[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Update on NetBSD and 3.3



On Thu, Apr 24, 2003 at 11:10:54PM -0600, Joel Baker wrote:
> Just a quick status update on gcc-3.3 and the various NetBSD ports.
> 
> 1) I'm working on a patch for 3.3; the netbsd-* patches from 3.2 don't
> even remotely apply cleanly, due to major changes in the files they modify
> (thankfully, said changes are in the direction of increased sanity!)

The patch is (almost) done.

> 2) Due to the changes from 3.2 -> 3.3, once the i386 stuff works, it SHOULD
> work with minimal extra changes for all other cpus. The patch (once done)
> is intended to have at least netbsd-alpha and netbsd-sparc stuff in it,
> since I have access to someone who can test it on those architectures
> (Matthew Garrett).

This can only be tested once 3-5 are completed.

> 3) Some stuff in rules.conf (the LIBC_DEV stuff, in particular) needs to be
> made more general, if netbsd-alpha and netbsd-sparc are to use it properly.
> However, given the number of times it's gotton broken in one fashion or
> another, I hesitate to muck with this myself, anymore...

Talking to Matthias about it.

> 4) There appears to be at least one very serious regression; Pascal now
> blows up with a build failure. I'm going to look into it.

Haven't checked this yet. It's next.

> 5) There also appear to be some regressions (or possibly new tests that
> just don't work at all) in the testsuite; thankfully, not many. I think.

See the attached file; it actually looks better, the 'regressions' are
(I *think*) stuff that broke before. If I'm reading the file right.

> 6) LIB_SPEC stuff is completely untested, as of yet; I want to fix the
> regressions first, before I go mucking with too much of the other stuff.
> However, I'll probably try a build with it changed, just to see if it
> persists in the 3.2 behavior of blowing up in unbelieveably impressive
> ways.

This comes after #4.
-- 
Joel Baker <fenton@debian.org>

Attachment: pgpJFstmRVXq_.pgp
Description: PGP signature


Reply to: