Bug 129420

Summary: Incorrect V_JITOperation_EJ call implementation in DFG for 32-bit ports.
Product: WebKit Reporter: Julien Brianceau <jbriance>
Component: JavaScriptCoreAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: commit-queue, ggaren, mark.lam, ossy, rgabor, yong.li.webkit, zherczeg
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Bug Depends on:    
Bug Blocks: 108645    
Attachments:
Description Flags
Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. none

Julien Brianceau
Reported 2014-02-27 05:03:12 PST
r162652 (http://trac.webkit.org/changeset/162652) introduced V_JITOperation_EJ prototype. Its implementation seems to be incorrect for 32-bit ports.
Attachments
Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. (1.32 KB, patch)
2014-02-27 05:09 PST, Julien Brianceau 		no flags
DetailsFormatted DiffDiff
View All Add attachment proposed patch, testcase, etc.
Julien Brianceau
Comment 1 2014-02-27 05:09:15 PST
Created attachment 225357 [details] Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports.
Zoltan Herczeg
Comment 2 2014-02-27 05:23:03 PST
Was this a speculative fix?
Julien Brianceau
Comment 3 2014-02-27 05:26:05 PST
(In reply to comment #2) > Was this a speculative fix? Yes, I only checked compilation. I'm going to retrieve my env to launch run-layout-jsc tests on my arm traditional board.
Julien Brianceau
Comment 4 2014-02-27 05:48:50 PST
(In reply to comment #3) > Yes, I only checked compilation. I'm going to retrieve my env to launch run-layout-jsc tests on my arm traditional board. run-layout-jsc reports the same results on my board with and without this patch: not better, not worse.
Zoltan Herczeg
Comment 5 2014-02-27 06:02:29 PST
> run-layout-jsc reports the same results on my board with and without this patch: not better, not worse. Interesting. So this patch is needed or not? I suspect this code is not triggered.
Julien Brianceau
Comment 6 2014-02-27 06:08:08 PST
(In reply to comment #5) > > run-layout-jsc reports the same results on my board with and without this patch: not better, not worse. > > Interesting. So this patch is needed or not? I suspect this code is not triggered. You're right. If I remove this function, compilation is still ok so it's not even compiled actually :) What's the best choice, removing it or fixing it in case we'd need it later?
Zoltan Herczeg
Comment 7 2014-02-27 06:16:06 PST
I think we should trigger it somehow, otherwise we don't know whether the fix is correct :(
Julien Brianceau
Comment 8 2014-02-27 07:40:42 PST
(In reply to comment #7) > I think we should trigger it somehow, otherwise we don't know whether the fix is correct :( As it's like http://trac.webkit.org/changeset/157797 for the tag/payload swap, I'm pretty sure the fix is correct. However if the code is not compiled, I won't be opposed to just remove it.
Yong Li
Comment 9 2014-02-27 08:05:58 PST
Comment on attachment 225357 [details] Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. Looks right to me
Yong Li
Comment 10 2014-02-27 08:08:46 PST
Comment on attachment 225357 [details] Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. Withdraw r+. A test case should have helped to find this problem.
Geoffrey Garen
Comment 11 2014-02-27 10:11:20 PST
Comment on attachment 225357 [details] Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. r=me
WebKit Commit Bot
Comment 12 2014-02-27 10:42:23 PST
Comment on attachment 225357 [details] Fix V_JITOperation_EJ call implementation in DFG for 32-bit ports. Clearing flags on attachment: 225357 Committed r164813: <http://trac.webkit.org/changeset/164813>
WebKit Commit Bot
Comment 13 2014-02-27 10:42:26 PST
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.