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] [PATCH 09/32] gfs2: Allow some glocks to be used during withdraw
Date: Wed, 13 Nov 2019 15:30:07 -0600	[thread overview]
Message-ID: <20191113213030.237431-10-rpeterso@redhat.com> (raw)
In-Reply-To: <20191113213030.237431-1-rpeterso@redhat.com>

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when we're withdrawn.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disallow the use of inode and rgrp glocks when withdrawn.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.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
---
 fs/gfs2/glock.c  | 29 +++++++++++++++++++++++++++--
 fs/gfs2/glops.c  |  4 ++--
 fs/gfs2/incore.h |  1 +
 3 files changed, 30 insertions(+), 4 deletions(-)

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 809f9cf4239d..15ff968e8192 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -133,6 +133,31 @@ static void gfs2_glock_dealloc(struct rcu_head *rcu)
 	}
 }
 
+/**
+ * allow_while_withdrawn - determine if we can use this glock while withdrawn
+ * @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+ * when we're withdrawn.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+ * disallow the use of inode and rgrp glocks when withdrawn.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 allow_while_withdrawn(struct gfs2_glock *gl)
+{
+	struct gfs2_sbd *sdp;
+
+	if (!(gl->gl_ops->go_flags & GLOF_JOURNALED))
+		return true;
+
+	sdp = gl->gl_name.ln_sbd;
+	if (!sdp->sd_jdesc || gl->gl_object == sdp->sd_jdesc->jd_inode)
+		return true;
+	return false;
+}
+
 void gfs2_glock_free(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
@@ -549,7 +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)) &&
+	if (unlikely(gfs2_withdrawn(sdp)) && !allow_while_withdrawn(gl) &&
 	    target != LM_ST_UNLOCKED)
 		return;
 	lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP |
@@ -1190,7 +1215,7 @@ int gfs2_glock_nq(struct gfs2_holder *gh)
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	int error = 0;
 
-	if (unlikely(gfs2_withdrawn(sdp)))
+	if (unlikely(gfs2_withdrawn(sdp)) && !allow_while_withdrawn(gl))
 		return -EIO;
 
 	if (test_bit(GLF_LRU, &gl->gl_flags))
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index dec5e245b991..04f55e5b8bf1 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -587,7 +587,7 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
 	.go_lock = inode_go_lock,
 	.go_dump = inode_go_dump,
 	.go_type = LM_TYPE_INODE,
-	.go_flags = GLOF_ASPACE | GLOF_LRU,
+	.go_flags = GLOF_ASPACE | GLOF_LRU | GLOF_JOURNALED,
 };
 
 const struct gfs2_glock_operations gfs2_rgrp_glops = {
@@ -596,7 +596,7 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_lock = gfs2_rgrp_go_lock,
 	.go_dump = gfs2_rgrp_dump,
 	.go_type = LM_TYPE_RGRP,
-	.go_flags = GLOF_LVB,
+	.go_flags = GLOF_LVB | GLOF_JOURNALED,
 };
 
 const struct gfs2_glock_operations gfs2_freeze_glops = {
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index d37698502d5d..f6ec52776408 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -247,6 +247,7 @@ struct gfs2_glock_operations {
 #define GLOF_ASPACE 1
 #define GLOF_LVB    2
 #define GLOF_LRU    4
+#define GLOF_JOURNALED 8 /* goes through the journal */
 };
 
 enum {
-- 
2.23.0



  parent reply	other threads:[~2019-11-13 21:30 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-13 21:29 [Cluster-devel] [PATCH 00/32] gfs2: misc recovery patch collection Bob Peterson
2019-11-13 21:29 ` [Cluster-devel] [PATCH 01/32] gfs2: Introduce concept of a pending withdraw Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 02/32] gfs2: clear ail1 list when gfs2 withdraws Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 03/32] gfs2: Rework how rgrp buffer_heads are managed Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 04/32] gfs2: fix infinite loop in gfs2_ail1_flush on io error Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 05/32] gfs2: log error reform Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 06/32] gfs2: Only complain the first time an io error occurs in quota or log Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 07/32] gfs2: Ignore dlm recovery requests if gfs2 is withdrawn Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 08/32] gfs2: move check_journal_clean to util.c for future use Bob Peterson
2019-11-13 21:30 ` Bob Peterson [this message]
2019-11-13 21:30 ` [Cluster-devel] [PATCH 10/32] gfs2: Don't loop forever in gfs2_freeze if withdrawn Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 11/32] gfs2: Make secondary withdrawers wait for first withdrawer Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 12/32] gfs2: Don't write log headers after file system withdraw Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 13/32] gfs2: Force withdraw to replay journals and wait for it to finish Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 14/32] gfs2: fix infinite loop when checking ail item count before go_inval Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 15/32] gfs2: Add verbose option to check_journal_clean Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 16/32] gfs2: Abort gfs2_freeze if io error is seen Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 17/32] gfs2: Issue revokes more intelligently Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 18/32] gfs2: Prepare to withdraw as soon as an IO error occurs in log write Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 19/32] gfs2: Check for log write errors before telling dlm to unlock Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 20/32] gfs2: new slab for transactions Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 21/32] gfs2: Close timing window with GLF_INVALIDATE_IN_PROGRESS Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 22/32] gfs2: Do log_flush in gfs2_ail_empty_gl even if ail list is empty Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 23/32] gfs2: Don't skip log flush if glock still has revokes Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 24/32] gfs2: initialize tr_ail1_list when creating transactions Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 25/32] gfs2: Withdraw in gfs2_ail1_flush if write_cache_pages returns error Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 26/32] gfs2: drain the ail2 list after io errors Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 27/32] gfs2: make gfs2_log_shutdown static Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 28/32] gfs2: Eliminate GFS2_RDF_UPTODATE flag in favor of buffer existence Bob Peterson
2019-11-14 10:42   ` Steven Whitehouse
2019-11-14 13:16     ` Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 29/32] gfs2: if finish_open returns error, clean up iopen glock mess Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 30/32] gfs2: Don't demote a glock until its revokes are written Bob Peterson
2019-11-14 10:45   ` Steven Whitehouse
2019-11-13 21:30 ` [Cluster-devel] [PATCH 31/32] gfs2: Do proper error checking for go_sync family of glops functions Bob Peterson
2019-11-13 21:30 ` [Cluster-devel] [PATCH 32/32] gfs2: fix glock reference problem in gfs2_trans_add_unrevoke Bob Peterson
2019-11-14 10:48 ` [Cluster-devel] [PATCH 00/32] gfs2: misc recovery patch collection Steven Whitehouse

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=20191113213030.237431-10-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).