Messages in this thread | | | Date | Tue, 17 Jan 2012 16:56:24 +0800 | From | Li Zefan <> | Subject | Re: Updated btrfs/crypto snappy interface ready for merging |
| |
Andi Kleen wrote: >> It's because decompressing inline extents always fails. I've fixed it >> and will send the patch out in a new mail thread. > > Thanks for fixing. > >> >> But seems there's bug in lib snappy code, which makes the decompressed >> data doesn't quite match the original data. >> >> Simply copy a file to a btrfs filesystem with snappy enabled, and clear >> page cache, and check the file: > > Hmm weird, I have never seen this. Do you have a reproducer? >
Just use some randomly chosen text files. For example:
# mount -t btrfs -o compress=snappy /dev/xxx /mnt # cp -r btrfs-progs-unstable/ /mnt # sync # echo 3 > /proc/sys/vm/drop_caches # diff -Nurp btrfs-progs-unstable /mnt/btrfs-progs-unstable
I've tested on both x86_32 and x86_64.
> The basic compression code is quite well tested, I have a reasonable > unit test. >
| |