linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: NeilBrown <neilb@suse.de>
To: Shaohua Li <shli@kernel.org>
Cc: linux-raid@vger.kernel.org
Subject: [PATCH 6/8] md/raid5: be more selective about distributing flags across batch.
Date: Fri, 22 May 2015 15:30:58 +1000	[thread overview]
Message-ID: <20150522053058.2117.62712.stgit@notabene.brown> (raw)
In-Reply-To: <20150522052802.2117.40527.stgit@notabene.brown>

When a batch of stripes is broken up, we keep some of the flags
that were per-stripe, and copy other flags from the head to all
others.

This only happens while a stripe is being handled, so many of the
flags are irrelevant.

The "SYNC_FLAGS" (which I've renamed to make it clear there are
several) and STRIPE_DEGRADED are set per-stripe and so need to be
preserved.  STRIPE_INSYNC is the only flag that is set on the head
that needs to be propagated to all others.

For safety, add a WARN_ON if others are set.

Signed-off-by: NeilBrown <neilb@suse.de>
---
 drivers/md/raid5.c |   57 ++++++++++++++++++++++++++++++++++++++++++----------
 drivers/md/raid5.h |    2 +-
 2 files changed, 47 insertions(+), 12 deletions(-)

diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 1f12a0bfa8bd..39901e2b5502 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3526,10 +3526,28 @@ unhash:
 				      struct stripe_head, batch_list);
 		list_del_init(&sh->batch_list);
 
-		set_mask_bits(&sh->state, ~STRIPE_EXPAND_SYNC_FLAG,
-			      head_sh->state & ~((1 << STRIPE_ACTIVE) |
-						 (1 << STRIPE_PREREAD_ACTIVE) |
-						 STRIPE_EXPAND_SYNC_FLAG));
+		WARN_ON_ONCE(sh->state & ((1 << STRIPE_ACTIVE) |
+					  (1 << STRIPE_SYNCING) |
+					  (1 << STRIPE_REPLACED) |
+					  (1 << STRIPE_PREREAD_ACTIVE) |
+					  (1 << STRIPE_DELAYED) |
+					  (1 << STRIPE_BIT_DELAY) |
+					  (1 << STRIPE_FULL_WRITE) |
+					  (1 << STRIPE_BIOFILL_RUN) |
+					  (1 << STRIPE_COMPUTE_RUN)  |
+					  (1 << STRIPE_OPS_REQ_PENDING) |
+					  (1 << STRIPE_ON_UNPLUG_LIST) |
+					  (1 << STRIPE_DISCARD) |
+					  (1 << STRIPE_ON_RELEASE_LIST) |
+					  (1 << STRIPE_BATCH_READY) |
+					  (1 << STRIPE_BATCH_ERR)));
+		WARN_ON_ONCE(head_sh->state & ((1 << STRIPE_DISCARD) |
+					      (1 << STRIPE_REPLACED)));
+
+		set_mask_bits(&sh->state, ~(STRIPE_EXPAND_SYNC_FLAGS |
+					    (1 << STRIPE_DEGRADED)),
+			      head_sh->state & (1 << STRIPE_INSYNC));
+
 		sh->check_state = head_sh->check_state;
 		sh->reconstruct_state = head_sh->reconstruct_state;
 		for (i = 0; i < sh->disks; i++) {
@@ -3541,7 +3559,7 @@ unhash:
 		spin_lock_irq(&sh->stripe_lock);
 		sh->batch_head = NULL;
 		spin_unlock_irq(&sh->stripe_lock);
-		if (sh->state & STRIPE_EXPAND_SYNC_FLAG)
+		if (sh->state & STRIPE_EXPAND_SYNC_FLAGS)
 			set_bit(STRIPE_HANDLE, &sh->state);
 		release_stripe(sh);
 	}
@@ -3550,7 +3568,7 @@ unhash:
 	head_sh->batch_head = NULL;
 	spin_unlock_irq(&head_sh->stripe_lock);
 	wake_up_nr(&conf->wait_for_overlap, wakeup_nr);
-	if (head_sh->state & STRIPE_EXPAND_SYNC_FLAG)
+	if (head_sh->state & STRIPE_EXPAND_SYNC_FLAGS)
 		set_bit(STRIPE_HANDLE, &head_sh->state);
 }
 
@@ -4237,11 +4255,28 @@ static void break_stripe_batch_list(struct stripe_head *head_sh,
 
 		list_del_init(&sh->batch_list);
 
-		set_mask_bits(&sh->state, ~STRIPE_EXPAND_SYNC_FLAG,
-			      head_sh->state & ~((1 << STRIPE_ACTIVE) |
-						 (1 << STRIPE_PREREAD_ACTIVE) |
-						 (1 << STRIPE_DEGRADED) |
-						 STRIPE_EXPAND_SYNC_FLAG));
+		WARN_ON_ONCE(sh->state & ((1 << STRIPE_ACTIVE) |
+					  (1 << STRIPE_SYNCING) |
+					  (1 << STRIPE_REPLACED) |
+					  (1 << STRIPE_PREREAD_ACTIVE) |
+					  (1 << STRIPE_DELAYED) |
+					  (1 << STRIPE_BIT_DELAY) |
+					  (1 << STRIPE_FULL_WRITE) |
+					  (1 << STRIPE_BIOFILL_RUN) |
+					  (1 << STRIPE_COMPUTE_RUN)  |
+					  (1 << STRIPE_OPS_REQ_PENDING) |
+					  (1 << STRIPE_ON_UNPLUG_LIST) |
+					  (1 << STRIPE_DISCARD) |
+					  (1 << STRIPE_ON_RELEASE_LIST) |
+					  (1 << STRIPE_BATCH_READY) |
+					  (1 << STRIPE_BATCH_ERR)));
+		WARN_ON_ONCE(head_sh->state & ((1 << STRIPE_DISCARD) |
+					      (1 << STRIPE_REPLACED)));
+
+		set_mask_bits(&sh->state, ~(STRIPE_EXPAND_SYNC_FLAGS |
+					    (1 << STRIPE_DEGRADED)),
+			      head_sh->state & (1 << STRIPE_INSYNC));
+
 		sh->check_state = head_sh->check_state;
 		sh->reconstruct_state = head_sh->reconstruct_state;
 		for (i = 0; i < sh->disks; i++) {
diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index 6307b904d318..d7b2bc8b756f 100644
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -339,7 +339,7 @@ enum {
 	STRIPE_BATCH_ERR,
 };
 
-#define STRIPE_EXPAND_SYNC_FLAG \
+#define STRIPE_EXPAND_SYNC_FLAGS \
 	((1 << STRIPE_EXPAND_SOURCE) |\
 	(1 << STRIPE_EXPAND_READY) |\
 	(1 << STRIPE_EXPANDING) |\



      parent reply	other threads:[~2015-05-22  5:30 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-22  5:30 [PATCH 0/8] Fixes for md/raid5 stripe batching code NeilBrown
2015-05-22  5:30 ` [PATCH 7/8] md/raid5: call break_stripe_batch_list from handle_stripe_clean_event NeilBrown
2015-05-22  5:30 ` [PATCH 5/8] md/raid5: add handle_flags arg to break_stripe_batch_list NeilBrown
2015-05-22  5:30 ` [PATCH 4/8] md/raid5: duplicate some more handle_stripe_clean_event code in break_stripe_batch_list NeilBrown
2015-05-22  5:30 ` [PATCH 3/8] md/raid5: remove condition test from check_break_stripe_batch_list NeilBrown
2015-05-22  5:30 ` [PATCH 2/8] md/raid5: Ensure a batch member is not handled prematurely NeilBrown
2015-05-22 23:44   ` Shaohua Li
2015-05-23  0:26     ` NeilBrown
2015-05-26 18:16       ` Shaohua Li
2015-05-26 22:35         ` NeilBrown
2015-05-26 23:21           ` Shaohua Li
2015-05-26 23:34           ` NeilBrown
2015-05-27  0:10             ` Shaohua Li
2015-05-27  0:36               ` NeilBrown
2015-05-22  5:30 ` [PATCH 8/8] md/raid5: break stripe-batches when the array has failed NeilBrown
2015-05-22  5:30 ` [PATCH 1/8] md/raid5: ensure whole batch is delayed for all required bitmap updates NeilBrown
2015-05-22  5:30 ` NeilBrown [this message]

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=20150522053058.2117.62712.stgit@notabene.brown \
    --to=neilb@suse.de \
    --cc=linux-raid@vger.kernel.org \
    --cc=shli@kernel.org \
    /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).