On 2018年07月19日 13:59, Lu Fengqi wrote: > On Thu, Jul 19, 2018 at 01:51:59PM +0800, Qu Wenruo wrote: >> >> >> On 2018年07月19日 13:15, Misono Tomohiro wrote: >>> Add -i (ignore offset) option to logical-resolve command >>> to show how BTRFS_IOC_LOGICAL_INO_V2 ioctl works >>> (returns every ref to the extent of given logical address). >>> >>> [Example] >>> $ mkfs.btrfs -f $DEV >>> $ mount $DEV /mnt >>> >>> $ dd if=ev/urandom of=/mnt/file bs=4k count0 >>> # split above extent >>> $ dd if=ev/urandom of=/mnt/file bs=4k seek count=1 conv=notrunc >>> $ btrfs filesystem sync >>> >>> # v1 >>> $ btrfs inspect-internal logical-resolve -P 13631488 /mnt >>> inode 257 offset 0 root 5 >>> >>> # v2 >>> $ btrfs inspect-internal logical-resolve -iP 13631488 /mnt >>> inode 257 offset 0 root 5 >>> inode 257 offset 45056 root 5 >>> >>> Signed-off-by: Misono Tomohiro <misono.tomohiro@xxxxxxxxxxxxxx> >>> --- >>> Documentation/btrfs-inspect-internal.asciidoc | 4 ++++ >>> cmds-inspect.c | 17 +++++++++++++++-- >>> ioctl.h | 10 +++++++++- >>> libbtrfsutil/btrfs.h | 10 +++++++++- >> >> Not related to this patch itself, but I'm wondering could we just use >> /usr/include/linux/btrfs.h? >> >> So we could save 2 same copies of headers here. > > If we want to compile the newest btrfs-progs, but the older > /usr/include/linux/btrfs.h maybe doesn't have already defined the ioctl, > this will cause the compile error. Makes sense. > I am curious about why the > libbtrfsutil need to copy the part of ioctl.h instead of including the > ioctl.h directly? Maybe for distribution purpose? Some binding may be delivered as independent package just as Omar tries to do. So in that case independent header makes sense. Thanks, Qu
Attachment:
signature.asc
Description: OpenPGP digital signature
