mbox

[pull,request] webkit: Version bump to 1.11.5

Message ID CAG2jQ8gLL4NdmHcFA=ELMbFfHY1oGPGaXkvSaPk66sesh-0+6w@mail.gmail.com
State Superseded
Headers show

Pull-request

https://github.com/hwoarang/buildroot.git webkit

Message

Markos Chandras Sept. 3, 2013, 1:59 p.m. UTC
Hi,

The following pull request bumps webkit to 1.11.5. A few upstream
patches are necessary to
make it work properly. It also restores the build on MIPS which is
currently (1.9.6 and 1.11.5)
broken due to missing assembly functions. The following patches have been
applied:

- upstream fix for 32-bit autotools
- upstream fix for conditional gstreamer build of gtklauncher
- upstream fix for detecting the latest harfbuzz
- Update existing patch for disabling docs
- Update existing patch for execinfo_h
- upstream patch for the MIPS DFG support. This fixes
compilation problems on MIPS.

The patch was tested with MIPS32r2, x86_64 and ARM cortex-a9 configurations.

The following changes since commit b7c0041c31754a5679eb0a3e4bcf41f80d473ae2:

  kmod: does not support static builds (2013-09-02 23:22:14 +0200)

are available in the git repository at:

  https://github.com/hwoarang/buildroot.git webkit

for you to fetch changes up to 26a8323eeed2a3cea5b53917565d4a23d7aa0889:

  webkit: Version bump to 1.11.5 (2013-09-03 13:19:32 +0100)

----------------------------------------------------------------
Markos Chandras (1):
      webkit: Version bump to 1.11.5

 package/webkit/Config.in                                   |   13 +-
 package/webkit/webkit-build-fix-for-32-bit-autotools.patch |   34 +++
 package/webkit/webkit-build-fix-for-gtklauncher.patch      |   51 ++++
 package/webkit/webkit-detect-harfbuzz-icu.patch            |   28 +++
 package/webkit/webkit-disable-docrebase.patch              |   13 +-
 package/webkit/webkit-execinfo_h.patch                     |   20 +-
 package/webkit/webkit-mips-dfg.patch                       | 1490
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 package/webkit/webkit.mk                                   |   18 +-
 8 files changed, 1637 insertions(+), 30 deletions(-)
 create mode 100644 package/webkit/webkit-build-fix-for-32-bit-autotools.patch
 create mode 100644 package/webkit/webkit-build-fix-for-gtklauncher.patch
 create mode 100644 package/webkit/webkit-detect-harfbuzz-icu.patch
 create mode 100644 package/webkit/webkit-mips-dfg.patch

Comments

Arnout Vandecappelle Sept. 4, 2013, 4:49 p.m. UTC | #1
On 09/03/13 15:59, Markos Chandras wrote:
> Hi,
>
> The following pull request bumps webkit to 1.11.5. A few upstream
> patches are necessary to
> make it work properly. It also restores the build on MIPS which is
> currently (1.9.6 and 1.11.5)
> broken due to missing assembly functions. The following patches have been
> applied:
>
> - upstream fix for 32-bit autotools
> - upstream fix for conditional gstreamer build of gtklauncher
> - upstream fix for detecting the latest harfbuzz
> - Update existing patch for disabling docs
> - Update existing patch for execinfo_h
> - upstream patch for the MIPS DFG support. This fixes
> compilation problems on MIPS.
>
> The patch was tested with MIPS32r2, x86_64 and ARM cortex-a9 configurations.
>
> The following changes since commit b7c0041c31754a5679eb0a3e4bcf41f80d473ae2:
>
>    kmod: does not support static builds (2013-09-02 23:22:14 +0200)
>
> are available in the git repository at:
>
>    https://github.com/hwoarang/buildroot.git webkit
>
> for you to fetch changes up to 26a8323eeed2a3cea5b53917565d4a23d7aa0889:
>
>    webkit: Version bump to 1.11.5 (2013-09-03 13:19:32 +0100)
>
> ----------------------------------------------------------------
> Markos Chandras (1):
>        webkit: Version bump to 1.11.5
>
>   package/webkit/Config.in                                   |   13 +-
>   package/webkit/webkit-build-fix-for-32-bit-autotools.patch |   34 +++
>   package/webkit/webkit-build-fix-for-gtklauncher.patch      |   51 ++++
>   package/webkit/webkit-detect-harfbuzz-icu.patch            |   28 +++
>   package/webkit/webkit-disable-docrebase.patch              |   13 +-
>   package/webkit/webkit-execinfo_h.patch                     |   20 +-
>   package/webkit/webkit-mips-dfg.patch                       | 1490
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>   package/webkit/webkit.mk                                   |   18 +-

  Hi Markos,

  Any chance to manually remove the huge patch so that you can post this 
on the list? Otherwise, there's not much chance for a review...

  Is there a reason why you cannot download the upstream patches with 
WEBKIT_PATCH ?

  Regards,
  Arnout

>   8 files changed, 1637 insertions(+), 30 deletions(-)
>   create mode 100644 package/webkit/webkit-build-fix-for-32-bit-autotools.patch
>   create mode 100644 package/webkit/webkit-build-fix-for-gtklauncher.patch
>   create mode 100644 package/webkit/webkit-detect-harfbuzz-icu.patch
>   create mode 100644 package/webkit/webkit-mips-dfg.patch
>
>
Thomas Petazzoni Sept. 5, 2013, 8:01 a.m. UTC | #2
Dear Arnout Vandecappelle,

On Wed, 04 Sep 2013 18:49:18 +0200, Arnout Vandecappelle wrote:

> >   package/webkit/webkit-mips-dfg.patch                       | 1490
> > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >   package/webkit/webkit.mk                                   |   18 +-
> 
>   Hi Markos,
> 
>   Any chance to manually remove the huge patch so that you can post this 
> on the list? Otherwise, there's not much chance for a review...

A 1490 line patch is not so huge, I believe it should make it on the
list, and if not, Peter will validate it manually. I agree that the
overall patch should be sent to the list for review.

>   Is there a reason why you cannot download the upstream patches with 
> WEBKIT_PATCH ?

When it's possible, it's indeed a good idea.

Best regards,

Thomas
Markos Chandras Sept. 5, 2013, 8:48 a.m. UTC | #3
Hello,

On 5 September 2013 09:01, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:
>>   Hi Markos,
>>
>>   Any chance to manually remove the huge patch so that you can post this
>> on the list? Otherwise, there's not much chance for a review...
>
> A 1490 line patch is not so huge, I believe it should make it on the
> list, and if not, Peter will validate it manually. I agree that the
> overall patch should be sent to the list for review.

Well, it's an upstream patch, so i don't think there is much to review
there but I can send it to the list anyway.

>
>>   Is there a reason why you cannot download the upstream patches with
>> WEBKIT_PATCH ?
>
> When it's possible, it's indeed a good idea.
>

Unfortunately, I can't (or I don't think I can) use that. I had to
tweak the upstream patches a bit,
for example delete the changes in the ChangeLog files since they don't apply.
For other patches, I had to merge 2-3 upstream patches into a single
one in order to fix the problem completely instead of providing
separate patches with 1-2 line fixes.
Thomas Petazzoni Sept. 5, 2013, 9:01 a.m. UTC | #4
Dear Markos Chandras,

On Thu, 5 Sep 2013 09:48:32 +0100, Markos Chandras wrote:

> > A 1490 line patch is not so huge, I believe it should make it on the
> > list, and if not, Peter will validate it manually. I agree that the
> > overall patch should be sent to the list for review.
> 
> Well, it's an upstream patch, so i don't think there is much to review
> there but I can send it to the list anyway.

It's not much to review this upstream patch, but rather the overall
changes in webkit.mk and so on.

> >>   Is there a reason why you cannot download the upstream patches with
> >> WEBKIT_PATCH ?
> >
> > When it's possible, it's indeed a good idea.
> 
> Unfortunately, I can't (or I don't think I can) use that. I had to
> tweak the upstream patches a bit,
> for example delete the changes in the ChangeLog files since they don't apply.
> For other patches, I had to merge 2-3 upstream patches into a single
> one in order to fix the problem completely instead of providing
> separate patches with 1-2 line fixes.

Ok. Since the patches are upstream, that's fine, since it means they
will be removed at some point in the future.

Thanks,

Thomas
Markos Chandras Sept. 5, 2013, 9:15 a.m. UTC | #5
Hi Thomas,

On 5 September 2013 10:01, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:
> Dear Markos Chandras,
>
> On Thu, 5 Sep 2013 09:48:32 +0100, Markos Chandras wrote:
>
>> > A 1490 line patch is not so huge, I believe it should make it on the
>> > list, and if not, Peter will validate it manually. I agree that the
>> > overall patch should be sent to the list for review.
>>
>> Well, it's an upstream patch, so i don't think there is much to review
>> there but I can send it to the list anyway.
>
> It's not much to review this upstream patch, but rather the overall
> changes in webkit.mk and so on.

Oh sorry I thought you were referring just to the MIPS patch.

>
>> >>   Is there a reason why you cannot download the upstream patches with
>> >> WEBKIT_PATCH ?
>> >
>> > When it's possible, it's indeed a good idea.
>>
>> Unfortunately, I can't (or I don't think I can) use that. I had to
>> tweak the upstream patches a bit,
>> for example delete the changes in the ChangeLog files since they don't apply.
>> For other patches, I had to merge 2-3 upstream patches into a single
>> one in order to fix the problem completely instead of providing
>> separate patches with 1-2 line fixes.
>
> Ok. Since the patches are upstream, that's fine, since it means they
> will be removed at some point in the future.
>
> Thanks,
>

Yes that's true.  I will send the patch to the list.