You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Investigate whether SO_TIMESTAMPING will work for AMP, and if so, what is required to implement it.
Is SO_TIMESTAMPING available in popular kernels? Do we need particular hardware to make it work? Will it work with a variety of socket types (raw, tcp, udp)?
Can we replace the current packet timestamping calls that use gettimeofday() with this and retrieve actual sending times from the socket? Should we replace the receive timestamp code using SO_TIMESTAMP with this also?
Get the software timestamp for when packets were actually sent rather than
relying on a call to gettimeofday() immediately before calling sendto().
See #35.
Looks like udpstream and tcpping use gettimefoday() and sendto() instead of delay_send_packet(), if possible they should use the same approach as the other similar tests.
Investigate whether SO_TIMESTAMPING will work for AMP, and if so, what is required to implement it.
Is SO_TIMESTAMPING available in popular kernels? Do we need particular hardware to make it work? Will it work with a variety of socket types (raw, tcp, udp)?
Can we replace the current packet timestamping calls that use gettimeofday() with this and retrieve actual sending times from the socket? Should we replace the receive timestamp code using SO_TIMESTAMP with this also?
See: https://www.kernel.org/doc/Documentation/networking/timestamping.txt
The text was updated successfully, but these errors were encountered: