(In reply to Jens Stutte [:jstutte] from comment #16) > Looking at [one of the crashes](https://crash-stats.mozilla.org/report/index/965a417e-9026-4568-94cb-1782d0210921), I arrived at [`nsThread::ShutdownInternal`](https://searchfox.org/mozilla-central/rev/50c3cf7a3c931409b54efa009795b69c19383541/xpcom/threads/nsThread.cpp#781-782) where I read: > > ``` > // XXXroc What if posting the event fails due to OOM? > mEvents->PutEvent(event.forget(), EventQueuePriority::Normal); > ```... > It also seems that we [sequentially shut down the threads](https://searchfox.org/mozilla-central/rev/50c3cf7a3c931409b54efa009795b69c19383541/xpcom/threads/nsThreadManager.cpp#411-413), waiting for each thread to join before we shut down the next. We might want to decouple the sending of the shutdown event to all threads from the waiting for the threads to join. > > Randell, WDYT ? Magnus, any concerns about doing this? (In reply to Jens Stutte [:jstutte] from comment #27) > ... > I [made a try push](https://treeherder.mozilla.org/#/jobs?repo=try&revision=f88333d5ff484cbef1cc6b3e02c06dc1ccb7e1b2) with a direct return (and warning) for that case to see what happens. There are also [other crash reports](https://crash-stats.mozilla.org/report/index/e657d3cd-c25f-4d6d-a208-b03820211011) that indicate this might happen also during `profile-before-change` for individual thread shutdowns. I suspect no one in the bug here dead any testing (I know I didn't - sorry). Is it worth sending this out to a few community members or development team, and if so what do we want to ask of them?
Bug 1629669 Comment 28 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
(In reply to Jens Stutte [:jstutte] from comment #16) > Looking at [one of the crashes](https://crash-stats.mozilla.org/report/index/965a417e-9026-4568-94cb-1782d0210921), I arrived at [`nsThread::ShutdownInternal`](https://searchfox.org/mozilla-central/rev/50c3cf7a3c931409b54efa009795b69c19383541/xpcom/threads/nsThread.cpp#781-782) where I read: > > ``` > // XXXroc What if posting the event fails due to OOM? > mEvents->PutEvent(event.forget(), EventQueuePriority::Normal); > ```... > It also seems that we [sequentially shut down the threads](https://searchfox.org/mozilla-central/rev/50c3cf7a3c931409b54efa009795b69c19383541/xpcom/threads/nsThreadManager.cpp#411-413), waiting for each thread to join before we shut down the next. We might want to decouple the sending of the shutdown event to all threads from the waiting for the threads to join. > > Randell, WDYT ? Magnus, any concerns about doing this? (In reply to Jens Stutte [:jstutte] from comment #27) > ... > I [made a try push](https://treeherder.mozilla.org/#/jobs?repo=try&revision=f88333d5ff484cbef1cc6b3e02c06dc1ccb7e1b2) with a direct return (and warning) for that case to see what happens. There are also [other crash reports](https://crash-stats.mozilla.org/report/index/e657d3cd-c25f-4d6d-a208-b03820211011) that indicate this might happen also during `profile-before-change` for individual thread shutdowns. I suspect no one in the bug here did any testing (I know I didn't - sorry). Is it worth sending this out to a few community members or development team, and if so what do we want to ask of them?