fs.close_range.v5.15

-----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYS3lFAAKCRCRxhvAZXjc
 opiiAQDydlquFUOP3/HAClwlBr1050GeFLnSVHFTrcaEjUIhLgD9H8hU77c+Q7O8
 ZxWmx3Ficn4fkVWN7zsOrcleN6RBtgk=
 =VXhQ
 -----END PGP SIGNATURE-----

Merge tag 'fs.close_range.v5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

Pull close_range() cleanup from Christian Brauner:
 "This is a cleanup for close_range() which was sent as part of a bugfix
  we did some time ago in commit 9b5b872215 ("file: fix close_range()
  for unshare+cloexec").

  We used to share more code between some helpers for close_range()
  which made retrieving the maximum number of open fds before calling
  into the helpers sensible. But with the introduction of
  CLOSE_RANGE_CLOEXEC and the need to retrieve the number of maximum fds
  once more for CLOSE_RANGE_CLOEXEC that stopped making sense. So the
  code was in a dumb in-limbo state.

  Fix this by simplifying the code a bit.

  The original idea was to only fix the bug itself and make backporting
  easy. And since the cleanup wasn't very pressing I left it in
  linux-next for a very long time. I didn't pull the patches from the
  list again back then which is why they don't have lore-links. So I'm
  listing them below explicitly"

Commit 03ba0fe4d0 ("file: simplify logic in __close_range()")
Link: https://lore.kernel.org/linux-fsdevel/20210402123548.108372-3-brauner@kernel.org

Commit f49fd6d3c0 ("file: let pick_file() tell caller it's done")
Link: https://lore.kernel.org/linux-fsdevel/20210402123548.108372-4-brauner@kernel.org

* tag 'fs.close_range.v5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  file: simplify logic in __close_range()
  file: let pick_file() tell caller it's done

fs/file.c

@@ -596,18 +596,32 @@ void fd_install(unsigned int fd, struct file *file)
 
 EXPORT_SYMBOL(fd_install);
 
+/**
+ * pick_file - return file associatd with fd
+ * @files: file struct to retrieve file from
+ * @fd: file descriptor to retrieve file for
+ *
+ * If this functions returns an EINVAL error pointer the fd was beyond the
+ * current maximum number of file descriptors for that fdtable.
+ *
+ * Returns: The file associated with @fd, on error returns an error pointer.
+ */
 static struct file *pick_file(struct files_struct *files, unsigned fd)
 {
-	struct file *file = NULL;
+	struct file *file;
 	struct fdtable *fdt;
 
 	spin_lock(&files->file_lock);
 	fdt = files_fdtable(files);
-	if (fd >= fdt->max_fds)
+	if (fd >= fdt->max_fds) {
+		file = ERR_PTR(-EINVAL);
 		goto out_unlock;
+	}
 	file = fdt->fd[fd];
-	if (!file)
+	if (!file) {
+		file = ERR_PTR(-EBADF);
 		goto out_unlock;
+	}
 	rcu_assign_pointer(fdt->fd[fd], NULL);
 	__put_unused_fd(files, fd);
 
@@ -622,7 +636,7 @@ int close_fd(unsigned fd)
 	struct file *file;
 
 	file = pick_file(files, fd);
-	if (!file)
+	if (IS_ERR(file))
 		return -EBADF;
 
 	return filp_close(file, files);
@@ -663,11 +677,16 @@ static inline void __range_close(struct files_struct *cur_fds, unsigned int fd,
 		struct file *file;
 
 		file = pick_file(cur_fds, fd++);
-		if (!file)
+		if (!IS_ERR(file)) {
+			/* found a valid file to close */
+			filp_close(file, cur_fds);
+			cond_resched();
 			continue;
+		}
 
-		filp_close(file, cur_fds);
-		cond_resched();
+		/* beyond the last fd in that table */
+		if (PTR_ERR(file) == -EINVAL)
+			return;
 	}
 }
 
@@ -682,7 +701,6 @@ static inline void __range_close(struct files_struct *cur_fds, unsigned int fd,
  */
 int __close_range(unsigned fd, unsigned max_fd, unsigned int flags)
 {
-	unsigned int cur_max;
 	struct task_struct *me = current;
 	struct files_struct *cur_fds = me->files, *fds = NULL;
 
@@ -692,26 +710,26 @@ int __close_range(unsigned fd, unsigned max_fd, unsigned int flags)
 	if (fd > max_fd)
 		return -EINVAL;
 
-	rcu_read_lock();
-	cur_max = files_fdtable(cur_fds)->max_fds;
-	rcu_read_unlock();
-
-	/* cap to last valid index into fdtable */
-	cur_max--;
-
 	if (flags & CLOSE_RANGE_UNSHARE) {
 		int ret;
 		unsigned int max_unshare_fds = NR_OPEN_MAX;
 
 		/*
-		 * If the requested range is greater than the current maximum,
-		 * we're closing everything so only copy all file descriptors
-		 * beneath the lowest file descriptor.
-		 * If the caller requested all fds to be made cloexec copy all
-		 * of the file descriptors since they still want to use them.
+		 * If the caller requested all fds to be made cloexec we always
+		 * copy all of the file descriptors since they still want to
+		 * use them.
 		 */
-		if (!(flags & CLOSE_RANGE_CLOEXEC) && (max_fd >= cur_max))
-			max_unshare_fds = fd;
+		if (!(flags & CLOSE_RANGE_CLOEXEC)) {
+			/*
+			 * If the requested range is greater than the current
+			 * maximum, we're closing everything so only copy all
+			 * file descriptors beneath the lowest file descriptor.
+			 */
+			rcu_read_lock();
+			if (max_fd >= last_fd(files_fdtable(cur_fds)))
+				max_unshare_fds = fd;
+			rcu_read_unlock();
+		}
 
 		ret = unshare_fd(CLONE_FILES, max_unshare_fds, &fds);
 		if (ret)
@@ -725,8 +743,6 @@ int __close_range(unsigned fd, unsigned max_fd, unsigned int flags)
 			swap(cur_fds, fds);
 	}
 
-	max_fd = min(max_fd, cur_max);
-
 	if (flags & CLOSE_RANGE_CLOEXEC)
 		__range_cloexec(cur_fds, fd, max_fd);
 	else