[PATCH 0/3] Eliminating __FD_*() functions from the kernel

[PATCH 0/3] Eliminating __FD_*() functions from the kernel

[David Howells]

Having examined H. Peter Anvin's patches to merge asm/posix_types.h together, I
think that we can go further: The FD_SET(), FD_CLR() and FD_ISSET() macros are
only used within the kernel itself in relation to the arrays of fd-is-open
flags and close-on-exec flags.  FD_ZERO() is not used at all within the kernel.

Now, the FD_SET() & co. wrappers must be exported to userspace, so they must be
kept, but the __FD_SET() & co. implementations behind them on most arches are
_not_ so exported.  Possibly PowerPC, MN10300 and Xtensa would survive their
removal entirely from the userspace API as per Peter's patches.

Examining the two remaining usages of struct fd_set, and FD_*() and __FD_*()
for the arrays of bits in struct fdtable, it would seem that we could just move
to using unsigned long directly rather than fd_set, and using __set_bit(),
__clear_bit() and test_bit() upon those.

This would also get rid of two abuses of the fd_set struct within this code:

 (1) Most of the time, the two fd_sets pointed to by struct fdtable are not the
     length they're supposed to be, and are, in fact, allocated short.  This
     means FD_ZERO() may not be used on them, for example.

     Basically, in effect, alloc_fdmem() allocates an array of unsigned longs
     just long enough and then casts it to an fd_set - so why bother with the
     fd_set at all?

 (2) Non-core VFS code (such as SELinux) sometimes wants to look through the
     array of unsigned longs inside the fd_set for the purposes of more
     efficient iteration over the entire set of open fds - so why not just
     eliminate the fd_set and go directly with an array of unsigned longs?

Furthermore, within the kernel, __FD_SET() and co. are redundant bitops and,
in the original implementation, may actually be less efficient.  For instance,
on x86/x86_64, __set_bit() uses the BTS instruction, but __FD_SET() may well
not (depending on the optimiser).  I note that Peter's patch does alter this.


So my patches do the following:

 (1) Wrap the normal set/bit/test operations on the open_fds and close_on_exec
     fd_sets so that the access to them is, for the most part, not open coded.

     There's still the issue of the things that poke around inside the fd_sets
     behind the scenes (SELinux for example).

 (2) Replace the fd_set with an unsigned long array and use __set_bit() and
     co. instead of FD_SET() and co.

     It might just be better to replace the uses of FD_SET() and co. with
     __set_bit() and co. directly, and not bother with wrappers, but the
     wrappers are more greppable.

 (3) Delete __FD_SET() and co. from linux/time.h.

And are implemented to go on top of Peter's posix_types patchset.

David
---
David Howells (3):
      Delete the __FD_*() funcs for operating on fd_set from linux/time.h
      Replace the fd_sets in struct fdtable with an array of unsigned longs
      Wrap accesses to the fd_sets in struct fdtable


 Documentation/filesystems/files.txt          |    4 +-
 arch/powerpc/platforms/cell/spufs/coredump.c |    2 -
 drivers/staging/android/binder.c             |   10 ++---
 fs/autofs4/dev-ioctl.c                       |    2 -
 fs/exec.c                                    |    8 ++--
 fs/fcntl.c                                   |   18 ++++-----
 fs/file.c                                    |   54 +++++++++++++-------------
 fs/open.c                                    |    4 +-
 fs/proc/base.c                               |    2 -
 fs/select.c                                  |    2 -
 include/linux/fdtable.h                      |   46 ++++++++++++++++------
 include/linux/time.h                         |   22 -----------
 kernel/exit.c                                |    2 -
 security/selinux/hooks.c                     |    2 -
 14 files changed, 88 insertions(+), 90 deletions(-)

[PATCH 1/3] Wrap accesses to the fd_sets in struct fdtable

[David Howells]

Wrap accesses to the fd_sets in struct fdtable (for recording open files and
close-on-exec flags) so that we can move away from using fd_sets since we
abuse the fd_set structs by not allocating the full-sized structure under
normal circumstances and by non-core code looking at the internals of the
fd_sets.

The first abuse means that use of FD_ZERO() on these fd_sets is not permitted,
since that cannot be told about their abnormal lengths.

This introduces six wrapper functions for setting, clearing and testing
close-on-exec flags and fd-is-open flags:

	void __set_close_on_exec(int fd, struct fdtable *fdt);
	void __clear_close_on_exec(int fd, struct fdtable *fdt);
	bool close_on_exec(int fd, const struct fdtable *fdt);
	void __set_open_fd(int fd, struct fdtable *fdt);
	void __clear_open_fd(int fd, struct fdtable *fdt);
	bool fd_is_open(int fd, const struct fdtable *fdt);

Note that I've prepended '__' to the names of the set/clear functions because
they require the caller to hold a lock to use them.

Note also that I haven't added wrappers for looking behind the scenes at the
the array.  Possibly that should exist too.

Signed-off-by: David Howells <dhowells@redhat.com>
---

 Documentation/filesystems/files.txt          |    4 ++-
 arch/powerpc/platforms/cell/spufs/coredump.c |    2 +-
 drivers/staging/android/binder.c             |   10 ++++-----
 fs/autofs4/dev-ioctl.c                       |    2 +-
 fs/exec.c                                    |    4 ++-
 fs/fcntl.c                                   |   18 ++++++++--------
 fs/file.c                                    |    8 +++----
 fs/open.c                                    |    4 ++-
 fs/proc/base.c                               |    2 +-
 include/linux/fdtable.h                      |   30 ++++++++++++++++++++++++++
 10 files changed, 57 insertions(+), 27 deletions(-)

diff --git a/Documentation/filesystems/files.txt b/Documentation/filesystems/files.txt
index ac2facc..46dfc6b 100644
--- a/Documentation/filesystems/files.txt
+++ b/Documentation/filesystems/files.txt
@@ -113,8 +113,8 @@ the fdtable structure -
 	if (fd >= 0) {
 		/* locate_fd() may have expanded fdtable, load the ptr */
 		fdt = files_fdtable(files);
-		FD_SET(fd, fdt->open_fds);
-		FD_CLR(fd, fdt->close_on_exec);
+		__set_open_fd(fd, fdt);
+		__clear_close_on_exec(fd, fdt);
 		spin_unlock(&files->file_lock);
 	.....
 
diff --git a/arch/powerpc/platforms/cell/spufs/coredump.c b/arch/powerpc/platforms/cell/spufs/coredump.c
index 03c5fce..c2c5b07 100644
--- a/arch/powerpc/platforms/cell/spufs/coredump.c
+++ b/arch/powerpc/platforms/cell/spufs/coredump.c
@@ -122,7 +122,7 @@ static struct spu_context *coredump_next_context(int *fd)
 	struct spu_context *ctx = NULL;
 
 	for (; *fd < fdt->max_fds; (*fd)++) {
-		if (!FD_ISSET(*fd, fdt->open_fds))
+		if (!fd_is_open(*fd, fdt))
 			continue;
 
 		file = fcheck(*fd);
diff --git a/drivers/staging/android/binder.c b/drivers/staging/android/binder.c
index f0b7e66..347bd6e 100644
--- a/drivers/staging/android/binder.c
+++ b/drivers/staging/android/binder.c
@@ -409,11 +409,11 @@ repeat:
 		goto repeat;
 	}
 
-	FD_SET(fd, fdt->open_fds);
+	__set_open_fd(fd, fdt);
 	if (flags & O_CLOEXEC)
-		FD_SET(fd, fdt->close_on_exec);
+		__set_close_on_exec(fd, fdt);
 	else
-		FD_CLR(fd, fdt->close_on_exec);
+		__clear_close_on_exec(fd, fdt);
 	files->next_fd = fd + 1;
 #if 1
 	/* Sanity check */
@@ -454,7 +454,7 @@ static void task_fd_install(
 static void __put_unused_fd(struct files_struct *files, unsigned int fd)
 {
 	struct fdtable *fdt = files_fdtable(files);
-	__FD_CLR(fd, fdt->open_fds);
+	__clear_open_fd(fd, fdt);
 	if (fd < files->next_fd)
 		files->next_fd = fd;
 }
@@ -480,7 +480,7 @@ static long task_close_fd(struct binder_proc *proc, unsigned int fd)
 	if (!filp)
 		goto out_unlock;
 	rcu_assign_pointer(fdt->fd[fd], NULL);
-	FD_CLR(fd, fdt->close_on_exec);
+	__clear_close_on_exec(fd, fdt);
 	__put_unused_fd(files, fd);
 	spin_unlock(&files->file_lock);
 	retval = filp_close(filp, files);
diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c
index 76741d8..3dfd615 100644
--- a/fs/autofs4/dev-ioctl.c
+++ b/fs/autofs4/dev-ioctl.c
@@ -230,7 +230,7 @@ static void autofs_dev_ioctl_fd_install(unsigned int fd, struct file *file)
 	fdt = files_fdtable(files);
 	BUG_ON(fdt->fd[fd] != NULL);
 	rcu_assign_pointer(fdt->fd[fd], file);
-	FD_SET(fd, fdt->close_on_exec);
+	__set_close_on_exec(fd, fdt);
 	spin_unlock(&files->file_lock);
 }
 
diff --git a/fs/exec.c b/fs/exec.c
index 92ce83a..22cc38d 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -2078,8 +2078,8 @@ static int umh_pipe_setup(struct subprocess_info *info, struct cred *new)
 	fd_install(0, rp);
 	spin_lock(&cf->file_lock);
 	fdt = files_fdtable(cf);
-	FD_SET(0, fdt->open_fds);
-	FD_CLR(0, fdt->close_on_exec);
+	__set_open_fd(0, fdt);
+	__clear_close_on_exec(0, fdt);
 	spin_unlock(&cf->file_lock);
 
 	/* and disallow core files too */
diff --git a/fs/fcntl.c b/fs/fcntl.c
index 22764c7..75e7c1f 100644
--- a/fs/fcntl.c
+++ b/fs/fcntl.c
@@ -32,20 +32,20 @@ void set_close_on_exec(unsigned int fd, int flag)
 	spin_lock(&files->file_lock);
 	fdt = files_fdtable(files);
 	if (flag)
-		FD_SET(fd, fdt->close_on_exec);
+		__set_close_on_exec(fd, fdt);
 	else
-		FD_CLR(fd, fdt->close_on_exec);
+		__clear_close_on_exec(fd, fdt);
 	spin_unlock(&files->file_lock);
 }
 
-static int get_close_on_exec(unsigned int fd)
+static bool get_close_on_exec(unsigned int fd)
 {
 	struct files_struct *files = current->files;
 	struct fdtable *fdt;
-	int res;
+	bool res;
 	rcu_read_lock();
 	fdt = files_fdtable(files);
-	res = FD_ISSET(fd, fdt->close_on_exec);
+	res = close_on_exec(fd, fdt);
 	rcu_read_unlock();
 	return res;
 }
@@ -90,15 +90,15 @@ SYSCALL_DEFINE3(dup3, unsigned int, oldfd, unsigned int, newfd, int, flags)
 	err = -EBUSY;
 	fdt = files_fdtable(files);
 	tofree = fdt->fd[newfd];
-	if (!tofree && FD_ISSET(newfd, fdt->open_fds))
+	if (!tofree && fd_is_open(newfd, fdt))
 		goto out_unlock;
 	get_file(file);
 	rcu_assign_pointer(fdt->fd[newfd], file);
-	FD_SET(newfd, fdt->open_fds);
+	__set_open_fd(newfd, fdt);
 	if (flags & O_CLOEXEC)
-		FD_SET(newfd, fdt->close_on_exec);
+		__set_close_on_exec(newfd, fdt);
 	else
-		FD_CLR(newfd, fdt->close_on_exec);
+		__clear_close_on_exec(newfd, fdt);
 	spin_unlock(&files->file_lock);
 
 	if (tofree)
diff --git a/fs/file.c b/fs/file.c
index 4c6992d..114fea0 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -366,7 +366,7 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
 			 * is partway through open().  So make sure that this
 			 * fd is available to the new process.
 			 */
-			FD_CLR(open_files - i, new_fdt->open_fds);
+			__clear_open_fd(open_files - i, new_fdt);
 		}
 		rcu_assign_pointer(*new_fds++, f);
 	}
@@ -460,11 +460,11 @@ repeat:
 	if (start <= files->next_fd)
 		files->next_fd = fd + 1;
 
-	FD_SET(fd, fdt->open_fds);
+	__set_open_fd(fd, fdt);
 	if (flags & O_CLOEXEC)
-		FD_SET(fd, fdt->close_on_exec);
+		__set_close_on_exec(fd, fdt);
 	else
-		FD_CLR(fd, fdt->close_on_exec);
+		__clear_close_on_exec(fd, fdt);
 	error = fd;
 #if 1
 	/* Sanity check */
diff --git a/fs/open.c b/fs/open.c
index 77becc0..5720854 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -836,7 +836,7 @@ EXPORT_SYMBOL(dentry_open);
 static void __put_unused_fd(struct files_struct *files, unsigned int fd)
 {
 	struct fdtable *fdt = files_fdtable(files);
-	__FD_CLR(fd, fdt->open_fds);
+	__clear_open_fd(fd, fdt);
 	if (fd < files->next_fd)
 		files->next_fd = fd;
 }
@@ -1080,7 +1080,7 @@ SYSCALL_DEFINE1(close, unsigned int, fd)
 	if (!filp)
 		goto out_unlock;
 	rcu_assign_pointer(fdt->fd[fd], NULL);
-	FD_CLR(fd, fdt->close_on_exec);
+	__clear_close_on_exec(fd, fdt);
 	__put_unused_fd(files, fd);
 	spin_unlock(&files->file_lock);
 	retval = filp_close(filp, files);
diff --git a/fs/proc/base.c b/fs/proc/base.c
index d4548dd..db6ab4b 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -1754,7 +1754,7 @@ static int proc_fd_info(struct inode *inode, struct path *path, char *info)
 
 			fdt = files_fdtable(files);
 			f_flags = file->f_flags & ~O_CLOEXEC;
-			if (FD_ISSET(fd, fdt->close_on_exec))
+			if (close_on_exec(fd, fdt))
 				f_flags |= O_CLOEXEC;
 
 			if (path) {
diff --git a/include/linux/fdtable.h b/include/linux/fdtable.h
index 82163c4..7675da2 100644
--- a/include/linux/fdtable.h
+++ b/include/linux/fdtable.h
@@ -38,6 +38,36 @@ struct fdtable {
 	struct fdtable *next;
 };
 
+static inline void __set_close_on_exec(int fd, struct fdtable *fdt)
+{
+	FD_SET(fd, fdt->close_on_exec);
+}
+
+static inline void __clear_close_on_exec(int fd, struct fdtable *fdt)
+{
+	FD_CLR(fd, fdt->close_on_exec);
+}
+
+static inline bool close_on_exec(int fd, const struct fdtable *fdt)
+{
+	return FD_ISSET(fd, fdt->close_on_exec);
+}
+
+static inline void __set_open_fd(int fd, struct fdtable *fdt)
+{
+	FD_SET(fd, fdt->open_fds);
+}
+
+static inline void __clear_open_fd(int fd, struct fdtable *fdt)
+{
+	FD_CLR(fd, fdt->open_fds);
+}
+
+static inline bool fd_is_open(int fd, const struct fdtable *fdt)
+{
+	return FD_ISSET(fd, fdt->open_fds);
+}
+
 /*
  * Open file table structure
  */

[PATCH 2/3] Replace the fd_sets in struct fdtable with an array of unsigned longs

[David Howells]

Replace the fd_sets in struct fdtable with an array of unsigned longs and then
use the standard non-atomic bit operations rather than the FD_* macros.

This:

 (1) Removes the abuses of struct fd_set:

     (a) Since we don't want to allocate a full fd_set the vast majority of the
     	 time, we actually, in effect, just allocate a just-big-enough array of
     	 unsigned longs and cast it to an fd_set type - so why bother with the
     	 fd_set at all?

     (b) Some places outside of the core fdtable handling code (such as
     	 SELinux) want to look inside the array of unsigned longs hidden inside
     	 the fd_set struct for more efficient iteration over the entire set.

 (2) Eliminates the use of FD_*() macros in the kernel completely.

 (3) Permits the __FD_*() macros to be deleted entirely where not exposed to
     userspace.

Signed-off-by: David Howells <dhowells@redhat.com>
---

 fs/exec.c                |    4 ++--
 fs/file.c                |   46 ++++++++++++++++++++++------------------------
 fs/select.c              |    2 +-
 include/linux/fdtable.h  |   28 ++++++++++------------------
 kernel/exit.c            |    2 +-
 security/selinux/hooks.c |    2 +-
 6 files changed, 37 insertions(+), 47 deletions(-)

diff --git a/fs/exec.c b/fs/exec.c
index 22cc38d..cfd5e30 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -1026,10 +1026,10 @@ static void flush_old_files(struct files_struct * files)
 		fdt = files_fdtable(files);
 		if (i >= fdt->max_fds)
 			break;
-		set = fdt->close_on_exec->fds_bits[j];
+		set = fdt->close_on_exec[j];
 		if (!set)
 			continue;
-		fdt->close_on_exec->fds_bits[j] = 0;
+		fdt->close_on_exec[j] = 0;
 		spin_unlock(&files->file_lock);
 		for ( ; set ; i++,set >>= 1) {
 			if (set & 1) {
diff --git a/fs/file.c b/fs/file.c
index 114fea0..2d479dd 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -40,7 +40,7 @@ int sysctl_nr_open_max = 1024 * 1024; /* raised later */
  */
 static DEFINE_PER_CPU(struct fdtable_defer, fdtable_defer_list);
 
-static void *alloc_fdmem(unsigned int size)
+static void *alloc_fdmem(size_t size)
 {
 	/*
 	 * Very large allocations can stress page reclaim, so fall back to
@@ -142,7 +142,7 @@ static void copy_fdtable(struct fdtable *nfdt, struct fdtable *ofdt)
 static struct fdtable * alloc_fdtable(unsigned int nr)
 {
 	struct fdtable *fdt;
-	char *data;
+	void *data;
 
 	/*
 	 * Figure out how many fds we actually want to support in this fdtable.
@@ -172,14 +172,15 @@ static struct fdtable * alloc_fdtable(unsigned int nr)
 	data = alloc_fdmem(nr * sizeof(struct file *));
 	if (!data)
 		goto out_fdt;
-	fdt->fd = (struct file **)data;
-	data = alloc_fdmem(max_t(unsigned int,
+	fdt->fd = data;
+
+	data = alloc_fdmem(max_t(size_t,
 				 2 * nr / BITS_PER_BYTE, L1_CACHE_BYTES));
 	if (!data)
 		goto out_arr;
-	fdt->open_fds = (fd_set *)data;
-	data += nr / BITS_PER_BYTE;
-	fdt->close_on_exec = (fd_set *)data;
+	fdt->open_fds = data;
+	data += nr / BITS_PER_LONG;
+	fdt->close_on_exec = data;
 	fdt->next = NULL;
 
 	return fdt;
@@ -275,11 +276,11 @@ static int count_open_files(struct fdtable *fdt)
 	int i;
 
 	/* Find the last open fd */
-	for (i = size/(8*sizeof(long)); i > 0; ) {
-		if (fdt->open_fds->fds_bits[--i])
+	for (i = size / BITS_PER_LONG; i > 0; ) {
+		if (fdt->open_fds[--i])
 			break;
 	}
-	i = (i+1) * 8 * sizeof(long);
+	i = (i + 1) * BITS_PER_LONG;
 	return i;
 }
 
@@ -306,8 +307,8 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
 	newf->next_fd = 0;
 	new_fdt = &newf->fdtab;
 	new_fdt->max_fds = NR_OPEN_DEFAULT;
-	new_fdt->close_on_exec = (fd_set *)&newf->close_on_exec_init;
-	new_fdt->open_fds = (fd_set *)&newf->open_fds_init;
+	new_fdt->close_on_exec = newf->close_on_exec_init;
+	new_fdt->open_fds = newf->open_fds_init;
 	new_fdt->fd = &newf->fd_array[0];
 	new_fdt->next = NULL;
 
@@ -350,10 +351,8 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
 	old_fds = old_fdt->fd;
 	new_fds = new_fdt->fd;
 
-	memcpy(new_fdt->open_fds->fds_bits,
-		old_fdt->open_fds->fds_bits, open_files/8);
-	memcpy(new_fdt->close_on_exec->fds_bits,
-		old_fdt->close_on_exec->fds_bits, open_files/8);
+	memcpy(new_fdt->open_fds, old_fdt->open_fds, open_files / 8);
+	memcpy(new_fdt->close_on_exec, old_fdt->close_on_exec, open_files / 8);
 
 	for (i = open_files; i != 0; i--) {
 		struct file *f = *old_fds++;
@@ -379,11 +378,11 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
 	memset(new_fds, 0, size);
 
 	if (new_fdt->max_fds > open_files) {
-		int left = (new_fdt->max_fds-open_files)/8;
-		int start = open_files / (8 * sizeof(unsigned long));
+		int left = (new_fdt->max_fds - open_files) / 8;
+		int start = open_files / BITS_PER_LONG;
 
-		memset(&new_fdt->open_fds->fds_bits[start], 0, left);
-		memset(&new_fdt->close_on_exec->fds_bits[start], 0, left);
+		memset(&new_fdt->open_fds[start], 0, left);
+		memset(&new_fdt->close_on_exec[start], 0, left);
 	}
 
 	rcu_assign_pointer(newf->fdt, new_fdt);
@@ -419,8 +418,8 @@ struct files_struct init_files = {
 	.fdtab		= {
 		.max_fds	= NR_OPEN_DEFAULT,
 		.fd		= &init_files.fd_array[0],
-		.close_on_exec	= (fd_set *)&init_files.close_on_exec_init,
-		.open_fds	= (fd_set *)&init_files.open_fds_init,
+		.close_on_exec	= init_files.close_on_exec_init,
+		.open_fds	= init_files.open_fds_init,
 	},
 	.file_lock	= __SPIN_LOCK_UNLOCKED(init_task.file_lock),
 };
@@ -443,8 +442,7 @@ repeat:
 		fd = files->next_fd;
 
 	if (fd < fdt->max_fds)
-		fd = find_next_zero_bit(fdt->open_fds->fds_bits,
-					   fdt->max_fds, fd);
+		fd = find_next_zero_bit(fdt->open_fds, fdt->max_fds, fd);
 
 	error = expand_files(files, fd);
 	if (error < 0)
diff --git a/fs/select.c b/fs/select.c
index d33418f..2e7fbe8 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -348,7 +348,7 @@ static int max_select_fd(unsigned long n, fd_set_bits *fds)
 	set = ~(~0UL << (n & (__NFDBITS-1)));
 	n /= __NFDBITS;
 	fdt = files_fdtable(current->files);
-	open_fds = fdt->open_fds->fds_bits+n;
+	open_fds = fdt->open_fds + n;
 	max = 0;
 	if (set) {
 		set &= BITS(fds, n);
diff --git a/include/linux/fdtable.h b/include/linux/fdtable.h
index 7675da2..158a41e 100644
--- a/include/linux/fdtable.h
+++ b/include/linux/fdtable.h
@@ -21,51 +21,43 @@
  */
 #define NR_OPEN_DEFAULT BITS_PER_LONG
 
-/*
- * The embedded_fd_set is a small fd_set,
- * suitable for most tasks (which open <= BITS_PER_LONG files)
- */
-struct embedded_fd_set {
-	unsigned long fds_bits[1];
-};
-
 struct fdtable {
 	unsigned int max_fds;
 	struct file __rcu **fd;      /* current fd array */
-	fd_set *close_on_exec;
-	fd_set *open_fds;
+	unsigned long *close_on_exec;
+	unsigned long *open_fds;
 	struct rcu_head rcu;
 	struct fdtable *next;
 };
 
 static inline void __set_close_on_exec(int fd, struct fdtable *fdt)
 {
-	FD_SET(fd, fdt->close_on_exec);
+	__set_bit(fd, fdt->close_on_exec);
 }
 
 static inline void __clear_close_on_exec(int fd, struct fdtable *fdt)
 {
-	FD_CLR(fd, fdt->close_on_exec);
+	__clear_bit(fd, fdt->close_on_exec);
 }
 
 static inline bool close_on_exec(int fd, const struct fdtable *fdt)
 {
-	return FD_ISSET(fd, fdt->close_on_exec);
+	return test_bit(fd, fdt->close_on_exec);
 }
 
 static inline void __set_open_fd(int fd, struct fdtable *fdt)
 {
-	FD_SET(fd, fdt->open_fds);
+	__set_bit(fd, fdt->open_fds);
 }
 
 static inline void __clear_open_fd(int fd, struct fdtable *fdt)
 {
-	FD_CLR(fd, fdt->open_fds);
+	__clear_bit(fd, fdt->open_fds);
 }
 
 static inline bool fd_is_open(int fd, const struct fdtable *fdt)
 {
-	return FD_ISSET(fd, fdt->open_fds);
+	return test_bit(fd, fdt->open_fds);
 }
 
 /*
@@ -83,8 +75,8 @@ struct files_struct {
    */
 	spinlock_t file_lock ____cacheline_aligned_in_smp;
 	int next_fd;
-	struct embedded_fd_set close_on_exec_init;
-	struct embedded_fd_set open_fds_init;
+	unsigned long close_on_exec_init[1];
+	unsigned long open_fds_init[1];
 	struct file __rcu * fd_array[NR_OPEN_DEFAULT];
 };
 
diff --git a/kernel/exit.c b/kernel/exit.c
index 4b4042f..4db0200 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -473,7 +473,7 @@ static void close_files(struct files_struct * files)
 		i = j * __NFDBITS;
 		if (i >= fdt->max_fds)
 			break;
-		set = fdt->open_fds->fds_bits[j++];
+		set = fdt->open_fds[j++];
 		while (set) {
 			if (set & 1) {
 				struct file * file = xchg(&fdt->fd[i], NULL);
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 6a3683e..421c990 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -2145,7 +2145,7 @@ static inline void flush_unauthorized_files(const struct cred *cred,
 		fdt = files_fdtable(files);
 		if (i >= fdt->max_fds)
 			break;
-		set = fdt->open_fds->fds_bits[j];
+		set = fdt->open_fds[j];
 		if (!set)
 			continue;
 		spin_unlock(&files->file_lock);

[PATCH 3/3] Delete the __FD_*() funcs for operating on fd_set from linux/time.h

[David Howells]

Delete the __FD_*() functions for operating on fd_set structs from
linux/time.h as they're no longer used within the kernel with the preceding
patch and are not exported to userspace.

Whilst linux/time.h *does* export the FD_*() equivalents as wrappers around
__FD_*(), userspace provides its own definition of __FD_*().

Note that the definition of FD_ZERO() in linux/time.h may not be used with the
fd_sets associated with struct fdtable as the fd_set may have been allocated in
a truncated fashion.

Signed-off-by: David Howells <dhowells@redhat.com>
---

 include/linux/time.h |   22 ----------------------
 1 files changed, 0 insertions(+), 22 deletions(-)

diff --git a/include/linux/time.h b/include/linux/time.h
index ca9b1e6..9f43487 100644
--- a/include/linux/time.h
+++ b/include/linux/time.h
@@ -4,9 +4,7 @@
 #include <linux/types.h>
 
 #ifdef __KERNEL__
-# include <linux/bitops.h>
 # include <linux/cache.h>
-# include <linux/posix_types.h>
 # include <linux/seqlock.h>
 # include <linux/math64.h>
 #endif
@@ -259,26 +257,6 @@ static __always_inline void timespec_add_ns(struct timespec *a, u64 ns)
 	a->tv_nsec = ns;
 }
 
-static inline void __FD_SET(unsigned long __fd, __kernel_fd_set *__fdsetp)
-{
-	__set_bit( __fd, __fdsetp->fds_bits);
-}
-
-static inline void __FD_CLR(unsigned long __fd, __kernel_fd_set *__fdsetp)
-{
-	__clear_bit(__fd, __fdsetp->fds_bits);
-}
-
-static inline int __FD_ISSET(unsigned long __fd, const __kernel_fd_set *__fdsetp)
-{
-	return test_bit(__fd, __fdsetp->fds_bits);
-}
-
-static inline void __FD_ZERO(__kernel_fd_set *__fdsetp)
-{
-	memset(__fdsetp->fds_bits, 0, sizeof __fdsetp->fds_bits);
-}
-
 #endif /* __KERNEL__ */
 
 #define NFDBITS			__NFDBITS