[webkit-dev] Changing data structures in webkit

2013-08-05 Thread Abhishek Bichhawat
Hi,

I tried changing the JSValue.h and JSCell.h classes by adding new fields. It 
runs fine in some of the cases, but in some other cases an assert fails when it 
checks if it is an m_cell in isString() function of JSString/JSValue. I 
suspect that it is related to some garbage collection as the website is loading 
and then it fails. I was wondering if there are some size checks that need to 
be fulfilled for this to go through fine and GC to run fine.

Thanks and Regards,
Abhishek___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] Windows Build Broken?

2013-08-05 Thread kwliu




--
View this message in context: 
http://mac-os-forge.2317878.n4.nabble.com/Windows-Build-Broken-tp223127p224861.html
Sent from the Webkit mailing list archive at Nabble.com.
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] Windows Build Broken?

2013-08-05 Thread kwliu
//I commented above lines , but i was encountering another problem 
msvcr80.dll missing .
You can Change the WinLauncher project's manifest File configuration from
Yes to No to avoid this error. 
I examined the generated mainfest file and found that msvc CRT 8.0 was
referenced in it. 


Brent Fulgham-4 wrote
 Hi Rakesh,
 
 On Aug 2, 2013, at 1:29 AM, Rakesh Sadhu lt;

 rakeshsadhu@

 gt; wrote:
 
 But I am facing one more  problem , my release winlauncher launches
 properly whereas 
 debug winlauncher fails with the message Failed to determine path to AAS
 Directory”.
 
 WebKit requires certain support DLLs to run properly (e.g., libxml2.dll,
 zlib1.dll, cairo.dll).  For the standard Apple build, these live in the
 Program Files directory and will be installed on your system if you have
 iTunes. We do not have an installer or canonical location for the WinCairo
 port’s support libraries; usually they are just copied to the same
 directory as the application.
 
 In the file DLLLauncherMain.cpp ::moidfypath 
 
 //#if defined(_M_X64)
 static const wstring pathPrefix = LC:\\Program Files\\Common
 Files\\Apple\\Apple Application Support;
 //#else
// static const wstring pathPrefix = LC:\\Program Files (x86)\\Common
 Files\\Apple\\Apple Application Support;
 //#endif
 
 I commented above lines , but i was encountering another problem 
 msvcr80.dll missing .
 
 You could also just return true without doing anything, which is probably
 the right thing to do for the WinCairo port.
 
 The missing “msvcr80.dll” makes me think you have some VS2005-based
 support library DLLs in your path. The current set of WinCairo support
 dlls are all build with VS2010, and do not link against this old runtime.
 
 -Brent
 
 
 
 
 
 
 So is it like , my windows somehow got corrupted and i need to redo all
 or is there any work around for this .
 
 regards
 RSadhu
 
 
 Subject: Re: [webkit-dev] Windows Build Broken?
 From: 

 bfulgham@

 Date: Thu, 1 Aug 2013 09:50:11 -0700
 CC: 

 mital.d.vora@

 ; 

 webkit-dev@.webkit

 To: 

 rakeshsadhu@

 
 Please review your build log.  The problem is the following:
 
 fatal error C1083: Cannot open include file: 'd3dx9.h': No such file or
 directory”
 
 You do not have the DirectX SDK installed, as documented in Step 4 of
 http://www.webkit.org/building/tools.html.
 
 -Brent
 
 
 On Aug 1, 2013, at 1:46 AM, Rakesh Sadhu lt;

 rakeshsadhu@

 gt; wrote:
 
 Hi ,
 I couldn't fix the problem  with vs2005.
 Now I am trying to build with vs2010 and it fails too.
 here is the pasebin link : http://pastebin.com/W0YyGXZG
 Kindly guide.
 
 
 regards
 RSadhu
 
 
 Date: Tue, 23 Jul 2013 07:21:23 +0530
 Subject: Re: [webkit-dev] Windows Build Broken?
 From: 

 mital.d.vora@

 To: 

 rakeshsadhu@

 CC: 

 webkit-dev@.webkit

 
 Sorry Its not exactly the same error. but Its similar error which is
 present in build-webkit
 
 
 I am trying to build using Visual Studio 2010 Express using commandline
 cygwin tools
 
 Thanks,
 
 Mital Vora.
 
 
 On Tue, Jul 23, 2013 at 7:17 AM, Mital Vora lt;

 mital.d.vora@

 gt; wrote:
 Hi, 
 
 For the WinCairo Build as well these are the exact errors coming in the
 buildbot
 
 Building results into:
 /cygdrive/c/Projects/BuildSlave/win-cairo-release/build/WebKitBuild
 WEBKIT_OUTPUTDIR is set to:
 C:\Projects\BuildSlave\win-cairo-release\build\WebKitBuild
 WEBKIT_LIBRARIES is set to: 
 Use of uninitialized value in print at
 /cygdrive/c/Projects/BuildSlave/win-cairo-release/build/Tools/Scripts/webkitdirs.pm
 line 1644.
 
 http://build.webkit.org/builders/WinCairo%20Release/builds/28414/steps/compile-webkit/logs/stdio
 
 
 
 Windows and WinCairo Build from commandline seems broken.. can anybody
 help fixing this.
 
 Regards,
 
 Mital Vora.
 
 
 On Sun, Jul 21, 2013 at 9:07 PM, Rakesh Sadhu lt;

 rakeshsadhu@

 gt; wrote:
 Hi ,
 I updated my webkit source code and tried to build it on windows machine
 , and it fails to compile, with following error
 
 Tools/Scripts/build-webkit --debug
 
 Installing WebKitSupportLibrary...
 The WebKitSupportLibrary has been sucessfully installed in
  /home/rsadhu/trunk/WebKitLibraries/win
 *
 Cannot find '/cygdrive/c/Program Files (x86)/Microsoft Visual Studio
 10.0/Common7/IDE/VCExpress.exe'
 Please execute the file 'vcvars32.bat' from
 'C:\Program Files (x86)\Microsoft Visual Studio 8\VC\bin\'
 to setup the necessary environment variables.
 *
 Died at /home/rsadhu/trunk/Tools/Scripts/webkitdirs.pm line 1627.
 
 
 Any Pointers on this?
 
 
 ___
 webkit-dev mailing list
 

 webkit-dev@.webkit

 https://lists.webkit.org/mailman/listinfo/webkit-dev
 
 
 
 ___
 webkit-dev mailing list
 

 webkit-dev@.webkit

 https://lists.webkit.org/mailman/listinfo/webkit-dev
 
 
 ___
 webkit-dev mailing 

Re: [webkit-dev] Changing data structures in webkit

2013-08-05 Thread Filip Pizlo
JSValue must be 64 bits, no more, no less. This is a hard requirement for all 
of our execution engines. 

-Filip

 On Aug 5, 2013, at 6:57 AM, Abhishek Bichhawat 
 abhishek_bichha...@yahoo.co.in wrote:
 
 Hi,
 
 I tried changing the JSValue.h and JSCell.h classes by adding new fields. It 
 runs fine in some of the cases, but in some other cases an assert fails when 
 it checks if it is an m_cell in isString() function of JSString/JSValue. I 
 suspect that it is related to some garbage collection as the website is 
 loading and then it fails. I was wondering if there are some size checks that 
 need to be fulfilled for this to go through fine and GC to run fine.
 
 Thanks and Regards,
 Abhishek
 ___
 webkit-dev mailing list
 webkit-dev@lists.webkit.org
 https://lists.webkit.org/mailman/listinfo/webkit-dev
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] WebGL on Windows

2013-08-05 Thread Alex Christensen
On Aug 4, 2013, at 8:07 PM, Xueqing Huang xqhuang.web...@gmail.com wrote:
 1. Do you have a plan to use ACCELERATED_COMPOSITING on Windows port? I had 
 enabled WebGL on Windows port via Angle, it seems like WebGL without 
 ACCELERATED_COMPOSITING has a poor performance.
The AppleWin port uses accelerated compositing, but WinCairo does not right 
now.  I have not worked on accelerated compositing.

 2. Do you want to make Angle run on seperated process like Chromium's GPU 
 process? If not, maybe WebCore will have more crash.
I have not looked into this.  I don’t think my work will make WebCore crash 
more unless you’re using it with unsafe WebGL content.

Alex Christensen
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] WebGL on Windows

2013-08-05 Thread Alex Christensen
 1. Do you have a plan to use ACCELERATED_COMPOSITING on Windows port? I had 
 enabled WebGL on Windows port via Angle, it seems like WebGL without 
 ACCELERATED_COMPOSITING has a poor performance.
One of my early experimental builds has much higher performance than what is 
currently in WebKit without using accelerated compositing.  I think I just 
forgot to enable optimization in my ANGLE project files, which can be easily 
fixed.

Alex
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


[webkit-dev] Proposal: Remove Microdata support

2013-08-05 Thread Sam Weinig
Hello,

I’d like to propose removing support for the Microdata JS API 
(http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html) 
from WebKit.  The specification has not gotten much adoption from either 
implementors or from site authors.  Additionally, its use cases can all be 
addressed in author scripts, and if we see it becoming popular we can always 
bring it back.

The Mac port has never enabled the feature, and Blink has recently ripped out 
their support. 

Thoughts?

-Sam

(I have done the work to rip it out, but it is not up for review yet - 
https://webkit.org/b/119480)
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] Proposal: Remove Microdata support

2013-08-05 Thread Brady Eidson
+1.

I see no benefit to keeping the code around right now.

~Brady

On Aug 5, 2013, at 1:16 PM, Sam Weinig wei...@apple.com wrote:

 Hello,
 
 I’d like to propose removing support for the Microdata JS API 
 (http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html) 
 from WebKit.  The specification has not gotten much adoption from either 
 implementors or from site authors.  Additionally, its use cases can all be 
 addressed in author scripts, and if we see it becoming popular we can always 
 bring it back.
 
 The Mac port has never enabled the feature, and Blink has recently ripped out 
 their support. 
 
 Thoughts?
 
 -Sam
 
 (I have done the work to rip it out, but it is not up for review yet - 
 https://webkit.org/b/119480)
 ___
 webkit-dev mailing list
 webkit-dev@lists.webkit.org
 https://lists.webkit.org/mailman/listinfo/webkit-dev

___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] WebGL on Windows

2013-08-05 Thread Xueqing Huang
2013/8/6 Alex Christensen achristen...@apple.com


  2. Do you want to make Angle run on seperated process like Chromium's
 GPU process? If not, maybe WebCore will have more crash.
 I have not looked into this.  I don’t think my work will make WebCore
 crash more unless you’re using it with unsafe WebGL content.


That what I mean, anyway, we should not make WebKit crash although web
developer write some unsafe WebGL code.
As I know, most WebGL crash because driver problem, which out of our
control.
___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev


Re: [webkit-dev] Upcoming changes to WebKit's logging functionality

2013-08-05 Thread Mark Rowe

On 2013-08-02, at 4:39 PM, Mark Rowe mr...@apple.com wrote:

 As part of improving WebKit's logging subsystem on OS X 
 (https://webkit.org/b/119031) I'm refactoring the code in WebCore, WebKit 
 and WebKit2 that initializes logging channels. I'm doing this to reduce the 
 need to touch several different pieces of code in order to add a new logging 
 channel, and so that more of the initialization code can be shared across 
 platforms. The result of this is that all ports will now share a common 
 syntax for the string specifying which logging channels to enable.
 
 Logging channels may now be enabled by providing a comma-separated list of 
 channel names, with the special all name enabling all channels. Channel 
 names prefixed with a leading - will result in the named channel being 
 disabled. For instance, specifying all,-history,-loading will result in all 
 logging channels except for history and loading being enabled.
 
 For OS X developers, this will also change the name of the user defaults used 
 to enable logging. This is done to allow the old user defaults to remain set 
 for those people that need to switch between version of WebKit before and 
 after this change. Where the old user default keys were WebCoreLogLevel, 
 WebKitLogLevel and WebKit2LogLevel, the new user default keys are 
 WebCoreLogging, WebKitLogging and WebKit2Logging.
 
 For developers of other ports, the syntax used to enable logging channels may 
 change for you. Some ports had been using a space-separated list of channel 
 names rather than a comma-separated list. Please let me know if you feel that 
 this change will cause you pain. I've made a best-effort attempt to keep 
 other ports compiling, but since my brain only supports a subset of C++11 
 it's likely that I've introduced some build breakages. The EWS appears to 
 build with logging disabled so I've not been able to use it for testing this 
 change. I'd appreciate it if maintainers of other ports could take a few 
 minutes to check whether the most recent patch attached to 
 https://webkit.org/b/119031 builds on their platform and let me know if it 
 works or what needs to be fixed.
 
 I'll be holding off on landing these changes for a few days to gives folks an 
 opportunity to provide feedback. I’ll also be keeping an eye on the bots 
 after I land the change in case I do happen to break any builds.

I landed this change in r153736, with Windows build fixes in subsequent 
revisions. I’ll continue to keep an eye on the other build bots in case it 
causes problems for any other ports.

- Mark

___
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev