move aio compat to fs/aio.c

... and fix the minor buglet in compat io_submit() - native one
kills ioctx as cleanup when put_user() fails.  Get rid of
bogus compat_... in !CONFIG_AIO case, while we are at it - they
should simply fail with ENOSYS, same as for native counterparts.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

fs/aio.c

@@ -1367,6 +1367,39 @@ out:
 	return ret;
 }
 
+#ifdef CONFIG_COMPAT
+COMPAT_SYSCALL_DEFINE2(io_setup, unsigned, nr_events, u32 __user *, ctx32p)
+{
+	struct kioctx *ioctx = NULL;
+	unsigned long ctx;
+	long ret;
+
+	ret = get_user(ctx, ctx32p);
+	if (unlikely(ret))
+		goto out;
+
+	ret = -EINVAL;
+	if (unlikely(ctx || nr_events == 0)) {
+		pr_debug("EINVAL: ctx %lu nr_events %u\n",
+		         ctx, nr_events);
+		goto out;
+	}
+
+	ioctx = ioctx_alloc(nr_events);
+	ret = PTR_ERR(ioctx);
+	if (!IS_ERR(ioctx)) {
+		/* truncating is ok because it's a user address */
+		ret = put_user((u32)ioctx->user_id, ctx32p);
+		if (ret)
+			kill_ioctx(current->mm, ioctx, NULL);
+		percpu_ref_put(&ioctx->users);
+	}
+
+out:
+	return ret;
+}
+#endif
+
 /* sys_io_destroy:
  *	Destroy the aio_context specified.  May cancel any outstanding 
  *	AIOs and block on completion.  Will fail with -ENOSYS if not
@@ -1591,8 +1624,8 @@ out_put_req:
 	return ret;
 }
 
-long do_io_submit(aio_context_t ctx_id, long nr,
-		  struct iocb __user *__user *iocbpp, bool compat)
+static long do_io_submit(aio_context_t ctx_id, long nr,
+			  struct iocb __user *__user *iocbpp, bool compat)
 {
 	struct kioctx *ctx;
 	long ret = 0;
@@ -1662,6 +1695,44 @@ SYSCALL_DEFINE3(io_submit, aio_context_t, ctx_id, long, nr,
 	return do_io_submit(ctx_id, nr, iocbpp, 0);
 }
 
+#ifdef CONFIG_COMPAT
+static inline long
+copy_iocb(long nr, u32 __user *ptr32, struct iocb __user * __user *ptr64)
+{
+	compat_uptr_t uptr;
+	int i;
+
+	for (i = 0; i < nr; ++i) {
+		if (get_user(uptr, ptr32 + i))
+			return -EFAULT;
+		if (put_user(compat_ptr(uptr), ptr64 + i))
+			return -EFAULT;
+	}
+	return 0;
+}
+
+#define MAX_AIO_SUBMITS 	(PAGE_SIZE/sizeof(struct iocb *))
+
+COMPAT_SYSCALL_DEFINE3(io_submit, compat_aio_context_t, ctx_id,
+		       int, nr, u32 __user *, iocb)
+{
+	struct iocb __user * __user *iocb64;
+	long ret;
+
+	if (unlikely(nr < 0))
+		return -EINVAL;
+
+	if (nr > MAX_AIO_SUBMITS)
+		nr = MAX_AIO_SUBMITS;
+
+	iocb64 = compat_alloc_user_space(nr * sizeof(*iocb64));
+	ret = copy_iocb(nr, iocb, iocb64);
+	if (!ret)
+		ret = do_io_submit(ctx_id, nr, iocb64, 1);
+	return ret;
+}
+#endif
+
 /* lookup_kiocb
  *	Finds a given iocb for cancellation.
  */
@@ -1761,3 +1832,25 @@ SYSCALL_DEFINE5(io_getevents, aio_context_t, ctx_id,
 	}
 	return ret;
 }
+
+#ifdef CONFIG_COMPAT
+COMPAT_SYSCALL_DEFINE5(io_getevents, compat_aio_context_t, ctx_id,
+		       compat_long_t, min_nr,
+		       compat_long_t, nr,
+		       struct io_event __user *, events,
+		       struct compat_timespec __user *, timeout)
+{
+	struct timespec t;
+	struct timespec __user *ut = NULL;
+
+	if (timeout) {
+		if (compat_get_timespec(&t, timeout))
+			return -EFAULT;
+
+		ut = compat_alloc_user_space(sizeof(*ut));
+		if (copy_to_user(ut, &t, sizeof(t)))
+			return -EFAULT;
+	}
+	return sys_io_getevents(ctx_id, min_nr, nr, events, ut);
+}
+#endif

fs/compat.c

@@ -487,45 +487,6 @@ COMPAT_SYSCALL_DEFINE3(fcntl, unsigned int, fd, unsigned int, cmd,
 	return compat_sys_fcntl64(fd, cmd, arg);
 }
 
-COMPAT_SYSCALL_DEFINE2(io_setup, unsigned, nr_reqs, u32 __user *, ctx32p)
-{
-	long ret;
-	aio_context_t ctx64;
-
-	mm_segment_t oldfs = get_fs();
-	if (unlikely(get_user(ctx64, ctx32p)))
-		return -EFAULT;
-
-	set_fs(KERNEL_DS);
-	/* The __user pointer cast is valid because of the set_fs() */
-	ret = sys_io_setup(nr_reqs, (aio_context_t __user *) &ctx64);
-	set_fs(oldfs);
-	/* truncating is ok because it's a user address */
-	if (!ret)
-		ret = put_user((u32) ctx64, ctx32p);
-	return ret;
-}
-
-COMPAT_SYSCALL_DEFINE5(io_getevents, compat_aio_context_t, ctx_id,
-		       compat_long_t, min_nr,
-		       compat_long_t, nr,
-		       struct io_event __user *, events,
-		       struct compat_timespec __user *, timeout)
-{
-	struct timespec t;
-	struct timespec __user *ut = NULL;
-
-	if (timeout) {
-		if (compat_get_timespec(&t, timeout))
-			return -EFAULT;
-
-		ut = compat_alloc_user_space(sizeof(*ut));
-		if (copy_to_user(ut, &t, sizeof(t)) )
-			return -EFAULT;
-	} 
-	return sys_io_getevents(ctx_id, min_nr, nr, events, ut);
-}
-
 /* A write operation does a read from user space and vice versa */
 #define vrfy_dir(type) ((type) == READ ? VERIFY_WRITE : VERIFY_READ)
 
@@ -602,42 +563,6 @@ out:
 	return ret;
 }
 
-static inline long
-copy_iocb(long nr, u32 __user *ptr32, struct iocb __user * __user *ptr64)
-{
-	compat_uptr_t uptr;
-	int i;
-
-	for (i = 0; i < nr; ++i) {
-		if (get_user(uptr, ptr32 + i))
-			return -EFAULT;
-		if (put_user(compat_ptr(uptr), ptr64 + i))
-			return -EFAULT;
-	}
-	return 0;
-}
-
-#define MAX_AIO_SUBMITS 	(PAGE_SIZE/sizeof(struct iocb *))
-
-COMPAT_SYSCALL_DEFINE3(io_submit, compat_aio_context_t, ctx_id,
-		       int, nr, u32 __user *, iocb)
-{
-	struct iocb __user * __user *iocb64; 
-	long ret;
-
-	if (unlikely(nr < 0))
-		return -EINVAL;
-
-	if (nr > MAX_AIO_SUBMITS)
-		nr = MAX_AIO_SUBMITS;
-	
-	iocb64 = compat_alloc_user_space(nr * sizeof(*iocb64));
-	ret = copy_iocb(nr, iocb, iocb64);
-	if (!ret)
-		ret = do_io_submit(ctx_id, nr, iocb64, 1);
-	return ret;
-}
-
 struct compat_ncp_mount_data {
 	compat_int_t version;
 	compat_uint_t ncp_fd;

include/linux/aio.h

@@ -14,14 +14,9 @@ typedef int (kiocb_cancel_fn)(struct kiocb *);
 /* prototypes */
 #ifdef CONFIG_AIO
 extern void exit_aio(struct mm_struct *mm);
-extern long do_io_submit(aio_context_t ctx_id, long nr,
-			 struct iocb __user *__user *iocbpp, bool compat);
 void kiocb_set_cancel_fn(struct kiocb *req, kiocb_cancel_fn *cancel);
 #else
 static inline void exit_aio(struct mm_struct *mm) { }
-static inline long do_io_submit(aio_context_t ctx_id, long nr,
-				struct iocb __user * __user *iocbpp,
-				bool compat) { return 0; }
 static inline void kiocb_set_cancel_fn(struct kiocb *req,
 				       kiocb_cancel_fn *cancel) { }
 #endif /* CONFIG_AIO */

kernel/sys_ni.c

@@ -150,6 +150,9 @@ cond_syscall(sys_io_destroy);
 cond_syscall(sys_io_submit);
 cond_syscall(sys_io_cancel);
 cond_syscall(sys_io_getevents);
+cond_syscall(compat_sys_io_setup);
+cond_syscall(compat_sys_io_submit);
+cond_syscall(compat_sys_io_getevents);
 cond_syscall(sys_sysfs);
 cond_syscall(sys_syslog);
 cond_syscall(sys_process_vm_readv);