diff --git a/fs/io_uring.c b/fs/io_uring.c index 34c6c2f757630e0744158077e31a50937faa10d3..b2566ae7790b17eb50d47981e6cf4211efc56cea 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -437,20 +437,20 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p) return NULL; } -static inline bool __io_sequence_defer(struct io_ring_ctx *ctx, - struct io_kiocb *req) +static inline bool __io_sequence_defer(struct io_kiocb *req) { + struct io_ring_ctx *ctx = req->ctx; + return req->sequence != ctx->cached_cq_tail + ctx->cached_sq_dropped + atomic_read(&ctx->cached_cq_overflow); } -static inline bool io_sequence_defer(struct io_ring_ctx *ctx, - struct io_kiocb *req) +static inline bool io_sequence_defer(struct io_kiocb *req) { if ((req->flags & (REQ_F_IO_DRAIN|REQ_F_IO_DRAINED)) != REQ_F_IO_DRAIN) return false; - return __io_sequence_defer(ctx, req); + return __io_sequence_defer(req); } static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx) @@ -458,7 +458,7 @@ static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx) struct io_kiocb *req; req = list_first_entry_or_null(&ctx->defer_list, struct io_kiocb, list); - if (req && !io_sequence_defer(ctx, req)) { + if (req && !io_sequence_defer(req)) { list_del_init(&req->list); return req; } @@ -471,7 +471,7 @@ static struct io_kiocb *io_get_timeout_req(struct io_ring_ctx *ctx) struct io_kiocb *req; req = list_first_entry_or_null(&ctx->timeout_list, struct io_kiocb, list); - if (req && !__io_sequence_defer(ctx, req)) { + if (req && !__io_sequence_defer(req)) { list_del_init(&req->list); return req; } @@ -534,10 +534,10 @@ static inline bool io_prep_async_work(struct io_kiocb *req) return do_hashed; } -static inline void io_queue_async_work(struct io_ring_ctx *ctx, - struct io_kiocb *req) +static inline void io_queue_async_work(struct io_kiocb *req) { bool do_hashed = io_prep_async_work(req); + struct io_ring_ctx *ctx = req->ctx; trace_io_uring_queue_async_work(ctx, do_hashed, req, &req->work, req->flags); @@ -588,7 +588,7 @@ static void io_commit_cqring(struct io_ring_ctx *ctx) continue; } req->flags |= REQ_F_IO_DRAINED; - io_queue_async_work(ctx, req); + io_queue_async_work(req); } } @@ -791,9 +791,9 @@ static void __io_free_req(struct io_kiocb *req) kmem_cache_free(req_cachep, req); } -static bool io_link_cancel_timeout(struct io_ring_ctx *ctx, - struct io_kiocb *req) +static bool io_link_cancel_timeout(struct io_kiocb *req) { + struct io_ring_ctx *ctx = req->ctx; int ret; ret = hrtimer_try_to_cancel(&req->timeout.timer); @@ -833,7 +833,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr) * in this context instead of having to queue up new async work. */ if (req->flags & REQ_F_LINK_TIMEOUT) { - wake_ev = io_link_cancel_timeout(ctx, nxt); + wake_ev = io_link_cancel_timeout(nxt); /* we dropped this link, get next */ nxt = list_first_entry_or_null(&req->link_list, @@ -842,7 +842,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr) *nxtptr = nxt; break; } else { - io_queue_async_work(req->ctx, nxt); + io_queue_async_work(nxt); break; } } @@ -870,7 +870,7 @@ static void io_fail_links(struct io_kiocb *req) if ((req->flags & REQ_F_LINK_TIMEOUT) && link->submit.sqe->opcode == IORING_OP_LINK_TIMEOUT) { - io_link_cancel_timeout(ctx, link); + io_link_cancel_timeout(link); } else { io_cqring_fill_event(link, -ECANCELED); io_double_put_req(link); @@ -939,7 +939,7 @@ static void io_put_req(struct io_kiocb *req, struct io_kiocb **nxtptr) if (nxtptr) *nxtptr = nxt; else - io_queue_async_work(nxt->ctx, nxt); + io_queue_async_work(nxt); } } @@ -1899,7 +1899,7 @@ static void io_poll_remove_one(struct io_kiocb *req) WRITE_ONCE(poll->canceled, true); if (!list_empty(&poll->wait.entry)) { list_del_init(&poll->wait.entry); - io_queue_async_work(req->ctx, req); + io_queue_async_work(req); } spin_unlock(&poll->head->lock); @@ -1951,9 +1951,10 @@ static int io_poll_remove(struct io_kiocb *req, const struct io_uring_sqe *sqe) return 0; } -static void io_poll_complete(struct io_ring_ctx *ctx, struct io_kiocb *req, - __poll_t mask) +static void io_poll_complete(struct io_kiocb *req, __poll_t mask) { + struct io_ring_ctx *ctx = req->ctx; + req->poll.done = true; io_cqring_fill_event(req, mangle_poll(mask)); io_commit_cqring(ctx); @@ -1989,7 +1990,7 @@ static void io_poll_complete_work(struct io_wq_work **workptr) return; } list_del_init(&req->list); - io_poll_complete(ctx, req, mask); + io_poll_complete(req, mask); spin_unlock_irq(&ctx->completion_lock); io_cqring_ev_posted(ctx); @@ -2017,13 +2018,13 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync, if (mask && spin_trylock_irqsave(&ctx->completion_lock, flags)) { list_del(&req->list); - io_poll_complete(ctx, req, mask); + io_poll_complete(req, mask); spin_unlock_irqrestore(&ctx->completion_lock, flags); io_cqring_ev_posted(ctx); io_put_req(req, NULL); } else { - io_queue_async_work(ctx, req); + io_queue_async_work(req); } return 1; @@ -2108,7 +2109,7 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe, } if (mask) { /* no async, we'd stolen it */ ipt.error = 0; - io_poll_complete(ctx, req, mask); + io_poll_complete(req, mask); } spin_unlock_irq(&ctx->completion_lock); @@ -2355,12 +2356,13 @@ static int io_async_cancel(struct io_kiocb *req, const struct io_uring_sqe *sqe, return 0; } -static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req) +static int io_req_defer(struct io_kiocb *req) { const struct io_uring_sqe *sqe = req->submit.sqe; struct io_uring_sqe *sqe_copy; + struct io_ring_ctx *ctx = req->ctx; - if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list)) + if (!io_sequence_defer(req) && list_empty(&ctx->defer_list)) return 0; sqe_copy = kmalloc(sizeof(*sqe_copy), GFP_KERNEL); @@ -2368,7 +2370,7 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req) return -EAGAIN; spin_lock_irq(&ctx->completion_lock); - if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list)) { + if (!io_sequence_defer(req) && list_empty(&ctx->defer_list)) { spin_unlock_irq(&ctx->completion_lock); kfree(sqe_copy); return 0; @@ -2383,11 +2385,12 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req) return -EIOCBQUEUED; } -static int __io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req, - struct io_kiocb **nxt, bool force_nonblock) +static int __io_submit_sqe(struct io_kiocb *req, struct io_kiocb **nxt, + bool force_nonblock) { int ret, opcode; struct sqe_submit *s = &req->submit; + struct io_ring_ctx *ctx = req->ctx; opcode = READ_ONCE(s->sqe->opcode); switch (opcode) { @@ -2467,7 +2470,6 @@ static void io_wq_submit_work(struct io_wq_work **workptr) { struct io_wq_work *work = *workptr; struct io_kiocb *req = container_of(work, struct io_kiocb, work); - struct io_ring_ctx *ctx = req->ctx; struct sqe_submit *s = &req->submit; const struct io_uring_sqe *sqe = s->sqe; struct io_kiocb *nxt = NULL; @@ -2483,7 +2485,7 @@ static void io_wq_submit_work(struct io_wq_work **workptr) s->has_user = (work->flags & IO_WQ_WORK_HAS_MM) != 0; s->in_async = true; do { - ret = __io_submit_sqe(ctx, req, &nxt, false); + ret = __io_submit_sqe(req, &nxt, false); /* * We can get EAGAIN for polled IO even though we're * forcing a sync submission from here, since we can't @@ -2537,10 +2539,10 @@ static inline struct file *io_file_from_index(struct io_ring_ctx *ctx, return table->files[index & IORING_FILE_TABLE_MASK]; } -static int io_req_set_file(struct io_ring_ctx *ctx, - struct io_submit_state *state, struct io_kiocb *req) +static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req) { struct sqe_submit *s = &req->submit; + struct io_ring_ctx *ctx = req->ctx; unsigned flags; int fd; @@ -2580,9 +2582,10 @@ static int io_req_set_file(struct io_ring_ctx *ctx, return 0; } -static int io_grab_files(struct io_ring_ctx *ctx, struct io_kiocb *req) +static int io_grab_files(struct io_kiocb *req) { int ret = -EBADF; + struct io_ring_ctx *ctx = req->ctx; rcu_read_lock(); spin_lock_irq(&ctx->inflight_lock); @@ -2698,7 +2701,7 @@ static inline struct io_kiocb *io_get_linked_timeout(struct io_kiocb *req) return NULL; } -static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) +static int __io_queue_sqe(struct io_kiocb *req) { struct io_kiocb *nxt; int ret; @@ -2710,7 +2713,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) goto err; } - ret = __io_submit_sqe(ctx, req, NULL, true); + ret = __io_submit_sqe(req, NULL, true); /* * We async punt it if the file wasn't marked NOWAIT, or if the file @@ -2725,7 +2728,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) if (sqe_copy) { s->sqe = sqe_copy; if (req->work.flags & IO_WQ_WORK_NEEDS_FILES) { - ret = io_grab_files(ctx, req); + ret = io_grab_files(req); if (ret) { kfree(sqe_copy); goto err; @@ -2736,7 +2739,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) * Queued up for async execution, worker will release * submit reference when the iocb is actually submitted. */ - io_queue_async_work(ctx, req); + io_queue_async_work(req); return 0; } } @@ -2756,11 +2759,11 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) return ret; } -static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) +static int io_queue_sqe(struct io_kiocb *req) { int ret; - ret = io_req_defer(ctx, req); + ret = io_req_defer(req); if (ret) { if (ret != -EIOCBQUEUED) { io_cqring_add_event(req, ret); @@ -2769,17 +2772,17 @@ static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req) return 0; } - return __io_queue_sqe(ctx, req); + return __io_queue_sqe(req); } -static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req, - struct io_kiocb *shadow) +static int io_queue_link_head(struct io_kiocb *req, struct io_kiocb *shadow) { int ret; int need_submit = false; + struct io_ring_ctx *ctx = req->ctx; if (!shadow) - return io_queue_sqe(ctx, req); + return io_queue_sqe(req); /* * Mark the first IO in link list as DRAIN, let all the following @@ -2787,7 +2790,7 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req, * list. */ req->flags |= REQ_F_IO_DRAIN; - ret = io_req_defer(ctx, req); + ret = io_req_defer(req); if (ret) { if (ret != -EIOCBQUEUED) { io_cqring_add_event(req, ret); @@ -2810,18 +2813,19 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req, spin_unlock_irq(&ctx->completion_lock); if (need_submit) - return __io_queue_sqe(ctx, req); + return __io_queue_sqe(req); return 0; } #define SQE_VALID_FLAGS (IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK) -static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req, - struct io_submit_state *state, struct io_kiocb **link) +static void io_submit_sqe(struct io_kiocb *req, struct io_submit_state *state, + struct io_kiocb **link) { struct io_uring_sqe *sqe_copy; struct sqe_submit *s = &req->submit; + struct io_ring_ctx *ctx = req->ctx; int ret; req->user_data = s->sqe->user_data; @@ -2832,7 +2836,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req, goto err_req; } - ret = io_req_set_file(ctx, state, req); + ret = io_req_set_file(state, req); if (unlikely(ret)) { err_req: io_cqring_add_event(req, ret); @@ -2869,7 +2873,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req, ret = -EINVAL; goto err_req; } else { - io_queue_sqe(ctx, req); + io_queue_sqe(req); } } @@ -3018,7 +3022,7 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr, req->submit.needs_fixed_file = async; trace_io_uring_submit_sqe(ctx, req->submit.sqe->user_data, true, async); - io_submit_sqe(ctx, req, statep, &link); + io_submit_sqe(req, statep, &link); submitted++; /* @@ -3026,14 +3030,14 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr, * that's the end of the chain. Submit the previous link. */ if (!(sqe_flags & IOSQE_IO_LINK) && link) { - io_queue_link_head(ctx, link, shadow_req); + io_queue_link_head(link, shadow_req); link = NULL; shadow_req = NULL; } } if (link) - io_queue_link_head(ctx, link, shadow_req); + io_queue_link_head(link, shadow_req); if (statep) io_submit_state_end(&state);