Ken Moffat wrote: > On Tue, Aug 20, 2013 at 11:10:45AM -0500, Bruce Dubbs wrote: >> Ken Moffat wrote: >>> I see you updated util-linux a little while ago. I'm now in xorg, >>> running the tests as a regular user: >>> bash tests/run.sh --srcdir=$PWD --builddir=$PWD >> >> sudo bash tests/run.sh --srcdir=$PWD --builddir=$PWD >> > OK, thanks. Taking my system's health in my hands, with a freshly > built source (as root) - > > It is now clear (to me) that "make check" builds the infrstructure > before reporting that all 0 tests passed - tests/run.sh told me I > needed to run 'make check'. > > Interesting. Looks as if recent fdisk might support GPT. > > I got 4 failures - > blkid: MD raid0 (whole-disks) ... IGNORE (missing mdadm) > blkid: MD raid1 (last partition) ... FAILED (blkid/md-raid1-part) > blkid: MD raid1 (whole-disks) ... IGNORE (missing mdadm) > - I guess that mdadm is needed.
Yes. > cramfs: mkfs doubles ...lt-mount: unknown filesystem type > 'cramfs' FAILED (cramfs/doubles) > cramfs: fsck endianness ... OK > cramfs: mkfs checksums ...lt-mount: unknown filesystem type > 'cramfs' FAILED (cramfs/mkfs) > cramfs: mkfs endianness ... OK > - and cramfs Enable cramfs in the kernel. Also minixfs. Make mtab a file. <rm /etc/mtab && touch /etc/mtab) > fdisk: align 512/4K +MD ... FAILED (fdisk/align-512-4K-md) > - no idea I don't recall that. > BTW that libmount: lock test is really annoying - seems to spawn a > lot of process and gives every appearance that the testsuite has > hung, then eventually completes. > > Complete log attached, in case anyone is interested. If you have a network connection, I think I recall the hwclock test worked. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
