Mail Archives: pgcc/1999/12/10/09:58:14
On Fri, 10 Dec 1999, "Michael" wrote:
> Date: Fri, 10 Dec 1999 10:14:55 +0800
> To: <pgcc AT delorie DOT com>
> From: "Michael" <mikejuni AT yahoo DOT com>
> Reply-To: pgcc AT delorie DOT com
> Subject: pgcc 2.95.3 can't compile glibc-2.1.2
>
> Hay,
>
> I still can't compile glibc-2.1.2 on pgcc-2.95.3. I know that gcc-2.95,
> gcc-2.95.1 have problem compiling glibc-2.1.2, make check fails on the Math
> test. So I don't seem suprise if pgcc-2.95, pgcc-2.95.1 can't compile
> glibc-2.1.2. But since gcc-2.95.2 solved those problems, and pgcc-2.95.2,
> pgcc-2.95.3 are based on gcc-2.95.2, they should not have problem too. But I
> am disappointed that this event still occurs on pgcc-2.95.3. I think that a
> stable compiler should at least be able to compile our C-Library without
> glitches.
>
> I use the following flags-
> -mcpu=pentiumpro -march=pentiumpro -O3 -malign-functions=4 -malign-loops=2 -
> malign-jumps=2 -fstrength-reduce -fschedule-insns2 -mstack-align-double
>
Well i did the compilation on mandrake and debian machines, and got an
working packages.
So i cant say in fact that pgcc-2.95.3 can compile glibc 2.1.2. I used flags:
On debian potato 2.2 host :
-O6 -march=k6 -mcpu=k6 -fschedule-insns2 -funroll-loops
On mandrake 6.1 :
-O6 -fomit-frame-pointer -pipe -s -mpentiumpro -mcpu=pentiumpro
-march=pentiumpro -ffast-math
-fexpensive-optimizations -malign-loops=2 -malign-jumps=2
-malign-functions=2
> Glibc 2.1.2 fails on make check, Math test, just like gcc-2.95 and
> gcc-2.95.1
>
> Michael
> I am not on the mailing list.
>
- Raw text -