All of lore.kernel.org
 help / color / mirror / Atom feed
From: "tip-bot for Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: linux-tip-commits@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, paulmck@linux.vnet.ibm.com,
	hpa@zytor.com, mingo@redhat.com, tglx@linutronix.de,
	mingo@elte.hu
Subject: [tip:core/rcu] rcu: Further cleanups of use of lastcomp
Date: Sat, 14 Nov 2009 10:18:55 GMT	[thread overview]
Message-ID: <tip-560d4bc0df9a5e63b980432282d8c2bd3559ec74@git.kernel.org> (raw)
In-Reply-To: <1258170699933-git-send-email->

Commit-ID:  560d4bc0df9a5e63b980432282d8c2bd3559ec74
Gitweb:     http://git.kernel.org/tip/560d4bc0df9a5e63b980432282d8c2bd3559ec74
Author:     Paul E. McKenney <paulmck@linux.vnet.ibm.com>
AuthorDate: Fri, 13 Nov 2009 19:51:38 -0800
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Sat, 14 Nov 2009 10:31:42 +0100

rcu: Further cleanups of use of lastcomp

Now that a copy of the rsp->completed flag is available in all
rcu_node structures, make full use of it.  It is still
legitimate to access rsp->completed while holding the root
rcu_node structure's lock, however.

Also, tighten up force_quiescent_state()'s checks for end of
current grace period.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
LKML-Reference: <1258170699933-git-send-email->
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 kernel/rcutree.c |   13 ++++++-------
 1 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index 3df0438..24bbf2c 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -817,7 +817,7 @@ cpu_quiet(int cpu, struct rcu_state *rsp, struct rcu_data *rdp, long lastcomp)
 
 	rnp = rdp->mynode;
 	spin_lock_irqsave(&rnp->lock, flags);
-	if (lastcomp != ACCESS_ONCE(rsp->completed)) {
+	if (lastcomp != rnp->completed) {
 
 		/*
 		 * Someone beat us to it for this grace period, so leave.
@@ -935,7 +935,6 @@ static void rcu_adopt_orphan_cbs(struct rcu_state *rsp)
 static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp)
 {
 	unsigned long flags;
-	long lastcomp;
 	unsigned long mask;
 	struct rcu_data *rdp = rsp->rda[cpu];
 	struct rcu_node *rnp;
@@ -971,7 +970,6 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp)
 		spin_unlock(&rnp->lock);	/* irqs remain disabled. */
 		rnp = rnp->parent;
 	} while (rnp != NULL);
-	lastcomp = rsp->completed;
 
 	spin_unlock_irqrestore(&rsp->onofflock, flags);
 
@@ -1145,7 +1143,7 @@ static int rcu_process_dyntick(struct rcu_state *rsp, long lastcomp,
 	rcu_for_each_leaf_node(rsp, rnp) {
 		mask = 0;
 		spin_lock_irqsave(&rnp->lock, flags);
-		if (rsp->completed != lastcomp) {
+		if (rnp->completed != lastcomp) {
 			spin_unlock_irqrestore(&rnp->lock, flags);
 			return 1;
 		}
@@ -1159,7 +1157,7 @@ static int rcu_process_dyntick(struct rcu_state *rsp, long lastcomp,
 			if ((rnp->qsmask & bit) != 0 && f(rsp->rda[cpu]))
 				mask |= bit;
 		}
-		if (mask != 0 && rsp->completed == lastcomp) {
+		if (mask != 0 && rnp->completed == lastcomp) {
 
 			/* cpu_quiet_msk() releases rnp->lock. */
 			cpu_quiet_msk(mask, rsp, rnp, flags);
@@ -1196,7 +1194,7 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
 	lastcomp = rsp->gpnum - 1;
 	signaled = rsp->signaled;
 	rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS;
-	if (lastcomp == rsp->gpnum) {
+	if(!rcu_gp_in_progress(rsp)) {
 		rsp->n_force_qs_ngp++;
 		spin_unlock(&rnp->lock);
 		goto unlock_ret;  /* no GP in progress, time updated. */
@@ -1224,7 +1222,8 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
 		/* Update state, record completion counter. */
 		forcenow = 0;
 		spin_lock(&rnp->lock);
-		if (lastcomp == rsp->completed &&
+		if (lastcomp + 1 == rsp->gpnum &&
+		    lastcomp == rsp->completed &&
 		    rsp->signaled == signaled) {
 			rsp->signaled = RCU_FORCE_QS;
 			rsp->completed_fqs = lastcomp;

  reply	other threads:[~2009-11-14 10:19 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-11-14  3:51 [PATCH tip/core/rcu 0/2] rcu: lastcomp cleanups, fix rcu_pending() false positives Paul E. McKenney
2009-11-14  3:51 ` [PATCH tip/core/rcu 1/2] rcu: further cleanups of use of lastcomp Paul E. McKenney
2009-11-14 10:18   ` tip-bot for Paul E. McKenney [this message]
2009-11-14  3:51 ` [PATCH tip/core/rcu 2/2] rcu: eliminate __rcu_pending() false positives Paul E. McKenney
2009-11-14 10:19   ` [tip:core/rcu] rcu: Eliminate " tip-bot for Paul E. McKenney

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=tip-560d4bc0df9a5e63b980432282d8c2bd3559ec74@git.kernel.org \
    --to=paulmck@linux.vnet.ibm.com \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-tip-commits@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=mingo@redhat.com \
    --cc=tglx@linutronix.de \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.