Bug 124733 - It should be possible for a LinkBuffer to outlive the MacroAssembler and still be useful
Summary: It should be possible for a LinkBuffer to outlive the MacroAssembler and stil...
Status: RESOLVED FIXED
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 treegraph
 
Reported: 2013-11-21 12:07 PST by Filip Pizlo
Modified: 2014-07-13 19:13 PDT (History)
10 users (show)

See Also:


Attachments
work in progress (7.30 KB, patch)
2013-11-21 12:17 PST, Filip Pizlo
no flags Details | Formatted Diff | Diff
work in progress (7.35 KB, patch)
2013-11-21 12:37 PST, Filip Pizlo
no flags Details | Formatted Diff | Diff
the patch (7.35 KB, patch)
2014-01-09 16:03 PST, Filip Pizlo
no flags Details | Formatted Diff | Diff
Patch (7.39 KB, patch)
2014-02-21 18:34 PST, Filip Pizlo
no flags Details | Formatted Diff | Diff
Patch (8.70 KB, patch)
2014-02-21 18:36 PST, Filip Pizlo
oliver: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Filip Pizlo 2013-11-21 12:07:25 PST
Patch forthcoming.
Comment 1 Filip Pizlo 2013-11-21 12:17:35 PST
Created attachment 217591 [details]
work in progress
Comment 2 Filip Pizlo 2013-11-21 12:37:21 PST
Created attachment 217599 [details]
work in progress
Comment 3 Filip Pizlo 2014-01-09 16:03:19 PST
Created attachment 220782 [details]
the patch
Comment 4 WebKit Commit Bot 2014-01-09 16:06:57 PST
Attachment 220782 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/JavaScriptCore/ChangeLog', u'Source/JavaScriptCore/assembler/ARM64Assembler.h', u'Source/JavaScriptCore/assembler/ARMv7Assembler.h', u'Source/JavaScriptCore/assembler/AssemblerBuffer.h', u'Source/JavaScriptCore/assembler/LinkBuffer.h', u'Source/JavaScriptCore/assembler/MacroAssemblerARM64.h', u'Source/JavaScriptCore/assembler/MacroAssemblerARMv7.h', '--commit-queue']" exit_code: 1
ERROR: Source/JavaScriptCore/assembler/AssemblerBuffer.h:65:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
Total errors found: 1 in 4 files


If any of these errors are false positives, please file a bug against check-webkit-style.
Comment 5 Filip Pizlo 2014-02-21 18:34:58 PST
Created attachment 224940 [details]
Patch
Comment 6 Filip Pizlo 2014-02-21 18:36:00 PST
Created attachment 224941 [details]
Patch
Comment 7 WebKit Commit Bot 2014-02-21 18:37:39 PST
Attachment 224941 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/assembler/AssemblerBuffer.h:65:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
Total errors found: 1 in 5 files


If any of these errors are false positives, please file a bug against check-webkit-style.
Comment 8 Filip Pizlo 2014-02-22 11:50:25 PST
Landed in http://trac.webkit.org/changeset/164548
Comment 10 Benjamin Poulain 2014-07-13 19:13:59 PDT
(In reply to comment #9)
> Looks like this was 7-10% regression on Dromaeo/dom-query?
> 
> https://perf.webkit.org/#mode=charts&chartList=%5B%5B%22mac-mavericks%22%2C%22Dromaeo%2Fdom-query%3ARuns%22%5D%2C%5B%22mac-mountainlion%22%2C%22Dromaeo%2Fdom-query%3ARuns%22%5D%5D&days=63&zoom=%5B1392794639427.6333%2C1393190255845.2073%5D

I looked into this today, this regression is very bad for PLT on ARM.

I have started fixing in https://bugs.webkit.org/show_bug.cgi?id=134659. I'll finish that tomorrow when I have access to all the hardware.