1 files changed, 142 insertions, 59 deletions

diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index cd9a94a6b5bb..6df65540e13d 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -1,10 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
  * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
  * Copyright (C) 2004-2006 Red Hat, Inc.  All rights reserved.
- *
- * This copyrighted material is made available to anyone wishing to use,
- * modify, copy, or redistribute it subject to the terms and conditions
- * of the GNU General Public License version 2.
  */
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
@@ -28,99 +25,144 @@
 #include "util.h"
 #include "trace_gfs2.h"
 
-int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
-		     unsigned int revokes)
+static void gfs2_print_trans(struct gfs2_sbd *sdp, const struct gfs2_trans *tr)
 {
-	struct gfs2_trans *tr;
-	int error;
+	fs_warn(sdp, "Transaction created at: %pSR\n", (void *)tr->tr_ip);
+	fs_warn(sdp, "blocks=%u revokes=%u reserved=%u touched=%u\n",
+		tr->tr_blocks, tr->tr_revokes, tr->tr_reserved,
+		test_bit(TR_TOUCHED, &tr->tr_flags));
+	fs_warn(sdp, "Buf %u/%u Databuf %u/%u Revoke %u\n",
+		tr->tr_num_buf_new, tr->tr_num_buf_rm,
+		tr->tr_num_databuf_new, tr->tr_num_databuf_rm,
+		tr->tr_num_revoke);
+}
+
+int __gfs2_trans_begin(struct gfs2_trans *tr, struct gfs2_sbd *sdp,
+		       unsigned int blocks, unsigned int revokes,
+		       unsigned long ip)
+{
+	unsigned int extra_revokes;
 
-	BUG_ON(current->journal_info);
+	if (current->journal_info) {
+		gfs2_print_trans(sdp, current->journal_info);
+		BUG();
+	}
 	BUG_ON(blocks == 0 && revokes == 0);
 
-	if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
+	if (gfs2_withdrawn(sdp))
 		return -EROFS;
 
-	tr = kzalloc(sizeof(struct gfs2_trans), GFP_NOFS);
-	if (!tr)
-		return -ENOMEM;
-
-	tr->tr_ip = _RET_IP_;
+	tr->tr_ip = ip;
 	tr->tr_blocks = blocks;
 	tr->tr_revokes = revokes;
-	tr->tr_reserved = 1;
-	set_bit(TR_ALLOCED, &tr->tr_flags);
-	if (blocks)
-		tr->tr_reserved += 6 + blocks;
-	if (revokes)
-		tr->tr_reserved += gfs2_struct2blk(sdp, revokes,
-						   sizeof(u64));
+	tr->tr_reserved = GFS2_LOG_FLUSH_MIN_BLOCKS;
+	if (blocks) {
+		/*
+		 * The reserved blocks are either used for data or metadata.
+		 * We can have mixed data and metadata, each with its own log
+		 * descriptor block; see calc_reserved().
+		 */
+		tr->tr_reserved += blocks + 1 + DIV_ROUND_UP(blocks - 1, databuf_limit(sdp));
+	}
 	INIT_LIST_HEAD(&tr->tr_databuf);
 	INIT_LIST_HEAD(&tr->tr_buf);
+	INIT_LIST_HEAD(&tr->tr_list);
+	INIT_LIST_HEAD(&tr->tr_ail1_list);
+	INIT_LIST_HEAD(&tr->tr_ail2_list);
+
+	if (gfs2_assert_warn(sdp, tr->tr_reserved <= sdp->sd_jdesc->jd_blocks))
+		return -EINVAL;
 
 	sb_start_intwrite(sdp->sd_vfs);
 
-	error = gfs2_log_reserve(sdp, tr->tr_reserved);
-	if (error)
-		goto fail;
+	/*
+	 * Try the reservations under sd_log_flush_lock to prevent log flushes
+	 * from creating inconsistencies between the number of allocated and
+	 * reserved revokes.  If that fails, do a full-block allocation outside
+	 * of the lock to avoid stalling log flushes.  Then, allot the
+	 * appropriate number of blocks to revokes, use as many revokes locally
+	 * as needed, and "release" the surplus into the revokes pool.
+	 */
+
+	down_read(&sdp->sd_log_flush_lock);
+	if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)))
+		goto out_not_live;
+	if (gfs2_log_try_reserve(sdp, tr, &extra_revokes))
+		goto reserved;
 
-	current->journal_info = tr;
+	up_read(&sdp->sd_log_flush_lock);
+	gfs2_log_reserve(sdp, tr, &extra_revokes);
+	down_read(&sdp->sd_log_flush_lock);
+	if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) {
+		revokes = tr->tr_revokes + extra_revokes;
+		gfs2_log_release_revokes(sdp, revokes);
+		gfs2_log_release(sdp, tr->tr_reserved);
+		goto out_not_live;
+	}
 
+reserved:
+	gfs2_log_release_revokes(sdp, extra_revokes);
+	current->journal_info = tr;
 	return 0;
 
-fail:
+out_not_live:
+	up_read(&sdp->sd_log_flush_lock);
 	sb_end_intwrite(sdp->sd_vfs);
-	kfree(tr);
-
-	return error;
+	return -EROFS;
 }
 
-static void gfs2_print_trans(struct gfs2_sbd *sdp, const struct gfs2_trans *tr)
+int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
+		     unsigned int revokes)
 {
-	fs_warn(sdp, "Transaction created at: %pSR\n", (void *)tr->tr_ip);
-	fs_warn(sdp, "blocks=%u revokes=%u reserved=%u touched=%u\n",
-		tr->tr_blocks, tr->tr_revokes, tr->tr_reserved,
-		test_bit(TR_TOUCHED, &tr->tr_flags));
-	fs_warn(sdp, "Buf %u/%u Databuf %u/%u Revoke %u/%u\n",
-		tr->tr_num_buf_new, tr->tr_num_buf_rm,
-		tr->tr_num_databuf_new, tr->tr_num_databuf_rm,
-		tr->tr_num_revoke, tr->tr_num_revoke_rm);
+	struct gfs2_trans *tr;
+	int error;
+
+	tr = kmem_cache_zalloc(gfs2_trans_cachep, GFP_NOFS);
+	if (!tr)
+		return -ENOMEM;
+	error = __gfs2_trans_begin(tr, sdp, blocks, revokes, _RET_IP_);
+	if (error)
+		kmem_cache_free(gfs2_trans_cachep, tr);
+	return error;
 }
 
 void gfs2_trans_end(struct gfs2_sbd *sdp)
 {
 	struct gfs2_trans *tr = current->journal_info;
 	s64 nbuf;
-	int alloced = test_bit(TR_ALLOCED, &tr->tr_flags);
 
 	current->journal_info = NULL;
 
 	if (!test_bit(TR_TOUCHED, &tr->tr_flags)) {
+		gfs2_log_release_revokes(sdp, tr->tr_revokes);
+		up_read(&sdp->sd_log_flush_lock);
 		gfs2_log_release(sdp, tr->tr_reserved);
-		if (alloced) {
-			kfree(tr);
-			sb_end_intwrite(sdp->sd_vfs);
-		}
+		if (!test_bit(TR_ONSTACK, &tr->tr_flags))
+			gfs2_trans_free(sdp, tr);
+		sb_end_intwrite(sdp->sd_vfs);
 		return;
 	}
 
+	gfs2_log_release_revokes(sdp, tr->tr_revokes - tr->tr_num_revoke);
+
 	nbuf = tr->tr_num_buf_new + tr->tr_num_databuf_new;
 	nbuf -= tr->tr_num_buf_rm;
 	nbuf -= tr->tr_num_databuf_rm;
 
-	if (gfs2_assert_withdraw(sdp, (nbuf <= tr->tr_blocks) &&
-				       (tr->tr_num_revoke <= tr->tr_revokes)))
+	if (gfs2_assert_withdraw(sdp, nbuf <= tr->tr_blocks) ||
+	    gfs2_assert_withdraw(sdp, tr->tr_num_revoke <= tr->tr_revokes))
 		gfs2_print_trans(sdp, tr);
 
 	gfs2_log_commit(sdp, tr);
-	if (alloced && !test_bit(TR_ATTACHED, &tr->tr_flags))
-		kfree(tr);
+	if (!test_bit(TR_ONSTACK, &tr->tr_flags) &&
+	    !test_bit(TR_ATTACHED, &tr->tr_flags))
+		gfs2_trans_free(sdp, tr);
 	up_read(&sdp->sd_log_flush_lock);
 
 	if (sdp->sd_vfs->s_flags & SB_SYNCHRONOUS)
 		gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
 			       GFS2_LFC_TRANS_END);
-	if (alloced)
-		sb_end_intwrite(sdp->sd_vfs);
+	sb_end_intwrite(sdp->sd_vfs);
 }
 
 static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl,
@@ -132,6 +174,8 @@ static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl,
 	bd->bd_bh = bh;
 	bd->bd_gl = gl;
 	INIT_LIST_HEAD(&bd->bd_list);
+	INIT_LIST_HEAD(&bd->bd_ail_st_list);
+	INIT_LIST_HEAD(&bd->bd_ail_gl_list);
 	bh->b_private = bd;
 	return bd;
 }
@@ -187,14 +231,35 @@ out:
 	unlock_buffer(bh);
 }
 
+void gfs2_trans_add_databufs(struct gfs2_glock *gl, struct folio *folio,
+			     size_t from, size_t len)
+{
+	struct buffer_head *head = folio_buffers(folio);
+	unsigned int bsize = head->b_size;
+	struct buffer_head *bh;
+	size_t to = from + len;
+	size_t start, end;
+
+	for (bh = head, start = 0; bh != head || !start;
+	     bh = bh->b_this_page, start = end) {
+		end = start + bsize;
+		if (end <= from)
+			continue;
+		if (start >= to)
+			break;
+		set_buffer_uptodate(bh);
+		gfs2_trans_add_data(gl, bh);
+	}
+}
+
 void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 {
 
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
+	struct super_block *sb = sdp->sd_vfs;
 	struct gfs2_bufdata *bd;
 	struct gfs2_meta_header *mh;
 	struct gfs2_trans *tr = current->journal_info;
-	enum gfs2_freeze_state state = atomic_read(&sdp->sd_freeze_state);
 
 	lock_buffer(bh);
 	if (buffer_pinned(bh)) {
@@ -206,12 +271,12 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 	if (bd == NULL) {
 		gfs2_log_unlock(sdp);
 		unlock_buffer(bh);
-		lock_page(bh->b_page);
+		folio_lock(bh->b_folio);
 		if (bh->b_private == NULL)
 			bd = gfs2_alloc_bufdata(gl, bh);
 		else
 			bd = bh->b_private;
-		unlock_page(bh->b_page);
+		folio_unlock(bh->b_folio);
 		lock_buffer(bh);
 		gfs2_log_lock(sdp);
 	}
@@ -228,9 +293,15 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 		       (unsigned long long)bd->bd_bh->b_blocknr);
 		BUG();
 	}
-	if (unlikely(state == SFS_FROZEN)) {
+	if (gfs2_withdrawn(sdp)) {
+		fs_info(sdp, "GFS2:adding buf while withdrawn! 0x%llx\n",
+			(unsigned long long)bd->bd_bh->b_blocknr);
+		goto out_unlock;
+	}
+	if (unlikely(sb->s_writers.frozen == SB_FREEZE_COMPLETE)) {
 		fs_info(sdp, "GFS2:adding buf while frozen\n");
-		gfs2_assert_withdraw(sdp, 0);
+		gfs2_withdraw(sdp);
+		goto out_unlock;
 	}
 	gfs2_pin(sdp, bd->bd_bh);
 	mh->__pad0 = cpu_to_be64(0);
@@ -253,20 +324,21 @@ void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd)
 	tr->tr_num_revoke++;
 }
 
-void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
+void gfs2_trans_remove_revoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
 {
 	struct gfs2_bufdata *bd, *tmp;
-	struct gfs2_trans *tr = current->journal_info;
 	unsigned int n = len;
 
 	gfs2_log_lock(sdp);
-	list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_list) {
+	list_for_each_entry_safe(bd, tmp, &sdp->sd_log_revokes, bd_list) {
 		if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) {
 			list_del_init(&bd->bd_list);
 			gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
 			sdp->sd_log_num_revoke--;
+			if (bd->bd_gl)
+				gfs2_glock_remove_revoke(bd->bd_gl);
 			kmem_cache_free(gfs2_bufdata_cachep, bd);
-			tr->tr_num_revoke_rm++;
+			gfs2_log_release_revokes(sdp, 1);
 			if (--n == 0)
 				break;
 		}
@@ -274,3 +346,14 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
 	gfs2_log_unlock(sdp);
 }
 
+void gfs2_trans_free(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
+{
+	if (tr == NULL)
+		return;
+
+	gfs2_assert_warn(sdp, list_empty(&tr->tr_ail1_list));
+	gfs2_assert_warn(sdp, list_empty(&tr->tr_ail2_list));
+	gfs2_assert_warn(sdp, list_empty(&tr->tr_databuf));
+	gfs2_assert_warn(sdp, list_empty(&tr->tr_buf));
+	kmem_cache_free(gfs2_trans_cachep, tr);
+}