Bug 121141

Summary:

Int32ToDouble should be predicted SpecInt48 and predictions should have nothing to do with constant folding

Product:

WebKit

Reporter:

Filip Pizlo <fpizlo>

Component:

JavaScriptCore

Assignee:

Filip Pizlo <fpizlo>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

barraclough, buildbot, ggaren, graouts, mark.lam, mhahnenberg, msaboff, oliver, rniwa, sam

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

All

Bug Depends on:

Bug Blocks:

121064

Attachments:

Description	Flags
the patch	ggaren: review+
even better patch	oliver: review+, buildbot: commit-queue-
Archive of layout-test-results from webkit-ews-10 for mac-mountainlion-wk2	none

Filip Pizlo

Reported 2013-09-10 22:43:35 PDT

Just changing Int32ToDouble to be predicted SpecInt48 breaks constant folding on that node because of soooper old code that prevented constant folding on mismatched predictions. Kill that code.

Attachments
the patch (26.05 KB, patch) 2013-09-10 22:48 PDT, Filip Pizlo	ggaren: review+	Details Formatted Diff Diff
even better patch (24.99 KB, patch) 2013-09-11 13:25 PDT, Filip Pizlo	oliver: review+ buildbot: commit-queue-	Details Formatted Diff Diff
Archive of layout-test-results from webkit-ews-10 for mac-mountainlion-wk2 (523.54 KB, application/zip) 2013-09-11 13:55 PDT, Build Bot	no flags	Details
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.

Filip Pizlo

Comment 1 2013-09-10 22:48:34 PDT

Created attachment 211282 [details] the patch

Geoffrey Garen

Comment 2 2013-09-11 08:59:04 PDT

Comment on attachment 211282 [details] the patch r=me

Filip Pizlo

Comment 3 2013-09-11 13:25:24 PDT

Created attachment 211345 [details] even better patch Introduce setConstant(node, value) method that does both of these things: forNode(node).set(m_graph, value) m_state.setFoundConstants(true)

Build Bot

Comment 4 2013-09-11 13:55:17 PDT

Comment on attachment 211345 [details] even better patch Attachment 211345 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.appspot.com/results/1738828 New failing tests: js/dfg-constant-fold-misprediction.html

Build Bot

Comment 5 2013-09-11 13:55:19 PDT

Created attachment 211347 [details] Archive of layout-test-results from webkit-ews-10 for mac-mountainlion-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: webkit-ews-10 Port: mac-mountainlion-wk2 Platform: Mac OS X 10.8.4

Filip Pizlo

Comment 6 2013-09-11 14:24:25 PDT

(In reply to comment #4) > (From update of attachment 211345 [details]) > Attachment 211345 [details] did not pass mac-wk2-ews (mac-wk2): > Output: http://webkit-queues.appspot.com/results/1738828 > > New failing tests: > js/dfg-constant-fold-misprediction.html Because I forgot to upload the test change with this patch like I did with the last one.

Filip Pizlo

Comment 7 2013-09-11 15:05:49 PDT

Landed in http://trac.webkit.org/changeset/155567

Antoine Quint

Comment 8 2013-11-02 06:47:23 PDT

I think this introduced a regression, see https://bugs.webkit.org/show_bug.cgi?id=123664.

Note You need to log in before you can comment on or make changes to this bug.