Bug 135719

Summary: REGRESSION(169139): LLINT intermittently fails JSC testapi tests
Product: WebKit Reporter: Mark Lam <mark.lam>
Component: JavaScriptCoreAssignee: Mark Lam <mark.lam>
Status: RESOLVED FIXED    
Severity: Normal CC: webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
the patch. ggaren: review+

Mark Lam
Reported 2014-08-07 13:38:21 PDT
See http://build.webkit.org/builders/Apple%20Mavericks%20LLINT%20CLoop%20%28BuildAndTest%29/builds/7450/steps/webkit-jsc-cloop-test/logs/stdio. Maybe this is just a time out thing, but we should look into fixing the issue so that the LLINT CLoop bot is more reliable.
Attachments
the patch. (3.64 KB, patch)
2015-03-25 16:10 PDT, Mark Lam 		ggaren: review+
DetailsFormatted DiffDiff
View All Add attachment proposed patch, testcase, etc.
Radar WebKit Bug Importer
Comment 1 2014-08-07 13:39:01 PDT
<rdar://problem/17949979>
Mark Lam
Comment 2 2015-03-25 15:35:09 PDT
This is a regression introduced in http://trac.webkit.org/changeset/169139 which changed VM::watchdog from an embedded field into a std::unique_ptr, but did not update the LLINT to access it as such. The issue has only manifested so far on the CLoop tests because those are LLINT only. In the non-CLoop cases, the JIT kicks in and does the right thing, thereby hiding the bug in the LLINT.
Mark Lam
Comment 3 2015-03-25 16:10:12 PDT
Created attachment 249438 [details] the patch.
Geoffrey Garen
Comment 4 2015-03-25 16:13:16 PDT
Comment on attachment 249438 [details] the patch. r=me
Mark Lam
Comment 5 2015-03-25 16:16:54 PDT
Thanks. Landed in r181981: <http://trac.webkit.org/r181981>.
Note You need to log in before you can comment on or make changes to this bug.