diff options
| author | Pavel Begunkov <asml.silence@gmail.com> | 2023-02-22 14:36:50 +0000 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2023-02-22 09:57:23 -0700 |
| commit | b000ae0ec2d709046ac1a3c5722fea417f8a067e (patch) | |
| tree | 91a0bf402c0a9b2eb5f5f9ba4471f3af3de956c6 /io_uring/rsrc.c | |
| parent | edd478269640b360c6f301f2baa04abdda563ef3 (diff) | |
io_uring/rsrc: optimise single entry advance
Iterating within the first bvec entry should be essentially free, but we
use iov_iter_advance() for that, which shows up in benchmark profiles
taking up to 0.5% of CPU. Replace it with a hand coded version.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'io_uring/rsrc.c')
| -rw-r--r-- | io_uring/rsrc.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c index 53845e496881..ebbd2cea7582 100644 --- a/io_uring/rsrc.c +++ b/io_uring/rsrc.c @@ -1364,7 +1364,10 @@ int io_import_fixed(int ddir, struct iov_iter *iter, const struct bio_vec *bvec = imu->bvec; if (offset <= bvec->bv_len) { - iov_iter_advance(iter, offset); + iter->bvec = bvec; + iter->nr_segs = bvec->bv_len; + iter->count -= offset; + iter->iov_offset = offset; } else { unsigned long seg_skip; |