kbd-1.12 types.h parse error

Ken Moffat ken at kenmoffat.uklinux.net
Sun Apr 3 11:18:00 PDT 2005


On Sun, 3 Apr 2005, Laurent Hugé wrote:

> Ken Moffat nous écrivait :
> >> > If the earlier headers are like 2.6.11.2, it should be defined
> >> > just above this in types.h : mine has
> >> >
> >> > #ifdef __CHECKER__
> >> > #define __bitwise __attribute__((bitwise))
> >> > #else
> >> > #define __bitwise
> >> > #endif
> >> >
> >> > How do your headers look ?
> >> Yes, it looks exactly the same.
> >>
> >
> > Hmm, in your original post you talked about compiler optimization.
> > What optimization flags did you pass to the toolchain (glibc, gcc,
> > binutils) and did you successfully run the testsuite for each of
> > them ?
> As the book states, I've unseted the optimization for those three builds
> (initializing it just before building coreutils in chapter 6).
> Each testsuite succedeed, except the gcc one which issued a "check not remade
> because of errors" ; anyway, I've search in LFS feedback, and found it was a
> "normal" error which can be left aside without consequences. As far as I
> didn't set optimization at this stage, I don't know what I can do else to
> make this gcc testsuite succeed.

Ok, I misunderstood - your file that breaks the compile looks ok, and
the whole toolchain sounds ok.  I did once have a very obscure bug (typo
in the POSIX version specified before building coreutils in chapter 5)
which only showed up when I failed to build QT a long way down the line,
so never rule things out.

I suppose I'd better ask what optimisations you are using, and indeed
what CPU you have.  At the moment, I'm casting around wildly trying to
guess what went wrong.

> I don't know if it can help, but I've tried to build kbd outside the chroot,
> and it builds well ; furthermore, the types.h of my source system (knoppix)
> and of my lfs are really different (I can post the diff file if you want).
>

I took a quick look at my own log from building kbd, but nothing of
interest shows around where you had the error.  Perhaps the way forward
is to run a fresh configure on a freshly-untarred kbd, both in chroot
and on the host system, and then diff both kbd directories with
diff -u on the host.  There will be a lot of differences because of
different compilers etc, but if you save the output as a file and then
look at it in 'view' with syntax highlighting you might be able to spot
something.  I'm guessing the Makefile in this subdirectory, and perhaps
the config.log in the top-level directory might be the most interesting
parts.

Other than that, I'm out of ideas.

Ken
-- 
 das eine Mal als Tragödie, das andere Mal als Farce




More information about the lfs-support mailing list