cluster-devel.redhat.com archive mirror
 help / color / mirror / Atom feed
From: Bob Peterson <rpeterso@redhat.com>
To: cluster-devel.redhat.com
Subject: [Cluster-devel] [GFS2 PATCH 25/28] gfs2: flesh out delayed withdraw for gfs2_log_flush
Date: Thu, 20 Feb 2020 13:53:26 -0600	[thread overview]
Message-ID: <20200220195329.952027-26-rpeterso@redhat.com> (raw)
In-Reply-To: <20200220195329.952027-1-rpeterso@redhat.com>

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 3771fb6451c1..b41273faab8d 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -318,13 +318,28 @@ 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



  parent reply	other threads:[~2020-02-20 19:53 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-20 19:53 [Cluster-devel] [GFS2 PATCH 00/28] GFS2 recovery patches v10 Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 01/28] gfs2: Split gfs2_lm_withdraw into two functions Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 02/28] gfs2: Report errors before withdraw Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 03/28] gfs2: Remove usused cluster_wide arguments of gfs2_consist functions Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 04/28] gfs2: Turn gfs2_consist into void functions Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 05/28] gfs2: Return bool from gfs2_assert functions Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 06/28] gfs2: Introduce concept of a pending withdraw Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 07/28] gfs2: clear ail1 list when gfs2 withdraws Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 08/28] gfs2: Rework how rgrp buffer_heads are managed Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 09/28] gfs2: log error reform Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 10/28] gfs2: Only complain the first time an io error occurs in quota or log Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 11/28] gfs2: Ignore dlm recovery requests if gfs2 is withdrawn Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 12/28] gfs2: move check_journal_clean to util.c for future use Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 13/28] gfs2: Allow some glocks to be used during withdraw Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 14/28] gfs2: Force withdraw to replay journals and wait for it to finish Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 15/28] gfs2: fix infinite loop when checking ail item count before go_inval Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 16/28] gfs2: Add verbose option to check_journal_clean Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 17/28] gfs2: Issue revokes more intelligently Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 18/28] gfs2: Prepare to withdraw as soon as an IO error occurs in log write Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 19/28] gfs2: Check for log write errors before telling dlm to unlock Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 20/28] gfs2: Do log_flush in gfs2_ail_empty_gl even if ail list is empty Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 21/28] gfs2: Withdraw in gfs2_ail1_flush if write_cache_pages fails Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 22/28] gfs2: drain the ail2 list after io errors Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 23/28] gfs2: Don't demote a glock until its revokes are written Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 24/28] gfs2: Do proper error checking for go_sync family of glops functions Bob Peterson
2020-02-20 19:53 ` Bob Peterson [this message]
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 26/28] fs: clean up __block_commit_write Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 27/28] gfs2: don't allow releasepage to free bd still used for revokes Bob Peterson
2020-02-20 19:53 ` [Cluster-devel] [GFS2 PATCH 28/28] gfs2: allow journal replay to hold sd_log_flush_lock Bob Peterson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200220195329.952027-26-rpeterso@redhat.com \
    --to=rpeterso@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).