[PATCH]: Fix GLX crashes on RISC cpus.

[David Miller]

From: Daniel Stone <daniel at fooishbar.org>
Date: Thu, 27 Dec 2007 08:22:18 +0000

> Oddly enough, I don't read every single bug report that comes through,
> so attributing this one to ignorance rather than deliberate malice might
> be acceptably charitable here.

It's a general sentiment and not directed at you specifically, sorry.

I've had to really make a big fuss in the past to get a patch
installed that kept the X server from eating people's IDE disks, and
based upon some other experiences that seems to be par for the course.

As another example, half the input drivers in the tree don't even
build, and it's been like this for over two years.  There has been
some level of forward progress, but the regression is still mostly
there.

The person who checked in the input driver API changes that broke the
driver builds gave a response something like "I implemented a real
important feature, I don't have the time to fix up all of these kooky
input drivers that no longer build."

This kind of stuff, to me, is miles away from acceptable and I hope
that this bug fix won't be handled like those cases.  These bugs being
fixed here are regressions of the modular tree conversion.

I'm happy to reimplement this specific fix to be cleaner, but let's
get the multi-year regression fixed to start off.

Thanks.