Subject: Re: X-build sparc64 => prep fails with "cc1 got fatal signal 11"
To: None <>
From: Jochen Kunz <>
List: tech-toolchain
Date: 05/13/2003 10:45:29
On 2003.05.13 08:44 Martin Husemann wrote:

> I don't see the point. Most host/target variants just work.
"most" =3D=3D i386 | powerpc?

> We fix problems as they are found - and besides powerpc and i386,=20
most =3D { i386 | powerpc };

> what are realistic cross compile hosts others would expect to work?=20
Any Platform we support?=20
Beside the already mentioned sparc64, alpha and amd64, there is
sgimips(64), hppa(64). (If we get this machines supported.) You can get
300 MHz R12k or PA8500 machines for a few hundred EUR. A SGI Octane or
HP9000 B1000 is a much more reliable workhorse than any PeeCee can be...
But this leads to an other discussion. ;-)=20
And even not so fast machines are still very usefull. Slow !=3D useless. =
have a MicroVAX 3100 [1] that I use for NetBSD developement. My current
compile host is a IBM RS/6000 43P-100. Really no speed daemon with its
100 MHz PPC 604e compared to "todays standards".

> Hopefully gcc 3.3 will improve things even more (and soon ;-) )
... and include hppa--netbsdelf? ;-)
Sorry if this is a FAQ: How to use gcc 3.3? Can it build 1.6-release?
(Last time I played with -cuurent was November.)=20

[1] OK. It is a 3100m95, 83MHz NVAX machine with 128 MB RAM. One of the
fastest VAXen ever build, but still much slower then the 300 MHz