On Sat, Aug 19, 2006 at 06:16:31PM +0100, I wrote: > > Of course, the fact that my 'bash-3.1: no such file' error didn't > show up makes me think that my earlier report of this on -m64 builds > is bogus, there was probably something strange in my temporary bash. > At least I'm used to making bogus bug reports!
Hmm, last week I built -rc5 as 'boot' and got this error again in 64-bit glibc. At the moment I'm rebuilding from chroot and not seeing this error. So, I have to conclude that something in my 'boot' build is breaking this, it could as easily be one of the extras (vim, ssh, nfs, ...) as one of the things in the book. The only obvious difference is that 'boot' builds all of util-linux instead of just mount and more. To be honest, I'm insufficently interested in this to investigate it any further, but if anybody else experiences the problem they can find this post in the archives. To summarise - if I go the 'boot' route on a 64-bit machine, it rather looks as if bash reports 'bash-3.1' instead of '/bin/bash' for $SHELL. This _only_ affects a glibc test. Why are the testsuites so fscking fragile ? Ken -- das eine Mal als Tragödie, das andere Mal als Farce _______________________________________________ Clfs-dev mailing list [email protected] http://lists.cross-lfs.org/cgi-bin/mailman/listinfo/clfs-dev
