Linux-Fsdevel Archive mirror
 help / color / mirror / Atom feed
From: Soheil Hassas Yeganeh <soheil.kdev@gmail.com>
To: torvalds@linux-foundation.org, viro@zeniv.linux.org.uk,
	linux-fsdevel@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, akpm@linux-foundation.org,
	dave@stgolabs.net, edumazet@google.com, willemb@google.com,
	khazhy@google.com, guantaol@google.com,
	Soheil Hassas Yeganeh <soheil@google.com>
Subject: [PATCH 8/8] epoll: eliminate unnecessary lock for zero timeout
Date: Fri,  6 Nov 2020 18:16:35 -0500	[thread overview]
Message-ID: <20201106231635.3528496-9-soheil.kdev@gmail.com> (raw)
In-Reply-To: <20201106231635.3528496-1-soheil.kdev@gmail.com>

From: Soheil Hassas Yeganeh <soheil@google.com>

We call ep_events_available() under lock when timeout is 0,
and then call it without locks in the loop for the other cases.

Instead, call ep_events_available() without lock for all cases.
For non-zero timeouts, we will recheck after adding the thread to
the wait queue. For zero timeout cases, by definition, user is
opportunistically polling and will have to call epoll_wait again
in the future.

Note that this lock was kept in c5a282e9635e9 because the whole
loop was historically under lock.

This patch results in a 1% CPU/RPC reduction in RPC benchmarks.

Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Khazhismel Kumykov <khazhy@google.com>
---
 fs/eventpoll.c | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index f4e1be7ada26..1aa23b0be72b 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -1830,7 +1830,7 @@ static inline struct timespec64 ep_set_mstimeout(long ms)
 static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
 		   int maxevents, long timeout)
 {
-	int res, eavail = 0, timed_out = 0;
+	int res, eavail, timed_out = 0;
 	u64 slack = 0;
 	wait_queue_entry_t wait;
 	ktime_t expires, *to = NULL;
@@ -1846,18 +1846,21 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
 	} else if (timeout == 0) {
 		/*
 		 * Avoid the unnecessary trip to the wait queue loop, if the
-		 * caller specified a non blocking operation. We still need
-		 * lock because we could race and not see an epi being added
-		 * to the ready list while in irq callback. Thus incorrectly
-		 * returning 0 back to userspace.
+		 * caller specified a non blocking operation.
 		 */
 		timed_out = 1;
-
-		write_lock_irq(&ep->lock);
-		eavail = ep_events_available(ep);
-		write_unlock_irq(&ep->lock);
 	}
 
+	/*
+	 * This call is racy: We may or may not see events that are being added
+	 * to the ready list under the lock (e.g., in IRQ callbacks). For, cases
+	 * with a non-zero timeout, this thread will check the ready list under
+	 * lock and will added to the wait queue.  For, cases with a zero
+	 * timeout, the user by definition should not care and will have to
+	 * recheck again.
+	 */
+	eavail = ep_events_available(ep);
+
 	while (1) {
 		if (eavail) {
 			/*
@@ -1873,10 +1876,6 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
 		if (timed_out)
 			return 0;
 
-		eavail = ep_events_available(ep);
-		if (eavail)
-			continue;
-
 		eavail = ep_busy_loop(ep, timed_out);
 		if (eavail)
 			continue;
-- 
2.29.1.341.ge80a0c044ae-goog


  parent reply	other threads:[~2020-11-06 23:18 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-06 23:16 [PATCH 0/8] simplify ep_poll Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 1/8] epoll: check for events when removing a timed out thread from the wait queue Soheil Hassas Yeganeh
2020-11-10  6:05   ` Davidlohr Bueso
2020-11-06 23:16 ` [PATCH 2/8] epoll: simplify signal handling Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 3/8] epoll: pull fatal signal checks into ep_send_events() Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 4/8] epoll: move eavail next to the list_empty_careful check Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 5/8] epoll: simplify and optimize busy loop logic Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 6/8] epoll: pull all code between fetch_events and send_event into the loop Soheil Hassas Yeganeh
2020-11-06 23:16 ` [PATCH 7/8] epoll: replace gotos with a proper loop Soheil Hassas Yeganeh
2020-11-06 23:16 ` Soheil Hassas Yeganeh [this message]
2020-11-06 23:35 ` [PATCH 0/8] simplify ep_poll Linus Torvalds
2020-11-08  1:43 ` Andrew Morton
2020-11-08  4:45   ` Soheil Hassas Yeganeh
2020-11-09 18:59     ` Davidlohr Bueso
2020-11-09 19:28       ` Soheil Hassas Yeganeh
2020-11-10 22:05     ` Andrew Morton
2020-11-11  3:37       ` Soheil Hassas Yeganeh

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20201106231635.3528496-9-soheil.kdev@gmail.com \
    --to=soheil.kdev@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=dave@stgolabs.net \
    --cc=edumazet@google.com \
    --cc=guantaol@google.com \
    --cc=khazhy@google.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=soheil@google.com \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    --cc=willemb@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).