Merge tag 'for-6.8/io_uring-2024-01-08' of git://git.kernel.dk/linux

Pull io_uring updates from Jens Axboe:
 "Mostly just come fixes and cleanups, but one feature as well. In
  detail:

   - Harden the check for handling IOPOLL based on return (Pavel)

   - Various minor optimizations (Pavel)

   - Drop remnants of SCM_RIGHTS fd passing support, now that it's no
     longer supported since 6.7 (me)

   - Fix for a case where bytes_done wasn't initialized properly on a
     failure condition for read/write requests (me)

   - Move the register related code to a separate file (me)

   - Add support for returning the provided ring buffer head (me)

   - Add support for adding a direct descriptor to the normal file table
     (me, Christian Brauner)

   - Fix for ensuring pending task_work for a ring with DEFER_TASKRUN is
     run even if we timeout waiting (me)"

* tag 'for-6.8/io_uring-2024-01-08' of git://git.kernel.dk/linux:
  io_uring: ensure local task_work is run on wait timeout
  io_uring/kbuf: add method for returning provided buffer ring head
  io_uring/rw: ensure io->bytes_done is always initialized
  io_uring: drop any code related to SCM_RIGHTS
  io_uring/unix: drop usage of io_uring socket
  io_uring/register: move io_uring_register(2) related code to register.c
  io_uring/openclose: add support for IORING_OP_FIXED_FD_INSTALL
  io_uring/cmd: inline io_uring_cmd_get_task
  io_uring/cmd: inline io_uring_cmd_do_in_task_lazy
  io_uring: split out cmd api into a separate header
  io_uring: optimise ltimeout for inline execution
  io_uring: don't check iopoll if request completes

include/linux/io_uring.h

@@ -6,66 +6,13 @@
 #include <linux/xarray.h>
 #include <uapi/linux/io_uring.h>
 
-enum io_uring_cmd_flags {
-	IO_URING_F_COMPLETE_DEFER	= 1,
-	IO_URING_F_UNLOCKED		= 2,
-	/* the request is executed from poll, it should not be freed */
-	IO_URING_F_MULTISHOT		= 4,
-	/* executed by io-wq */
-	IO_URING_F_IOWQ			= 8,
-	/* int's last bit, sign checks are usually faster than a bit test */
-	IO_URING_F_NONBLOCK		= INT_MIN,
-
-	/* ctx state flags, for URING_CMD */
-	IO_URING_F_SQE128		= (1 << 8),
-	IO_URING_F_CQE32		= (1 << 9),
-	IO_URING_F_IOPOLL		= (1 << 10),
-
-	/* set when uring wants to cancel a previously issued command */
-	IO_URING_F_CANCEL		= (1 << 11),
-	IO_URING_F_COMPAT		= (1 << 12),
-};
-
-/* only top 8 bits of sqe->uring_cmd_flags for kernel internal use */
-#define IORING_URING_CMD_CANCELABLE	(1U << 30)
-
-struct io_uring_cmd {
-	struct file	*file;
-	const struct io_uring_sqe *sqe;
-	/* callback to defer completions to task context */
-	void (*task_work_cb)(struct io_uring_cmd *cmd, unsigned);
-	u32		cmd_op;
-	u32		flags;
-	u8		pdu[32]; /* available inline for free use */
-};
-
-static inline const void *io_uring_sqe_cmd(const struct io_uring_sqe *sqe)
-{
-	return sqe->cmd;
-}
-
 #if defined(CONFIG_IO_URING)
-int io_uring_cmd_import_fixed(u64 ubuf, unsigned long len, int rw,
-			      struct iov_iter *iter, void *ioucmd);
-void io_uring_cmd_done(struct io_uring_cmd *cmd, ssize_t ret, ssize_t res2,
-			unsigned issue_flags);
-struct sock *io_uring_get_socket(struct file *file);
 void __io_uring_cancel(bool cancel_all);
 void __io_uring_free(struct task_struct *tsk);
 void io_uring_unreg_ringfd(void);
 const char *io_uring_get_opcode(u8 opcode);
-void __io_uring_cmd_do_in_task(struct io_uring_cmd *ioucmd,
-			    void (*task_work_cb)(struct io_uring_cmd *, unsigned),
-			    unsigned flags);
-/* users should follow semantics of IOU_F_TWQ_LAZY_WAKE */
-void io_uring_cmd_do_in_task_lazy(struct io_uring_cmd *ioucmd,
-			void (*task_work_cb)(struct io_uring_cmd *, unsigned));
-
-static inline void io_uring_cmd_complete_in_task(struct io_uring_cmd *ioucmd,
-			void (*task_work_cb)(struct io_uring_cmd *, unsigned))
-{
-	__io_uring_cmd_do_in_task(ioucmd, task_work_cb, 0);
-}
+int io_uring_cmd_sock(struct io_uring_cmd *cmd, unsigned int issue_flags);
+bool io_is_uring_fops(struct file *file);
 
 static inline void io_uring_files_cancel(void)
 {
@@ -84,32 +31,7 @@ static inline void io_uring_free(struct task_struct *tsk)
 	if (tsk->io_uring)
 		__io_uring_free(tsk);
 }
-int io_uring_cmd_sock(struct io_uring_cmd *cmd, unsigned int issue_flags);
-void io_uring_cmd_mark_cancelable(struct io_uring_cmd *cmd,
-		unsigned int issue_flags);
-struct task_struct *io_uring_cmd_get_task(struct io_uring_cmd *cmd);
 #else
-static inline int io_uring_cmd_import_fixed(u64 ubuf, unsigned long len, int rw,
-			      struct iov_iter *iter, void *ioucmd)
-{
-	return -EOPNOTSUPP;
-}
-static inline void io_uring_cmd_done(struct io_uring_cmd *cmd, ssize_t ret,
-		ssize_t ret2, unsigned issue_flags)
-{
-}
-static inline void io_uring_cmd_complete_in_task(struct io_uring_cmd *ioucmd,
-			void (*task_work_cb)(struct io_uring_cmd *, unsigned))
-{
-}
-static inline void io_uring_cmd_do_in_task_lazy(struct io_uring_cmd *ioucmd,
-			void (*task_work_cb)(struct io_uring_cmd *, unsigned))
-{
-}
-static inline struct sock *io_uring_get_socket(struct file *file)
-{
-	return NULL;
-}
 static inline void io_uring_task_cancel(void)
 {
 }
@@ -128,13 +50,9 @@ static inline int io_uring_cmd_sock(struct io_uring_cmd *cmd,
 {
 	return -EOPNOTSUPP;
 }
-static inline void io_uring_cmd_mark_cancelable(struct io_uring_cmd *cmd,
-		unsigned int issue_flags)
+static inline bool io_is_uring_fops(struct file *file)
 {
-}
-static inline struct task_struct *io_uring_cmd_get_task(struct io_uring_cmd *cmd)
-{
-	return NULL;
+	return false;
 }
 #endif
 

include/linux/io_uring/cmd.h

@@ -0,0 +1,77 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+#ifndef _LINUX_IO_URING_CMD_H
+#define _LINUX_IO_URING_CMD_H
+
+#include <uapi/linux/io_uring.h>
+#include <linux/io_uring_types.h>
+
+/* only top 8 bits of sqe->uring_cmd_flags for kernel internal use */
+#define IORING_URING_CMD_CANCELABLE	(1U << 30)
+
+struct io_uring_cmd {
+	struct file	*file;
+	const struct io_uring_sqe *sqe;
+	/* callback to defer completions to task context */
+	void (*task_work_cb)(struct io_uring_cmd *cmd, unsigned);
+	u32		cmd_op;
+	u32		flags;
+	u8		pdu[32]; /* available inline for free use */
+};
+
+static inline const void *io_uring_sqe_cmd(const struct io_uring_sqe *sqe)
+{
+	return sqe->cmd;
+}
+
+#if defined(CONFIG_IO_URING)
+int io_uring_cmd_import_fixed(u64 ubuf, unsigned long len, int rw,
+			      struct iov_iter *iter, void *ioucmd);
+void io_uring_cmd_done(struct io_uring_cmd *cmd, ssize_t ret, ssize_t res2,
+			unsigned issue_flags);
+void __io_uring_cmd_do_in_task(struct io_uring_cmd *ioucmd,
+			    void (*task_work_cb)(struct io_uring_cmd *, unsigned),
+			    unsigned flags);
+
+void io_uring_cmd_mark_cancelable(struct io_uring_cmd *cmd,
+		unsigned int issue_flags);
+
+#else
+static inline int io_uring_cmd_import_fixed(u64 ubuf, unsigned long len, int rw,
+			      struct iov_iter *iter, void *ioucmd)
+{
+	return -EOPNOTSUPP;
+}
+static inline void io_uring_cmd_done(struct io_uring_cmd *cmd, ssize_t ret,
+		ssize_t ret2, unsigned issue_flags)
+{
+}
+static inline void __io_uring_cmd_do_in_task(struct io_uring_cmd *ioucmd,
+			    void (*task_work_cb)(struct io_uring_cmd *, unsigned),
+			    unsigned flags)
+{
+}
+static inline void io_uring_cmd_mark_cancelable(struct io_uring_cmd *cmd,
+		unsigned int issue_flags)
+{
+}
+#endif
+
+/* users must follow the IOU_F_TWQ_LAZY_WAKE semantics */
+static inline void io_uring_cmd_do_in_task_lazy(struct io_uring_cmd *ioucmd,
+			void (*task_work_cb)(struct io_uring_cmd *, unsigned))
+{
+	__io_uring_cmd_do_in_task(ioucmd, task_work_cb, IOU_F_TWQ_LAZY_WAKE);
+}
+
+static inline void io_uring_cmd_complete_in_task(struct io_uring_cmd *ioucmd,
+			void (*task_work_cb)(struct io_uring_cmd *, unsigned))
+{
+	__io_uring_cmd_do_in_task(ioucmd, task_work_cb, 0);
+}
+
+static inline struct task_struct *io_uring_cmd_get_task(struct io_uring_cmd *cmd)
+{
+	return cmd_to_io_kiocb(cmd)->task;
+}
+
+#endif /* _LINUX_IO_URING_CMD_H */

include/linux/io_uring_types.h

@@ -7,6 +7,37 @@
 #include <linux/llist.h>
 #include <uapi/linux/io_uring.h>
 
+enum {
+	/*
+	 * A hint to not wake right away but delay until there are enough of
+	 * tw's queued to match the number of CQEs the task is waiting for.
+	 *
+	 * Must not be used wirh requests generating more than one CQE.
+	 * It's also ignored unless IORING_SETUP_DEFER_TASKRUN is set.
+	 */
+	IOU_F_TWQ_LAZY_WAKE			= 1,
+};
+
+enum io_uring_cmd_flags {
+	IO_URING_F_COMPLETE_DEFER	= 1,
+	IO_URING_F_UNLOCKED		= 2,
+	/* the request is executed from poll, it should not be freed */
+	IO_URING_F_MULTISHOT		= 4,
+	/* executed by io-wq */
+	IO_URING_F_IOWQ			= 8,
+	/* int's last bit, sign checks are usually faster than a bit test */
+	IO_URING_F_NONBLOCK		= INT_MIN,
+
+	/* ctx state flags, for URING_CMD */
+	IO_URING_F_SQE128		= (1 << 8),
+	IO_URING_F_CQE32		= (1 << 9),
+	IO_URING_F_IOPOLL		= (1 << 10),
+
+	/* set when uring wants to cancel a previously issued command */
+	IO_URING_F_CANCEL		= (1 << 11),
+	IO_URING_F_COMPAT		= (1 << 12),
+};
+
 struct io_wq_work_node {
 	struct io_wq_work_node *next;
 };
@@ -358,9 +389,6 @@ struct io_ring_ctx {
 	struct wait_queue_head		rsrc_quiesce_wq;
 	unsigned			rsrc_quiesce;
 
-	#if defined(CONFIG_UNIX)
-		struct socket		*ring_sock;
-	#endif
 	/* hashed buffered write serialization */
 	struct io_wq_hash		*hash_map;
 

include/uapi/linux/io_uring.h

@@ -71,6 +71,7 @@ struct io_uring_sqe {
 		__u32		uring_cmd_flags;
 		__u32		waitid_flags;
 		__u32		futex_flags;
+		__u32		install_fd_flags;
 	};
 	__u64	user_data;	/* data to be passed back at completion time */
 	/* pack this to avoid bogus arm OABI complaints */
@@ -253,6 +254,7 @@ enum io_uring_op {
 	IORING_OP_FUTEX_WAIT,
 	IORING_OP_FUTEX_WAKE,
 	IORING_OP_FUTEX_WAITV,
+	IORING_OP_FIXED_FD_INSTALL,
 
 	/* this goes last, obviously */
 	IORING_OP_LAST,
@@ -386,6 +388,13 @@ enum {
 /* Pass through the flags from sqe->file_index to cqe->flags */
 #define IORING_MSG_RING_FLAGS_PASS	(1U << 1)
 
+/*
+ * IORING_OP_FIXED_FD_INSTALL flags (sqe->install_fd_flags)
+ *
+ * IORING_FIXED_FD_NO_CLOEXEC	Don't mark the fd as O_CLOEXEC
+ */
+#define IORING_FIXED_FD_NO_CLOEXEC	(1U << 0)
+
 /*
  * IO completion data structure (Completion Queue Entry)
  */
@@ -558,6 +567,9 @@ enum {
 	/* register a range of fixed file slots for automatic slot allocation */
 	IORING_REGISTER_FILE_ALLOC_RANGE	= 25,
 
+	/* return status information for a buffer group */
+	IORING_REGISTER_PBUF_STATUS		= 26,
+
 	/* this goes last */
 	IORING_REGISTER_LAST,
 
@@ -684,6 +696,13 @@ struct io_uring_buf_reg {
 	__u64	resv[3];
 };
 
+/* argument for IORING_REGISTER_PBUF_STATUS */
+struct io_uring_buf_status {
+	__u32	buf_group;	/* input */
+	__u32	head;		/* output */
+	__u32	resv[8];
+};
+
 /*
  * io_uring_restriction->opcode values
  */