This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: Patch for i386-darwin: test suite results
- From: Andrew Pinski <pinskia at physics dot uc dot edu>
- To: tolaini at libero dot it (Sandro Tolaini)
- Cc: toa at pop dot agri dot ch (Andreas Tobler), java-patches at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org, java at gcc dot gnu dot org, aph at redhat dot com (Andrew Haley), pinskia at physics dot uc dot edu (Andrew Pinski)
- Date: Thu, 16 Mar 2006 19:09:10 -0500 (EST)
- Subject: Re: Patch for i386-darwin: test suite results
>
>
> On 17/mar/2006, at 00:57, Andrew Pinski wrote:
>
> > Actually it looks like libffi is still not working as you no longer
> > have eh info for x86-darwin. This is why all the gij tests fail.
>
> Is that the block at the end of sysv.S which I commented out? :)
Yep :).
>
> The problem is that the Apple assembler (actually a rather old gas:
> "Apple Computer, Inc. version cctools-590.23.2.obj~30, GNU assembler
> version 1.38") doesn't support most of the directives used in that
> part of the source. And I'm not an asm wizard, I had to learn it in
> two days to fix the annoying 16-byte stack aligning requirement of
> the Darwin-i386 ABI...
Don't talk to me about Apple's cctools but they just don't cut anything
for me recently, making GCC developers update cctools every release
is a pain in the butt. The recent change for __cxx_atexit is good
for C++ developers but it causes a failure unless you have a new cctools.
I just found out the hard way that this failed. Please Apple think
next time you are going to change something like this.
I was going to finish testing my performance patch and bootstrap
failed because not having the new cctools which is piece of crap.
The cctools release which was made was known to be not tested and
really should have never been released.
-- Pinski