Merge tag 'for-linus-v3.8-rc1' of git://oss.sgi.com/xfs/xfs

Pull xfs update from Ben Myers:
 "There is plenty going on, including the cleanup of xfssyncd, metadata
  verifiers, CRC infrastructure for the log, tracking of inodes with
  speculative allocation, a cleanup of xfs_fs_subr.c, fixes for
  XFS_IOC_ZERO_RANGE, and important fix related to log replay (only
  update the last_sync_lsn when a transaction completes), a fix for
  deadlock on AGF buffers, documentation and comment updates, and a few
  more cleanups and fixes.

  Details:
   - remove the xfssyncd mess
   - only update the last_sync_lsn when a transaction completes
   - zero allocation_args on the kernel stack
   - fix AGF/alloc workqueue deadlock
   - silence uninitialised f.file warning
   - Update inode alloc comments
   - Update mount options documentation
   - report projid32bit feature in geometry call
   - speculative preallocation inode tracking
   - fix attr tree double split corruption
   - fix broken error handling in xfs_vm_writepage
   - drop buffer io reference when a bad bio is built
   - add more attribute tree trace points
   - growfs infrastructure changes for 3.8
   - fs/xfs/xfs_fs_subr.c die die die
   - add CRC infrastructure
   - add CRC checks to the log
   - Remove description of nodelaylog mount option from xfs.txt
   - inode allocation should use unmapped buffers
   - byte range granularity for XFS_IOC_ZERO_RANGE
   - fix direct IO nested transaction deadlock
   - fix stray dquot unlock when reclaiming dquots
   - fix sparse reported log CRC endian issue"

Fix up trivial conflict in fs/xfs/xfs_fsops.c due to the same patch
having been applied twice (commits eaef854335ce and 1375cb65e87b: "xfs:
growfs: don't read garbage for new secondary superblocks") with later
updates to the affected code in the XFS tree.

* tag 'for-linus-v3.8-rc1' of git://oss.sgi.com/xfs/xfs: (78 commits)
  xfs: fix sparse reported log CRC endian issue
  xfs: fix stray dquot unlock when reclaiming dquots
  xfs: fix direct IO nested transaction deadlock.
  xfs: byte range granularity for XFS_IOC_ZERO_RANGE
  xfs: inode allocation should use unmapped buffers.
  xfs: Remove the description of nodelaylog mount option from xfs.txt
  xfs: add CRC checks to the log
  xfs: add CRC infrastructure
  xfs: convert buffer verifiers to an ops structure.
  xfs: connect up write verifiers to new buffers
  xfs: add pre-write metadata buffer verifier callbacks
  xfs: add buffer pre-write callback
  xfs: Add verifiers to dir2 data readahead.
  xfs: add xfs_da_node verification
  xfs: factor and verify attr leaf reads
  xfs: factor dir2 leaf read
  xfs: factor out dir2 data block reading
  xfs: factor dir2 free block reading
  xfs: verify dir2 block format buffers
  xfs: factor dir2 block read operations
  ...

1 files changed, 72 insertions, 74 deletions

diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index d308749fabf..96fcbb85ff8 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -41,7 +41,9 @@
 #include "xfs_trans_priv.h"
 #include "xfs_quota.h"
 #include "xfs_utils.h"
+#include "xfs_cksum.h"
 #include "xfs_trace.h"
+#include "xfs_icache.h"
 
 STATIC int
 xlog_find_zeroed(
@@ -2143,7 +2145,7 @@ xlog_recover_buffer_pass2(
 		buf_flags |= XBF_UNMAPPED;
 
 	bp = xfs_buf_read(mp->m_ddev_targp, buf_f->blf_blkno, buf_f->blf_len,
-			  buf_flags);
+			  buf_flags, NULL);
 	if (!bp)
 		return XFS_ERROR(ENOMEM);
 	error = bp->b_error;
@@ -2236,7 +2238,8 @@ xlog_recover_inode_pass2(
 	}
 	trace_xfs_log_recover_inode_recover(log, in_f);
 
-	bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0);
+	bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0,
+			  NULL);
 	if (!bp) {
 		error = ENOMEM;
 		goto error;
@@ -2547,7 +2550,8 @@ xlog_recover_dquot_pass2(
 	ASSERT(dq_f->qlf_len == 1);
 
 	error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dq_f->qlf_blkno,
-				   XFS_FSB_TO_BB(mp, dq_f->qlf_len), 0, &bp);
+				   XFS_FSB_TO_BB(mp, dq_f->qlf_len), 0, &bp,
+				   NULL);
 	if (error)
 		return error;
 
@@ -3213,80 +3217,58 @@ xlog_recover_process_iunlinks(
 	mp->m_dmevmask = mp_dmevmask;
 }
 
-
-#ifdef DEBUG
-STATIC void
-xlog_pack_data_checksum(
-	struct xlog		*log,
-	struct xlog_in_core	*iclog,
-	int			size)
-{
-	int		i;
-	__be32		*up;
-	uint		chksum = 0;
-
-	up = (__be32 *)iclog->ic_datap;
-	/* divide length by 4 to get # words */
-	for (i = 0; i < (size >> 2); i++) {
-		chksum ^= be32_to_cpu(*up);
-		up++;
-	}
-	iclog->ic_header.h_chksum = cpu_to_be32(chksum);
-}
-#else
-#define xlog_pack_data_checksum(log, iclog, size)
-#endif
-
 /*
- * Stamp cycle number in every block
+ * Upack the log buffer data and crc check it. If the check fails, issue a
+ * warning if and only if the CRC in the header is non-zero. This makes the
+ * check an advisory warning, and the zero CRC check will prevent failure
+ * warnings from being emitted when upgrading the kernel from one that does not
+ * add CRCs by default.
+ *
+ * When filesystems are CRC enabled, this CRC mismatch becomes a fatal log
+ * corruption failure
  */
-void
-xlog_pack_data(
-	struct xlog		*log,
-	struct xlog_in_core	*iclog,
-	int			roundoff)
+STATIC int
+xlog_unpack_data_crc(
+	struct xlog_rec_header	*rhead,
+	xfs_caddr_t		dp,
+	struct xlog		*log)
 {
-	int			i, j, k;
-	int			size = iclog->ic_offset + roundoff;
-	__be32			cycle_lsn;
-	xfs_caddr_t		dp;
-
-	xlog_pack_data_checksum(log, iclog, size);
-
-	cycle_lsn = CYCLE_LSN_DISK(iclog->ic_header.h_lsn);
-
-	dp = iclog->ic_datap;
-	for (i = 0; i < BTOBB(size) &&
-		i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) {
-		iclog->ic_header.h_cycle_data[i] = *(__be32 *)dp;
-		*(__be32 *)dp = cycle_lsn;
-		dp += BBSIZE;
-	}
-
-	if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) {
-		xlog_in_core_2_t *xhdr = iclog->ic_data;
-
-		for ( ; i < BTOBB(size); i++) {
-			j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
-			k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
-			xhdr[j].hic_xheader.xh_cycle_data[k] = *(__be32 *)dp;
-			*(__be32 *)dp = cycle_lsn;
-			dp += BBSIZE;
+	__le32			crc;
+
+	crc = xlog_cksum(log, rhead, dp, be32_to_cpu(rhead->h_len));
+	if (crc != rhead->h_crc) {
+		if (rhead->h_crc || xfs_sb_version_hascrc(&log->l_mp->m_sb)) {
+			xfs_alert(log->l_mp,
+		"log record CRC mismatch: found 0x%x, expected 0x%x.\n",
+					le32_to_cpu(rhead->h_crc),
+					le32_to_cpu(crc));
+			xfs_hex_dump(dp, 32);
 		}
 
-		for (i = 1; i < log->l_iclog_heads; i++) {
-			xhdr[i].hic_xheader.xh_cycle = cycle_lsn;
-		}
+		/*
+		 * If we've detected a log record corruption, then we can't
+		 * recover past this point. Abort recovery if we are enforcing
+		 * CRC protection by punting an error back up the stack.
+		 */
+		if (xfs_sb_version_hascrc(&log->l_mp->m_sb))
+			return EFSCORRUPTED;
 	}
+
+	return 0;
 }
 
-STATIC void
+STATIC int
 xlog_unpack_data(
 	struct xlog_rec_header	*rhead,
 	xfs_caddr_t		dp,
 	struct xlog		*log)
 {
 	int			i, j, k;
+	int			error;
+
+	error = xlog_unpack_data_crc(rhead, dp, log);
+	if (error)
+		return error;
 
 	for (i = 0; i < BTOBB(be32_to_cpu(rhead->h_len)) &&
 		  i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) {
@@ -3303,6 +3285,8 @@ xlog_unpack_data(
 			dp += BBSIZE;
 		}
 	}
+
+	return 0;
 }
 
 STATIC int
@@ -3434,9 +3418,13 @@ xlog_do_recovery_pass(
 			if (error)
 				goto bread_err2;
 
-			xlog_unpack_data(rhead, offset, log);
-			if ((error = xlog_recover_process_data(log,
-						rhash, rhead, offset, pass)))
+			error = xlog_unpack_data(rhead, offset, log);
+			if (error)
+				goto bread_err2;
+
+			error = xlog_recover_process_data(log,
+						rhash, rhead, offset, pass);
+			if (error)
 				goto bread_err2;
 			blk_no += bblks + hblks;
 		}
@@ -3546,9 +3534,14 @@ xlog_do_recovery_pass(
 				if (error)
 					goto bread_err2;
 			}
-			xlog_unpack_data(rhead, offset, log);
-			if ((error = xlog_recover_process_data(log, rhash,
-							rhead, offset, pass)))
+
+			error = xlog_unpack_data(rhead, offset, log);
+			if (error)
+				goto bread_err2;
+
+			error = xlog_recover_process_data(log, rhash,
+							rhead, offset, pass);
+			if (error)
 				goto bread_err2;
 			blk_no += bblks;
 		}
@@ -3573,9 +3566,13 @@ xlog_do_recovery_pass(
 			if (error)
 				goto bread_err2;
 
-			xlog_unpack_data(rhead, offset, log);
-			if ((error = xlog_recover_process_data(log, rhash,
-							rhead, offset, pass)))
+			error = xlog_unpack_data(rhead, offset, log);
+			if (error)
+				goto bread_err2;
+
+			error = xlog_recover_process_data(log, rhash,
+							rhead, offset, pass);
+			if (error)
 				goto bread_err2;
 			blk_no += bblks + hblks;
 		}
@@ -3689,13 +3686,14 @@ xlog_do_recover(
 
 	/*
 	 * Now that we've finished replaying all buffer and inode
-	 * updates, re-read in the superblock.
+	 * updates, re-read in the superblock and reverify it.
 	 */
 	bp = xfs_getsb(log->l_mp, 0);
 	XFS_BUF_UNDONE(bp);
 	ASSERT(!(XFS_BUF_ISWRITE(bp)));
 	XFS_BUF_READ(bp);
 	XFS_BUF_UNASYNC(bp);
+	bp->b_ops = &xfs_sb_buf_ops;
 	xfsbdstrat(log->l_mp, bp);
 	error = xfs_buf_iowait(bp);
 	if (error) {
@@ -3707,7 +3705,7 @@ xlog_do_recover(
 
 	/* Convert superblock from on-disk format */
 	sbp = &log->l_mp->m_sb;
-	xfs_sb_from_disk(log->l_mp, XFS_BUF_TO_SBP(bp));
+	xfs_sb_from_disk(sbp, XFS_BUF_TO_SBP(bp));
 	ASSERT(sbp->sb_magicnum == XFS_SB_MAGIC);
 	ASSERT(xfs_sb_good_version(sbp));
 	xfs_buf_relse(bp);