_stat)) ); diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index c4a566737085c..18734e70c5ddb 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -986,7 +986,7 @@ svcauth_gss_unwrap_priv(struct svc_rqst *rqstp, u32 seq, struct gss_ctx *ctx) return -EINVAL; } -static int +static enum svc_auth_status svcauth_gss_set_client(struct svc_rqst *rqstp) { struct gss_svc_data *svcdata = rqstp->rq_auth_data; @@ -1634,7 +1634,7 @@ svcauth_gss_decode_credbody(struct xdr_stream *xdr, * * The rqstp->rq_auth_stat field is also set (see RFCs 2203 and 5531). */ -static int +static enum svc_auth_status svcauth_gss_accept(struct svc_rqst *rqstp) { struct gss_svc_data *svcdata = rqstp->rq_auth_data; @@ -1945,9 +1945,6 @@ static int svcauth_gss_wrap_priv(struct svc_rqst *rqstp) * %0: the Reply is ready to be sent * %-ENOMEM: failed to allocate memory * %-EINVAL: encoding error - * - * XXX: These return values do not match the return values documented - * for the auth_ops ->release method in linux/sunrpc/svcauth.h. */ static int svcauth_gss_release(struct svc_rqst *rqstp) diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 030f8c759ee60..7873a3ff24a83 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1275,8 +1275,9 @@ svc_process_common(struct svc_rqst *rqstp) const struct svc_procedure *procp = NULL; struct svc_serv *serv = rqstp->rq_server; struct svc_process_info process; - int auth_res, rc; + enum svc_auth_status auth_res; unsigned int aoffset; + int rc; __be32 *p; /* Will be turned off by GSS integrity and privacy services */ @@ -1331,6 +1332,9 @@ svc_process_common(struct svc_rqst *rqstp) goto dropit; case SVC_COMPLETE: goto sendit; + default: + pr_warn_once("Unexpected svc_auth_status (%d)\n", auth_res); + goto err_system_err; } if (progp == NULL) diff --git a/net/sunrpc/svcauth.c b/net/sunrpc/svcauth.c index 67d8245a08afb..aa4429d0b8106 100644 --- a/net/sunrpc/svcauth.c +++ b/net/sunrpc/svcauth.c @@ -60,8 +60,19 @@ svc_put_auth_ops(struct auth_ops *aops) module_put(aops->owner); } -int -svc_authenticate(struct svc_rqst *rqstp) +/** + * svc_authenticate - Initialize an outgoing credential + * @rqstp: RPC execution context + * + * Return values: + * %SVC_OK: XDR encoding of the result can begin + * %SVC_DENIED: Credential or verifier is not valid + * %SVC_GARBAGE: Failed to decode credential or verifier + * %SVC_COMPLETE: GSS context lifetime event; no further action + * %SVC_DROP: Drop this request; no further action + * %SVC_CLOSE: Like drop, but also close transport connection + */ +enum svc_auth_status svc_authenticate(struct svc_rqst *rqstp) { struct auth_ops *aops; u32 flavor; @@ -89,16 +100,28 @@ svc_authenticate(struct svc_rqst *rqstp) } EXPORT_SYMBOL_GPL(svc_authenticate); -int svc_set_client(struct svc_rqst *rqstp) +/** + * svc_set_client - Assign an appropriate 'auth_domain' as the client + * @rqstp: RPC execution context + * + * Return values: + * %SVC_OK: Client was found and assigned + * %SVC_DENY: Client was explicitly denied + * %SVC_DROP: Ignore this request + * %SVC_CLOSE: Ignore this request and close the connection + */ +enum svc_auth_status svc_set_client(struct svc_rqst *rqstp) { rqstp->rq_client = NULL; return rqstp->rq_authop->set_client(rqstp); } EXPORT_SYMBOL_GPL(svc_set_client); -/* A request, which was authenticated, has now executed. - * Time to finalise the credentials and verifier - * and release and resources +/** + * svc_authorise - Finalize credentials/verifier and release resources + * @rqstp: RPC execution context + * + * Returns zero on success, or a negative errno. */ int svc_authorise(struct svc_rqst *rqstp) { diff --git a/net/sunrpc/svcauth_unix.c b/net/sunrpc/svcauth_unix.c index 174783f804fa5..04b45588ae6fe 100644 --- a/net/sunrpc/svcauth_unix.c +++ b/net/sunrpc/svcauth_unix.c @@ -665,7 +665,7 @@ static struct group_info *unix_gid_find(kuid_t uid, struct svc_rqst *rqstp) } } -int +enum svc_auth_status svcauth_unix_set_client(struct svc_rqst *rqstp) { struct sockaddr_in *sin; @@ -736,7 +736,6 @@ svcauth_unix_set_client(struct svc_rqst *rqstp) rqstp->rq_auth_stat = rpc_auth_ok; return SVC_OK; } - EXPORT_SYMBOL_GPL(svcauth_unix_set_client); /** @@ -751,7 +750,7 @@ EXPORT_SYMBOL_GPL(svcauth_unix_set_client); * * rqstp->rq_auth_stat is set as mandated by RFC 5531. */ -static int +static enum svc_auth_status svcauth_null_accept(struct svc_rqst *rqstp) { struct xdr_stream *xdr = &rqstp->rq_arg_stream; @@ -828,7 +827,7 @@ struct auth_ops svcauth_null = { * * rqstp->rq_auth_stat is set as mandated by RFC 5531. */ -static int +static enum svc_auth_status svcauth_tls_accept(struct svc_rqst *rqstp) { struct xdr_stream *xdr = &rqstp->rq_arg_stream; @@ -913,7 +912,7 @@ struct auth_ops svcauth_tls = { * * rqstp->rq_auth_stat is set as mandated by RFC 5531. */ -static int +static enum svc_auth_status svcauth_unix_accept(struct svc_rqst *rqstp) { struct xdr_stream *xdr = &rqstp->rq_arg_stream; From 82e5d82a45741839bd9dcb6636cfcf67747a5af5 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jul 2023 15:24:42 -0400 Subject: [PATCH 170/186] SUNRPC: Move trace_svc_xprt_enqueue The xpt_flags field frequently changes between the time that svc_xprt_ready() grabs a copy and execution flow arrives at the tracepoint at the tail of svc_xprt_enqueue(). In fact, there's usually a sleep/wake-up in there, so those flags are almost guaranteed to be different. It would be more useful to record the exact flags that were used to decide whether the transport is ready, so move the tracepoint. Moving it means the tracepoint can't pick up the waker's pid. That can be added to struct svc_rqst if it turns out that is important. Signed-off-by: Chuck Lever --- include/trace/events/sunrpc.h | 18 +++++++++--------- net/sunrpc/svc_xprt.c | 2 +- 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 55716b62ce914..6beb38c1dcb5e 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -2014,25 +2014,25 @@ TRACE_EVENT(svc_xprt_create_err, TRACE_EVENT(svc_xprt_enqueue, TP_PROTO( const struct svc_xprt *xprt, - const struct svc_rqst *rqst + unsigned long flags ), - TP_ARGS(xprt, rqst), + TP_ARGS(xprt, flags), TP_STRUCT__entry( SVC_XPRT_ENDPOINT_FIELDS(xprt) - - __field(int, pid) ), TP_fast_assign( - SVC_XPRT_ENDPOINT_ASSIGNMENTS(xprt); - - __entry->pid = rqst? rqst->rq_task->pid : 0; + __assign_sockaddr(server, &xprt->xpt_local, + xprt->xpt_locallen); + __assign_sockaddr(client, &xprt->xpt_remote, + xprt->xpt_remotelen); + __entry->flags = flags; + __entry->netns_ino = xprt->xpt_net->ns.inum; ), - TP_printk(SVC_XPRT_ENDPOINT_FORMAT " pid=%d", - SVC_XPRT_ENDPOINT_VARARGS, __entry->pid) + TP_printk(SVC_XPRT_ENDPOINT_FORMAT, SVC_XPRT_ENDPOINT_VARARGS) ); TRACE_EVENT(svc_xprt_dequeue, diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 9bdcdd8401b86..cdea4b49cbc5d 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -434,6 +434,7 @@ static bool svc_xprt_ready(struct svc_xprt *xprt) smp_rmb(); xpt_flags = READ_ONCE(xprt->xpt_flags); + trace_svc_xprt_enqueue(xprt, xpt_flags); if (xpt_flags & BIT(XPT_BUSY)) return false; if (xpt_flags & (BIT(XPT_CONN) | BIT(XPT_CLOSE) | BIT(XPT_HANDSHAKE))) @@ -490,7 +491,6 @@ void svc_xprt_enqueue(struct svc_xprt *xprt) rqstp = NULL; out_unlock: rcu_read_unlock(); - trace_svc_xprt_enqueue(xprt, rqstp); } EXPORT_SYMBOL_GPL(svc_xprt_enqueue); From 850bac3ae4a636e9e6bb8de62fe697ac171cb221 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 10 Jul 2023 12:42:00 -0400 Subject: [PATCH 171/186] SUNRPC: Deduplicate thread wake-up code Refactor: Extract the loop that finds an idle service thread from svc_xprt_enqueue() and svc_wake_up(). Both functions do just about the same thing. Note that svc_wake_up() currently does not hold the RCU read lock while waking the target thread. It indeed should hold the lock, just as svc_xprt_enqueue() does, to ensure the rqstp does not vanish during the wake-up. This patch adds the RCU lock for svc_wake_up(). Note that shrinking the pool thread count is rare, and calls to svc_wake_up() are also quite infrequent. In practice, this race is very unlikely to be hit, so we are not marking the lock fix for stable backport at this time. Reviewed-by: Jeff Layton Reviewed-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 1 + net/sunrpc/svc.c | 34 +++++++++++++++++++++++++++++ net/sunrpc/svc_xprt.c | 44 +++++++------------------------------- 3 files changed, 43 insertions(+), 36 deletions(-) diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 1c491f02efc87..2b9c6df230784 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -419,6 +419,7 @@ int svc_register(const struct svc_serv *, struct net *, const int, void svc_wake_up(struct svc_serv *); void svc_reserve(struct svc_rqst *rqstp, int space); +bool svc_pool_wake_idle_thread(struct svc_pool *pool); struct svc_pool *svc_pool_for_cpu(struct svc_serv *serv); char * svc_print_addr(struct svc_rqst *, char *, size_t); const char * svc_proc_name(const struct svc_rqst *rqstp); diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 7873a3ff24a83..ee9d55b2f2752 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -687,6 +687,40 @@ svc_prepare_thread(struct svc_serv *serv, struct svc_pool *pool, int node) return rqstp; } +/** + * svc_pool_wake_idle_thread - Awaken an idle thread in @pool + * @pool: service thread pool + * + * Can be called from soft IRQ or process context. Finding an idle + * service thread and marking it BUSY is atomic with respect to + * other calls to svc_pool_wake_idle_thread(). + * + * Return value: + * %true: An idle thread was awoken + * %false: No idle thread was found + */ +bool svc_pool_wake_idle_thread(struct svc_pool *pool) +{ + struct svc_rqst *rqstp; + + rcu_read_lock(); + list_for_each_entry_rcu(rqstp, &pool->sp_all_threads, rq_all) { + if (test_and_set_bit(RQ_BUSY, &rqstp->rq_flags)) + continue; + + WRITE_ONCE(rqstp->rq_qtime, ktime_get()); + wake_up_process(rqstp->rq_task); + rcu_read_unlock(); + percpu_counter_inc(&pool->sp_threads_woken); + trace_svc_wake_up(rqstp->rq_task->pid); + return true; + } + rcu_read_unlock(); + + set_bit(SP_CONGESTED, &pool->sp_flags); + return false; +} + /* * Choose a pool in which to create a new thread, for svc_set_num_threads */ diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index cdea4b49cbc5d..20c66f6591133 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -457,7 +457,6 @@ static bool svc_xprt_ready(struct svc_xprt *xprt) void svc_xprt_enqueue(struct svc_xprt *xprt) { struct svc_pool *pool; - struct svc_rqst *rqstp = NULL; if (!svc_xprt_ready(xprt)) return; @@ -477,20 +476,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt) list_add_tail(&xprt->xpt_ready, &pool->sp_sockets); spin_unlock_bh(&pool->sp_lock); - /* find a thread for this xprt */ - rcu_read_lock(); - list_for_each_entry_rcu(rqstp, &pool->sp_all_threads, rq_all) { - if (test_and_set_bit(RQ_BUSY, &rqstp->rq_flags)) - continue; - percpu_counter_inc(&pool->sp_threads_woken); - rqstp->rq_qtime = ktime_get(); - wake_up_process(rqstp->rq_task); - goto out_unlock; - } - set_bit(SP_CONGESTED, &pool->sp_flags); - rqstp = NULL; -out_unlock: - rcu_read_unlock(); + svc_pool_wake_idle_thread(pool); } EXPORT_SYMBOL_GPL(svc_xprt_enqueue); @@ -581,7 +567,10 @@ static void svc_xprt_release(struct svc_rqst *rqstp) svc_xprt_put(xprt); } -/* +/** + * svc_wake_up - Wake up a service thread for non-transport work + * @serv: RPC service + * * Some svc_serv's will have occasional work to do, even when a xprt is not * waiting to be serviced. This function is there to "kick" a task in one of * those services so that it can wake up and do that work. Note that we only @@ -590,27 +579,10 @@ static void svc_xprt_release(struct svc_rqst *rqstp) */ void svc_wake_up(struct svc_serv *serv) { - struct svc_rqst *rqstp; - struct svc_pool *pool; - - pool = &serv->sv_pools[0]; - - rcu_read_lock(); - list_for_each_entry_rcu(rqstp, &pool->sp_all_threads, rq_all) { - /* skip any that aren't queued */ - if (test_bit(RQ_BUSY, &rqstp->rq_flags)) - continue; - rcu_read_unlock(); - wake_up_process(rqstp->rq_task); - trace_svc_wake_up(rqstp->rq_task->pid); - return; - } - rcu_read_unlock(); + struct svc_pool *pool = &serv->sv_pools[0]; - /* No free entries available */ - set_bit(SP_TASK_PENDING, &pool->sp_flags); - smp_wmb(); - trace_svc_wake_up(0); + if (!svc_pool_wake_idle_thread(pool)) + set_bit(SP_TASK_PENDING, &pool->sp_flags); } EXPORT_SYMBOL_GPL(svc_wake_up); From f208e9508ace01864f2b37a45d07cda0641ff3ea Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 10 Jul 2023 12:42:20 -0400 Subject: [PATCH 172/186] SUNRPC: Count ingress RPC messages per svc_pool svc_xprt_enqueue() can be costly, since it involves selecting and waking up a process. More than one enqueue is done per incoming RPC. For example, svc_data_ready() enqueues, and so does svc_xprt_receive(). Also, if an RPC message requires more than one call to ->recvfrom() to receive it fully, each one of those calls does an enqueue. To get a sense of the average number of transport enqueue operations needed to process an incoming RPC message, re-use the "packets" pool stat. Track the number of complete RPC messages processed by each thread pool. Reviewed-by: Jeff Layton Reviewed-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 1 + net/sunrpc/svc.c | 2 ++ net/sunrpc/svc_xprt.c | 3 ++- 3 files changed, 5 insertions(+), 1 deletion(-) diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 2b9c6df230784..7838b37bcfa8a 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -39,6 +39,7 @@ struct svc_pool { struct list_head sp_all_threads; /* all server threads */ /* statistics on pool operation */ + struct percpu_counter sp_messages_arrived; struct percpu_counter sp_sockets_queued; struct percpu_counter sp_threads_woken; diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index ee9d55b2f2752..98dff8dfbac89 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -513,6 +513,7 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools, INIT_LIST_HEAD(&pool->sp_all_threads); spin_lock_init(&pool->sp_lock); + percpu_counter_init(&pool->sp_messages_arrived, 0, GFP_KERNEL); percpu_counter_init(&pool->sp_sockets_queued, 0, GFP_KERNEL); percpu_counter_init(&pool->sp_threads_woken, 0, GFP_KERNEL); } @@ -587,6 +588,7 @@ svc_destroy(struct kref *ref) for (i = 0; i < serv->sv_nrpools; i++) { struct svc_pool *pool = &serv->sv_pools[i]; + percpu_counter_destroy(&pool->sp_messages_arrived); percpu_counter_destroy(&pool->sp_sockets_queued); percpu_counter_destroy(&pool->sp_threads_woken); } diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 20c66f6591133..d3280ae70e365 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -871,6 +871,7 @@ void svc_recv(struct svc_rqst *rqstp) if (serv->sv_stats) serv->sv_stats->netcnt++; + percpu_counter_inc(&rqstp->rq_pool->sp_messages_arrived); rqstp->rq_stime = ktime_get(); svc_process(rqstp); out: @@ -1420,7 +1421,7 @@ static int svc_pool_stats_show(struct seq_file *m, void *p) seq_printf(m, "%u %llu %llu %llu 0\n", pool->sp_id, - percpu_counter_sum_positive(&pool->sp_sockets_queued), + percpu_counter_sum_positive(&pool->sp_messages_arrived), percpu_counter_sum_positive(&pool->sp_sockets_queued), percpu_counter_sum_positive(&pool->sp_threads_woken)); From d2f0ef1cbf37e396ef9c57a30c004ebe65cdbca9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 10 Jul 2023 12:42:33 -0400 Subject: [PATCH 173/186] SUNRPC: Clean up svc_set_num_threads Document the API contract and remove stale or obvious comments. Reviewed-by: Jeff Layton Reviewed-by: NeilBrown Signed-off-by: Chuck Lever --- net/sunrpc/svc.c | 60 ++++++++++++++++++++---------------------------- 1 file changed, 25 insertions(+), 35 deletions(-) diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 98dff8dfbac89..af692bff44ab6 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -723,23 +723,14 @@ bool svc_pool_wake_idle_thread(struct svc_pool *pool) return false; } -/* - * Choose a pool in which to create a new thread, for svc_set_num_threads - */ -static inline struct svc_pool * -choose_pool(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) +static struct svc_pool * +svc_pool_next(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) { - if (pool != NULL) - return pool; - - return &serv->sv_pools[(*state)++ % serv->sv_nrpools]; + return pool ? pool : &serv->sv_pools[(*state)++ % serv->sv_nrpools]; } -/* - * Choose a thread to kill, for svc_set_num_threads - */ -static inline struct task_struct * -choose_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) +static struct task_struct * +svc_pool_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) { unsigned int i; struct task_struct *task = NULL; @@ -747,7 +738,6 @@ choose_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) if (pool != NULL) { spin_lock_bh(&pool->sp_lock); } else { - /* choose a pool in round-robin fashion */ for (i = 0; i < serv->sv_nrpools; i++) { pool = &serv->sv_pools[--(*state) % serv->sv_nrpools]; spin_lock_bh(&pool->sp_lock); @@ -762,21 +752,15 @@ choose_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state) if (!list_empty(&pool->sp_all_threads)) { struct svc_rqst *rqstp; - /* - * Remove from the pool->sp_all_threads list - * so we don't try to kill it again. - */ rqstp = list_entry(pool->sp_all_threads.next, struct svc_rqst, rq_all); set_bit(RQ_VICTIM, &rqstp->rq_flags); list_del_rcu(&rqstp->rq_all); task = rqstp->rq_task; } spin_unlock_bh(&pool->sp_lock); - return task; } -/* create new threads */ static int svc_start_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) { @@ -788,13 +772,12 @@ svc_start_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) do { nrservs--; - chosen_pool = choose_pool(serv, pool, &state); - + chosen_pool = svc_pool_next(serv, pool, &state); node = svc_pool_map_get_node(chosen_pool->sp_id); + rqstp = svc_prepare_thread(serv, chosen_pool, node); if (IS_ERR(rqstp)) return PTR_ERR(rqstp); - task = kthread_create_on_node(serv->sv_threadfn, rqstp, node, "%s", serv->sv_name); if (IS_ERR(task)) { @@ -813,15 +796,6 @@ svc_start_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) return 0; } -/* - * Create or destroy enough new threads to make the number - * of threads the given number. If `pool' is non-NULL, applies - * only to threads in that pool, otherwise round-robins between - * all pools. Caller must ensure that mutual exclusion between this and - * server startup or shutdown. - */ - -/* destroy old threads */ static int svc_stop_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) { @@ -829,9 +803,8 @@ svc_stop_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) struct task_struct *task; unsigned int state = serv->sv_nrthreads-1; - /* destroy old threads */ do { - task = choose_victim(serv, pool, &state); + task = svc_pool_victim(serv, pool, &state); if (task == NULL) break; rqstp = kthread_data(task); @@ -843,6 +816,23 @@ svc_stop_kthreads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) return 0; } +/** + * svc_set_num_threads - adjust number of threads per RPC service + * @serv: RPC service to adjust + * @pool: Specific pool from which to choose threads, or NULL + * @nrservs: New number of threads for @serv (0 or less means kill all threads) + * + * Create or destroy threads to make the number of threads for @serv the + * given number. If @pool is non-NULL, change only threads in that pool; + * otherwise, round-robin between all pools for @serv. @serv's + * sv_nrthreads is adjusted for each thread created or destroyed. + * + * Caller must ensure mutual exclusion between this and server startup or + * shutdown. + * + * Returns zero on success or a negative errno if an error occurred while + * starting a thread. + */ int svc_set_num_threads(struct svc_serv *serv, struct svc_pool *pool, int nrservs) { From 6859d1f2902c600f6b1c587c91408a91e05cdc02 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 31 Jul 2023 16:48:28 +1000 Subject: [PATCH 174/186] SUNRPC: make rqst_should_sleep() idempotent() Based on its name you would think that rqst_should_sleep() would be read-only, not changing anything. But in fact it will clear SP_TASK_PENDING if that was set. This is surprising, and it blurs the line between "check for work to do" and "dequeue work to do". So change the "test_and_clear" to simple "test" and clear the bit once the thread has decided to wake up and return to the caller. With this, it makes sense to *always* set SP_TASK_PENDING when asked, rather than to set it only if no thread could be woken up. [ cel: Previously TASK_PENDING indicated there is work waiting but no idle threads were found to pick up that work. After this patch, it acts as an XPT_BUSY flag for wake-ups that have no associated xprt. ] Signed-off-by: NeilBrown Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- net/sunrpc/svc_xprt.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index d3280ae70e365..4cfe9640df481 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -581,8 +581,8 @@ void svc_wake_up(struct svc_serv *serv) { struct svc_pool *pool = &serv->sv_pools[0]; - if (!svc_pool_wake_idle_thread(pool)) - set_bit(SP_TASK_PENDING, &pool->sp_flags); + set_bit(SP_TASK_PENDING, &pool->sp_flags); + svc_pool_wake_idle_thread(pool); } EXPORT_SYMBOL_GPL(svc_wake_up); @@ -704,7 +704,7 @@ rqst_should_sleep(struct svc_rqst *rqstp) struct svc_pool *pool = rqstp->rq_pool; /* did someone call svc_wake_up? */ - if (test_and_clear_bit(SP_TASK_PENDING, &pool->sp_flags)) + if (test_bit(SP_TASK_PENDING, &pool->sp_flags)) return false; /* was a socket queued? */ @@ -748,6 +748,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp) set_bit(RQ_BUSY, &rqstp->rq_flags); smp_mb__after_atomic(); + clear_bit(SP_TASK_PENDING, &pool->sp_flags); rqstp->rq_xprt = svc_xprt_dequeue(pool); if (rqstp->rq_xprt) goto out_found; @@ -756,6 +757,7 @@ static struct svc_xprt *svc_get_next_xprt(struct svc_rqst *rqstp) return NULL; return NULL; out_found: + clear_bit(SP_TASK_PENDING, &pool->sp_flags); /* Normally we will wait up to 5 seconds for any required * cache information to be provided. */ From 2a4557452aacf9e7168cb83bc102467094ff9391 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 31 Jul 2023 16:48:29 +1000 Subject: [PATCH 175/186] SUNRPC: Remove return value of svc_pool_wake_idle_thread() The returned value is not used (any more), so don't return it. Signed-off-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 2 +- net/sunrpc/svc.c | 8 ++------ 2 files changed, 3 insertions(+), 7 deletions(-) diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 7838b37bcfa8a..dbf5b21feafe4 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -420,7 +420,7 @@ int svc_register(const struct svc_serv *, struct net *, const int, void svc_wake_up(struct svc_serv *); void svc_reserve(struct svc_rqst *rqstp, int space); -bool svc_pool_wake_idle_thread(struct svc_pool *pool); +void svc_pool_wake_idle_thread(struct svc_pool *pool); struct svc_pool *svc_pool_for_cpu(struct svc_serv *serv); char * svc_print_addr(struct svc_rqst *, char *, size_t); const char * svc_proc_name(const struct svc_rqst *rqstp); diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index af692bff44ab6..dc21e6c732db3 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -697,11 +697,8 @@ svc_prepare_thread(struct svc_serv *serv, struct svc_pool *pool, int node) * service thread and marking it BUSY is atomic with respect to * other calls to svc_pool_wake_idle_thread(). * - * Return value: - * %true: An idle thread was awoken - * %false: No idle thread was found */ -bool svc_pool_wake_idle_thread(struct svc_pool *pool) +void svc_pool_wake_idle_thread(struct svc_pool *pool) { struct svc_rqst *rqstp; @@ -715,12 +712,11 @@ bool svc_pool_wake_idle_thread(struct svc_pool *pool) rcu_read_unlock(); percpu_counter_inc(&pool->sp_threads_woken); trace_svc_wake_up(rqstp->rq_task->pid); - return true; + return; } rcu_read_unlock(); set_bit(SP_CONGESTED, &pool->sp_flags); - return false; } static struct svc_pool * From 6372e2ee629894433fe6107d7048536a3280a284 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 16 Aug 2023 10:20:52 -0400 Subject: [PATCH 176/186] NFSD: da_addr_body field missing in some GETDEVICEINFO replies The XDR specification in RFC 8881 looks like this: struct device_addr4 { layouttype4 da_layout_type; opaque da_addr_body<>; }; struct GETDEVICEINFO4resok { device_addr4 gdir_device_addr; bitmap4 gdir_notification; }; union GETDEVICEINFO4res switch (nfsstat4 gdir_status) { case NFS4_OK: GETDEVICEINFO4resok gdir_resok4; case NFS4ERR_TOOSMALL: count4 gdir_mincount; default: void; }; Looking at nfsd4_encode_getdeviceinfo() .... When the client provides a zero gd_maxcount, then the Linux NFS server implementation encodes the da_layout_type field and then skips the da_addr_body field completely, proceeding directly to encode gdir_notification field. There does not appear to be an option in the specification to skip encoding da_addr_body. Moreover, Section 18.40.3 says: > If the client wants to just update or turn off notifications, it > MAY send a GETDEVICEINFO operation with gdia_maxcount set to zero. > In that event, if the device ID is valid, the reply's da_addr_body > field of the gdir_device_addr field will be of zero length. Since the layout drivers are responsible for encoding the da_addr_body field, put this fix inside the ->encode_getdeviceinfo methods. Fixes: 9cf514ccfacb ("nfsd: implement pNFS operations") Reviewed-by: Christoph Hellwig Cc: Tom Haynes Signed-off-by: Chuck Lever --- fs/nfsd/blocklayoutxdr.c | 9 +++++++++ fs/nfsd/flexfilelayoutxdr.c | 9 +++++++++ fs/nfsd/nfs4xdr.c | 25 +++++++++++-------------- 3 files changed, 29 insertions(+), 14 deletions(-) diff --git a/fs/nfsd/blocklayoutxdr.c b/fs/nfsd/blocklayoutxdr.c index 8e9c1a0f8d380..1ed2f691ebb90 100644 --- a/fs/nfsd/blocklayoutxdr.c +++ b/fs/nfsd/blocklayoutxdr.c @@ -83,6 +83,15 @@ nfsd4_block_encode_getdeviceinfo(struct xdr_stream *xdr, int len = sizeof(__be32), ret, i; __be32 *p; + /* + * See paragraph 5 of RFC 8881 S18.40.3. + */ + if (!gdp->gd_maxcount) { + if (xdr_stream_encode_u32(xdr, 0) != XDR_UNIT) + return nfserr_resource; + return nfs_ok; + } + p = xdr_reserve_space(xdr, len + sizeof(__be32)); if (!p) return nfserr_resource; diff --git a/fs/nfsd/flexfilelayoutxdr.c b/fs/nfsd/flexfilelayoutxdr.c index e81d2a5cf381e..bb205328e043d 100644 --- a/fs/nfsd/flexfilelayoutxdr.c +++ b/fs/nfsd/flexfilelayoutxdr.c @@ -85,6 +85,15 @@ nfsd4_ff_encode_getdeviceinfo(struct xdr_stream *xdr, int addr_len; __be32 *p; + /* + * See paragraph 5 of RFC 8881 S18.40.3. + */ + if (!gdp->gd_maxcount) { + if (xdr_stream_encode_u32(xdr, 0) != XDR_UNIT) + return nfserr_resource; + return nfs_ok; + } + /* len + padding for two strings */ addr_len = 16 + da->netaddr.netid_len + da->netaddr.addr_len; ver_len = 20; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index d4de39404cde8..2e40c74d2f727 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -4686,20 +4686,17 @@ nfsd4_encode_getdeviceinfo(struct nfsd4_compoundres *resp, __be32 nfserr, *p++ = cpu_to_be32(gdev->gd_layout_type); - /* If maxcount is 0 then just update notifications */ - if (gdev->gd_maxcount != 0) { - ops = nfsd4_layout_ops[gdev->gd_layout_type]; - nfserr = ops->encode_getdeviceinfo(xdr, gdev); - if (nfserr) { - /* - * We don't bother to burden the layout drivers with - * enforcing gd_maxcount, just tell the client to - * come back with a bigger buffer if it's not enough. - */ - if (xdr->buf->len + 4 > gdev->gd_maxcount) - goto toosmall; - return nfserr; - } + ops = nfsd4_layout_ops[gdev->gd_layout_type]; + nfserr = ops->encode_getdeviceinfo(xdr, gdev); + if (nfserr) { + /* + * We don't bother to burden the layout drivers with + * enforcing gd_maxcount, just tell the client to + * come back with a bigger buffer if it's not enough. + */ + if (xdr->buf->len + 4 > gdev->gd_maxcount) + goto toosmall; + return nfserr; } if (gdev->gd_notify_types) { From 07dc19dbd1d194397d7ae1c4781e203f8419b3c6 Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Mon, 21 Aug 2023 20:33:46 +0800 Subject: [PATCH 177/186] SUNRPC: Remove unused declarations Commit c7d7ec8f043e ("SUNRPC: Remove svc_shutdown_net()") removed svc_close_net() implementation but left declaration in place. Remove it. Commit 1f11a034cdc4 ("SUNRPC new transport for the NFSv4.1 shared back channel") removed svc_sock_create()/svc_sock_destroy() but not the declarations. Signed-off-by: Yue Haibing Signed-off-by: Chuck Lever --- include/linux/sunrpc/svcsock.h | 3 --- 1 file changed, 3 deletions(-) diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index d4a173c5b3be1..7c78ec6356b92 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -56,7 +56,6 @@ static inline u32 svc_sock_final_rec(struct svc_sock *svsk) /* * Function prototypes. */ -void svc_close_net(struct svc_serv *, struct net *); void svc_recv(struct svc_rqst *rqstp); void svc_send(struct svc_rqst *rqstp); void svc_drop(struct svc_rqst *); @@ -66,8 +65,6 @@ int svc_addsock(struct svc_serv *serv, struct net *net, const struct cred *cred); void svc_init_xprt_sock(void); void svc_cleanup_xprt_sock(void); -struct svc_xprt *svc_sock_create(struct svc_serv *serv, int prot); -void svc_sock_destroy(struct svc_xprt *); /* * svc_makesock socket characteristics From 899525e892dd165d2bb2e41f9f9d9d82574b31b5 Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Wed, 9 Aug 2023 22:14:26 +0800 Subject: [PATCH 178/186] SUNRPC: Remove unused declaration rpc_modcount() These declarations are never implemented since the beginning of git history. Remove these, then merge the two #ifdef block for simplification. Signed-off-by: Yue Haibing Reviewed-by: NeilBrown Signed-off-by: Chuck Lever --- include/linux/sunrpc/stats.h | 23 +++++++---------------- 1 file changed, 7 insertions(+), 16 deletions(-) diff --git a/include/linux/sunrpc/stats.h b/include/linux/sunrpc/stats.h index d94d4f4105074..3ce1550d1beb3 100644 --- a/include/linux/sunrpc/stats.h +++ b/include/linux/sunrpc/stats.h @@ -43,22 +43,6 @@ struct net; #ifdef CONFIG_PROC_FS int rpc_proc_init(struct net *); void rpc_proc_exit(struct net *); -#else -static inline int rpc_proc_init(struct net *net) -{ - return 0; -} - -static inline void rpc_proc_exit(struct net *net) -{ -} -#endif - -#ifdef MODULE -void rpc_modcount(struct inode *, int); -#endif - -#ifdef CONFIG_PROC_FS struct proc_dir_entry * rpc_proc_register(struct net *,struct rpc_stat *); void rpc_proc_unregister(struct net *,const char *); void rpc_proc_zero(const struct rpc_program *); @@ -69,7 +53,14 @@ void svc_proc_unregister(struct net *, const char *); void svc_seq_show(struct seq_file *, const struct svc_stat *); #else +static inline int rpc_proc_init(struct net *net) +{ + return 0; +} +static inline void rpc_proc_exit(struct net *net) +{ +} static inline struct proc_dir_entry *rpc_proc_register(struct net *net, struct rpc_stat *s) { return NULL; } static inline void rpc_proc_unregister(struct net *net, const char *p) {} static inline void rpc_proc_zero(const struct rpc_program *p) {} From b38a6023da6a12b561f0421c6a5a1f7624a1529c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 25 Aug 2023 15:04:23 -0400 Subject: [PATCH 179/186] Documentation: Add missing documentation for EXPORT_OP flags The commits that introduced these flags neglected to update the Documentation/filesystems/nfs/exporting.rst file. Signed-off-by: Chuck Lever --- Documentation/filesystems/nfs/exporting.rst | 26 +++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/Documentation/filesystems/nfs/exporting.rst b/Documentation/filesystems/nfs/exporting.rst index 3d97b8d8f7354..4b30daee399af 100644 --- a/Documentation/filesystems/nfs/exporting.rst +++ b/Documentation/filesystems/nfs/exporting.rst @@ -215,3 +215,29 @@ following flags are defined: This flag causes nfsd to close any open files for this inode _before_ calling into the vfs to do an unlink or a rename that would replace an existing file. + + EXPORT_OP_REMOTE_FS - Backing storage for this filesystem is remote + PF_LOCAL_THROTTLE exists for loopback NFSD, where a thread needs to + write to one bdi (the final bdi) in order to free up writes queued + to another bdi (the client bdi). Such threads get a private balance + of dirty pages so that dirty pages for the client bdi do not imact + the daemon writing to the final bdi. For filesystems whose durable + storage is not local (such as exported NFS filesystems), this + constraint has negative consequences. EXPORT_OP_REMOTE_FS enables + an export to disable writeback throttling. + + EXPORT_OP_NOATOMIC_ATTR - Filesystem does not update attributes atomically + EXPORT_OP_NOATOMIC_ATTR indicates that the exported filesystem + cannot provide the semantics required by the "atomic" boolean in + NFSv4's change_info4. This boolean indicates to a client whether the + returned before and after change attributes were obtained atomically + with the respect to the requested metadata operation (UNLINK, + OPEN/CREATE, MKDIR, etc). + + EXPORT_OP_FLUSH_ON_CLOSE - Filesystem flushes file data on close(2) + On most filesystems, inodes can remain under writeback after the + file is closed. NFSD relies on client activity or local flusher + threads to handle writeback. Certain filesystems, such as NFS, flush + all of an inode's dirty data on last close. Exports that behave this + way should set EXPORT_OP_FLUSH_ON_CLOSE so that NFSD knows to skip + waiting for writeback when closing such files. From f67b55b6588bcf9316a1e6e8d529100a5aa3ebe6 Mon Sep 17 00:00:00 2001 From: Benjamin Coddington Date: Tue, 22 Aug 2023 14:22:38 -0400 Subject: [PATCH 180/186] NFS: Guard against READDIR loop when entry names exceed MAXNAMELEN Commit 64cfca85bacd asserts the only valid return values for nfs2/3_decode_dirent should not include -ENAMETOOLONG, but for a server that sends a filename3 which exceeds MAXNAMELEN in a READDIR response the client's behavior will be to endlessly retry the operation. We could map -ENAMETOOLONG into -EBADCOOKIE, but that would produce truncated listings without any error. The client should return an error for this case to clearly assert that the server implementation must be corrected. Fixes: 64cfca85bacd ("NFS: Return valid errors from nfs2/3_decode_dirent()") Signed-off-by: Benjamin Coddington Signed-off-by: Anna Schumaker --- fs/nfs/nfs2xdr.c | 2 +- fs/nfs/nfs3xdr.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 05c3b4b2b3dd8..c190938142960 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -949,7 +949,7 @@ int nfs2_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, error = decode_filename_inline(xdr, &entry->name, &entry->len); if (unlikely(error)) - return -EAGAIN; + return error == -ENAMETOOLONG ? -ENAMETOOLONG : -EAGAIN; /* * The type (size and byte order) of nfscookie isn't defined in diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 3b0b650c9c5ab..60f032be805ae 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -1991,7 +1991,7 @@ int nfs3_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, error = decode_inline_filename3(xdr, &entry->name, &entry->len); if (unlikely(error)) - return -EAGAIN; + return error == -ENAMETOOLONG ? -ENAMETOOLONG : -EAGAIN; error = decode_cookie3(xdr, &new_cookie); if (unlikely(error)) From 5690eed941ab7e33c3c3d6b850100cabf740f075 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Thu, 24 Aug 2023 16:43:53 -0400 Subject: [PATCH 181/186] NFSv4.2: fix handling of COPY ERR_OFFLOAD_NO_REQ If the client sent a synchronous copy and the server replied with ERR_OFFLOAD_NO_REQ indicating that it wants an asynchronous copy instead, the client should retry with asynchronous copy. Fixes: 539f57b3e0fd ("NFS handle COPY ERR_OFFLOAD_NO_REQS") Signed-off-by: Olga Kornievskaia Signed-off-by: Anna Schumaker --- fs/nfs/nfs42proc.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c index 49f78e23b34c0..063e00aff87ed 100644 --- a/fs/nfs/nfs42proc.c +++ b/fs/nfs/nfs42proc.c @@ -471,8 +471,9 @@ ssize_t nfs42_proc_copy(struct file *src, loff_t pos_src, continue; } break; - } else if (err == -NFS4ERR_OFFLOAD_NO_REQS && !args.sync) { - args.sync = true; + } else if (err == -NFS4ERR_OFFLOAD_NO_REQS && + args.sync != res.synchronous) { + args.sync = res.synchronous; dst_exception.retry = 1; continue; } else if ((err == -ESTALE || From c4a123d2e8c4dc91d581ee7d05c0cd51a0273fab Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Wed, 30 Aug 2023 14:31:31 -0400 Subject: [PATCH 182/186] pNFS: Fix assignment of xprtdata.cred The comma at the end of the line was leftover from an earlier refactor of the _nfs4_pnfs_v3_ds_connect() function. This is technically valid C, so the compilers didn't catch it, but if I'm understanding how it works correctly it assigns the return value of rpc_clnt_add_xprtr() to xprtdata.cred. Reported-by: Olga Kornievskaia Fixes: a12f996d3413 ("NFSv4/pNFS: Use connections to a DS that are all of the same protocol family") Signed-off-by: Anna Schumaker --- fs/nfs/pnfs_nfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c index a08cfda6fff1f..afd23910f3bff 100644 --- a/fs/nfs/pnfs_nfs.c +++ b/fs/nfs/pnfs_nfs.c @@ -946,7 +946,7 @@ static int _nfs4_pnfs_v4_ds_connect(struct nfs_server *mds_srv, * Test this address for session trunking and * add as an alias */ - xprtdata.cred = nfs4_get_clid_cred(clp), + xprtdata.cred = nfs4_get_clid_cred(clp); rpc_clnt_add_xprt(clp->cl_rpcclient, &xprt_args, rpc_clnt_setup_test_and_add_xprt, &rpcdata); From 69881be3d9a00cca770886af40913cfc5274b2d0 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Tue, 29 Aug 2023 17:23:56 +0200 Subject: [PATCH 183/186] fs: export sget_dev() They will be used for mtd devices as well. Acked-by: Richard Weinberger Reviewed-by: Jan Kara Message-Id: <20230829-vfs-super-mtd-v1-1-fecb572e5df3@kernel.org> Signed-off-by: Christian Brauner Reviewed-by: Christoph Hellwig Signed-off-by: Christian Brauner --- fs/super.c | 64 ++++++++++++++++++++++++++++++++-------------- include/linux/fs.h | 1 + 2 files changed, 46 insertions(+), 19 deletions(-) diff --git a/fs/super.c b/fs/super.c index bd8dcfc822c39..2d762ce67f6e6 100644 --- a/fs/super.c +++ b/fs/super.c @@ -1373,6 +1373,50 @@ int get_tree_keyed(struct fs_context *fc, } EXPORT_SYMBOL(get_tree_keyed); +static int set_bdev_super(struct super_block *s, void *data) +{ + s->s_dev = *(dev_t *)data; + return 0; +} + +static int super_s_dev_set(struct super_block *s, struct fs_context *fc) +{ + return set_bdev_super(s, fc->sget_key); +} + +static int super_s_dev_test(struct super_block *s, struct fs_context *fc) +{ + return !(s->s_iflags & SB_I_RETIRED) && + s->s_dev == *(dev_t *)fc->sget_key; +} + +/** + * sget_dev - Find or create a superblock by device number + * @fc: Filesystem context. + * @dev: device number + * + * Find or create a superblock using the provided device number that + * will be stored in fc->sget_key. + * + * If an extant superblock is matched, then that will be returned with + * an elevated reference count that the caller must transfer or discard. + * + * If no match is made, a new superblock will be allocated and basic + * initialisation will be performed (s_type, s_fs_info, s_id, s_dev will + * be set). The superblock will be published and it will be returned in + * a partially constructed state with SB_BORN and SB_ACTIVE as yet + * unset. + * + * Return: an existing or newly created superblock on success, an error + * pointer on failure. + */ +struct super_block *sget_dev(struct fs_context *fc, dev_t dev) +{ + fc->sget_key = &dev; + return sget_fc(fc, super_s_dev_test, super_s_dev_set); +} +EXPORT_SYMBOL(sget_dev); + #ifdef CONFIG_BLOCK /* * Lock a super block that the callers holds a reference to. @@ -1431,23 +1475,6 @@ const struct blk_holder_ops fs_holder_ops = { }; EXPORT_SYMBOL_GPL(fs_holder_ops); -static int set_bdev_super(struct super_block *s, void *data) -{ - s->s_dev = *(dev_t *)data; - return 0; -} - -static int set_bdev_super_fc(struct super_block *s, struct fs_context *fc) -{ - return set_bdev_super(s, fc->sget_key); -} - -static int test_bdev_super_fc(struct super_block *s, struct fs_context *fc) -{ - return !(s->s_iflags & SB_I_RETIRED) && - s->s_dev == *(dev_t *)fc->sget_key; -} - int setup_bdev_super(struct super_block *sb, int sb_flags, struct fs_context *fc) { @@ -1525,8 +1552,7 @@ int get_tree_bdev(struct fs_context *fc, } fc->sb_flags |= SB_NOSEC; - fc->sget_key = &dev; - s = sget_fc(fc, test_bdev_super_fc, set_bdev_super_fc); + s = sget_dev(fc, dev); if (IS_ERR(s)) return PTR_ERR(s); diff --git a/include/linux/fs.h b/include/linux/fs.h index c8ff4156a0a15..4aeb3fa119277 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2397,6 +2397,7 @@ struct super_block *sget(struct file_system_type *type, int (*test)(struct super_block *,void *), int (*set)(struct super_block *,void *), int flags, void *data); +struct super_block *sget_dev(struct fs_context *fc, dev_t dev); /* Alas, no aliases. Too much hassle with bringing module.h everywhere */ #define fops_get(fops) \ From ec952aa253c0f49a70d9de7b44b5f5c93e2dfe54 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Tue, 29 Aug 2023 17:23:57 +0200 Subject: [PATCH 184/186] mtd: key superblock by device number The mtd driver has similar problems than the one that was fixed in commit dc3216b14160 ("super: ensure valid info"). The kill_mtd_super() helper calls shuts the superblock down but leaves the superblock on fs_supers as the devices are still in use but puts the mtd device and cleans out the superblock's s_mtd field. This means another mounter can find the superblock on the list accessing its s_mtd field while it is curently in the process of being freed or already freed. Prevent that from happening by keying superblock by dev_t just as we do in the generic code. Link: https://lore.kernel.org/linux-fsdevel/20230829-weitab-lauwarm-49c40fc85863@brauner Acked-by: Richard Weinberger Reviewed-by: Jan Kara Reviewed-by: Christoph Hellwig Message-Id: <20230829-vfs-super-mtd-v1-2-fecb572e5df3@kernel.org> Signed-off-by: Christian Brauner --- drivers/mtd/mtdsuper.c | 45 +++++++++++------------------------------- 1 file changed, 11 insertions(+), 34 deletions(-) diff --git a/drivers/mtd/mtdsuper.c b/drivers/mtd/mtdsuper.c index 5ff001140ef4a..b7e3763c47f0c 100644 --- a/drivers/mtd/mtdsuper.c +++ b/drivers/mtd/mtdsuper.c @@ -19,38 +19,6 @@ #include #include "mtdcore.h" -/* - * compare superblocks to see if they're equivalent - * - they are if the underlying MTD device is the same - */ -static int mtd_test_super(struct super_block *sb, struct fs_context *fc) -{ - struct mtd_info *mtd = fc->sget_key; - - if (sb->s_mtd == fc->sget_key) { - pr_debug("MTDSB: Match on device %d (\"%s\")\n", - mtd->index, mtd->name); - return 1; - } - - pr_debug("MTDSB: No match, device %d (\"%s\"), device %d (\"%s\")\n", - sb->s_mtd->index, sb->s_mtd->name, mtd->index, mtd->name); - return 0; -} - -/* - * mark the superblock by the MTD device it is using - * - set the device number to be the correct MTD block device for pesuperstence - * of NFS exports - */ -static int mtd_set_super(struct super_block *sb, struct fs_context *fc) -{ - sb->s_mtd = fc->sget_key; - sb->s_dev = MKDEV(MTD_BLOCK_MAJOR, sb->s_mtd->index); - sb->s_bdi = bdi_get(mtd_bdi); - return 0; -} - /* * get a superblock on an MTD-backed filesystem */ @@ -62,8 +30,7 @@ static int mtd_get_sb(struct fs_context *fc, struct super_block *sb; int ret; - fc->sget_key = mtd; - sb = sget_fc(fc, mtd_test_super, mtd_set_super); + sb = sget_dev(fc, MKDEV(MTD_BLOCK_MAJOR, mtd->index)); if (IS_ERR(sb)) return PTR_ERR(sb); @@ -77,6 +44,16 @@ static int mtd_get_sb(struct fs_context *fc, pr_debug("MTDSB: New superblock for device %d (\"%s\")\n", mtd->index, mtd->name); + /* + * Would usually have been set with @sb_lock held but in + * contrast to sb->s_bdev that's checked with only + * @sb_lock held, nothing checks sb->s_mtd without also + * holding sb->s_umount and we're holding sb->s_umount + * here. + */ + sb->s_mtd = mtd; + sb->s_bdi = bdi_get(mtd_bdi); + ret = fill_super(sb, fc); if (ret < 0) goto error_sb; From 5069ba84b5e67873a2dfa4bf73a24506950fa1bf Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 31 Aug 2023 07:29:40 +0200 Subject: [PATCH 185/186] NFS: switch back to using kill_anon_super NFS switch to open coding kill_anon_super in 7b14a213890a ("nfs: don't call bdi_unregister") to avoid the extra bdi_unregister call. At that point bdi_destroy was called in nfs_free_server and thus it required a later freeing of the anon dev_t. But since 0db10944a76b ("nfs: Convert to separately allocated bdi") the bdi has been free implicitly by the sb destruction, so this isn't needed anymore. By not open coding kill_anon_super, nfs now inherits the fix in dc3216b14160 ("super: ensure valid info"), and we remove the only open coded version of kill_anon_super. Signed-off-by: Christoph Hellwig Reviewed-by: Jan Kara Message-Id: <20230831052940.256193-1-hch@lst.de> Signed-off-by: Christian Brauner --- fs/nfs/super.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 2284f749d8924..0d6473cb00cb3 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -1339,15 +1339,13 @@ int nfs_get_tree_common(struct fs_context *fc) void nfs_kill_super(struct super_block *s) { struct nfs_server *server = NFS_SB(s); - dev_t dev = s->s_dev; nfs_sysfs_move_sb_to_server(server); - generic_shutdown_super(s); + kill_anon_super(s); nfs_fscache_release_super_cookie(s); nfs_free_server(server); - free_anon_bdev(dev); } EXPORT_SYMBOL_GPL(nfs_kill_super); From f441ff73f1ec568acef03f0ce4d5088c7e65c106 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Thu, 31 Aug 2023 23:27:31 +0200 Subject: [PATCH 186/186] powerpc: Fix pud_mkwrite() definition after pte_mkwrite() API changes Fix up missed semantic mis-merge between commits 161e393c0f63 ("mm: Make pte_mkwrite() take a VMA") 27af67f35631 ("powerpc/book3s64/mm: enable transparent pud hugepage") where the newly introduced powerpc use of 'pte_mkwrite()' needs to use the 'novma()' versions as per commit 2f0584f3f4bd ("mm: Rename arch pte_mkwrite()'s to pte_mkwrite_novma()"). Fixes: df57721f9a63 ("Merge tag 'x86_shstk_for_6.6-rc1' of [...]") Signed-off-by: Ingo Molnar Signed-off-by: Linus Torvalds --- arch/powerpc/include/asm/book3s/64/pgtable.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h index 136232a897392..5c497c862d757 100644 --- a/arch/powerpc/include/asm/book3s/64/pgtable.h +++ b/arch/powerpc/include/asm/book3s/64/pgtable.h @@ -931,7 +931,7 @@ static inline pte_t *pudp_ptep(pud_t *pud) #define pud_mkdirty(pud) pte_pud(pte_mkdirty(pud_pte(pud))) #define pud_mkclean(pud) pte_pud(pte_mkclean(pud_pte(pud))) #define pud_mkyoung(pud) pte_pud(pte_mkyoung(pud_pte(pud))) -#define pud_mkwrite(pud) pte_pud(pte_mkwrite(pud_pte(pud))) +#define pud_mkwrite(pud) pte_pud(pte_mkwrite_novma(pud_pte(pud))) #define pud_write(pud) pte_write(pud_pte(pud)) #ifdef CONFIG_HAVE_ARCH_SOFT_DIRTY