Related articles |
---|
64-bit ELF Cross Compiler heckler@my-deja.com (Mike S) (2000-07-18) |
Re: 64-bit ELF Cross Compiler pat@jantar.org (Patryk Zadarnowski) (2000-07-23) |
Re: 64-bit ELF Cross Compiler pedwards@dmapub.dma.org (2000-07-27) |
From: | pedwards@dmapub.dma.org (Phil Edwards) |
Newsgroups: | comp.compilers |
Date: | 27 Jul 2000 21:37:44 -0400 |
Organization: | $home is where the core is |
References: | 00-07-042 |
Keywords: | GCC, architecture |
Patryk Zadarnowski <pat@jantar.org> wrote:
+ - the 64bit MIPS (and possibly others) GNU linker is completely broken
+ (it patches pointers to almost random values.)
I wondered what the call to rand() was doing in there...
+ Also, I would very much
+ like to hear of experiences people had using GCC on SPARC v9.
Take a look at the GCC mailing list archives or the Sun groups. Summary:
all the experiences are bad, since the assembly generated is sometimes
illegal (the integer long division IIRC). The GCC FAQ has an entry on
this subject, which I expect will be updated as progress is made.
+ assembler listing, so I'm staying clear of sparc64 GCC untill I hear
Using "sparc64" can be misleading. Apparently this was a mis-applied name
for some variation of Linux at one point. I'm fuzzy on the details (and
the GCC installation instructions get it wrong a couple of times, too),
but "sparcv9" is the best term to use when talking about GCC. In the
current source code, that is the platform name being generated as well.
Phil
Return to the
comp.compilers page.
Search the
comp.compilers archives again.