Re: [PATCH v13 06/10] xfs: Add xfs_attr_set_deferred and xfs_attr_remove_deferred

From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: Allison Henderson <allison.henderson@oracle.com>
Cc: linux-xfs@vger.kernel.org
Subject: Re: [PATCH v13 06/10] xfs: Add xfs_attr_set_deferred and xfs_attr_remove_deferred
Date: Fri, 13 Nov 2020 18:03:25 -0800	[thread overview]
Message-ID: <20201114020325.GO9695@magnolia> (raw)
In-Reply-To: <2f9cc387-fca2-7978-76e5-235e095740bc@oracle.com>

On Thu, Nov 12, 2020 at 06:27:38PM -0700, Allison Henderson wrote:
> 
> 
> On 11/10/20 1:15 PM, Darrick J. Wong wrote:
> > On Thu, Oct 22, 2020 at 11:34:31PM -0700, Allison Henderson wrote:
> > > From: Allison Collins <allison.henderson@oracle.com>
> > > 
> > > These routines to set up and start a new deferred attribute operations.
> > > These functions are meant to be called by any routine needing to
> > > initiate a deferred attribute operation as opposed to the existing
> > > inline operations. New helper function xfs_attr_item_init also added.
> > > 
> > > Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
> > > ---
> > >   fs/xfs/libxfs/xfs_attr.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++
> > >   fs/xfs/libxfs/xfs_attr.h |  2 ++
> > >   2 files changed, 56 insertions(+)
> > > 
> > > diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
> > > index 760383c..7fe5554 100644
> > > --- a/fs/xfs/libxfs/xfs_attr.c
> > > +++ b/fs/xfs/libxfs/xfs_attr.c
> > > @@ -25,6 +25,7 @@
> > >   #include "xfs_trans_space.h"
> > >   #include "xfs_trace.h"
> > >   #include "xfs_attr_item.h"
> > > +#include "xfs_attr.h"
> > >   /*
> > >    * xfs_attr.c
> > > @@ -643,6 +644,59 @@ xfs_attr_set(
> > >   	goto out_unlock;
> > >   }
> > > +STATIC int
> > > +xfs_attr_item_init(
> > > +	struct xfs_da_args	*args,
> > > +	unsigned int		op_flags,	/* op flag (set or remove) */
> > > +	struct xfs_attr_item	**attr)		/* new xfs_attr_item */
> > > +{
> > > +
> > > +	struct xfs_attr_item	*new;
> > > +
> > > +	new = kmem_alloc_large(sizeof(struct xfs_attr_item), KM_NOFS);
> > 
> > I don't think we need _large allocations for struct xfs_attr_item, right?
> I will try it and see, I think it should be ok, one of the new test cases
> I'm using does try to progressively add larger and larger attrs. If it
> doesnt work, I'll make a note of it though.

Ok.  Note that kmem_alloc will only return heap objects from memory
that's directly addressable by the kerneli (and can't be larger than a
page), whereas kmem_alloc_large can fall back to virtually mapped
memory, which is scarce on 32-bit systems.

--D

> > 
> > > +	memset(new, 0, sizeof(struct xfs_attr_item));
> > 
> > Use kmem_zalloc and you won't have to memset.  Better yet, zalloc will
> > get you memory that's been pre-zeroed in the background.
> > 
> > > +	new->xattri_op_flags = op_flags;
> > > +	new->xattri_dac.da_args = args;
> > > +
> > > +	*attr = new;
> > > +	return 0;
> > > +}
> > > +
> > > +/* Sets an attribute for an inode as a deferred operation */
> > > +int
> > > +xfs_attr_set_deferred(
> > > +	struct xfs_da_args	*args)
> > > +{
> > > +	struct xfs_attr_item	*new;
> > > +	int			error = 0;
> > > +
> > > +	error = xfs_attr_item_init(args, XFS_ATTR_OP_FLAGS_SET, &new);
> > > +	if (error)
> > > +		return error;
> > > +
> > > +	xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new->xattri_list);
> > > +
> > > +	return 0;
> > > +}
> > 
> > The changes in "xfs: enable delayed attributes" should be moved to this
> > patch so that these new functions immediately have callers.
> Sure, will merge those patches together then
> 
> > 
> > (Also see the reply I sent to the next patch, which will avoid weird
> > regressions if someone's bisect lands in the middle of this series...)
> > 
> > --D
> > 
> > > +
> > > +/* Removes an attribute for an inode as a deferred operation */
> > > +int
> > > +xfs_attr_remove_deferred(
> > > +	struct xfs_da_args	*args)
> > > +{
> > > +
> > > +	struct xfs_attr_item	*new;
> > > +	int			error;
> > > +
> > > +	error  = xfs_attr_item_init(args, XFS_ATTR_OP_FLAGS_REMOVE, &new);
> > > +	if (error)
> > > +		return error;
> > > +
> > > +	xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new->xattri_list);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > >   /*========================================================================
> > >    * External routines when attribute list is inside the inode
> > >    *========================================================================*/
> > > diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
> > > index 5b4a1ca..8a08411 100644
> > > --- a/fs/xfs/libxfs/xfs_attr.h
> > > +++ b/fs/xfs/libxfs/xfs_attr.h
> > > @@ -307,5 +307,7 @@ bool xfs_attr_namecheck(const void *name, size_t length);
> > >   void xfs_delattr_context_init(struct xfs_delattr_context *dac,
> > >   			      struct xfs_da_args *args);
> > >   int xfs_attr_calc_size(struct xfs_da_args *args, int *local);
> > > +int xfs_attr_set_deferred(struct xfs_da_args *args);
> > > +int xfs_attr_remove_deferred(struct xfs_da_args *args);
> > >   #endif	/* __XFS_ATTR_H__ */
> > > -- 
> > > 2.7.4
> > > 