[Webkit-unassigned] [Bug 128928] Improve GDB backtrace generation for GTK/EFL

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=128928

--- Comment #5 from Carlos Alberto Lopez Perez <clopez at igalia.com> ---
(In reply to comment #4)
> Nice investigation indeed.
> 
> The right thing to do is to get cores from coredumpctl so we don't have to
> mess around with changing the core_pattern. I'd be interested in working on
> this eventually (for some large value of eventually).

I guess that would require the users to run systemd. This can be problematic on the bots (they run on unprivileged containers, and systemd don't plays nice there)


Also I'm not sure if it will solve the main problem, that is:

 * Due to WebKit multiprocess model, our tooling for the tests can only know the pid of WebKitTestRunner, but not the pid of any of the childs (WebKitWebProcess, WebKitNetworkProcess, etc).

So, when you get a coredump for WebKitWebProcess (for example) with pid NNNN you can't tell which test caused the crash.

The only reasonable way of solving this that I see is setting an environment variable (that the childs will inherit) with the test name.

When it crashes, the program invoked via the pipe in core_pattern will have access to /proc/$pid/environ and can read this.

With coredumpctl, such program will be systemd-coredump.

So...  the main question.... is coredumpctl able to get and print the environ variables that a program had defined when it crashed for a given coredump? If is not, I don't think it will be of any help here.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20151223/5ad3d0f9/attachment.html>