bcachefs: Fix safe errors by default

i.e. the start of automatic self healing:

If errors=continue or fix_safe, we now automatically fix simple errors
without user intervention.

New error action option: fix_safe

This replaces the existing errors=ro option, which gets a new slot, i.e.
existing errors=ro users now get errors=fix_safe.

This is currently only enabled for a limited set of errors - initially
just disk accounting; errors we would never not want to fix, and we
don't want to require user intervention (i.e. to make sure a bug report
gets filed).

Errors will still be counted in the superblock, so we (developers) will
still know they've been occuring if a bug report gets filed (as bug
reports typically include the errors superblock section).

Eventually we'll be enabling this for a much wider set of errors, after
we've done thorough error injection testing.

Signed-off-by: Kent Overstreet <[email protected]>

Author: Kent Overstreet

fs/bcachefs/bcachefs_format.h

@@ -990,8 +990,9 @@ enum bch_version_upgrade_opts {
 
 #define BCH_ERROR_ACTIONS()		\
 	x(continue,		0)	\
-	x(ro,			1)	\
-	x(panic,		2)
+	x(fix_safe,		1)	\
+	x(panic,		2)	\
+	x(ro,			3)
 
 enum bch_error_actions {
 #define x(t, n) BCH_ON_ERROR_##t = n,

fs/bcachefs/error.c

@@ -15,6 +15,7 @@ bool bch2_inconsistent_error(struct bch_fs *c)
 	switch (c->opts.errors) {
 	case BCH_ON_ERROR_continue:
 		return false;
+	case BCH_ON_ERROR_fix_safe:
 	case BCH_ON_ERROR_ro:
 		if (bch2_fs_emergency_read_only(c))
 			bch_err(c, "inconsistency detected - emergency read only at journal seq %llu",
@@ -191,6 +192,12 @@ static void prt_actioning(struct printbuf *out, const char *action)
 	prt_str(out, "ing");
 }
 
+static const u8 fsck_flags_extra[] = {
+#define x(t, n, flags)		[BCH_FSCK_ERR_##t] = flags,
+	BCH_SB_ERRS()
+#undef x
+};
+
 int bch2_fsck_err(struct bch_fs *c,
 		  enum bch_fsck_flags flags,
 		  enum bch_sb_error_id err,
@@ -203,6 +210,9 @@ int bch2_fsck_err(struct bch_fs *c,
 	int ret = -BCH_ERR_fsck_ignore;
 	const char *action_orig = "fix?", *action = action_orig;
 
+	if (!WARN_ON(err >= ARRAY_SIZE(fsck_flags_extra)))
+		flags |= fsck_flags_extra[err];
+
 	if ((flags & FSCK_CAN_FIX) &&
 	    test_bit(err, c->sb.errors_silent))
 		return -BCH_ERR_fsck_fix;
@@ -265,7 +275,14 @@ int bch2_fsck_err(struct bch_fs *c,
 		prt_printf(out, bch2_log_msg(c, ""));
 #endif
 
-	if (!test_bit(BCH_FS_fsck_running, &c->flags)) {
+	if ((flags & FSCK_CAN_FIX) &&
+	    (flags & FSCK_AUTOFIX) &&
+	    (c->opts.errors == BCH_ON_ERROR_continue ||
+	     c->opts.errors == BCH_ON_ERROR_fix_safe)) {
+		prt_str(out, ", ");
+		prt_actioning(out, action);
+		ret = -BCH_ERR_fsck_fix;
+	} else if (!test_bit(BCH_FS_fsck_running, &c->flags)) {
 		if (c->opts.errors != BCH_ON_ERROR_continue ||
 		    !(flags & (FSCK_CAN_FIX|FSCK_CAN_IGNORE))) {
 			prt_str(out, ", shutting down");

fs/bcachefs/error.h

@@ -108,13 +108,6 @@ struct fsck_err_state {
 	char			*last_msg;
 };
 
-enum bch_fsck_flags {
-	FSCK_CAN_FIX		= 1 << 0,
-	FSCK_CAN_IGNORE		= 1 << 1,
-	FSCK_NEED_FSCK		= 1 << 2,
-	FSCK_NO_RATELIMIT	= 1 << 3,
-};
-
 #define fsck_err_count(_c, _err)	bch2_sb_err_count(_c, BCH_FSCK_ERR_##_err)
 
 __printf(4, 5) __cold

fs/bcachefs/opts.h

@@ -137,7 +137,7 @@ enum fsck_err_opts {
 	x(errors,			u8,				\
 	  OPT_FS|OPT_FORMAT|OPT_MOUNT|OPT_RUNTIME,			\
 	  OPT_STR(bch2_error_actions),					\
-	  BCH_SB_ERROR_ACTION,		BCH_ON_ERROR_ro,		\
+	  BCH_SB_ERROR_ACTION,		BCH_ON_ERROR_fix_safe,		\
 	  NULL,		"Action to take on filesystem error")		\
 	x(metadata_replicas,		u8,				\
 	  OPT_FS|OPT_FORMAT|OPT_MOUNT|OPT_RUNTIME,			\