Closed Bug 1197941 Opened 8 years ago Closed 8 years ago

Intermittent subprocess.CalledProcessError: Command '['/builds/slave/l64-br-haz_m-in_dep-0000000000/build/obj-opt-js/dist/bin/js', '/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/js/src/devtools/rootAnalysis/computeCallgraph.js']' returned non-

Categories

(Core :: JavaScript Engine, defect)

Product:
Core
Component:
JavaScript Engine
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED
Milestone:
mozilla44
Tracking Flags:
Tracking Status
firefox43 --- affected
firefox44 --- fixed

People

(Reporter: RyanVM, Assigned: sfink)

Details

(Keywords: intermittent-failure)

Attachments

(1 file, 1 obsolete file)

Allow getline() to malloc its own buffer to avoid intermittent crashes
1.89 KB, patch
shu
: review+
Details | Diff | Splinter Review 
08:41:38 INFO - Running callgraph to generate callgraph.txt
08:41:38 INFO - PATH="/builds/slave/l64-br-haz_m-in_dep-0000000000/build/gcc:/builds/slave/l64-br-haz_m-in_dep-0000000000/build/sixgill/usr/bin:${PATH}" ANALYZED_OBJDIR='/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/obj-analyzed' XDB='/builds/slave/l64-br-haz_m-in_dep-0000000000/build/sixgill/usr/bin/xdb.so' SOURCE='/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source' /builds/slave/l64-br-haz_m-in_dep-0000000000/build/obj-opt-js/dist/bin/js /builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/js/src/devtools/rootAnalysis/computeCallgraph.js > callgraph.txt
08:41:38 INFO - xdb/xdb.h: 99: MinDataStream: Assertion 'm_fd != -1 && !m_has_error' failed.
08:41:38 INFO - Redirecting call to abort() to mozalloc_abort
08:41:38 ERROR - Traceback (most recent call last):
08:41:38 INFO - File "/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/js/src/devtools/rootAnalysis/analyze.py", line 279, in <module>
08:41:38 INFO - run_job(step, data)
08:41:38 INFO - File "/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/js/src/devtools/rootAnalysis/analyze.py", line 175, in run_job
08:41:38 INFO - subprocess.check_call(command, stdout=output, env=env(config))
08:41:38 INFO - File "/tools/python27/lib/python2.7/subprocess.py", line 540, in check_call
08:41:38 INFO - raise CalledProcessError(retcode, cmd)
08:41:38 INFO - subprocess.CalledProcessError: Command '['/builds/slave/l64-br-haz_m-in_dep-0000000000/build/obj-opt-js/dist/bin/js', '/builds/slave/l64-br-haz_m-in_dep-0000000000/build/source/js/src/devtools/rootAnalysis/computeCallgraph.js']' returned non-zero exit status -11
08:41:38 ERROR - Return code: 1
08:41:38 FATAL - 'failed to build'
08:41:38 FATAL - Running post_fatal callback...
08:41:38 FATAL - Exiting 2
program finished with exit code 2
It'd be very handy to have an assertion failure instead of a silent SIGSEGV. Note that this will bump the browser hazard job from random_choice([100,40]) to 160 minutes.
Attachment #8667558 - Flags: review?(shu)
Assignee: nobody → sphink
Status: NEW → ASSIGNED
Attachment #8667558 - Flags: review?(shu) → review+
Comment on attachment 8667605 [details] [diff] [review]
Allow getline() to malloc its own buffer to avoid intermittent crashes

Review of attachment 8667605 [details] [diff] [review]:
-----------------------------------------------------------------

Yay

::: js/src/devtools/rootAnalysis/utility.js
@@ +176,2 @@
>    libc.fclose(fp);
> +  printErr("At end, bufsize = " + bufsize.value.toString());

I guess this printf doesn't hurt.
Attachment #8667605 - Flags: review?(shu) → review+
Multiple failures on inbound since the push in comment 40.
Flags: needinfo?(sphink)
What, you don't like how he made it permanent instead of intermittent? Backed out in https://hg.mozilla.org/integration/mozilla-inbound/rev/e5ee1b65e448
Re-landed with a version that I, y'know, actually ran once.

Oddly, one severe bug (my ctypes fclose() call was actually invoking fopen()) hasn't seemed to particularly bother it all this time.
Comment on attachment 8667558 [details] [diff] [review]
Temporarily switch hazard shell to opt+debug to diagnose

No longer needed, and wouldn't have helped in the first place (unless it turned on jemalloc debugging?)
Attachment #8667558 - Attachment is obsolete: true
Flags: needinfo?(sphink)
https://hg.mozilla.org/mozilla-central/rev/e535d0587fda
Status: ASSIGNED → RESOLVED
Closed: 8 years ago
status-firefox44: --- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla44
You need to log in before you can comment on or make changes to this bug.