[lfs-dev] Possible problem with current glibc (LFS 7.2 cant recompile LFS 7.2)
zarniwhoop at ntlworld.com
Sun Aug 26 15:53:53 PDT 2012
On Sun, Aug 26, 2012 at 05:24:42PM -0500, Bruce Dubbs wrote:
> Ken Moffat wrote:
> It would now appear there is an error in the sunrpc portion of the
> Makefile that is calling the wrong gcc. I guess it didn't fail for me
> because I did install libtirpc for testing samba.
I surmise that Jasmine failed to install the rpc headers, which we
do in chapter 6. But, that has highlighted the issue for those of
us who did install them.
> I don't know if changing the gcc to x86_64-lfs-linux-gnu-gcc would be
> sufficient or not, but I think we need to do that as a minimum.
Your reply came in while I was composing my follow-up. I refer you
back to my initial posting where I pointed to a post on libc-alpha -
glibc-2.16.0 appears to have been changed so that it will install
some rpc stuff even if cross-compiling, so from upstream's POV it
iis using the _right_ gcc.
My current theory is that overriding the compiler *and* using the
internal headers in chapter 5 glibs will suffice, but I'll use
--enable-obsolete-rpc headers in chapter 6. I'm doing a fresh
build, it's just got to ch5 glibc.
> -- Bruce
> `-MF FILE'
> When used with the driver options `-MD' or `-MMD', `-MF' overrides
> the default dependency output file.
> This option instructs CPP to add a phony target for each dependency
> other than the main file, causing each to depend on nothing. These
> dummy rules work around errors `make' gives if you remove header
> files without updating the `Makefile' to match.
> Like `-MD' except mention only user header files, not system
> header files.
> `-MT TARGET' An `-MT' option will set the target to be exactly the
> string you specify.
Heh, I'll leave these options to you, I'm too stupid to get htem
das eine Mal als Tragödie, das andere Mal als Farce
More information about the lfs-dev