I don't understand how we're getting two instances of this crash in version 1:11.1-1ubuntu7.3.
It sounds like either someone built their own package by the same name, or errors.ubuntu.com is not grouping the crashes correctly.
I say this because the fix does:
+++ pulseaudio/src/modules/alsa/module-alsa-card.c
@@ -625,7 +625,8 @@ static pa_hook_result_t card_suspend_changed
if (card->suspend_cause == 0) {
/* We were unsuspended, update jack state in case it changed while we were suspended */ PA_HASHMAP_FOREACH(jack, u->jacks, state) {
- report_jack_state(jack->melem, 0);
+ if (jack->melem)
+ report_jack_state(jack->melem, 0);
}
}
Yet errors.ubuntu.com is saying it is still crashing due to a NULL melem:
I think most likely errors.ubuntu.com is grouping similar assertions in mixer.c:929 incorrectly. And it only shows the full symbolic stack trace from some older instance of the crash -- not the ones from version 1:11.1-1ubuntu7.3.
I don't understand how we're getting two instances of this crash in version 1:11.1-1ubuntu7.3.
It sounds like either someone built their own package by the same name, or errors.ubuntu.com is not grouping the crashes correctly.
I say this because the fix does:
+++ pulseaudio/ src/modules/ alsa/module- alsa-card. c changed suspend_ cause == 0) {
PA_HASHMAP_ FOREACH( jack, u->jacks, state) { jack_state( jack->melem, 0); jack_state( jack->melem, 0);
@@ -625,7 +625,8 @@ static pa_hook_result_t card_suspend_
if (card->
/* We were unsuspended, update jack state in case it changed while we were suspended */
- report_
+ if (jack->melem)
+ report_
}
}
Yet errors.ubuntu.com is saying it is still crashing due to a NULL melem:
report_jack_state (melem=0x0, mask=mask@entry=0) suspend_ changed
card_
That NULL should be impossible with the patch.
I think most likely errors.ubuntu.com is grouping similar assertions in mixer.c:929 incorrectly. And it only shows the full symbolic stack trace from some older instance of the crash -- not the ones from version 1:11.1-1ubuntu7.3.