Breakpad Linux: Fix crash handler writing to disk on the wrong thread....

Breakpad Linux: Fix crash handler writing to disk on the wrong thread.

BUG=none
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=75325

[Lei Zhang, 2011-02-17 22:49:08]

Official build in debug mode triggered the thread restriction code because we
were doing writes on the IO thread.

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
File chrome/browser/crash_handler_host_linux.cc (left):

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
chrome/browser/crash_handler_host_linux.cc:297: HANDLE_EINTR(close(signal_fd));
On failure, we closed |signal_fd| twice.

[agl, 2011-02-17 22:57:41]

LGTM

[piman, 2011-04-15 05:17:16]

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
File chrome/browser/crash_handler_host_linux.cc (right):

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
chrome/browser/crash_handler_host_linux.cc:309:
reinterpret_cast<char*>(&crash_context),
Woah, are we passing a reference to a buffer on the stack here ? It's invalid
when WriteDumpFile executes

Also, is it really necessary to hop threads ? We are already on a worker thread.

[Lei Zhang, 2011-04-15 05:25:07]

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
File chrome/browser/crash_handler_host_linux.cc (right):

http://codereview.chromium.org/6538033/diff/1/chrome/browser/crash_handler_ho...
chrome/browser/crash_handler_host_linux.cc:309:
reinterpret_cast<char*>(&crash_context),
On 2011/04/15 05:17:17, piman wrote:
> Woah, are we passing a reference to a buffer on the stack here ? It's invalid
> when WriteDumpFile executes
> 
> Also, is it really necessary to hop threads ? We are already on a worker
thread.

Doh. I'll fix it in a bit.