From f05b558d7e0368e0511a098c4beb9654e41246c4 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 3 Jun 2014 15:33:27 +0800 Subject: workqueue: clear POOL_DISASSOCIATED in rebind_workers() The commit a9ab775bcadf ("workqueue: directly restore CPU affinity of workers from CPU_ONLINE") moved the pool->lock into rebind_workers() without also moving "pool->flags &= ~POOL_DISASSOCIATED". There is nothing wrong with "pool->flags &= ~POOL_DISASSOCIATED" not being moved together, but there isn't any benefit either. We move it into rebind_workers() and achieve these benefits: 1) Better readability. POOL_DISASSOCIATED is cleared in rebind_workers() as expected. 2) When POOL_DISASSOCIATED is cleared, we can ensure that all the running workers of the pool are on the local CPU (pool->cpu). tj: Cosmetic updates to the code and description. Signed-off-by: Lai Jiangshan Signed-off-by: Tejun Heo --- kernel/workqueue.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 6203d2900877..129e4c3452ee 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -4542,6 +4542,7 @@ static void rebind_workers(struct worker_pool *pool) pool->attrs->cpumask) < 0); spin_lock_irq(&pool->lock); + pool->flags &= ~POOL_DISASSOCIATED; for_each_pool_worker(worker, pool) { unsigned int worker_flags = worker->flags; @@ -4643,15 +4644,10 @@ static int workqueue_cpu_up_callback(struct notifier_block *nfb, for_each_pool(pool, pi) { mutex_lock(&pool->attach_mutex); - if (pool->cpu == cpu) { - spin_lock_irq(&pool->lock); - pool->flags &= ~POOL_DISASSOCIATED; - spin_unlock_irq(&pool->lock); - + if (pool->cpu == cpu) rebind_workers(pool); - } else if (pool->cpu < 0) { + else if (pool->cpu < 0) restore_unbound_workers_cpumask(pool, cpu); - } mutex_unlock(&pool->attach_mutex); } -- cgit v1.2.3 From 807407c0a29b1e9fe35565f5a671ef3a9dd9f00c Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 3 Jun 2014 15:33:28 +0800 Subject: workqueue: stronger test in process_one_work() After the recent changes, when POOL_DISASSOCIATED is cleared, the running worker's local CPU should be the same as pool->cpu without any exception even during cpu-hotplug. Update the sanity check in process_one_work() accordingly. This patch changes "(proposition_A && proposition_B && proposition_C)" to "(proposition_B && proposition_C)", so if the old compound proposition is true, the new one must be true too. so this will not hide any possible bug which can be caught by the old test. tj: Minor updates to the description. CC: Jason J. Herne CC: Sasha Levin Signed-off-by: Lai Jiangshan Signed-off-by: Tejun Heo --- kernel/workqueue.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 129e4c3452ee..a3021d63f62d 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2020,13 +2020,8 @@ __acquires(&pool->lock) lockdep_copy_map(&lockdep_map, &work->lockdep_map); #endif - /* - * Ensure we're on the correct CPU. DISASSOCIATED test is - * necessary to avoid spurious warnings from rescuers servicing the - * unbound or a disassociated pool. - */ - WARN_ON_ONCE(!(worker->flags & WORKER_UNBOUND) && - !(pool->flags & POOL_DISASSOCIATED) && + /* ensure we're on the correct CPU */ + WARN_ON_ONCE(!(pool->flags & POOL_DISASSOCIATED) && raw_smp_processor_id() != pool->cpu); /* -- cgit v1.2.3 From 9a1049da9bd2cd83fe11d46433e603c193aa9c71 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 28 Jun 2014 08:10:14 -0400 Subject: percpu-refcount: require percpu_ref to be exited explicitly Currently, a percpu_ref undoes percpu_ref_init() automatically by freeing the allocated percpu area when the percpu_ref is killed. While seemingly convenient, this has the following niggles. * It's impossible to re-init a released reference counter without going through re-allocation. * In the similar vein, it's impossible to initialize a percpu_ref count with static percpu variables. * We need and have an explicit destructor anyway for failure paths - percpu_ref_cancel_init(). This patch removes the automatic percpu counter freeing in percpu_ref_kill_rcu() and repurposes percpu_ref_cancel_init() into a generic destructor now named percpu_ref_exit(). percpu_ref_destroy() is considered but it gets confusing with percpu_ref_kill() while "exit" clearly indicates that it's the counterpart of percpu_ref_init(). All percpu_ref_cancel_init() users are updated to invoke percpu_ref_exit() instead and explicit percpu_ref_exit() calls are added to the destruction path of all percpu_ref users. Signed-off-by: Tejun Heo Acked-by: Benjamin LaHaise Cc: Kent Overstreet Cc: Christoph Lameter Cc: Benjamin LaHaise Cc: Nicholas A. Bellinger Cc: Li Zefan --- drivers/target/target_core_tpg.c | 4 +++- fs/aio.c | 6 ++++-- include/linux/percpu-refcount.h | 6 ++---- kernel/cgroup.c | 8 +++++--- lib/percpu-refcount.c | 33 ++++++++++----------------------- 5 files changed, 24 insertions(+), 33 deletions(-) (limited to 'kernel') diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c index c036595b17cf..fddfae61222f 100644 --- a/drivers/target/target_core_tpg.c +++ b/drivers/target/target_core_tpg.c @@ -825,7 +825,7 @@ int core_tpg_add_lun( ret = core_dev_export(dev, tpg, lun); if (ret < 0) { - percpu_ref_cancel_init(&lun->lun_ref); + percpu_ref_exit(&lun->lun_ref); return ret; } @@ -880,5 +880,7 @@ int core_tpg_post_dellun( lun->lun_status = TRANSPORT_LUN_STATUS_FREE; spin_unlock(&tpg->tpg_lun_lock); + percpu_ref_exit(&lun->lun_ref); + return 0; } diff --git a/fs/aio.c b/fs/aio.c index 5e0d7f9cb693..ea1bc2e8f4f3 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -506,6 +506,8 @@ static void free_ioctx(struct work_struct *work) aio_free_ring(ctx); free_percpu(ctx->cpu); + percpu_ref_exit(&ctx->reqs); + percpu_ref_exit(&ctx->users); kmem_cache_free(kioctx_cachep, ctx); } @@ -715,8 +717,8 @@ err_ctx: err: mutex_unlock(&ctx->ring_lock); free_percpu(ctx->cpu); - percpu_ref_cancel_init(&ctx->reqs); - percpu_ref_cancel_init(&ctx->users); + percpu_ref_exit(&ctx->reqs); + percpu_ref_exit(&ctx->users); kmem_cache_free(kioctx_cachep, ctx); pr_debug("error allocating ioctx %d\n", err); return ERR_PTR(err); diff --git a/include/linux/percpu-refcount.h b/include/linux/percpu-refcount.h index 6f8cd4c0546c..0ddd2839ca84 100644 --- a/include/linux/percpu-refcount.h +++ b/include/linux/percpu-refcount.h @@ -57,9 +57,7 @@ struct percpu_ref { atomic_t count; /* * The low bit of the pointer indicates whether the ref is in percpu - * mode; if set, then get/put will manipulate the atomic_t (this is a - * hack because we need to keep the pointer around for - * percpu_ref_kill_rcu()) + * mode; if set, then get/put will manipulate the atomic_t. */ unsigned long pcpu_count_ptr; percpu_ref_func_t *release; @@ -69,7 +67,7 @@ struct percpu_ref { int __must_check percpu_ref_init(struct percpu_ref *ref, percpu_ref_func_t *release); -void percpu_ref_cancel_init(struct percpu_ref *ref); +void percpu_ref_exit(struct percpu_ref *ref); void percpu_ref_kill_and_confirm(struct percpu_ref *ref, percpu_ref_func_t *confirm_kill); diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 7868fc3c0bc5..c06aa5e257a8 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1638,7 +1638,7 @@ destroy_root: exit_root_id: cgroup_exit_root_id(root); cancel_ref: - percpu_ref_cancel_init(&root_cgrp->self.refcnt); + percpu_ref_exit(&root_cgrp->self.refcnt); out: free_cgrp_cset_links(&tmp_links); return ret; @@ -4133,6 +4133,8 @@ static void css_free_work_fn(struct work_struct *work) container_of(work, struct cgroup_subsys_state, destroy_work); struct cgroup *cgrp = css->cgroup; + percpu_ref_exit(&css->refcnt); + if (css->ss) { /* css free path */ if (css->parent) @@ -4330,7 +4332,7 @@ err_list_del: err_free_id: cgroup_idr_remove(&ss->css_idr, css->id); err_free_percpu_ref: - percpu_ref_cancel_init(&css->refcnt); + percpu_ref_exit(&css->refcnt); err_free_css: call_rcu(&css->rcu_head, css_free_rcu_fn); return err; @@ -4441,7 +4443,7 @@ static int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, out_free_id: cgroup_idr_remove(&root->cgroup_idr, cgrp->id); out_cancel_ref: - percpu_ref_cancel_init(&cgrp->self.refcnt); + percpu_ref_exit(&cgrp->self.refcnt); out_free_cgrp: kfree(cgrp); out_unlock: diff --git a/lib/percpu-refcount.c b/lib/percpu-refcount.c index 94e5b624de64..ac4299120087 100644 --- a/lib/percpu-refcount.c +++ b/lib/percpu-refcount.c @@ -61,36 +61,25 @@ int percpu_ref_init(struct percpu_ref *ref, percpu_ref_func_t *release) EXPORT_SYMBOL_GPL(percpu_ref_init); /** - * percpu_ref_cancel_init - cancel percpu_ref_init() - * @ref: percpu_ref to cancel init for + * percpu_ref_exit - undo percpu_ref_init() + * @ref: percpu_ref to exit * - * Once a percpu_ref is initialized, its destruction is initiated by - * percpu_ref_kill() and completes asynchronously, which can be painful to - * do when destroying a half-constructed object in init failure path. - * - * This function destroys @ref without invoking @ref->release and the - * memory area containing it can be freed immediately on return. To - * prevent accidental misuse, it's required that @ref has finished - * percpu_ref_init(), whether successful or not, but never used. - * - * The weird name and usage restriction are to prevent people from using - * this function by mistake for normal shutdown instead of - * percpu_ref_kill(). + * This function exits @ref. The caller is responsible for ensuring that + * @ref is no longer in active use. The usual places to invoke this + * function from are the @ref->release() callback or in init failure path + * where percpu_ref_init() succeeded but other parts of the initialization + * of the embedding object failed. */ -void percpu_ref_cancel_init(struct percpu_ref *ref) +void percpu_ref_exit(struct percpu_ref *ref) { unsigned __percpu *pcpu_count = pcpu_count_ptr(ref); - int cpu; - - WARN_ON_ONCE(atomic_read(&ref->count) != 1 + PCPU_COUNT_BIAS); if (pcpu_count) { - for_each_possible_cpu(cpu) - WARN_ON_ONCE(*per_cpu_ptr(pcpu_count, cpu)); free_percpu(pcpu_count); + ref->pcpu_count_ptr = PCPU_REF_DEAD; } } -EXPORT_SYMBOL_GPL(percpu_ref_cancel_init); +EXPORT_SYMBOL_GPL(percpu_ref_exit); static void percpu_ref_kill_rcu(struct rcu_head *rcu) { @@ -102,8 +91,6 @@ static void percpu_ref_kill_rcu(struct rcu_head *rcu) for_each_possible_cpu(cpu) count += *per_cpu_ptr(pcpu_count, cpu); - free_percpu(pcpu_count); - pr_debug("global %i pcpu %i", atomic_read(&ref->count), (int) count); /* -- cgit v1.2.3