On Tue, Nov 29, 2022 at 12:34:53PM +1100, Dave Chinner wrote:
> On Mon, Nov 28, 2022 at 05:21:18PM -0800, Darrick J. Wong wrote:
> > From: Darrick J. Wong <djwong@xxxxxxxxxx>
> >
> > Add a new error injection knob so that we can arbitrarily slow down
> > writeback to test for race conditions and aberrant reclaim behavior if
> > the writeback mechanisms are slow to issue writeback. This will enable
> > functional testing for the ifork sequence counters introduced in commit
> > 745b3f76d1c8 ("xfs: maintain a sequence count for inode fork
> > manipulations").
> >
> > Signed-off-by: Darrick J. Wong <djwong@xxxxxxxxxx>
> > ---
> > v2: this time with tracepoints
> > ---
> .....
>
> > @@ -267,6 +270,14 @@ xfs_errortag_valid(
> > return true;
> > }
> >
> > +bool
> > +xfs_errortag_enabled(
> > + struct xfs_mount *mp,
> > + unsigned int tag)
> > +{
> > + return mp->m_errortag && mp->m_errortag[tag] != 0;
> > +}
>
> Perhaps consider using the new xfs_errortag_valid() helper? i.e.
>
> {
> if (!mp->errortag)
> return false;
> if (!xfs_errortag_valid(tag))
> return false;
> return mp->m_errortag[tag] != 0;
Fixed.
> }
>
> > +
> > bool
> > xfs_errortag_test(
> > struct xfs_mount *mp,
> > diff --git a/fs/xfs/xfs_error.h b/fs/xfs/xfs_error.h
> > index 5191e9145e55..936d0c52d6af 100644
> > --- a/fs/xfs/xfs_error.h
> > +++ b/fs/xfs/xfs_error.h
> > @@ -45,6 +45,17 @@ extern bool xfs_errortag_test(struct xfs_mount *mp, const char *expression,
> > const char *file, int line, unsigned int error_tag);
> > #define XFS_TEST_ERROR(expr, mp, tag) \
> > ((expr) || xfs_errortag_test((mp), #expr, __FILE__, __LINE__, (tag)))
> > +bool xfs_errortag_enabled(struct xfs_mount *mp, unsigned int tag);
> > +#define XFS_ERRORTAG_DELAY(mp, tag) \
> > + do { \
> > + if (!xfs_errortag_enabled((mp), (tag))) \
> > + break; \
> > + xfs_warn_ratelimited((mp), \
> > +"Injecting %ums delay at file %s, line %d, on filesystem \"%s\"", \
> > + (mp)->m_errortag[(tag)], __FILE__, __LINE__, \
> > + (mp)->m_super->s_id); \
> > + mdelay((mp)->m_errortag[(tag)]); \
> > + } while (0)
>
> Putting a might_sleep() in this macro might be a good idea - that
> will catch delays being added inside spin lock contexts...
Done. Thanks for the review!
--D
> Other than that, it looks fine.
>
> Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>
>
> --
> Dave Chinner
> david@xxxxxxxxxxxxx
