1033954 - [MTBF] crash at eglsubAndroid.so

Status: RESOLVED DUPLICATE of bug 1031355

(Core :: Graphics, defect)

Description

[Paul Yang [:pyang] (away)]

Crash Report: https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-1059d2140703
Device: flame
Gaia      9e9b52c0260863be9451e557767713103ecfed40
Gecko     https://hg.mozilla.org/releases/mozilla-aurora/rev/4851cd40200f
BuildID   20140630160201
Version   32.0a2
ro.build.version.incremental=96
ro.build.date=Fri May 23 11:17:40 CST 2014

Comment 1

[Sotaro Ikeda [:sotaro]]

(In reply to Paul Yang [: pyang] from comment #0)
> Crash Report:
> https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> 1059d2140703
> Device: flame

Crash thread stack do not have useful information. Only information is egl related. From it, Bug 1031553 might be related to this bug.

Comment 2

[bhavana bajaj [:bajaj]]

(In reply to Sotaro Ikeda [:sotaro] from comment #1)
> (In reply to Paul Yang [: pyang] from comment #0)
> > Crash Report:
> > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > 1059d2140703
> > Device: flame
> 
> Crash thread stack do not have useful information. Only information is egl
> related. From it, Bug 1031553 might be related to this bug.

Can we add more logging/debugging information to get better data ? Or do you other ideas here ?

Paul, how often do you hit this ? what sort of test cases are being run when you hit this crash ?

Comment 3

[Paul Yang [:pyang] (away)]

(In reply to Sotaro Ikeda [:sotaro] from comment #1)
> (In reply to Paul Yang [: pyang] from comment #0)
> > Crash Report:
> > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > 1059d2140703
> > Device: flame
> 
> Crash thread stack do not have useful information. Only information is egl
> related. From it, Bug 1031553 might be related to this bug.

Bug 1031553 should be related but no good STR as well.

I have logged STR from framework but it take several hours; will replay it today and see if it's reproducible.

Comment 4

[bhavana bajaj [:bajaj]]

(In reply to Paul Yang [: pyang] from comment #3)
> (In reply to Sotaro Ikeda [:sotaro] from comment #1)
> > (In reply to Paul Yang [: pyang] from comment #0)
> > > Crash Report:
> > > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > > 1059d2140703
> > > Device: flame
> > 
> > Crash thread stack do not have useful information. Only information is egl
> > related. From it, Bug 1031553 might be related to this bug.
> 
> Bug 1031553 should be related but no good STR as well.
> 
> I have logged STR from framework but it take several hours; will replay it
> today and see if it's reproducible.

Lets try to do this and get a reduced reproducible testcase. We also are close to fixing 1031553, after which we can verify if this issue still occurs and confirm they are related or DUP's

Comment 5

[Jason Smith [:jsmith]]

(In reply to bhavana bajaj [:bajaj] [NOT reading Bugmail, needInfo please] from comment #4)
> (In reply to Paul Yang [: pyang] from comment #3)
> > (In reply to Sotaro Ikeda [:sotaro] from comment #1)
> > > (In reply to Paul Yang [: pyang] from comment #0)
> > > > Crash Report:
> > > > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > > > 1059d2140703
> > > > Device: flame
> > > 
> > > Crash thread stack do not have useful information. Only information is egl
> > > related. From it, Bug 1031553 might be related to this bug.
> > 
> > Bug 1031553 should be related but no good STR as well.
> > 
> > I have logged STR from framework but it take several hours; will replay it
> > today and see if it's reproducible.
> 
> Lets try to do this and get a reduced reproducible testcase. We also are
> close to fixing 1031553, after which we can verify if this issue still
> occurs and confirm they are related or DUP's

This isn't what MTBF based bugs aim to do. Those type of bugs are filed to indicate blockers to keeping phones alive during a MTBF run. There isn't likely going to be STR associated with these style of bugs.

Comment 6

[bhavana bajaj [:bajaj]]

(In reply to Jason Smith [:jsmith] from comment #5)
> (In reply to bhavana bajaj [:bajaj] [NOT reading Bugmail, needInfo please]
> from comment #4)
> > (In reply to Paul Yang [: pyang] from comment #3)
> > > (In reply to Sotaro Ikeda [:sotaro] from comment #1)
> > > > (In reply to Paul Yang [: pyang] from comment #0)
> > > > > Crash Report:
> > > > > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > > > > 1059d2140703
> > > > > Device: flame
> > > > 
> > > > Crash thread stack do not have useful information. Only information is egl
> > > > related. From it, Bug 1031553 might be related to this bug.
> > > 
> > > Bug 1031553 should be related but no good STR as well.
> > > 
> > > I have logged STR from framework but it take several hours; will replay it
> > > today and see if it's reproducible.
> > 
> > Lets try to do this and get a reduced reproducible testcase. We also are
> > close to fixing 1031553, after which we can verify if this issue still
> > occurs and confirm they are related or DUP's
> 
> This isn't what MTBF based bugs aim to do. Those type of bugs are filed to
> indicate blockers to keeping phones alive during a MTBF run. 
I do understand that, but I think we were trying to get something narrowed down in comment #3 . correct ?

>There isn't
> likely going to be STR associated with these style of bugs.

Comment 7

[Jason Smith [:jsmith]]

(In reply to bhavana bajaj [:bajaj] [NOT reading Bugmail, needInfo please] from comment #6)
> (In reply to Jason Smith [:jsmith] from comment #5)
> > (In reply to bhavana bajaj [:bajaj] [NOT reading Bugmail, needInfo please]
> > from comment #4)
> > > (In reply to Paul Yang [: pyang] from comment #3)
> > > > (In reply to Sotaro Ikeda [:sotaro] from comment #1)
> > > > > (In reply to Paul Yang [: pyang] from comment #0)
> > > > > > Crash Report:
> > > > > > https://crash-stats.mozilla.com/report/index/216cfcd8-418c-44c7-a6a6-
> > > > > > 1059d2140703
> > > > > > Device: flame
> > > > > 
> > > > > Crash thread stack do not have useful information. Only information is egl
> > > > > related. From it, Bug 1031553 might be related to this bug.
> > > > 
> > > > Bug 1031553 should be related but no good STR as well.
> > > > 
> > > > I have logged STR from framework but it take several hours; will replay it
> > > > today and see if it's reproducible.
> > > 
> > > Lets try to do this and get a reduced reproducible testcase. We also are
> > > close to fixing 1031553, after which we can verify if this issue still
> > > occurs and confirm they are related or DUP's
> > 
> > This isn't what MTBF based bugs aim to do. Those type of bugs are filed to
> > indicate blockers to keeping phones alive during a MTBF run. 
> I do understand that, but I think we were trying to get something narrowed
> down in comment #3 . correct ?

Not reduced, but a way to re-trigger the crash if need be to get additional logs. Trying to reduce here would be far too much an overhead cost anyways if the bug takes several hours to reproduce.

Comment 8

[Walter Chen[:ypwalter][:wachen]]

I agree with what Jason says. That is not a reduced STR but a way to completely run through the same test order. In that case, you need to hook the gdb or any other tools of your choice to see how memory changes and other information. When people want to help, they can just setup it to run in the exact same way we did and try to figure it out.

There is not a known solution to reduce the STR as for now. We will try to work toward it as time goes by. Any suggestions are welcomed.

Comment 9

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

This might be a duplicate of bug 1031355?