FAILED: patch "[PATCH] io_uring: fix CQ waiting timeout handling" failed to apply to 5.15-stable tree

The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

Possible dependencies:

12521a5d5cb7 ("io_uring: fix CQ waiting timeout handling") 35d90f95cfa7 ("io_uring: include task_work run after scheduling in wait for events") 3a08576b96e3 ("io_uring: remove check_cq checking from hot paths") ed29b0b4fd83 ("io_uring: move to separate directory")
155bc9505dbd ("io_uring: return an error when cqe is dropped") 10988a0a67ba ("io_uring: use constants for cq_overflow bitfield") 3e813c902672 ("io_uring: rework io_uring_enter to simplify return value") cef216fc32d7 ("io_uring: explicitly keep a CQE in io_kiocb") b4f20bb4e6d5 ("io_uring: move finish_wait() outside of loop in cqring_wait()") d487b43cd327 ("io_uring: optimise mutex locking for submit+iopoll") 773697b610bf ("io_uring: pre-calculate syscall iopolling decision") f81440d33cc6 ("io_uring: split off IOPOLL argument verifiction") b605a7fabb60 ("io_uring: move poll recycling later in compl flushing") a538be5be328 ("io_uring: optimise io_free_batch_list")
c0713540f6d5 ("io_uring: fix leaks on IOPOLL and CQE_SKIP") 323b190ba2de ("io_uring: free iovec if file assignment fails") 7179c3ce3dbf ("io_uring: fix poll error reporting")
cce64ef01308 ("io_uring: fix poll file assign deadlock")
82733d168cbd ("io_uring: stop using io_wq_work as an fd placeholder") 2804ecd8d3e3 ("io_uring: move apoll->events cache")

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 12521a5d5cb7ff0ad43eadfc9c135d86e1131fa8 Mon Sep 17 00:00:00 2001
From: Pavel Begunkov <[email protected]>
Date: Thu, 5 Jan 2023 10:49:15 +0000
Subject: [PATCH] io_uring: fix CQ waiting timeout handling

Jiffy to ktime CQ waiting conversion broke how we treat timeouts, in particular we rearm it anew every time we get into io_cqring_wait_schedule() without adjusting the timeout. Waiting for 2 CQEs and getting a task_work in the middle may double the timeout value, or even worse in some cases task may wait indefinitely.

index 472574192dd6..2ac1cd8d23ea 100644
@@ -2470,7 +2470,7 @@ int io_run_task_work_sig(struct io_ring_ctx *ctx)
 /* when returns >0, the caller should retry */
 static inline int io_cqring_wait_schedule(struct io_ring_ctx *ctx,
 					  struct io_wait_queue *iowq,
-					  ktime_t timeout)
+					  ktime_t *timeout)
 {
 	int ret;
 	unsigned long check_cq;

@@ -2488,7 +2488,7 @@ static inline int io_cqring_wait_schedule(struct io_ring_ctx *ctx,
 		if (check_cq & BIT(IO_CHECK_CQ_DROPPED_BIT))
 			return -EBADR;
 	}
-	if (!schedule_hrtimeout(&timeout, HRTIMER_MODE_ABS))
+	if (!schedule_hrtimeout(timeout, HRTIMER_MODE_ABS))
 		return -ETIME;
 
 	/*

@@ -2564,7 +2564,7 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
 		}
 		prepare_to_wait_exclusive(&ctx->cq_wait, &iowq.wq,
 						TASK_INTERRUPTIBLE);
-		ret = io_cqring_wait_schedule(ctx, &iowq, timeout);
+		ret = io_cqring_wait_schedule(ctx, &iowq, &timeout);
 		if (__io_cqring_events_user(ctx) >= min_events)
 			break;
 		cond_resched();

Nevermind, I fixed this up for all 3 trees.

greg k-h

Awesome, thanks Greg!

Pavel Begunkov