[PATCH] files: break up files struct

In order for the RCU to work, the file table array, sets and their sizes must
be updated atomically.  Instead of ensuring this through too many memory
barriers, we put the arrays and their sizes in a separate structure.  This
patch takes the first step of putting the file table elements in a separate
structure fdtable that is embedded withing files_struct.  It also changes all
the users to refer to the file table using files_fdtable() macro.  Subsequent
applciation of RCU becomes easier after this.

Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

8 files changed, 104 insertions, 63 deletions

diff --git a/fs/exec.c b/fs/exec.c
index 222ab1c..14dd039 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -798,6 +798,7 @@ no_thread_group:
 static inline void flush_old_files(struct files_struct * files)
 {
 	long j = -1;
+	struct fdtable *fdt;
 
 	spin_lock(&files->file_lock);
 	for (;;) {
@@ -805,12 +806,13 @@ static inline void flush_old_files(struct files_struct * files)
 
 		j++;
 		i = j * __NFDBITS;
-		if (i >= files->max_fds || i >= files->max_fdset)
+		fdt = files_fdtable(files);
+		if (i >= fdt->max_fds || i >= fdt->max_fdset)
 			break;
-		set = files->close_on_exec->fds_bits[j];
+		set = fdt->close_on_exec->fds_bits[j];
 		if (!set)
 			continue;
-		files->close_on_exec->fds_bits[j] = 0;
+		fdt->close_on_exec->fds_bits[j] = 0;
 		spin_unlock(&files->file_lock);
 		for ( ; set ; i++,set >>= 1) {
 			if (set & 1) {
diff --git a/fs/fcntl.c b/fs/fcntl.c
index 6fbc9d8..bfecc62 100644
--- a/fs/fcntl.c
+++ b/fs/fcntl.c
@@ -24,20 +24,24 @@
 void fastcall set_close_on_exec(unsigned int fd, int flag)
 {
 	struct files_struct *files = current->files;
+	struct fdtable *fdt;
 	spin_lock(&files->file_lock);
+	fdt = files_fdtable(files);
 	if (flag)
-		FD_SET(fd, files->close_on_exec);
+		FD_SET(fd, fdt->close_on_exec);
 	else
-		FD_CLR(fd, files->close_on_exec);
+		FD_CLR(fd, fdt->close_on_exec);
 	spin_unlock(&files->file_lock);
 }
 
 static inline int get_close_on_exec(unsigned int fd)
 {
 	struct files_struct *files = current->files;
+	struct fdtable *fdt;
 	int res;
 	spin_lock(&files->file_lock);
-	res = FD_ISSET(fd, files->close_on_exec);
+	fdt = files_fdtable(files);
+	res = FD_ISSET(fd, fdt->close_on_exec);
 	spin_unlock(&files->file_lock);
 	return res;
 }
@@ -54,24 +58,26 @@ static int locate_fd(struct files_struct *files,
 	unsigned int newfd;
 	unsigned int start;
 	int error;
+	struct fdtable *fdt;
 
 	error = -EINVAL;
 	if (orig_start >= current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
 		goto out;
 
+	fdt = files_fdtable(files);
 repeat:
 	/*
 	 * Someone might have closed fd's in the range
-	 * orig_start..files->next_fd
+	 * orig_start..fdt->next_fd
 	 */
 	start = orig_start;
-	if (start < files->next_fd)
-		start = files->next_fd;
+	if (start < fdt->next_fd)
+		start = fdt->next_fd;
 
 	newfd = start;
-	if (start < files->max_fdset) {
-		newfd = find_next_zero_bit(files->open_fds->fds_bits,
-			files->max_fdset, start);
+	if (start < fdt->max_fdset) {
+		newfd = find_next_zero_bit(fdt->open_fds->fds_bits,
+			fdt->max_fdset, start);
 	}
 	
 	error = -EMFILE;
@@ -89,8 +95,8 @@ repeat:
 	if (error)
 		goto repeat;
 
-	if (start <= files->next_fd)
-		files->next_fd = newfd + 1;
+	if (start <= fdt->next_fd)
+		fdt->next_fd = newfd + 1;
 	
 	error = newfd;
 	
@@ -101,13 +107,16 @@ out:
 static int dupfd(struct file *file, unsigned int start)
 {
 	struct files_struct * files = current->files;
+	struct fdtable *fdt;
 	int fd;
 
 	spin_lock(&files->file_lock);
 	fd = locate_fd(files, file, start);
 	if (fd >= 0) {
-		FD_SET(fd, files->open_fds);
-		FD_CLR(fd, files->close_on_exec);
+		/* 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);
 		spin_unlock(&files->file_lock);
 		fd_install(fd, file);
 	} else {
@@ -123,6 +132,7 @@ asmlinkage long sys_dup2(unsigned int oldfd, unsigned int newfd)
 	int err = -EBADF;
 	struct file * file, *tofree;
 	struct files_struct * files = current->files;
+	struct fdtable *fdt;
 
 	spin_lock(&files->file_lock);
 	if (!(file = fcheck(oldfd)))
@@ -148,13 +158,14 @@ asmlinkage long sys_dup2(unsigned int oldfd, unsigned int newfd)
 
 	/* Yes. It's a race. In user space. Nothing sane to do */
 	err = -EBUSY;
-	tofree = files->fd[newfd];
-	if (!tofree && FD_ISSET(newfd, files->open_fds))
+	fdt = files_fdtable(files);
+	tofree = fdt->fd[newfd];
+	if (!tofree && FD_ISSET(newfd, fdt->open_fds))
 		goto out_fput;
 
-	files->fd[newfd] = file;
-	FD_SET(newfd, files->open_fds);
-	FD_CLR(newfd, files->close_on_exec);
+	fdt->fd[newfd] = file;
+	FD_SET(newfd, fdt->open_fds);
+	FD_CLR(newfd, fdt->close_on_exec);
 	spin_unlock(&files->file_lock);
 
 	if (tofree)
diff --git a/fs/file.c b/fs/file.c
index 92b5f25..f5926ce 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -59,13 +59,15 @@ static int expand_fd_array(struct files_struct *files, int nr)
 {
 	struct file **new_fds;
 	int error, nfds;
+	struct fdtable *fdt;
 
 	
 	error = -EMFILE;
-	if (files->max_fds >= NR_OPEN || nr >= NR_OPEN)
+	fdt = files_fdtable(files);
+	if (fdt->max_fds >= NR_OPEN || nr >= NR_OPEN)
 		goto out;
 
-	nfds = files->max_fds;
+	nfds = fdt->max_fds;
 	spin_unlock(&files->file_lock);
 
 	/* 
@@ -95,13 +97,14 @@ static int expand_fd_array(struct files_struct *files, int nr)
 		goto out;
 
 	/* Copy the existing array and install the new pointer */
+	fdt = files_fdtable(files);
 
-	if (nfds > files->max_fds) {
+	if (nfds > fdt->max_fds) {
 		struct file **old_fds;
 		int i;
 		
-		old_fds = xchg(&files->fd, new_fds);
-		i = xchg(&files->max_fds, nfds);
+		old_fds = xchg(&fdt->fd, new_fds);
+		i = xchg(&fdt->max_fds, nfds);
 
 		/* Don't copy/clear the array if we are creating a new
 		   fd array for fork() */
@@ -164,12 +167,14 @@ static int expand_fdset(struct files_struct *files, int nr)
 {
 	fd_set *new_openset = NULL, *new_execset = NULL;
 	int error, nfds = 0;
+	struct fdtable *fdt;
 
 	error = -EMFILE;
-	if (files->max_fdset >= NR_OPEN || nr >= NR_OPEN)
+	fdt = files_fdtable(files);
+	if (fdt->max_fdset >= NR_OPEN || nr >= NR_OPEN)
 		goto out;
 
-	nfds = files->max_fdset;
+	nfds = fdt->max_fdset;
 	spin_unlock(&files->file_lock);
 
 	/* Expand to the max in easy steps */
@@ -193,24 +198,25 @@ static int expand_fdset(struct files_struct *files, int nr)
 	error = 0;
 	
 	/* Copy the existing tables and install the new pointers */
-	if (nfds > files->max_fdset) {
-		int i = files->max_fdset / (sizeof(unsigned long) * 8);
-		int count = (nfds - files->max_fdset) / 8;
+	fdt = files_fdtable(files);
+	if (nfds > fdt->max_fdset) {
+		int i = fdt->max_fdset / (sizeof(unsigned long) * 8);
+		int count = (nfds - fdt->max_fdset) / 8;
 		
 		/* 
 		 * Don't copy the entire array if the current fdset is
 		 * not yet initialised.  
 		 */
 		if (i) {
-			memcpy (new_openset, files->open_fds, files->max_fdset/8);
-			memcpy (new_execset, files->close_on_exec, files->max_fdset/8);
+			memcpy (new_openset, fdt->open_fds, fdt->max_fdset/8);
+			memcpy (new_execset, fdt->close_on_exec, fdt->max_fdset/8);
 			memset (&new_openset->fds_bits[i], 0, count);
 			memset (&new_execset->fds_bits[i], 0, count);
 		}
 		
-		nfds = xchg(&files->max_fdset, nfds);
-		new_openset = xchg(&files->open_fds, new_openset);
-		new_execset = xchg(&files->close_on_exec, new_execset);
+		nfds = xchg(&fdt->max_fdset, nfds);
+		new_openset = xchg(&fdt->open_fds, new_openset);
+		new_execset = xchg(&fdt->close_on_exec, new_execset);
 		spin_unlock(&files->file_lock);
 		free_fdset (new_openset, nfds);
 		free_fdset (new_execset, nfds);
@@ -237,13 +243,15 @@ out:
 int expand_files(struct files_struct *files, int nr)
 {
 	int err, expand = 0;
+	struct fdtable *fdt;
 
-	if (nr >= files->max_fdset) {
+	fdt = files_fdtable(files);
+	if (nr >= fdt->max_fdset) {
 		expand = 1;
 		if ((err = expand_fdset(files, nr)))
 			goto out;
 	}
-	if (nr >= files->max_fds) {
+	if (nr >= fdt->max_fds) {
 		expand = 1;
 		if ((err = expand_fd_array(files, nr)))
 			goto out;
diff --git a/fs/locks.c b/fs/locks.c
index 11956b6..c2c09b4 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -2198,21 +2198,23 @@ void steal_locks(fl_owner_t from)
 {
 	struct files_struct *files = current->files;
 	int i, j;
+	struct fdtable *fdt;
 
 	if (from == files)
 		return;
 
 	lock_kernel();
 	j = 0;
+	fdt = files_fdtable(files);
 	for (;;) {
 		unsigned long set;
 		i = j * __NFDBITS;
-		if (i >= files->max_fdset || i >= files->max_fds)
+		if (i >= fdt->max_fdset || i >= fdt->max_fds)
 			break;
-		set = files->open_fds->fds_bits[j++];
+		set = fdt->open_fds->fds_bits[j++];
 		while (set) {
 			if (set & 1) {
-				struct file *file = files->fd[i];
+				struct file *file = fdt->fd[i];
 				if (file)
 					__steal_locks(file, from);
 			}
diff --git a/fs/open.c b/fs/open.c
index 4ee2dcc..b654251 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -842,14 +842,16 @@ int get_unused_fd(void)
 {
 	struct files_struct * files = current->files;
 	int fd, error;
+	struct fdtable *fdt;
 
   	error = -EMFILE;
 	spin_lock(&files->file_lock);
 
 repeat:
- 	fd = find_next_zero_bit(files->open_fds->fds_bits, 
-				files->max_fdset, 
-				files->next_fd);
+	fdt = files_fdtable(files);
+ 	fd = find_next_zero_bit(fdt->open_fds->fds_bits,
+				fdt->max_fdset,
+				fdt->next_fd);
 
 	/*
 	 * N.B. For clone tasks sharing a files structure, this test
@@ -872,14 +874,14 @@ repeat:
 		goto repeat;
 	}
 
-	FD_SET(fd, files->open_fds);
-	FD_CLR(fd, files->close_on_exec);
-	files->next_fd = fd + 1;
+	FD_SET(fd, fdt->open_fds);
+	FD_CLR(fd, fdt->close_on_exec);
+	fdt->next_fd = fd + 1;
 #if 1
 	/* Sanity check */
-	if (files->fd[fd] != NULL) {
+	if (fdt->fd[fd] != NULL) {
 		printk(KERN_WARNING "get_unused_fd: slot %d not NULL!\n", fd);
-		files->fd[fd] = NULL;
+		fdt->fd[fd] = NULL;
 	}
 #endif
 	error = fd;
@@ -893,9 +895,10 @@ EXPORT_SYMBOL(get_unused_fd);
 
 static inline void __put_unused_fd(struct files_struct *files, unsigned int fd)
 {
-	__FD_CLR(fd, files->open_fds);
-	if (fd < files->next_fd)
-		files->next_fd = fd;
+	struct fdtable *fdt = files_fdtable(files);
+	__FD_CLR(fd, fdt->open_fds);
+	if (fd < fdt->next_fd)
+		fdt->next_fd = fd;
 }
 
 void fastcall put_unused_fd(unsigned int fd)
@@ -924,10 +927,12 @@ EXPORT_SYMBOL(put_unused_fd);
 void fastcall fd_install(unsigned int fd, struct file * file)
 {
 	struct files_struct *files = current->files;
+	struct fdtable *fdt;
 	spin_lock(&files->file_lock);
-	if (unlikely(files->fd[fd] != NULL))
+	fdt = files_fdtable(files);
+	if (unlikely(fdt->fd[fd] != NULL))
 		BUG();
-	files->fd[fd] = file;
+	fdt->fd[fd] = file;
 	spin_unlock(&files->file_lock);
 }
 
@@ -1010,15 +1015,17 @@ asmlinkage long sys_close(unsigned int fd)
 {
 	struct file * filp;
 	struct files_struct *files = current->files;
+	struct fdtable *fdt;
 
 	spin_lock(&files->file_lock);
-	if (fd >= files->max_fds)
+	fdt = files_fdtable(files);
+	if (fd >= fdt->max_fds)
 		goto out_unlock;
-	filp = files->fd[fd];
+	filp = fdt->fd[fd];
 	if (!filp)
 		goto out_unlock;
-	files->fd[fd] = NULL;
-	FD_CLR(fd, files->close_on_exec);
+	fdt->fd[fd] = NULL;
+	FD_CLR(fd, fdt->close_on_exec);
 	__put_unused_fd(files, fd);
 	spin_unlock(&files->file_lock);
 	return filp_close(filp, files);
diff --git a/fs/proc/array.c b/fs/proc/array.c
index 37668fe..d88d518 100644
--- a/fs/proc/array.c
+++ b/fs/proc/array.c
@@ -159,6 +159,7 @@ static inline char * task_state(struct task_struct *p, char *buffer)
 {
 	struct group_info *group_info;
 	int g;
+	struct fdtable *fdt = NULL;
 
 	read_lock(&tasklist_lock);
 	buffer += sprintf(buffer,
@@ -179,10 +180,12 @@ static inline char * task_state(struct task_struct *p, char *buffer)
 		p->gid, p->egid, p->sgid, p->fsgid);
 	read_unlock(&tasklist_lock);
 	task_lock(p);
+	if (p->files)
+		fdt = files_fdtable(p->files);
 	buffer += sprintf(buffer,
 		"FDSize:\t%d\n"
 		"Groups:\t",
-		p->files ? p->files->max_fds : 0);
+		fdt ? fdt->max_fds : 0);
 
 	group_info = p->group_info;
 	get_group_info(group_info);
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 84751f3..d0087a0 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -1039,6 +1039,7 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir)
 	int retval;
 	char buf[NUMBUF];
 	struct files_struct * files;
+	struct fdtable *fdt;
 
 	retval = -ENOENT;
 	if (!pid_alive(p))
@@ -1062,8 +1063,9 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir)
 			if (!files)
 				goto out;
 			spin_lock(&files->file_lock);
+			fdt = files_fdtable(files);
 			for (fd = filp->f_pos-2;
-			     fd < files->max_fds;
+			     fd < fdt->max_fds;
 			     fd++, filp->f_pos++) {
 				unsigned int i,j;
 
diff --git a/fs/select.c b/fs/select.c
index b80e7eb..2e56325 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -132,11 +132,13 @@ static int max_select_fd(unsigned long n, fd_set_bits *fds)
 	unsigned long *open_fds;
 	unsigned long set;
 	int max;
+	struct fdtable *fdt;
 
 	/* handle last in-complete long-word first */
 	set = ~(~0UL << (n & (__NFDBITS-1)));
 	n /= __NFDBITS;
-	open_fds = current->files->open_fds->fds_bits+n;
+	fdt = files_fdtable(current->files);
+	open_fds = fdt->open_fds->fds_bits+n;
 	max = 0;
 	if (set) {
 		set &= BITS(fds, n);
@@ -299,6 +301,7 @@ sys_select(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, s
 	char *bits;
 	long timeout;
 	int ret, size, max_fdset;
+	struct fdtable *fdt;
 
 	timeout = MAX_SCHEDULE_TIMEOUT;
 	if (tvp) {
@@ -326,7 +329,8 @@ sys_select(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, s
 		goto out_nofds;
 
 	/* max_fdset can increase, so grab it once to avoid race */
-	max_fdset = current->files->max_fdset;
+	fdt = files_fdtable(current->files);
+	max_fdset = fdt->max_fdset;
 	if (n > max_fdset)
 		n = max_fdset;
 
@@ -464,9 +468,11 @@ asmlinkage long sys_poll(struct pollfd __user * ufds, unsigned int nfds, long ti
  	unsigned int i;
 	struct poll_list *head;
  	struct poll_list *walk;
+	struct fdtable *fdt;
 
 	/* Do a sanity check on nfds ... */
-	if (nfds > current->files->max_fdset && nfds > OPEN_MAX)
+	fdt = files_fdtable(current->files);
+	if (nfds > fdt->max_fdset && nfds > OPEN_MAX)
 		return -EINVAL;
 
 	if (timeout) {