xfs: extent size hints can round up extents past MAXEXTLEN

This results in BMBT corruption, as seen by this test:

# mkfs.xfs -f -d size=40051712b,agcount=4 /dev/vdc
....
# mount /dev/vdc /mnt/scratch
# xfs_io -ft -c "extsize 16m" -c "falloc 0 30g" -c "bmap -vp" /mnt/scratch/foo

which results in this failure on a debug kernel:

XFS: Assertion failed: (blockcount & xfs_mask64hi(64-BMBT_BLOCKCOUNT_BITLEN)) == 0, file: fs/xfs/libxfs/xfs_bmap_btree.c, line: 211
....
Call Trace:
 [<ffffffff814cf0ff>] xfs_bmbt_set_allf+0x8f/0x100
 [<ffffffff814cf18d>] xfs_bmbt_set_all+0x1d/0x20
 [<ffffffff814f2efe>] xfs_iext_insert+0x9e/0x120
 [<ffffffff814c7956>] ? xfs_bmap_add_extent_hole_real+0x1c6/0xc70
 [<ffffffff814c7956>] xfs_bmap_add_extent_hole_real+0x1c6/0xc70
 [<ffffffff814caaab>] xfs_bmapi_write+0x72b/0xed0
 [<ffffffff811c72ac>] ? kmem_cache_alloc+0x15c/0x170
 [<ffffffff814fe070>] xfs_alloc_file_space+0x160/0x400
 [<ffffffff81ddcc29>] ? down_write+0x29/0x60
 [<ffffffff815063eb>] xfs_file_fallocate+0x29b/0x310
 [<ffffffff811d2bc8>] ? __sb_start_write+0x58/0x120
 [<ffffffff811e3e18>] ? do_vfs_ioctl+0x318/0x570
 [<ffffffff811cd680>] vfs_fallocate+0x140/0x260
 [<ffffffff811ce6f8>] SyS_fallocate+0x48/0x80
 [<ffffffff81ddec09>] system_call_fastpath+0x12/0x17

The tracepoint that indicates the extent that triggered the assert
failure is:

xfs_iext_insert:   idx 0 offset 0 block 16777224 count 2097152 flag 1

Clearly indicating that the extent length is greater than MAXEXTLEN,
which is 2097151. A prior trace point shows the allocation was an
exact size match and that a length greater than MAXEXTLEN was asked
for:

xfs_alloc_size_done:  agno 1 agbno 8 minlen 2097152 maxlen 2097152
					    ^^^^^^^        ^^^^^^^

We don't see this problem with extent size hints through the IO path
because we can't do single IOs large enough to trigger MAXEXTLEN
allocation. fallocate(), OTOH, is not limited in it's allocation
sizes and so needs help here.

The issue is that the extent size hint alignment is rounding up the
extent size past MAXEXTLEN, because xfs_bmapi_write() is not taking
into account extent size hints when calculating the maximum extent
length to allocate. xfs_bmapi_reserve_delalloc() is already doing
this, but direct extent allocation is not.

Unfortunately, the calculation in xfs_bmapi_reserve_delalloc() is
wrong, and it works only because delayed allocation extents are not
limited in size to MAXEXTLEN in the in-core extent tree. hence this
calculation does not work for direct allocation, and the delalloc
code needs fixing. This may, in fact be the underlying bug that
occassionally causes transaction overruns in delayed allocation
extent conversion, so now we know it's wrong we should fix it, too.
Many thanks to Brian Foster for finding this problem during review
of this patch.

Hence the fix, after much code reading, is to allow
xfs_bmap_extsize_align() to align partial extents when full
alignment would extend the alignment past MAXEXTLEN. We can safely
do this because all callers have higher layer allocation loops that
already handle short allocations, and so will simply run another
allocation to cover the remainder of the requested allocation range
that we ignored during alignment. The advantage of this approach is
that it also removes the need for callers to do anything other than
limit their requests to MAXEXTLEN - they don't really need to be
aware of extent size hints at all.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>

fs/xfs/libxfs/xfs_bmap.c

@@ -3224,12 +3224,24 @@ xfs_bmap_extsize_align(
 		align_alen += temp;
 		align_off -= temp;
 	}
+
+	/* Same adjustment for the end of the requested area. */
+	temp = (align_alen % extsz);
+	if (temp)
+		align_alen += extsz - temp;
+
 	/*
-	 * Same adjustment for the end of the requested area.
+	 * For large extent hint sizes, the aligned extent might be larger than
+	 * MAXEXTLEN. In that case, reduce the size by an extsz so that it pulls
+	 * the length back under MAXEXTLEN. The outer allocation loops handle
+	 * short allocation just fine, so it is safe to do this. We only want to
+	 * do it when we are forced to, though, because it means more allocation
+	 * operations are required.
 	 */
-	if ((temp = (align_alen % extsz))) {
-		align_alen += extsz - temp;
-	}
+	while (align_alen > MAXEXTLEN)
+		align_alen -= extsz;
+	ASSERT(align_alen <= MAXEXTLEN);
+
 	/*
 	 * If the previous block overlaps with this proposed allocation
 	 * then move the start forward without adjusting the length.
@@ -3318,7 +3330,9 @@ xfs_bmap_extsize_align(
 			return -EINVAL;
 	} else {
 		ASSERT(orig_off >= align_off);
-		ASSERT(orig_end <= align_off + align_alen);
+		/* see MAXEXTLEN handling above */
+		ASSERT(orig_end <= align_off + align_alen ||
+		       align_alen + extsz > MAXEXTLEN);
 	}
 
 #ifdef DEBUG
@@ -4099,13 +4113,6 @@ xfs_bmapi_reserve_delalloc(
 	/* Figure out the extent size, adjust alen */
 	extsz = xfs_get_extsz_hint(ip);
 	if (extsz) {
-		/*
-		 * Make sure we don't exceed a single extent length when we
-		 * align the extent by reducing length we are going to
-		 * allocate by the maximum amount extent size aligment may
-		 * require.
-		 */
-		alen = XFS_FILBLKS_MIN(len, MAXEXTLEN - (2 * extsz - 1));
 		error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof,
 					       1, 0, &aoff, &alen);
 		ASSERT(!error);