PPC64 Compiler bug !!

[Mike Stump]

On Tuesday, June 17, 2003, at 04:00 PM, linas at austin.ibm.com wrote:
> One of these fabled programmers was sent to me for help with thier
> kernel
> crash...
>
>> If the kernel requires it and the guide doesn't mention it, that is a
>> bug in the guide.
>
> What kernel guide?

You know, the piece of documentation or Makefile that prevents the
kernel crash that you saw.  If it doesn't exist, well, you guys will
either have to fight kernel crashes all the time, or document it or
produce a kernel build environment that sets -msoft-float.  That's up
to you'll.

>> I understand sympathize with your pain.  :-(  It sounds as if
>> -mno-implicit-fp matches the programmers expectations from your point
>> of view.
>
> Well, it wasn't my pain, it was someone elses.   And it wasn't just
> my point of view; more'n half-a-dozen other kernel programmers heard
> about this problem and not a single one of them knew about this
> 'feature.'

I'm not a kernel programmer and I've known about it for years.  :-)

Maybe a line like, OS kernels (Linux on PPC, VxWorks on PPC) that don't
otherwise provide an FP context in the kernel require that all such
code be compiled with -msoft-float on the PPC as otherwise the compiler
will use the FP registers for integer code.  Simple, direct, to the
point.

Anyway, since this email includes the linux ppc list, all ppc linux
developers now know, don't they.  :-)

[ I don't mean to discourage you from pressing the gcc folks to provide
a solution, if there is one we can provide; I think that would be good ]


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/