On 6/26/18 9:48 PM, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
>
> Now that we've ported the extent size hint verifiers to libxfs, call
*almost* ;)
> them from xfs_repair instead of open-coding the checks.
*nod*
> Tweak the
> repair message slightly to reflect the fact that we zero the field and
> clear the inode flag if the hint is garbage or is inconsistent with the
> inode flags. Also clear the extent size hints when we're zapping the
> inode core since we clear the associated flags.
So for that last clause, the fix to clear_dinode_core, I wonder if it should
be a separate patch, possibly even merged before the libxfs changes.
I guess I can't prove that it's an existing bug prior to adding the extent
size hints to the verifier, but it seems weird for clear_dinode_core to
/not/ clear the extent size hint, doesn't it?
The only advantage is we could avoid any regression point for xfs/033
(which is how I found this) if we fixed up clear_dinode_core, then
added the extent size hint validation to the verifier, and then taught
xfs_repair to use the new extent size hint validation function.
Does that seem worth it?
The patch all looks fine, just wondering if we should structure/sequence
the changes more nicely.
-Eric
> Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
> ---
> libxfs/libxfs_api_defs.h | 2 +
> repair/dinode.c | 85 +++++++++++++++++++---------------------------
> 2 files changed, 38 insertions(+), 49 deletions(-)
>
>
> diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
> index e5cf1554..fe8336ab 100644
> --- a/libxfs/libxfs_api_defs.h
> +++ b/libxfs/libxfs_api_defs.h
> @@ -98,6 +98,8 @@
> #define xfs_dinode_calc_crc libxfs_dinode_calc_crc
> #define xfs_idata_realloc libxfs_idata_realloc
> #define xfs_idestroy_fork libxfs_idestroy_fork
> +#define xfs_inode_validate_extsize libxfs_inode_validate_extsize
> +#define xfs_inode_validate_cowextsize libxfs_inode_validate_cowextsize
>
> #define xfs_rmap_ag_owner libxfs_rmap_ag_owner
> #define xfs_rmap_alloc libxfs_rmap_alloc
> diff --git a/repair/dinode.c b/repair/dinode.c
> index f9b2bac0..14c055bd 100644
> --- a/repair/dinode.c
> +++ b/repair/dinode.c
> @@ -194,6 +194,11 @@ clear_dinode_core(struct xfs_mount *mp, xfs_dinode_t *dinoc, xfs_ino_t ino_num)
> dinoc->di_anextents = 0;
> }
>
> + if (be32_to_cpu(dinoc->di_extsize) != 0) {
> + __dirty_no_modify_ret(dirty);
> + dinoc->di_extsize = 0;
> + }
> +
> if (dinoc->di_version > 1 &&
> be32_to_cpu(dinoc->di_nlink) != 0) {
> __dirty_no_modify_ret(dirty);
> @@ -237,6 +242,11 @@ clear_dinode_core(struct xfs_mount *mp, xfs_dinode_t *dinoc, xfs_ino_t ino_num)
> dinoc->di_changecount = 0;
> }
>
> + if (be32_to_cpu(dinoc->di_cowextsize) != 0) {
> + __dirty_no_modify_ret(dirty);
> + dinoc->di_cowextsize = 0;
> + }
> +
> return dirty;
> }
>
> @@ -2843,24 +2853,21 @@ _("bad (negative) size %" PRId64 " on inode %" PRIu64 "\n"),
> * only regular files with REALTIME or EXTSIZE flags set can have
> * extsize set, or directories with EXTSZINHERIT.
> */
> - if (be32_to_cpu(dino->di_extsize) != 0) {
> - if ((type == XR_INO_RTDATA) ||
> - (type == XR_INO_DIR && (be16_to_cpu(dino->di_flags) &
> - XFS_DIFLAG_EXTSZINHERIT)) ||
> - (type == XR_INO_DATA && (be16_to_cpu(dino->di_flags) &
> - XFS_DIFLAG_EXTSIZE))) {
> - /* s'okay */ ;
> - } else {
> - do_warn(
> -_("bad non-zero extent size %u for non-realtime/extsize inode %" PRIu64 ", "),
> - be32_to_cpu(dino->di_extsize), lino);
> - if (!no_modify) {
> - do_warn(_("resetting to zero\n"));
> - dino->di_extsize = 0;
> - *dirty = 1;
> - } else
> - do_warn(_("would reset to zero\n"));
> - }
> + if (libxfs_inode_validate_extsize(mp,
> + be32_to_cpu(dino->di_extsize),
> + be16_to_cpu(dino->di_mode),
> + be16_to_cpu(dino->di_flags)) != NULL) {
> + do_warn(
> +_("Bad extent size %u for non-realtime/extsize inode %" PRIu64 ", "),
> + be32_to_cpu(dino->di_extsize), lino);
> + if (!no_modify) {
> + do_warn(_("resetting to zero\n"));
> + dino->di_extsize = 0;
> + dino->di_flags &= ~cpu_to_be16(XFS_DIFLAG_EXTSIZE |
> + XFS_DIFLAG_EXTSZINHERIT);
> + *dirty = 1;
> + } else
> + do_warn(_("would reset to zero\n"));
> }
>
> /*
> @@ -2868,41 +2875,21 @@ _("bad non-zero extent size %u for non-realtime/extsize inode %" PRIu64 ", "),
> * set can have extsize set.
> */
> if (dino->di_version >= 3 &&
> - be32_to_cpu(dino->di_cowextsize) != 0) {
> - if ((type == XR_INO_DIR || type == XR_INO_DATA) &&
> - (be64_to_cpu(dino->di_flags2) &
> - XFS_DIFLAG2_COWEXTSIZE)) {
> - /* s'okay */ ;
> - } else {
> - do_warn(
> -_("Cannot have non-zero CoW extent size %u on non-cowextsize inode %" PRIu64 ", "),
> - be32_to_cpu(dino->di_cowextsize), lino);
> - if (!no_modify) {
> - do_warn(_("resetting to zero\n"));
> - dino->di_flags2 &= ~cpu_to_be64(XFS_DIFLAG2_COWEXTSIZE);
> - dino->di_cowextsize = 0;
> - *dirty = 1;
> - } else
> - do_warn(_("would reset to zero\n"));
> - }
> - }
> -
> - /*
> - * Can't have the COWEXTSIZE flag set with no hint.
> - */
> - if (dino->di_version >= 3 &&
> - be32_to_cpu(dino->di_cowextsize) == 0 &&
> - (be64_to_cpu(dino->di_flags2) & XFS_DIFLAG2_COWEXTSIZE)) {
> + libxfs_inode_validate_cowextsize(mp,
> + be32_to_cpu(dino->di_cowextsize),
> + be16_to_cpu(dino->di_mode),
> + be16_to_cpu(dino->di_flags),
> + be64_to_cpu(dino->di_flags2)) != NULL) {
> do_warn(
> -_("Cannot have CoW extent size of zero on cowextsize inode %" PRIu64 ", "),
> - lino);
> +_("Bad CoW extent size %u on non-cowextsize inode %" PRIu64 ", "),
> + be32_to_cpu(dino->di_cowextsize), lino);
> if (!no_modify) {
> - do_warn(_("clearing cowextsize flag\n"));
> + do_warn(_("resetting to zero\n"));
> dino->di_flags2 &= ~cpu_to_be64(XFS_DIFLAG2_COWEXTSIZE);
> + dino->di_cowextsize = 0;
> *dirty = 1;
> - } else {
> - do_warn(_("would clear cowextsize flag\n"));
> - }
> + } else
> + do_warn(_("would reset to zero\n"));
> }
>
> /* nsec fields cannot be larger than 1 billion */
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
