[uClibc]Latest toolchain: ix86 --> powerpc problems

Pedro Sanchez psanchez at nortelnetworks.com
Tue Dec 3 20:23:53 UTC 2002


Thanks Erik,

Question, when you say "cvs up and try again," which tree in the
toolchain source (3.2, 3.2.1, etc) are you updating? I just cvs up'ed
the whole toolchain tree and got nothing new, and ViewCVS (web) doesn't
show any update either.

BTW, another question: What is the best way to get the toolchain
optimized for a particular processor? So far, what I'm doing is editing
the make file to include something like

	CONFIG_OPTS=--with-cpu=403 --nfp --without-fp

and then adding this variable to the configure statements used to build
gcc. Is there another way to do it?

-- 
Pedro

On Tue, 2002-12-03 at 15:08, Erik Andersen wrote:
> On Tue Dec 03, 2002 at 02:25:28PM -0500, Pedro Sanchez wrote:
> > Hello,
> > 
> > I'm trying to get the toolchain (latest from CVS, with fixes for the
> > cross-build) to compile and run well. But I'm not there yet. Here are a
> > couple of comments on what I'm seeing:
> > 
> > 1. There is this line in the Makefile in the section that configures the
> > kernel headers:
> > 
> > (cd $(LINUX_DIR)/include; ln -fs asm-$(ARCH)$(NOMMU) asm;)
> > 
> > ARCH=powerpc in my case, and there is no "asm-powerpc" dir in the kernel
> > headers. It should be "asm-ppc". I believe this should be fixed.
> 
> Ok, I've fixed this now...
> 
> > 2. I get the following when using the latest toolchains from CVS (both
> > 3.2 and 3.2.1)
> > 
> > Applying gcc-3.2.1/sources/gcc-001-disable-mathf.patch using plaintext: 
> > can't find file to patch at input line 3
> > Perhaps you used the wrong -p or --strip option?
> > The text leading up to this was:
> 
> This was a secondary error, resulting from the earlier failure.
> Do a 'cvs up', then 'make clean' and try again...
> 
> > --------------------------
> > |--- gcc-3.1/libstdc++-v3/configure	Tue May  7 23:27:59 2002
> > |+++ gcc-3.1-patched/libstdc++-v3/configure	Thu Aug  8 11:01:35 2002
> > --------------------------
> > File to patch: 
> > 
> > hmm, why gcc-3.1?
> 
> Because I havn't updated the patch since it applies fine.
> 
>  -Erik
> 
> --
> Erik B. Andersen             http://codepoet-consulting.com/
> --This message was written using 73% post-consumer electrons--





More information about the uClibc mailing list