github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-11-07 06:17:02 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph

io_uring/notif: count pending zerocopy notifications per ring

Browse Source 
SEND_ZC notifications are io_kiocbs like any other as far as request
and task accounting goes, but they are special in that they cannot be
canceled and only complete once the network stack is done with the
data. With an unresponsive peer, that can take a while.

The exit and release paths are about to start waiting for requests to
finish so that the files they pin get released in a timely manner, and
notifications neither pin files nor finish in a timely manner. Keep a
per-ring count of them so those paths can leave them out.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
This commit is contained in:
Jens Axboe 2026-09-08 17:03:12 +00:00
parent 8d1c3a1f2f
commit fffb79350f
2 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
include/linux/io_uring_types.h
View File

@ -522,6 +522,8 @@ struct io_ring_ctx {
/* protected by ->completion_lock */
unsigned nr_req_allocated;
/* pending SEND_ZC notifications, protected by ->uring_lock */
unsigned nr_notifs;
#ifdef CONFIG_NET_RX_BUSY_POLL
struct list_head napi_list; /* track busy poll napi_id */

2
io_uring/notif.c
View File

@ -36,6 +36,7 @@ static void io_notif_tw_complete(struct io_tw_req tw_req, io_tw_token_t tw)
}
nd = nd->next;
ctx->nr_notifs--;
io_req_task_complete((struct io_tw_req){notif}, tw);
} while (nd);
}
@ -119,6 +120,7 @@ struct io_kiocb *io_alloc_notif(struct io_ring_ctx *ctx)
if (unlikely(!io_alloc_req(ctx, &notif)))
return NULL;
ctx->nr_notifs++;
notif->ctx = ctx;
notif->opcode = IORING_OP_NOP;
notif->flags = 0;