IBM POWER6/AIX6 current state?
Sat Jan 23 16:22:00 GMT 2010
-----BEGIN PGP SIGNED MESSAGE-----
Thanks, i think i will give it a try as soon as gcc 4.5 is released.
In the meantime i would like to ask you for some additional background
Last time i went this route it was with gcc 4.4.0 and some binutils 2.20
prerelease (it could have been a post 2.19 snapshot from the trunk as well).
The gcc build came to a grinding halt, when in stage 2 the gcc compiler
built in stage 1 advised gas to generate POWER6 machine instructions
(the flag was -pwr6 i think) and gas choked because id did not
understand this command line parameter (IIRC only up to -pwr2 was
possible then, but i am not 100% sure about this anymore).
So my question is, what has changed in the meantime:
1) Is it that gas now recognises -pwr6 and generates POWER6 machine code
when handed down this option?
2) Or gcc 4.5 will no longer insist on gas generating POWER6 opcodes,
but rather fall back to the latest instruction set supported by gas
(which one btw?)?
3) Or did binutils/gas as used by me already support POWER6, only the
- -pwr6 flag was wrong?
4) other explanations?
Am 23.01.2010 13:13, schrieb Richard Sandiford:
> Franz Fehringer <firstname.lastname@example.org> writes:
>> Hi all,
>> Are binutils usable on IBM POWER6/AIX6 machines?
>> I would like to use them in current gcc (4.4.x / 4.5 snapshot) builds.
>> Last time i tried, gas only supported older processors and gprof was not
>> supported at all.
> Binutils 2.20 should support POWER6 and AIX6, provided that you use
> a recent GCC 4.5 snapshot. However, I've only tested it in a cross
> configuration. Please file a bug report if you find any problems
> using them natively.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.12 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
-----END PGP SIGNATURE-----
More information about the Binutils