Source: webkit2gtk
Version: 2.54.2-2
Severity: normal
X-Debbugs-Cc: debian-68k@lists.debian.org, debian-powerpc@lists.debian.org, debian-s390@lists.debian.org, debian-sparc@lists.debian.org
User: debian-68k@lists.debian.org
Usertags: m68k
User: debian-powerpc@lists.debian.org
Usertags: powerpc
User: debian-s390@lists.debian.org
Usertags: s390x
User: debian-sparc@lists.debian.org
Usertags: sparc64

Hello,

I just noticed that webkit2gtk was disabled on big-endian targets without reaching
out to porters asking for help and I think this change should be reverted.

The reason for disabling big-endian targets was that Skia is supposedly not supported
on big-endian targets. However, Skia has always had some issues on big-endian targets
but there were always easy to fix.

Oracle Solaris always ships a recent version of Firefox also for SPARC and there are
definitely patches for fixing Skia on big-endian patches [1].

Please let's work out fixing Skia and webkit2gtk on big-endian targets and not just
disable such an important package so many architectures for just a bug that can
be fixed.

Thanks,
Adrian

Hi John,

I am currently working on a patch that fixes Skia on big endian
architectures. I have got it build and have fixed some test-cases and is
currently working on the rest. I'll update you as I go forward.

Thanks,
Pranav

Hi Pranav,

Great, thanks a lot! I'm testing this patch [1] by Oracle now.

Adrian

My opinion is that this should be sent upstream (not to Google,
because it's unlikely to be accepted, but to WebKitGTK) and someone
would need to commit to maintain it there. The WebKitGTK team does not
have resources for that.

Berto

Swift toolchain in the future (it's unclear when exactly, but possibly
during 2027).

At this point swiftlang in Debian only supports amd64 and arm64, so
having Skia won't be enough.

More details here:

https://lists.debian.org/debian-devel/2026/09/msg00139.html

Berto