gcc 3.3.1 on AIX 5.2?

gcc 3.3.1 on AIX 5.2?

[Paul Ripke]

Anyone running recent a recent gcc release on AIX 5.2? I've tried
building gcc 3.3.1 with a mostly-working IBM compiled gcc, and gnumake.
The error is during the libgcc2 build. I have a vague recollection
building gcc on MacBSD 0.8 where some m68k gcc libraries can't be
natively built by gcc... Is this something similar?

Build error is:

./xgcc -B./ -B/app/gcc/powerpc-ibm-aix5.2.0.0/bin/ -isystem 
/app/gcc/powerpc-ibm-aix5.2.0.0/include -isystem 
/app/gcc/powerpc-ibm-aix5.2.0.0/sys-include -O2  -DIN_GCC    -W -Wall 
-Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -isystem 
./include   -g -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 
-D__GCC_FLOAT_NOT_NEEDED  -I. -I. -I../../gcc-3.3.1/gcc 
-I../../gcc-3.3.1/gcc/. -I../../gcc-3.3.1/gcc/config 
-I../../gcc-3.3.1/gcc/../include  -DL_fixunsdfsi -c 
../../gcc-3.3.1/gcc/libgcc2.c -o libgcc/./_fixunsdfsi.o
../../gcc-3.3.1/gcc/libgcc2.c: In function `__fixunsdfsi':
../../gcc-3.3.1/gcc/libgcc2.c:1161: error: unrecognizable insn:
(insn:HI 11 10 12 0 0 (set (reg:CCEQ 122)
         (compare:CCEQ (ior:SI (gt:SI (reg:CCFP 121)
                     (const_int 0 [0x0]))
                 (eq:SI (reg:CCFP 121)
                     (const_int 0 [0x0])))
             (const_int 1 [0x1]))) -1 (insn_list 10 (nil))
     (expr_list:REG_DEAD (reg:CCFP 121)
         (nil)))
../../gcc-3.3.1/gcc/libgcc2.c:1161: internal compiler error: in 
extract_insn, at recog.c:2175
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
gnumake[3]: *** [libgcc/./_fixunsdfsi.o] Error 1

Mostly working IBM built gcc is:

Reading specs from 
/usr/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/2.9-aix51-020209/specs
gcc version 2.9-aix51-020209

Cheers,
--
Paul Ripke
Unix/OpenVMS/TSM/DBA
I love deadlines. I like the whooshing sound they make as they fly by.
-- Douglas Adams

Re: gcc 3.3.1 on AIX 5.2?

[David Edelsohn]

>>>>> Paul Ripke writes:

> Anyone running recent a recent gcc release on AIX 5.2? I've tried
> building gcc 3.3.1 with a mostly-working IBM compiled gcc, and gnumake.
> The error is during the libgcc2 build.
> 
> ../../gcc-3.3.1/gcc/libgcc2.c:1161: internal compiler error: in 
> extract_insn, at recog.c:2175
> 
> Mostly working IBM built gcc is:
> 
> Reading specs from 
> /usr/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/2.9-aix51-020209/specs
> gcc version 2.9-aix51-020209

	Note that you are using GCC built for AIX 5.1 to build GCC for AIX
5.2 *ON AN AIX 5.2 system*.

	You need to delete the cache of "fixed" AIX 5.1 header files in
the GCC installation.

	AIX 5.2 includes the atoll() function and the GCC configuration
finds it, but the AIX 5.1 header file in the cache does not declare it.
GCC assumes that atoll() returns an "int" instead of "long long" and
things go downhill from there.

David

Re: gcc 3.3.1 on AIX 5.2?

[Paul Ripke]

On Thursday, Oct 9, 2003, at 05:04 Australia/Sydney, David Edelsohn 
wrote:

>> Paul Ripke writes:
>>
>> Anyone running recent a recent gcc release on AIX 5.2? I've tried
>> building gcc 3.3.1 with a mostly-working IBM compiled gcc, and 
>> gnumake.
>> The error is during the libgcc2 build.
>>
>> ../../gcc-3.3.1/gcc/libgcc2.c:1161: internal compiler error: in
>> extract_insn, at recog.c:2175
>>
>> Mostly working IBM built gcc is:
>>
>> Reading specs from
>> /usr/bin/../lib/gcc-lib/powerpc-ibm-aix5.1.0.0/2.9-aix51-020209/specs
>> gcc version 2.9-aix51-020209
>
> 	Note that you are using GCC built for AIX 5.1 to build GCC for AIX
> 5.2 *ON AN AIX 5.2 system*.
>
> 	You need to delete the cache of "fixed" AIX 5.1 header files in
> the GCC installation.
>
> 	AIX 5.2 includes the atoll() function and the GCC configuration
> finds it, but the AIX 5.1 header file in the cache does not declare it.
> GCC assumes that atoll() returns an "int" instead of "long long" and
> things go downhill from there.

Ended up finding one of your mails in the archives describing the 
atoll() issue: <http://gcc.gnu.org/ml/gcc-bugs/2003-06/msg02965.html>.
Added a definition to the gcc "fixed" headers from 2.95.2. After
fixing that, it left me with a bad header after the gcc fix headers
step. Some hand tweaking there, and I seem to have a working compiler.

If anyone's interested, I'll work out what I needed to do, and pass it
on.

BTW: I wasn't using bash.

Thanks for the reply,
--
Paul Ripke
Unix/OpenVMS/TSM/DBA
I love deadlines. I like the whooshing sound they make as they fly by.
-- Douglas Adams