On Sun, Oct 12, 2014 at 10:55:54PM +0000, Duncan wrote:
[...]
> It's the raid-factor. =:^)
>
> Btrfs in the kernel is apparently accounting for raid-factor in used
> space in whatever function standard df is using, but not in available
> space, even where that available space is already chunk-allocated (btrfs
> fi show, individual devices, size vs. used, where "used" simply means
> chunk-allocated for show) and thus the raid-factor known.
Nice analysis. There must be a bug then. The raid factor is taken into account
when calculating the usable space:
super.c:btrfs_statfs():
1827 buf->f_bavail = total_free_data;
1828 ret = btrfs_calc_avail_data_space(fs_info->tree_root, &total_free_data);
1829 if (ret) {
1830 mutex_unlock(&fs_info->chunk_mutex);
1831 mutex_unlock(&fs_info->fs_devices->device_list_mutex);
1832 return ret;
1833 }
1834 buf->f_bavail += div_u64(total_free_data, factor);
but it looks like the free space is counted twice, lines 1827 and 1834. The
function btrfs_calc_avail_data_space does the guesswork "how much space could
be still allocated" in the logical units, basically simulating the allocator
logic. I'll have a look as I've touched btrfs_statfs last.
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
