Merge tag 'gfs2-v5.17-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2

Pull gfs2 fixes from Andreas Gruenbacher:

 - To avoid deadlocks, actively cancel dlm locking requests when we give
   up on them.

   Further dlm operations on the same lock will return -EBUSY until the
   cancel has been completed, so in that case, wait and repeat. (This is
   rare.)

 - Lock inversion fixes in gfs2_inode_lookup() and gfs2_create_inode().

 - Some more fallout from the gfs2 mmap + page fault deadlock fixes
   (merged in commit c03098d4b9ad7: "Merge tag 'gfs2-v5.15-rc5-mmap-fault'").

 - Various other minor bug fixes and cleanups.

* tag 'gfs2-v5.17-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
  gfs2: Make sure FITRIM minlen is rounded up to fs block size
  gfs2: Make sure not to return short direct writes
  gfs2: Remove dead code in gfs2_file_read_iter
  gfs2: Fix gfs2_file_buffered_write endless loop workaround
  gfs2: Minor retry logic cleanup
  gfs2: Disable page faults during lockless buffered reads
  gfs2: Fix should_fault_in_pages() logic
  gfs2: Remove return value for gfs2_indirect_init
  gfs2: Initialize gh_error in gfs2_glock_nq
  gfs2: Make use of list_is_first
  gfs2: Switch lock order of inode and iopen glock
  gfs2: cancel timed-out glock requests
  gfs2: Expect -EBUSY after canceling dlm locking requests
  gfs2: gfs2_setattr_size error path fix
  gfs2: assign rgrp glock before compute_bitstructs

8 files changed, 104 insertions, 76 deletions

diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index d67108489148..39080b2d6cf8 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -606,9 +606,9 @@ out:
 	return ret;
 }
 
-static inline __be64 *gfs2_indirect_init(struct metapath *mp,
-					 struct gfs2_glock *gl, unsigned int i,
-					 unsigned offset, u64 bn)
+static inline void gfs2_indirect_init(struct metapath *mp,
+				      struct gfs2_glock *gl, unsigned int i,
+				      unsigned offset, u64 bn)
 {
 	__be64 *ptr = (__be64 *)(mp->mp_bh[i - 1]->b_data +
 		       ((i > 1) ? sizeof(struct gfs2_meta_header) :
@@ -621,7 +621,6 @@ static inline __be64 *gfs2_indirect_init(struct metapath *mp,
 	gfs2_buffer_clear_tail(mp->mp_bh[i], sizeof(struct gfs2_meta_header));
 	ptr += offset;
 	*ptr = cpu_to_be64(bn);
-	return ptr;
 }
 
 enum alloc_state {
@@ -2146,7 +2145,7 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize)
 
 	ret = do_shrink(inode, newsize);
 out:
-	gfs2_rs_delete(ip, NULL);
+	gfs2_rs_delete(ip);
 	gfs2_qa_put(ip);
 	return ret;
 }
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 8c39a8571b1f..22b41acfbbc3 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -706,7 +706,7 @@ static int gfs2_release(struct inode *inode, struct file *file)
 
 	if (file->f_mode & FMODE_WRITE) {
 		if (gfs2_rs_active(&ip->i_res))
-			gfs2_rs_delete(ip, &inode->i_writecount);
+			gfs2_rs_delete(ip);
 		gfs2_qa_put(ip);
 	}
 	return 0;
@@ -775,8 +775,7 @@ static inline bool should_fault_in_pages(ssize_t ret, struct iov_iter *i,
 					 size_t *window_size)
 {
 	size_t count = iov_iter_count(i);
-	char __user *p;
-	int pages = 1;
+	size_t size, offs;
 
 	if (likely(!count))
 		return false;
@@ -785,18 +784,20 @@ static inline bool should_fault_in_pages(ssize_t ret, struct iov_iter *i,
 	if (!iter_is_iovec(i))
 		return false;
 
+	size = PAGE_SIZE;
+	offs = offset_in_page(i->iov[0].iov_base + i->iov_offset);
 	if (*prev_count != count || !*window_size) {
-		int pages, nr_dirtied;
+		size_t nr_dirtied;
 
-		pages = min_t(int, BIO_MAX_VECS, DIV_ROUND_UP(count, PAGE_SIZE));
+		size = ALIGN(offs + count, PAGE_SIZE);
+		size = min_t(size_t, size, SZ_1M);
 		nr_dirtied = max(current->nr_dirtied_pause -
-				 current->nr_dirtied, 1);
-		pages = min(pages, nr_dirtied);
+				 current->nr_dirtied, 8);
+		size = min(size, nr_dirtied << PAGE_SHIFT);
 	}
 
 	*prev_count = count;
-	p = i->iov[0].iov_base + i->iov_offset;
-	*window_size = (size_t)PAGE_SIZE * pages - offset_in_page(p);
+	*window_size = size - offs;
 	return true;
 }
 
@@ -851,9 +852,9 @@ retry_under_glock:
 		leftover = fault_in_iov_iter_writeable(to, window_size);
 		gfs2_holder_disallow_demote(gh);
 		if (leftover != window_size) {
-			if (!gfs2_holder_queued(gh))
-				goto retry;
-			goto retry_under_glock;
+			if (gfs2_holder_queued(gh))
+				goto retry_under_glock;
+			goto retry;
 		}
 	}
 	if (gfs2_holder_queued(gh))
@@ -920,9 +921,9 @@ retry_under_glock:
 		leftover = fault_in_iov_iter_readable(from, window_size);
 		gfs2_holder_disallow_demote(gh);
 		if (leftover != window_size) {
-			if (!gfs2_holder_queued(gh))
-				goto retry;
-			goto retry_under_glock;
+			if (gfs2_holder_queued(gh))
+				goto retry_under_glock;
+			goto retry;
 		}
 	}
 out:
@@ -950,20 +951,19 @@ static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 	 * and retry.
 	 */
 
-	if (iocb->ki_flags & IOCB_DIRECT) {
-		ret = gfs2_file_direct_read(iocb, to, &gh);
-		if (likely(ret != -ENOTBLK))
-			return ret;
-		iocb->ki_flags &= ~IOCB_DIRECT;
-	}
+	if (iocb->ki_flags & IOCB_DIRECT)
+		return gfs2_file_direct_read(iocb, to, &gh);
+
+	pagefault_disable();
 	iocb->ki_flags |= IOCB_NOIO;
 	ret = generic_file_read_iter(iocb, to);
 	iocb->ki_flags &= ~IOCB_NOIO;
+	pagefault_enable();
 	if (ret >= 0) {
 		if (!iov_iter_count(to))
 			return ret;
 		written = ret;
-	} else {
+	} else if (ret != -EFAULT) {
 		if (ret != -EAGAIN)
 			return ret;
 		if (iocb->ki_flags & IOCB_NOWAIT)
@@ -989,12 +989,11 @@ retry_under_glock:
 		leftover = fault_in_iov_iter_writeable(to, window_size);
 		gfs2_holder_disallow_demote(&gh);
 		if (leftover != window_size) {
-			if (!gfs2_holder_queued(&gh)) {
-				if (written)
-					goto out_uninit;
-				goto retry;
-			}
-			goto retry_under_glock;
+			if (gfs2_holder_queued(&gh))
+				goto retry_under_glock;
+			if (written)
+				goto out_uninit;
+			goto retry;
 		}
 	}
 	if (gfs2_holder_queued(&gh))
@@ -1068,12 +1067,11 @@ retry_under_glock:
 		gfs2_holder_disallow_demote(gh);
 		if (leftover != window_size) {
 			from->count = min(from->count, window_size - leftover);
-			if (!gfs2_holder_queued(gh)) {
-				if (read)
-					goto out_uninit;
-				goto retry;
-			}
-			goto retry_under_glock;
+			if (gfs2_holder_queued(gh))
+				goto retry_under_glock;
+			if (read && !(iocb->ki_flags & IOCB_DIRECT))
+				goto out_uninit;
+			goto retry;
 		}
 	}
 out_unlock:
@@ -1083,6 +1081,7 @@ out_uninit:
 	gfs2_holder_uninit(gh);
 	if (statfs_gh)
 		kfree(statfs_gh);
+	from->count = orig_count - read;
 	return read ? read : ret;
 }
 
@@ -1497,7 +1496,6 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 		if (error != GLR_TRYFAILED)
 			break;
 		fl_gh->gh_flags = LM_FLAG_TRY | GL_EXACT;
-		fl_gh->gh_error = 0;
 		msleep(sleeptime);
 	}
 	if (error) {
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 6b23399eaee0..630c6550eacf 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -542,7 +542,7 @@ restart:
 			 * some reason. If this holder is the head of the list, it
 			 * means we have a blocked holder at the head, so return 1.
 			 */
-			if (gh->gh_list.prev == &gl->gl_holders)
+			if (list_is_first(&gh->gh_list, &gl->gl_holders))
 				return 1;
 			do_error(gl, 0);
 			break;
@@ -669,6 +669,8 @@ static void finish_xmote(struct gfs2_glock *gl, unsigned int ret)
 
 	/* Check for state != intended state */
 	if (unlikely(state != gl->gl_target)) {
+		if (gh && (ret & LM_OUT_CANCELED))
+			gfs2_holder_wake(gh);
 		if (gh && !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)) {
 			/* move to back of queue and try next entry */
 			if (ret & LM_OUT_CANCELED) {
@@ -1259,7 +1261,6 @@ void __gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, u16 flags,
 	gh->gh_owner_pid = get_pid(task_pid(current));
 	gh->gh_state = state;
 	gh->gh_flags = flags;
-	gh->gh_error = 0;
 	gh->gh_iflags = 0;
 	gfs2_glock_hold(gl);
 }
@@ -1565,6 +1566,7 @@ int gfs2_glock_nq(struct gfs2_holder *gh)
 	if (test_bit(GLF_LRU, &gl->gl_flags))
 		gfs2_glock_remove_from_lru(gl);
 
+	gh->gh_error = 0;
 	spin_lock(&gl->gl_lockref.lock);
 	add_to_queue(gh);
 	if (unlikely((LM_FLAG_NOEXP & gh->gh_flags) &&
@@ -1691,6 +1693,14 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
 	struct gfs2_glock *gl = gh->gh_gl;
 
 	spin_lock(&gl->gl_lockref.lock);
+	if (list_is_first(&gh->gh_list, &gl->gl_holders) &&
+	    !test_bit(HIF_HOLDER, &gh->gh_iflags)) {
+		spin_unlock(&gl->gl_lockref.lock);
+		gl->gl_name.ln_sbd->sd_lockstruct.ls_ops->lm_cancel(gl);
+		wait_on_bit(&gh->gh_iflags, HIF_WAIT, TASK_UNINTERRUPTIBLE);
+		spin_lock(&gl->gl_lockref.lock);
+	}
+
 	__gfs2_glock_dq(gh);
 	spin_unlock(&gl->gl_lockref.lock);
 }
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 89905f4f29bb..c8ec876f33ea 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -131,7 +131,21 @@ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type,
 		struct gfs2_sbd *sdp = GFS2_SB(inode);
 		struct gfs2_glock *io_gl;
 
-		error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, &ip->i_gl);
+		error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE,
+				       &ip->i_gl);
+		if (unlikely(error))
+			goto fail;
+
+		error = gfs2_glock_get(sdp, no_addr, &gfs2_iopen_glops, CREATE,
+				       &io_gl);
+		if (unlikely(error))
+			goto fail;
+
+		if (blktype != GFS2_BLKST_UNLINKED)
+			gfs2_cancel_delete_work(io_gl);
+		error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT,
+					   &ip->i_iopen_gh);
+		gfs2_glock_put(io_gl);
 		if (unlikely(error))
 			goto fail;
 
@@ -161,16 +175,6 @@ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type,
 
 		set_bit(GLF_INSTANTIATE_NEEDED, &ip->i_gl->gl_flags);
 
-		error = gfs2_glock_get(sdp, no_addr, &gfs2_iopen_glops, CREATE, &io_gl);
-		if (unlikely(error))
-			goto fail;
-		if (blktype != GFS2_BLKST_UNLINKED)
-			gfs2_cancel_delete_work(io_gl);
-		error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh);
-		gfs2_glock_put(io_gl);
-		if (unlikely(error))
-			goto fail;
-
 		/* Lowest possible timestamp; will be overwritten in gfs2_dinode_in. */
 		inode->i_atime.tv_sec = 1LL << (8 * sizeof(inode->i_atime.tv_sec) - 1);
 		inode->i_atime.tv_nsec = 0;
@@ -716,13 +720,17 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
 	error = insert_inode_locked4(inode, ip->i_no_addr, iget_test, &ip->i_no_addr);
 	BUG_ON(error);
 
-	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1);
+	error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh);
 	if (error)
 		goto fail_gunlock2;
 
+	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1);
+	if (error)
+		goto fail_gunlock3;
+
 	error = gfs2_trans_begin(sdp, blocks, 0);
 	if (error)
-		goto fail_gunlock2;
+		goto fail_gunlock3;
 
 	if (blocks > 1) {
 		ip->i_eattr = ip->i_no_addr + 1;
@@ -731,10 +739,6 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
 	init_dinode(dip, ip, symname);
 	gfs2_trans_end(sdp);
 
-	error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh);
-	if (error)
-		goto fail_gunlock2;
-
 	glock_set_object(ip->i_gl, ip);
 	glock_set_object(io_gl, ip);
 	gfs2_set_iop(inode);
@@ -745,14 +749,14 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
 	if (default_acl) {
 		error = __gfs2_set_acl(inode, default_acl, ACL_TYPE_DEFAULT);
 		if (error)
-			goto fail_gunlock3;
+			goto fail_gunlock4;
 		posix_acl_release(default_acl);
 		default_acl = NULL;
 	}
 	if (acl) {
 		error = __gfs2_set_acl(inode, acl, ACL_TYPE_ACCESS);
 		if (error)
-			goto fail_gunlock3;
+			goto fail_gunlock4;
 		posix_acl_release(acl);
 		acl = NULL;
 	}
@@ -760,11 +764,11 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
 	error = security_inode_init_security(&ip->i_inode, &dip->i_inode, name,
 					     &gfs2_initxattrs, NULL);
 	if (error)
-		goto fail_gunlock3;
+		goto fail_gunlock4;
 
 	error = link_dinode(dip, name, ip, &da);
 	if (error)
-		goto fail_gunlock3;
+		goto fail_gunlock4;
 
 	mark_inode_dirty(inode);
 	d_instantiate(dentry, inode);
@@ -782,9 +786,10 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry,
 	unlock_new_inode(inode);
 	return error;
 
-fail_gunlock3:
+fail_gunlock4:
 	glock_clear_object(ip->i_gl, ip);
 	glock_clear_object(io_gl, ip);
+fail_gunlock3:
 	gfs2_glock_dq_uninit(&ip->i_iopen_gh);
 fail_gunlock2:
 	gfs2_glock_put(io_gl);
@@ -793,7 +798,7 @@ fail_free_inode:
 		if (free_vfs_inode) /* else evict will do the put for us */
 			gfs2_glock_put(ip->i_gl);
 	}
-	gfs2_rs_delete(ip, NULL);
+	gfs2_rs_deltree(&ip->i_res);
 	gfs2_qa_put(ip);
 fail_free_acls:
 	posix_acl_release(default_acl);
diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c
index 50578f881e6d..2559a79cf14b 100644
--- a/fs/gfs2/lock_dlm.c
+++ b/fs/gfs2/lock_dlm.c
@@ -261,6 +261,7 @@ static int gdlm_lock(struct gfs2_glock *gl, unsigned int req_state,
 	int req;
 	u32 lkf;
 	char strname[GDLM_STRNAME_BYTES] = "";
+	int error;
 
 	req = make_mode(gl->gl_name.ln_sbd, req_state);
 	lkf = make_flags(gl, flags, req);
@@ -279,8 +280,14 @@ static int gdlm_lock(struct gfs2_glock *gl, unsigned int req_state,
 	 * Submit the actual lock request.
 	 */
 
-	return dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, strname,
+again:
+	error = dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, strname,
 			GDLM_STRNAME_BYTES - 1, 0, gdlm_ast, gl, gdlm_bast);
+	if (error == -EBUSY) {
+		msleep(20);
+		goto again;
+	}
+	return error;
 }
 
 static void gdlm_put_lock(struct gfs2_glock *gl)
@@ -312,8 +319,14 @@ static void gdlm_put_lock(struct gfs2_glock *gl)
 		return;
 	}
 
+again:
 	error = dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_VALBLK,
 			   NULL, gl);
+	if (error == -EBUSY) {
+		msleep(20);
+		goto again;
+	}
+
 	if (error) {
 		fs_err(sdp, "gdlm_unlock %x,%llx err=%d\n",
 		       gl->gl_name.ln_type,
diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index 0fb3c01bc557..801ad9f4f2be 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -680,13 +680,14 @@ void gfs2_rs_deltree(struct gfs2_blkreserv *rs)
 /**
  * gfs2_rs_delete - delete a multi-block reservation
  * @ip: The inode for this reservation
- * @wcount: The inode's write count, or NULL
  *
  */
-void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount)
+void gfs2_rs_delete(struct gfs2_inode *ip)
 {
+	struct inode *inode = &ip->i_inode;
+
 	down_write(&ip->i_rw_mutex);
-	if ((wcount == NULL) || (atomic_read(wcount) <= 1))
+	if (atomic_read(&inode->i_writecount) <= 1)
 		gfs2_rs_deltree(&ip->i_res);
 	up_write(&ip->i_rw_mutex);
 }
@@ -922,15 +923,15 @@ static int read_rindex_entry(struct gfs2_inode *ip)
 	spin_lock_init(&rgd->rd_rsspin);
 	mutex_init(&rgd->rd_mutex);
 
-	error = compute_bitstructs(rgd);
-	if (error)
-		goto fail;
-
 	error = gfs2_glock_get(sdp, rgd->rd_addr,
 			       &gfs2_rgrp_glops, CREATE, &rgd->rd_gl);
 	if (error)
 		goto fail;
 
+	error = compute_bitstructs(rgd);
+	if (error)
+		goto fail_glock;
+
 	rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lksb.sb_lvbptr;
 	rgd->rd_flags &= ~GFS2_RDF_PREFERRED;
 	if (rgd->rd_data > sdp->sd_max_rg_data)
@@ -944,6 +945,7 @@ static int read_rindex_entry(struct gfs2_inode *ip)
 	}
 
 	error = 0; /* someone else read in the rgrp; free it and ignore it */
+fail_glock:
 	gfs2_glock_put(rgd->rd_gl);
 
 fail:
@@ -1415,7 +1417,8 @@ int gfs2_fitrim(struct file *filp, void __user *argp)
 
 	start = r.start >> bs_shift;
 	end = start + (r.len >> bs_shift);
-	minlen = max_t(u64, r.minlen,
+	minlen = max_t(u64, r.minlen, sdp->sd_sb.sb_bsize);
+	minlen = max_t(u64, minlen,
 		       q->limits.discard_granularity) >> bs_shift;
 
 	if (end <= start || minlen > sdp->sd_max_rg_data)
diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h
index 3e2ca1fb4305..46dd94e9e085 100644
--- a/fs/gfs2/rgrp.h
+++ b/fs/gfs2/rgrp.h
@@ -45,7 +45,7 @@ extern int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *n,
 			     bool dinode, u64 *generation);
 
 extern void gfs2_rs_deltree(struct gfs2_blkreserv *rs);
-extern void gfs2_rs_delete(struct gfs2_inode *ip, atomic_t *wcount);
+extern void gfs2_rs_delete(struct gfs2_inode *ip);
 extern void __gfs2_free_blocks(struct gfs2_inode *ip, struct gfs2_rgrpd *rgd,
 			       u64 bstart, u32 blen, int meta);
 extern void gfs2_free_meta(struct gfs2_inode *ip, struct gfs2_rgrpd *rgd,
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index cf9cf66522b3..bdb773e5c88f 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -1396,7 +1396,7 @@ out:
 	truncate_inode_pages_final(&inode->i_data);
 	if (ip->i_qadata)
 		gfs2_assert_warn(sdp, ip->i_qadata->qa_ref == 0);
-	gfs2_rs_delete(ip, NULL);
+	gfs2_rs_deltree(&ip->i_res);
 	gfs2_ordered_del_inode(ip);
 	clear_inode(inode);
 	gfs2_dir_hash_inval(ip);