All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andrew Morton <akpm@linux-foundation.org>
To: akpm@linux-foundation.org, jbaron@akamai.com, khazhy@google.com,
	linux-mm@kvack.org, mm-commits@vger.kernel.org, r@hev.cc,
	rpenyaev@suse.de, stable@vger.kernel.org,
	torvalds@linux-foundation.org, viro@zeniv.linux.org.uk
Subject: [patch 07/15] eventpoll: fix missing wakeup for ovflist in ep_poll_callback
Date: Thu, 07 May 2020 18:35:59 -0700	[thread overview]
Message-ID: <20200508013559.g5WZPjEJv%akpm@linux-foundation.org> (raw)
In-Reply-To: <20200507183509.c5ef146c5aaeb118a25a39a8@linux-foundation.org>

From: Khazhismel Kumykov <khazhy@google.com>
Subject: eventpoll: fix missing wakeup for ovflist in ep_poll_callback

In the event that we add to ovflist, before 339ddb53d373 we would be woken
up by ep_scan_ready_list, and did no wakeup in ep_poll_callback.  With
that wakeup removed, if we add to ovflist here, we may never wake up. 
Rather than adding back the ep_scan_ready_list wakeup - which was
resulting in unnecessary wakeups, trigger a wake-up in ep_poll_callback.

We noticed that one of our workloads was missing wakeups starting with
339ddb53d373 and upon manual inspection, this wakeup seemed missing to me.
With this patch added, we no longer see missing wakeups.  I haven't yet
tried to make a small reproducer, but the existing kselftests in
filesystem/epoll passed for me with this patch.

[khazhy@google.com: use if/elif instead of goto + cleanup suggested by Roman]
  Link: http://lkml.kernel.org/r/20200424190039.192373-1-khazhy@google.com
Link: http://lkml.kernel.org/r/20200424025057.118641-1-khazhy@google.com
Fixes: 339ddb53d373 ("fs/epoll: remove unnecessary wakeups of nested epoll")
Signed-off-by: Khazhismel Kumykov <khazhy@google.com>
Reviewed-by: Roman Penyaev <rpenyaev@suse.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Roman Penyaev <rpenyaev@suse.de>
Cc: Heiher <r@hev.cc>
Cc: Jason Baron <jbaron@akamai.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/eventpoll.c |   18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

--- a/fs/eventpoll.c~eventpoll-fix-missing-wakeup-for-ovflist-in-ep_poll_callback
+++ a/fs/eventpoll.c
@@ -1171,6 +1171,10 @@ static inline bool chain_epi_lockless(st
 {
 	struct eventpoll *ep = epi->ep;
 
+	/* Fast preliminary check */
+	if (epi->next != EP_UNACTIVE_PTR)
+		return false;
+
 	/* Check that the same epi has not been just chained from another CPU */
 	if (cmpxchg(&epi->next, EP_UNACTIVE_PTR, NULL) != EP_UNACTIVE_PTR)
 		return false;
@@ -1237,16 +1241,12 @@ static int ep_poll_callback(wait_queue_e
 	 * chained in ep->ovflist and requeued later on.
 	 */
 	if (READ_ONCE(ep->ovflist) != EP_UNACTIVE_PTR) {
-		if (epi->next == EP_UNACTIVE_PTR &&
-		    chain_epi_lockless(epi))
+		if (chain_epi_lockless(epi))
+			ep_pm_stay_awake_rcu(epi);
+	} else if (!ep_is_linked(epi)) {
+		/* In the usual case, add event to ready list. */
+		if (list_add_tail_lockless(&epi->rdllink, &ep->rdllist))
 			ep_pm_stay_awake_rcu(epi);
-		goto out_unlock;
-	}

WARNING: multiple messages have this Message-ID (diff)
From: Andrew Morton <akpm@linux-foundation.org>
To: akpm@linux-foundation.org, jbaron@akamai.com, khazhy@google.com,
	linux-mm@kvack.org, mm-commits@vger.kernel.org, r@hev.cc,
	rpenyaev@suse.de, stable@vger.kernel.org,
	torvalds@linux-foundation.org, viro@zeniv.linux.org.uk
Subject: [patch 07/15] eventpoll: fix missing wakeup for ovflist in ep_poll_callback
Date: Thu, 07 May 2020 18:35:59 -0700	[thread overview]
Message-ID: <20200508013559.g5WZPjEJv%akpm@linux-foundation.org> (raw)
In-Reply-To: <20200507183509.c5ef146c5aaeb118a25a39a8@linux-foundation.org>

From: Khazhismel Kumykov <khazhy@google.com>
Subject: eventpoll: fix missing wakeup for ovflist in ep_poll_callback

In the event that we add to ovflist, before 339ddb53d373 we would be woken
up by ep_scan_ready_list, and did no wakeup in ep_poll_callback.  With
that wakeup removed, if we add to ovflist here, we may never wake up. 
Rather than adding back the ep_scan_ready_list wakeup - which was
resulting in unnecessary wakeups, trigger a wake-up in ep_poll_callback.

We noticed that one of our workloads was missing wakeups starting with
339ddb53d373 and upon manual inspection, this wakeup seemed missing to me.
With this patch added, we no longer see missing wakeups.  I haven't yet
tried to make a small reproducer, but the existing kselftests in
filesystem/epoll passed for me with this patch.

[khazhy@google.com: use if/elif instead of goto + cleanup suggested by Roman]
  Link: http://lkml.kernel.org/r/20200424190039.192373-1-khazhy@google.com
Link: http://lkml.kernel.org/r/20200424025057.118641-1-khazhy@google.com
Fixes: 339ddb53d373 ("fs/epoll: remove unnecessary wakeups of nested epoll")
Signed-off-by: Khazhismel Kumykov <khazhy@google.com>
Reviewed-by: Roman Penyaev <rpenyaev@suse.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Roman Penyaev <rpenyaev@suse.de>
Cc: Heiher <r@hev.cc>
Cc: Jason Baron <jbaron@akamai.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/eventpoll.c |   18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

--- a/fs/eventpoll.c~eventpoll-fix-missing-wakeup-for-ovflist-in-ep_poll_callback
+++ a/fs/eventpoll.c
@@ -1171,6 +1171,10 @@ static inline bool chain_epi_lockless(st
 {
 	struct eventpoll *ep = epi->ep;
 
+	/* Fast preliminary check */
+	if (epi->next != EP_UNACTIVE_PTR)
+		return false;
+
 	/* Check that the same epi has not been just chained from another CPU */
 	if (cmpxchg(&epi->next, EP_UNACTIVE_PTR, NULL) != EP_UNACTIVE_PTR)
 		return false;
@@ -1237,16 +1241,12 @@ static int ep_poll_callback(wait_queue_e
 	 * chained in ep->ovflist and requeued later on.
 	 */
 	if (READ_ONCE(ep->ovflist) != EP_UNACTIVE_PTR) {
-		if (epi->next == EP_UNACTIVE_PTR &&
-		    chain_epi_lockless(epi))
+		if (chain_epi_lockless(epi))
+			ep_pm_stay_awake_rcu(epi);
+	} else if (!ep_is_linked(epi)) {
+		/* In the usual case, add event to ready list. */
+		if (list_add_tail_lockless(&epi->rdllink, &ep->rdllist))
 			ep_pm_stay_awake_rcu(epi);
-		goto out_unlock;
-	}
-
-	/* If this file is already in the ready list we exit soon */
-	if (!ep_is_linked(epi) &&
-	    list_add_tail_lockless(&epi->rdllink, &ep->rdllist)) {
-		ep_pm_stay_awake_rcu(epi);
 	}
 
 	/*
_

  parent reply	other threads:[~2020-05-08  1:36 UTC|newest]

Thread overview: 99+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-08  1:35 incoming Andrew Morton
2020-05-08  1:35 ` [patch 01/15] ipc/mqueue.c: change __do_notify() to bypass check_kill_permission() Andrew Morton
2020-05-08  1:35   ` Andrew Morton
2020-05-09 12:30   ` Sasha Levin
2020-05-08  1:35 ` [patch 02/15] mm, memcg: fix error return value of mem_cgroup_css_alloc() Andrew Morton
2020-05-08  1:35 ` [patch 03/15] mm/page_alloc: fix watchdog soft lockups during set_zone_contiguous() Andrew Morton
2020-05-08  1:35 ` [patch 04/15] kernel/kcov.c: fix typos in kcov_remote_start documentation Andrew Morton
2020-05-08  1:35 ` [patch 05/15] scripts/decodecode: fix trapping instruction formatting Andrew Morton
2020-05-08  1:35 ` [patch 06/15] arch/x86/kvm/svm/sev.c: change flag passed to GUP fast in sev_pin_memory() Andrew Morton
2020-05-08  1:35 ` Andrew Morton [this message]
2020-05-08  1:35   ` [patch 07/15] eventpoll: fix missing wakeup for ovflist in ep_poll_callback Andrew Morton
2020-05-08  1:36 ` [patch 08/15] scripts/gdb: repair rb_first() and rb_last() Andrew Morton
2020-05-08  1:36 ` [patch 09/15] mm/slub: fix incorrect interpretation of s->offset Andrew Morton
2020-05-08  1:36 ` [patch 10/15] percpu: make pcpu_alloc() aware of current gfp context Andrew Morton
2020-05-08  1:36 ` [patch 11/15] kselftests: introduce new epoll60 testcase for catching lost wakeups Andrew Morton
2020-05-08  1:36 ` [patch 12/15] epoll: atomically remove wait entry on wake up Andrew Morton
2020-05-08  1:36   ` Andrew Morton
2020-05-08  1:36 ` [patch 13/15] mm/vmscan: remove unnecessary argument description of isolate_lru_pages() Andrew Morton
2020-05-08  1:36 ` [patch 14/15] ubsan: disable UBSAN_ALIGNMENT under COMPILE_TEST Andrew Morton
2020-05-08  1:36   ` Andrew Morton
2020-05-08  1:36 ` [patch 15/15] mm: limit boost_watermark on small zones Andrew Morton
2020-05-08  2:08 ` + arch-kunmap-remove-duplicate-kunmap-implementations-fix.patch added to -mm tree Andrew Morton
2020-05-08 20:46 ` + ipc-utilc-sysvipc_find_ipc-incorrectly-updates-position-index-fix.patch " Andrew Morton
2020-05-08 20:50 ` + mm-introduce-external-memory-hinting-api-fix-2.patch " Andrew Morton
2020-05-08 21:32 ` + checkpatch-use-patch-subject-when-reading-from-stdin-fix.patch " Andrew Morton
2020-05-08 23:19 ` + mm-fix-numa-node-file-count-error-in-replace_page_cache.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-fix-stat-corrupting-race-in-charge-moving.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-drop-compound-parameter-from-memcg-charging-api.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-move-out-cgroup-swaprate-throttling.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-convert-page-cache-to-a-new-mem_cgroup_charge-api.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-prepare-uncharging-for-removal-of-private-page-type-counters.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-prepare-move_account-for-removal-of-private-page-type-counters.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-prepare-cgroup-vmstat-infrastructure-for-native-anon-counters.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-switch-to-native-nr_file_pages-and-nr_shmem-counters.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-switch-to-native-nr_anon_mapped-counter.patch " Andrew Morton
2020-05-08 23:19 ` + mm-memcontrol-switch-to-native-nr_anon_thps-counter.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-convert-anon-and-file-thp-to-new-mem_cgroup_charge-api.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-drop-unused-try-commit-cancel-charge-api.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-prepare-swap-controller-setup-for-integration.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-make-swap-tracking-an-integral-part-of-memory-control.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-charge-swapin-pages-on-instantiation.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-document-the-new-swap-control-behavior.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-delete-unused-lrucare-handling.patch " Andrew Morton
2020-05-08 23:20 ` + mm-memcontrol-update-page-mem_cgroup-stability-rules.patch " Andrew Morton
2020-05-08 23:38 ` + selftests-vm-pkeys-introduce-powerpc-support-fix.patch " Andrew Morton
2020-05-08 23:38 ` + selftests-vm-pkeys-override-access-right-definitions-on-powerpc-fix.patch " Andrew Morton
2020-05-08 23:40 ` + memcg-expose-root-cgroups-memorystat.patch " Andrew Morton
2020-05-08 23:43 ` + doc-cgroup-update-note-about-conditions-when-oom-killer-is-invoked.patch " Andrew Morton
2020-05-08 23:43 ` + doc-cgroup-update-note-about-conditions-when-oom-killer-is-invoked-fix.patch " Andrew Morton
2020-05-08 23:48 ` + zcomp-use-array_size-for-backends-list.patch " Andrew Morton
2020-05-08 23:53 ` + device-dax-dont-leak-kernel-memory-to-user-space-after-unloading-kmem.patch " Andrew Morton
2020-05-08 23:56 ` + mm-memory_hotplug-introduce-add_memory_driver_managed.patch " Andrew Morton
2020-05-08 23:56 ` + kexec_file-dont-place-kexec-images-on-ioresource_mem_driver_managed.patch " Andrew Morton
2020-05-08 23:56 ` + device-dax-add-memory-via-add_memory_driver_managed.patch " Andrew Morton
2020-05-09  0:45 ` + mm-memory_hotplug-allow-arch-override-of-non-boot-memory-resource-names-fix.patch " Andrew Morton
2020-05-11 19:08 ` + mm-reset-numa-stats-for-boot-pagesets-v3.patch " Andrew Morton
2020-05-11 19:54 ` + mm-shmem-remove-rare-optimization-when-swapin-races-with-hole-punching.patch " Andrew Morton
2020-05-11 20:31 ` [to-be-updated] kexec-prevent-removal-of-memory-in-use-by-a-loaded-kexec-image.patch removed from " Andrew Morton
2020-05-11 20:31 ` [to-be-updated] mm-memory_hotplug-allow-arch-override-of-non-boot-memory-resource-names.patch " Andrew Morton
2020-05-11 20:31 ` [to-be-updated] mm-memory_hotplug-allow-arch-override-of-non-boot-memory-resource-names-fix.patch " Andrew Morton
2020-05-11 20:31 ` [to-be-updated] arm64-memory-give-hotplug-memory-a-different-resource-name.patch " Andrew Morton
2020-05-11 20:38 ` + arm-add-support-for-folded-p4d-page-tables-fix.patch added to " Andrew Morton
2020-05-11 20:40 ` + mm-add-debug_wx-support-fix-2.patch " Andrew Morton
2020-05-11 20:41 ` + mm-add-debug_wx-support-fix-3.patch " Andrew Morton
2020-05-11 20:50 ` + arm64-mm-drop-__have_arch_huge_ptep_get.patch " Andrew Morton
2020-05-11 20:50 ` + mm-hugetlb-define-a-generic-fallback-for-is_hugepage_only_range.patch " Andrew Morton
2020-05-11 20:50 ` + mm-hugetlb-define-a-generic-fallback-for-arch_clear_hugepage_flags.patch " Andrew Morton
2020-05-11 21:00 ` + mm-introduce-external-memory-hinting-api-fix-2-fix.patch " Andrew Morton
2020-05-11 21:02 ` + mm-support-vector-address-ranges-for-process_madvise-fix-fix-fix-fix-fix.patch " Andrew Morton
2020-05-11 21:12 ` [alternative-merged] mm-vmscan-consistent-update-to-pgsteal-and-pgscan.patch removed from " Andrew Morton
2020-05-11 21:21 ` + seq_file-introduce-define_seq_attribute-helper-macro.patch added to " Andrew Morton
2020-05-11 21:21 ` + mm-vmstat-convert-to-use-define_seq_attribute-macro.patch " Andrew Morton
2020-05-11 21:21 ` + kernel-kprobes-convert-to-use-define_seq_attribute-macro.patch " Andrew Morton
2020-05-11 21:22 ` + seq_file-introduce-define_seq_attribute-helper-macro-checkpatch-fixes.patch " Andrew Morton
2020-05-11 21:23 ` + lib-flex_proportionsc-cleanup-__fprop_inc_percpu_max.patch " Andrew Morton
2020-05-11 21:31 ` [alternative-merged] lib-flex_proportionsc-aging-counts-when-fraction-smaller-than-max_frac-fprop_frac_base.patch removed from " Andrew Morton
2020-05-11 21:44 ` + xtensa-add-loglvl-to-show_trace-fix.patch added to " Andrew Morton
2020-05-11 22:44 ` mmotm 2020-05-11-15-43 uploaded Andrew Morton
2020-05-12  2:12   ` mmotm 2020-05-11-15-43 uploaded (ethernet/ti/ti_cpsw) Randy Dunlap
2020-05-13  9:20     ` Grygorii Strashko
2020-05-13 15:18       ` Randy Dunlap
2020-05-12  4:41   ` mmotm 2020-05-11-15-43 uploaded (mm/memcontrol.c, huge pages) Randy Dunlap
2020-05-12 12:17     ` Johannes Weiner
2020-05-12 15:27       ` Randy Dunlap
2020-05-12 15:37       ` Naresh Kamboju
2020-05-12 15:37         ` Naresh Kamboju
2020-05-12 17:16       ` Geert Uytterhoeven
2020-05-12 17:16         ` Geert Uytterhoeven
2020-05-12 15:11     ` Geert Uytterhoeven
2020-05-12 15:11       ` Geert Uytterhoeven
2020-05-12 21:03 ` + kasan-consistently-disable-debugging-features.patch added to -mm tree Andrew Morton
2020-05-12 21:03 ` + kasan-add-missing-functions-declarations-to-kasanh.patch " Andrew Morton
2020-05-12 21:04 ` + kasan-move-kasan_report-into-reportc.patch " Andrew Morton
2020-05-13 18:31 ` + mm-compaction-avoid-vm_bug_onpageslab-in-page_mapcount.patch " Andrew Morton
2020-05-13 20:56 ` + kernel-sysctl-ignore-out-of-range-taint-bits-introduced-via-kerneltainted.patch " Andrew Morton
2020-05-13 21:26 ` + vfs-keep-inodes-with-page-cache-off-the-inode-shrinker-lru.patch " Andrew Morton
2020-05-13 22:00 ` + mm-memcontrol-convert-anon-and-file-thp-to-new-mem_cgroup_charge-api-fix.patch " Andrew Morton
2020-05-13 22:48 ` + mm-swap-use-prandom_u32_max.patch " Andrew Morton
2020-05-13 22:54 ` + mm-memcontrol-switch-to-native-nr_anon_thps-counter-fix.patch " Andrew Morton

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=20200508013559.g5WZPjEJv%akpm@linux-foundation.org \
    --to=akpm@linux-foundation.org \
    --cc=jbaron@akamai.com \
    --cc=khazhy@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mm-commits@vger.kernel.org \
    --cc=r@hev.cc \
    --cc=rpenyaev@suse.de \
    --cc=stable@vger.kernel.org \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.