Bug#647992: chromium crashes

Jonathan Nieder jrnieder at gmail.com
Wed Nov 9 08:51:56 UTC 2011


Hi,

Hideki Yamane wrote:

>  After upgrading to 15.0.874.106~r107270-1, chromium cannot show any pages
>  (setting, local and remote).

Yep.

>  I'll attach backtrace as http://wiki.debian.org/Chromium/Debugging instructs.

Nice!  How did you do that? :)

When I tried to get a backtrace here, I have no luck:

chromium -g::
	Renderer dies, chrome stays alive --- no backtrace.

chromium -g --temp-profile --single-process::
	Renderer doesn't die.

chromium --temp-profile \
  --renderer-cmd-prefix='xterm -title renderer -e gdb --args'::
	Renderer doesn't die.

chromium --temp-profile --renderer-cmd-prefix='xterm -e'::
	Renderer doesn't die.

chromium --temp-profile --renderer-cmd-prefix=''::
	Renderer dies, chrome stays alive --- no backtrace.

Which makes sense: as the upstream LinuxDebugging doc says,

| Note: using the --renderer-cmd-prefix option bypasses the zygote
| launcher, so the renderers won't be sandboxed. It is generally not an
| issue, except when you are trying to debug interactions with the
| sandbox.

I'm just surprised it was possible to easily do that without
artificially inserting a sleep() somewhere.

> Program received signal SIGSEGV, Segmentation fault.
> 0x00007ffff8fb02a0 in ?? ()
> (gdb) bt
> #0  0x00007ffff8fb02a0 in ?? ()
> #1  0x00007ffff5097971 in BalloonViewHost::UpdateActualSize (this=0x7ffff8cb60e0, new_size=...) at chrome/browser/ui/gtk/notifications/balloon_view_host_gtk.cc:22
[...]

Would it be possible to get a full backtrace ("bt full") and
disassembly in the top frame near the ip at the time of the crash
("disassemble")?

Thanks much,
Jonathan





More information about the Pkg-games-devel mailing list