diff options
author | Jens Axboe <axboe@kernel.dk> | 2024-02-27 02:43:01 +0300 |
---|---|---|
committer | Jens Axboe <axboe@kernel.dk> | 2024-02-27 21:16:00 +0300 |
commit | 792060de8b3e9ca11fab4afc0c3c5927186152a2 (patch) | |
tree | 0b7cfc28dea4b940da185fce48c15942c0b0720f /scripts/bpf_doc.py | |
parent | c55978024d123d43808ab393a0a4ce3ce8568150 (diff) | |
download | linux-792060de8b3e9ca11fab4afc0c3c5927186152a2.tar.xz |
io_uring/net: improve the usercopy for sendmsg/recvmsg
We're spending a considerable amount of the sendmsg/recvmsg time just
copying in the message header. And for provided buffers, the known
single entry iovec.
Be a bit smarter about it and enable/disable user access around our
copying. In a test case that does both sendmsg and recvmsg, the
runtime before this change (averaged over multiple runs, very stable
times however):
Kernel Time Diff
====================================
-git 4720 usec
-git+commit 4311 usec -8.7%
and looking at a profile diff, we see the following:
0.25% +9.33% [kernel.kallsyms] [k] _copy_from_user
4.47% -3.32% [kernel.kallsyms] [k] __io_msg_copy_hdr.constprop.0
where we drop more than 9% of _copy_from_user() time, and consequently
add time to __io_msg_copy_hdr() where the copies are now attributed to,
but with a net win of 6%.
In comparison, the same test case with send/recv runs in 3745 usec, which
is (expectedly) still quite a bit faster. But at least sendmsg/recvmsg is
now only ~13% slower, where it was ~21% slower before.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'scripts/bpf_doc.py')
0 files changed, 0 insertions, 0 deletions