Bug 91296 - LLInt fails to mark structures stored in the bytecode
Summary: LLInt fails to mark structures stored in the bytecode
Status: REOPENED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Oliver Hunt
URL:
Keywords:
Depends on: 91298
Blocks:
  Show dependency tree / graph
 
Reported: 2012-07-13 15:59 PDT by Oliver Hunt
Modified: 2012-07-13 16:33 PDT (History)
1 user (show)

See Also:

Attachments
Patch (4.21 KB, patch)
2012-07-13 16:03 PDT, Oliver Hunt 		ggaren: review+
Details | Formatted Diff | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Oliver Hunt 2012-07-13 15:59:26 PDT 
LLInt fails to mark structures stored in the bytecode
Comment 1 Oliver Hunt 2012-07-13 16:03:23 PDT 
Created attachment 152362 [details]
Patch
Comment 2 Geoffrey Garen 2012-07-13 16:07:25 PDT 
Comment on attachment 152362 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=152362&action=review

r=me

> Source/JavaScriptCore/bytecode/CodeBlock.cpp:2252
> +    if (!m_globalData->interpreter->classicEnabled() && !!numberOfInstructions() && getJITType() < JITCode::bottomTierJIT()) {

Can we merge this with the CLASSIC_INTERPRETER case?
Comment 3 Oliver Hunt 2012-07-13 16:12:22 PDT 
Committed r122640: <http://trac.webkit.org/changeset/122640>
Comment 4 WebKit Review Bot 2012-07-13 16:25:28 PDT 
Re-opened since this is blocked by 91298
Comment 5 Oliver Hunt 2012-07-13 16:33:20 PDT 
Okay, my understanding of what the LLInt expects to do is clearly flawed.  Back to trying to work out how it's failing to unlink the resolves correctly.