From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org,
"Peter Zijlstra (Intel)" <peterz@infradead.org>,
Song Liu <songliubraving@fb.com>, Sasha Levin <sashal@kernel.org>,
Lucian Grijincu <lucian@fb.com>
Subject: [PATCH 5.10 20/22] perf/core: fix userpage->time_enabled of inactive events
Date: Thu, 14 Oct 2021 16:54:26 +0200 [thread overview]
Message-ID: <20211014145208.627318874@linuxfoundation.org> (raw)
In-Reply-To: <20211014145207.979449962@linuxfoundation.org>
From: Song Liu <songliubraving@fb.com>
[ Upstream commit f792565326825ed806626da50c6f9a928f1079c1 ]
Users of rdpmc rely on the mmapped user page to calculate accurate
time_enabled. Currently, userpage->time_enabled is only updated when the
event is added to the pmu. As a result, inactive event (due to counter
multiplexing) does not have accurate userpage->time_enabled. This can
be reproduced with something like:
/* open 20 task perf_event "cycles", to create multiplexing */
fd = perf_event_open(); /* open task perf_event "cycles" */
userpage = mmap(fd); /* use mmap and rdmpc */
while (true) {
time_enabled_mmap = xxx; /* use logic in perf_event_mmap_page */
time_enabled_read = read(fd).time_enabled;
if (time_enabled_mmap > time_enabled_read)
BUG();
}
Fix this by updating userpage for inactive events in merge_sched_in.
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reported-and-tested-by: Lucian Grijincu <lucian@fb.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210929194313.2398474-1-songliubraving@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
include/linux/perf_event.h | 4 +++-
kernel/events/core.c | 34 ++++++++++++++++++++++++++++++----
2 files changed, 33 insertions(+), 5 deletions(-)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 072ac6c1ef2b..c095e713cf08 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -682,7 +682,9 @@ struct perf_event {
/*
* timestamp shadows the actual context timing but it can
* be safely used in NMI interrupt context. It reflects the
- * context time as it was when the event was last scheduled in.
+ * context time as it was when the event was last scheduled in,
+ * or when ctx_sched_in failed to schedule the event because we
+ * run out of PMC.
*
* ctx_time already accounts for ctx->timestamp. Therefore to
* compute ctx_time for a sample, simply add perf_clock().
diff --git a/kernel/events/core.c b/kernel/events/core.c
index c677f934353a..c81151926171 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -3695,6 +3695,29 @@ static noinline int visit_groups_merge(struct perf_cpu_context *cpuctx,
return 0;
}
+static inline bool event_update_userpage(struct perf_event *event)
+{
+ if (likely(!atomic_read(&event->mmap_count)))
+ return false;
+
+ perf_event_update_time(event);
+ perf_set_shadow_time(event, event->ctx);
+ perf_event_update_userpage(event);
+
+ return true;
+}
+
+static inline void group_update_userpage(struct perf_event *group_event)
+{
+ struct perf_event *event;
+
+ if (!event_update_userpage(group_event))
+ return;
+
+ for_each_sibling_event(event, group_event)
+ event_update_userpage(event);
+}
+
static int merge_sched_in(struct perf_event *event, void *data)
{
struct perf_event_context *ctx = event->ctx;
@@ -3713,14 +3736,15 @@ static int merge_sched_in(struct perf_event *event, void *data)
}
if (event->state == PERF_EVENT_STATE_INACTIVE) {
+ *can_add_hw = 0;
if (event->attr.pinned) {
perf_cgroup_event_disable(event, ctx);
perf_event_set_state(event, PERF_EVENT_STATE_ERROR);
+ } else {
+ ctx->rotate_necessary = 1;
+ perf_mux_hrtimer_restart(cpuctx);
+ group_update_userpage(event);
}
-
- *can_add_hw = 0;
- ctx->rotate_necessary = 1;
- perf_mux_hrtimer_restart(cpuctx);
}
return 0;
@@ -6239,6 +6263,8 @@ accounting:
ring_buffer_attach(event, rb);
+ perf_event_update_time(event);
+ perf_set_shadow_time(event, event->ctx);
perf_event_init_userpage(event);
perf_event_update_userpage(event);
} else {
--
2.33.0
next prev parent reply other threads:[~2021-10-14 15:02 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-14 14:54 [PATCH 5.10 00/22] 5.10.74-rc1 review Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 01/22] ext4: check and update i_disksize properly Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 02/22] ext4: correct the error path of ext4_write_inline_data_end() Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 03/22] ASoC: Intel: sof_sdw: tag SoundWire BEs as non-atomic Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 04/22] HID: apple: Fix logical maximum and usage maximum of Magic Keyboard JIS Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 05/22] netfilter: ip6_tables: zero-initialize fragment offset Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 06/22] HID: wacom: Add new Intuos BT (CTL-4100WL/CTL-6100WL) device IDs Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 07/22] ASoC: SOF: loader: release_firmware() on load failure to avoid batching Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 08/22] netfilter: nf_nat_masquerade: make async masq_inet6_event handling generic Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 09/22] netfilter: nf_nat_masquerade: defer conntrack walk to work queue Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 10/22] mac80211: Drop frames from invalid MAC address in ad-hoc mode Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 11/22] m68k: Handle arrivals of multiple signals correctly Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 12/22] hwmon: (ltc2947) Properly handle errors when looking for the external clock Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 13/22] net: prevent user from passing illegal stab size Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 14/22] mac80211: check return value of rhashtable_init Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 15/22] vboxfs: fix broken legacy mount signature checking Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 16/22] net: sun: SUNVNET_COMMON should depend on INET Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 17/22] drm/amdgpu: fix gart.bo pin_count leak Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 18/22] scsi: ses: Fix unsigned comparison with less than zero Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 19/22] scsi: virtio_scsi: Fix spelling mistake "Unsupport" -> "Unsupported" Greg Kroah-Hartman
2021-10-14 14:54 ` Greg Kroah-Hartman [this message]
2021-10-14 14:54 ` [PATCH 5.10 21/22] sched: Always inline is_percpu_thread() Greg Kroah-Hartman
2021-10-14 14:54 ` [PATCH 5.10 22/22] hwmon: (pmbus/ibm-cffps) max_power_out swap changes Greg Kroah-Hartman
2021-10-14 18:20 ` [PATCH 5.10 00/22] 5.10.74-rc1 review Fox Chen
2021-10-14 21:08 ` Pavel Machek
2021-10-14 22:04 ` Florian Fainelli
2021-10-14 22:38 ` Shuah Khan
2021-10-15 7:21 ` Samuel Zou
2021-10-15 14:30 ` Sudip Mukherjee
2021-10-15 16:28 ` Daniel Díaz
2021-10-15 22:07 ` Guenter Roeck
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=20211014145208.627318874@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lucian@fb.com \
--cc=peterz@infradead.org \
--cc=sashal@kernel.org \
--cc=songliubraving@fb.com \
--cc=stable@vger.kernel.org \
/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 NNTP newsgroup(s).