Summary: | run-perf-tests should support content_shell | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Adam Barth <abarth> | ||||||||
Component: | New Bugs | Assignee: | Adam Barth <abarth> | ||||||||
Status: | RESOLVED FIXED | ||||||||||
Severity: | Normal | CC: | dpranke, eric, jochen, rniwa, webkit.review.bot | ||||||||
Priority: | P2 | ||||||||||
Version: | 528+ (Nightly build) | ||||||||||
Hardware: | Unspecified | ||||||||||
OS: | Unspecified | ||||||||||
Attachments: |
|
Description
Adam Barth
2013-03-13 14:30:22 PDT
Created attachment 192992 [details]
work in progress
Created attachment 192997 [details]
Patch
Comment on attachment 192997 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=192997&action=review > Tools/Scripts/webkitpy/performance_tests/perftest.py:238 > + re.compile(r'INFO:SkFontHost_fontconfig.cpp'), > + re.compile(r'Running without the SUID sandbox'), You might want to mark that these are from content_shell in a comment. Comment on attachment 192997 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=192997&action=review >> Tools/Scripts/webkitpy/performance_tests/perftest.py:238 >> + re.compile(r'Running without the SUID sandbox'), > > You might want to mark that these are from content_shell in a comment. Or just install the SUID sandbox... https://code.google.com/p/chromium/wiki/LinuxSUIDSandboxDevelopment > Or just install the SUID sandbox... https://code.google.com/p/chromium/wiki/LinuxSUIDSandboxDevelopment
That would solve the problem for me, but other folks would run into the error.
Created attachment 192998 [details]
Patch for landing
Doesn't the error include the link how to fix it? I think the idea is that running any browser without a sandbox is bad. However, since content_shell in DRT mode doesn't allow outgoing connections anyway, I guess it doesn't really matter. > Doesn't the error include the link how to fix it? Yes, but I want to run performance tests, not fix configurations problems. :) > However, since content_shell in DRT mode doesn't allow outgoing connections anyway, I guess it doesn't really matter. Maybe I should fix it, but it should block me from profiling "<html>". Comment on attachment 192998 [details] Patch for landing Clearing flags on attachment: 192998 Committed r145764: <http://trac.webkit.org/changeset/145764> All reviewed patches have been landed. Closing bug. |