Merge tag 'xfs-5.7-merge-8' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull xfs updates from Darrick Wong:
 "There's a lot going on this cycle with cleanups in the log code, the
  btree code, and the xattr code.

  We're tightening of metadata validation and online fsck checking, and
  introducing a common btree rebuilding library so that we can refactor
  xfs_repair and introduce online repair in a future cycle.

  We also fixed a few visible bugs -- most notably there's one in
  getdents that we introduced in 5.6; and a fix for hangs when disabling
  quotas.

  This series has been running fstests & other QA in the background for
  over a week and looks good so far.

  I anticipate sending a second pull request next week. That batch will
  change how xfs interacts with memory reclaim; how the log batches and
  throttles log items; how hard writes near ENOSPC will try to squeeze
  more space out of the filesystem; and hopefully fix the last of the
  umount hangs after a catastrophic failure. That should ease a lot of
  problems when running at the limits, but for now I'm leaving that in
  for-next for another week to make sure we got all the subtleties
  right.

  Summary:

   - Fix a hard to trigger race between iclog error checking and log
     shutdown.

   - Strengthen the AGF verifier.

   - Ratelimit some of the more spammy error messages.

   - Remove the icdinode uid/gid members and just use the ones in the
     vfs inode.

   - Hold ILOCK across insert/collapse range.

   - Clean up the extended attribute interfaces.

   - Clean up the attr flags mess.

   - Restore PF_MEMALLOC after exiting xfsaild thread to avoid
     triggering warnings in the process accounting code.

   - Remove the flexibly-sized array from struct xfs_agfl to eliminate
     compiler warnings about unaligned pointers and packed structures.

   - Various macro and typedef removals.

   - Stale metadata buffers if we decide they're corrupt outside of a
     verifier.

   - Check directory data/block/free block owners.

   - Fix a UAF when aborting inactivation of a corrupt xattr fork.

   - Teach online scrub to report failed directory and attr name lookups
     as a metadata corruption instead of a runtime error.

   - Avoid potential buffer overflows in sysfs files by using scnprintf.

   - Fix a regression in getdents lookups due to a mistake in pointer
     arithmetic.

   - Refactor btree cursor private data structures to use anonymous
     unions.

   - Cleanups in the log unmounting code.

   - Fix a potential mishandling of ENOMEM errors on multi-block
     directory buffer lookups.

   - Fix an incorrect test in the block allocation code.

   - Cleanups and name prefix shortening in the scrub code.

   - Introduce btree bulk loading code for online repair and scrub.

   - Fix a quotaoff log item leak (and hang) when the fs goes down
     midway through a quotaoff operation.

   - Remove di_version from the incore inode.

   - Refactor some of the log shutdown checking code.

   - Record the forcing of the log unmount records in the log force
     counters.

   - Fix a longstanding bug where quotacheck would purge the
     administrator's default quota grace interval and warning limits.

   - Reduce memory usage when scrubbing directory and xattr trees.

   - Don't let fsfreeze race with GETFSMAP or online scrub.

   - Handle bio_add_page failures more gracefully in xlog_write_iclog"

* tag 'xfs-5.7-merge-8' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (108 commits)
  xfs: prohibit fs freezing when using empty transactions
  xfs: shutdown on failure to add page to log bio
  xfs: directory bestfree check should release buffers
  xfs: drop all altpath buffers at the end of the sibling check
  xfs: preserve default grace interval during quotacheck
  xfs: remove xlog_state_want_sync
  xfs: move the ioerror check out of xlog_state_clean_iclog
  xfs: refactor xlog_state_clean_iclog
  xfs: remove the aborted parameter to xlog_state_done_syncing
  xfs: simplify log shutdown checking in xfs_log_release_iclog
  xfs: simplify the xfs_log_release_iclog calling convention
  xfs: factor out a xlog_wait_on_iclog helper
  xfs: merge xlog_cil_push into xlog_cil_push_work
  xfs: remove the di_version field from struct icdinode
  xfs: simplify a check in xfs_ioctl_setattr_check_cowextsize
  xfs: simplify di_flags2 inheritance in xfs_ialloc
  xfs: only check the superblock version for dinode size calculation
  xfs: add a new xfs_sb_version_has_v3inode helper
  xfs: fix unmount hang and memory leak on shutdown during quotaoff
  xfs: factor out quotaoff intent AIL removal and memory free
  ...

1 files changed, 77 insertions, 27 deletions

diff --git a/fs/xfs/libxfs/xfs_refcount_btree.c b/fs/xfs/libxfs/xfs_refcount_btree.c
index 38529dbacd55..7fd6044a4f78 100644
--- a/fs/xfs/libxfs/xfs_refcount_btree.c
+++ b/fs/xfs/libxfs/xfs_refcount_btree.c
@@ -12,6 +12,7 @@
 #include "xfs_sb.h"
 #include "xfs_mount.h"
 #include "xfs_btree.h"
+#include "xfs_btree_staging.h"
 #include "xfs_refcount_btree.h"
 #include "xfs_alloc.h"
 #include "xfs_error.h"
@@ -25,7 +26,7 @@ xfs_refcountbt_dup_cursor(
 	struct xfs_btree_cur	*cur)
 {
 	return xfs_refcountbt_init_cursor(cur->bc_mp, cur->bc_tp,
-			cur->bc_private.a.agbp, cur->bc_private.a.agno);
+			cur->bc_ag.agbp, cur->bc_ag.agno);
 }
 
 STATIC void
@@ -34,8 +35,8 @@ xfs_refcountbt_set_root(
 	union xfs_btree_ptr	*ptr,
 	int			inc)
 {
-	struct xfs_buf		*agbp = cur->bc_private.a.agbp;
-	struct xfs_agf		*agf = XFS_BUF_TO_AGF(agbp);
+	struct xfs_buf		*agbp = cur->bc_ag.agbp;
+	struct xfs_agf		*agf = agbp->b_addr;
 	xfs_agnumber_t		seqno = be32_to_cpu(agf->agf_seqno);
 	struct xfs_perag	*pag = xfs_perag_get(cur->bc_mp, seqno);
 
@@ -57,8 +58,8 @@ xfs_refcountbt_alloc_block(
 	union xfs_btree_ptr	*new,
 	int			*stat)
 {
-	struct xfs_buf		*agbp = cur->bc_private.a.agbp;
-	struct xfs_agf		*agf = XFS_BUF_TO_AGF(agbp);
+	struct xfs_buf		*agbp = cur->bc_ag.agbp;
+	struct xfs_agf		*agf = agbp->b_addr;
 	struct xfs_alloc_arg	args;		/* block allocation args */
 	int			error;		/* error return value */
 
@@ -66,7 +67,7 @@ xfs_refcountbt_alloc_block(
 	args.tp = cur->bc_tp;
 	args.mp = cur->bc_mp;
 	args.type = XFS_ALLOCTYPE_NEAR_BNO;
-	args.fsbno = XFS_AGB_TO_FSB(cur->bc_mp, cur->bc_private.a.agno,
+	args.fsbno = XFS_AGB_TO_FSB(cur->bc_mp, cur->bc_ag.agno,
 			xfs_refc_block(args.mp));
 	args.oinfo = XFS_RMAP_OINFO_REFC;
 	args.minlen = args.maxlen = args.prod = 1;
@@ -75,13 +76,13 @@ xfs_refcountbt_alloc_block(
 	error = xfs_alloc_vextent(&args);
 	if (error)
 		goto out_error;
-	trace_xfs_refcountbt_alloc_block(cur->bc_mp, cur->bc_private.a.agno,
+	trace_xfs_refcountbt_alloc_block(cur->bc_mp, cur->bc_ag.agno,
 			args.agbno, 1);
 	if (args.fsbno == NULLFSBLOCK) {
 		*stat = 0;
 		return 0;
 	}
-	ASSERT(args.agno == cur->bc_private.a.agno);
+	ASSERT(args.agno == cur->bc_ag.agno);
 	ASSERT(args.len == 1);
 
 	new->s = cpu_to_be32(args.agbno);
@@ -101,12 +102,12 @@ xfs_refcountbt_free_block(
 	struct xfs_buf		*bp)
 {
 	struct xfs_mount	*mp = cur->bc_mp;
-	struct xfs_buf		*agbp = cur->bc_private.a.agbp;
-	struct xfs_agf		*agf = XFS_BUF_TO_AGF(agbp);
+	struct xfs_buf		*agbp = cur->bc_ag.agbp;
+	struct xfs_agf		*agf = agbp->b_addr;
 	xfs_fsblock_t		fsbno = XFS_DADDR_TO_FSB(mp, XFS_BUF_ADDR(bp));
 	int			error;
 
-	trace_xfs_refcountbt_free_block(cur->bc_mp, cur->bc_private.a.agno,
+	trace_xfs_refcountbt_free_block(cur->bc_mp, cur->bc_ag.agno,
 			XFS_FSB_TO_AGBNO(cur->bc_mp, fsbno), 1);
 	be32_add_cpu(&agf->agf_refcount_blocks, -1);
 	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_REFCOUNT_BLOCKS);
@@ -169,9 +170,9 @@ xfs_refcountbt_init_ptr_from_cur(
 	struct xfs_btree_cur	*cur,
 	union xfs_btree_ptr	*ptr)
 {
-	struct xfs_agf		*agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp);
+	struct xfs_agf		*agf = cur->bc_ag.agbp->b_addr;
 
-	ASSERT(cur->bc_private.a.agno == be32_to_cpu(agf->agf_seqno));
+	ASSERT(cur->bc_ag.agno == be32_to_cpu(agf->agf_seqno));
 
 	ptr->s = agf->agf_refcount_root;
 }
@@ -311,41 +312,90 @@ static const struct xfs_btree_ops xfs_refcountbt_ops = {
 };
 
 /*
- * Allocate a new refcount btree cursor.
+ * Initialize a new refcount btree cursor.
  */
-struct xfs_btree_cur *
-xfs_refcountbt_init_cursor(
+static struct xfs_btree_cur *
+xfs_refcountbt_init_common(
 	struct xfs_mount	*mp,
 	struct xfs_trans	*tp,
-	struct xfs_buf		*agbp,
 	xfs_agnumber_t		agno)
 {
-	struct xfs_agf		*agf = XFS_BUF_TO_AGF(agbp);
 	struct xfs_btree_cur	*cur;
 
 	ASSERT(agno != NULLAGNUMBER);
 	ASSERT(agno < mp->m_sb.sb_agcount);
-	cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_NOFS);
 
+	cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_NOFS);
 	cur->bc_tp = tp;
 	cur->bc_mp = mp;
 	cur->bc_btnum = XFS_BTNUM_REFC;
 	cur->bc_blocklog = mp->m_sb.sb_blocklog;
-	cur->bc_ops = &xfs_refcountbt_ops;
 	cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_refcbt_2);
 
-	cur->bc_nlevels = be32_to_cpu(agf->agf_refcount_level);
-
-	cur->bc_private.a.agbp = agbp;
-	cur->bc_private.a.agno = agno;
+	cur->bc_ag.agno = agno;
 	cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
 
-	cur->bc_private.a.priv.refc.nr_ops = 0;
-	cur->bc_private.a.priv.refc.shape_changes = 0;
+	cur->bc_ag.refc.nr_ops = 0;
+	cur->bc_ag.refc.shape_changes = 0;
+	cur->bc_ops = &xfs_refcountbt_ops;
+	return cur;
+}
 
+/* Create a btree cursor. */
+struct xfs_btree_cur *
+xfs_refcountbt_init_cursor(
+	struct xfs_mount	*mp,
+	struct xfs_trans	*tp,
+	struct xfs_buf		*agbp,
+	xfs_agnumber_t		agno)
+{
+	struct xfs_agf		*agf = agbp->b_addr;
+	struct xfs_btree_cur	*cur;
+
+	cur = xfs_refcountbt_init_common(mp, tp, agno);
+	cur->bc_nlevels = be32_to_cpu(agf->agf_refcount_level);
+	cur->bc_ag.agbp = agbp;
 	return cur;
 }
 
+/* Create a btree cursor with a fake root for staging. */
+struct xfs_btree_cur *
+xfs_refcountbt_stage_cursor(
+	struct xfs_mount	*mp,
+	struct xbtree_afakeroot	*afake,
+	xfs_agnumber_t		agno)
+{
+	struct xfs_btree_cur	*cur;
+
+	cur = xfs_refcountbt_init_common(mp, NULL, agno);
+	xfs_btree_stage_afakeroot(cur, afake);
+	return cur;
+}
+
+/*
+ * Swap in the new btree root.  Once we pass this point the newly rebuilt btree
+ * is in place and we have to kill off all the old btree blocks.
+ */
+void
+xfs_refcountbt_commit_staged_btree(
+	struct xfs_btree_cur	*cur,
+	struct xfs_trans	*tp,
+	struct xfs_buf		*agbp)
+{
+	struct xfs_agf		*agf = agbp->b_addr;
+	struct xbtree_afakeroot	*afake = cur->bc_ag.afake;
+
+	ASSERT(cur->bc_flags & XFS_BTREE_STAGING);
+
+	agf->agf_refcount_root = cpu_to_be32(afake->af_root);
+	agf->agf_refcount_level = cpu_to_be32(afake->af_levels);
+	agf->agf_refcount_blocks = cpu_to_be32(afake->af_blocks);
+	xfs_alloc_log_agf(tp, agbp, XFS_AGF_REFCOUNT_BLOCKS |
+				    XFS_AGF_REFCOUNT_ROOT |
+				    XFS_AGF_REFCOUNT_LEVEL);
+	xfs_btree_commit_afakeroot(cur, tp, agbp, &xfs_refcountbt_ops);
+}
+
 /*
  * Calculate the number of records in a refcount btree block.
  */
@@ -420,7 +470,7 @@ xfs_refcountbt_calc_reserves(
 	if (error)
 		return error;
 
-	agf = XFS_BUF_TO_AGF(agbp);
+	agf = agbp->b_addr;
 	agblocks = be32_to_cpu(agf->agf_length);
 	tree_len = be32_to_cpu(agf->agf_refcount_blocks);
 	xfs_trans_brelse(tp, agbp);