summaryrefslogtreecommitdiff
path: root/include
diff options
context:
space:
mode:
authorMing Lei <ming.lei@redhat.com>2026-01-16 15:46:38 +0800
committerJens Axboe <axboe@kernel.dk>2026-01-20 10:18:01 -0700
commitf7bc22ca0d55bdcb59e3a4a028fb811d23e53959 (patch)
treecc2e8ed9bad2d693a32f28b7aaf73754758188d2 /include
parent5e2fde1a9433efc484a5feec36f748aa3ea58c85 (diff)
nvme/io_uring: optimize IOPOLL completions for local ring context
When multiple io_uring rings poll on the same NVMe queue, one ring can find completions belonging to another ring. The current code always uses task_work to handle this, but this adds overhead for the common single-ring case. This patch passes the polling io_ring_ctx through io_comp_batch's new poll_ctx field. In io_do_iopoll(), the polling ring's context is stored in iob.poll_ctx before calling the iopoll callbacks. In nvme_uring_cmd_end_io(), we now compare iob->poll_ctx with the request's owning io_ring_ctx (via io_uring_cmd_ctx_handle()). If they match (local context), we complete inline with io_uring_cmd_done32(). If they differ (remote context) or iob is NULL (non-iopoll path), we use task_work as before. This optimization eliminates task_work scheduling overhead for the common case where a ring polls and finds its own completions. ~10% IOPS improvement is observed in the following benchmark: fio/t/io_uring -b512 -d128 -c32 -s32 -p1 -F1 -O0 -P1 -u1 -n1 /dev/ng0n1 Signed-off-by: Ming Lei <ming.lei@redhat.com> Reviewed-by: Kanchan Joshi <joshi.k@samsung.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'include')
-rw-r--r--include/linux/blkdev.h1
1 files changed, 1 insertions, 0 deletions
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 438c4946b6e5..251e0f538c4c 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -1822,6 +1822,7 @@ struct io_comp_batch {
struct rq_list req_list;
bool need_ts;
void (*complete)(struct io_comp_batch *);
+ void *poll_ctx;
};
static inline bool blk_atomic_write_start_sect_aligned(sector_t sector,