Re: [PATCH V5 12/12] xfs: Introduce error injection to allocate only minlen size extents for files

[Chandan Babu R <chandanrlinux@gmail.com>]

On Saturday 3 October 2020 11:26:33 AM IST Chandan Babu R wrote:
> This commit adds XFS_ERRTAG_BMAP_ALLOC_MINLEN_EXTENT error tag which
> helps userspace test programs to get xfs_bmap_btalloc() to always
> allocate minlen sized extents.
> 
> This is required for test programs which need a guarantee that minlen
> extents allocated for a file do not get merged with their existing
> neighbours in the inode's BMBT. "Inode fork extent overflow check" for
> Directories, Xattrs and extension of realtime inodes need this since the
> file offset at which the extents are being allocated cannot be
> explicitly controlled from userspace.
> 
> One way to use this error tag is to,
> 1. Consume all of the free space by sequentially writing to a file.
> 2. Punch alternate blocks of the file. This causes CNTBT to contain
>    sufficient number of one block sized extent records.
> 3. Inject XFS_ERRTAG_BMAP_ALLOC_MINLEN_EXTENT error tag.
> After step 3, xfs_bmap_btalloc() will issue space allocation
> requests for minlen sized extents only.
> 
> ENOSPC error code is returned to userspace when there aren't any "one
> block sized" extents left in any of the AGs.
> 
> Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com>
> ---
>  fs/xfs/libxfs/xfs_alloc.c    | 46 ++++++++++++++++++++++++++++++++++++
>  fs/xfs/libxfs/xfs_alloc.h    |  1 +
>  fs/xfs/libxfs/xfs_bmap.c     | 26 ++++++++++++++------
>  fs/xfs/libxfs/xfs_errortag.h |  4 +++-
>  fs/xfs/xfs_error.c           |  3 +++
>  5 files changed, 72 insertions(+), 8 deletions(-)
> 
> diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c
> index 852b536551b5..d8d8ab1478db 100644
> --- a/fs/xfs/libxfs/xfs_alloc.c
> +++ b/fs/xfs/libxfs/xfs_alloc.c
> @@ -2473,6 +2473,45 @@ xfs_defer_agfl_block(
>  	xfs_defer_add(tp, XFS_DEFER_OPS_TYPE_AGFL_FREE, &new->xefi_list);
>  }
>  
> +STATIC int
> +minlen_freespace_available(
> +	struct xfs_alloc_arg	*args,
> +	struct xfs_buf		*agbp,
> +	int			*stat)
> +{
> +	xfs_btree_cur_t		*cnt_cur;
> +	xfs_agblock_t		fbno;
> +	xfs_extlen_t		flen;
> +	int			btree_error = XFS_BTREE_NOERROR;
> +	int			error = 0;
> +
> +	cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, agbp,
> +			args->agno, XFS_BTNUM_CNT);
> +	error = xfs_alloc_lookup_ge(cnt_cur, 0, args->minlen, stat);
> +	if (error) {
> +		btree_error = XFS_BTREE_ERROR;
> +		goto out;
> +	}
> +
> +	ASSERT(*stat == 1);
> +
> +	error = xfs_alloc_get_rec(cnt_cur, &fbno, &flen, stat);
> +	if (error) {
> +		btree_error = XFS_BTREE_ERROR;
> +		goto out;
> +	}
> +
> +	if (flen == args->minlen)
> +		*stat = 1;
> +	else
> +		*stat = 0;
> +
> +out:
> +	xfs_btree_del_cursor(cnt_cur, btree_error);
> +
> +	return error;
> +}
> +
>  /*
>   * Decide whether to use this allocation group for this allocation.
>   * If so, fix up the btree freelist's size.
> @@ -2490,6 +2529,7 @@ xfs_alloc_fix_freelist(
>  	struct xfs_alloc_arg	targs;	/* local allocation arguments */
>  	xfs_agblock_t		bno;	/* freelist block */
>  	xfs_extlen_t		need;	/* total blocks needed in freelist */
> +	int			i;
>  	int			error = 0;
>  
>  	/* deferred ops (AGFL block frees) require permanent transactions */
> @@ -2544,6 +2584,12 @@ xfs_alloc_fix_freelist(
>  	if (!xfs_alloc_space_available(args, need, flags))
>  		goto out_agbp_relse;
>  
> +	if (args->alloc_minlen_only) {
> +		error = minlen_freespace_available(args, agbp, &i);
> +		if (error || !i)
> +			goto out_agbp_relse;
> +	}
> +
>  	/*
>  	 * Make the freelist shorter if it's too long.
>  	 *
> diff --git a/fs/xfs/libxfs/xfs_alloc.h b/fs/xfs/libxfs/xfs_alloc.h
> index 6c22b12176b8..1d04089b7fb4 100644
> --- a/fs/xfs/libxfs/xfs_alloc.h
> +++ b/fs/xfs/libxfs/xfs_alloc.h
> @@ -75,6 +75,7 @@ typedef struct xfs_alloc_arg {
>  	char		wasfromfl;	/* set if allocation is from freelist */
>  	struct xfs_owner_info	oinfo;	/* owner of blocks being allocated */
>  	enum xfs_ag_resv_type	resv;	/* block reservation to use */
> +	bool		alloc_minlen_only;
>  } xfs_alloc_arg_t;
>  
>  /*
> diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
> index 5156cbd476f2..fab4097e7492 100644
> --- a/fs/xfs/libxfs/xfs_bmap.c
> +++ b/fs/xfs/libxfs/xfs_bmap.c
> @@ -3510,12 +3510,19 @@ xfs_bmap_btalloc(
>  		ASSERT(ap->length);
>  	}
>  
> +	memset(&args, 0, sizeof(args));
> +
> +	args.alloc_minlen_only = XFS_TEST_ERROR(false, mp,
> +					XFS_ERRTAG_BMAP_ALLOC_MINLEN_EXTENT);
>  
>  	nullfb = ap->tp->t_firstblock == NULLFSBLOCK;
>  	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp,
>  							ap->tp->t_firstblock);
>  	if (nullfb) {
> -		if ((ap->datatype & XFS_ALLOC_USERDATA) &&
> +		if (args.alloc_minlen_only) {
> +			ag = 0;
> +			ap->blkno = XFS_AGB_TO_FSB(mp, ag, 0);
> +		} else if ((ap->datatype & XFS_ALLOC_USERDATA) &&
>  		    xfs_inode_is_filestream(ap->ip)) {
>  			ag = xfs_filestream_lookup_ag(ap->ip);
>  			ag = (ag != NULLAGNUMBER) ? ag : 0;
> @@ -3523,10 +3530,12 @@ xfs_bmap_btalloc(
>  		} else {
>  			ap->blkno = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
>  		}
> -	} else
> +	} else {
>  		ap->blkno = ap->tp->t_firstblock;
> +	}
>  
> -	xfs_bmap_adjacent(ap);
> +	if (!args.alloc_minlen_only)
> +		xfs_bmap_adjacent(ap);
>  
>  	/*
>  	 * If allowed, use ap->blkno; otherwise must use firstblock since
> @@ -3540,7 +3549,6 @@ xfs_bmap_btalloc(
>  	 * Normal allocation, done through xfs_alloc_vextent.
>  	 */
>  	tryagain = isaligned = 0;
> -	memset(&args, 0, sizeof(args));
>  	args.tp = ap->tp;
>  	args.mp = mp;
>  	args.fsbno = ap->blkno;
> @@ -3549,7 +3557,10 @@ xfs_bmap_btalloc(
>  	/* Trim the allocation back to the maximum an AG can fit. */
>  	args.maxlen = min(ap->length, mp->m_ag_max_usable);
>  	blen = 0;
> -	if (nullfb) {
> +	if (args.alloc_minlen_only) {
> +		args.type = XFS_ALLOCTYPE_START_AG;

The above should have been,

args.type = XFS_ALLOCTYPE_FIRST_AG;

In my experiments, I had introduced a new args.type value and had later
realized that XFS_ALLOCTYPE_FIRST_AG would suffice for my requirements. I had
tested the changed version (which was in my git stash) and forgot to apply
that to this commit after testing was completed. Hence I ended up sending a
slightly stale patch. I am sorry about this. I will resend the series.

-- 
chandan