On Wed, Nov 09, 2022 at 02:22:36PM +0800, Qu Wenruo wrote:
> [FAILURES]
> The following btrfs test cases failed with newer btrfs-progs:
>
> - btrfs/197
> - btrfs/198
> - btrfs/254
>
> They all fail due to output mismatch like the following:
>
> Label: none uuid: <UUID>
> Total devices <NUM> FS bytes used <SIZE>
> devid <DEVID> size <SIZE> used <SIZE> path SCRATCH_DEV
> - *** Some devices missing
> + devid <DEVID> size 0 used 0 path MISSING
>
> [CAUSE]
> Since btrfs-progs commit 957a79c9b016 ("btrfs-progs: fi show: print
> missing device for a mounted file system"), we output the detailed info
> of a missing device if "btrfs filesystem show" is executed using
> "-m <mnt>" option.
>
> Such detailed output no longer follows the old format, thus causing the
> output mismatch.
>
> [FIX]
> Update _filter_btrfs_filesystem_show() to handle detailed missing
> device by:
>
> - Buffer the output first
>
> - Output the first two lines
> Which is always label/uuid and the total device accounting.
>
> - Replace the detailed missing device line with old output
>
> - Sort (in reverse order) and uniq the device list
>
> By this we can handle both old and new output correctly.
> Although this means we lacks the ability to detect mutltiple missing
> devices, thankfully the involved test cases are not checking this yet.
>
> Signed-off-by: Qu Wenruo <wqu@xxxxxxxx>
> ---
> common/filter.btrfs | 12 +++++++++++-
> 1 file changed, 11 insertions(+), 1 deletion(-)
>
> diff --git a/common/filter.btrfs b/common/filter.btrfs
> index d4169cc6..c570d366 100644
> --- a/common/filter.btrfs
> +++ b/common/filter.btrfs
> @@ -35,7 +35,17 @@ _filter_btrfs_filesystem_show()
> _filter_size | _filter_btrfs_version | _filter_devid | \
> _filter_zero_size | \
> sed -e "s/\(Total devices\) $NUMDEVS/\1 $NUM_SUBST/g" | \
> - uniq
> + uniq > $tmp.btrfs_filesystem_show
> +
> + # The first two lines are Label/UUID and total devices
> + head -n 2 $tmp.btrfs_filesystem_show
> +
> + # The remaining is the device list, first filter out the detailed
> + # missing device to the generic one.
> + # Then sort and uniq the result
> + tail -n +3 $tmp.btrfs_filesystem_show | \
> + sed -e "s/devid <DEVID> size 0 used 0 path MISSING/*** Some devices missing/" | \
> + sort -r | uniq
Ah, I had updated to btrfs-progs 6.0 last week, and was wondering how
was it possible no one had fixed this before - either no one is testing
with 5.19+ or no one cares.
It looks good to me, and it works for me.
I had a local patch here as well, but you beat me to it.
My version is based on awk only, no sorting needed:
diff --git a/common/filter.btrfs b/common/filter.btrfs
index d4169cc6..a27a7276 100644
--- a/common/filter.btrfs
+++ b/common/filter.btrfs
@@ -30,11 +30,22 @@ _filter_btrfs_filesystem_show()
UUID=$2
fi
- # the uniq collapses all device lines into 1
+ # The awk script converts between the output format of btrfs-progs 5.19+
+ # to the output format of older versions when there are missing devices.
+ # That format changed in the btrfs-progs commit:
+ #
+ # 957a79c9b016 ("btrfs-progs: fi show: print missing device for a mounted file system")
+ #
+ # The uniq at the end collapses all device lines into 1.
_filter_uuid $UUID | _filter_scratch | _filter_scratch_pool | \
_filter_size | _filter_btrfs_version | _filter_devid | \
_filter_zero_size | \
sed -e "s/\(Total devices\) $NUMDEVS/\1 $NUM_SUBST/g" | \
+ $AWK_PROG -e 'BEGIN { missing = 0 }' \
+ -e '/\sMISSING$/ { missing = 1 }' \
+ -e '/^\s*$/ { next }' \
+ -e '!/\sMISSING$/ { print }' \
+ -e 'END { if (missing) print "\t*** Some devices missing\n" }' | \
uniq
}
Either way, it looks good to me, thanks!
Reviewed-by: Filipe Manana <fdmanana@xxxxxxxx>
> }
>
> # This eliminates all numbers, and shows only unique lines,
> --
> 2.38.0
>
