[WIP] Use the DpexTargetContext inside dpjit #1027
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
FYA @chudur-budur the changes in my PR will impact your SyclQueue work. I will merge my changes only after you are done. |
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
b6c2365
to
66772be
Compare
diptorupd
changed the title
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
d0b424d
to
49813a3
Compare
diptorupd
changed the base branch from
main
to
feature/remove_numpy_array_kernelargs
|
Documentation preview: show. |
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
49813a3
to
b1c9142
Compare
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
2 times, most recently
from
23aaf11
to
407ee35
Compare
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
3 times, most recently
from
4af7777
to
2219ad0
Compare
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
2219ad0
to
5c9db85
Compare
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
5c9db85
to
b4282fb
Compare
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
b4282fb
to
6fc374a
Compare
- Moved all dpnp overload and intrinsic functions
into the dpex target.
- The dpjit decorator now uses the DpexTargetContext
instead of the CPUTarget.
- All initialization of the DpjitDispatcher now do a
target_override to "dpex".
- The PreParfor pass can invoke NumPy overloads such as
numpy.dtype. The changes here introduced a dpex-specific
PreParfor pass that runs the upstream pass after doing a
target override to the dpex target.
diptorupd
force-pushed
the
feature/DpexTargetContext
branch
from
6fc374a
to
8c480da
Compare
Closed
5 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The PR is a draft for the set of changes to ensure that numba_dpex.dpjit uses its own target that is isolated from Numba's
CPUTarget. The goal of the PR is to ensure numba-dpex does not have unintended side effects on Numba, and also free numba-dpex to implement custom dispatch, caching, function registry support.
Have you added a test, reproducer or referred to an issue with a reproducer?
Have you tested your changes locally for CPU and GPU devices?
As yet minimally.
Have you made sure that new changes do not introduce compiler warnings?
If this PR is a work in progress, are you filing the PR as a draft?'
Fixes #1024