crash in RegisterStaticAtoms due to bad hardware (disk? memory?)

NEW
Unassigned

Status

()

Core
XPCOM
--
critical
2 years ago
a year ago

People

(Reporter: calixte, Unassigned)

Tracking

({crash})

unspecified
x86
Windows NT
crash
Points:
---

Firefox Tracking Flags

(firefox46 affected, firefox47 affected, firefox48 affected, firefox49 affected, firefox-esr45 affected, firefox50 affected, firefox51 affected, firefox52 wontfix, firefox53 affected)

Details

(crash signature)

(Reporter)

Description

2 years ago
This bug was filed from the Socorro interface and is 
report bp-023ebcb2-cc34-4c83-b378-e113e2160426.
=============================================================

It's a startup crash and it just appeared the 2016-04-25 on aurora.
There have been a few spikes, from a single user, on both aurora and nightly.

I looked at bp-b9973864-024d-401d-8398-9e30b2160429.  Essentially we're crashing inside HashString(), allegedly 0x1917FC characters into the string, and with 0x403fc4a8 left to go.  We crash when we cross a page boundary.  (The crash addresses on x86, where crash addresses are reliable, all end in three zeros.)  The actual string should be "itemref", which is 7 characters.

This makes me suspect disk corruption of the binary.
(Each spike from a single user, not necessarily the same for the different spikes.)
bp-286bc19f-dd8d-4905-b324-3b6932160407, from the same user, 7 seconds earlier, has a crash that looks like a bit flip (RBX == 0x200007fec6900028), where the 2 in the high bit seems wrong.  That makes it seem more like bad memory, which could also explain the other.
Indeed, this one is just iterating over static data. There's no obvious possibility for error without some kind of problem beyond our control, such as disk corruption or hardware faults.
Crash volume for signature 'RegisterStaticAtoms':
 - nightly (version 50): 0 crash from 2016-06-06.
 - aurora  (version 49): 6 crashes from 2016-06-07.
 - beta    (version 48): 279 crashes from 2016-06-06.
 - release (version 47): 650 crashes from 2016-05-31.
 - esr     (version 45): 37 crashes from 2016-04-07.

Crash volume on the last weeks:
             Week N-1   Week N-2   Week N-3   Week N-4   Week N-5   Week N-6   Week N-7
 - nightly          0          0          0          0          0          0          0
 - aurora           1          0          0          5          0          0          0
 - beta            50         47         66         57         17         34          2
 - release         88         71         96        109        102        109         38
 - esr              1          1          5          0         12          8          1

Affected platforms: Windows, Mac OS X, Linux
status-firefox49: --- → affected
status-firefox-esr45: --- → affected
Crash volume for signature 'RegisterStaticAtoms':
 - nightly(version 50):1 crash from 2016-06-06.
 - aurora (version 49):6 crashes from 2016-06-07.
 - beta   (version 48):286 crashes from 2016-06-06.
 - release(version 47):703 crashes from 2016-05-31.
 - esr    (version 45):37 crashes from 2016-04-07.

Crash volume on the last weeks:
            W. N-1  W. N-2  W. N-3  W. N-4  W. N-5  W. N-6  W. N-7
 - nightly       1       0       0       0       0       0       0
 - aurora        0       1       0       0       5       0       0
 - beta         13      50      47      66      57      17      34
 - release      87      88      71      96     109     102     109
 - esr           2       1       1       5       0      12       8

Affected platforms: Windows, Mac OS X, Linux
status-firefox50: --- → affected
Crash volume for signature 'RegisterStaticAtoms':
 - nightly (version 52): 5 crashes from 2016-09-19.
 - aurora  (version 51): 0 crashes from 2016-09-19.
 - beta    (version 50): 36 crashes from 2016-09-20.
 - release (version 49): 88 crashes from 2016-09-05.
 - esr     (version 45): 127 crashes from 2016-06-01.

Crash volume on the last weeks (Week N is from 10-03 to 10-09):
            W. N-1  W. N-2
 - nightly       2       3
 - aurora        0       0
 - beta         30       6
 - release      60      18
 - esr           3       1

Affected platforms: Windows, Mac OS X, Linux

Crash rank on the last 7 days:
           Browser   Content     Plugin
 - nightly #372
 - aurora
 - beta    #406
 - release #800      #1300
 - esr     #1892
status-firefox52: --- → affected
Crash volume for signature 'RegisterStaticAtoms':
 - nightly (version 53): 2 crashes from 2016-11-14.
 - aurora  (version 52): 8 crashes from 2016-11-14.
 - beta    (version 51): 292 crashes from 2016-11-14.
 - release (version 50): 724 crashes from 2016-11-01.
 - esr     (version 45): 215 crashes from 2016-07-22.

Crash volume on the last weeks (Week N is from 01-16 to 01-22):
            W. N-1  W. N-2  W. N-3  W. N-4  W. N-5  W. N-6  W. N-7
 - nightly       1       0       0       0       1       0       0
 - aurora        0       0       1       0       0       0       0
 - beta         76      50      24      18       3      38       9
 - release      88      65      65      81      82      77     107
 - esr           9       3       0      20       1      12      12

Affected platforms: Windows, Mac OS X, Linux

Crash rank on the last 7 days:
           Browser   Content   Plugin
 - nightly #2762     #1519
 - aurora  #800
 - beta    #357      #1682
 - release #681      #2418
 - esr     #917
status-firefox51: --- → affected
status-firefox53: --- → affected
Too late for firefox 52, mass-wontfix.
status-firefox52: affected → wontfix
Is this even worth keeping open if we know/suspect we can't fix it per comment 4?
Flags: needinfo?(dbaron)
I tend to think it's better to keep it open, given that we don't have a better mechanism for tracking crashes that we can't fix, and an open bug is a sign that it's not fixed.
Flags: needinfo?(dbaron)
Summary: crash in RegisterStaticAtoms → crash in RegisterStaticAtoms due to bad hardware (disk? memory?)
You need to log in before you can comment on or make changes to this bug.