Closed
Bug 1191592
Opened 9 years ago
Closed 8 years ago
Crashing GTests don't print a readable stack on linux, and print no stack at all on OS X
Categories
(Testing :: General, defect)
Testing
General
Tracking
(Not tracked)
RESOLVED
DUPLICATE
of bug 1055224
People
(Reporter: seth, Unassigned)
References
Details
Right now, GTests don't print a readable stack on linux when they crash - we get addresses, but no symbols. (Quite possibly we just need to postprocess the stack.) On OS X, we don't seem to print any stack at all.
It'd be nice to fix this so debugging crashing GTests is feasible.
Reporter | ||
Updated•9 years ago
|
Component: ImageLib → General
Product: Core → Testing
Reporter | ||
Comment 1•9 years ago
|
||
Bug 1191347 is an excellent demonstration of this phenomenon.
Reporter | ||
Comment 2•9 years ago
|
||
I hate to keep making work for you, Ted, but I think this one is up your alley too, right?
Flags: needinfo?(ted)
Comment 3•9 years ago
|
||
Chris, is this something you could help with otherwise? It's actively blocking a rather irritating orange bug.
Flags: needinfo?(cmanchester)
Comment 4•9 years ago
|
||
And given that "usable crash stacks" is a visibility requirement, there's that too.
Comment 5•9 years ago
|
||
I think this is a dupe of bug 1055224. To quote myself that bug "The problem here is that we actually build the gtest libxul as part of `make check` which is after we've run `make buildsymbols`."
I think we should figure out how to fix this after we fix bug 992983.
There's also bug 1136841 on not getting a stack at all sometimes.
Flags: needinfo?(ted)
Comment 6•9 years ago
|
||
I may be able to look at bug 1055224 after bug 992983, which I'm trying to land this week.
Flags: needinfo?(cmanchester)
Updated•8 years ago
|
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → DUPLICATE
You need to log in
before you can comment on or make changes to this bug.
Description
•