Snap for 9769031 from 32991d5d99 to android13-5.15-keystone-qcom-release

Change-Id: I5f9e3d1946f114fa957983c1194bc8a45a498cb8

MAINTAINERS

@@ -7244,9 +7244,6 @@ F:	include/linux/fs.h
 F:	include/linux/fs_types.h
 F:	include/uapi/linux/fs.h
 F:	include/uapi/linux/openat2.h
-X:	fs/io-wq.c
-X:	fs/io-wq.h
-X:	fs/io_uring.c
 
 FINTEK F75375S HARDWARE MONITOR AND FAN CONTROLLER DRIVER
 M:	Riku Voipio <riku.voipio@iki.fi>
@@ -9825,9 +9822,7 @@ L:	io-uring@vger.kernel.org
 S:	Maintained
 T:	git git://git.kernel.dk/linux-block
 T:	git git://git.kernel.dk/liburing
-F:	fs/io-wq.c
-F:	fs/io-wq.h
-F:	fs/io_uring.c
+F:	io_uring/
 F:	include/linux/io_uring.h
 F:	include/uapi/linux/io_uring.h
 F:	tools/io_uring/

Makefile

@@ -1212,7 +1212,7 @@ endif
 	$(Q)$(MAKE) $(hdr-inst)=$(hdr-prefix)arch/$(SRCARCH)/include/uapi
 
 ifeq ($(KBUILD_EXTMOD),)
-core-y		+= kernel/ certs/ mm/ fs/ ipc/ security/ crypto/ block/
+core-y		+= kernel/ certs/ mm/ fs/ ipc/ security/ crypto/ block/ io_uring/
 
 vmlinux-dirs	:= $(patsubst %/,%,$(filter %/, \
 		     $(core-y) $(core-m) $(drivers-y) $(drivers-m) \

fs/Makefile

@@ -34,8 +34,6 @@ obj-$(CONFIG_TIMERFD)		+= timerfd.o
 obj-$(CONFIG_EVENTFD)		+= eventfd.o
 obj-$(CONFIG_USERFAULTFD)	+= userfaultfd.o
 obj-$(CONFIG_AIO)               += aio.o
-obj-$(CONFIG_IO_URING)		+= io_uring.o
-obj-$(CONFIG_IO_WQ)		+= io-wq.o
 obj-$(CONFIG_FS_DAX)		+= dax.o
 obj-$(CONFIG_FS_ENCRYPTION)	+= crypto/
 obj-$(CONFIG_FS_VERITY)		+= verity/

fs/eventfd.c

@@ -43,21 +43,7 @@ struct eventfd_ctx {
 	int id;
 };
 
-/**
- * eventfd_signal - Adds @n to the eventfd counter.
- * @ctx: [in] Pointer to the eventfd context.
- * @n: [in] Value of the counter to be added to the eventfd internal counter.
- *          The value cannot be negative.
- *
- * This function is supposed to be called by the kernel in paths that do not
- * allow sleeping. In this function we allow the counter to reach the ULLONG_MAX
- * value, and we signal this as overflow condition by returning a EPOLLERR
- * to poll(2).
- *
- * Returns the amount by which the counter was incremented.  This will be less
- * than @n if the counter has overflowed.
- */
-__u64 eventfd_signal(struct eventfd_ctx *ctx, __u64 n)
+__u64 eventfd_signal_mask(struct eventfd_ctx *ctx, __u64 n, unsigned mask)
 {
 	unsigned long flags;
 
@@ -78,12 +64,31 @@ __u64 eventfd_signal(struct eventfd_ctx *ctx, __u64 n)
 		n = ULLONG_MAX - ctx->count;
 	ctx->count += n;
 	if (waitqueue_active(&ctx->wqh))
-		wake_up_locked_poll(&ctx->wqh, EPOLLIN);
+		wake_up_locked_poll(&ctx->wqh, EPOLLIN | mask);
 	current->in_eventfd_signal = 0;
 	spin_unlock_irqrestore(&ctx->wqh.lock, flags);
 
 	return n;
 }
+
+/**
+ * eventfd_signal - Adds @n to the eventfd counter.
+ * @ctx: [in] Pointer to the eventfd context.
+ * @n: [in] Value of the counter to be added to the eventfd internal counter.
+ *          The value cannot be negative.
+ *
+ * This function is supposed to be called by the kernel in paths that do not
+ * allow sleeping. In this function we allow the counter to reach the ULLONG_MAX
+ * value, and we signal this as overflow condition by returning a EPOLLERR
+ * to poll(2).
+ *
+ * Returns the amount by which the counter was incremented.  This will be less
+ * than @n if the counter has overflowed.
+ */
+__u64 eventfd_signal(struct eventfd_ctx *ctx, __u64 n)
+{
+	return eventfd_signal_mask(ctx, n, 0);
+}
 EXPORT_SYMBOL_GPL(eventfd_signal);
 
 static void eventfd_free_ctx(struct eventfd_ctx *ctx)

fs/eventpoll.c

@@ -485,7 +485,8 @@ static inline void ep_set_busy_poll_napi_id(struct epitem *epi)
  */
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 
-static void ep_poll_safewake(struct eventpoll *ep, struct epitem *epi)
+static void ep_poll_safewake(struct eventpoll *ep, struct epitem *epi,
+			     unsigned pollflags)
 {
 	struct eventpoll *ep_src;
 	unsigned long flags;
@@ -516,16 +517,17 @@ static void ep_poll_safewake(struct eventpoll *ep, struct epitem *epi)
 	}
 	spin_lock_irqsave_nested(&ep->poll_wait.lock, flags, nests);
 	ep->nests = nests + 1;
-	wake_up_locked_poll(&ep->poll_wait, EPOLLIN);
+	wake_up_locked_poll(&ep->poll_wait, EPOLLIN | pollflags);
 	ep->nests = 0;
 	spin_unlock_irqrestore(&ep->poll_wait.lock, flags);
 }
 
 #else
 
-static void ep_poll_safewake(struct eventpoll *ep, struct epitem *epi)
+static void ep_poll_safewake(struct eventpoll *ep, struct epitem *epi,
+			     unsigned pollflags)
 {
-	wake_up_poll(&ep->poll_wait, EPOLLIN);
+	wake_up_poll(&ep->poll_wait, EPOLLIN | pollflags);
 }
 
 #endif
@@ -736,7 +738,7 @@ static void ep_free(struct eventpoll *ep)
 
 	/* We need to release all tasks waiting for these file */
 	if (waitqueue_active(&ep->poll_wait))
-		ep_poll_safewake(ep, NULL);
+		ep_poll_safewake(ep, NULL, 0);
 
 	/*
 	 * We need to lock this because we could be hit by
@@ -1202,7 +1204,7 @@ static int ep_poll_callback(wait_queue_entry_t *wait, unsigned mode, int sync, v
 
 	/* We have to call this outside the lock */
 	if (pwake)
-		ep_poll_safewake(ep, epi);
+		ep_poll_safewake(ep, epi, pollflags & EPOLL_URING_WAKE);
 
 	if (!(epi->event.events & EPOLLEXCLUSIVE))
 		ewake = 1;
@@ -1547,7 +1549,7 @@ static int ep_insert(struct eventpoll *ep, const struct epoll_event *event,
 
 	/* We have to call this outside the lock */
 	if (pwake)
-		ep_poll_safewake(ep, NULL);
+		ep_poll_safewake(ep, NULL, 0);
 
 	return 0;
 }
@@ -1623,7 +1625,7 @@ static int ep_modify(struct eventpoll *ep, struct epitem *epi,
 
 	/* We have to call this outside the lock */
 	if (pwake)
-		ep_poll_safewake(ep, NULL);
+		ep_poll_safewake(ep, NULL, 0);
 
 	return 0;
 }

include/linux/eventfd.h

@@ -40,6 +40,7 @@ struct file *eventfd_fget(int fd);
 struct eventfd_ctx *eventfd_ctx_fdget(int fd);
 struct eventfd_ctx *eventfd_ctx_fileget(struct file *file);
 __u64 eventfd_signal(struct eventfd_ctx *ctx, __u64 n);
+__u64 eventfd_signal_mask(struct eventfd_ctx *ctx, __u64 n, unsigned mask);
 int eventfd_ctx_remove_wait_queue(struct eventfd_ctx *ctx, wait_queue_entry_t *wait,
 				  __u64 *cnt);
 void eventfd_ctx_do_read(struct eventfd_ctx *ctx, __u64 *cnt);
@@ -66,6 +67,12 @@ static inline int eventfd_signal(struct eventfd_ctx *ctx, int n)
 	return -ENOSYS;
 }
 
+static inline int eventfd_signal_mask(struct eventfd_ctx *ctx, __u64 n,
+				      unsigned mask)
+{
+	return -ENOSYS;
+}
+
 static inline void eventfd_ctx_put(struct eventfd_ctx *ctx)
 {
 

include/uapi/linux/eventpoll.h

@@ -41,6 +41,12 @@
 #define EPOLLMSG	(__force __poll_t)0x00000400
 #define EPOLLRDHUP	(__force __poll_t)0x00002000
 
+/*
+ * Internal flag - wakeup generated by io_uring, used to detect recursion back
+ * into the io_uring poll handler.
+ */
+#define EPOLL_URING_WAKE	((__force __poll_t)(1U << 27))
+
 /* Set exclusive wakeup mode for the target file descriptor */
 #define EPOLLEXCLUSIVE	((__force __poll_t)(1U << 28))
 

io_uring/Makefile

@@ -0,0 +1,6 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for io_uring
+
+obj-$(CONFIG_IO_URING)		+= io_uring.o
+obj-$(CONFIG_IO_WQ)		+= io-wq.o

io_uring/io-wq.c

@@ -513,7 +513,7 @@ static struct io_wq_work *io_get_next_work(struct io_wqe_acct *acct,
 
 static bool io_flush_signals(void)
 {
-	if (unlikely(test_thread_flag(TIF_NOTIFY_SIGNAL))) {
+	if (test_thread_flag(TIF_NOTIFY_SIGNAL) || current->task_works) {
 		__set_current_state(TASK_RUNNING);
 		tracehook_notify_signal();
 		return true;
@@ -1217,6 +1217,12 @@ static void io_wq_cancel_tw_create(struct io_wq *wq)
 
 		worker = container_of(cb, struct io_worker, create_work);
 		io_worker_cancel_cb(worker);
+		/*
+		 * Only the worker continuation helper has worker allocated and
+		 * hence needs freeing.
+		 */
+		if (cb->func == create_worker_cont)
+			kfree(worker);
 	}
 }
 

io_uring/io_uring.c

@@ -85,7 +85,7 @@
 
 #include <uapi/linux/io_uring.h>
 
-#include "internal.h"
+#include "../fs/internal.h"
 #include "io-wq.h"
 
 #define IORING_MAX_ENTRIES	32768
@@ -578,6 +578,7 @@ struct io_sr_msg {
 	int				msg_flags;
 	int				bgid;
 	size_t				len;
+	size_t				done_io;
 	struct io_buffer		*kbuf;
 };
 
@@ -739,6 +740,7 @@ enum {
 	REQ_F_CREDS_BIT,
 	REQ_F_REFCOUNT_BIT,
 	REQ_F_ARM_LTIMEOUT_BIT,
+	REQ_F_PARTIAL_IO_BIT,
 	/* keep async read/write and isreg together and in order */
 	REQ_F_NOWAIT_READ_BIT,
 	REQ_F_NOWAIT_WRITE_BIT,
@@ -794,6 +796,8 @@ enum {
 	REQ_F_REFCOUNT		= BIT(REQ_F_REFCOUNT_BIT),
 	/* there is a linked timeout that has to be armed */
 	REQ_F_ARM_LTIMEOUT	= BIT(REQ_F_ARM_LTIMEOUT_BIT),
+	/* request has already done partial IO */
+	REQ_F_PARTIAL_IO	= BIT(REQ_F_PARTIAL_IO_BIT),
 };
 
 struct async_poll {
@@ -1091,7 +1095,8 @@ static int __io_register_rsrc_update(struct io_ring_ctx *ctx, unsigned type,
 				     unsigned nr_args);
 static void io_clean_op(struct io_kiocb *req);
 static struct file *io_file_get(struct io_ring_ctx *ctx,
-				struct io_kiocb *req, int fd, bool fixed);
+				struct io_kiocb *req, int fd, bool fixed,
+				unsigned int issue_flags);
 static void __io_queue_sqe(struct io_kiocb *req);
 static void io_rsrc_put_work(struct work_struct *work);
 
@@ -1629,13 +1634,15 @@ static void io_cqring_ev_posted(struct io_ring_ctx *ctx)
 	 * wake as many waiters as we need to.
 	 */
 	if (wq_has_sleeper(&ctx->cq_wait))
-		wake_up_all(&ctx->cq_wait);
+		__wake_up(&ctx->cq_wait, TASK_NORMAL, 0,
+				poll_to_key(EPOLL_URING_WAKE | EPOLLIN));
 	if (ctx->sq_data && waitqueue_active(&ctx->sq_data->wait))
 		wake_up(&ctx->sq_data->wait);
 	if (io_should_trigger_evfd(ctx))
-		eventfd_signal(ctx->cq_ev_fd, 1);
+		eventfd_signal_mask(ctx->cq_ev_fd, 1, EPOLL_URING_WAKE);
 	if (waitqueue_active(&ctx->poll_wait))
-		wake_up_interruptible(&ctx->poll_wait);
+		__wake_up(&ctx->poll_wait, TASK_INTERRUPTIBLE, 0,
+				poll_to_key(EPOLL_URING_WAKE | EPOLLIN));
 }
 
 static void io_cqring_ev_posted_iopoll(struct io_ring_ctx *ctx)
@@ -1645,12 +1652,14 @@ static void io_cqring_ev_posted_iopoll(struct io_ring_ctx *ctx)
 
 	if (ctx->flags & IORING_SETUP_SQPOLL) {
 		if (waitqueue_active(&ctx->cq_wait))
-			wake_up_all(&ctx->cq_wait);
+			__wake_up(&ctx->cq_wait, TASK_NORMAL, 0,
+				  poll_to_key(EPOLL_URING_WAKE | EPOLLIN));
 	}
 	if (io_should_trigger_evfd(ctx))
-		eventfd_signal(ctx->cq_ev_fd, 1);
+		eventfd_signal_mask(ctx->cq_ev_fd, 1, EPOLL_URING_WAKE);
 	if (waitqueue_active(&ctx->poll_wait))
-		wake_up_interruptible(&ctx->poll_wait);
+		__wake_up(&ctx->poll_wait, TASK_INTERRUPTIBLE, 0,
+				poll_to_key(EPOLL_URING_WAKE | EPOLLIN));
 }
 
 /* Returns true if there are no backlogged entries after the flush */
@@ -2477,12 +2486,26 @@ static void io_iopoll_complete(struct io_ring_ctx *ctx, unsigned int *nr_events,
 
 	io_init_req_batch(&rb);
 	while (!list_empty(done)) {
+		struct io_uring_cqe *cqe;
+		unsigned cflags;
+
 		req = list_first_entry(done, struct io_kiocb, inflight_entry);
 		list_del(&req->inflight_entry);
-
-		io_fill_cqe_req(req, req->result, io_put_rw_kbuf(req));
+		cflags = io_put_rw_kbuf(req);
 		(*nr_events)++;
 
+		cqe = io_get_cqe(ctx);
+		if (cqe) {
+			WRITE_ONCE(cqe->user_data, req->user_data);
+			WRITE_ONCE(cqe->res, req->result);
+			WRITE_ONCE(cqe->flags, cflags);
+		} else {
+			spin_lock(&ctx->completion_lock);
+			io_cqring_event_overflow(ctx, req->user_data,
+							req->result, cflags);
+			spin_unlock(&ctx->completion_lock);
+		}
+
 		if (req_ref_put_and_test(req))
 			io_req_free_batch(&rb, req, &ctx->submit_state);
 	}
@@ -2681,17 +2704,32 @@ static bool io_rw_should_reissue(struct io_kiocb *req)
 }
 #endif
 
-static bool __io_complete_rw_common(struct io_kiocb *req, long res)
+/*
+ * Trigger the notifications after having done some IO, and finish the write
+ * accounting, if any.
+ */
+static void io_req_io_end(struct io_kiocb *req)
 {
-	if (req->rw.kiocb.ki_flags & IOCB_WRITE) {
+	struct io_rw *rw = &req->rw;
+
+	if (rw->kiocb.ki_flags & IOCB_WRITE) {
 		kiocb_end_write(req);
 		fsnotify_modify(req->file);
 	} else {
 		fsnotify_access(req->file);
 	}
+}
+
+static bool __io_complete_rw_common(struct io_kiocb *req, long res)
+{
 	if (res != req->result) {
 		if ((res == -EAGAIN || res == -EOPNOTSUPP) &&
 		    io_rw_should_reissue(req)) {
+			/*
+			 * Reissue will start accounting again, finish the
+			 * current cycle.
+			 */
+			io_req_io_end(req);
 			req->flags |= REQ_F_REISSUE;
 			return true;
 		}
@@ -2701,7 +2739,7 @@ static bool __io_complete_rw_common(struct io_kiocb *req, long res)
 	return false;
 }
 
-static inline int io_fixup_rw_res(struct io_kiocb *req, unsigned res)
+static inline int io_fixup_rw_res(struct io_kiocb *req, long res)
 {
 	struct io_async_rw *io = req->async_data;
 
@@ -2733,12 +2771,10 @@ static void io_req_task_complete(struct io_kiocb *req, bool *locked)
 	}
 }
 
-static void __io_complete_rw(struct io_kiocb *req, long res, long res2,
-			     unsigned int issue_flags)
+static void io_req_rw_complete(struct io_kiocb *req, bool *locked)
 {
-	if (__io_complete_rw_common(req, res))
-		return;
-	__io_req_complete(req, issue_flags, io_fixup_rw_res(req, res), io_put_rw_kbuf(req));
+	io_req_io_end(req);
+	io_req_task_complete(req, locked);
 }
 
 static void io_complete_rw(struct kiocb *kiocb, long res, long res2)
@@ -2748,7 +2784,7 @@ static void io_complete_rw(struct kiocb *kiocb, long res, long res2)
 	if (__io_complete_rw_common(req, res))
 		return;
 	req->result = io_fixup_rw_res(req, res);
-	req->io_task_work.func = io_req_task_complete;
+	req->io_task_work.func = io_req_rw_complete;
 	io_req_task_work_add(req);
 }
 
@@ -2900,14 +2936,6 @@ static int io_prep_rw(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 		req->flags |= REQ_F_ISREG;
 
 	kiocb->ki_pos = READ_ONCE(sqe->off);
-	if (kiocb->ki_pos == -1) {
-		if (!(file->f_mode & FMODE_STREAM)) {
-			req->flags |= REQ_F_CUR_POS;
-			kiocb->ki_pos = file->f_pos;
-		} else {
-			kiocb->ki_pos = 0;
-		}
-	}
 	kiocb->ki_hint = ki_hint_validate(file_write_hint(kiocb->ki_filp));
 	kiocb->ki_flags = iocb_flags(kiocb->ki_filp);
 	ret = kiocb_set_rw_flags(kiocb, READ_ONCE(sqe->rw_flags));
@@ -2989,6 +3017,23 @@ static inline void io_rw_done(struct kiocb *kiocb, ssize_t ret)
 	}
 }
 
+static inline loff_t *io_kiocb_update_pos(struct io_kiocb *req)
+{
+	struct kiocb *kiocb = &req->rw.kiocb;
+
+	if (kiocb->ki_pos != -1)
+		return &kiocb->ki_pos;
+
+	if (!(req->file->f_mode & FMODE_STREAM)) {
+		req->flags |= REQ_F_CUR_POS;
+		kiocb->ki_pos = req->file->f_pos;
+		return &kiocb->ki_pos;
+	}
+
+	kiocb->ki_pos = 0;
+	return NULL;
+}
+
 static void kiocb_done(struct kiocb *kiocb, ssize_t ret,
 		       unsigned int issue_flags)
 {
@@ -2996,10 +3041,20 @@ static void kiocb_done(struct kiocb *kiocb, ssize_t ret,
 
 	if (req->flags & REQ_F_CUR_POS)
 		req->file->f_pos = kiocb->ki_pos;
-	if (ret >= 0 && (kiocb->ki_complete == io_complete_rw))
-		__io_complete_rw(req, ret, 0, issue_flags);
-	else
+	if (ret >= 0 && (kiocb->ki_complete == io_complete_rw)) {
+		if (!__io_complete_rw_common(req, ret)) {
+			/*
+			 * Safe to call io_end from here as we're inline
+			 * from the submission path.
+			 */
+			io_req_io_end(req);
+			__io_req_complete(req, issue_flags,
+					  io_fixup_rw_res(req, ret),
+					  io_put_rw_kbuf(req));
+		}
+	} else {
 		io_rw_done(kiocb, ret);
+	}
 
 	if (req->flags & REQ_F_REISSUE) {
 		req->flags &= ~REQ_F_REISSUE;
@@ -3281,6 +3336,7 @@ static ssize_t loop_rw_iter(int rw, struct io_kiocb *req, struct iov_iter *iter)
 	struct kiocb *kiocb = &req->rw.kiocb;
 	struct file *file = req->file;
 	ssize_t ret = 0;
+	loff_t *ppos;
 
 	/*
 	 * Don't support polled IO through this interface, and we can't
@@ -3292,6 +3348,8 @@ static ssize_t loop_rw_iter(int rw, struct io_kiocb *req, struct iov_iter *iter)
 	if (kiocb->ki_flags & IOCB_NOWAIT)
 		return -EAGAIN;
 
+	ppos = io_kiocb_ppos(kiocb);
+
 	while (iov_iter_count(iter)) {
 		struct iovec iovec;
 		ssize_t nr;
@@ -3305,10 +3363,10 @@ static ssize_t loop_rw_iter(int rw, struct io_kiocb *req, struct iov_iter *iter)
 
 		if (rw == READ) {
 			nr = file->f_op->read(file, iovec.iov_base,
-					      iovec.iov_len, io_kiocb_ppos(kiocb));
+					      iovec.iov_len, ppos);
 		} else {
 			nr = file->f_op->write(file, iovec.iov_base,
-					       iovec.iov_len, io_kiocb_ppos(kiocb));
+					       iovec.iov_len, ppos);
 		}
 
 		if (nr < 0) {
@@ -3509,6 +3567,7 @@ static int io_read(struct io_kiocb *req, unsigned int issue_flags)
 	bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK;
 	struct iov_iter_state __state, *state;
 	ssize_t ret, ret2;
+	loff_t *ppos;
 
 	if (rw) {
 		iter = &rw->iter;
@@ -3541,7 +3600,9 @@ static int io_read(struct io_kiocb *req, unsigned int issue_flags)
 		return ret ?: -EAGAIN;
 	}
 
-	ret = rw_verify_area(READ, req->file, io_kiocb_ppos(kiocb), req->result);
+	ppos = io_kiocb_update_pos(req);
+
+	ret = rw_verify_area(READ, req->file, ppos, req->result);
 	if (unlikely(ret)) {
 		kfree(iovec);
 		return ret;
@@ -3645,6 +3706,7 @@ static int io_write(struct io_kiocb *req, unsigned int issue_flags)
 	bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK;
 	struct iov_iter_state __state, *state;
 	ssize_t ret, ret2;
+	loff_t *ppos;
 
 	if (rw) {
 		iter = &rw->iter;
@@ -3675,7 +3737,9 @@ static int io_write(struct io_kiocb *req, unsigned int issue_flags)
 	    (req->flags & REQ_F_ISREG))
 		goto copy_iov;
 
-	ret = rw_verify_area(WRITE, req->file, io_kiocb_ppos(kiocb), req->result);
+	ppos = io_kiocb_update_pos(req);
+
+	ret = rw_verify_area(WRITE, req->file, ppos, req->result);
 	if (unlikely(ret))
 		goto out_free;
 
@@ -4058,7 +4122,7 @@ static int io_tee(struct io_kiocb *req, unsigned int issue_flags)
 		return -EAGAIN;
 
 	in = io_file_get(req->ctx, req, sp->splice_fd_in,
-				  (sp->flags & SPLICE_F_FD_IN_FIXED));
+			 (sp->flags & SPLICE_F_FD_IN_FIXED), issue_flags);
 	if (!in) {
 		ret = -EBADF;
 		goto done;
@@ -4098,7 +4162,7 @@ static int io_splice(struct io_kiocb *req, unsigned int issue_flags)
 		return -EAGAIN;
 
 	in = io_file_get(req->ctx, req, sp->splice_fd_in,
-				  (sp->flags & SPLICE_F_FD_IN_FIXED));
+			 (sp->flags & SPLICE_F_FD_IN_FIXED), issue_flags);
 	if (!in) {
 		ret = -EBADF;
 		goto done;
@@ -4755,6 +4819,13 @@ static int io_sync_file_range(struct io_kiocb *req, unsigned int issue_flags)
 }
 
 #if defined(CONFIG_NET)
+static bool io_net_retry(struct socket *sock, int flags)
+{
+	if (!(flags & MSG_WAITALL))
+		return false;
+	return sock->type == SOCK_STREAM || sock->type == SOCK_SEQPACKET;
+}
+
 static int io_setup_async_msg(struct io_kiocb *req,
 			      struct io_async_msghdr *kmsg)
 {
@@ -4772,8 +4843,10 @@ static int io_setup_async_msg(struct io_kiocb *req,
 	if (async_msg->msg.msg_name)
 		async_msg->msg.msg_name = &async_msg->addr;
 	/* if were using fast_iov, set it to the new one */
-	if (!async_msg->free_iov)
-		async_msg->msg.msg_iter.iov = async_msg->fast_iov;
+	if (!kmsg->free_iov) {
+		size_t fast_idx = kmsg->msg.msg_iter.iov - kmsg->fast_iov;
+		async_msg->msg.msg_iter.iov = &async_msg->fast_iov[fast_idx];
+	}
 
 	return -EAGAIN;
 }
@@ -4818,12 +4891,14 @@ static int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	if (req->ctx->compat)
 		sr->msg_flags |= MSG_CMSG_COMPAT;
 #endif
+	sr->done_io = 0;
 	return 0;
 }
 
 static int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct io_async_msghdr iomsg, *kmsg;
+	struct io_sr_msg *sr = &req->sr_msg;
 	struct socket *sock;
 	unsigned flags;
 	int min_ret = 0;
@@ -4848,17 +4923,27 @@ static int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 		min_ret = iov_iter_count(&kmsg->msg.msg_iter);
 
 	ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags);
-	if ((issue_flags & IO_URING_F_NONBLOCK) && ret == -EAGAIN)
-		return io_setup_async_msg(req, kmsg);
-	if (ret == -ERESTARTSYS)
-		ret = -EINTR;
 
+	if (ret < min_ret) {
+		if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK))
+			return io_setup_async_msg(req, kmsg);
+		if (ret == -ERESTARTSYS)
+			ret = -EINTR;
+		if (ret > 0 && io_net_retry(sock, flags)) {
+			sr->done_io += ret;
+			req->flags |= REQ_F_PARTIAL_IO;
+			return io_setup_async_msg(req, kmsg);
+		}
+		req_set_fail(req);
+	}
 	/* fast path, check for non-NULL to avoid function call */
 	if (kmsg->free_iov)
 		kfree(kmsg->free_iov);
 	req->flags &= ~REQ_F_NEED_CLEANUP;
-	if (ret < min_ret)
-		req_set_fail(req);
+	if (ret >= 0)
+		ret += sr->done_io;
+	else if (sr->done_io)
+		ret = sr->done_io;
 	__io_req_complete(req, issue_flags, ret, 0);
 	return 0;
 }
@@ -4894,13 +4979,24 @@ static int io_send(struct io_kiocb *req, unsigned int issue_flags)
 
 	msg.msg_flags = flags;
 	ret = sock_sendmsg(sock, &msg);
-	if ((issue_flags & IO_URING_F_NONBLOCK) && ret == -EAGAIN)
-		return -EAGAIN;
-	if (ret == -ERESTARTSYS)
-		ret = -EINTR;
-
-	if (ret < min_ret)
+	if (ret < min_ret) {
+		if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK))
+			return -EAGAIN;
+		if (ret == -ERESTARTSYS)
+			ret = -EINTR;
+		if (ret > 0 && io_net_retry(sock, flags)) {
+			sr->len -= ret;
+			sr->buf += ret;
+			sr->done_io += ret;
+			req->flags |= REQ_F_PARTIAL_IO;
+			return -EAGAIN;
+		}
 		req_set_fail(req);
+	}
+	if (ret >= 0)
+		ret += sr->done_io;
+	else if (sr->done_io)
+		ret = sr->done_io;
 	__io_req_complete(req, issue_flags, ret, 0);
 	return 0;
 }
@@ -5044,12 +5140,14 @@ static int io_recvmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	if (req->ctx->compat)
 		sr->msg_flags |= MSG_CMSG_COMPAT;
 #endif
+	sr->done_io = 0;
 	return 0;
 }
 
 static int io_recvmsg(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct io_async_msghdr iomsg, *kmsg;
+	struct io_sr_msg *sr = &req->sr_msg;
 	struct socket *sock;
 	struct io_buffer *kbuf;
 	unsigned flags;
@@ -5087,10 +5185,20 @@ static int io_recvmsg(struct io_kiocb *req, unsigned int issue_flags)
 
 	ret = __sys_recvmsg_sock(sock, &kmsg->msg, req->sr_msg.umsg,
 					kmsg->uaddr, flags);
-	if (force_nonblock && ret == -EAGAIN)
-		return io_setup_async_msg(req, kmsg);
-	if (ret == -ERESTARTSYS)
-		ret = -EINTR;
+	if (ret < min_ret) {
+		if (ret == -EAGAIN && force_nonblock)
+			return io_setup_async_msg(req, kmsg);
+		if (ret == -ERESTARTSYS)
+			ret = -EINTR;
+		if (ret > 0 && io_net_retry(sock, flags)) {
+			sr->done_io += ret;
+			req->flags |= REQ_F_PARTIAL_IO;
+			return io_setup_async_msg(req, kmsg);
+		}
+		req_set_fail(req);
+	} else if ((flags & MSG_WAITALL) && (kmsg->msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))) {
+		req_set_fail(req);
+	}
 
 	if (req->flags & REQ_F_BUFFER_SELECTED)
 		cflags = io_put_recv_kbuf(req);
@@ -5098,8 +5206,10 @@ static int io_recvmsg(struct io_kiocb *req, unsigned int issue_flags)
 	if (kmsg->free_iov)
 		kfree(kmsg->free_iov);
 	req->flags &= ~REQ_F_NEED_CLEANUP;
-	if (ret < min_ret || ((flags & MSG_WAITALL) && (kmsg->msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))))
-		req_set_fail(req);
+	if (ret >= 0)
+		ret += sr->done_io;
+	else if (sr->done_io)
+		ret = sr->done_io;
 	__io_req_complete(req, issue_flags, ret, cflags);
 	return 0;
 }
@@ -5146,15 +5256,29 @@ static int io_recv(struct io_kiocb *req, unsigned int issue_flags)
 		min_ret = iov_iter_count(&msg.msg_iter);
 
 	ret = sock_recvmsg(sock, &msg, flags);
-	if (force_nonblock && ret == -EAGAIN)
-		return -EAGAIN;
-	if (ret == -ERESTARTSYS)
-		ret = -EINTR;
+	if (ret < min_ret) {
+		if (ret == -EAGAIN && force_nonblock)
+			return -EAGAIN;
+		if (ret == -ERESTARTSYS)
+			ret = -EINTR;
+		if (ret > 0 && io_net_retry(sock, flags)) {
+			sr->len -= ret;
+			sr->buf += ret;
+			sr->done_io += ret;
+			req->flags |= REQ_F_PARTIAL_IO;
+			return -EAGAIN;
+		}
+		req_set_fail(req);
+	} else if ((flags & MSG_WAITALL) && (msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))) {
 out_free:
+		req_set_fail(req);
+	}
 	if (req->flags & REQ_F_BUFFER_SELECTED)
 		cflags = io_put_recv_kbuf(req);
-	if (ret < min_ret || ((flags & MSG_WAITALL) && (msg.msg_flags & (MSG_TRUNC | MSG_CTRUNC))))
-		req_set_fail(req);
+	if (ret >= 0)
+		ret += sr->done_io;
+	else if (sr->done_io)
+		ret = sr->done_io;
 	__io_req_complete(req, issue_flags, ret, cflags);
 	return 0;
 }
@@ -5192,9 +5316,6 @@ static int io_accept(struct io_kiocb *req, unsigned int issue_flags)
 	struct file *file;
 	int ret, fd;
 
-	if (req->file->f_flags & O_NONBLOCK)
-		req->flags |= REQ_F_NOWAIT;
-
 	if (!fixed) {
 		fd = __get_unused_fd_flags(accept->flags, accept->nofile);
 		if (unlikely(fd < 0))
@@ -5206,6 +5327,8 @@ static int io_accept(struct io_kiocb *req, unsigned int issue_flags)
 		if (!fixed)
 			put_unused_fd(fd);
 		ret = PTR_ERR(file);
+		/* safe to retry */
+		req->flags |= REQ_F_PARTIAL_IO;
 		if (ret == -EAGAIN && force_nonblock)
 			return -EAGAIN;
 		if (ret == -ERESTARTSYS)
@@ -5322,7 +5445,29 @@ struct io_poll_table {
 };
 
 #define IO_POLL_CANCEL_FLAG	BIT(31)
-#define IO_POLL_REF_MASK	GENMASK(30, 0)
+#define IO_POLL_RETRY_FLAG	BIT(30)
+#define IO_POLL_REF_MASK	GENMASK(29, 0)
+
+/*
+ * We usually have 1-2 refs taken, 128 is more than enough and we want to
+ * maximise the margin between this amount and the moment when it overflows.
+ */
+#define IO_POLL_REF_BIAS       128
+
+static bool io_poll_get_ownership_slowpath(struct io_kiocb *req)
+{
+	int v;
+
+	/*
+	 * poll_refs are already elevated and we don't have much hope for
+	 * grabbing the ownership. Instead of incrementing set a retry flag
+	 * to notify the loop that there might have been some change.
+	 */
+	v = atomic_fetch_or(IO_POLL_RETRY_FLAG, &req->poll_refs);
+	if (v & IO_POLL_REF_MASK)
+		return false;
+	return !(atomic_fetch_inc(&req->poll_refs) & IO_POLL_REF_MASK);
+}
 
 /*
  * If refs part of ->poll_refs (see IO_POLL_REF_MASK) is 0, it's free. We can
@@ -5332,6 +5477,8 @@ struct io_poll_table {
  */
 static inline bool io_poll_get_ownership(struct io_kiocb *req)
 {
+	if (unlikely(atomic_read(&req->poll_refs) >= IO_POLL_REF_BIAS))
+		return io_poll_get_ownership_slowpath(req);
 	return !(atomic_fetch_inc(&req->poll_refs) & IO_POLL_REF_MASK);
 }
 
@@ -5440,6 +5587,23 @@ static int io_poll_check_events(struct io_kiocb *req)
 			return 0;
 		if (v & IO_POLL_CANCEL_FLAG)
 			return -ECANCELED;
+		/*
+		 * cqe.res contains only events of the first wake up
+		 * and all others are be lost. Redo vfs_poll() to get
+		 * up to date state.
+		 */
+		if ((v & IO_POLL_REF_MASK) != 1)
+			req->result = 0;
+		if (v & IO_POLL_RETRY_FLAG) {
+			req->result = 0;
+			/*
+			 * We won't find new events that came in between
+			 * vfs_poll and the ref put unless we clear the
+			 * flag in advance.
+			 */
+			atomic_andnot(IO_POLL_RETRY_FLAG, &req->poll_refs);
+			v &= ~IO_POLL_RETRY_FLAG;
+		}
 
 		if (!req->result) {
 			struct poll_table_struct pt = { ._key = poll->events };
@@ -5464,11 +5628,15 @@ static int io_poll_check_events(struct io_kiocb *req)
 			return 0;
 		}
 
+		/* force the next iteration to vfs_poll() */
+		req->result = 0;
+
 		/*
 		 * Release all references, retry if someone tried to restart
 		 * task_work while we were executing it.
 		 */
-	} while (atomic_sub_return(v & IO_POLL_REF_MASK, &req->poll_refs));
+	} while (atomic_sub_return(v & IO_POLL_REF_MASK, &req->poll_refs) &
+					IO_POLL_REF_MASK);
 
 	return 1;
 }
@@ -5577,8 +5745,17 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
 	if (mask && !(mask & poll->events))
 		return 0;
 
-	if (io_poll_get_ownership(req))
+	if (io_poll_get_ownership(req)) {
+		/*
+		 * If we trigger a multishot poll off our own wakeup path,
+		 * disable multishot as there is a circular dependency between
+		 * CQ posting and triggering the event.
+		 */
+		if (mask & EPOLL_URING_WAKE)
+			poll->events |= EPOLLONESHOT;
+
 		__io_poll_execute(req, mask);
+	}
 	return 1;
 }
 
@@ -5640,7 +5817,6 @@ static int __io_arm_poll_handler(struct io_kiocb *req,
 				 struct io_poll_table *ipt, __poll_t mask)
 {
 	struct io_ring_ctx *ctx = req->ctx;
-	int v;
 
 	INIT_HLIST_NODE(&req->hash_node);
 	io_init_poll_iocb(poll, mask, io_poll_wake);
@@ -5686,11 +5862,10 @@ static int __io_arm_poll_handler(struct io_kiocb *req,
 	}
 
 	/*
-	 * Release ownership. If someone tried to queue a tw while it was
-	 * locked, kick it off for them.
+	 * Try to release ownership. If we see a change of state, e.g.
+	 * poll was waken up, queue up a tw, it'll deal with it.
 	 */
-	v = atomic_dec_return(&req->poll_refs);
-	if (unlikely(v & IO_POLL_REF_MASK))
+	if (atomic_cmpxchg(&req->poll_refs, 1, 0) != 1)
 		__io_poll_execute(req, 0);
 	return 0;
 }
@@ -5721,7 +5896,7 @@ static int io_arm_poll_handler(struct io_kiocb *req)
 
 	if (!req->file || !file_can_poll(req->file))
 		return IO_APOLL_ABORTED;
-	if (req->flags & REQ_F_POLLED)
+	if ((req->flags & (REQ_F_POLLED|REQ_F_PARTIAL_IO)) == REQ_F_POLLED)
 		return IO_APOLL_ABORTED;
 	if (!def->pollin && !def->pollout)
 		return IO_APOLL_ABORTED;
@@ -5737,7 +5912,12 @@ static int io_arm_poll_handler(struct io_kiocb *req)
 		mask |= POLLOUT | POLLWRNORM;
 	}
 
-	apoll = kmalloc(sizeof(*apoll), GFP_ATOMIC);
+	if (req->flags & REQ_F_POLLED) {
+		apoll = req->apoll;
+		kfree(apoll->double_poll);
+	} else {
+		apoll = kmalloc(sizeof(*apoll), GFP_ATOMIC);
+	}
 	if (unlikely(!apoll))
 		return IO_APOLL_ABORTED;
 	apoll->double_poll = NULL;
@@ -6868,13 +7048,16 @@ static void io_fixed_file_set(struct io_fixed_file *file_slot, struct file *file
 }
 
 static inline struct file *io_file_get_fixed(struct io_ring_ctx *ctx,
-					     struct io_kiocb *req, int fd)
+					     struct io_kiocb *req, int fd,
+					     unsigned int issue_flags)
 {
-	struct file *file;
+	struct file *file = NULL;
 	unsigned long file_ptr;
 
+	io_ring_submit_lock(ctx, !(issue_flags & IO_URING_F_NONBLOCK));
+
 	if (unlikely((unsigned int)fd >= ctx->nr_user_files))
-		return NULL;
+		goto out;
 	fd = array_index_nospec(fd, ctx->nr_user_files);
 	file_ptr = io_fixed_file_slot(&ctx->file_table, fd)->file_ptr;
 	file = (struct file *) (file_ptr & FFS_MASK);
@@ -6882,6 +7065,8 @@ static inline struct file *io_file_get_fixed(struct io_ring_ctx *ctx,
 	/* mask in overlapping REQ_F and FFS bits */
 	req->flags |= (file_ptr << REQ_F_NOWAIT_READ_BIT);
 	io_req_set_rsrc_node(req);
+out:
+	io_ring_submit_unlock(ctx, !(issue_flags & IO_URING_F_NONBLOCK));
 	return file;
 }
 
@@ -6899,10 +7084,11 @@ static struct file *io_file_get_normal(struct io_ring_ctx *ctx,
 }
 
 static inline struct file *io_file_get(struct io_ring_ctx *ctx,
-				       struct io_kiocb *req, int fd, bool fixed)
+				       struct io_kiocb *req, int fd, bool fixed,
+				       unsigned int issue_flags)
 {
 	if (fixed)
-		return io_file_get_fixed(ctx, req, fd);
+		return io_file_get_fixed(ctx, req, fd, issue_flags);
 	else
 		return io_file_get_normal(ctx, req, fd);
 }
@@ -7124,7 +7310,8 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
 
 	if (io_op_defs[req->opcode].needs_file) {
 		req->file = io_file_get(ctx, req, READ_ONCE(sqe->fd),
-					(sqe_flags & IOSQE_FIXED_FILE));
+					(sqe_flags & IOSQE_FIXED_FILE),
+					IO_URING_F_NONBLOCK);
 		if (unlikely(!req->file))
 			ret = -EBADF;
 	}
@@ -7555,7 +7742,7 @@ static int io_run_task_work_sig(void)
 /* when returns >0, the caller should retry */
 static inline int io_cqring_wait_schedule(struct io_ring_ctx *ctx,
 					  struct io_wait_queue *iowq,
-					  ktime_t timeout)
+					  ktime_t *timeout)
 {
 	int ret;
 
@@ -7567,7 +7754,7 @@ static inline int io_cqring_wait_schedule(struct io_ring_ctx *ctx,
 	if (test_bit(0, &ctx->check_cq_overflow))
 		return 1;
 
-	if (!schedule_hrtimeout(&timeout, HRTIMER_MODE_ABS))
+	if (!schedule_hrtimeout(timeout, HRTIMER_MODE_ABS))
 		return -ETIME;
 	return 1;
 }
@@ -7630,7 +7817,7 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
 		}
 		prepare_to_wait_exclusive(&ctx->cq_wait, &iowq.wq,
 						TASK_INTERRUPTIBLE);
-		ret = io_cqring_wait_schedule(ctx, &iowq, timeout);
+		ret = io_cqring_wait_schedule(ctx, &iowq, &timeout);
 		finish_wait(&ctx->cq_wait, &iowq.wq);
 		cond_resched();
 	} while (ret > 0);
@@ -9424,8 +9611,10 @@ static void io_tctx_exit_cb(struct callback_head *cb)
 	/*
 	 * When @in_idle, we're in cancellation and it's racy to remove the
 	 * node. It'll be removed by the end of cancellation, just ignore it.
+	 * tctx can be NULL if the queueing of this task_work raced with
+	 * work cancelation off the exec path.
 	 */
-	if (!atomic_read(&tctx->in_idle))
+	if (tctx && !atomic_read(&tctx->in_idle))
 		io_uring_del_tctx_node((unsigned long)work->ctx);
 	complete(&work->completion);
 }
@@ -9548,6 +9737,10 @@ static void io_ring_ctx_wait_and_kill(struct io_ring_ctx *ctx)
 	/* if we failed setting up the ctx, we might not have any rings */
 	io_iopoll_try_reap_events(ctx);
 
+	/* drop cached put refs after potentially doing completions */
+	if (current->io_uring)
+		io_uring_drop_tctx_refs(current);
+
 	INIT_WORK(&ctx->exit_work, io_ring_exit_work);
 	/*
 	 * Use system_unbound_wq to avoid spawning tons of event kworkers
@@ -10850,8 +11043,6 @@ static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
 		return -ENXIO;
 
 	if (ctx->restricted) {
-		if (opcode >= IORING_REGISTER_LAST)
-			return -EINVAL;
 		opcode = array_index_nospec(opcode, IORING_REGISTER_LAST);
 		if (!test_bit(opcode, ctx->restrictions.register_op))
 			return -EACCES;
@@ -10983,6 +11174,9 @@ SYSCALL_DEFINE4(io_uring_register, unsigned int, fd, unsigned int, opcode,
 	long ret = -EBADF;
 	struct fd f;
 
+	if (opcode >= IORING_REGISTER_LAST)
+		return -EINVAL;
+
 	f = fdget(fd);
 	if (!f.file)
 		return -EBADF;

kernel/sched/core.c

@@ -21,7 +21,7 @@
 #include <asm/tlb.h>
 
 #include "../workqueue_internal.h"
-#include "../../fs/io-wq.h"
+#include "../../io_uring/io-wq.h"
 #include "../smpboot.h"
 
 #include "pelt.h"