btrfs: store the error that turned the fs into error state

Currently when we turn the fs into an error state, typically after a
transaction abort, we don't store the error anywhere, we just set a bit
(BTRFS_FS_STATE_ERROR) at struct btrfs_fs_info::fs_state to signal the
error state.

There are cases where it would be useful to have access to the specific
error in order to provide a more meaningful error to users/applications.
This change adds a member to struct btrfs_fs_info to store the error and
removes the BTRFS_FS_STATE_ERROR bit. When there's no error, the new
member (fs_error) has a value of 0, otherwise its value is a negative
errno value.

Followup changes will make use of this new member.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>

fs/btrfs/disk-io.c

@@ -3222,7 +3222,7 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 
 	/* check FS state, whether FS is broken. */
 	if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_ERROR)
-		set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state);
+		WRITE_ONCE(fs_info->fs_error, -EUCLEAN);
 
 	/*
 	 * In the long term, we'll store the compression type in the super

fs/btrfs/fs.h

@@ -46,8 +46,6 @@ static_assert(sizeof(struct btrfs_super_block) == BTRFS_SUPER_INFO_SIZE);
  * Runtime (in-memory) states of filesystem
  */
 enum {
-	/* Global indicator of serious filesystem errors */
-	BTRFS_FS_STATE_ERROR,
 	/*
 	 * Filesystem is being remounted, allow to skip some operations, like
 	 * defrag
@@ -686,6 +684,12 @@ struct btrfs_fs_info {
 	bool qgroup_rescan_running;
 	u8 qgroup_drop_subtree_thres;
 
+	/*
+	 * If this is not 0, then it indicates a serious filesystem error has
+	 * happened and it contains that error (negative errno value).
+	 */
+	int fs_error;
+
 	/* Filesystem state */
 	unsigned long fs_state;
 
@@ -962,8 +966,8 @@ static inline void btrfs_wake_unfinished_drop(struct btrfs_fs_info *fs_info)
 	clear_and_wake_up_bit(BTRFS_FS_UNFINISHED_DROPS, &fs_info->flags);
 }
 
-#define BTRFS_FS_ERROR(fs_info)	(unlikely(test_bit(BTRFS_FS_STATE_ERROR, \
-						   &(fs_info)->fs_state)))
+#define BTRFS_FS_ERROR(fs_info)	(READ_ONCE((fs_info)->fs_error))
+
 #define BTRFS_FS_LOG_CLEANUP_ERROR(fs_info)				\
 	(unlikely(test_bit(BTRFS_FS_STATE_LOG_CLEANUP_ERROR,		\
 			   &(fs_info)->fs_state)))

fs/btrfs/messages.c

@@ -10,14 +10,13 @@
 #ifdef CONFIG_PRINTK
 
 #define STATE_STRING_PREFACE	": state "
-#define STATE_STRING_BUF_LEN	(sizeof(STATE_STRING_PREFACE) + BTRFS_FS_STATE_COUNT)
+#define STATE_STRING_BUF_LEN	(sizeof(STATE_STRING_PREFACE) + BTRFS_FS_STATE_COUNT + 1)
 
 /*
  * Characters to print to indicate error conditions or uncommon filesystem state.
  * RO is not an error.
  */
 static const char fs_state_chars[] = {
-	[BTRFS_FS_STATE_ERROR]			= 'E',
 	[BTRFS_FS_STATE_REMOUNTING]		= 'M',
 	[BTRFS_FS_STATE_RO]			= 0,
 	[BTRFS_FS_STATE_TRANS_ABORTED]		= 'A',
@@ -37,6 +36,11 @@ static void btrfs_state_to_string(const struct btrfs_fs_info *info, char *buf)
 	memcpy(curr, STATE_STRING_PREFACE, sizeof(STATE_STRING_PREFACE));
 	curr += sizeof(STATE_STRING_PREFACE) - 1;
 
+	if (BTRFS_FS_ERROR(info)) {
+		*curr++ = 'E';
+		states_printed = true;
+	}
+
 	for_each_set_bit(bit, &fs_state, sizeof(fs_state)) {
 		WARN_ON_ONCE(bit >= BTRFS_FS_STATE_COUNT);
 		if ((bit < BTRFS_FS_STATE_COUNT) && fs_state_chars[bit]) {
@@ -155,7 +159,7 @@ void __btrfs_handle_fs_error(struct btrfs_fs_info *fs_info, const char *function
 	 * Today we only save the error info to memory.  Long term we'll also
 	 * send it down to the disk.
 	 */
-	set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state);
+	WRITE_ONCE(fs_info->fs_error, errno);
 
 	/* Don't go through full error handling during mount. */
 	if (!(sb->s_flags & SB_BORN))