[Automated review] Coverity tasks fail with "catastrophic signal: 11 (SIGSEGV)" followed by "connectionpool.py:791: InsecureRequestWarning: Unverified HTTPS request is being made."
Categories
(Developer Infrastructure :: Source Code Analysis, defect, P1)
Tracking
(Not tracked)
People
(Reporter: dholbert, Assigned: andi)
References
(Blocks 1 open bug)
Details
Phabricator URL: https://phabricator.services.mozilla.com/D83456
Reviewbot try push with coverity tasks (which I retriggered, and it seems to be a perma-failure):
https://treeherder.mozilla.org/#/jobs?repo=try&revision=042bb30193e322de5f738fb2b394305fe7bfe01d&selectedTaskRun=WscAkouWSluJmCpm0beeXg-0
The failure looks like this (I shortened one line with [...SNIP...]:
[task 2020-07-21T15:18:32.409Z] [STATUS] Compiling Unified_cpp_layout_generic0.cpp
[task 2020-07-21T15:18:32.409Z] /builds/worker/workspace/coverity/cov-analysis-linux64-2020.03/bin/cov-internal-emit-clang [...SNIP...] Unified_cpp_layout_generic0.cpp
[task 2020-07-21T15:19:03.939Z] catastrophic signal: 11 (SIGSEGV)
[task 2020-07-21T15:19:03.940Z] faulting data pointer: 0x0000000000000030
[task 2020-07-21T15:19:03.940Z] call stack backtrace:
[task 2020-07-21T15:19:03.940Z] cov-internal-emit-clang linux64 2020.03
[task 2020-07-21T15:19:03.940Z] 0x000000000093d6d0
[task 2020-07-21T15:19:03.940Z] 0x00000000005af76c
[task 2020-07-21T15:19:03.940Z] 0x000000000066dee8
[task 2020-07-21T15:19:03.940Z] 0x000000000058c08b
[task 2020-07-21T15:19:03.940Z] 0x000000000066e068
[task 2020-07-21T15:19:03.940Z] 0x00000000005972ad
[task 2020-07-21T15:19:03.940Z] 0x000000000066ebe8
[task 2020-07-21T15:19:03.940Z] 0x0000000000a34bae
[task 2020-07-21T15:19:03.940Z] 0x000000000061784a
[task 2020-07-21T15:19:03.940Z] 0x0000000000606eba
[task 2020-07-21T15:19:03.940Z] 0x000000000057e220
[task 2020-07-21T15:19:03.940Z] 0x000000000042aa3e
[task 2020-07-21T15:19:03.940Z] 0x000000000042efff
[task 2020-07-21T15:19:03.940Z] 0x0000000000554b04
[task 2020-07-21T15:19:03.940Z] 0x000000000054e0c2
[task 2020-07-21T15:19:03.940Z] 0x0000000000a5c455
[task 2020-07-21T15:19:03.940Z] 0x000000000043b59e
[task 2020-07-21T15:19:03.940Z] libc.so.6 linux64 2020.03
[task 2020-07-21T15:19:03.940Z] 0x000000000002409b
[task 2020-07-21T15:19:03.958Z] WARNING: cov-internal-emit-clang returned with code 4 for Unified_cpp_layout_generic0.cpp
[task 2020-07-21T15:19:03.960Z]
[task 2020-07-21T15:19:03.974Z] Coverity Desktop Analysis version 2020.03 on Linux 4.4.0-1014-aws x86_64
[task 2020-07-21T15:19:03.974Z] [STATUS] Getting snapshot for code version date 2020-07-21T15:17:12+00:00...
[task 2020-07-21T15:19:05.513Z] Selected 1 translation unit for analysis:
[task 2020-07-21T15:19:05.513Z] ! /builds/worker/checkouts/gecko/obj-x86_64-pc-linux-gnu/layout/generic/Unified_cpp_layout_generic0.cpp
[task 2020-07-21T15:19:05.513Z]
[task 2020-07-21T15:19:05.513Z] [WARNING] 1 TU selected for analysis (100%, marked with '!') failed to emit
[task 2020-07-21T15:19:05.513Z] properly and will produce no analysis results.
[task 2020-07-21T15:19:05.513Z]
[task 2020-07-21T15:19:05.513Z] [STATUS] Parsing source files...
[task 2020-07-21T15:19:32.261Z] Compilation failed with code 4
[task 2020-07-21T15:19:32.264Z]
[task 2020-07-21T15:19:32.266Z] /builds/worker/checkouts/gecko/third_party/python/requests/requests/packages/urllib3/connectionpool.py:791: InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.org/en/latest/security.html
[task 2020-07-21T15:19:32.266Z] InsecureRequestWarning)
[task 2020-07-21T15:19:32.266Z] Configure complete!
https://firefoxci.taskcluster-artifacts.net/Af0nc-RqQECSzua_tZrthw/0/public/logs/live_backing.log
Notably, there's both a "signal 11" segfault as well as an HTTPS error there; I'm not sure if they're at all related (perhaps not).
Reporter | ||
Comment 1•5 years ago
|
||
Note that my patch does add a new .cpp file (which is pretty trivial, just some "stub" to-be-implemented functions).
That presumably changes the unified-compilation bucketing a little bit, with respect to current mozilla-central. Not sure if that's relevant or something that confuses coverity.
Reporter | ||
Comment 2•5 years ago
|
||
One (or more) of boris's patches hit the same issue, BTW:
https://phabricator.services.mozilla.com/D79337#2619485
https://treeherder.mozilla.org/#/jobs?repo=try&selectedTaskRun=QYKoYRn_SH6p5F8Jb7e4lA-0&revision=6643253c7b107900861b5f56901ab260d9722a63
https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=310649833&repo=try&lineNumber=1150
Assignee | ||
Comment 3•5 years ago
|
||
Daniel would it be possible to re-push the revisions that have issues for the bot to re-analyze them. We did an update to the Coverity Analysis and hopefully this update fixes the crashes from Coverity.
Reporter | ||
Comment 4•5 years ago
|
||
Sure - I've re-pushed it (after some minor rebasing). Same phab URL. Let's see how reviewbot deals with it - thanks!
Reporter | ||
Comment 5•5 years ago
•
|
||
Looks like the try push for my resubmission is:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=08e62f5ed44c28004013ff4761007fc9d6968bed
(Still running for now)
Reporter | ||
Comment 6•5 years ago
|
||
That one also failed, with same issues as in comment 0:
https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=310726883&repo=try&lineNumber=837
Comment 7•5 years ago
|
||
The severity field is not set for this bug.
:andi, could you have a look please?
For more information, please visit auto_nag documentation.
Assignee | ||
Updated•5 years ago
|
Reporter | ||
Comment 8•5 years ago
|
||
Maybe it's just me, but it's feeling like this task is just a perma-fail right now.
If that's the case (not sure), could we disable the task, or at least disable the reviewbot comments until that's fixed? (Or is it just me?)
Reporter | ||
Comment 9•4 years ago
|
||
The crash is always in Unified_cpp_layout_generic0.cpp
, I think -- so maybe this is only an issue for patches that touch that compilation unit. (i.e. maybe it's only a perma-fail for folks whose patches are touching layout code.)
Is there somewhere we can report this upstream to Coverity folks? It seems highly reproducible, and I suspect they'd be interested in investigating & fixing their bug (or reporting it further upstream e.g. as a clang bug or whatever)
Assignee | ||
Updated•4 years ago
|
Assignee | ||
Comment 10•4 years ago
|
||
I'm gonna take care of this issue.
Assignee | ||
Comment 12•4 years ago
|
||
The issue is still under investigation by synopsys.
Assignee | ||
Updated•4 years ago
|
Assignee | ||
Comment 13•4 years ago
|
||
I've pinged them today again, hopefully we will have a resolution from them. What I see from the logs and what I get when trying this locally it seems to be a bad pointer and I think inside their code this address is hard-coded 0x0000000000000030
.
Assignee | ||
Comment 14•4 years ago
|
||
Just submitted again some extra data that Synopsys required in order to fix this issue.
Assignee | ||
Comment 15•4 years ago
|
||
This has been fixed in the latest coverity release.
Updated•3 years ago
|
Description
•