Bug 119962

Summary: DFG should inline typedArray.byteOffset
Product: WebKit Reporter: Filip Pizlo <fpizlo>
Component: JavaScriptCoreAssignee: Filip Pizlo <fpizlo>
Status: RESOLVED FIXED    
Severity: Normal CC: barraclough, buildbot, ggaren, mark.lam, mhahnenberg, msaboff, oliver, rniwa, sam
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: All   
OS: All   
Bug Depends on: 119961    
Bug Blocks:    
Attachments:
Description Flags
work in progress
none
the patch
oliver: review+
almost patch for landing
buildbot: commit-queue-
Archive of layout-test-results from webkit-ews-16 for mac-mountainlion-wk2 none

Filip Pizlo
Reported 2013-08-17 14:45:46 PDT
Patch forthcoming.
Attachments
work in progress (22.10 KB, patch)
2013-08-17 14:46 PDT, Filip Pizlo 		no flags
DetailsFormatted DiffDiff
the patch (25.53 KB, patch)
2013-08-17 21:23 PDT, Filip Pizlo 		oliver: review+
DetailsFormatted DiffDiff
almost patch for landing (27.37 KB, patch)
2013-08-18 22:00 PDT, Filip Pizlo 		buildbot: commit-queue-
DetailsFormatted DiffDiff
Archive of layout-test-results from webkit-ews-16 for mac-mountainlion-wk2 (846.11 KB, application/zip)
2013-08-19 03:11 PDT, Build Bot 		no flags
Details
Show Obsolete (2) View All Add attachment proposed patch, testcase, etc.
Filip Pizlo
Comment 1 2013-08-17 14:46:35 PDT
Created attachment 209011 [details] work in progress Almost done.
Filip Pizlo
Comment 2 2013-08-17 21:23:17 PDT
Created attachment 209017 [details] the patch
Oliver Hunt
Comment 3 2013-08-18 10:04:08 PDT
Comment on attachment 209017 [details] the patch View in context: https://bugs.webkit.org/attachment.cgi?id=209017&action=review > LayoutTests/fast/js/regress/script-tests/ArrayBuffer-Int32Array-byteOffset.js:9 > +for (var i = 0; i < 1000000; ++i) { > + result += array1.byteOffset; > + result += array2.byteOffset; > +} Is it possible to have a similar test that neuters midway through?
Filip Pizlo
Comment 4 2013-08-18 21:48:05 PDT
(In reply to comment #3) > (From update of attachment 209017 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=209017&action=review > > > LayoutTests/fast/js/regress/script-tests/ArrayBuffer-Int32Array-byteOffset.js:9 > > +for (var i = 0; i < 1000000; ++i) { > > + result += array1.byteOffset; > > + result += array2.byteOffset; > > +} > > Is it possible to have a similar test that neuters midway through? Oh yeah! I'm not sure it's meaningful to have a performance test (all operations on views run the same code whether it's neutered or not), but I definitely should add a correctness test to make sure that when neutered, the pointer indirections and subtractions work.
Filip Pizlo
Comment 5 2013-08-18 22:00:39 PDT
Created attachment 209053 [details] almost patch for landing
Filip Pizlo
Comment 6 2013-08-18 22:01:14 PDT
(In reply to comment #5) > Created an attachment (id=209053) [details] > almost patch for landing Still need to generate expected file for one of the new tests.
Build Bot
Comment 7 2013-08-19 03:11:16 PDT
Comment on attachment 209053 [details] almost patch for landing Attachment 209053 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/1478819 New failing tests: fast/js/dfg-byteOffset-neuter.html
Build Bot
Comment 8 2013-08-19 03:11:18 PDT
Created attachment 209070 [details] Archive of layout-test-results from webkit-ews-16 for mac-mountainlion-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: webkit-ews-16 Port: mac-mountainlion-wk2 Platform: Mac OS X 10.8.4
Filip Pizlo
Comment 9 2013-08-19 16:15:43 PDT
Landed in http://trac.webkit.org/changeset/154305
Note You need to log in before you can comment on or make changes to this bug.