Bug 64866 - Optimizing shadow blur filter to ARM-neon SIMD instruction set
Summary: Optimizing shadow blur filter to ARM-neon SIMD instruction set
Status: RESOLVED WONTFIX
Alias: None
Product: WebKit
Classification: Unclassified
Component: SVG (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Zoltan Herczeg
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-07-20 06:15 PDT by Zoltan Herczeg
Modified: 2012-02-01 23:28 PST (History)
4 users (show)

See Also:

Attachments
patch (30.71 KB, patch)
2011-07-20 06:30 PDT, Zoltan Herczeg 		oliver: review+
Details | Formatted Diff | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Zoltan Herczeg 2011-07-20 06:15:40 PDT 
Since many ports are moving to ShadowBlur, it is time to make it fast!
Comment 1 Zoltan Herczeg 2011-07-20 06:30:00 PDT 
Created attachment 101458 [details]
patch
Comment 2 WebKit Review Bot 2011-07-20 06:33:25 PDT 
Attachment 101458 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/CMakeLists.txt', u'Source/W..." exit_code: 1

Source/WebCore/platform/graphics/filters/arm/ShadowBlurNEON.cpp:28:  Found other header before a header this file implements. Should be: config.h, primary header, blank line, and then alphabetically sorted.  [build/include_order] [4]
Source/WebCore/platform/graphics/filters/arm/ShadowBlurNEON.h:27:  #ifndef header guard has wrong style, please use: ShadowBlurNEON_h  [build/header_guard] [5]
Total errors found: 2 in 10 files


If any of these errors are false positives, please file a bug against check-webkit-style.
Comment 3 Oliver Hunt 2011-09-23 19:05:55 PDT 
Comment on attachment 101458 [details]
patch

Does the compiler not have intrinsics for NEON instructions?
Comment 4 Oliver Hunt 2011-09-23 19:08:59 PDT 
Comment on attachment 101458 [details]
patch

whoops, r=me regardless of the answer to that question.
Comment 5 Dirk Schulze 2012-02-01 22:00:27 PST 
What is the status of this bug. Wasn't it landed? Can we close it now? Would be great if you can add the commit message.
Comment 6 Zoltan Herczeg 2012-02-01 23:28:25 PST 
According to our recent measurements the gain is not enough to worth this work, so unless we find a better optimization I would not land it.