From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932985Ab3CECRd (ORCPT ); Mon, 4 Mar 2013 21:17:33 -0500 Received: from mail-ia0-f170.google.com ([209.85.210.170]:55434 "EHLO mail-ia0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932948Ab3CECRb (ORCPT ); Mon, 4 Mar 2013 21:17:31 -0500 From: Michel Lespinasse To: linux-kernel@vger.kernel.org, Peter Zijlstra , Ingo Molnar , Oleg Nesterov , Lai Jiangshan , "Srivatsa S. Bhat" , paulmck@linux.vnet.ibm.com, Rusty Russell , rostedt@goodmis.org, tglx@linutronix.de, Andrew Morton , Andi Kleen Subject: [PATCH 2/2] lglock: update lockdep annotations to report recursive local locks Date: Mon, 4 Mar 2013 18:17:25 -0800 Message-Id: <1362449845-7492-3-git-send-email-walken@google.com> X-Mailer: git-send-email 1.8.1.3 In-Reply-To: <1362449845-7492-1-git-send-email-walken@google.com> References: <1362449845-7492-1-git-send-email-walken@google.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Oleg Nesterov recently noticed that the lockdep annotations in lglock.c are not sufficient to detect some obvious deadlocks, such as lg_local_lock(LOCK) + lg_local_lock(LOCK) or spin_lock(X) + lg_local_lock(Y) vs lg_local_lock(Y) + spin_lock(X). Both issues are easily fixed by indicating to lockdep that lglock's local locks are not recursive. We shouldn't use the rwlock acquire/release functions here, as lglock doesn't share the same semantics. Instead we can base our lockdep annotations on the lock_acquire_shared (for local lglock) and lock_acquire_exclusive (for global lglock) helpers. I am not proposing new lglock specific helpers as I don't see the point of the existing second level of helpers :) Signed-off-by: Michel Lespinasse --- kernel/lglock.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/kernel/lglock.c b/kernel/lglock.c index 6535a667a5a7..86ae2aebf004 100644 --- a/kernel/lglock.c +++ b/kernel/lglock.c @@ -21,7 +21,7 @@ void lg_local_lock(struct lglock *lg) arch_spinlock_t *lock; preempt_disable(); - rwlock_acquire_read(&lg->lock_dep_map, 0, 0, _RET_IP_); + lock_acquire_shared(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); lock = this_cpu_ptr(lg->lock); arch_spin_lock(lock); } @@ -31,7 +31,7 @@ void lg_local_unlock(struct lglock *lg) { arch_spinlock_t *lock; - rwlock_release(&lg->lock_dep_map, 1, _RET_IP_); + lock_release(&lg->lock_dep_map, 1, _RET_IP_); lock = this_cpu_ptr(lg->lock); arch_spin_unlock(lock); preempt_enable(); @@ -43,7 +43,7 @@ void lg_local_lock_cpu(struct lglock *lg, int cpu) arch_spinlock_t *lock; preempt_disable(); - rwlock_acquire_read(&lg->lock_dep_map, 0, 0, _RET_IP_); + lock_acquire_shared(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); lock = per_cpu_ptr(lg->lock, cpu); arch_spin_lock(lock); } @@ -53,7 +53,7 @@ void lg_local_unlock_cpu(struct lglock *lg, int cpu) { arch_spinlock_t *lock; - rwlock_release(&lg->lock_dep_map, 1, _RET_IP_); + lock_release(&lg->lock_dep_map, 1, _RET_IP_); lock = per_cpu_ptr(lg->lock, cpu); arch_spin_unlock(lock); preempt_enable(); @@ -65,7 +65,7 @@ void lg_global_lock(struct lglock *lg) int i; preempt_disable(); - rwlock_acquire(&lg->lock_dep_map, 0, 0, _RET_IP_); + lock_acquire_exclusive(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); for_each_possible_cpu(i) { arch_spinlock_t *lock; lock = per_cpu_ptr(lg->lock, i); @@ -78,7 +78,7 @@ void lg_global_unlock(struct lglock *lg) { int i; - rwlock_release(&lg->lock_dep_map, 1, _RET_IP_); + lock_release(&lg->lock_dep_map, 1, _RET_IP_); for_each_possible_cpu(i) { arch_spinlock_t *lock; lock = per_cpu_ptr(lg->lock, i); -- 1.8.1.3