diff options
Diffstat (limited to 'fs/io-wq.c')
-rw-r--r-- | fs/io-wq.c | 117 |
1 files changed, 92 insertions, 25 deletions
diff --git a/fs/io-wq.c b/fs/io-wq.c index 0a182f1333e8..02894df7656d 100644 --- a/fs/io-wq.c +++ b/fs/io-wq.c @@ -18,7 +18,10 @@ #include <linux/fs_struct.h> #include <linux/task_work.h> #include <linux/blk-cgroup.h> +#include <linux/audit.h> +#include <linux/cpu.h> +#include "../kernel/sched/sched.h" #include "io-wq.h" #define WORKER_IDLE_TIMEOUT (5 * HZ) @@ -122,9 +125,13 @@ struct io_wq { refcount_t refs; struct completion done; + struct hlist_node cpuhp_node; + refcount_t use_refs; }; +static enum cpuhp_state io_wq_online; + static bool io_worker_get(struct io_worker *worker) { return refcount_inc_not_zero(&worker->ref); @@ -186,7 +193,8 @@ static bool __io_worker_unuse(struct io_wqe *wqe, struct io_worker *worker) worker->blkcg_css = NULL; } #endif - + if (current->signal->rlim[RLIMIT_FSIZE].rlim_cur != RLIM_INFINITY) + current->signal->rlim[RLIMIT_FSIZE].rlim_cur = RLIM_INFINITY; return dropped_lock; } @@ -429,14 +437,10 @@ static void io_wq_switch_mm(struct io_worker *worker, struct io_wq_work *work) mmput(worker->mm); worker->mm = NULL; } - if (!work->mm) - return; - if (mmget_not_zero(work->mm)) { - kthread_use_mm(work->mm); - worker->mm = work->mm; - /* hang on to this mm */ - work->mm = NULL; + if (mmget_not_zero(work->identity->mm)) { + kthread_use_mm(work->identity->mm); + worker->mm = work->identity->mm; return; } @@ -448,9 +452,11 @@ static inline void io_wq_switch_blkcg(struct io_worker *worker, struct io_wq_work *work) { #ifdef CONFIG_BLK_CGROUP - if (work->blkcg_css != worker->blkcg_css) { - kthread_associate_blkcg(work->blkcg_css); - worker->blkcg_css = work->blkcg_css; + if (!(work->flags & IO_WQ_WORK_BLKCG)) + return; + if (work->identity->blkcg_css != worker->blkcg_css) { + kthread_associate_blkcg(work->identity->blkcg_css); + worker->blkcg_css = work->identity->blkcg_css; } #endif } @@ -458,9 +464,9 @@ static inline void io_wq_switch_blkcg(struct io_worker *worker, static void io_wq_switch_creds(struct io_worker *worker, struct io_wq_work *work) { - const struct cred *old_creds = override_creds(work->creds); + const struct cred *old_creds = override_creds(work->identity->creds); - worker->cur_creds = work->creds; + worker->cur_creds = work->identity->creds; if (worker->saved_creds) put_cred(old_creds); /* creds set by previous switch */ else @@ -470,20 +476,29 @@ static void io_wq_switch_creds(struct io_worker *worker, static void io_impersonate_work(struct io_worker *worker, struct io_wq_work *work) { - if (work->files && current->files != work->files) { + if ((work->flags & IO_WQ_WORK_FILES) && + current->files != work->identity->files) { task_lock(current); - current->files = work->files; - current->nsproxy = work->nsproxy; + current->files = work->identity->files; + current->nsproxy = work->identity->nsproxy; task_unlock(current); } - if (work->fs && current->fs != work->fs) - current->fs = work->fs; - if (work->mm != worker->mm) + if ((work->flags & IO_WQ_WORK_FS) && current->fs != work->identity->fs) + current->fs = work->identity->fs; + if ((work->flags & IO_WQ_WORK_MM) && work->identity->mm != worker->mm) io_wq_switch_mm(worker, work); - if (worker->cur_creds != work->creds) + if ((work->flags & IO_WQ_WORK_CREDS) && + worker->cur_creds != work->identity->creds) io_wq_switch_creds(worker, work); - current->signal->rlim[RLIMIT_FSIZE].rlim_cur = work->fsize; + if (work->flags & IO_WQ_WORK_FSIZE) + current->signal->rlim[RLIMIT_FSIZE].rlim_cur = work->identity->fsize; + else if (current->signal->rlim[RLIMIT_FSIZE].rlim_cur != RLIM_INFINITY) + current->signal->rlim[RLIMIT_FSIZE].rlim_cur = RLIM_INFINITY; io_wq_switch_blkcg(worker, work); +#ifdef CONFIG_AUDIT + current->loginuid = work->identity->loginuid; + current->sessionid = work->identity->sessionid; +#endif } static void io_assign_current_work(struct io_worker *worker, @@ -496,6 +511,11 @@ static void io_assign_current_work(struct io_worker *worker, cond_resched(); } +#ifdef CONFIG_AUDIT + current->loginuid = KUIDT_INIT(AUDIT_UID_UNSET); + current->sessionid = AUDIT_SID_UNSET; +#endif + spin_lock_irq(&worker->lock); worker->cur_work = work; spin_unlock_irq(&worker->lock); @@ -676,6 +696,7 @@ static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index) kfree(worker); return false; } + kthread_bind_mask(worker->task, cpumask_of_node(wqe->node)); raw_spin_lock_irq(&wqe->lock); hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list); @@ -1076,10 +1097,12 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data) return ERR_PTR(-ENOMEM); wq->wqes = kcalloc(nr_node_ids, sizeof(struct io_wqe *), GFP_KERNEL); - if (!wq->wqes) { - kfree(wq); - return ERR_PTR(-ENOMEM); - } + if (!wq->wqes) + goto err_wq; + + ret = cpuhp_state_add_instance_nocalls(io_wq_online, &wq->cpuhp_node); + if (ret) + goto err_wqes; wq->free_work = data->free_work; wq->do_work = data->do_work; @@ -1087,6 +1110,7 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data) /* caller must already hold a reference to this */ wq->user = data->user; + ret = -ENOMEM; for_each_node(node) { struct io_wqe *wqe; int alloc_node = node; @@ -1130,9 +1154,12 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data) ret = PTR_ERR(wq->manager); complete(&wq->done); err: + cpuhp_state_remove_instance_nocalls(io_wq_online, &wq->cpuhp_node); for_each_node(node) kfree(wq->wqes[node]); +err_wqes: kfree(wq->wqes); +err_wq: kfree(wq); return ERR_PTR(ret); } @@ -1149,6 +1176,8 @@ static void __io_wq_destroy(struct io_wq *wq) { int node; + cpuhp_state_remove_instance_nocalls(io_wq_online, &wq->cpuhp_node); + set_bit(IO_WQ_BIT_EXIT, &wq->state); if (wq->manager) kthread_stop(wq->manager); @@ -1176,3 +1205,41 @@ struct task_struct *io_wq_get_task(struct io_wq *wq) { return wq->manager; } + +static bool io_wq_worker_affinity(struct io_worker *worker, void *data) +{ + struct task_struct *task = worker->task; + struct rq_flags rf; + struct rq *rq; + + rq = task_rq_lock(task, &rf); + do_set_cpus_allowed(task, cpumask_of_node(worker->wqe->node)); + task->flags |= PF_NO_SETAFFINITY; + task_rq_unlock(rq, task, &rf); + return false; +} + +static int io_wq_cpu_online(unsigned int cpu, struct hlist_node *node) +{ + struct io_wq *wq = hlist_entry_safe(node, struct io_wq, cpuhp_node); + int i; + + rcu_read_lock(); + for_each_node(i) + io_wq_for_each_worker(wq->wqes[i], io_wq_worker_affinity, NULL); + rcu_read_unlock(); + return 0; +} + +static __init int io_wq_init(void) +{ + int ret; + + ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, "io-wq/online", + io_wq_cpu_online, NULL); + if (ret < 0) + return ret; + io_wq_online = ret; + return 0; +} +subsys_initcall(io_wq_init); |