Re: Reverse reversed_comparison_code_parts patch

[Jeffrey Oldham <oldham at codesourcery dot com>]

On Thu, Mar 15, 2001 at 09:53:15AM +0100, Jan Hubicka wrote:
> > On Wed, Mar 14, 2001 at 07:04:27PM +0100, Jan Hubicka wrote:
> > > > Jan,
> > > > 
> > > > 	I compiled your proposed patch to gcc/jump.c + gcc/optabs.c +
> > > > MIPS changes on a mips-sgi-irix6.5.  The code succeeded up to the same
> > > > point where the unpatched code also dies so I presume it eliminates
> > > > the problem for mips-sgi-irix6.5.  ): Because building the Java
> > > > libraries dies, I cannot run the test cases to see if the resulting
> > > > executables still work.  I would be reluctant to check in the patch
> > > > until this is done. :(
> > > I've sent you pointer to two patches - you need to install both of them.
> > > It seems to me that you've installed just first one.
> > 
> > Jan,
> > 
> > 	I am having difficulty understanding exactly what code you
> > want tested.  Please be as precise as possible.  Sending exactly one
> > patch with a ChangeLog entry would be optimal.  When I resume work
> > Monday, I will bootstrap using this patch.  After that you can use a
> > cross-compiler or a SUSE SGI machine for testing.
> I am attaching both patches - please install them both and try the bootstrap.
> I've already tested that they fixes the cross compiler crash before
> sending them to the list. So unless I messed up the runtime tests, they
> should do the job.
> 
> Honza

The patches you sent do not include your original changes to
gcc/jump.c.  Thus, I did not test them.

I have tried to be helpful, but this change is taking too much time.
Please use a SUSE mips-sgi-irix6.5 machine to test your proposed
patches if you want to submit them.  I cannot spend several more hours
of CodeSourcery's time to test your work.

Thanks,
Jeffrey D. Oldham
oldham@codesourcery.com