Bug 102938 – BrowserMarshal.h build fail

• Component: WebKitGTK
• Status: RESOLVED FIXED

[YuTeh Shen]

Created attachment 175433 [details]
patch to fix the build error

Encounter build error in Webkit gtk as the following:

  CXXLD  Programs/ImageDiff                                                                                                                                                       
  CC     Tools/MiniBrowser/gtk/Programs_MiniBrowser-BrowserCellRendererVariant.o
../../Tools/MiniBrowser/gtk/BrowserCellRendererVariant.c:27:28: fatal error: BrowserMarshal.h: No such file or directory
compilation terminated.
make[1]: *** [Tools/MiniBrowser/gtk/Programs_MiniBrowser-BrowserCellRendererVariant.o] Error 1
make[1]: *** Waiting for unfinished jobs....
make[1]: Leaving directory `/home/ytshen/thirdparty/WebKit/WebKitBuild/Debug'
make: *** [all] Error 2
   
Failed to build WebKit using 'make'!

[Alberto Garcia]

Created attachment 211526 [details]
Simplify the makefile

Thanks for spotting the typo in the makefile, although I don't quite
see how that fixes the whole thing.

The stamp-mini-browser-marshal files seem unnecessary to me, it should
be enough to just list the dependencies directly needed to build
BrowserMarshal.h and BrowserMarshal.c without creating the
intermediate files.

[Gustavo Noronha (kov)]

Comment on attachment 211526 [details]
Simplify the makefile

Might wanna take a peek at the commit that added this indirection just to be sure it was not working around some other crazy behaviour, but it looks unnecessary indeed.

[Alberto Garcia]

(In reply to comment #2)
> Might wanna take a peek at the commit that added this indirection
> just to be sure it was not working around some other crazy
> behaviour, but it looks unnecessary indeed.

Yeah, Carlos was mentioning this comment:

https://git.gnome.org/browse/gtk+/tree/gtk/Makefile.am#n1148

I took a quick look but I didn't quite get why it is necessary. I'll
double check before committing this patch.

[Carlos Garcia Campos]

(In reply to comment #3)
> (In reply to comment #2)
> > Might wanna take a peek at the commit that added this indirection
> > just to be sure it was not working around some other crazy
> > behaviour, but it looks unnecessary indeed.
> 
> Yeah, Carlos was mentioning this comment:
> 
> https://git.gnome.org/browse/gtk+/tree/gtk/Makefile.am#n1148
> 
> I took a quick look but I didn't quite get why it is necessary. I'll
> double check before committing this patch.

Yes, please, I think it would be enough to make sure make distcheck passes.

[Alberto Garcia]

Looks like distcheck is failing for me even without the patch:

In file included from ../../Source/WebCore/platform/graphics/Font.h:28:0,
                 from ../../Source/WebCore/platform/graphics/GraphicsContext.h:33,
                 from ../../Source/WebCore/platform/graphics/ImageBuffer.h:34,
                 from ../../Source/WebCore/platform/graphics/GeneratorGeneratedImage.h:32,
                 from ../../Source/WebCore/platform/graphics/GeneratorGeneratedImage.cpp:27:
../../Source/WebCore/platform/graphics/FontDescription.h:36:35: fatal error: WebKitFontFamilyNames.h: No such file or directory
compilation terminated.

either I'm doing something wrong or this is bug 76388 again ?

[Carlos Garcia Campos]

(In reply to comment #5)
> Looks like distcheck is failing for me even without the patch:
> 
> In file included from ../../Source/WebCore/platform/graphics/Font.h:28:0,
>                  from ../../Source/WebCore/platform/graphics/GraphicsContext.h:33,
>                  from ../../Source/WebCore/platform/graphics/ImageBuffer.h:34,
>                  from ../../Source/WebCore/platform/graphics/GeneratorGeneratedImage.h:32,
>                  from ../../Source/WebCore/platform/graphics/GeneratorGeneratedImage.cpp:27:
> ../../Source/WebCore/platform/graphics/FontDescription.h:36:35: fatal error: WebKitFontFamilyNames.h: No such file or directory
> compilation terminated.
> 
> either I'm doing something wrong or this is bug 76388 again ?

Nah, distcheck is almost always broken, I'll try to find some time to fix it.

[Alberto Garcia]

(In reply to comment #6)
> Nah, distcheck is almost always broken, I'll try to find some time to fix it.

Ah, that's because I tried to do distcheck from a clean build.

(which I guess it should anyway work? but that's a different issue)

[Alberto Garcia]

Created attachment 211882 [details]
Patch

I see that we're using the stamp method in other places so I think we
should just fix the variable name now.

If the stamp thing is worth changing I would leave that for a separate
bug.

[Alberto Garcia]

Committed r155945: <http://trac.webkit.org/changeset/155945>