Remove $outputDir and $outputHeadersDir from global variables of code generators

[Kentaro Hara]

$outputDir and $outputHeadersDir are used in WriteData() only. We don't need to use global variables for them.

[Kentaro Hara]

Created attachment 174662 [details]
Patch

[Early Warning System Bot]

Comment on attachment 174662 [details]
Patch

Attachment 174662 [details] did not pass qt-wk2-ews (qt):
Output: http://queues.webkit.org/results/14858472

[Build Bot]

Comment on attachment 174662 [details]
Patch

Attachment 174662 [details] did not pass win-ews (win):
Output: http://queues.webkit.org/results/14873136

[Adam Barth]

Comment on attachment 174662 [details]
Patch

Looks like you have come compile problems...

[Kentaro Hara]

Created attachment 174955 [details]
another try

[Early Warning System Bot]

Comment on attachment 174955 [details]
another try

Attachment 174955 [details] did not pass qt-wk2-ews (qt):
Output: http://queues.webkit.org/results/14901167

[Kentaro Hara]

Created attachment 174958 [details]
another try

[Kentaro Hara]

Comment on attachment 174958 [details]
another try

Now build bots look happy. r?

[Adam Barth]

Based on what timothy wrote in another bug, this shouldn't be a problem for the Safari builds because they don't use CodeGenerator.pm.

[Kentaro Hara]

Created attachment 175311 [details]
patch for landing

[WebKit Review Bot]

Comment on attachment 175311 [details]
patch for landing

Clearing flags on attachment: 175311

Committed r135340: <http://trac.webkit.org/changeset/135340>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Tim Horton]

(In reply to comment #9)
> Based on what timothy wrote in another bug, this shouldn't be a problem for the Safari builds because they don't use CodeGenerator.pm.

Erg, this broke stuff too. I'll look and see if it's easy to fix, otherwise I might have to roll this out till next week.

[Kentaro Hara]

(In reply to comment #13)
> (In reply to comment #9)
> > Based on what timothy wrote in another bug, this shouldn't be a problem for the Safari builds because they don't use CodeGenerator.pm.
> 
> Erg, this broke stuff too. I'll look and see if it's easy to fix, otherwise I might have to roll this out till next week.

sorry, I can roll out the patch. Ping me if needed.

[Kentaro Hara]

Reverted r135340 for reason:

It broke Safari's internal build

Committed r135387: <http://trac.webkit.org/changeset/135387>

[Kentaro Hara]

I will reland the patch next week when Apple guys can take care of Safari's internal build.

[Csaba Osztrogonác]

(In reply to comment #15)
> Reverted r135340 for reason:
> 
> It broke Safari's internal build
> 
> Committed r135387: <http://trac.webkit.org/changeset/135387>

And the rollout broke the bining tests. Could you update it too?

[Adam Barth]

> Erg, this broke stuff too.

I'm sorry this patch broke things.  I must have misunderstood what you wrote in <https://bugs.webkit.org/show_bug.cgi?id=102671#c6>.  :(

[Timothy Hatcher]

It turns out we have two scripts that use WebCore binging scripts. One uses generate-bindings.pl. The other uses CodeGenerator.pm.

[Adam Barth]

Thanks for the clarification and your patience.

[Timothy Hatcher]

More details. We have our own CodeGeneratorSafari.pm too. So we need to keep up will any changes to CodeGenerator.pm and its interface. I'm working on making this work with TOT WebCore now.

[Timothy Hatcher]

Go ahead and land this again. I know the fixes we need to make.

[WebKit Review Bot]

Comment on attachment 175311 [details]
patch for landing

Clearing flags on attachment: 175311

Committed r135426: <http://trac.webkit.org/changeset/135426>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Kentaro Hara]

Thanks for the fix!