(No comment was entered for this change.)

Issue 21682 has been merged into this issue.

Saw this again. Build details here:
http://chromeos-botmaster.mtv.corp.google.com:8026/builders/x86-alex-binary/builds/1635

two instances:
http://chromeos-botmaster.mtv.corp.google.com:8026/builders/stumpy-binary/builds/945/steps/VMTest/logs/stdio
http://chromeos-botmaster.mtv.corp.google.com:8026/builders/x86-alex-binary/builds/1659/steps/VMTest/logs/stdio

crosbug.com/20638 might be related to this issue.

Last two comments are  Issue 21559 .

This bug should only track crashes with the following signature:

Crashes detected during testing:
Traceback (most recent call last):
  File "/usr/bin/generate_test_report", line 380, in <module>
    main()
  File "/usr/bin/generate_test_report", line 376, in main
    generator.Run()
  File "/usr/bin/generate_test_report", line 335, in Run
    self._GenerateReportText()
  File "/usr/bin/generate_test_report", line 295, in _GenerateReportText
    print line
NameError: global name 'line' is not defined

But the comments that caused this to be initially filed were _not_ that.

Do we have any reason to believe that this error is causing us not to get chrome stacks?  if not, why are we re-purposing this issue?

What branch is that generate_test_report from? The stack does not match against ToT. I don't even see a "print line" statement.

http://git.chromium.org/gitweb/?p=chromiumos/platform/crostestutils.git;a=blob;f=utils_py/generate_test_report.py

Also stacks are generated inside of Autotest not in generate_test_report, so this error would have to be unrelated if the question is "Is this causing us to lose stack traces?"

The two failures in Chris's original report have similar messages in /var/log/messages, and indicate that it couldn't read the process from /proc at all:

From http://chromeos-botmaster.mtv.corp.google.com:8026/builders/lumpy-binary/builds/343

2011-10-13T01:45:57.299047+00:00 localhost kernel: [  116.298562] chrome[9094]: segfault at 70363bcb ip 741607c2 sp 6f774230 error 4 in chrome[72f63000+4893000]
2011-10-13T01:45:57.440727+00:00 localhost crash_reporter[9107]: Received crash notification for chrome[9082] sig 11 (developer build - not testing - always dumping)
2011-10-13T01:45:57.446238+00:00 localhost crash_reporter[9107]: Could not read status file
2011-10-13T01:45:57.446251+00:00 localhost crash_reporter[9107]: Path /proc/9082 DirectoryExists: 0
2011-10-13T01:45:57.446258+00:00 localhost crash_reporter[9107]: Unable to find/create process-specific crash path
2011-10-13T01:45:57.446265+00:00 localhost crash_reporter[9107]: Writing conversion problems as separate crash report.

From http://chromeos-botmaster.mtv.corp.google.com:8026/builders/x86-zgb-binary/builds/1496

2011-10-12T20:45:36.422048-07:00 localhost kernel: [  301.421562] chrome[10898]: segfault at 70284bcb ip 740837c2 sp 6f695230 error 4 in chrome[72e86000+4893000]
2011-10-12T20:45:37.226279-07:00 localhost crash_reporter[10911]: Received crash notification for chrome[10886] sig 11 (developer build - not testing - always dumping)
2011-10-12T20:45:37.234274-07:00 localhost crash_reporter[10911]: Could not read status file
2011-10-12T20:45:37.234290-07:00 localhost crash_reporter[10911]: Path /proc/10886 DirectoryExists: 0
2011-10-12T20:45:37.234298-07:00 localhost crash_reporter[10911]: Unable to find/create process-specific crash path
2011-10-12T20:45:37.234303-07:00 localhost crash_reporter[10911]: Writing conversion problems as separate crash report.

For comparison, the supplied_chrome crashes look slightly different:

2011-10-14T20:49:51.285039+00:00 localhost kernel: [  230.284578] chrome[10877]: segfault at 70249bd7 ip 740c7ef2 sp 6f656230 error 4 in chrome[72eb5000+485e000]
2011-10-14T20:49:51.319395+00:00 localhost crash_reporter[10890]: Readlink failed on /proc/10865/exe with 2
2011-10-14T20:49:51.319526+00:00 localhost crash_reporter[10890]: GetSymlinkTarget failed - Path /proc/10865 DirectoryExists: 0
2011-10-14T20:49:51.319620+00:00 localhost crash_reporter[10890]: stat /proc/10865/exe failed: -1 2
2011-10-14T20:49:51.325540+00:00 localhost crash_reporter[10890]: Received crash notification for supplied_chrome[10865] sig 11 (developer build - not testing - always dumping)
2011-10-14T20:49:51.333302+00:00 localhost crash_reporter[10890]: Could not read status file
2011-10-14T20:49:51.333366+00:00 localhost crash_reporter[10890]: Path /proc/10865 DirectoryExists: 0
2011-10-14T20:49:51.333374+00:00 localhost crash_reporter[10890]: Unable to find/create process-specific crash path
2011-10-14T20:49:51.333380+00:00 localhost crash_reporter[10890]: Writing conversion problems as separate crash report.

If we're going to track this as just covering the stackless chrome crashes, I'll mark this as a duplicate of 21559 which has a bit more info.

(No comment was entered for this change.)

(No comment was entered for this change.)