Bug 53469
Summary: | [chromium] compositing/scroll-painted-composited-content.html crashes on Mac | ||
---|---|---|---|
Product: | WebKit | Reporter: | Ryosuke Niwa <rniwa> |
Component: | WebGL | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED FIXED | ||
Severity: | Normal | CC: | jamesr, kbr, mihaip, senorblanco, tony, vangelis, zmo |
Priority: | P2 | ||
Version: | 528+ (Nightly build) | ||
Hardware: | PC | ||
OS: | OS X 10.5 |
Ryosuke Niwa
http://trac.webkit.org/changeset/77172 has been crashing on Mac canary bot since http://trac.webkit.org/changeset/77172 was landed.
Attachments | ||
---|---|---|
Add attachment proposed patch, testcase, etc. |
Mihai Parparita
This appears to be an interaction with another test; I can only reproduce the crash if I run:
new-run-webkit-tests --platform=chromium-gpu --verbose compositing/repaint/layer-repaint-rects.html compositing/scroll-painted-composited-content.html
I'm not getting a stack trace, and the one from a debug bot (http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Mac10.5%20%28dbg%29%282%29/builds/258/steps/webkit_gpu_tests/logs/stdio) looks mangled:
[80962:267:1311695793449825:ERROR:process_util_posix.cc(106)] Received signal 10
0 DumpRenderTree 0x0000468a _mh_execute_header + 13962
1 DumpRenderTree 0x005b5e0f start + 5865355
2 libSystem.B.dylib 0x90f1d2bb _sigtramp + 43
3 ??? 0xffffffff 0x0 + 4294967295
4 DumpRenderTree 0x00061456 start + 275922
5 DumpRenderTree 0x0006106f start + 274923
6 DumpRenderTree 0x0008c4a6 start + 452130
7 DumpRenderTree 0x00d620d9 start + 13910613
8 DumpRenderTree 0x00e9b9d2 float WebCore::narrowPrecisionToCGFloat<double>(double) + 1195046
9 DumpRenderTree 0x00e93b01 float WebCore::narrowPrecisionToCGFloat<double>(double) + 1162581
10 DumpRenderTree 0x000e0225 start + 795553
11 DumpRenderTree 0x00d8d2e7 float WebCore::narrowPrecisionToCGFloat<double>(double) + 87355
12 DumpRenderTree 0x00d8703b float WebCore::narrowPrecisionToCGFloat<double>(double) + 62095
13 DumpRenderTree 0x00d875ad float WebCore::narrowPrecisionToCGFloat<double>(double) + 63489
14 DumpRenderTree 0x000d65d7 start + 755539
15 DumpRenderTree 0x000d9f41 start + 770237
16 DumpRenderTree 0x000586d4 start + 239696
17 DumpRenderTree 0x00058952 start + 240334
18 DumpRenderTree 0x00038a1a start + 109462
19 DumpRenderTree 0x00042af4 start + 150640
20 DumpRenderTree 0x0002574f start + 30923
21 DumpRenderTree 0x000257d2 start + 31054
22 DumpRenderTree 0x00f3fcd9 float WebCore::narrowPrecisionToCGFloat<double>(double) + 1867565
23 DumpRenderTree 0x00f3fe6f float WebCore::narrowPrecisionToCGFloat<double>(double) + 1867971
24 DumpRenderTree 0x009fe625 start + 10356641
25 DumpRenderTree 0x009fe6e0 start + 10356828
26 DumpRenderTree 0x009fe705 start + 10356865
27 ??? 0x0647228e 0x0 + 105325198
+ more GPU folks, since I don't know much about this (all http://trac.webkit.org/changeset/77172 did was to make it so we actually pass --enable-accelerated-compositing --enable-accelerated-2d-canvas on the Mac too).
Stephen White
This test stopped crashing, I believe due to http://trac.webkit.org/changeset/89206/ which simply changed the sort order of tests, thus moving the crashes around.
See also http://crbug.com/86714 for where the crashiness ended up.