summaryrefslogtreecommitdiff
path: root/fs/xfs/xfs_trans_buf.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/xfs/xfs_trans_buf.c')
-rw-r--r--fs/xfs/xfs_trans_buf.c195
1 files changed, 141 insertions, 54 deletions
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 21c5a5e3700d..aa5a04b844d6 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -41,20 +41,26 @@ STATIC struct xfs_buf *
xfs_trans_buf_item_match(
struct xfs_trans *tp,
struct xfs_buftarg *target,
- xfs_daddr_t blkno,
- int len)
+ struct xfs_buf_map *map,
+ int nmaps)
{
struct xfs_log_item_desc *lidp;
struct xfs_buf_log_item *blip;
+ int len = 0;
+ int i;
+
+ for (i = 0; i < nmaps; i++)
+ len += map[i].bm_len;
- len = BBTOB(len);
list_for_each_entry(lidp, &tp->t_items, lid_trans) {
blip = (struct xfs_buf_log_item *)lidp->lid_item;
if (blip->bli_item.li_type == XFS_LI_BUF &&
blip->bli_buf->b_target == target &&
- XFS_BUF_ADDR(blip->bli_buf) == blkno &&
- BBTOB(blip->bli_buf->b_length) == len)
+ XFS_BUF_ADDR(blip->bli_buf) == map[0].bm_bn &&
+ blip->bli_buf->b_length == len) {
+ ASSERT(blip->bli_buf->b_map_count == nmaps);
return blip->bli_buf;
+ }
}
return NULL;
@@ -87,7 +93,7 @@ _xfs_trans_bjoin(
xfs_buf_item_init(bp, tp->t_mountp);
bip = bp->b_fspriv;
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
- ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL));
ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
if (reset_recur)
bip->bli_recur = 0;
@@ -128,21 +134,19 @@ xfs_trans_bjoin(
* If the transaction pointer is NULL, make this just a normal
* get_buf() call.
*/
-xfs_buf_t *
-xfs_trans_get_buf(xfs_trans_t *tp,
- xfs_buftarg_t *target_dev,
- xfs_daddr_t blkno,
- int len,
- uint flags)
+struct xfs_buf *
+xfs_trans_get_buf_map(
+ struct xfs_trans *tp,
+ struct xfs_buftarg *target,
+ struct xfs_buf_map *map,
+ int nmaps,
+ xfs_buf_flags_t flags)
{
xfs_buf_t *bp;
xfs_buf_log_item_t *bip;
- /*
- * Default to a normal get_buf() call if the tp is NULL.
- */
- if (tp == NULL)
- return xfs_buf_get(target_dev, blkno, len, flags);
+ if (!tp)
+ return xfs_buf_get_map(target, map, nmaps, flags);
/*
* If we find the buffer in the cache with this transaction
@@ -150,7 +154,7 @@ xfs_trans_get_buf(xfs_trans_t *tp,
* have it locked. In this case we just increment the lock
* recursion count and return the buffer to the caller.
*/
- bp = xfs_trans_buf_item_match(tp, target_dev, blkno, len);
+ bp = xfs_trans_buf_item_match(tp, target, map, nmaps);
if (bp != NULL) {
ASSERT(xfs_buf_islocked(bp));
if (XFS_FORCED_SHUTDOWN(tp->t_mountp)) {
@@ -167,7 +171,7 @@ xfs_trans_get_buf(xfs_trans_t *tp,
return (bp);
}
- bp = xfs_buf_get(target_dev, blkno, len, flags);
+ bp = xfs_buf_get_map(target, map, nmaps, flags);
if (bp == NULL) {
return NULL;
}
@@ -246,26 +250,23 @@ int xfs_error_mod = 33;
* read_buf() call.
*/
int
-xfs_trans_read_buf(
- xfs_mount_t *mp,
- xfs_trans_t *tp,
- xfs_buftarg_t *target,
- xfs_daddr_t blkno,
- int len,
- uint flags,
- xfs_buf_t **bpp)
+xfs_trans_read_buf_map(
+ struct xfs_mount *mp,
+ struct xfs_trans *tp,
+ struct xfs_buftarg *target,
+ struct xfs_buf_map *map,
+ int nmaps,
+ xfs_buf_flags_t flags,
+ struct xfs_buf **bpp,
+ const struct xfs_buf_ops *ops)
{
xfs_buf_t *bp;
xfs_buf_log_item_t *bip;
int error;
*bpp = NULL;
-
- /*
- * Default to a normal get_buf() call if the tp is NULL.
- */
- if (tp == NULL) {
- bp = xfs_buf_read(target, blkno, len, flags);
+ if (!tp) {
+ bp = xfs_buf_read_map(target, map, nmaps, flags, ops);
if (!bp)
return (flags & XBF_TRYLOCK) ?
EAGAIN : XFS_ERROR(ENOMEM);
@@ -303,7 +304,7 @@ xfs_trans_read_buf(
* If the buffer is not yet read in, then we read it in, increment
* the lock recursion count, and return it to the caller.
*/
- bp = xfs_trans_buf_item_match(tp, target, blkno, len);
+ bp = xfs_trans_buf_item_match(tp, target, map, nmaps);
if (bp != NULL) {
ASSERT(xfs_buf_islocked(bp));
ASSERT(bp->b_transp == tp);
@@ -312,7 +313,9 @@ xfs_trans_read_buf(
if (!(XFS_BUF_ISDONE(bp))) {
trace_xfs_trans_read_buf_io(bp, _RET_IP_);
ASSERT(!XFS_BUF_ISASYNC(bp));
+ ASSERT(bp->b_iodone == NULL);
XFS_BUF_READ(bp);
+ bp->b_ops = ops;
xfsbdstrat(tp->t_mountp, bp);
error = xfs_buf_iowait(bp);
if (error) {
@@ -349,7 +352,7 @@ xfs_trans_read_buf(
return 0;
}
- bp = xfs_buf_read(target, blkno, len, flags);
+ bp = xfs_buf_read_map(target, map, nmaps, flags, ops);
if (bp == NULL) {
*bpp = NULL;
return (flags & XBF_TRYLOCK) ?
@@ -394,7 +397,6 @@ shutdown_abort:
return XFS_ERROR(EIO);
}
-
/*
* Release the buffer bp which was previously acquired with one of the
* xfs_trans_... buffer allocation routines if the buffer has not
@@ -429,7 +431,7 @@ xfs_trans_brelse(xfs_trans_t *tp,
bip = bp->b_fspriv;
ASSERT(bip->bli_item.li_type == XFS_LI_BUF);
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
- ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL));
ASSERT(atomic_read(&bip->bli_refcount) > 0);
trace_xfs_trans_brelse(bip);
@@ -516,7 +518,7 @@ xfs_trans_bhold(xfs_trans_t *tp,
ASSERT(bp->b_transp == tp);
ASSERT(bip != NULL);
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
- ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL));
ASSERT(atomic_read(&bip->bli_refcount) > 0);
bip->bli_flags |= XFS_BLI_HOLD;
@@ -536,7 +538,7 @@ xfs_trans_bhold_release(xfs_trans_t *tp,
ASSERT(bp->b_transp == tp);
ASSERT(bip != NULL);
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
- ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL));
ASSERT(atomic_read(&bip->bli_refcount) > 0);
ASSERT(bip->bli_flags & XFS_BLI_HOLD);
@@ -595,13 +597,19 @@ xfs_trans_log_buf(xfs_trans_t *tp,
bip->bli_flags &= ~XFS_BLI_STALE;
ASSERT(XFS_BUF_ISSTALE(bp));
XFS_BUF_UNSTALE(bp);
- bip->bli_format.blf_flags &= ~XFS_BLF_CANCEL;
+ bip->__bli_format.blf_flags &= ~XFS_BLF_CANCEL;
}
tp->t_flags |= XFS_TRANS_DIRTY;
bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY;
- bip->bli_flags |= XFS_BLI_LOGGED;
- xfs_buf_item_log(bip, first, last);
+
+ /*
+ * If we have an ordered buffer we are not logging any dirty range but
+ * it still needs to be marked dirty and that it has been logged.
+ */
+ bip->bli_flags |= XFS_BLI_DIRTY | XFS_BLI_LOGGED;
+ if (!(bip->bli_flags & XFS_BLI_ORDERED))
+ xfs_buf_item_log(bip, first, last);
}
@@ -640,6 +648,7 @@ xfs_trans_binval(
xfs_buf_t *bp)
{
xfs_buf_log_item_t *bip = bp->b_fspriv;
+ int i;
ASSERT(bp->b_transp == tp);
ASSERT(bip != NULL);
@@ -654,8 +663,9 @@ xfs_trans_binval(
*/
ASSERT(XFS_BUF_ISSTALE(bp));
ASSERT(!(bip->bli_flags & (XFS_BLI_LOGGED | XFS_BLI_DIRTY)));
- ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_INODE_BUF));
- ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL);
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_INODE_BUF));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLFT_MASK));
+ ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL);
ASSERT(bip->bli_item.li_desc->lid_flags & XFS_LID_DIRTY);
ASSERT(tp->t_flags & XFS_TRANS_DIRTY);
return;
@@ -665,10 +675,13 @@ xfs_trans_binval(
bip->bli_flags |= XFS_BLI_STALE;
bip->bli_flags &= ~(XFS_BLI_INODE_BUF | XFS_BLI_LOGGED | XFS_BLI_DIRTY);
- bip->bli_format.blf_flags &= ~XFS_BLF_INODE_BUF;
- bip->bli_format.blf_flags |= XFS_BLF_CANCEL;
- memset((char *)(bip->bli_format.blf_data_map), 0,
- (bip->bli_format.blf_map_size * sizeof(uint)));
+ bip->__bli_format.blf_flags &= ~XFS_BLF_INODE_BUF;
+ bip->__bli_format.blf_flags |= XFS_BLF_CANCEL;
+ bip->__bli_format.blf_flags &= ~XFS_BLFT_MASK;
+ for (i = 0; i < bip->bli_format_count; i++) {
+ memset(bip->bli_formats[i].blf_data_map, 0,
+ (bip->bli_formats[i].blf_map_size * sizeof(uint)));
+ }
bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY;
tp->t_flags |= XFS_TRANS_DIRTY;
}
@@ -696,12 +709,13 @@ xfs_trans_inode_buf(
ASSERT(atomic_read(&bip->bli_refcount) > 0);
bip->bli_flags |= XFS_BLI_INODE_BUF;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLFT_DINO_BUF);
}
/*
* This call is used to indicate that the buffer is going to
* be staled and was an inode buffer. This means it gets
- * special processing during unpin - where any inodes
+ * special processing during unpin - where any inodes
* associated with the buffer should be removed from ail.
* There is also special processing during recovery,
* any replay of the inodes in the buffer needs to be
@@ -720,6 +734,7 @@ xfs_trans_stale_inode_buf(
bip->bli_flags |= XFS_BLI_STALE_INODE;
bip->bli_item.li_cb = xfs_buf_iodone;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLFT_DINO_BUF);
}
/*
@@ -743,8 +758,66 @@ xfs_trans_inode_alloc_buf(
ASSERT(atomic_read(&bip->bli_refcount) > 0);
bip->bli_flags |= XFS_BLI_INODE_ALLOC_BUF;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLFT_DINO_BUF);
}
+/*
+ * Mark the buffer as ordered for this transaction. This means
+ * that the contents of the buffer are not recorded in the transaction
+ * but it is tracked in the AIL as though it was. This allows us
+ * to record logical changes in transactions rather than the physical
+ * changes we make to the buffer without changing writeback ordering
+ * constraints of metadata buffers.
+ */
+void
+xfs_trans_ordered_buf(
+ struct xfs_trans *tp,
+ struct xfs_buf *bp)
+{
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ ASSERT(bp->b_transp == tp);
+ ASSERT(bip != NULL);
+ ASSERT(atomic_read(&bip->bli_refcount) > 0);
+
+ bip->bli_flags |= XFS_BLI_ORDERED;
+ trace_xfs_buf_item_ordered(bip);
+}
+
+/*
+ * Set the type of the buffer for log recovery so that it can correctly identify
+ * and hence attach the correct buffer ops to the buffer after replay.
+ */
+void
+xfs_trans_buf_set_type(
+ struct xfs_trans *tp,
+ struct xfs_buf *bp,
+ enum xfs_blft type)
+{
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ if (!tp)
+ return;
+
+ ASSERT(bp->b_transp == tp);
+ ASSERT(bip != NULL);
+ ASSERT(atomic_read(&bip->bli_refcount) > 0);
+
+ xfs_blft_to_flags(&bip->__bli_format, type);
+}
+
+void
+xfs_trans_buf_copy_type(
+ struct xfs_buf *dst_bp,
+ struct xfs_buf *src_bp)
+{
+ struct xfs_buf_log_item *sbip = src_bp->b_fspriv;
+ struct xfs_buf_log_item *dbip = dst_bp->b_fspriv;
+ enum xfs_blft type;
+
+ type = xfs_blft_from_flags(&sbip->__bli_format);
+ xfs_blft_to_flags(&dbip->__bli_format, type);
+}
/*
* Similar to xfs_trans_inode_buf(), this marks the buffer as a cluster of
@@ -763,14 +836,28 @@ xfs_trans_dquot_buf(
xfs_buf_t *bp,
uint type)
{
- xfs_buf_log_item_t *bip = bp->b_fspriv;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
- ASSERT(bp->b_transp == tp);
- ASSERT(bip != NULL);
ASSERT(type == XFS_BLF_UDQUOT_BUF ||
type == XFS_BLF_PDQUOT_BUF ||
type == XFS_BLF_GDQUOT_BUF);
- ASSERT(atomic_read(&bip->bli_refcount) > 0);
- bip->bli_format.blf_flags |= type;
+ bip->__bli_format.blf_flags |= type;
+
+ switch (type) {
+ case XFS_BLF_UDQUOT_BUF:
+ type = XFS_BLFT_UDQUOT_BUF;
+ break;
+ case XFS_BLF_PDQUOT_BUF:
+ type = XFS_BLFT_PDQUOT_BUF;
+ break;
+ case XFS_BLF_GDQUOT_BUF:
+ type = XFS_BLFT_GDQUOT_BUF;
+ break;
+ default:
+ type = XFS_BLFT_UNKNOWN_BUF;
+ break;
+ }
+
+ xfs_trans_buf_set_type(tp, bp, type);
}