[Cluster-devel] [GFS2 PATCH 00/25] GFS2 recovery patches v9

[Cluster-devel] [GFS2 PATCH 00/25] GFS2 recovery patches v9

[Bob Peterson]

Hi,

As you may recall, Andreas and I have been working to clean up the extensive
recovery patch set. Here, then, is the most recent version. Note that some
patches were removed, and others were restructered a bit.
I hope to push this to for-next and merge it in soon.

Many thanks to Andreas Gruenbacher for his suggestions and clean-ups.

Regards,

Bob Peterson

Andreas Gruenbacher (5):
  gfs2: Split gfs2_lm_withdraw into two functions
  gfs2: Report errors before withdraw
  gfs2: Remove usused cluster_wide arguments of gfs2_consist functions
  gfs2: Turn gfs2_consist into void functions
  gfs2: Return bool from gfs2_assert functions

Bob Peterson (20):
  gfs2: Introduce concept of a pending withdraw
  gfs2: clear ail1 list when gfs2 withdraws
  gfs2: Rework how rgrp buffer_heads are managed
  gfs2: log error reform
  gfs2: Only complain the first time an io error occurs in quota or log
  gfs2: Ignore dlm recovery requests if gfs2 is withdrawn
  gfs2: move check_journal_clean to util.c for future use
  gfs2: Allow some glocks to be used during withdraw
  gfs2: Make secondary withdrawers wait for first withdrawer
  gfs2: fix infinite loop when checking ail item count before go_inval
  gfs2: Add verbose option to check_journal_clean
  gfs2: Issue revokes more intelligently
  gfs2: Prepare to withdraw as soon as an IO error occurs in log write
  gfs2: Check for log write errors before telling dlm to unlock
  gfs2: Do log_flush in gfs2_ail_empty_gl even if ail list is empty
  gfs2: Withdraw in gfs2_ail1_flush if write_cache_pages fails
  gfs2: drain the ail2 list after io errors
  gfs2: Don't demote a glock until its revokes are written
  gfs2: Do proper error checking for go_sync family of glops functions
  gfs2: flesh out delayed withdraw for gfs2_log_flush

 fs/gfs2/glock.c      | 125 +++++++++++---
 fs/gfs2/glops.c      | 157 ++++++++++++++---
 fs/gfs2/incore.h     |  25 ++-
 fs/gfs2/lock_dlm.c   |  52 ++++++
 fs/gfs2/log.c        | 238 +++++++++++++++++---------
 fs/gfs2/log.h        |   1 +
 fs/gfs2/lops.c       |   8 +-
 fs/gfs2/meta_io.c    |   2 +-
 fs/gfs2/ops_fstype.c |  55 +-----
 fs/gfs2/quota.c      |   7 +-
 fs/gfs2/recovery.c   |   5 +
 fs/gfs2/rgrp.c       |  71 +++-----
 fs/gfs2/rgrp.h       |   1 -
 fs/gfs2/super.c      |  73 +++++---
 fs/gfs2/sys.c        |   3 +-
 fs/gfs2/trans.c      |   4 +
 fs/gfs2/util.c       | 396 +++++++++++++++++++++++++++++++++----------
 fs/gfs2/util.h       |  76 ++++++---
 18 files changed, 934 insertions(+), 365 deletions(-)

-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 01/25] gfs2: Split gfs2_lm_withdraw into two functions

[Bob Peterson]

From: Andreas Gruenbacher <agruenba@redhat.com>

Split gfs2_lm_withdraw into a function that prints an error message and a
function that withdraws the filesystem.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/glops.c |   3 +-
 fs/gfs2/log.c   |  19 ++++----
 fs/gfs2/sys.c   |   3 +-
 fs/gfs2/util.c  | 124 +++++++++++++++++++++++++-----------------------
 fs/gfs2/util.h  |   3 +-
 5 files changed, 82 insertions(+), 70 deletions(-)

diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 061d22e1ceb6..58431f67665e 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -39,7 +39,8 @@ static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
 	fs_err(gl->gl_name.ln_sbd, "AIL glock %u:%llu mapping %p\n",
 	       gl->gl_name.ln_type, gl->gl_name.ln_number,
 	       gfs2_glock2aspace(gl));
-	gfs2_lm_withdraw(gl->gl_name.ln_sbd, "AIL error\n");
+	gfs2_lm(gl->gl_name.ln_sbd, "AIL error\n");
+	gfs2_withdraw(gl->gl_name.ln_sbd);
 }
 
 /**
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 00a2e721a374..c4c7c013f7a7 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -165,7 +165,7 @@ void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc)
 	spin_unlock(&sdp->sd_ail_lock);
 	blk_finish_plug(&plug);
 	if (withdraw)
-		gfs2_lm_withdraw(sdp, NULL);
+		gfs2_withdraw(sdp);
 	trace_gfs2_ail_flush(sdp, wbc, 0);
 }
 
@@ -239,8 +239,10 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp)
 	ret = list_empty(&sdp->sd_ail1_list);
 	spin_unlock(&sdp->sd_ail_lock);
 
-	if (withdraw)
-		gfs2_lm_withdraw(sdp, "fatal: I/O error(s)\n");
+	if (withdraw) {
+		gfs2_lm(sdp, "fatal: I/O error(s)\n");
+		gfs2_withdraw(sdp);
+	}
 
 	return ret;
 }
@@ -1016,11 +1018,12 @@ int gfs2_logd(void *data)
 
 		/* Check for errors writing to the journal */
 		if (sdp->sd_log_error) {
-			gfs2_lm_withdraw(sdp,
-					 "GFS2: fsid=%s: error %d: "
-					 "withdrawing the file system to "
-					 "prevent further damage.\n",
-					 sdp->sd_fsname, sdp->sd_log_error);
+			gfs2_lm(sdp,
+				"GFS2: fsid=%s: error %d: "
+				"withdrawing the file system to "
+				"prevent further damage.\n",
+				sdp->sd_fsname, sdp->sd_log_error);
+			gfs2_withdraw(sdp);
 		}
 
 		did_flush = false;
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index 8ccb68f4ed16..a2eae5c578d6 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -136,7 +136,8 @@ static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
 	if (val != 1)
 		return -EINVAL;
 
-	gfs2_lm_withdraw(sdp, "withdrawing from cluster at user's request\n");
+	gfs2_lm(sdp, "withdrawing from cluster at user's request\n");
+	gfs2_withdraw(sdp);
 
 	return len;
 }
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index ec600b487498..322012e2064e 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -33,28 +33,31 @@ void gfs2_assert_i(struct gfs2_sbd *sdp)
 	fs_emerg(sdp, "fatal assertion failed\n");
 }
 
-int gfs2_lm_withdraw(struct gfs2_sbd *sdp, const char *fmt, ...)
+void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...)
+{
+	struct va_format vaf;
+	va_list args;
+
+	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW &&
+	    test_bit(SDF_WITHDRAWN, &sdp->sd_flags))
+		return;
+
+	va_start(args, fmt);
+	vaf.fmt = fmt;
+	vaf.va = &args;
+	fs_err(sdp, "%pV", &vaf);
+	va_end(args);
+}
+
+int gfs2_withdraw(struct gfs2_sbd *sdp)
 {
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 	const struct lm_lockops *lm = ls->ls_ops;
-	va_list args;
-	struct va_format vaf;
 
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW &&
 	    test_and_set_bit(SDF_WITHDRAWN, &sdp->sd_flags))
 		return 0;
 
-	if (fmt) {
-		va_start(args, fmt);
-
-		vaf.fmt = fmt;
-		vaf.va = &args;
-
-		fs_err(sdp, "%pV", &vaf);
-
-		va_end(args);
-	}
-
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) {
 		fs_err(sdp, "about to withdraw this file system\n");
 		BUG_ON(sdp->sd_args.ar_debug);
@@ -89,10 +92,12 @@ int gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
 			   const char *function, char *file, unsigned int line)
 {
 	int me;
-	me = gfs2_lm_withdraw(sdp,
-			      "fatal: assertion \"%s\" failed\n"
-			      "   function = %s, file = %s, line = %u\n",
-			      assertion, function, file, line);
+
+	gfs2_lm(sdp,
+		"fatal: assertion \"%s\" failed\n"
+		"   function = %s, file = %s, line = %u\n",
+		assertion, function, file, line);
+	me = gfs2_withdraw(sdp);
 	dump_stack();
 	return (me) ? -1 : -2;
 }
@@ -140,11 +145,10 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
 int gfs2_consist_i(struct gfs2_sbd *sdp, int cluster_wide, const char *function,
 		   char *file, unsigned int line)
 {
-	int rv;
-	rv = gfs2_lm_withdraw(sdp,
-			      "fatal: filesystem consistency error - function = %s, file = %s, line = %u\n",
-			      function, file, line);
-	return rv;
+	gfs2_lm(sdp,
+		"fatal: filesystem consistency error - function = %s, file = %s, line = %u\n",
+		function, file, line);
+	return gfs2_withdraw(sdp);
 }
 
 /**
@@ -157,15 +161,15 @@ int gfs2_consist_inode_i(struct gfs2_inode *ip, int cluster_wide,
 			 const char *function, char *file, unsigned int line)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	int rv;
-	rv = gfs2_lm_withdraw(sdp,
-			      "fatal: filesystem consistency error\n"
-			      "  inode = %llu %llu\n"
-			      "  function = %s, file = %s, line = %u\n",
-			      (unsigned long long)ip->i_no_formal_ino,
-			      (unsigned long long)ip->i_no_addr,
-			      function, file, line);
-	return rv;
+
+	gfs2_lm(sdp,
+		"fatal: filesystem consistency error\n"
+		"  inode = %llu %llu\n"
+		"  function = %s, file = %s, line = %u\n",
+		(unsigned long long)ip->i_no_formal_ino,
+		(unsigned long long)ip->i_no_addr,
+		function, file, line);
+	return gfs2_withdraw(sdp);
 }
 
 /**
@@ -179,17 +183,16 @@ int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd, int cluster_wide,
 {
 	struct gfs2_sbd *sdp = rgd->rd_sbd;
 	char fs_id_buf[sizeof(sdp->sd_fsname) + 7];
-	int rv;
 
 	sprintf(fs_id_buf, "fsid=%s: ", sdp->sd_fsname);
 	gfs2_rgrp_dump(NULL, rgd->rd_gl, fs_id_buf);
-	rv = gfs2_lm_withdraw(sdp,
-			      "fatal: filesystem consistency error\n"
-			      "  RG = %llu\n"
-			      "  function = %s, file = %s, line = %u\n",
-			      (unsigned long long)rgd->rd_addr,
-			      function, file, line);
-	return rv;
+	gfs2_lm(sdp,
+		"fatal: filesystem consistency error\n"
+		"  RG = %llu\n"
+		"  function = %s, file = %s, line = %u\n",
+		(unsigned long long)rgd->rd_addr,
+		function, file, line);
+	return gfs2_withdraw(sdp);
 }
 
 /**
@@ -203,12 +206,14 @@ int gfs2_meta_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh,
 		       unsigned int line)
 {
 	int me;
-	me = gfs2_lm_withdraw(sdp,
-			      "fatal: invalid metadata block\n"
-			      "  bh = %llu (%s)\n"
-			      "  function = %s, file = %s, line = %u\n",
-			      (unsigned long long)bh->b_blocknr, type,
-			      function, file, line);
+
+	gfs2_lm(sdp,
+		"fatal: invalid metadata block\n"
+		"  bh = %llu (%s)\n"
+		"  function = %s, file = %s, line = %u\n",
+		(unsigned long long)bh->b_blocknr, type,
+		function, file, line);
+	me = gfs2_withdraw(sdp);
 	return (me) ? -1 : -2;
 }
 
@@ -223,12 +228,14 @@ int gfs2_metatype_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh,
 			   char *file, unsigned int line)
 {
 	int me;
-	me = gfs2_lm_withdraw(sdp,
-			      "fatal: invalid metadata block\n"
-			      "  bh = %llu (type: exp=%u, found=%u)\n"
-			      "  function = %s, file = %s, line = %u\n",
-			      (unsigned long long)bh->b_blocknr, type, t,
-			      function, file, line);
+
+	gfs2_lm(sdp,
+		"fatal: invalid metadata block\n"
+		"  bh = %llu (type: exp=%u, found=%u)\n"
+		"  function = %s, file = %s, line = %u\n",
+		(unsigned long long)bh->b_blocknr, type, t,
+		function, file, line);
+	me = gfs2_withdraw(sdp);
 	return (me) ? -1 : -2;
 }
 
@@ -241,12 +248,11 @@ int gfs2_metatype_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh,
 int gfs2_io_error_i(struct gfs2_sbd *sdp, const char *function, char *file,
 		    unsigned int line)
 {
-	int rv;
-	rv = gfs2_lm_withdraw(sdp,
-			      "fatal: I/O error\n"
-			      "  function = %s, file = %s, line = %u\n",
-			      function, file, line);
-	return rv;
+	gfs2_lm(sdp,
+		"fatal: I/O error\n"
+		"  function = %s, file = %s, line = %u\n",
+		function, file, line);
+	return gfs2_withdraw(sdp);
 }
 
 /**
@@ -266,6 +272,6 @@ void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh,
 		       (unsigned long long)bh->b_blocknr,
 		       function, file, line);
 	if (withdraw)
-		gfs2_lm_withdraw(sdp, NULL);
+		gfs2_withdraw(sdp);
 }
 
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index f2702bc9837c..fdc218a28609 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -177,6 +177,7 @@ static inline bool gfs2_withdrawn(struct gfs2_sbd *sdp)
 gfs2_tune_get_i(&(sdp)->sd_tune, &(sdp)->sd_tune.field)
 
 __printf(2, 3)
-int gfs2_lm_withdraw(struct gfs2_sbd *sdp, const char *fmt, ...);
+void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...);
+int gfs2_withdraw(struct gfs2_sbd *sdp);
 
 #endif /* __UTIL_DOT_H__ */
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 02/25] gfs2: Report errors before withdraw

[Bob Peterson]

From: Andreas Gruenbacher <agruenba@redhat.com>

In gfs2_rgrp_verify and compute_bitstructs, make sure to report errors before
withdrawing the filesystem: otherwise, when we withdraw first and withdraw is
configured to panic, we'll never get to the error reporting.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/rgrp.c | 48 +++++++++++++++++++++++-------------------------
 1 file changed, 23 insertions(+), 25 deletions(-)

diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index e7bf91ec231c..2bdd662deff5 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -457,24 +457,24 @@ void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd)
 	}
 
 	if (count[0] != rgd->rd_free) {
-		if (gfs2_consist_rgrpd(rgd))
-			fs_err(sdp, "free data mismatch:  %u != %u\n",
-			       count[0], rgd->rd_free);
+		gfs2_lm(sdp, "free data mismatch:  %u != %u\n",
+			count[0], rgd->rd_free);
+		gfs2_consist_rgrpd(rgd);
 		return;
 	}
 
 	tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes;
 	if (count[1] != tmp) {
-		if (gfs2_consist_rgrpd(rgd))
-			fs_err(sdp, "used data mismatch:  %u != %u\n",
-			       count[1], tmp);
+		gfs2_lm(sdp, "used data mismatch:  %u != %u\n",
+			count[1], tmp);
+		gfs2_consist_rgrpd(rgd);
 		return;
 	}
 
 	if (count[2] + count[3] != rgd->rd_dinodes) {
-		if (gfs2_consist_rgrpd(rgd))
-			fs_err(sdp, "used metadata mismatch:  %u != %u\n",
-			       count[2] + count[3], rgd->rd_dinodes);
+		gfs2_lm(sdp, "used metadata mismatch:  %u != %u\n",
+			count[2] + count[3], rgd->rd_dinodes);
+		gfs2_consist_rgrpd(rgd);
 		return;
 	}
 }
@@ -733,17 +733,6 @@ void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
 	}
 }
 
-static void gfs2_rindex_print(const struct gfs2_rgrpd *rgd)
-{
-	struct gfs2_sbd *sdp = rgd->rd_sbd;
-
-	fs_info(sdp, "ri_addr = %llu\n", (unsigned long long)rgd->rd_addr);
-	fs_info(sdp, "ri_length = %u\n", rgd->rd_length);
-	fs_info(sdp, "ri_data0 = %llu\n", (unsigned long long)rgd->rd_data0);
-	fs_info(sdp, "ri_data = %u\n", rgd->rd_data);
-	fs_info(sdp, "ri_bitbytes = %u\n", rgd->rd_bitbytes);
-}
-
 /**
  * gfs2_compute_bitstructs - Compute the bitmap sizes
  * @rgd: The resource group descriptor
@@ -814,11 +803,20 @@ static int compute_bitstructs(struct gfs2_rgrpd *rgd)
 	}
 	bi = rgd->rd_bits + (length - 1);
 	if ((bi->bi_start + bi->bi_bytes) * GFS2_NBBY != rgd->rd_data) {
-		if (gfs2_consist_rgrpd(rgd)) {
-			gfs2_rindex_print(rgd);
-			fs_err(sdp, "start=%u len=%u offset=%u\n",
-			       bi->bi_start, bi->bi_bytes, bi->bi_offset);
-		}
+		gfs2_lm(sdp,
+			"ri_addr = %llu\n"
+			"ri_length = %u\n"
+			"ri_data0 = %llu\n"
+			"ri_data = %u\n"
+			"ri_bitbytes = %u\n"
+			"start=%u len=%u offset=%u\n",
+			(unsigned long long)rgd->rd_addr,
+			rgd->rd_length,
+			(unsigned long long)rgd->rd_data0,
+			rgd->rd_data,
+			rgd->rd_bitbytes,
+			bi->bi_start, bi->bi_bytes, bi->bi_offset);
+		gfs2_consist_rgrpd(rgd);
 		return -EIO;
 	}
 
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 03/25] gfs2: Remove usused cluster_wide arguments of gfs2_consist functions

[Bob Peterson]

From: Andreas Gruenbacher <agruenba@redhat.com>

These arguments are always passed as 0, and they are never evaluated.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/util.c |  6 +++---
 fs/gfs2/util.h | 12 ++++++------
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 322012e2064e..74247d686cf7 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -142,7 +142,7 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
  *          0 if it was already withdrawn
  */
 
-int gfs2_consist_i(struct gfs2_sbd *sdp, int cluster_wide, const char *function,
+int gfs2_consist_i(struct gfs2_sbd *sdp, const char *function,
 		   char *file, unsigned int line)
 {
 	gfs2_lm(sdp,
@@ -157,7 +157,7 @@ int gfs2_consist_i(struct gfs2_sbd *sdp, int cluster_wide, const char *function,
  *          0 if it was already withdrawn
  */
 
-int gfs2_consist_inode_i(struct gfs2_inode *ip, int cluster_wide,
+int gfs2_consist_inode_i(struct gfs2_inode *ip,
 			 const char *function, char *file, unsigned int line)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
@@ -178,7 +178,7 @@ int gfs2_consist_inode_i(struct gfs2_inode *ip, int cluster_wide,
  *          0 if it was already withdrawn
  */
 
-int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd, int cluster_wide,
+int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
 			 const char *function, char *file, unsigned int line)
 {
 	struct gfs2_sbd *sdp = rgd->rd_sbd;
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index fdc218a28609..2c2940d55a35 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -52,25 +52,25 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
 					__func__, __FILE__, __LINE__))
 
 
-int gfs2_consist_i(struct gfs2_sbd *sdp, int cluster_wide,
+int gfs2_consist_i(struct gfs2_sbd *sdp,
 		   const char *function, char *file, unsigned int line);
 
 #define gfs2_consist(sdp) \
-gfs2_consist_i((sdp), 0, __func__, __FILE__, __LINE__)
+gfs2_consist_i((sdp), __func__, __FILE__, __LINE__)
 
 
-int gfs2_consist_inode_i(struct gfs2_inode *ip, int cluster_wide,
+int gfs2_consist_inode_i(struct gfs2_inode *ip,
 			 const char *function, char *file, unsigned int line);
 
 #define gfs2_consist_inode(ip) \
-gfs2_consist_inode_i((ip), 0, __func__, __FILE__, __LINE__)
+gfs2_consist_inode_i((ip), __func__, __FILE__, __LINE__)
 
 
-int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd, int cluster_wide,
+int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
 			 const char *function, char *file, unsigned int line);
 
 #define gfs2_consist_rgrpd(rgd) \
-gfs2_consist_rgrpd_i((rgd), 0, __func__, __FILE__, __LINE__)
+gfs2_consist_rgrpd_i((rgd), __func__, __FILE__, __LINE__)
 
 
 int gfs2_meta_check_ii(struct gfs2_sbd *sdp, struct buffer_head *bh,
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 04/25] gfs2: Turn gfs2_consist into void functions

[Bob Peterson]

From: Andreas Gruenbacher <agruenba@redhat.com>

Change the various gfs2_consist functions to return void.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/util.c | 24 +++++++++---------------
 fs/gfs2/util.h | 12 ++++++------
 2 files changed, 15 insertions(+), 21 deletions(-)

diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 74247d686cf7..698eb5952438 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -138,27 +138,23 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
 
 /**
  * gfs2_consist_i - Flag a filesystem consistency error and withdraw
- * Returns: -1 if this call withdrew the machine,
- *          0 if it was already withdrawn
  */
 
-int gfs2_consist_i(struct gfs2_sbd *sdp, const char *function,
-		   char *file, unsigned int line)
+void gfs2_consist_i(struct gfs2_sbd *sdp, const char *function,
+		    char *file, unsigned int line)
 {
 	gfs2_lm(sdp,
 		"fatal: filesystem consistency error - function = %s, file = %s, line = %u\n",
 		function, file, line);
-	return gfs2_withdraw(sdp);
+	gfs2_withdraw(sdp);
 }
 
 /**
  * gfs2_consist_inode_i - Flag an inode consistency error and withdraw
- * Returns: -1 if this call withdrew the machine,
- *          0 if it was already withdrawn
  */
 
-int gfs2_consist_inode_i(struct gfs2_inode *ip,
-			 const char *function, char *file, unsigned int line)
+void gfs2_consist_inode_i(struct gfs2_inode *ip,
+			  const char *function, char *file, unsigned int line)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 
@@ -169,17 +165,15 @@ int gfs2_consist_inode_i(struct gfs2_inode *ip,
 		(unsigned long long)ip->i_no_formal_ino,
 		(unsigned long long)ip->i_no_addr,
 		function, file, line);
-	return gfs2_withdraw(sdp);
+	gfs2_withdraw(sdp);
 }
 
 /**
  * gfs2_consist_rgrpd_i - Flag a RG consistency error and withdraw
- * Returns: -1 if this call withdrew the machine,
- *          0 if it was already withdrawn
  */
 
-int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
-			 const char *function, char *file, unsigned int line)
+void gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
+			  const char *function, char *file, unsigned int line)
 {
 	struct gfs2_sbd *sdp = rgd->rd_sbd;
 	char fs_id_buf[sizeof(sdp->sd_fsname) + 7];
@@ -192,7 +186,7 @@ int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
 		"  function = %s, file = %s, line = %u\n",
 		(unsigned long long)rgd->rd_addr,
 		function, file, line);
-	return gfs2_withdraw(sdp);
+	gfs2_withdraw(sdp);
 }
 
 /**
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 2c2940d55a35..93e089327216 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -52,22 +52,22 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
 					__func__, __FILE__, __LINE__))
 
 
-int gfs2_consist_i(struct gfs2_sbd *sdp,
-		   const char *function, char *file, unsigned int line);
+void gfs2_consist_i(struct gfs2_sbd *sdp,
+		    const char *function, char *file, unsigned int line);
 
 #define gfs2_consist(sdp) \
 gfs2_consist_i((sdp), __func__, __FILE__, __LINE__)
 
 
-int gfs2_consist_inode_i(struct gfs2_inode *ip,
-			 const char *function, char *file, unsigned int line);
+void gfs2_consist_inode_i(struct gfs2_inode *ip,
+			  const char *function, char *file, unsigned int line);
 
 #define gfs2_consist_inode(ip) \
 gfs2_consist_inode_i((ip), __func__, __FILE__, __LINE__)
 
 
-int gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
-			 const char *function, char *file, unsigned int line);
+void gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
+			  const char *function, char *file, unsigned int line);
 
 #define gfs2_consist_rgrpd(rgd) \
 gfs2_consist_rgrpd_i((rgd), __func__, __FILE__, __LINE__)
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 05/25] gfs2: Return bool from gfs2_assert functions

[Bob Peterson]

From: Andreas Gruenbacher <agruenba@redhat.com>

The gfs2_assert functions only print messages when the filesystem hasn't been
withdrawn yet, and they indicate whether or not they've printed something in
their return value.  However, none of the callers use that information, so
simply return whether or not the assert has failed.

(The gfs2_assert functions are still backwards; they return false when an
assertion is true.)

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/util.c | 21 ++++++---------------
 fs/gfs2/util.h | 28 ++++++++++++++++++----------
 2 files changed, 24 insertions(+), 25 deletions(-)

diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 698eb5952438..ec8e8c5ce848 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -84,37 +84,30 @@ int gfs2_withdraw(struct gfs2_sbd *sdp)
 
 /**
  * gfs2_assert_withdraw_i - Cause the machine to withdraw if @assertion is false
- * Returns: -1 if this call withdrew the machine,
- *          -2 if it was already withdrawn
  */
 
-int gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
-			   const char *function, char *file, unsigned int line)
+void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
+			    const char *function, char *file, unsigned int line)
 {
-	int me;
-
 	gfs2_lm(sdp,
 		"fatal: assertion \"%s\" failed\n"
 		"   function = %s, file = %s, line = %u\n",
 		assertion, function, file, line);
-	me = gfs2_withdraw(sdp);
+	gfs2_withdraw(sdp);
 	dump_stack();
-	return (me) ? -1 : -2;
 }
 
 /**
  * gfs2_assert_warn_i - Print a message to the console if @assertion is false
- * Returns: -1 if we printed something
- *          -2 if we didn't
  */
 
-int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
-		       const char *function, char *file, unsigned int line)
+void gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
+			const char *function, char *file, unsigned int line)
 {
 	if (time_before(jiffies,
 			sdp->sd_last_warning +
 			gfs2_tune_get(sdp, gt_complain_secs) * HZ))
-		return -2;
+		return;
 
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW)
 		fs_warn(sdp, "warning: assertion \"%s\" failed at function = %s, file = %s, line = %u\n",
@@ -132,8 +125,6 @@ int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
 		      sdp->sd_fsname, function, file, line);
 
 	sdp->sd_last_warning = jiffies;
-
-	return -1;
 }
 
 /**
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 93e089327216..572399e75ce6 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -36,21 +36,29 @@ do { \
 } while (0)
 
 
-int gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
-			   const char *function, char *file, unsigned int line);
+void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
+			    const char *function, char *file, unsigned int line);
 
 #define gfs2_assert_withdraw(sdp, assertion) \
-((likely(assertion)) ? 0 : gfs2_assert_withdraw_i((sdp), #assertion, \
-					__func__, __FILE__, __LINE__))
+	({ \
+		bool _bool = (assertion); \
+		if (unlikely(!_bool)) \
+			gfs2_assert_withdraw_i((sdp), #assertion, \
+					__func__, __FILE__, __LINE__); \
+		!_bool; \
+	})
 
-
-int gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
-		       const char *function, char *file, unsigned int line);
+void gfs2_assert_warn_i(struct gfs2_sbd *sdp, char *assertion,
+			const char *function, char *file, unsigned int line);
 
 #define gfs2_assert_warn(sdp, assertion) \
-((likely(assertion)) ? 0 : gfs2_assert_warn_i((sdp), #assertion, \
-					__func__, __FILE__, __LINE__))
-
+	({ \
+		bool _bool = (assertion); \
+		if (unlikely(!_bool)) \
+			gfs2_assert_warn_i((sdp), #assertion, \
+					__func__, __FILE__, __LINE__); \
+		!_bool; \
+	})
 
 void gfs2_consist_i(struct gfs2_sbd *sdp,
 		    const char *function, char *file, unsigned int line);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 06/25] gfs2: Introduce concept of a pending withdraw

[Bob Peterson]

File system withdraws can be delayed when inconsistencies are
discovered when we cannot withdraw immediately, for example, when
critical spin_locks are held. But delaying the withdraw can cause
gfs2 to ignore the error and keep running for a short period of time.
For example, an rgrp glock may be dequeued and demoted while there
are still buffers that haven't been properly revoked, due to io
errors writing to the journal.

This patch introduces a new concept of a pending withdraw, which
means an inconsistency has been discovered and we need to withdraw
at the earliest possible opportunity. In these cases, we aren't
quite withdrawn yet, but we still need to not dequeue glocks and
other critical things. If we dequeue the glocks and the withdraw
results in our journal being replayed, the replay could overwrite
data that's been modified by a different node that acquired the
glock in the meantime.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/incore.h |  1 +
 fs/gfs2/log.c    | 21 ++++++++-------------
 fs/gfs2/util.c   | 14 +++++++-------
 fs/gfs2/util.h   | 12 +++++++++++-
 4 files changed, 27 insertions(+), 21 deletions(-)

diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 9fd88ed18807..3cd2de3db40a 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -618,6 +618,7 @@ enum {
 	SDF_FORCE_AIL_FLUSH     = 9,
 	SDF_AIL1_IO_ERROR	= 10,
 	SDF_FS_FROZEN           = 11,
+	SDF_WITHDRAWING		= 12, /* Will withdraw eventually */
 };
 
 enum gfs2_freeze_state {
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index c4c7c013f7a7..d1ab04135b2f 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -88,8 +88,7 @@ static void gfs2_remove_from_ail(struct gfs2_bufdata *bd)
 
 static int gfs2_ail1_start_one(struct gfs2_sbd *sdp,
 			       struct writeback_control *wbc,
-			       struct gfs2_trans *tr,
-			       bool *withdraw)
+			       struct gfs2_trans *tr)
 __releases(&sdp->sd_ail_lock)
 __acquires(&sdp->sd_ail_lock)
 {
@@ -108,7 +107,7 @@ __acquires(&sdp->sd_ail_lock)
 			    !test_and_set_bit(SDF_AIL1_IO_ERROR,
 					      &sdp->sd_flags)) {
 				gfs2_io_error_bh(sdp, bh);
-				*withdraw = true;
+				gfs2_withdraw_delayed(sdp);
 			}
 			list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list);
 			continue;
@@ -149,7 +148,6 @@ void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc)
 	struct list_head *head = &sdp->sd_ail1_list;
 	struct gfs2_trans *tr;
 	struct blk_plug plug;
-	bool withdraw = false;
 
 	trace_gfs2_ail_flush(sdp, wbc, 1);
 	blk_start_plug(&plug);
@@ -158,13 +156,12 @@ void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc)
 	list_for_each_entry_reverse(tr, head, tr_list) {
 		if (wbc->nr_to_write <= 0)
 			break;
-		if (gfs2_ail1_start_one(sdp, wbc, tr, &withdraw) &&
-		    !gfs2_withdrawn(sdp))
+		if (gfs2_ail1_start_one(sdp, wbc, tr) && !gfs2_withdrawn(sdp))
 			goto restart;
 	}
 	spin_unlock(&sdp->sd_ail_lock);
 	blk_finish_plug(&plug);
-	if (withdraw)
+	if (test_bit(SDF_WITHDRAWING, &sdp->sd_flags))
 		gfs2_withdraw(sdp);
 	trace_gfs2_ail_flush(sdp, wbc, 0);
 }
@@ -193,8 +190,7 @@ static void gfs2_ail1_start(struct gfs2_sbd *sdp)
  *
  */
 
-static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
-				bool *withdraw)
+static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
 {
 	struct gfs2_bufdata *bd, *s;
 	struct buffer_head *bh;
@@ -208,7 +204,7 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
 		if (!buffer_uptodate(bh) &&
 		    !test_and_set_bit(SDF_AIL1_IO_ERROR, &sdp->sd_flags)) {
 			gfs2_io_error_bh(sdp, bh);
-			*withdraw = true;
+			gfs2_withdraw_delayed(sdp);
 		}
 		list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list);
 	}
@@ -226,11 +222,10 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp)
 	struct gfs2_trans *tr, *s;
 	int oldest_tr = 1;
 	int ret;
-	bool withdraw = false;
 
 	spin_lock(&sdp->sd_ail_lock);
 	list_for_each_entry_safe_reverse(tr, s, &sdp->sd_ail1_list, tr_list) {
-		gfs2_ail1_empty_one(sdp, tr, &withdraw);
+		gfs2_ail1_empty_one(sdp, tr);
 		if (list_empty(&tr->tr_ail1_list) && oldest_tr)
 			list_move(&tr->tr_list, &sdp->sd_ail2_list);
 		else
@@ -239,7 +234,7 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp)
 	ret = list_empty(&sdp->sd_ail1_list);
 	spin_unlock(&sdp->sd_ail_lock);
 
-	if (withdraw) {
+	if (test_bit(SDF_WITHDRAWING, &sdp->sd_flags)) {
 		gfs2_lm(sdp, "fatal: I/O error(s)\n");
 		gfs2_withdraw(sdp);
 	}
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index ec8e8c5ce848..47cd40de08b1 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -249,13 +249,13 @@ void gfs2_io_error_bh_i(struct gfs2_sbd *sdp, struct buffer_head *bh,
 			const char *function, char *file, unsigned int line,
 			bool withdraw)
 {
-	if (!gfs2_withdrawn(sdp))
-		fs_err(sdp,
-		       "fatal: I/O error\n"
-		       "  block = %llu\n"
-		       "  function = %s, file = %s, line = %u\n",
-		       (unsigned long long)bh->b_blocknr,
-		       function, file, line);
+	if (gfs2_withdrawn(sdp))
+		return;
+
+	fs_err(sdp, "fatal: I/O error\n"
+	       "  block = %llu\n"
+	       "  function = %s, file = %s, line = %u\n",
+	       (unsigned long long)bh->b_blocknr, function, file, line);
 	if (withdraw)
 		gfs2_withdraw(sdp);
 }
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 572399e75ce6..16b2cc6c4560 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -172,13 +172,23 @@ static inline unsigned int gfs2_tune_get_i(struct gfs2_tune *gt,
 	return x;
 }
 
+/**
+ * gfs2_withdraw_delayed - withdraw as soon as possible without deadlocks
+ * @sdp: the superblock
+ */
+static inline void gfs2_withdraw_delayed(struct gfs2_sbd *sdp)
+{
+	set_bit(SDF_WITHDRAWING, &sdp->sd_flags);
+}
+
 /**
  * gfs2_withdrawn - test whether the file system is withdrawing or withdrawn
  * @sdp: the superblock
  */
 static inline bool gfs2_withdrawn(struct gfs2_sbd *sdp)
 {
-	return test_bit(SDF_WITHDRAWN, &sdp->sd_flags);
+	return test_bit(SDF_WITHDRAWN, &sdp->sd_flags) ||
+		test_bit(SDF_WITHDRAWING, &sdp->sd_flags);
 }
 
 #define gfs2_tune_get(sdp, field) \
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 07/25] gfs2: clear ail1 list when gfs2 withdraws

[Bob Peterson]

This patch fixes a bug in which function gfs2_log_flush can get into
an infinite loop when a gfs2 file system is withdrawn. The problem
is the infinite loop "for (;;)" in gfs2_log_flush which would never
finish because the io error and subsequent withdraw prevented the
items from being taken off the ail list.

This patch tries to clean up the mess by allowing withdraw situations
to move not-in-flight buffer_heads to the ail2 list, where they will
be dealt with later.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/log.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index d1ab04135b2f..9ebec6f93fa3 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -103,16 +103,22 @@ __acquires(&sdp->sd_ail_lock)
 		gfs2_assert(sdp, bd->bd_tr == tr);
 
 		if (!buffer_busy(bh)) {
-			if (!buffer_uptodate(bh) &&
-			    !test_and_set_bit(SDF_AIL1_IO_ERROR,
+			if (buffer_uptodate(bh)) {
+				list_move(&bd->bd_ail_st_list,
+					  &tr->tr_ail2_list);
+				continue;
+			}
+			if (!test_and_set_bit(SDF_AIL1_IO_ERROR,
 					      &sdp->sd_flags)) {
 				gfs2_io_error_bh(sdp, bh);
 				gfs2_withdraw_delayed(sdp);
 			}
-			list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list);
-			continue;
 		}
 
+		if (gfs2_withdrawn(sdp)) {
+			gfs2_remove_from_ail(bd);
+			continue;
+		}
 		if (!buffer_dirty(bh))
 			continue;
 		if (gl == bd->bd_gl)
@@ -859,6 +865,8 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 				if (gfs2_ail1_empty(sdp))
 					break;
 			}
+			if (gfs2_withdrawn(sdp))
+				goto out;
 			atomic_dec(&sdp->sd_log_blks_free); /* Adjust for unreserved buffer */
 			trace_gfs2_log_blocks(sdp, -1);
 			log_write_header(sdp, flags);
@@ -871,6 +879,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 			atomic_set(&sdp->sd_freeze_state, SFS_FROZEN);
 	}
 
+out:
 	trace_gfs2_log_flush(sdp, 0, flags);
 	up_write(&sdp->sd_log_flush_lock);
 
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 08/25] gfs2: Rework how rgrp buffer_heads are managed

[Bob Peterson]

Before this patch, the rgrp code had a serious problem related to
how it managed buffer_heads for resource groups. The problem caused
file system corruption, especially in cases of journal replay.

When an rgrp glock was demoted to transfer ownership to a
different cluster node, do_xmote() first calls rgrp_go_sync and then
rgrp_go_inval, as expected. When it calls rgrp_go_sync, that called
gfs2_rgrp_brelse() that dropped the buffer_head reference count.
In most cases, the reference count went to zero, which is right.
However, there were other places where the buffers are handled
differently.

After rgrp_go_sync, do_xmote called rgrp_go_inval which called
gfs2_rgrp_brelse a second time, then rgrp_go_inval's call to
truncate_inode_pages_range would get rid of the pages in memory,
but only if the reference count drops to 0.

Unfortunately, gfs2_rgrp_brelse was setting bi->bi_bh = NULL.
So when rgrp_go_sync called gfs2_rgrp_brelse, it lost the pointer
to the buffer_heads in cases where the reference count was still 1.
Therefore, when rgrp_go_inval called gfs2_rgrp_brelse a second time,
it failed the check for "if (bi->bi_bh)" and thus failed to call
brelse a second time. Because of that, the reference count on those
buffers sometimes failed to drop from 1 to 0. And that caused
function truncate_inode_pages_range to keep the pages in page cache
rather than freeing them.

The next time the rgrp glock was acquired, the metadata read of
the rgrp buffers re-used the pages in memory, which were now
wrong because they were likely modified by the other node who
acquired the glock in EX (which is why we demoted the glock).
This re-use of the page cache caused corruption because changes
made by the other nodes were never seen, so the bitmaps were
inaccurate.

For some reason, the problem became most apparent when journal
replay forced the replay of rgrps in memory, which caused newer
rgrp data to be overwritten by the older in-core pages.

A big part of the problem was that the rgrp buffer were released
in multiple places: The go_unlock function would release them when
the glock was released rather than when the glock is demoted,
which is clearly wrong because our intent was to cache them until
the glock is demoted from SH or EX.

This patch attempts to clean up the mess and make one consistent
and centralized mechanism for managing the rgrp buffer_heads by
implementing several changes:

1. It eliminates the call to gfs2_rgrp_brelse() from rgrp_go_sync.
   We don't want to release the buffers or zero the pointers when
   syncing for the reasons stated above. It only makes sense to
   release them when the glock is actually invalidated (go_inval).
   And when we do, then we set the bh pointers to NULL.
2. The go_unlock function (which was only used for rgrps) is
   eliminated, as we've talked about doing many times before.
   The go_unlock function was called too early in the glock dq
   process, and should not happen until the glock is invalidated.
3. It also eliminates the call to rgrp_brelse in gfs2_clear_rgrpd.
   That will now happen automatically when the rgrp glocks are
   demoted, and shouldn't happen any sooner or later than that.
   Instead, function gfs2_clear_rgrpd has been modified to demote
   the rgrp glocks, and therefore, free those pages, before the
   remaining glocks are culled by gfs2_gl_hash_clear. This
   prevents the gl_object from hanging around when the glocks are
   culled.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c  |  8 --------
 fs/gfs2/glops.c  |  9 +--------
 fs/gfs2/incore.h |  1 -
 fs/gfs2/rgrp.c   | 23 +++++------------------
 fs/gfs2/rgrp.h   |  1 -
 5 files changed, 6 insertions(+), 36 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index d0eceaff3cea..1cb471a8bc87 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -1241,7 +1241,6 @@ int gfs2_glock_poll(struct gfs2_holder *gh)
 void gfs2_glock_dq(struct gfs2_holder *gh)
 {
 	struct gfs2_glock *gl = gh->gh_gl;
-	const struct gfs2_glock_operations *glops = gl->gl_ops;
 	unsigned delay = 0;
 	int fast_path = 0;
 
@@ -1252,13 +1251,6 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
 	list_del_init(&gh->gh_list);
 	clear_bit(HIF_HOLDER, &gh->gh_iflags);
 	if (find_first_holder(gl) == NULL) {
-		if (glops->go_unlock) {
-			GLOCK_BUG_ON(gl, test_and_set_bit(GLF_LOCK, &gl->gl_flags));
-			spin_unlock(&gl->gl_lockref.lock);
-			glops->go_unlock(gh);
-			spin_lock(&gl->gl_lockref.lock);
-			clear_bit(GLF_LOCK, &gl->gl_flags);
-		}
 		if (list_empty(&gl->gl_holders) &&
 		    !test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
 		    !test_bit(GLF_DEMOTE, &gl->gl_flags))
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 58431f67665e..1c557457c753 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -145,15 +145,9 @@ static void rgrp_go_sync(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	struct address_space *mapping = &sdp->sd_aspace;
-	struct gfs2_rgrpd *rgd;
+	struct gfs2_rgrpd *rgd = gfs2_glock2rgrp(gl);
 	int error;
 
-	spin_lock(&gl->gl_lockref.lock);
-	rgd = gl->gl_object;
-	if (rgd)
-		gfs2_rgrp_brelse(rgd);
-	spin_unlock(&gl->gl_lockref.lock);
-
 	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
 		return;
 	GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE);
@@ -601,7 +595,6 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_sync = rgrp_go_sync,
 	.go_inval = rgrp_go_inval,
 	.go_lock = gfs2_rgrp_go_lock,
-	.go_unlock = gfs2_rgrp_go_unlock,
 	.go_dump = gfs2_rgrp_dump,
 	.go_type = LM_TYPE_RGRP,
 	.go_flags = GLOF_LVB,
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 3cd2de3db40a..b95c8a31d309 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -239,7 +239,6 @@ struct gfs2_glock_operations {
 	void (*go_inval) (struct gfs2_glock *gl, int flags);
 	int (*go_demote_ok) (const struct gfs2_glock *gl);
 	int (*go_lock) (struct gfs2_holder *gh);
-	void (*go_unlock) (struct gfs2_holder *gh);
 	void (*go_dump)(struct seq_file *seq, struct gfs2_glock *gl,
 			const char *fs_id_buf);
 	void (*go_callback)(struct gfs2_glock *gl, bool remote);
diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index 2bdd662deff5..2ee2f7d48bc1 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -720,8 +720,12 @@ void gfs2_clear_rgrpd(struct gfs2_sbd *sdp)
 		rb_erase(n, &sdp->sd_rindex_tree);
 
 		if (gl) {
-			glock_clear_object(gl, rgd);
+			if (gl->gl_state != LM_ST_UNLOCKED) {
+				gfs2_glock_cb(gl, LM_ST_UNLOCKED);
+				flush_delayed_work(&gl->gl_work);
+			}
 			gfs2_rgrp_brelse(rgd);
+			glock_clear_object(gl, rgd);
 			gfs2_glock_put(gl);
 		}
 
@@ -1284,23 +1288,6 @@ void gfs2_rgrp_brelse(struct gfs2_rgrpd *rgd)
 			bi->bi_bh = NULL;
 		}
 	}
-
-}
-
-/**
- * gfs2_rgrp_go_unlock - Unlock a rgrp glock
- * @gh: The glock holder for the resource group
- *
- */
-
-void gfs2_rgrp_go_unlock(struct gfs2_holder *gh)
-{
-	struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object;
-	int demote_requested = test_bit(GLF_DEMOTE, &gh->gh_gl->gl_flags) |
-		test_bit(GLF_PENDING_DEMOTE, &gh->gh_gl->gl_flags);
-
-	if (rgd && demote_requested)
-		gfs2_rgrp_brelse(rgd);
 }
 
 int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset,
diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h
index c14a673ae36f..a584f3096418 100644
--- a/fs/gfs2/rgrp.h
+++ b/fs/gfs2/rgrp.h
@@ -33,7 +33,6 @@ extern int gfs2_rindex_update(struct gfs2_sbd *sdp);
 extern void gfs2_free_clones(struct gfs2_rgrpd *rgd);
 extern int gfs2_rgrp_go_lock(struct gfs2_holder *gh);
 extern void gfs2_rgrp_brelse(struct gfs2_rgrpd *rgd);
-extern void gfs2_rgrp_go_unlock(struct gfs2_holder *gh);
 
 extern struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip);
 
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 09/25] gfs2: log error reform

[Bob Peterson]

Before this patch, gfs2 kept track of journal io errors in two
places sd_log_error and the SDF_AIL1_IO_ERROR flag in sd_flags.
This patch consolidates the two into sd_log_error so that it
reflects the first error encountered writing to the journal.
In future patches, we will take advantage of this by checking
this value rather than having to check both when reacting to
io errors.

In addition, this fixes a tight loop in unmount: If buffers
get on the ail1 list and an io error occurs elsewhere, the
ail1 list would never be cleared because they were always busy.
So unmount would hang, waiting for the ail1 list to empty.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/incore.h |  7 +++----
 fs/gfs2/log.c    | 16 ++++++++++++----
 fs/gfs2/quota.c  |  2 +-
 3 files changed, 16 insertions(+), 9 deletions(-)

diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index b95c8a31d309..ab89f746b3b6 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -615,9 +615,8 @@ enum {
 	SDF_RORECOVERY		= 7, /* read only recovery */
 	SDF_SKIP_DLM_UNLOCK	= 8,
 	SDF_FORCE_AIL_FLUSH     = 9,
-	SDF_AIL1_IO_ERROR	= 10,
-	SDF_FS_FROZEN           = 11,
-	SDF_WITHDRAWING		= 12, /* Will withdraw eventually */
+	SDF_FS_FROZEN           = 10,
+	SDF_WITHDRAWING		= 11, /* Will withdraw eventually */
 };
 
 enum gfs2_freeze_state {
@@ -828,7 +827,7 @@ struct gfs2_sbd {
 	atomic_t sd_log_in_flight;
 	struct bio *sd_log_bio;
 	wait_queue_head_t sd_log_flush_wait;
-	int sd_log_error;
+	int sd_log_error; /* First log error */
 
 	atomic_t sd_reserving_log;
 	wait_queue_head_t sd_reserving_log_wait;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 9ebec6f93fa3..584bb7ce15bf 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -108,8 +108,7 @@ __acquires(&sdp->sd_ail_lock)
 					  &tr->tr_ail2_list);
 				continue;
 			}
-			if (!test_and_set_bit(SDF_AIL1_IO_ERROR,
-					      &sdp->sd_flags)) {
+			if (!cmpxchg(&sdp->sd_log_error, 0, -EIO)) {
 				gfs2_io_error_bh(sdp, bh);
 				gfs2_withdraw_delayed(sdp);
 			}
@@ -205,10 +204,19 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
 					 bd_ail_st_list) {
 		bh = bd->bd_bh;
 		gfs2_assert(sdp, bd->bd_tr == tr);
-		if (buffer_busy(bh))
+		/*
+		 * If another process flagged an io error, e.g. writing to the
+		 * journal, error all other bhs and move them off the ail1 to
+		 * prevent a tight loop when unmount tries to flush ail1,
+		 * regardless of whether they're still busy. If no outside
+		 * errors were found and the buffer is busy, move to the next.
+		 * If the ail buffer is not busy and caught an error, flag it
+		 * for others.
+		 */
+		if (!sdp->sd_log_error && buffer_busy(bh))
 			continue;
 		if (!buffer_uptodate(bh) &&
-		    !test_and_set_bit(SDF_AIL1_IO_ERROR, &sdp->sd_flags)) {
+		    !cmpxchg(&sdp->sd_log_error, 0, -EIO)) {
 			gfs2_io_error_bh(sdp, bh);
 			gfs2_withdraw_delayed(sdp);
 		}
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index e9f93045eb01..ca2194cfa38e 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -1477,7 +1477,7 @@ static void quotad_error(struct gfs2_sbd *sdp, const char *msg, int error)
 		return;
 	if (!gfs2_withdrawn(sdp)) {
 		fs_err(sdp, "gfs2_quotad: %s error %d\n", msg, error);
-		sdp->sd_log_error = error;
+		cmpxchg(&sdp->sd_log_error, 0, error);
 		wake_up(&sdp->sd_logd_waitq);
 	}
 }
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 10/25] gfs2: Only complain the first time an io error occurs in quota or log

[Bob Peterson]

Before this patch, all io errors received by the quota daemon or the
logd daemon would cause a complaint message to be issued, such as:

   gfs2: fsid=dm-13.0: Error 10 writing to journal, jid=0

This patch changes it so that the error message is only issued the
first time the error is encountered.

Also, before this patch function gfs2_end_log_write did not set the
sd_log_error value, so log errors would not cause the file system to
be withdrawn. This patch sets the error code so the file system is
properly withdrawn if an io error is encountered writing to the journal.

WARNING: This change in function breaks check xfstests generic/441
and causes it to fail: io errors writing to the log should cause a
file system to be withdrawn, and no further operations are tolerated.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/lops.c  | 5 +++--
 fs/gfs2/quota.c | 4 ++--
 2 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index d9431724b788..45fa5cd88feb 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -203,8 +203,9 @@ static void gfs2_end_log_write(struct bio *bio)
 	struct bvec_iter_all iter_all;
 
 	if (bio->bi_status) {
-		fs_err(sdp, "Error %d writing to journal, jid=%u\n",
-		       bio->bi_status, sdp->sd_jdesc->jd_jid);
+		if (!cmpxchg(&sdp->sd_log_error, 0, (int)bio->bi_status))
+			fs_err(sdp, "Error %d writing to journal, jid=%u\n",
+			       bio->bi_status, sdp->sd_jdesc->jd_jid);
 		wake_up(&sdp->sd_logd_waitq);
 	}
 
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index ca2194cfa38e..dbe87b2b55af 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -1476,8 +1476,8 @@ static void quotad_error(struct gfs2_sbd *sdp, const char *msg, int error)
 	if (error == 0 || error == -EROFS)
 		return;
 	if (!gfs2_withdrawn(sdp)) {
-		fs_err(sdp, "gfs2_quotad: %s error %d\n", msg, error);
-		cmpxchg(&sdp->sd_log_error, 0, error);
+		if (!cmpxchg(&sdp->sd_log_error, 0, error))
+			fs_err(sdp, "gfs2_quotad: %s error %d\n", msg, error);
 		wake_up(&sdp->sd_logd_waitq);
 	}
 }
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 11/25] gfs2: Ignore dlm recovery requests if gfs2 is withdrawn

[Bob Peterson]

When a node fails, user space informs dlm of the node failure,
and dlm instructs gfs2 on the surviving nodes to perform journal
recovery. It does this by calling various callback functions in
lock_dlm.c. To mark its progress, it keeps generation numbers
and recover bits in a dlm "control" lock lvb, which is seen by
all nodes to determine which journals need to be replayed.

The gfs2 on all nodes get the same recovery requests from dlm,
so they all try to do the recovery, but only one will be
granted the exclusive lock on the journal. The others fail
with a "Busy" message on their "try lock."

However, when a node is withdrawn, it cannot safely do any
recovery or replay any journals. To make matters worse,
gfs2 might withdraw as a result of attempting recovery. For
example, this might happen if the device goes offline, or if
an hba fails. But in today's gfs2 code, it doesn't check for
being withdrawn at any step in the recovery process. What's
worse is that these callbacks from dlm have no return code,
so there is no way to indicate failure back to dlm. We can
send a "Recovery failed" uevent eventually, but that tells
user space what happened, not dlm's kernel code.

Before this patch, lock_dlm would perform its recovery steps but
ignore the result, and eventually it would still update its
generation number in the lvb, despite the fact that it may have
withdrawn or encountered an error. The other nodes would then
see the newer generation number in the lvb and conclude that
they don't need to do recovery because the generation number
is newer than the last one they saw. They think a different
node has already recovered the journal.

This patch adds checks to several of the callbacks used by dlm
in its recovery state machine so that the functions are ignored
and skipped if an io error has occurred or if the file system
is withdrawn. That prevents the lvb bits from being updated, and
therefore dlm and user space still see the need for recovery to
take place.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/lock_dlm.c | 18 ++++++++++++++++++
 fs/gfs2/recovery.c |  5 +++++
 2 files changed, 23 insertions(+)

diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c
index 7c7197343ee2..57fdf53d2246 100644
--- a/fs/gfs2/lock_dlm.c
+++ b/fs/gfs2/lock_dlm.c
@@ -1079,6 +1079,10 @@ static void gdlm_recover_prep(void *arg)
 	struct gfs2_sbd *sdp = arg;
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 
+	if (gfs2_withdrawn(sdp)) {
+		fs_err(sdp, "recover_prep ignored due to withdraw.\n");
+		return;
+	}
 	spin_lock(&ls->ls_recover_spin);
 	ls->ls_recover_block = ls->ls_recover_start;
 	set_bit(DFL_DLM_RECOVERY, &ls->ls_recover_flags);
@@ -1101,6 +1105,11 @@ static void gdlm_recover_slot(void *arg, struct dlm_slot *slot)
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 	int jid = slot->slot - 1;
 
+	if (gfs2_withdrawn(sdp)) {
+		fs_err(sdp, "recover_slot jid %d ignored due to withdraw.\n",
+		       jid);
+		return;
+	}
 	spin_lock(&ls->ls_recover_spin);
 	if (ls->ls_recover_size < jid + 1) {
 		fs_err(sdp, "recover_slot jid %d gen %u short size %d\n",
@@ -1125,6 +1134,10 @@ static void gdlm_recover_done(void *arg, struct dlm_slot *slots, int num_slots,
 	struct gfs2_sbd *sdp = arg;
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 
+	if (gfs2_withdrawn(sdp)) {
+		fs_err(sdp, "recover_done ignored due to withdraw.\n");
+		return;
+	}
 	/* ensure the ls jid arrays are large enough */
 	set_recover_size(sdp, slots, num_slots);
 
@@ -1152,6 +1165,11 @@ static void gdlm_recovery_result(struct gfs2_sbd *sdp, unsigned int jid,
 {
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 
+	if (gfs2_withdrawn(sdp)) {
+		fs_err(sdp, "recovery_result jid %d ignored due to withdraw.\n",
+		       jid);
+		return;
+	}
 	if (test_bit(DFL_NO_DLM_OPS, &ls->ls_recover_flags))
 		return;
 
diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c
index 85f830e56945..8cc26bef4e64 100644
--- a/fs/gfs2/recovery.c
+++ b/fs/gfs2/recovery.c
@@ -305,6 +305,11 @@ void gfs2_recover_func(struct work_struct *work)
 	int error = 0;
 	int jlocked = 0;
 
+	if (gfs2_withdrawn(sdp)) {
+		fs_err(sdp, "jid=%u: Recovery not attempted due to withdraw.\n",
+		       jd->jd_jid);
+		goto fail;
+	}
 	t_start = ktime_get();
 	if (sdp->sd_args.ar_spectator)
 		goto fail;
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 12/25] gfs2: move check_journal_clean to util.c for future use

[Bob Peterson]

Before this patch function check_journal_clean was in ops_fstype.c.
This patch moves it to util.c so we can make use of it elsewhere
in a future patch.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/ops_fstype.c | 42 -----------------------------------------
 fs/gfs2/util.c       | 45 ++++++++++++++++++++++++++++++++++++++++++++
 fs/gfs2/util.h       |  1 +
 3 files changed, 46 insertions(+), 42 deletions(-)

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index b3e904bcc02c..d19ee57c99ce 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -600,48 +600,6 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
 	return error;
 }
 
-/**
- * check_journal_clean - Make sure a journal is clean for a spectator mount
- * @sdp: The GFS2 superblock
- * @jd: The journal descriptor
- *
- * Returns: 0 if the journal is clean or locked, else an error
- */
-static int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
-{
-	int error;
-	struct gfs2_holder j_gh;
-	struct gfs2_log_header_host head;
-	struct gfs2_inode *ip;
-
-	ip = GFS2_I(jd->jd_inode);
-	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_NOEXP |
-				   GL_EXACT | GL_NOCACHE, &j_gh);
-	if (error) {
-		fs_err(sdp, "Error locking journal for spectator mount.\n");
-		return -EPERM;
-	}
-	error = gfs2_jdesc_check(jd);
-	if (error) {
-		fs_err(sdp, "Error checking journal for spectator mount.\n");
-		goto out_unlock;
-	}
-	error = gfs2_find_jhead(jd, &head, false);
-	if (error) {
-		fs_err(sdp, "Error parsing journal for spectator mount.\n");
-		goto out_unlock;
-	}
-	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
-		error = -EPERM;
-		fs_err(sdp, "jid=%u: Journal is dirty, so the first mounter "
-		       "must not be a spectator.\n", jd->jd_jid);
-	}
-
-out_unlock:
-	gfs2_glock_dq_uninit(&j_gh);
-	return error;
-}
-
 static int init_journal(struct gfs2_sbd *sdp, int undo)
 {
 	struct inode *master = d_inode(sdp->sd_master_dir);
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 47cd40de08b1..86965e6089c6 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -16,7 +16,10 @@
 #include "gfs2.h"
 #include "incore.h"
 #include "glock.h"
+#include "lops.h"
+#include "recovery.h"
 #include "rgrp.h"
+#include "super.h"
 #include "util.h"
 
 struct kmem_cache *gfs2_glock_cachep __read_mostly;
@@ -33,6 +36,48 @@ void gfs2_assert_i(struct gfs2_sbd *sdp)
 	fs_emerg(sdp, "fatal assertion failed\n");
 }
 
+/**
+ * check_journal_clean - Make sure a journal is clean for a spectator mount
+ * @sdp: The GFS2 superblock
+ * @jd: The journal descriptor
+ *
+ * Returns: 0 if the journal is clean or locked, else an error
+ */
+int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
+{
+	int error;
+	struct gfs2_holder j_gh;
+	struct gfs2_log_header_host head;
+	struct gfs2_inode *ip;
+
+	ip = GFS2_I(jd->jd_inode);
+	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_NOEXP |
+				   GL_EXACT | GL_NOCACHE, &j_gh);
+	if (error) {
+		fs_err(sdp, "Error locking journal for spectator mount.\n");
+		return -EPERM;
+	}
+	error = gfs2_jdesc_check(jd);
+	if (error) {
+		fs_err(sdp, "Error checking journal for spectator mount.\n");
+		goto out_unlock;
+	}
+	error = gfs2_find_jhead(jd, &head, false);
+	if (error) {
+		fs_err(sdp, "Error parsing journal for spectator mount.\n");
+		goto out_unlock;
+	}
+	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
+		error = -EPERM;
+		fs_err(sdp, "jid=%u: Journal is dirty, so the first mounter "
+		       "must not be a spectator.\n", jd->jd_jid);
+	}
+
+out_unlock:
+	gfs2_glock_dq_uninit(&j_gh);
+	return error;
+}
+
 void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...)
 {
 	struct va_format vaf;
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 16b2cc6c4560..cf613497a20e 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -136,6 +136,7 @@ static inline void gfs2_metatype_set(struct buffer_head *bh, u16 type,
 
 int gfs2_io_error_i(struct gfs2_sbd *sdp, const char *function,
 		    char *file, unsigned int line);
+int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd);
 
 #define gfs2_io_error(sdp) \
 gfs2_io_error_i((sdp), __func__, __FILE__, __LINE__);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 13/25] gfs2: Allow some glocks to be used during withdraw

[Bob Peterson]

We need to allow some glocks to be enqueued, dequeued, promoted, and demoted
when we're withdrawn. For example, to maintain metadata integrity, we should
disallow the use of inode and rgrp glocks when withdrawn. Other glocks, like
iopen or the transaction glocks may be safely used because none of their
metadata goes through the journal. So in general, we should disallow all
glocks with an address space, and allow all the others. One exception is:
we need to allow our active journal to be demoted so others may recover it.

Allowing glocks after withdraw gives us the ability to take appropriate
action (in a following patch) to have our journal properly replayed by
another node rather than just abandoning the current transactions and
pretending nothing bad happened, leaving the other nodes free to modify
the blocks we had in our journal, which may result in file system
corruption.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c  | 31 +++++++++++++++++++++++++++----
 fs/gfs2/glops.c  | 10 +++++++---
 fs/gfs2/incore.h |  7 ++++---
 3 files changed, 38 insertions(+), 10 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 1cb471a8bc87..69c0c388e7be 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -133,6 +133,31 @@ static void gfs2_glock_dealloc(struct rcu_head *rcu)
 	}
 }
 
+/**
+ * glock_blocked_by_withdraw - determine if we can still use a glock
+ * @gl: the glock
+ *
+ * We need to allow some glocks to be enqueued, dequeued, promoted, and demoted
+ * when we're withdrawn. For example, to maintain metadata integrity, we should
+ * disallow the use of inode and rgrp glocks when withdrawn. Other glocks, like
+ * iopen or the transaction glocks may be safely used because none of their
+ * metadata goes through the journal. So in general, we should disallow all
+ * glocks that are journaled, and allow all the others. One exception is:
+ * we need to allow our active journal to be demoted so others may recover it.
+ */
+static bool glock_blocked_by_withdraw(struct gfs2_glock *gl)
+{
+	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
+
+	if (likely(!gfs2_withdrawn(sdp)))
+		return false;
+	if (gl->gl_ops->go_flags & GLOF_NONDISK)
+		return false;
+	if (!sdp->sd_jdesc || gl->gl_object == sdp->sd_jdesc->jd_inode)
+		return false;
+	return true;
+}
+
 void gfs2_glock_free(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
@@ -549,8 +574,7 @@ __acquires(&gl->gl_lockref.lock)
 	unsigned int lck_flags = (unsigned int)(gh ? gh->gh_flags : 0);
 	int ret;
 
-	if (unlikely(gfs2_withdrawn(sdp)) &&
-	    target != LM_ST_UNLOCKED)
+	if (target != LM_ST_UNLOCKED && glock_blocked_by_withdraw(gl))
 		return;
 	lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP |
 		      LM_FLAG_PRIORITY);
@@ -1194,10 +1218,9 @@ __acquires(&gl->gl_lockref.lock)
 int gfs2_glock_nq(struct gfs2_holder *gh)
 {
 	struct gfs2_glock *gl = gh->gh_gl;
-	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	int error = 0;
 
-	if (unlikely(gfs2_withdrawn(sdp)))
+	if (glock_blocked_by_withdraw(gl))
 		return -EIO;
 
 	if (test_bit(GLF_LRU, &gl->gl_flags))
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 1c557457c753..3553ca939337 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -579,6 +579,7 @@ static void iopen_go_callback(struct gfs2_glock *gl, bool remote)
 
 const struct gfs2_glock_operations gfs2_meta_glops = {
 	.go_type = LM_TYPE_META,
+	.go_flags = GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_inode_glops = {
@@ -605,30 +606,33 @@ const struct gfs2_glock_operations gfs2_freeze_glops = {
 	.go_xmote_bh = freeze_go_xmote_bh,
 	.go_demote_ok = freeze_go_demote_ok,
 	.go_type = LM_TYPE_NONDISK,
+	.go_flags = GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_iopen_glops = {
 	.go_type = LM_TYPE_IOPEN,
 	.go_callback = iopen_go_callback,
-	.go_flags = GLOF_LRU,
+	.go_flags = GLOF_LRU | GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_flock_glops = {
 	.go_type = LM_TYPE_FLOCK,
-	.go_flags = GLOF_LRU,
+	.go_flags = GLOF_LRU | GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_nondisk_glops = {
 	.go_type = LM_TYPE_NONDISK,
+	.go_flags = GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_quota_glops = {
 	.go_type = LM_TYPE_QUOTA,
-	.go_flags = GLOF_LVB | GLOF_LRU,
+	.go_flags = GLOF_LVB | GLOF_LRU | GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations gfs2_journal_glops = {
 	.go_type = LM_TYPE_JOURNAL,
+	.go_flags = GLOF_NONDISK,
 };
 
 const struct gfs2_glock_operations *gfs2_glops_list[] = {
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index ab89f746b3b6..0d1cca2ac3b9 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -244,9 +244,10 @@ struct gfs2_glock_operations {
 	void (*go_callback)(struct gfs2_glock *gl, bool remote);
 	const int go_type;
 	const unsigned long go_flags;
-#define GLOF_ASPACE 1
-#define GLOF_LVB    2
-#define GLOF_LRU    4
+#define GLOF_ASPACE 1 /* address space attached */
+#define GLOF_LVB    2 /* Lock Value Block attached */
+#define GLOF_LRU    4 /* LRU managed */
+#define GLOF_NONDISK   8 /* not I/O related */
 };
 
 enum {
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 14/25] gfs2: Make secondary withdrawers wait for first withdrawer

[Bob Peterson]

Before this patch, if a process encountered an error and decided to
withdraw, if another process was already in the process of withdrawing,
the secondary withdraw would be silently ignored, which set it free
to proceed with its processing, unlock any locks, etc. That's correct
behavior if the original withdrawer encounters further errors down
the road. However, second withdrawers need to wait for the first
withdrawer to finish its withdraw before proceeding. If we don't wait
we could end up assuming everything is alright, unlock glocks and
telling other nodes they can have the glock, despite the fact that
a withdraw is still ongoing and may require a journal replay before
any locks are released. For example, if an rgrp glock is freed
by a process that didn't wait for the withdraw, a journal replay
could introduce file system corruption by replaying a rgrp block
that has already been granted to another node.

This patch makes secondary withdrawers wait with their processing until
the primary withdrawer is finished.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c      |  23 ++++++-
 fs/gfs2/glops.c      |  77 ++++++++++++++++++++-
 fs/gfs2/incore.h     |   9 +++
 fs/gfs2/lock_dlm.c   |  34 +++++++++
 fs/gfs2/meta_io.c    |   2 +-
 fs/gfs2/ops_fstype.c |  11 ++-
 fs/gfs2/quota.c      |   3 +
 fs/gfs2/super.c      |  73 +++++++++++++++-----
 fs/gfs2/util.c       | 160 ++++++++++++++++++++++++++++++++++++++++++-
 9 files changed, 362 insertions(+), 30 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 69c0c388e7be..7094ef07003c 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -269,7 +269,7 @@ static void __gfs2_glock_put(struct gfs2_glock *gl)
 	gfs2_glock_remove_from_lru(gl);
 	spin_unlock(&gl->gl_lockref.lock);
 	GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
-	GLOCK_BUG_ON(gl, mapping && mapping->nrpages);
+	GLOCK_BUG_ON(gl, mapping && mapping->nrpages && !gfs2_withdrawn(sdp));
 	trace_gfs2_glock_put(gl);
 	sdp->sd_lockstruct.ls_ops->lm_put_lock(gl);
 }
@@ -574,7 +574,8 @@ __acquires(&gl->gl_lockref.lock)
 	unsigned int lck_flags = (unsigned int)(gh ? gh->gh_flags : 0);
 	int ret;
 
-	if (target != LM_ST_UNLOCKED && glock_blocked_by_withdraw(gl))
+	if (target != LM_ST_UNLOCKED && glock_blocked_by_withdraw(gl) &&
+	    gh && !(gh->gh_flags & LM_FLAG_NOEXP))
 		return;
 	lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP |
 		      LM_FLAG_PRIORITY);
@@ -1220,7 +1221,7 @@ int gfs2_glock_nq(struct gfs2_holder *gh)
 	struct gfs2_glock *gl = gh->gh_gl;
 	int error = 0;
 
-	if (glock_blocked_by_withdraw(gl))
+	if (glock_blocked_by_withdraw(gl) && !(gh->gh_flags & LM_FLAG_NOEXP))
 		return -EIO;
 
 	if (test_bit(GLF_LRU, &gl->gl_flags))
@@ -1264,10 +1265,26 @@ int gfs2_glock_poll(struct gfs2_holder *gh)
 void gfs2_glock_dq(struct gfs2_holder *gh)
 {
 	struct gfs2_glock *gl = gh->gh_gl;
+	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	unsigned delay = 0;
 	int fast_path = 0;
 
 	spin_lock(&gl->gl_lockref.lock);
+	/*
+	 * If we're in the process of file system withdraw, we cannot just
+	 * dequeue any glocks until our journal is recovered, lest we
+	 * introduce file system corruption. We need two exceptions to this
+	 * rule: We need to allow unlocking of nondisk glocks and the glock
+	 * for our own journal that needs recovery.
+	 */
+	if (test_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags) &&
+	    glock_blocked_by_withdraw(gl) &&
+	    gh->gh_gl != sdp->sd_jinode_gl) {
+		sdp->sd_glock_dqs_held++;
+		might_sleep();
+		wait_on_bit(&sdp->sd_flags, SDF_WITHDRAW_RECOVERY,
+			    TASK_UNINTERRUPTIBLE);
+	}
 	if (gh->gh_flags & GL_NOCACHE)
 		handle_callback(gl, LM_ST_UNLOCKED, 0, false);
 
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 3553ca939337..7cfacbe35e59 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -29,6 +29,8 @@
 
 struct workqueue_struct *gfs2_freeze_wq;
 
+extern struct workqueue_struct *gfs2_control_wq;
+
 static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
 {
 	fs_err(gl->gl_name.ln_sbd,
@@ -496,13 +498,17 @@ static void freeze_go_sync(struct gfs2_glock *gl)
 	int error = 0;
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 
-	if (gl->gl_state == LM_ST_SHARED &&
+	if (gl->gl_state == LM_ST_SHARED && !gfs2_withdrawn(sdp) &&
 	    test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
 		atomic_set(&sdp->sd_freeze_state, SFS_STARTING_FREEZE);
 		error = freeze_super(sdp->sd_vfs);
 		if (error) {
 			fs_info(sdp, "GFS2: couldn't freeze filesystem: %d\n",
 				error);
+			if (gfs2_withdrawn(sdp)) {
+				atomic_set(&sdp->sd_freeze_state, SFS_UNFROZEN);
+				return;
+			}
 			gfs2_assert_withdraw(sdp, 0);
 		}
 		queue_work(gfs2_freeze_wq, &sdp->sd_freeze_work);
@@ -577,6 +583,73 @@ static void iopen_go_callback(struct gfs2_glock *gl, bool remote)
 	}
 }
 
+/**
+ * inode_go_free - wake up anyone waiting for dlm's unlock ast to free it
+ * @gl: glock being freed
+ *
+ * For now, this is only used for the journal inode glock. In withdraw
+ * situations, we need to wait for the glock to be freed so that we know
+ * other nodes may proceed with recovery / journal replay.
+ */
+static void inode_go_free(struct gfs2_glock *gl)
+{
+	/* Note that we cannot reference gl_object because it's already set
+	 * to NULL by this point in its lifecycle. */
+	if (!test_bit(GLF_FREEING, &gl->gl_flags))
+		return;
+	clear_bit_unlock(GLF_FREEING, &gl->gl_flags);
+	wake_up_bit(&gl->gl_flags, GLF_FREEING);
+}
+
+/**
+ * nondisk_go_callback - used to signal when a node did a withdraw
+ * @gl: the nondisk glock
+ * @remote: true if this came from a different cluster node
+ *
+ */
+static void nondisk_go_callback(struct gfs2_glock *gl, bool remote)
+{
+	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
+
+	/* Ignore the callback unless it's from another node, and it's the
+	   live lock. */
+	if (!remote || gl->gl_name.ln_number != GFS2_LIVE_LOCK)
+		return;
+
+	/* First order of business is to cancel the demote request. We don't
+	 * really want to demote a nondisk glock. At best it's just to inform
+	 * us of another node's withdraw. We'll keep it in SH mode. */
+	clear_bit(GLF_DEMOTE, &gl->gl_flags);
+	clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
+
+	/* Ignore the unlock if we're withdrawn, unmounting, or in recovery. */
+	if (test_bit(SDF_NORECOVERY, &sdp->sd_flags) ||
+	    test_bit(SDF_WITHDRAWN, &sdp->sd_flags) ||
+	    test_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags))
+		return;
+
+	/* We only care when a node wants us to unlock, because that means
+	 * they want a journal recovered. */
+	if (gl->gl_demote_state != LM_ST_UNLOCKED)
+		return;
+
+	if (sdp->sd_args.ar_spectator) {
+		fs_warn(sdp, "Spectator node cannot recover journals.\n");
+		return;
+	}
+
+	fs_warn(sdp, "Some node has withdrawn; checking for recovery.\n");
+	set_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags);
+	/*
+	 * We can't call remote_withdraw directly here or gfs2_recover_journal
+	 * because this is called from the glock unlock function and the
+	 * remote_withdraw needs to enqueue and dequeue the same "live" glock
+	 * we were called from. So we queue it to the control work queue in
+	 * lock_dlm.
+	 */
+	queue_delayed_work(gfs2_control_wq, &sdp->sd_control_work, 0);
+}
+
 const struct gfs2_glock_operations gfs2_meta_glops = {
 	.go_type = LM_TYPE_META,
 	.go_flags = GLOF_NONDISK,
@@ -590,6 +663,7 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
 	.go_dump = inode_go_dump,
 	.go_type = LM_TYPE_INODE,
 	.go_flags = GLOF_ASPACE | GLOF_LRU,
+	.go_free = inode_go_free,
 };
 
 const struct gfs2_glock_operations gfs2_rgrp_glops = {
@@ -623,6 +697,7 @@ const struct gfs2_glock_operations gfs2_flock_glops = {
 const struct gfs2_glock_operations gfs2_nondisk_glops = {
 	.go_type = LM_TYPE_NONDISK,
 	.go_flags = GLOF_NONDISK,
+	.go_callback = nondisk_go_callback,
 };
 
 const struct gfs2_glock_operations gfs2_quota_glops = {
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 0d1cca2ac3b9..259b1956e990 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -242,6 +242,7 @@ struct gfs2_glock_operations {
 	void (*go_dump)(struct seq_file *seq, struct gfs2_glock *gl,
 			const char *fs_id_buf);
 	void (*go_callback)(struct gfs2_glock *gl, bool remote);
+	void (*go_free)(struct gfs2_glock *gl);
 	const int go_type;
 	const unsigned long go_flags;
 #define GLOF_ASPACE 1 /* address space attached */
@@ -343,6 +344,7 @@ enum {
 	GLF_OBJECT			= 14, /* Used only for tracing */
 	GLF_BLOCKING			= 15,
 	GLF_INODE_CREATING		= 16, /* Inode creation occurring */
+	GLF_FREEING			= 18, /* Wait for glock to be freed */
 };
 
 struct gfs2_glock {
@@ -618,6 +620,10 @@ enum {
 	SDF_FORCE_AIL_FLUSH     = 9,
 	SDF_FS_FROZEN           = 10,
 	SDF_WITHDRAWING		= 11, /* Will withdraw eventually */
+	SDF_WITHDRAW_IN_PROG	= 12, /* Withdraw is in progress */
+	SDF_REMOTE_WITHDRAW	= 13, /* Performing remote recovery */
+	SDF_WITHDRAW_RECOVERY	= 14, /* Wait for journal recovery when we are
+					 withdrawing */
 };
 
 enum gfs2_freeze_state {
@@ -768,6 +774,7 @@ struct gfs2_sbd {
 	struct gfs2_jdesc *sd_jdesc;
 	struct gfs2_holder sd_journal_gh;
 	struct gfs2_holder sd_jinode_gh;
+	struct gfs2_glock *sd_jinode_gl;
 
 	struct gfs2_holder sd_sc_gh;
 	struct gfs2_holder sd_qc_gh;
@@ -829,6 +836,7 @@ struct gfs2_sbd {
 	struct bio *sd_log_bio;
 	wait_queue_head_t sd_log_flush_wait;
 	int sd_log_error; /* First log error */
+	wait_queue_head_t sd_withdraw_wait;
 
 	atomic_t sd_reserving_log;
 	wait_queue_head_t sd_reserving_log_wait;
@@ -852,6 +860,7 @@ struct gfs2_sbd {
 
 	unsigned long sd_last_warning;
 	struct dentry *debugfs_dir;    /* debugfs directory */
+	unsigned long sd_glock_dqs_held;
 };
 
 static inline void gfs2_glstats_inc(struct gfs2_glock *gl, int which)
diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c
index 57fdf53d2246..9f2b5609f225 100644
--- a/fs/gfs2/lock_dlm.c
+++ b/fs/gfs2/lock_dlm.c
@@ -16,6 +16,8 @@
 
 #include "incore.h"
 #include "glock.h"
+#include "glops.h"
+#include "recovery.h"
 #include "util.h"
 #include "sys.h"
 #include "trace_gfs2.h"
@@ -124,6 +126,8 @@ static void gdlm_ast(void *arg)
 
 	switch (gl->gl_lksb.sb_status) {
 	case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */
+		if (gl->gl_ops->go_free)
+			gl->gl_ops->go_free(gl);
 		gfs2_glock_free(gl);
 		return;
 	case -DLM_ECANCEL: /* Cancel while getting lock */
@@ -323,6 +327,7 @@ static void gdlm_cancel(struct gfs2_glock *gl)
 /*
  * dlm/gfs2 recovery coordination using dlm_recover callbacks
  *
+ *  0. gfs2 checks for another cluster node withdraw, needing journal replay
  *  1. dlm_controld sees lockspace members change
  *  2. dlm_controld blocks dlm-kernel locking activity
  *  3. dlm_controld within dlm-kernel notifies gfs2 (recover_prep)
@@ -571,6 +576,28 @@ static int control_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags)
 			 &ls->ls_control_lksb, "control_lock");
 }
 
+/**
+ * remote_withdraw - react to a node withdrawing from the file system
+ * @sdp: The superblock
+ */
+static void remote_withdraw(struct gfs2_sbd *sdp)
+{
+	struct gfs2_jdesc *jd;
+	int ret = 0, count = 0;
+
+	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
+		if (jd->jd_jid == sdp->sd_lockstruct.ls_jid)
+			continue;
+		ret = gfs2_recover_journal(jd, true);
+		if (ret)
+			break;
+		count++;
+	}
+
+	/* Now drop the additional reference we acquired */
+	fs_err(sdp, "Journals checked: %d, ret = %d.\n", count, ret);
+}
+
 static void gfs2_control_func(struct work_struct *work)
 {
 	struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_control_work.work);
@@ -581,6 +608,13 @@ static void gfs2_control_func(struct work_struct *work)
 	int recover_size;
 	int i, error;
 
+	/* First check for other nodes that may have done a withdraw. */
+	if (test_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags)) {
+		remote_withdraw(sdp);
+		clear_bit(SDF_REMOTE_WITHDRAW, &sdp->sd_flags);
+		return;
+	}
+
 	spin_lock(&ls->ls_recover_spin);
 	/*
 	 * No MOUNT_DONE means we're still mounting; control_mount()
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index 0c3772974030..6bdbdb9b3cd5 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -251,7 +251,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
 	struct buffer_head *bh, *bhs[2];
 	int num = 0;
 
-	if (unlikely(gfs2_withdrawn(sdp))) {
+	if (unlikely(gfs2_withdrawn(sdp)) && gl != sdp->sd_jinode_gl) {
 		*bhp = NULL;
 		return -EIO;
 	}
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index d19ee57c99ce..a38d153f1815 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -652,7 +652,8 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 
 		error = gfs2_glock_nq_num(sdp, sdp->sd_lockstruct.ls_jid,
 					  &gfs2_journal_glops,
-					  LM_ST_EXCLUSIVE, LM_FLAG_NOEXP,
+					  LM_ST_EXCLUSIVE,
+					  LM_FLAG_NOEXP | GL_NOCACHE,
 					  &sdp->sd_journal_gh);
 		if (error) {
 			fs_err(sdp, "can't acquire journal glock: %d\n", error);
@@ -660,6 +661,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 		}
 
 		ip = GFS2_I(sdp->sd_jdesc->jd_inode);
+		sdp->sd_jinode_gl = ip->i_gl;
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED,
 					   LM_FLAG_NOEXP | GL_EXACT | GL_NOCACHE,
 					   &sdp->sd_jinode_gh);
@@ -720,10 +722,13 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 	return 0;
 
 fail_jinode_gh:
-	if (!sdp->sd_args.ar_spectator)
+	/* A withdraw may have done dq/uninit so now we need to check it */
+	if (!sdp->sd_args.ar_spectator &&
+	    gfs2_holder_initialized(&sdp->sd_jinode_gh))
 		gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
 fail_journal_gh:
-	if (!sdp->sd_args.ar_spectator)
+	if (!sdp->sd_args.ar_spectator &&
+	    gfs2_holder_initialized(&sdp->sd_journal_gh))
 		gfs2_glock_dq_uninit(&sdp->sd_journal_gh);
 fail_jindex:
 	gfs2_jindex_free(sdp);
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index dbe87b2b55af..43ffe5997098 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -1541,6 +1541,8 @@ int gfs2_quotad(void *data)
 
 	while (!kthread_should_stop()) {
 
+		if (gfs2_withdrawn(sdp))
+			goto bypass;
 		/* Update the master statfs file */
 		if (sdp->sd_statfs_force_sync) {
 			int error = gfs2_statfs_sync(sdp->sd_vfs, 0);
@@ -1561,6 +1563,7 @@ int gfs2_quotad(void *data)
 
 		try_to_freeze();
 
+bypass:
 		t = min(quotad_timeo, statfs_timeo);
 
 		prepare_to_wait(&sdp->sd_quota_wait, &wait, TASK_INTERRUPTIBLE);
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index 68cc7c291a81..782463d64ee9 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -171,9 +171,13 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
 		goto fail_threads;
 
 	j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
+	if (gfs2_withdrawn(sdp)) {
+		error = -EIO;
+		goto fail;
+	}
 
 	error = gfs2_find_jhead(sdp->sd_jdesc, &head, false);
-	if (error)
+	if (error || gfs2_withdrawn(sdp))
 		goto fail;
 
 	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
@@ -187,7 +191,7 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
 	gfs2_log_pointers_init(sdp, head.lh_blkno);
 
 	error = gfs2_quota_init(sdp);
-	if (error)
+	if (error || gfs2_withdrawn(sdp))
 		goto fail;
 
 	set_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
@@ -599,34 +603,63 @@ static void gfs2_dirty_inode(struct inode *inode, int flags)
 int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
 {
 	struct gfs2_holder freeze_gh;
-	int error;
-
-	error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, GL_NOCACHE,
-				   &freeze_gh);
-	if (error && !gfs2_withdrawn(sdp))
-		return error;
+	int error = 0;
+	int log_write_allowed = test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
+
+	gfs2_holder_mark_uninitialized(&freeze_gh);
+	if (sdp->sd_freeze_gl &&
+	    !gfs2_glock_is_locked_by_me(sdp->sd_freeze_gl)) {
+		if (!log_write_allowed) {
+			error = gfs2_glock_nq_init(sdp->sd_freeze_gl,
+						   LM_ST_SHARED, GL_NOCACHE |
+						   LM_FLAG_TRY, &freeze_gh);
+			if (error == GLR_TRYFAILED)
+				error = 0;
+		} else {
+			error = gfs2_glock_nq_init(sdp->sd_freeze_gl,
+						   LM_ST_SHARED, GL_NOCACHE,
+						   &freeze_gh);
+			if (error && !gfs2_withdrawn(sdp))
+				return error;
+		}
+	}
 
 	flush_workqueue(gfs2_delete_workqueue);
-	if (sdp->sd_quotad_process)
+	if (!log_write_allowed && current == sdp->sd_quotad_process)
+		fs_warn(sdp, "The quotad daemon is withdrawing.\n");
+	else if (sdp->sd_quotad_process)
 		kthread_stop(sdp->sd_quotad_process);
 	sdp->sd_quotad_process = NULL;
-	if (sdp->sd_logd_process)
+
+	if (!log_write_allowed && current == sdp->sd_logd_process)
+		fs_warn(sdp, "The logd daemon is withdrawing.\n");
+	else if (sdp->sd_logd_process)
 		kthread_stop(sdp->sd_logd_process);
 	sdp->sd_logd_process = NULL;
 
-	gfs2_quota_sync(sdp->sd_vfs, 0);
-	gfs2_statfs_sync(sdp->sd_vfs, 0);
-
-	gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_SHUTDOWN |
-		       GFS2_LFC_MAKE_FS_RO);
-	wait_event(sdp->sd_reserving_log_wait, atomic_read(&sdp->sd_reserving_log) == 0);
-	gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_blks_free) == sdp->sd_jdesc->jd_blocks);
+	if (log_write_allowed) {
+		gfs2_quota_sync(sdp->sd_vfs, 0);
+		gfs2_statfs_sync(sdp->sd_vfs, 0);
 
+		gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_SHUTDOWN |
+			       GFS2_LFC_MAKE_FS_RO);
+		wait_event(sdp->sd_reserving_log_wait,
+			   atomic_read(&sdp->sd_reserving_log) == 0);
+		gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_blks_free) ==
+				 sdp->sd_jdesc->jd_blocks);
+	} else {
+		wait_event_timeout(sdp->sd_reserving_log_wait,
+				   atomic_read(&sdp->sd_reserving_log) == 0,
+				   HZ * 5);
+	}
 	if (gfs2_holder_initialized(&freeze_gh))
 		gfs2_glock_dq_uninit(&freeze_gh);
 
 	gfs2_quota_cleanup(sdp);
 
+	if (!log_write_allowed)
+		sdp->sd_vfs->s_flags |= SB_RDONLY;
+
 	return error;
 }
 
@@ -677,8 +710,10 @@ static void gfs2_put_super(struct super_block *sb)
 	gfs2_glock_put(sdp->sd_freeze_gl);
 
 	if (!sdp->sd_args.ar_spectator) {
-		gfs2_glock_dq_uninit(&sdp->sd_journal_gh);
-		gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
+		if (gfs2_holder_initialized(&sdp->sd_journal_gh))
+			gfs2_glock_dq_uninit(&sdp->sd_journal_gh);
+		if (gfs2_holder_initialized(&sdp->sd_jinode_gh))
+			gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
 		gfs2_glock_dq_uninit(&sdp->sd_sc_gh);
 		gfs2_glock_dq_uninit(&sdp->sd_qc_gh);
 		iput(sdp->sd_sc_inode);
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 86965e6089c6..efcaad2fc5b8 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -11,11 +11,14 @@
 #include <linux/buffer_head.h>
 #include <linux/crc32.h>
 #include <linux/gfs2_ondisk.h>
+#include <linux/delay.h>
 #include <linux/uaccess.h>
 
 #include "gfs2.h"
 #include "incore.h"
 #include "glock.h"
+#include "glops.h"
+#include "log.h"
 #include "lops.h"
 #include "recovery.h"
 #include "rgrp.h"
@@ -78,6 +81,144 @@ int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
 	return error;
 }
 
+static void signal_our_withdraw(struct gfs2_sbd *sdp)
+{
+	struct gfs2_glock *gl = sdp->sd_live_gh.gh_gl;
+	int ret = 0;
+	int tries;
+
+	if (test_bit(SDF_NORECOVERY, &sdp->sd_flags))
+		return;
+
+	/* Prevent any glock dq until withdraw recovery is complete */
+	set_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
+	/*
+	 * Don't tell dlm we're bailing until we have no more buffers in the
+	 * wind. If journal had an IO error, the log code should just purge
+	 * the outstanding buffers rather than submitting new IO. Making the
+	 * file system read-only will flush the journal, etc.
+	 *
+	 * During a normal unmount, gfs2_make_fs_ro calls gfs2_log_shutdown
+	 * which clears SDF_JOURNAL_LIVE. In a withdraw, we must not write
+	 * any UNMOUNT log header, so we can't call gfs2_log_shutdown, and
+	 * therefore we need to clear SDF_JOURNAL_LIVE manually.
+	 */
+	clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
+	if (sdp->sd_vfs && !sb_rdonly(sdp->sd_vfs))
+		ret = gfs2_make_fs_ro(sdp);
+
+	/*
+	 * Drop the glock for our journal so another node can recover it.
+	 */
+	if (gfs2_holder_initialized(&sdp->sd_journal_gh)) {
+		gfs2_glock_dq_wait(&sdp->sd_journal_gh);
+		gfs2_holder_uninit(&sdp->sd_journal_gh);
+	}
+	sdp->sd_jinode_gh.gh_flags |= GL_NOCACHE;
+	gfs2_glock_dq(&sdp->sd_jinode_gh);
+	if (test_bit(SDF_FS_FROZEN, &sdp->sd_flags)) {
+		/* Make sure gfs2_unfreeze works if partially-frozen */
+		flush_workqueue(gfs2_freeze_wq);
+		atomic_set(&sdp->sd_freeze_state, SFS_FROZEN);
+		thaw_super(sdp->sd_vfs);
+	} else {
+		wait_on_bit(&gl->gl_flags, GLF_DEMOTE, TASK_UNINTERRUPTIBLE);
+	}
+
+	/*
+	 * holder_uninit to force glock_put, to force dlm to let go
+	 */
+	gfs2_holder_uninit(&sdp->sd_jinode_gh);
+
+	/*
+	 * Note: We need to be careful here:
+	 * Func gfs2_jindex_free calls iput on jd_inode, which will evict it.
+	 * The evict, in turn, will dequeue its glock, but the glock dq will
+	 * wait for the withdraw unless we have exception code in glock_dq.
+	 */
+	gfs2_jindex_free(sdp);
+	/*
+	 * Wait until the journal inode's glock is freed. This allows try locks
+	 * on other nodes to be successful, otherwise we remain the owner of
+	 * the glock as far as dlm is concerned.
+	 */
+	if (gl->gl_ops->go_free) {
+		set_bit(GLF_FREEING, &gl->gl_flags);
+		wait_on_bit(&gl->gl_flags, GLF_FREEING, TASK_UNINTERRUPTIBLE);
+	}
+
+	if (sdp->sd_lockstruct.ls_ops->lm_lock == NULL) { /* lock_nolock */
+		clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
+		goto skip_recovery;
+	}
+	/*
+	 * Dequeue the "live" glock, but keep a reference so it's never freed.
+	 */
+	gfs2_glock_hold(gl);
+	gfs2_glock_dq_wait(&sdp->sd_live_gh);
+	/*
+	 * We enqueue the "live" glock in EX so that all other nodes
+	 * get a demote request and act on it. We don't really want the
+	 * lock in EX, so we send a "try" lock with 1CB to produce a callback.
+	 */
+	fs_warn(sdp, "Requesting recovery of jid %d.\n",
+		sdp->sd_lockstruct.ls_jid);
+	gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB | LM_FLAG_NOEXP,
+			   &sdp->sd_live_gh);
+	msleep(GL_GLOCK_MAX_HOLD);
+	/*
+	 * This will likely fail in a cluster, but succeed standalone:
+	 */
+	ret = gfs2_glock_nq(&sdp->sd_live_gh);
+
+	/*
+	 * If we actually got the "live" lock in EX mode, there are no other
+	 * nodes available to replay our journal. So we try to replay it
+	 * ourselves. We hold the "live" glock to prevent other mounters
+	 * during recovery, then just dequeue it and reacquire it in our
+	 * normal SH mode. Just in case the problem that caused us to
+	 * withdraw prevents us from recovering our journal (e.g. io errors
+	 * and such) we still check if the journal is clean before proceeding
+	 * but we may wait forever until another mounter does the recovery.
+	 */
+	if (ret == 0) {
+		fs_warn(sdp, "No other mounters found. Trying to recover our "
+			"own journal jid %d.\n", sdp->sd_lockstruct.ls_jid);
+		if (gfs2_recover_journal(sdp->sd_jdesc, 1))
+			fs_warn(sdp, "Unable to recover our journal jid %d.\n",
+				sdp->sd_lockstruct.ls_jid);
+		gfs2_glock_dq_wait(&sdp->sd_live_gh);
+		gfs2_holder_reinit(LM_ST_SHARED, LM_FLAG_NOEXP | GL_EXACT,
+				   &sdp->sd_live_gh);
+		gfs2_glock_nq(&sdp->sd_live_gh);
+	}
+
+	gfs2_glock_queue_put(gl); /* drop the extra reference we acquired */
+	clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
+
+	/*
+	 * Now wait until recovery complete.
+	 */
+	for (tries = 0; tries < 10; tries++) {
+		ret = check_journal_clean(sdp, sdp->sd_jdesc);
+		if (!ret)
+			break;
+		msleep(HZ);
+		fs_warn(sdp, "Waiting for journal recovery jid %d.\n",
+			sdp->sd_lockstruct.ls_jid);
+	}
+skip_recovery:
+	if (!ret)
+		fs_warn(sdp, "Journal recovery complete for jid %d.\n",
+			sdp->sd_lockstruct.ls_jid);
+	else
+		fs_warn(sdp, "Journal recovery skipped for %d until next "
+			"mount.\n", sdp->sd_lockstruct.ls_jid);
+	fs_warn(sdp, "Glock dequeues delayed: %lu\n", sdp->sd_glock_dqs_held);
+	sdp->sd_glock_dqs_held = 0;
+	wake_up_bit(&sdp->sd_flags, SDF_WITHDRAW_RECOVERY);
+}
+
 void gfs2_lm(struct gfs2_sbd *sdp, const char *fmt, ...)
 {
 	struct va_format vaf;
@@ -100,13 +241,23 @@ int gfs2_withdraw(struct gfs2_sbd *sdp)
 	const struct lm_lockops *lm = ls->ls_ops;
 
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW &&
-	    test_and_set_bit(SDF_WITHDRAWN, &sdp->sd_flags))
-		return 0;
+	    test_and_set_bit(SDF_WITHDRAWN, &sdp->sd_flags)) {
+		if (!test_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags))
+			return -1;
+
+		wait_on_bit(&sdp->sd_flags, SDF_WITHDRAW_IN_PROG,
+			    TASK_UNINTERRUPTIBLE);
+		return -1;
+	}
+
+	set_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags);
 
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_WITHDRAW) {
 		fs_err(sdp, "about to withdraw this file system\n");
 		BUG_ON(sdp->sd_args.ar_debug);
 
+		signal_our_withdraw(sdp);
+
 		kobject_uevent(&sdp->sd_kobj, KOBJ_OFFLINE);
 
 		if (!strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm"))
@@ -117,8 +268,11 @@ int gfs2_withdraw(struct gfs2_sbd *sdp)
 			lm->lm_unmount(sdp);
 		}
 		set_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags);
-		fs_err(sdp, "withdrawn\n");
+		fs_err(sdp, "File system withdrawn\n");
 		dump_stack();
+		clear_bit(SDF_WITHDRAW_IN_PROG, &sdp->sd_flags);
+		smp_mb__after_atomic();
+		wake_up_bit(&sdp->sd_flags, SDF_WITHDRAW_IN_PROG);
 	}
 
 	if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC)
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 15/25] gfs2: fix infinite loop when checking ail item count before go_inval

[Bob Peterson]

Before this patch, the rgrp_go_inval and inode_go_inval functions each
checked if there were any items left on the ail count (by way of a
count), and if so, did a withdraw. But the withdraw code now uses
glocks when changing the file system to read-only status. So we can
not have glock functions withdrawing or a hang will likely result:
The glocks can't be serviced by the work_func if the work_func is
busy doing its own withdraw.

This patch removes the checks from the go_inval functions and adds
a centralized check in do_xmote to warn about the problem and not
withdraw, but flag the error so it's eventually caught when the logd
daemon eventually runs.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c | 17 +++++++++++++++--
 fs/gfs2/glops.c |  3 ---
 2 files changed, 15 insertions(+), 5 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 7094ef07003c..233189d47afc 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -602,9 +602,22 @@ __acquires(&gl->gl_lockref.lock)
 	spin_unlock(&gl->gl_lockref.lock);
 	if (glops->go_sync)
 		glops->go_sync(gl);
-	if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags))
+	if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags)) {
+		/*
+		 * The call to go_sync should have cleared out the ail list.
+		 * If there are still items, we have a problem. We ought to
+		 * withdraw, but we can't because the withdraw code also uses
+		 * glocks. Warn about the error, dump the glock, then fall
+		 * through and wait for logd to do the withdraw for us.
+		 */
+		if ((atomic_read(&gl->gl_ail_count) != 0) &&
+		    (!cmpxchg(&sdp->sd_log_error, 0, -EIO))) {
+			gfs2_assert_warn(sdp, !atomic_read(&gl->gl_ail_count));
+			gfs2_dump_glock(NULL, gl, true);
+		}
 		glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA);
-	clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags);
+		clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags);
+	}
 
 	gfs2_glock_hold(gl);
 	if (sdp->sd_lockstruct.ls_ops->lm_lock)	{
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 7cfacbe35e59..b58924482d9a 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -188,7 +188,6 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
 		gfs2_rgrp_brelse(rgd);
 
 	WARN_ON_ONCE(!(flags & DIO_METADATA));
-	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
 	truncate_inode_pages_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
 
 	if (rgd)
@@ -288,8 +287,6 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags)
 {
 	struct gfs2_inode *ip = gfs2_glock2inode(gl);
 
-	gfs2_assert_withdraw(gl->gl_name.ln_sbd, !atomic_read(&gl->gl_ail_count));
-
 	if (flags & DIO_METADATA) {
 		struct address_space *mapping = gfs2_glock2aspace(gl);
 		truncate_inode_pages(mapping, 0);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 16/25] gfs2: Add verbose option to check_journal_clean

[Bob Peterson]

Before this patch, function check_journal_clean would give messages
related to journal recovery. That's fine for mount time, but when a
node withdraws and forces replay that way, we don't want all those
distracting and misleading messages. This patch adds a new parameter
to make those messages optional.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/ops_fstype.c |  2 +-
 fs/gfs2/util.c       | 23 ++++++++++++++++-------
 fs/gfs2/util.h       |  4 +++-
 3 files changed, 20 insertions(+), 9 deletions(-)

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index a38d153f1815..b4d717ac4526 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -692,7 +692,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 			struct gfs2_jdesc *jd = gfs2_jdesc_find(sdp, x);
 
 			if (sdp->sd_args.ar_spectator) {
-				error = check_journal_clean(sdp, jd);
+				error = check_journal_clean(sdp, jd, true);
 				if (error)
 					goto fail_jinode_gh;
 				continue;
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index efcaad2fc5b8..37a89916c755 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -46,7 +46,8 @@ void gfs2_assert_i(struct gfs2_sbd *sdp)
  *
  * Returns: 0 if the journal is clean or locked, else an error
  */
-int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
+int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd,
+			bool verbose)
 {
 	int error;
 	struct gfs2_holder j_gh;
@@ -57,23 +58,31 @@ int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd)
 	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_NOEXP |
 				   GL_EXACT | GL_NOCACHE, &j_gh);
 	if (error) {
-		fs_err(sdp, "Error locking journal for spectator mount.\n");
+		if (verbose)
+			fs_err(sdp, "Error %d locking journal for spectator "
+			       "mount.\n", error);
 		return -EPERM;
 	}
 	error = gfs2_jdesc_check(jd);
 	if (error) {
-		fs_err(sdp, "Error checking journal for spectator mount.\n");
+		if (verbose)
+			fs_err(sdp, "Error checking journal for spectator "
+			       "mount.\n");
 		goto out_unlock;
 	}
 	error = gfs2_find_jhead(jd, &head, false);
 	if (error) {
-		fs_err(sdp, "Error parsing journal for spectator mount.\n");
+		if (verbose)
+			fs_err(sdp, "Error parsing journal for spectator "
+			       "mount.\n");
 		goto out_unlock;
 	}
 	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
 		error = -EPERM;
-		fs_err(sdp, "jid=%u: Journal is dirty, so the first mounter "
-		       "must not be a spectator.\n", jd->jd_jid);
+		if (verbose)
+			fs_err(sdp, "jid=%u: Journal is dirty, so the first "
+			       "mounter must not be a spectator.\n",
+			       jd->jd_jid);
 	}
 
 out_unlock:
@@ -200,7 +209,7 @@ static void signal_our_withdraw(struct gfs2_sbd *sdp)
 	 * Now wait until recovery complete.
 	 */
 	for (tries = 0; tries < 10; tries++) {
-		ret = check_journal_clean(sdp, sdp->sd_jdesc);
+		ret = check_journal_clean(sdp, sdp->sd_jdesc, false);
 		if (!ret)
 			break;
 		msleep(HZ);
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index cf613497a20e..97117a766bde 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -136,7 +136,9 @@ static inline void gfs2_metatype_set(struct buffer_head *bh, u16 type,
 
 int gfs2_io_error_i(struct gfs2_sbd *sdp, const char *function,
 		    char *file, unsigned int line);
-int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd);
+
+extern int check_journal_clean(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd,
+			       bool verbose);
 
 #define gfs2_io_error(sdp) \
 gfs2_io_error_i((sdp), __func__, __FILE__, __LINE__);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 17/25] gfs2: Issue revokes more intelligently

[Bob Peterson]

Before this patch, function gfs2_write_revokes would call
gfs2_ail1_empty, then traverse the sd_ail1_list looking for
transactions that had bds which were no longer queued to a glock.
And if it found some, it would try to issue revokes for them, up to
a predetermined maximum. There were two problems with how it did
this. First was the fact that gfs2_ail1_empty moves transactions
which have nothing remaining on the ail1 list from the sd_ail1_list
to the sd_ail2_list, thus making its traversal of sd_ail1_list
miss them completely, and therefore, never issue revokes for them.
Second was the fact that there were three traversals (or partial
traversals) of the sd_ail1_list, each of which took and then
released the sd_ail_lock lock: First inside gfs2_ail1_empty,
second to determine if there are any revokes to be issued, and
third to actually issue them. All this taking and releasing of the
sd_ail_lock meant other processes could modify the lists and the
conditions in which we're working.

This patch simplies the whole process by adding a new parameter
to function gfs2_ail1_empty, max_revokes. For normal calls, this
is passed in as 0, meaning we don't want to issue any revokes.
For function gfs2_write_revokes, we pass in the maximum number
of revokes we can, thus allowing gfs2_ail1_empty to add the
revokes where needed. This simplies the code, allows for a single
holding of the sd_ail_lock, and allows gfs2_ail1_empty to add
revokes for all the necessary bd items without missing any.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/log.c | 74 +++++++++++++++++++++++++--------------------------
 1 file changed, 36 insertions(+), 38 deletions(-)

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 584bb7ce15bf..578c1e0cd415 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -191,11 +191,13 @@ static void gfs2_ail1_start(struct gfs2_sbd *sdp)
 /**
  * gfs2_ail1_empty_one - Check whether or not a trans in the AIL has been synced
  * @sdp: the filesystem
- * @ai: the AIL entry
+ * @tr: the transaction
+ * @max_revokes: If nonzero, issue revokes for the bd items for written buffers
  *
  */
 
-static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
+static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
+				int *max_revokes)
 {
 	struct gfs2_bufdata *bd, *s;
 	struct buffer_head *bh;
@@ -220,6 +222,17 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
 			gfs2_io_error_bh(sdp, bh);
 			gfs2_withdraw_delayed(sdp);
 		}
+		/*
+		 * If we have space for revokes and the bd is no longer on any
+		 * buf list, we can just add a revoke for it immediately and
+		 * avoid having to put it on the ail2 list, where it would need
+		 * to be revoked later.
+		 */
+		if (*max_revokes && list_empty(&bd->bd_list)) {
+			gfs2_add_revoke(sdp, bd);
+			(*max_revokes)--;
+			continue;
+		}
 		list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list);
 	}
 }
@@ -227,11 +240,12 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
 /**
  * gfs2_ail1_empty - Try to empty the ail1 lists
  * @sdp: The superblock
+ * @max_revokes: If non-zero, add revokes where appropriate
  *
  * Tries to empty the ail1 lists, starting with the oldest first
  */
 
-static int gfs2_ail1_empty(struct gfs2_sbd *sdp)
+static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int max_revokes)
 {
 	struct gfs2_trans *tr, *s;
 	int oldest_tr = 1;
@@ -239,7 +253,7 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp)
 
 	spin_lock(&sdp->sd_ail_lock);
 	list_for_each_entry_safe_reverse(tr, s, &sdp->sd_ail1_list, tr_list) {
-		gfs2_ail1_empty_one(sdp, tr);
+		gfs2_ail1_empty_one(sdp, tr, &max_revokes);
 		if (list_empty(&tr->tr_ail1_list) && oldest_tr)
 			list_move(&tr->tr_list, &sdp->sd_ail2_list);
 		else
@@ -627,27 +641,24 @@ void gfs2_glock_remove_revoke(struct gfs2_glock *gl)
 	}
 }
 
+/**
+ * gfs2_write_revokes - Add as many revokes to the system transaction as we can
+ * @sdp: The GFS2 superblock
+ *
+ * Our usual strategy is to defer writing revokes as much as we can in the hope
+ * that we'll eventually overwrite the journal, which will make those revokes
+ * go away.  This changes when we flush the log: at that point, there will
+ * likely be some left-over space in the last revoke block of that transaction.
+ * We can fill that space with additional revokes for blocks that have already
+ * been written back.  This will basically come at no cost now, and will save
+ * us from having to keep track of those blocks on the AIL2 list later.
+ */
 void gfs2_write_revokes(struct gfs2_sbd *sdp)
 {
-	struct gfs2_trans *tr;
-	struct gfs2_bufdata *bd, *tmp;
-	int have_revokes = 0;
+	/* number of revokes we still have room for */
 	int max_revokes = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_log_descriptor)) / sizeof(u64);
 
-	gfs2_ail1_empty(sdp);
-	spin_lock(&sdp->sd_ail_lock);
-	list_for_each_entry_reverse(tr, &sdp->sd_ail1_list, tr_list) {
-		list_for_each_entry(bd, &tr->tr_ail2_list, bd_ail_st_list) {
-			if (list_empty(&bd->bd_list)) {
-				have_revokes = 1;
-				goto done;
-			}
-		}
-	}
-done:
-	spin_unlock(&sdp->sd_ail_lock);
-	if (have_revokes == 0)
-		return;
+	gfs2_log_lock(sdp);
 	while (sdp->sd_log_num_revoke > max_revokes)
 		max_revokes += (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / sizeof(u64);
 	max_revokes -= sdp->sd_log_num_revoke;
@@ -658,20 +669,7 @@ void gfs2_write_revokes(struct gfs2_sbd *sdp)
 		if (!sdp->sd_log_blks_reserved)
 			atomic_dec(&sdp->sd_log_blks_free);
 	}
-	gfs2_log_lock(sdp);
-	spin_lock(&sdp->sd_ail_lock);
-	list_for_each_entry_reverse(tr, &sdp->sd_ail1_list, tr_list) {
-		list_for_each_entry_safe(bd, tmp, &tr->tr_ail2_list, bd_ail_st_list) {
-			if (max_revokes == 0)
-				goto out_of_blocks;
-			if (!list_empty(&bd->bd_list))
-				continue;
-			gfs2_add_revoke(sdp, bd);
-			max_revokes--;
-		}
-	}
-out_of_blocks:
-	spin_unlock(&sdp->sd_ail_lock);
+	gfs2_ail1_empty(sdp, max_revokes);
 	gfs2_log_unlock(sdp);
 
 	if (!sdp->sd_log_num_revoke) {
@@ -870,7 +868,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 			for (;;) {
 				gfs2_ail1_start(sdp);
 				gfs2_ail1_wait(sdp);
-				if (gfs2_ail1_empty(sdp))
+				if (gfs2_ail1_empty(sdp, 0))
 					break;
 			}
 			if (gfs2_withdrawn(sdp))
@@ -1040,7 +1038,7 @@ int gfs2_logd(void *data)
 
 		did_flush = false;
 		if (gfs2_jrnl_flush_reqd(sdp) || t == 0) {
-			gfs2_ail1_empty(sdp);
+			gfs2_ail1_empty(sdp, 0);
 			gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
 				       GFS2_LFC_LOGD_JFLUSH_REQD);
 			did_flush = true;
@@ -1049,7 +1047,7 @@ int gfs2_logd(void *data)
 		if (gfs2_ail_flush_reqd(sdp)) {
 			gfs2_ail1_start(sdp);
 			gfs2_ail1_wait(sdp);
-			gfs2_ail1_empty(sdp);
+			gfs2_ail1_empty(sdp, 0);
 			gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
 				       GFS2_LFC_LOGD_AIL_FLUSH_REQD);
 			did_flush = true;
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 18/25] gfs2: Prepare to withdraw as soon as an IO error occurs in log write

[Bob Peterson]

Before this patch, function gfs2_end_log_write would detect any IO
errors writing to the journal and put out an appropriate message,
but it never set a withdrawing condition. Eventually, the log daemon
would see the error and determine it was time to withdraw, but in
the meantime, other processes could continue running as if nothing
bad ever happened. The biggest consequence is that __gfs2_glock_put
would BUG() when it saw that there were still unwritten items.

This patch sets the WITHDRAWING status as soon as an IO error is
detected, and that way, the BUG will be avoided so the file system
can be properly withdrawn and unmounted.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/lops.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 45fa5cd88feb..3862c1f5c463 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -206,6 +206,9 @@ static void gfs2_end_log_write(struct bio *bio)
 		if (!cmpxchg(&sdp->sd_log_error, 0, (int)bio->bi_status))
 			fs_err(sdp, "Error %d writing to journal, jid=%u\n",
 			       bio->bi_status, sdp->sd_jdesc->jd_jid);
+		gfs2_withdraw_delayed(sdp);
+		/* prevent more writes to the journal */
+		clear_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);
 		wake_up(&sdp->sd_logd_waitq);
 	}
 
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 19/25] gfs2: Check for log write errors before telling dlm to unlock

[Bob Peterson]

Before this patch, function do_xmote just assumed all the writes
submitted to the journal were finished and successful, and it
called the go_unlock function to release the dlm lock. But if
they're not, and a revoke failed to make its way to the journal,
a journal replay on another node will cause corruption if we
let the go_inval function continue and tell dlm to release the
glock to another node. This patch adds a couple checks for errors
in do_xmote after the calls to go_sync and go_inval. If an error
is found, we cannot withdraw yet, because the withdraw itself
uses glocks to make the file system read-only. Instead, we flag
the error. Later, asserts should cause another node to replay
the journal before continuing, thus protecting rgrp and dinode
glocks and maintaining the integrity of the metadata. Note that
we only need to do this for journaled glocks. System glocks
should be able to progress even under withdrawn conditions.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c | 31 ++++++++++++++++++++++++++++---
 1 file changed, 28 insertions(+), 3 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 233189d47afc..70a97a40f3db 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -620,6 +620,32 @@ __acquires(&gl->gl_lockref.lock)
 	}
 
 	gfs2_glock_hold(gl);
+	/*
+	 * Check for an error encountered since we called go_sync and go_inval.
+	 * If so, we can't withdraw from the glock code because the withdraw
+	 * code itself uses glocks (see function signal_our_withdraw) to
+	 * change the mount to read-only. Most importantly, we must not call
+	 * dlm to unlock the glock until the journal is in a known good state
+	 * (after journal replay) otherwise other nodes may use the object
+	 * (rgrp or dinode) and then later, journal replay will corrupt the
+	 * file system. The best we can do here is wait for the logd daemon
+	 * to see sd_log_error and withdraw, and in the meantime, requeue the
+	 * work for later.
+	 *
+	 * However, if we're just unlocking the lock (say, for unmount, when
+	 * gfs2_gl_hash_clear calls clear_glock) and recovery is complete
+	 * then it's okay to tell dlm to unlock it.
+	 */
+	if (unlikely(sdp->sd_log_error && !gfs2_withdrawn(sdp)))
+		gfs2_withdraw_delayed(sdp);
+	if (glock_blocked_by_withdraw(gl)) {
+		if (target != LM_ST_UNLOCKED ||
+		    test_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags)) {
+			gfs2_glock_queue_work(gl, GL_GLOCK_DFT_HOLD);
+			goto out;
+		}
+	}
+
 	if (sdp->sd_lockstruct.ls_ops->lm_lock)	{
 		/* lock_dlm */
 		ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags);
@@ -628,8 +654,7 @@ __acquires(&gl->gl_lockref.lock)
 		    test_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags)) {
 			finish_xmote(gl, target);
 			gfs2_glock_queue_work(gl, 0);
-		}
-		else if (ret) {
+		} else if (ret) {
 			fs_err(sdp, "lm_lock ret %d\n", ret);
 			GLOCK_BUG_ON(gl, !gfs2_withdrawn(sdp));
 		}
@@ -637,7 +662,7 @@ __acquires(&gl->gl_lockref.lock)
 		finish_xmote(gl, target);
 		gfs2_glock_queue_work(gl, 0);
 	}
-
+out:
 	spin_lock(&gl->gl_lockref.lock);
 }
 
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 20/25] gfs2: Do log_flush in gfs2_ail_empty_gl even if ail list is empty

[Bob Peterson]

Before this patch, if gfs2_ail_empty_gl saw there was nothing on
the ail list, it would return and not flush the log. The problem
is that there could still be a revoke for the rgrp sitting on the
sd_log_le_revoke list that's been recently taken off the ail list.
But that revoke still needs to be written, and the rgrp_go_inval
still needs to call log_flush_wait to ensure the revokes are all
properly written to the journal before we relinquish control of
the glock to another node. If we give the glock to another node
before we have this knowledge, the node might crash and its journal
replayed, in which case the missing revoke would allow the journal
replay to replay the rgrp over top of the rgrp we already gave to
another node, thus overwriting its changes and corrupting the
file system.

This patch makes gfs2_ail_empty_gl still call gfs2_log_flush rather
than returning.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glops.c | 27 ++++++++++++++++++++++++++-
 fs/gfs2/log.c   |  2 +-
 fs/gfs2/log.h   |  1 +
 3 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index b58924482d9a..bbbcae8d853c 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -92,8 +92,32 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 	INIT_LIST_HEAD(&tr.tr_databuf);
 	tr.tr_revokes = atomic_read(&gl->gl_ail_count);
 
-	if (!tr.tr_revokes)
+	if (!tr.tr_revokes) {
+		bool have_revokes;
+		bool log_in_flight;
+
+		/*
+		 * We have nothing on the ail, but there could be revokes on
+		 * the sdp revoke queue, in which case, we still want to flush
+		 * the log and wait for it to finish.
+		 *
+		 * If the sdp revoke list is empty too, we might still have an
+		 * io outstanding for writing revokes, so we should wait for
+		 * it before returning.
+		 *
+		 * If none of these conditions are true, our revokes are all
+		 * flushed and we can return.
+		 */
+		gfs2_log_lock(sdp);
+		have_revokes = !list_empty(&sdp->sd_log_revokes);
+		log_in_flight = atomic_read(&sdp->sd_log_in_flight);
+		gfs2_log_unlock(sdp);
+		if (have_revokes)
+			goto flush;
+		if (log_in_flight)
+			log_flush_wait(sdp);
 		return;
+	}
 
 	/* A shortened, inline version of gfs2_trans_begin()
          * tr->alloced is not set since the transaction structure is
@@ -108,6 +132,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 	__gfs2_ail_flush(gl, 0, tr.tr_revokes);
 
 	gfs2_trans_end(sdp);
+flush:
 	gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
 		       GFS2_LFC_AIL_EMPTY_GL);
 }
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 578c1e0cd415..c37f81470792 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -537,7 +537,7 @@ static void log_pull_tail(struct gfs2_sbd *sdp, unsigned int new_tail)
 }
 
 
-static void log_flush_wait(struct gfs2_sbd *sdp)
+void log_flush_wait(struct gfs2_sbd *sdp)
 {
 	DEFINE_WAIT(wait);
 
diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h
index c0a65e5a126b..c1cd6ae17659 100644
--- a/fs/gfs2/log.h
+++ b/fs/gfs2/log.h
@@ -73,6 +73,7 @@ extern void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl,
 			   u32 type);
 extern void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *trans);
 extern void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc);
+extern void log_flush_wait(struct gfs2_sbd *sdp);
 
 extern int gfs2_logd(void *data);
 extern void gfs2_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 21/25] gfs2: Withdraw in gfs2_ail1_flush if write_cache_pages fails

[Bob Peterson]

Before this patch, function gfs2_ail1_start_one would return any
errors it received from write_cache_pages (except -EBUSY) but it did
not withdraw. Since function gfs2_ail1_flush just checks for the bad
return code and loops, the loop might potentially never end.
This patch adds some logic to allow it to exit the loop and withdraw
properly when errors are received from write_cache_pages.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/log.c | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index c37f81470792..ed80ef8e5c33 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -96,6 +96,7 @@ __acquires(&sdp->sd_ail_lock)
 	struct address_space *mapping;
 	struct gfs2_bufdata *bd, *s;
 	struct buffer_head *bh;
+	int ret = 0;
 
 	list_for_each_entry_safe_reverse(bd, s, &tr->tr_ail1_list, bd_ail_st_list) {
 		bh = bd->bd_bh;
@@ -128,14 +129,14 @@ __acquires(&sdp->sd_ail_lock)
 		if (!mapping)
 			continue;
 		spin_unlock(&sdp->sd_ail_lock);
-		generic_writepages(mapping, wbc);
+		ret = generic_writepages(mapping, wbc);
 		spin_lock(&sdp->sd_ail_lock);
-		if (wbc->nr_to_write <= 0)
+		if (ret || wbc->nr_to_write <= 0)
 			break;
-		return 1;
+		return -EBUSY;
 	}
 
-	return 0;
+	return ret;
 }
 
 
@@ -153,6 +154,7 @@ void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc)
 	struct list_head *head = &sdp->sd_ail1_list;
 	struct gfs2_trans *tr;
 	struct blk_plug plug;
+	int ret = 0;
 
 	trace_gfs2_ail_flush(sdp, wbc, 1);
 	blk_start_plug(&plug);
@@ -161,12 +163,16 @@ void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc)
 	list_for_each_entry_reverse(tr, head, tr_list) {
 		if (wbc->nr_to_write <= 0)
 			break;
-		if (gfs2_ail1_start_one(sdp, wbc, tr) && !gfs2_withdrawn(sdp))
-			goto restart;
+		ret = gfs2_ail1_start_one(sdp, wbc, tr);
+		if (ret) {
+			if (ret == -EBUSY)
+				goto restart;
+			break;
+		}
 	}
 	spin_unlock(&sdp->sd_ail_lock);
 	blk_finish_plug(&plug);
-	if (test_bit(SDF_WITHDRAWING, &sdp->sd_flags))
+	if (ret)
 		gfs2_withdraw(sdp);
 	trace_gfs2_ail_flush(sdp, wbc, 0);
 }
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 22/25] gfs2: drain the ail2 list after io errors

[Bob Peterson]

Before this patch, gfs2_logd continually tried to flush its journal
log, after the file system is withdrawn. We don't want to write anything
to the journal, lest we add corruption. Best course of action is to
drain the ail1 into the ail2 list (via gfs2_ail1_empty) then drain the
ail2 list with a new function, ail2_drain.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/log.c   | 71 ++++++++++++++++++++++++++++++++++++++++++-------
 fs/gfs2/trans.c |  4 +++
 2 files changed, 65 insertions(+), 10 deletions(-)

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index ed80ef8e5c33..010c319caade 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -299,20 +299,17 @@ static void gfs2_ail1_wait(struct gfs2_sbd *sdp)
 }
 
 /**
- * gfs2_ail2_empty_one - Check whether or not a trans in the AIL has been synced
- * @sdp: the filesystem
- * @ai: the AIL entry
- *
+ * gfs2_ail_empty_tr - empty one of the ail lists for a transaction
  */
 
-static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
+static void gfs2_ail_empty_tr(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
+			      struct list_head *head)
 {
-	struct list_head *head = &tr->tr_ail2_list;
 	struct gfs2_bufdata *bd;
 
 	while (!list_empty(head)) {
-		bd = list_entry(head->prev, struct gfs2_bufdata,
-				bd_ail_st_list);
+		bd = list_first_entry(head, struct gfs2_bufdata,
+				      bd_ail_st_list);
 		gfs2_assert(sdp, bd->bd_tr == tr);
 		gfs2_remove_from_ail(bd);
 	}
@@ -334,7 +331,7 @@ static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail)
 		if (!rm)
 			continue;
 
-		gfs2_ail2_empty_one(sdp, tr);
+		gfs2_ail_empty_tr(sdp, tr, &tr->tr_ail2_list);
 		list_del(&tr->tr_list);
 		gfs2_assert_warn(sdp, list_empty(&tr->tr_ail1_list));
 		gfs2_assert_warn(sdp, list_empty(&tr->tr_ail2_list));
@@ -801,6 +798,40 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags)
 		log_pull_tail(sdp, tail);
 }
 
+/**
+ * ail_drain - drain the ail lists after a withdraw
+ * @sdp: Pointer to GFS2 superblock
+ */
+static void ail_drain(struct gfs2_sbd *sdp)
+{
+	struct gfs2_trans *tr;
+
+	spin_lock(&sdp->sd_ail_lock);
+	/*
+	 * For transactions on the sd_ail1_list we need to drain both the
+	 * ail1 and ail2 lists. That's because function gfs2_ail1_start_one
+	 * (temporarily) moves items from its tr_ail1 list to tr_ail2 list
+	 * before revokes are sent for that block. Items on the sd_ail2_list
+	 * should have already gotten beyond that point, so no need.
+	 */
+	while (!list_empty(&sdp->sd_ail1_list)) {
+		tr = list_first_entry(&sdp->sd_ail1_list, struct gfs2_trans,
+				      tr_list);
+		gfs2_ail_empty_tr(sdp, tr, &tr->tr_ail1_list);
+		gfs2_ail_empty_tr(sdp, tr, &tr->tr_ail2_list);
+		list_del(&tr->tr_list);
+		kfree(tr);
+	}
+	while (!list_empty(&sdp->sd_ail2_list)) {
+		tr = list_first_entry(&sdp->sd_ail2_list, struct gfs2_trans,
+				      tr_list);
+		gfs2_ail_empty_tr(sdp, tr, &tr->tr_ail2_list);
+		list_del(&tr->tr_list);
+		kfree(tr);
+	}
+	spin_unlock(&sdp->sd_ail_lock);
+}
+
 /**
  * gfs2_log_flush - flush incore transaction(s)
  * @sdp: the filesystem
@@ -811,11 +842,18 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags)
 
 void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 {
-	struct gfs2_trans *tr;
+	struct gfs2_trans *tr = NULL;
 	enum gfs2_freeze_state state = atomic_read(&sdp->sd_freeze_state);
 
 	down_write(&sdp->sd_log_flush_lock);
 
+	/*
+	 * Do this check while holding the log_flush_lock to prevent new
+	 * buffers from being added to the ail via gfs2_pin()
+	 */
+	if (gfs2_withdrawn(sdp))
+		goto out;
+
 	/* Log might have been flushed while we waited for the flush lock */
 	if (gl && !test_bit(GLF_LFLUSH, &gl->gl_flags)) {
 		up_write(&sdp->sd_log_flush_lock);
@@ -843,8 +881,14 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 			sdp->sd_log_num_revoke == sdp->sd_log_committed_revoke);
 
 	gfs2_ordered_write(sdp);
+	if (gfs2_withdrawn(sdp))
+		goto out;
 	lops_before_commit(sdp, tr);
+	if (gfs2_withdrawn(sdp))
+		goto out;
 	gfs2_log_submit_bio(&sdp->sd_log_bio, REQ_OP_WRITE);
+	if (gfs2_withdrawn(sdp))
+		goto out;
 
 	if (sdp->sd_log_head != sdp->sd_log_flush_head) {
 		log_flush_wait(sdp);
@@ -854,6 +898,8 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		trace_gfs2_log_blocks(sdp, -1);
 		log_write_header(sdp, flags);
 	}
+	if (gfs2_withdrawn(sdp))
+		goto out;
 	lops_after_commit(sdp, tr);
 
 	gfs2_log_lock(sdp);
@@ -892,6 +938,11 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 	}
 
 out:
+	if (gfs2_withdrawn(sdp)) {
+		ail_drain(sdp); /* frees all transactions */
+		tr = NULL;
+	}
+
 	trace_gfs2_log_flush(sdp, 0, flags);
 	up_write(&sdp->sd_log_flush_lock);
 
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index a685637a5b55..ffe840505082 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -228,6 +228,10 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
 		fs_info(sdp, "GFS2:adding buf while frozen\n");
 		gfs2_assert_withdraw(sdp, 0);
 	}
+	if (unlikely(gfs2_withdrawn(sdp))) {
+		fs_info(sdp, "GFS2:adding buf while withdrawn! 0x%llx\n",
+			(unsigned long long)bd->bd_bh->b_blocknr);
+	}
 	gfs2_pin(sdp, bd->bd_bh);
 	mh->__pad0 = cpu_to_be64(0);
 	mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 23/25] gfs2: Don't demote a glock until its revokes are written

[Bob Peterson]

Before this patch, run_queue would demote glocks based on whether
there are any more holders. But if the glock has pending revokes that
haven't been written to the media, giving up the glock might end in
file system corruption if the revokes never get written due to
io errors, node crashes and fences, etc. In that case, another node
will replay the metadata blocks associated with the glock, but
because the revoke was never written, it could replay that block
even though the glock had since been granted to another node who
might have made changes.

This patch changes the logic in run_queue so that it never demotes
a glock until its count of pending revokes reaches zero.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 70a97a40f3db..2b43a8488aec 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -708,6 +708,9 @@ __acquires(&gl->gl_lockref.lock)
 			goto out_unlock;
 		if (nonblock)
 			goto out_sched;
+		smp_mb();
+		if (atomic_read(&gl->gl_revokes) != 0)
+			goto out_sched;
 		set_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
 		GLOCK_BUG_ON(gl, gl->gl_demote_state == LM_ST_EXCLUSIVE);
 		gl->gl_target = gl->gl_demote_state;
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 24/25] gfs2: Do proper error checking for go_sync family of glops functions

[Bob Peterson]

Before this patch, function do_xmote would try to sync out the glock
dirty data by calling the appropriate glops function XXX_go_sync()
but it did not check for a good return code. If the sync was not
possible due to an io error or whatever, do_xmote would continue on
and call go_inval and release the glock to other cluster nodes.
When those nodes go to replay the journal, they may already be holding
glocks for the journal records that should have been synced, but were
not due to the ignored error.

This patch introduces proper error code checking to the go_sync
family of glops functions.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/glock.c  | 16 ++++++++++++++--
 fs/gfs2/glops.c  | 30 +++++++++++++++++++-----------
 fs/gfs2/incore.h |  2 +-
 3 files changed, 34 insertions(+), 14 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 2b43a8488aec..a09ffc7b439b 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -600,8 +600,20 @@ __acquires(&gl->gl_lockref.lock)
 	    (lck_flags & (LM_FLAG_TRY|LM_FLAG_TRY_1CB)))
 		clear_bit(GLF_BLOCKING, &gl->gl_flags);
 	spin_unlock(&gl->gl_lockref.lock);
-	if (glops->go_sync)
-		glops->go_sync(gl);
+	if (glops->go_sync) {
+		ret = glops->go_sync(gl);
+		/* If we had a problem syncing (due to io errors or whatever,
+		 * we should not invalidate the metadata or tell dlm to
+		 * release the glock to other nodes.
+		 */
+		if (ret) {
+			if (cmpxchg(&sdp->sd_log_error, 0, ret)) {
+				fs_err(sdp, "Error %d syncing glock \n", ret);
+				gfs2_dump_glock(NULL, gl, true);
+			}
+			return;
+		}
+	}
 	if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags)) {
 		/*
 		 * The call to go_sync should have cleared out the ail list.
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index bbbcae8d853c..9e9c7a4b8c66 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -82,10 +82,11 @@ static void __gfs2_ail_flush(struct gfs2_glock *gl, bool fsync,
 }
 
 
-static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
+static int gfs2_ail_empty_gl(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	struct gfs2_trans tr;
+	int ret;
 
 	memset(&tr, 0, sizeof(tr));
 	INIT_LIST_HEAD(&tr.tr_buf);
@@ -116,7 +117,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 			goto flush;
 		if (log_in_flight)
 			log_flush_wait(sdp);
-		return;
+		return 0;
 	}
 
 	/* A shortened, inline version of gfs2_trans_begin()
@@ -124,8 +125,9 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
          * on the stack */
 	tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes);
 	tr.tr_ip = _RET_IP_;
-	if (gfs2_log_reserve(sdp, tr.tr_reserved) < 0)
-		return;
+	ret = gfs2_log_reserve(sdp, tr.tr_reserved);
+	if (ret < 0)
+		return ret;
 	WARN_ON_ONCE(current->journal_info);
 	current->journal_info = &tr;
 
@@ -135,6 +137,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 flush:
 	gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_NORMAL |
 		       GFS2_LFC_AIL_EMPTY_GL);
+	return 0;
 }
 
 void gfs2_ail_flush(struct gfs2_glock *gl, bool fsync)
@@ -168,7 +171,7 @@ void gfs2_ail_flush(struct gfs2_glock *gl, bool fsync)
  * return to caller to demote/unlock the glock until I/O is complete.
  */
 
-static void rgrp_go_sync(struct gfs2_glock *gl)
+static int rgrp_go_sync(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	struct address_space *mapping = &sdp->sd_aspace;
@@ -176,21 +179,24 @@ static void rgrp_go_sync(struct gfs2_glock *gl)
 	int error;
 
 	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
-		return;
+		return 0;
 	GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE);
 
 	gfs2_log_flush(sdp, gl, GFS2_LOG_HEAD_FLUSH_NORMAL |
 		       GFS2_LFC_RGRP_GO_SYNC);
 	filemap_fdatawrite_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
 	error = filemap_fdatawait_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
+	WARN_ON_ONCE(error);
 	mapping_set_error(mapping, error);
-	gfs2_ail_empty_gl(gl);
+	if (!error)
+		error = gfs2_ail_empty_gl(gl);
 
 	spin_lock(&gl->gl_lockref.lock);
 	rgd = gl->gl_object;
 	if (rgd)
 		gfs2_free_clones(rgd);
 	spin_unlock(&gl->gl_lockref.lock);
+	return error;
 }
 
 /**
@@ -257,12 +263,12 @@ static void gfs2_clear_glop_pending(struct gfs2_inode *ip)
  *
  */
 
-static void inode_go_sync(struct gfs2_glock *gl)
+static int inode_go_sync(struct gfs2_glock *gl)
 {
 	struct gfs2_inode *ip = gfs2_glock2inode(gl);
 	int isreg = ip && S_ISREG(ip->i_inode.i_mode);
 	struct address_space *metamapping = gfs2_glock2aspace(gl);
-	int error;
+	int error = 0;
 
 	if (isreg) {
 		if (test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags))
@@ -295,6 +301,7 @@ static void inode_go_sync(struct gfs2_glock *gl)
 
 out:
 	gfs2_clear_glop_pending(ip);
+	return error;
 }
 
 /**
@@ -515,7 +522,7 @@ static void inode_go_dump(struct seq_file *seq, struct gfs2_glock *gl,
  *
  */
 
-static void freeze_go_sync(struct gfs2_glock *gl)
+static int freeze_go_sync(struct gfs2_glock *gl)
 {
 	int error = 0;
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
@@ -529,7 +536,7 @@ static void freeze_go_sync(struct gfs2_glock *gl)
 				error);
 			if (gfs2_withdrawn(sdp)) {
 				atomic_set(&sdp->sd_freeze_state, SFS_UNFROZEN);
-				return;
+				return 0;
 			}
 			gfs2_assert_withdraw(sdp, 0);
 		}
@@ -537,6 +544,7 @@ static void freeze_go_sync(struct gfs2_glock *gl)
 		gfs2_log_flush(sdp, NULL, GFS2_LOG_HEAD_FLUSH_FREEZE |
 			       GFS2_LFC_FREEZE_GO_SYNC);
 	}
+	return 0;
 }
 
 /**
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 259b1956e990..acb4aa691e81 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -234,7 +234,7 @@ struct lm_lockname {
 
 
 struct gfs2_glock_operations {
-	void (*go_sync) (struct gfs2_glock *gl);
+	int (*go_sync) (struct gfs2_glock *gl);
 	int (*go_xmote_bh) (struct gfs2_glock *gl, struct gfs2_holder *gh);
 	void (*go_inval) (struct gfs2_glock *gl, int flags);
 	int (*go_demote_ok) (const struct gfs2_glock *gl);
-- 
2.24.1

[Cluster-devel] [GFS2 PATCH 25/25] gfs2: flesh out delayed withdraw for gfs2_log_flush

[Bob Peterson]

Function gfs2_log_flush() had a few places where it tried to withdraw
from the file system when errors were encountered. The problem is,
it should delay those withdraws until the log flush lock is no longer
held.

This patch creates a new function just for delayed withdraws for
situations like this. If errors=panic was specified on mount, we
still want to do it the old fashioned way because the panic it does
not help to delay in that situation.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/log.c  | 12 ++++++++----
 fs/gfs2/util.c | 27 +++++++++++++++++++++------
 fs/gfs2/util.h | 14 ++++++++++++--
 3 files changed, 41 insertions(+), 12 deletions(-)

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 010c319caade..67465a34954e 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -872,13 +872,17 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		INIT_LIST_HEAD(&tr->tr_ail2_list);
 		tr->tr_first = sdp->sd_log_flush_head;
 		if (unlikely (state == SFS_FROZEN))
-			gfs2_assert_withdraw(sdp, !tr->tr_num_buf_new && !tr->tr_num_databuf_new);
+			if (gfs2_assert_withdraw_delayed(sdp,
+			       !tr->tr_num_buf_new && !tr->tr_num_databuf_new))
+				goto out;
 	}
 
 	if (unlikely(state == SFS_FROZEN))
-		gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
-	gfs2_assert_withdraw(sdp,
-			sdp->sd_log_num_revoke == sdp->sd_log_committed_revoke);
+		if (gfs2_assert_withdraw_delayed(sdp, !sdp->sd_log_num_revoke))
+			goto out;
+	if (gfs2_assert_withdraw_delayed(sdp,
+			sdp->sd_log_num_revoke == sdp->sd_log_committed_revoke))
+		goto out;
 
 	gfs2_ordered_write(sdp);
 	if (gfs2_withdrawn(sdp))
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 37a89916c755..0250f32445c2 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -295,13 +295,28 @@ int gfs2_withdraw(struct gfs2_sbd *sdp)
  */
 
 void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
-			    const char *function, char *file, unsigned int line)
+			    const char *function, char *file, unsigned int line,
+			    bool delayed)
 {
-	gfs2_lm(sdp,
-		"fatal: assertion \"%s\" failed\n"
-		"   function = %s, file = %s, line = %u\n",
-		assertion, function, file, line);
-	gfs2_withdraw(sdp);
+	if (gfs2_withdrawn(sdp))
+		return;
+
+	fs_err(sdp,
+	       "fatal: assertion \"%s\" failed\n"
+	       "   function = %s, file = %s, line = %u\n",
+	       assertion, function, file, line);
+
+	/*
+	 * If errors=panic was specified on mount, it won't help to delay the
+	 * withdraw.
+	 */
+	if (sdp->sd_args.ar_errors == GFS2_ERRORS_PANIC)
+		delayed = false;
+
+	if (delayed)
+		gfs2_withdraw_delayed(sdp);
+	else
+		gfs2_withdraw(sdp);
 	dump_stack();
 }
 
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 97117a766bde..a3542560da6f 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -37,14 +37,24 @@ do { \
 
 
 void gfs2_assert_withdraw_i(struct gfs2_sbd *sdp, char *assertion,
-			    const char *function, char *file, unsigned int line);
+			    const char *function, char *file, unsigned int line,
+			    bool delayed);
 
 #define gfs2_assert_withdraw(sdp, assertion) \
 	({ \
 		bool _bool = (assertion); \
 		if (unlikely(!_bool)) \
 			gfs2_assert_withdraw_i((sdp), #assertion, \
-					__func__, __FILE__, __LINE__); \
+					__func__, __FILE__, __LINE__, false); \
+		!_bool; \
+	})
+
+#define gfs2_assert_withdraw_delayed(sdp, assertion) \
+	({ \
+		bool _bool = (assertion); \
+		if (unlikely(!_bool)) \
+			gfs2_assert_withdraw_i((sdp), #assertion, \
+					__func__, __FILE__, __LINE__, true); \
 		!_bool; \
 	})
 
-- 
2.24.1