[Btrfs-devel] Re: mkfs/mount troubles with the latest btrfs-unstable.

[Chris Mason]

On Tuesday 25 March 2008, Evgeniy Polyakov wrote:
> Here is first, when mkfs fails:
>
> #0  0xffffe410 in __kernel_vsyscall ()
> #1  0xb7e85520 in raise () from /lib/i686/cmov/libc.so.6
> #2  0xb7e86ee8 in abort () from /lib/i686/cmov/libc.so.6
> #3  0x08050d41 in open_ctree_fd (fp=5, path=0xbf848e73 "/dev/sdb1",
> sb_bytenr=16384) at disk-io.c:502 #4  0x08066825 in make_root_dir (fd=5,
> device_name=0xbf848e73 "/dev/sdb1") at mkfs.c:71 #5  0x080670ce in main
> (ac=0, av=0xbf8475a4) at mkfs.c:220
>
> Here is a patch to fix this issue, which brings us to second problem.
>

Ouch, thanks.  I'm not sure why gcc doesn't warn about uninit variables in the 
btrfs progs, I would expect -Wall to cover that.
> Starting program: /root/btrfs-progs-unstable/mkfs.btrfs /dev/sdb1
> found device 1 on /dev/sdb1
> lowest devid now 1
> found Btrfs on /dev/sdb1 with 1 devices
> opening /dev/sdb1 devid 1 fd 7
> btrfs_alloc_extent: ret: 0.
> btrfs_alloc_extent: ret: 0.
> alloc chunk size 134217728 from dev 1
> btrfs_alloc_extent: ret: 0.
> btrfs_alloc_extent: ret: -28.
>
> It is -ENOSPC after find_free_extent().

Very strange.  Looks like you're on 32 bit, which I haven't tested on yet.  
I'll test 32 bit later this week, but there are probably silly mistakes in 
there.

-chris