Bug 150655

Summary: we're incorrectly adjusting a stack location with respect to the localsOffset in FTLCompile
Product: WebKit Reporter: Saam Barati <saam>
Component: JavaScriptCoreAssignee: Saam Barati <saam>
Status: RESOLVED FIXED    
Severity: Normal CC: benjamin, commit-queue, fpizlo, ggaren, gskachkov, keith_miller, mark.lam, msaboff, oliver, sukolsak, ysuzuki
Priority: P2    
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
patch none

Saam Barati
Reported 2015-10-28 18:42:56 PDT
we're recomputing this for a *OSRExitDescriptor* for every one of its *OSRExits*. This is having a multiplicative effect on offsets. I'm amazed this didn't break any of our tests. I think I introduced this in: https://bugs.webkit.org/show_bug.cgi?id=149970
Attachments
patch (2.55 KB, patch)
2015-10-28 19:12 PDT, Saam Barati 		no flags
DetailsFormatted DiffDiff
View All Add attachment proposed patch, testcase, etc.
Saam Barati
Comment 1 2015-10-28 19:12:06 PDT
Created attachment 264287 [details] patch
Filip Pizlo
Comment 2 2015-10-28 19:19:21 PDT
Nice catch!
Geoffrey Garen
Comment 3 2015-10-28 19:29:04 PDT
Comment on attachment 264287 [details] patch Can haz regression test?
Saam Barati
Comment 4 2015-10-28 22:20:21 PDT
(In reply to comment #3) > Comment on attachment 264287 [details] > patch > > Can haz regression test? I'll try to see if I can come up with one before landing this. We will definitely have one once FTL try/catch lands but it will be nice if I can come up with a small and simple reproducible program. This might be a bit tricky to get right. If I can't come up with a small program, I think coming up with a small program for FTL try/catch will be easier because of the nature of OSR exits from from calls and get/put_by_ids.
Saam Barati
Comment 5 2015-10-29 11:28:53 PDT
(In reply to comment #4) > (In reply to comment #3) > > Comment on attachment 264287 [details] > > patch > > > > Can haz regression test? > I'll try to see if I can come up with one before landing this. > We will definitely have one once FTL try/catch lands but > it will be nice if I can come up with a small and simple > reproducible program. This might be a bit tricky > to get right. If I can't come up with a small program, > I think coming up with a small program for FTL try/catch > will be easier because of the nature of OSR exits from > from calls and get/put_by_ids. I couldn't get a test case to reproduce this easily. I'm going to defer until FTL try/catch where we definitely will have a test case.
Geoffrey Garen
Comment 6 2015-10-29 11:35:40 PDT
Okeedokee.
WebKit Commit Bot
Comment 7 2015-10-29 12:15:37 PDT
Comment on attachment 264287 [details] patch Clearing flags on attachment: 264287 Committed r191744: <http://trac.webkit.org/changeset/191744>
WebKit Commit Bot
Comment 8 2015-10-29 12:15:46 PDT
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.