Bug 114930 - fourthTier: FTL should call LLVMFinalizeAllObjects()
Summary: fourthTier: FTL should call LLVMFinalizeAllObjects()
Status: RESOLVED INVALID
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: All All
: P2 Normal
Assignee: Filip Pizlo
URL:
Keywords:
Depends on:
Blocks: 112840
  Show dependency tree / graph
 
Reported: 2013-04-21 19:05 PDT by Filip Pizlo
Modified: 2013-04-29 21:51 PDT (History)
7 users (show)

See Also:

Attachments
the patch (1.60 KB, patch)
2013-04-21 19:07 PDT, Filip Pizlo 		no flags Details | Formatted Diff | Diff
corresponding llvm patch (6.04 KB, patch)
2013-04-21 19:08 PDT, Filip Pizlo 		no flags 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 Filip Pizlo 2013-04-21 19:05:46 PDT 
This is needed for the corresponding MCJIT llvm-c patch I'm submitting to llvm-commits.
Comment 1 Filip Pizlo 2013-04-21 19:07:30 PDT 
Created attachment 198980 [details]
the patch

I will wait with landing this until there is a clear story for my llvm-c patch.
Comment 2 Filip Pizlo 2013-04-21 19:08:34 PDT 
Created attachment 198981 [details]
corresponding llvm patch
Comment 3 Filip Pizlo 2013-04-22 15:00:21 PDT 
Comment on attachment 198980 [details]
the patch

It looks like LLVMFinalizeAllObjects() is not going to be a thing.  Clearing r?.
Comment 4 Filip Pizlo 2013-04-22 15:01:22 PDT 
It looks like this is again taking a different direction; this time, the LLVM C API will ensure that the MCJIT finalizes all objects before you use them.  I'm leaving this bug open for now in case this again changes, but probably this will get resolved as wontfix.
Comment 5 Filip Pizlo 2013-04-29 21:51:00 PDT 
Nope, it looks like LLVM isn't going to require this.