Bug 111440

Summary:

Chromium DumpRenderTree dumps a core every time it's run (Linux/Debug)

Product:

WebKit

Reporter:

Tommy Widenflycht <tommyw>

Component:

Tools / Tests

Assignee:

James Robinson <jamesr>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

jamesr

Priority:

Version:

528+ (Nightly build)

Hardware:

Unspecified

OS:

Linux

Attachments:

Description	Flags
Patch	enne: review+

Description Tommy Widenflycht 2013-03-05 07:34:43 PST

Call stack:

$ out/Debug/DumpRenderTree 
#EOF
[29529:29529:2182963651329:FATAL:web_compositor_support_impl.cc(76)] Check failed: initialized_. 
 [0x7ff9bada6e6e] base::debug::StackTrace::StackTrace()
 [0x7ff9bade5f8f] logging::LogMessage::~LogMessage()
 [0x7ff9bb191df3] webkit::WebCompositorSupportImpl::shutdown()
 [0x00000043f6cf] main
 [0x7ff9b039076d] __libc_start_main
 [0x00000043cc29] <unknown>

Trace/breakpoint trap (core dumped)

Comment 1 Tommy Widenflycht 2013-03-05 08:39:44 PST

Is the call to
    WebKit::Platform::current()->compositorSupport()->shutdown();
in
    Tools/DumpRenderTree/chromium/DumpRenderTree.cpp:296
really necessary? Considering that 
    WebKit::Platform::current()->compositorSupport()->initialize()
has never been called?

Comment 2 James Robinson 2013-03-05 10:16:10 PST

Neither call are necessary - the assert is just too eager.  Will fix.

Comment 3 James Robinson 2013-03-05 10:18:56 PST

Created attachment 191511 [details]
Patch

Comment 4 Adrienne Walker 2013-03-05 10:20:12 PST

Comment on attachment 191511 [details]
Patch

R=me.

Comment 5 James Robinson 2013-03-05 10:20:41 PST

Committed r144775: <http://trac.webkit.org/changeset/144775>