musl: backport patches to fix pthread_cond_wait with PI mutex

https://github.com/pulseaudio/pulseaudio/blob/master/src/pulsecore/mutex-posix.c#L55
https://github.com/pulseaudio/pulseaudio/blob/master/src/pulse/thread-mainloop.c#L118

413a8f8917

pulseaudio uses a prio-inheriting mutex since the above revision
that is paised with a condition variable; the behavior of this
was broken in musl until the above patches, which would result
in pulseaudio deadlocking with gstreamer pulsesink (most often
with webkit) and possibly other things.

Fixes https://github.com/void-linux/void-packages/issues/15631
This commit is contained in:
q66 2021-08-02 22:05:48 +02:00
parent f482179eba
commit 53d00fffc2
4 changed files with 133 additions and 1 deletions

View File

@ -0,0 +1,56 @@
From 2d0bbe6c788938d1332609c014eeebc1dff966ac Mon Sep 17 00:00:00 2001
From: Rich Felker <dalias@aerifal.cx>
Date: Mon, 26 Oct 2020 15:56:25 -0400
Subject: fix pthread_cond_wait paired with with priority-inheritance mutex
pthread_cond_wait arranged for requeued waiters to wake when the mutex
is unlocked by temporarily adjusting the mutex's waiter count. commit
54ca677983d47529bab8752315ac1a2b49888870 broke this when introducing
PI mutexes by repurposing the waiter count field of the mutex
structure. since then, for PI mutexes, the waiter count adjustment was
misinterpreted by the mutex locking code as indicating that the mutex
is non a non-recoverable state.
it would be possible to special-case PI mutexes here, but instead just
drop all adjustment of the waiters count, and instead use the lock
word waiters bit for all mutex types. since the mutex is either held
by the caller or in unrecoverable state at the time the bit is set, it
will necessarily still be set at the time of any subsequent valid
unlock operation, and this will produce the desired effect of waking
the next waiter.
if waiter counts are entirely dropped at some point in the future this
code should still work without modification.
---
src/thread/pthread_cond_timedwait.c | 11 +++++------
1 file changed, 5 insertions(+), 6 deletions(-)
(limited to 'src/thread/pthread_cond_timedwait.c')
diff --git a/src/thread/pthread_cond_timedwait.c b/src/thread/pthread_cond_timedwait.c
index d1501240..f5f37af1 100644
--- a/src/thread/pthread_cond_timedwait.c
+++ b/src/thread/pthread_cond_timedwait.c
@@ -146,14 +146,13 @@ relock:
if (oldstate == WAITING) goto done;
- if (!node.next) a_inc(&m->_m_waiters);
-
/* Unlock the barrier that's holding back the next waiter, and
* either wake it or requeue it to the mutex. */
- if (node.prev)
- unlock_requeue(&node.prev->barrier, &m->_m_lock, m->_m_type & 128);
- else
- a_dec(&m->_m_waiters);
+ if (node.prev) {
+ int val = m->_m_lock;
+ if (val>0) a_cas(&m->_m_lock, val, val|0x80000000);
+ unlock_requeue(&node.prev->barrier, &m->_m_lock, m->_m_type & (8|128));
+ }
/* Since a signal was consumed, cancellation is not permitted. */
if (e == ECANCELED) e = 0;
--
cgit v1.2.1

View File

@ -0,0 +1,48 @@
From 27b2fc9d6db956359727a66c262f1e69995660aa Mon Sep 17 00:00:00 2001
From: Rich Felker <dalias@aerifal.cx>
Date: Fri, 30 Oct 2020 11:21:06 -0400
Subject: fix missing-wake regression in pthread_cond_wait
the reasoning in commit 2d0bbe6c788938d1332609c014eeebc1dff966ac was
not entirely correct. while it's true that setting the waiters flag
ensures that the next unlock will perform a wake, it's possible that
the wake is consumed by a mutex waiter that has no relationship with
the condvar wait queue being processed, which then takes the mutex.
when that thread subsequently unlocks, it sees no waiters, and leaves
the rest of the condvar queue stuck.
bring back the waiter count adjustment, but skip it for PI mutexes,
for which a successful lock-after-waiting always sets the waiters bit.
if future changes are made to bring this same waiters-bit contract to
all lock types, this can be reverted.
---
src/thread/pthread_cond_timedwait.c | 5 +++++
1 file changed, 5 insertions(+)
(limited to 'src/thread/pthread_cond_timedwait.c')
diff --git a/src/thread/pthread_cond_timedwait.c b/src/thread/pthread_cond_timedwait.c
index f5f37af1..a0cd4904 100644
--- a/src/thread/pthread_cond_timedwait.c
+++ b/src/thread/pthread_cond_timedwait.c
@@ -146,12 +146,17 @@ relock:
if (oldstate == WAITING) goto done;
+ if (!node.next && !(m->_m_type & 8))
+ a_inc(&m->_m_waiters);
+
/* Unlock the barrier that's holding back the next waiter, and
* either wake it or requeue it to the mutex. */
if (node.prev) {
int val = m->_m_lock;
if (val>0) a_cas(&m->_m_lock, val, val|0x80000000);
unlock_requeue(&node.prev->barrier, &m->_m_lock, m->_m_type & (8|128));
+ } else if (!!(m->_m_type & 8)) {
+ a_dec(&m->_m_waiters);
}
/* Since a signal was consumed, cancellation is not permitted. */
--
cgit v1.2.1

View File

@ -0,0 +1,28 @@
From d91a6cf6e369a79587c5665fce9635e5634ca201 Mon Sep 17 00:00:00 2001
From: Rich Felker <dalias@aerifal.cx>
Date: Fri, 30 Oct 2020 16:50:08 -0400
Subject: fix erroneous pthread_cond_wait mutex waiter count logic due to typo
introduced in commit 27b2fc9d6db956359727a66c262f1e69995660aa.
---
src/thread/pthread_cond_timedwait.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
(limited to 'src/thread/pthread_cond_timedwait.c')
diff --git a/src/thread/pthread_cond_timedwait.c b/src/thread/pthread_cond_timedwait.c
index a0cd4904..6b761455 100644
--- a/src/thread/pthread_cond_timedwait.c
+++ b/src/thread/pthread_cond_timedwait.c
@@ -155,7 +155,7 @@ relock:
int val = m->_m_lock;
if (val>0) a_cas(&m->_m_lock, val, val|0x80000000);
unlock_requeue(&node.prev->barrier, &m->_m_lock, m->_m_type & (8|128));
- } else if (!!(m->_m_type & 8)) {
+ } else if (!(m->_m_type & 8)) {
a_dec(&m->_m_waiters);
}
--
cgit v1.2.1

View File

@ -2,7 +2,7 @@
pkgname=musl
reverts="1.2.0_1"
version=1.1.24
revision=8
revision=9
archs="*-musl"
bootstrap=yes
build_style=gnu-configure