Re: Weird sysinst behavior - port amd64

[Andreas Gustafsson <gson%gson.org@localhost>]

Paul Goyette wrote:
> While trying to check out something else, I was installing 6.99.6 in a 
> qemu VM.  While sysinst was unpacking comp.tgz, I encountered the 
> following error several times:
> 
>   91% |*******************************    |   275 MiB    4.01 MiB/s   00:06 
> ETA/
> targetroot: directory block of inode 0xd3eb extends across I/O block 
> boundary, which shouldn't happen!
>   92% |********************************   |   277 MiB    3.82 MiB/s   00:05 
> ETA/
> targetroot: directory block of inode 0xd3eb extends across I/O block 
> boundary, which shouldn't happen!

I see the same messages in the logs on the TNF test system, on both
amd64 and i386.  They appeared quite recently, with the following commit:

  commit 2012.05.09.00.21.17 riastradh src/sys/rump/fs/lib/libext2fs/Makefile 
1.4
  commit 2012.05.09.00.21.17 riastradh src/sys/rump/fs/lib/libffs/Makefile 1.14
  commit 2012.05.09.00.21.17 riastradh src/sys/ufs/files.ufs 1.29
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ext2fs/ext2fs_dir.h 1.19
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ext2fs/ext2fs_extern.h 1.44
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ext2fs/ext2fs_rename.c 1.1
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ext2fs/ext2fs_vnops.c 1.104
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/lfs/lfs_vnops.c 1.242
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ufs/ufs_extern.h 1.72
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ufs/ufs_lookup.c 1.115
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ufs/ufs_rename.c 1.1
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ufs/ufs_vnops.c 1.209
  commit 2012.05.09.00.21.18 riastradh src/sys/ufs/ufs/ufsmount.h 1.38

-- 
Andreas Gustafsson, gson%gson.org@localhost