----- Original Message ----- From: "Bruce Dubbs" <[email protected]> To: "LFS Developers Mailinglist" <[email protected]> Sent: Sunday, February 06, 2011 10:07 PM Subject: Re: LFS SVN-20110204 error report
> Gilles Espinasse wrote: > > > The script I send run until test fail. This vary from 1 to 150 attempts. > > OK, I took a look. The files differ by one character in a timestamp. > > http://en.wikipedia.org/wiki/Ar_(Unix) > > cmp is the wrong tool for comparison. There is a race condition and > thus a but in the glibc configure.in file. > > Perhaps it should be: > > cmp -si 34 conftest.a conftest2.a > > I'll file a bug, but I doubt it will be addressed. > > -- Bruce > > Posix don't know -i option. http://pubs.opengroup.org/onlinepubs/009695399/utilities/cmp.html The bug only trigger an unneeded operation, that does not break anything. So that may be good enought to just add this comment : "Test may fail due to a race but if ranlib does not run the same second the archive was created. Test failure cause no harm, just an unneeded ranlib run." Gilles -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
