Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull vfs update from Al Viro:

 - big one - consolidation of descriptor-related logics; almost all of
   that is moved to fs/file.c

   (BTW, I'm seriously tempted to rename the result to fd.c.  As it is,
   we have a situation when file_table.c is about handling of struct
   file and file.c is about handling of descriptor tables; the reasons
   are historical - file_table.c used to be about a static array of
   struct file we used to have way back).

   A lot of stray ends got cleaned up and converted to saner primitives,
   disgusting mess in android/binder.c is still disgusting, but at least
   doesn't poke so much in descriptor table guts anymore.  A bunch of
   relatively minor races got fixed in process, plus an ext4 struct file
   leak.

 - related thing - fget_light() partially unuglified; see fdget() in
   there (and yes, it generates the code as good as we used to have).

 - also related - bits of Cyrill's procfs stuff that got entangled into
   that work; _not_ all of it, just the initial move to fs/proc/fd.c and
   switch of fdinfo to seq_file.

 - Alex's fs/coredump.c spiltoff - the same story, had been easier to
   take that commit than mess with conflicts.  The rest is a separate
   pile, this was just a mechanical code movement.

 - a few misc patches all over the place.  Not all for this cycle,
   there'll be more (and quite a few currently sit in akpm's tree)."

Fix up trivial conflicts in the android binder driver, and some fairly
simple conflicts due to two different changes to the sock_alloc_file()
interface ("take descriptor handling from sock_alloc_file() to callers"
vs "net: Providing protocol type via system.sockprotoname xattr of
/proc/PID/fd entries" adding a dentry name to the socket)

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (72 commits)
  MAX_LFS_FILESIZE should be a loff_t
  compat: fs: Generic compat_sys_sendfile implementation
  fs: push rcu_barrier() from deactivate_locked_super() to filesystems
  btrfs: reada_extent doesn't need kref for refcount
  coredump: move core dump functionality into its own file
  coredump: prevent double-free on an error path in core dumper
  usb/gadget: fix misannotations
  fcntl: fix misannotations
  ceph: don't abuse d_delete() on failure exits
  hypfs: ->d_parent is never NULL or negative
  vfs: delete surplus inode NULL check
  switch simple cases of fget_light to fdget
  new helpers: fdget()/fdput()
  switch o2hb_region_dev_write() to fget_light()
  proc_map_files_readdir(): don't bother with grabbing files
  make get_file() return its argument
  vhost_set_vring(): turn pollstart/pollstop into bool
  switch prctl_set_mm_exe_file() to fget_light()
  switch xfs_find_handle() to fget_light()
  switch xfs_swapext() to fget_light()
  ...

arch/alpha/kernel/osf_sys.c

@@ -145,27 +145,24 @@ SYSCALL_DEFINE4(osf_getdirentries, unsigned int, fd,
 		long __user *, basep)
 {
 	int error;
-	struct file *file;
+	struct fd arg = fdget(fd);
 	struct osf_dirent_callback buf;
 
-	error = -EBADF;
-	file = fget(fd);
-	if (!file)
-		goto out;
+	if (!arg.file)
+		return -EBADF;
 
 	buf.dirent = dirent;
 	buf.basep = basep;
 	buf.count = count;
 	buf.error = 0;
 
-	error = vfs_readdir(file, osf_filldir, &buf);
+	error = vfs_readdir(arg.file, osf_filldir, &buf);
 	if (error >= 0)
 		error = buf.error;
 	if (count != buf.count)
 		error = count - buf.count;
 
-	fput(file);
- out:
+	fdput(arg);
 	return error;
 }
 

arch/ia64/kernel/perfmon.c

@@ -2306,7 +2306,7 @@ pfm_smpl_buffer_alloc(struct task_struct *task, struct file *filp, pfm_context_t
 	 * partially initialize the vma for the sampling buffer
 	 */
 	vma->vm_mm	     = mm;
-	vma->vm_file	     = filp;
+	vma->vm_file	     = get_file(filp);
 	vma->vm_flags	     = VM_READ| VM_MAYREAD |VM_RESERVED;
 	vma->vm_page_prot    = PAGE_READONLY; /* XXX may need to change */
 
@@ -2345,8 +2345,6 @@ pfm_smpl_buffer_alloc(struct task_struct *task, struct file *filp, pfm_context_t
 		goto error;
 	}
 
-	get_file(filp);
-
 	/*
 	 * now insert the vma in the vm list for the process, must be
 	 * done with mmap lock held
@@ -4782,7 +4780,7 @@ recheck:
 asmlinkage long
 sys_perfmonctl (int fd, int cmd, void __user *arg, int count)
 {
-	struct file *file = NULL;
+	struct fd f = {NULL, 0};
 	pfm_context_t *ctx = NULL;
 	unsigned long flags = 0UL;
 	void *args_k = NULL;
@@ -4879,17 +4877,17 @@ restart_args:
 
 	ret = -EBADF;
 
-	file = fget(fd);
-	if (unlikely(file == NULL)) {
+	f = fdget(fd);
+	if (unlikely(f.file == NULL)) {
 		DPRINT(("invalid fd %d\n", fd));
 		goto error_args;
 	}
-	if (unlikely(PFM_IS_FILE(file) == 0)) {
+	if (unlikely(PFM_IS_FILE(f.file) == 0)) {
 		DPRINT(("fd %d not related to perfmon\n", fd));
 		goto error_args;
 	}
 
-	ctx = file->private_data;
+	ctx = f.file->private_data;
 	if (unlikely(ctx == NULL)) {
 		DPRINT(("no context for fd %d\n", fd));
 		goto error_args;
@@ -4919,8 +4917,8 @@ abort_locked:
 	if (call_made && PFM_CMD_RW_ARG(cmd) && copy_to_user(arg, args_k, base_sz*count)) ret = -EFAULT;
 
 error_args:
-	if (file)
-		fput(file);
+	if (f.file)
+		fdput(f);
 
 	kfree(args_k);
 

arch/parisc/hpux/fs.c

@@ -109,33 +109,32 @@ Efault:
 
 int hpux_getdents(unsigned int fd, struct hpux_dirent __user *dirent, unsigned int count)
 {
-	struct file * file;
+	struct fd arg;
 	struct hpux_dirent __user * lastdirent;
 	struct getdents_callback buf;
-	int error = -EBADF;
+	int error;
 
-	file = fget(fd);
-	if (!file)
-		goto out;
+	arg = fdget(fd);
+	if (!arg.file)
+		return -EBADF;
 
 	buf.current_dir = dirent;
 	buf.previous = NULL;
 	buf.count = count;
 	buf.error = 0;
 
-	error = vfs_readdir(file, filldir, &buf);
+	error = vfs_readdir(arg.file, filldir, &buf);
 	if (error >= 0)
 		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
-		if (put_user(file->f_pos, &lastdirent->d_off))
+		if (put_user(arg.file->f_pos, &lastdirent->d_off))
 			error = -EFAULT;
 		else
 			error = count - buf.count;
 	}
 
-	fput(file);
-out:
+	fdput(arg);
 	return error;
 }
 

arch/powerpc/include/asm/systbl.h

@@ -189,7 +189,7 @@ SYSCALL_SPU(getcwd)
 SYSCALL_SPU(capget)
 SYSCALL_SPU(capset)
 COMPAT_SYS(sigaltstack)
-SYSX_SPU(sys_sendfile64,compat_sys_sendfile,sys_sendfile)
+SYSX_SPU(sys_sendfile,compat_sys_sendfile_wrapper,sys_sendfile)
 SYSCALL(ni_syscall)
 SYSCALL(ni_syscall)
 PPC_SYS(vfork)
@@ -229,7 +229,7 @@ COMPAT_SYS_SPU(sched_setaffinity)
 COMPAT_SYS_SPU(sched_getaffinity)
 SYSCALL(ni_syscall)
 SYSCALL(ni_syscall)
-SYS32ONLY(sendfile64)
+SYSX(sys_ni_syscall,compat_sys_sendfile64_wrapper,sys_sendfile64)
 COMPAT_SYS_SPU(io_setup)
 SYSCALL_SPU(io_destroy)
 COMPAT_SYS_SPU(io_getevents)

arch/powerpc/include/asm/unistd.h

@@ -419,6 +419,7 @@
 #define __ARCH_WANT_COMPAT_SYS_TIME
 #define __ARCH_WANT_COMPAT_SYS_RT_SIGSUSPEND
 #define __ARCH_WANT_SYS_NEWFSTATAT
+#define __ARCH_WANT_COMPAT_SYS_SENDFILE
 #endif
 
 /*

arch/powerpc/kernel/sys_ppc32.c

@@ -143,48 +143,17 @@ long compat_sys_ipc(u32 call, u32 first, u32 second, u32 third, compat_uptr_t pt
  * proper conversion (sign extension) between the register representation of a signed int (msr in 32-bit mode)
  * and the register representation of a signed int (msr in 64-bit mode) is performed.
  */
-asmlinkage long compat_sys_sendfile(u32 out_fd, u32 in_fd, compat_off_t __user * offset, u32 count)
+asmlinkage long compat_sys_sendfile_wrapper(u32 out_fd, u32 in_fd,
+					    compat_off_t __user *offset, u32 count)
 {
-	mm_segment_t old_fs = get_fs();
-	int ret;
-	off_t of;
-	off_t __user *up;
-
-	if (offset && get_user(of, offset))
-		return -EFAULT;
-
-	/* The __user pointer cast is valid because of the set_fs() */		
-	set_fs(KERNEL_DS);
-	up = offset ? (off_t __user *) &of : NULL;
-	ret = sys_sendfile((int)out_fd, (int)in_fd, up, count);
-	set_fs(old_fs);
-	
-	if (offset && put_user(of, offset))
-		return -EFAULT;
-		
-	return ret;
+	return compat_sys_sendfile((int)out_fd, (int)in_fd, offset, count);
 }
 
-asmlinkage int compat_sys_sendfile64(int out_fd, int in_fd, compat_loff_t __user *offset, s32 count)
+asmlinkage long compat_sys_sendfile64_wrapper(u32 out_fd, u32 in_fd,
+					      compat_loff_t __user *offset, u32 count)
 {
-	mm_segment_t old_fs = get_fs();
-	int ret;
-	loff_t lof;
-	loff_t __user *up;
-	
-	if (offset && get_user(lof, offset))
-		return -EFAULT;
-		
-	/* The __user pointer cast is valid because of the set_fs() */		
-	set_fs(KERNEL_DS);
-	up = offset ? (loff_t __user *) &lof : NULL;
-	ret = sys_sendfile64(out_fd, in_fd, up, count);
-	set_fs(old_fs);
-	
-	if (offset && put_user(lof, offset))
-		return -EFAULT;
-		
-	return ret;
+	return sys_sendfile((int)out_fd, (int)in_fd,
+			    (off_t __user *)offset, count);
 }
 
 long compat_sys_execve(unsigned long a0, unsigned long a1, unsigned long a2,

arch/powerpc/platforms/cell/spu_syscalls.c

@@ -69,8 +69,6 @@ SYSCALL_DEFINE4(spu_create, const char __user *, name, unsigned int, flags,
 	umode_t, mode, int, neighbor_fd)
 {
 	long ret;
-	struct file *neighbor;
-	int fput_needed;
 	struct spufs_calls *calls;
 
 	calls = spufs_calls_get();
@@ -78,11 +76,11 @@ SYSCALL_DEFINE4(spu_create, const char __user *, name, unsigned int, flags,
 		return -ENOSYS;
 
 	if (flags & SPU_CREATE_AFFINITY_SPU) {
+		struct fd neighbor = fdget(neighbor_fd);
 		ret = -EBADF;
-		neighbor = fget_light(neighbor_fd, &fput_needed);
-		if (neighbor) {
-			ret = calls->create_thread(name, flags, mode, neighbor);
-			fput_light(neighbor, fput_needed);
+		if (neighbor.file) {
+			ret = calls->create_thread(name, flags, mode, neighbor.file);
+			fdput(neighbor);
 		}
 	} else
 		ret = calls->create_thread(name, flags, mode, NULL);
@@ -94,8 +92,7 @@ SYSCALL_DEFINE4(spu_create, const char __user *, name, unsigned int, flags,
 asmlinkage long sys_spu_run(int fd, __u32 __user *unpc, __u32 __user *ustatus)
 {
 	long ret;
-	struct file *filp;
-	int fput_needed;
+	struct fd arg;
 	struct spufs_calls *calls;
 
 	calls = spufs_calls_get();
@@ -103,10 +100,10 @@ asmlinkage long sys_spu_run(int fd, __u32 __user *unpc, __u32 __user *ustatus)
 		return -ENOSYS;
 
 	ret = -EBADF;
-	filp = fget_light(fd, &fput_needed);
-	if (filp) {
-		ret = calls->spu_run(filp, unpc, ustatus);
-		fput_light(filp, fput_needed);
+	arg = fdget(fd);
+	if (arg.file) {
+		ret = calls->spu_run(arg.file, unpc, ustatus);
+		fdput(arg);
 	}
 
 	spufs_calls_put(calls);

arch/powerpc/platforms/cell/spufs/coredump.c

@@ -106,6 +106,17 @@ static int spufs_ctx_note_size(struct spu_context *ctx, int dfd)
 	return total;
 }
 
+static int match_context(const void *v, struct file *file, unsigned fd)
+{
+	struct spu_context *ctx;
+	if (file->f_op != &spufs_context_fops)
+		return 0;
+	ctx = SPUFS_I(file->f_dentry->d_inode)->i_ctx;
+	if (ctx->flags & SPU_CREATE_NOSCHED)
+		return 0;
+	return fd + 1;
+}
+
 /*
  * The additional architecture-specific notes for Cell are various
  * context files in the spu context.
@@ -115,29 +126,18 @@ static int spufs_ctx_note_size(struct spu_context *ctx, int dfd)
  * internal functionality to dump them without needing to actually
  * open the files.
  */
+/*
+ * descriptor table is not shared, so files can't change or go away.
+ */
 static struct spu_context *coredump_next_context(int *fd)
 {
-	struct fdtable *fdt = files_fdtable(current->files);
 	struct file *file;
-	struct spu_context *ctx = NULL;
-
-	for (; *fd < fdt->max_fds; (*fd)++) {
-		if (!fd_is_open(*fd, fdt))
-			continue;
-
-		file = fcheck(*fd);
-
-		if (!file || file->f_op != &spufs_context_fops)
-			continue;
-
-		ctx = SPUFS_I(file->f_dentry->d_inode)->i_ctx;
-		if (ctx->flags & SPU_CREATE_NOSCHED)
-			continue;
-
-		break;
-	}
-
-	return ctx;
+	int n = iterate_fd(current->files, *fd, match_context, NULL);
+	if (!n)
+		return NULL;
+	*fd = n - 1;
+	file = fcheck(*fd);
+	return SPUFS_I(file->f_dentry->d_inode)->i_ctx;
 }
 
 int spufs_coredump_extra_notes_size(void)

arch/s390/hypfs/inode.c

@@ -72,8 +72,6 @@ static void hypfs_remove(struct dentry *dentry)
 	struct dentry *parent;
 
 	parent = dentry->d_parent;
-	if (!parent || !parent->d_inode)
-		return;
 	mutex_lock(&parent->d_inode->i_mutex);
 	if (hypfs_positive(dentry)) {
 		if (S_ISDIR(dentry->d_inode->i_mode))

arch/sparc/include/asm/unistd.h

@@ -447,6 +447,7 @@
 #else
 #define __ARCH_WANT_COMPAT_SYS_TIME
 #define __ARCH_WANT_COMPAT_SYS_RT_SIGSUSPEND
+#define __ARCH_WANT_COMPAT_SYS_SENDFILE
 #endif
 
 /*

arch/sparc/kernel/sys32.S

@@ -90,7 +90,7 @@ SIGN1(sys32_mkdir, sys_mkdir, %o1)
 SIGN3(sys32_futex, compat_sys_futex, %o1, %o2, %o5)
 SIGN1(sys32_sysfs, compat_sys_sysfs, %o0)
 SIGN2(sys32_sendfile, compat_sys_sendfile, %o0, %o1)
-SIGN2(sys32_sendfile64, compat_sys_sendfile64, %o0, %o1)
+SIGN2(sys32_sendfile64, sys_sendfile, %o0, %o1)
 SIGN1(sys32_prctl, sys_prctl, %o0)
 SIGN1(sys32_sched_rr_get_interval, compat_sys_sched_rr_get_interval, %o0)
 SIGN2(sys32_waitpid, sys_waitpid, %o0, %o2)

arch/sparc/kernel/sys_sparc32.c

@@ -506,52 +506,6 @@ long compat_sys_fadvise64_64(int fd,
 				advice);
 }
 
-asmlinkage long compat_sys_sendfile(int out_fd, int in_fd,
-				    compat_off_t __user *offset,
-				    compat_size_t count)
-{
-	mm_segment_t old_fs = get_fs();
-	int ret;
-	off_t of;
-	
-	if (offset && get_user(of, offset))
-		return -EFAULT;
-		
-	set_fs(KERNEL_DS);
-	ret = sys_sendfile(out_fd, in_fd,
-			   offset ? (off_t __user *) &of : NULL,
-			   count);
-	set_fs(old_fs);
-	
-	if (offset && put_user(of, offset))
-		return -EFAULT;
-		
-	return ret;
-}
-
-asmlinkage long compat_sys_sendfile64(int out_fd, int in_fd,
-				      compat_loff_t __user *offset,
-				      compat_size_t count)
-{
-	mm_segment_t old_fs = get_fs();
-	int ret;
-	loff_t lof;
-	
-	if (offset && get_user(lof, offset))
-		return -EFAULT;
-		
-	set_fs(KERNEL_DS);
-	ret = sys_sendfile64(out_fd, in_fd,
-			     offset ? (loff_t __user *) &lof : NULL,
-			     count);
-	set_fs(old_fs);
-	
-	if (offset && put_user(lof, offset))
-		return -EFAULT;
-		
-	return ret;
-}
-
 /* This is just a version for 32-bit applications which does
  * not force O_LARGEFILE on.
  */

arch/um/drivers/mconsole_kern.c

@@ -21,6 +21,9 @@
 #include <linux/un.h>
 #include <linux/workqueue.h>
 #include <linux/mutex.h>
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/file.h>
 #include <asm/uaccess.h>
 #include <asm/switch_to.h>
 
@@ -118,18 +121,14 @@ void mconsole_log(struct mc_request *req)
 	mconsole_reply(req, "", 0, 0);
 }
 
-/* This is a more convoluted version of mconsole_proc, which has some stability
- * problems; however, we need it fixed, because it is expected that UML users
- * mount HPPFS instead of procfs on /proc. And we want mconsole_proc to still
- * show the real procfs content, not the ones from hppfs.*/
-#if 0
 void mconsole_proc(struct mc_request *req)
 {
 	struct vfsmount *mnt = current->nsproxy->pid_ns->proc_mnt;
+	char *buf;
+	int len;
 	struct file *file;
-	int n;
-	char *ptr = req->request.data, *buf;
-	mm_segment_t old_fs = get_fs();
+	int first_chunk = 1;
+	char *ptr = req->request.data;
 
 	ptr += strlen("proc");
 	ptr = skip_spaces(ptr);
@@ -137,6 +136,7 @@ void mconsole_proc(struct mc_request *req)
 	file = file_open_root(mnt->mnt_root, mnt, ptr, O_RDONLY);
 	if (IS_ERR(file)) {
 		mconsole_reply(req, "Failed to open file", 1, 0);
+		printk(KERN_ERR "open /proc/%s: %ld\n", ptr, PTR_ERR(file));
 		goto out;
 	}
 
@@ -146,62 +146,13 @@ void mconsole_proc(struct mc_request *req)
 		goto out_fput;
 	}
 
-	if (file->f_op->read) {
-		do {
-			loff_t pos;
-			set_fs(KERNEL_DS);
-			n = vfs_read(file, buf, PAGE_SIZE - 1, &pos);
-			file_pos_write(file, pos);
-			set_fs(old_fs);
-			if (n >= 0) {
-				buf[n] = '\0';
-				mconsole_reply(req, buf, 0, (n > 0));
-			}
-			else {
-				mconsole_reply(req, "Read of file failed",
-					       1, 0);
-				goto out_free;
-			}
-		} while (n > 0);
-	}
-	else mconsole_reply(req, "", 0, 0);
-
- out_free:
-	kfree(buf);
- out_fput:
-	fput(file);
- out: ;
-}
-#endif
-
-void mconsole_proc(struct mc_request *req)
-{
-	char path[64];
-	char *buf;
-	int len;
-	int fd;
-	int first_chunk = 1;
-	char *ptr = req->request.data;
-
-	ptr += strlen("proc");
-	ptr = skip_spaces(ptr);
-	snprintf(path, sizeof(path), "/proc/%s", ptr);
-
-	fd = sys_open(path, 0, 0);
-	if (fd < 0) {
-		mconsole_reply(req, "Failed to open file", 1, 0);
-		printk(KERN_ERR "open %s: %d\n",path,fd);
-		goto out;
-	}
-
-	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
-	if (buf == NULL) {
-		mconsole_reply(req, "Failed to allocate buffer", 1, 0);
-		goto out_close;
-	}
-
-	for (;;) {
-		len = sys_read(fd, buf, PAGE_SIZE-1);
+	do {
+		loff_t pos;
+		mm_segment_t old_fs = get_fs();
+		set_fs(KERNEL_DS);
+		len = vfs_read(file, buf, PAGE_SIZE - 1, &pos);
+		set_fs(old_fs);
+		file->f_pos = pos;
 		if (len < 0) {
 			mconsole_reply(req, "Read of file failed", 1, 0);
 			goto out_free;
@@ -211,22 +162,14 @@ void mconsole_proc(struct mc_request *req)
 			mconsole_reply(req, "\n", 0, 1);
 			first_chunk = 0;
 		}
-		if (len == PAGE_SIZE-1) {
-			buf[len] = '\0';
-			mconsole_reply(req, buf, 0, 1);
-		} else {
-			buf[len] = '\0';
-			mconsole_reply(req, buf, 0, 0);
-			break;
-		}
-	}
-
+		buf[len] = '\0';
+		mconsole_reply(req, buf, 0, (len > 0));
+	} while (len > 0);
  out_free:
 	kfree(buf);
- out_close:
-	sys_close(fd);
- out:
-	/* nothing */;
+ out_fput:
+	fput(file);
+ out: ;
 }
 
 #define UML_MCONSOLE_HELPTEXT \

drivers/base/dma-buf.c

@@ -460,8 +460,7 @@ int dma_buf_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma,
 	if (vma->vm_file)
 		fput(vma->vm_file);
 
-	vma->vm_file = dmabuf->file;
-	get_file(vma->vm_file);
+	vma->vm_file = get_file(dmabuf->file);
 
 	vma->vm_pgoff = pgoff;
 

drivers/infiniband/core/ucma.c

@@ -1183,7 +1183,7 @@ static ssize_t ucma_migrate_id(struct ucma_file *new_file,
 	struct rdma_ucm_migrate_id cmd;
 	struct rdma_ucm_migrate_resp resp;
 	struct ucma_context *ctx;
-	struct file *filp;
+	struct fd f;
 	struct ucma_file *cur_file;
 	int ret = 0;
 
@@ -1191,12 +1191,12 @@ static ssize_t ucma_migrate_id(struct ucma_file *new_file,
 		return -EFAULT;
 
 	/* Get current fd to protect against it being closed */
-	filp = fget(cmd.fd);
-	if (!filp)
+	f = fdget(cmd.fd);
+	if (!f.file)
 		return -ENOENT;
 
 	/* Validate current fd and prevent destruction of id. */
-	ctx = ucma_get_ctx(filp->private_data, cmd.id);
+	ctx = ucma_get_ctx(f.file->private_data, cmd.id);
 	if (IS_ERR(ctx)) {
 		ret = PTR_ERR(ctx);
 		goto file_put;
@@ -1230,7 +1230,7 @@ response:
 
 	ucma_put_ctx(ctx);
 file_put:
-	fput(filp);
+	fdput(f);
 	return ret;
 }
 

drivers/infiniband/core/uverbs_cmd.c

@@ -705,7 +705,7 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file,
 	struct ib_udata			udata;
 	struct ib_uxrcd_object         *obj;
 	struct ib_xrcd                 *xrcd = NULL;
-	struct file                    *f = NULL;
+	struct fd			f = {NULL, 0};
 	struct inode                   *inode = NULL;
 	int				ret = 0;
 	int				new_xrcd = 0;
@@ -724,18 +724,13 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file,
 
 	if (cmd.fd != -1) {
 		/* search for file descriptor */
-		f = fget(cmd.fd);
-		if (!f) {
-			ret = -EBADF;
-			goto err_tree_mutex_unlock;
-		}
-
-		inode = f->f_dentry->d_inode;
-		if (!inode) {
+		f = fdget(cmd.fd);
+		if (!f.file) {
 			ret = -EBADF;
 			goto err_tree_mutex_unlock;
 		}
 
+		inode = f.file->f_path.dentry->d_inode;
 		xrcd = find_xrcd(file->device, inode);
 		if (!xrcd && !(cmd.oflags & O_CREAT)) {
 			/* no file descriptor. Need CREATE flag */
@@ -800,8 +795,8 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file,
 		goto err_copy;
 	}
 
-	if (f)
-		fput(f);
+	if (f.file)
+		fdput(f);
 
 	mutex_lock(&file->mutex);
 	list_add_tail(&obj->uobject.list, &file->ucontext->xrcd_list);
@@ -830,8 +825,8 @@ err:
 	put_uobj_write(&obj->uobject);
 
 err_tree_mutex_unlock:
-	if (f)
-		fput(f);
+	if (f.file)
+		fdput(f);
 
 	mutex_unlock(&file->device->xrcd_tree_mutex);
 

drivers/infiniband/core/uverbs_main.c

@@ -541,16 +541,15 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file,
 struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd)
 {
 	struct ib_uverbs_event_file *ev_file = NULL;
-	struct file *filp;
+	struct fd f = fdget(fd);
 
-	filp = fget(fd);
-	if (!filp)
+	if (!f.file)
 		return NULL;
 
-	if (filp->f_op != &uverbs_event_fops)
+	if (f.file->f_op != &uverbs_event_fops)
 		goto out;
 
-	ev_file = filp->private_data;
+	ev_file = f.file->private_data;
 	if (ev_file->is_async) {
 		ev_file = NULL;
 		goto out;
@@ -559,7 +558,7 @@ struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd)
 	kref_get(&ev_file->ref);
 
 out:
-	fput(filp);
+	fdput(f);
 	return ev_file;
 }
 

drivers/staging/android/binder.c

@@ -362,71 +362,22 @@ struct binder_transaction {
 static void
 binder_defer_work(struct binder_proc *proc, enum binder_deferred_state defer);
 
-/*
- * copied from get_unused_fd_flags
- */
 static int task_get_unused_fd_flags(struct binder_proc *proc, int flags)
 {
 	struct files_struct *files = proc->files;
-	int fd, error;
-	struct fdtable *fdt;
 	unsigned long rlim_cur;
 	unsigned long irqs;
 
 	if (files == NULL)
 		return -ESRCH;
 
-	error = -EMFILE;
-	spin_lock(&files->file_lock);
+	if (!lock_task_sighand(proc->tsk, &irqs))
+		return -EMFILE;
 
-repeat:
-	fdt = files_fdtable(files);
-	fd = find_next_zero_bit(fdt->open_fds, fdt->max_fds, files->next_fd);
+	rlim_cur = task_rlimit(proc->tsk, RLIMIT_NOFILE);
+	unlock_task_sighand(proc->tsk, &irqs);
 
-	/*
-	 * N.B. For clone tasks sharing a files structure, this test
-	 * will limit the total number of files that can be opened.
-	 */
-	rlim_cur = 0;
-	if (lock_task_sighand(proc->tsk, &irqs)) {
-		rlim_cur = proc->tsk->signal->rlim[RLIMIT_NOFILE].rlim_cur;
-		unlock_task_sighand(proc->tsk, &irqs);
-	}
-	if (fd >= rlim_cur)
-		goto out;
-
-	/* Do we need to expand the fd array or fd set?  */
-	error = expand_files(files, fd);
-	if (error < 0)
-		goto out;
-
-	if (error) {
-		/*
-		 * If we needed to expand the fs array we
-		 * might have blocked - try again.
-		 */
-		error = -EMFILE;
-		goto repeat;
-	}
-
-	__set_open_fd(fd, fdt);
-	if (flags & O_CLOEXEC)
-		__set_close_on_exec(fd, fdt);
-	else
-		__clear_close_on_exec(fd, fdt);
-	files->next_fd = fd + 1;
-
-	/* Sanity check */
-	if (fdt->fd[fd] != NULL) {
-		pr_warn("get_unused_fd: slot %d not NULL!\n", fd);
-		fdt->fd[fd] = NULL;
-	}
-
-	error = fd;
-
-out:
-	spin_unlock(&files->file_lock);
-	return error;
+	return __alloc_fd(files, 0, rlim_cur, flags);
 }
 
 /*
@@ -435,28 +386,8 @@ out:
 static void task_fd_install(
 	struct binder_proc *proc, unsigned int fd, struct file *file)
 {
-	struct files_struct *files = proc->files;
-	struct fdtable *fdt;
-
-	if (files == NULL)
-		return;
-
-	spin_lock(&files->file_lock);
-	fdt = files_fdtable(files);
-	BUG_ON(fdt->fd[fd] != NULL);
-	rcu_assign_pointer(fdt->fd[fd], file);
-	spin_unlock(&files->file_lock);
-}
-
-/*
- * copied from __put_unused_fd in open.c
- */
-static void __put_unused_fd(struct files_struct *files, unsigned int fd)
-{
-	struct fdtable *fdt = files_fdtable(files);
-	__clear_open_fd(fd, fdt);
-	if (fd < files->next_fd)
-		files->next_fd = fd;
+	if (proc->files)
+		__fd_install(proc->files, fd, file);
 }
 
 /*
@@ -464,27 +395,12 @@ static void __put_unused_fd(struct files_struct *files, unsigned int fd)
  */
 static long task_close_fd(struct binder_proc *proc, unsigned int fd)
 {
-	struct file *filp;
-	struct files_struct *files = proc->files;
-	struct fdtable *fdt;
 	int retval;
 
-	if (files == NULL)
+	if (proc->files == NULL)
 		return -ESRCH;
 
-	spin_lock(&files->file_lock);
-	fdt = files_fdtable(files);
-	if (fd >= fdt->max_fds)
-		goto out_unlock;
-	filp = fdt->fd[fd];
-	if (!filp)
-		goto out_unlock;
-	rcu_assign_pointer(fdt->fd[fd], NULL);
-	__clear_close_on_exec(fd, fdt);
-	__put_unused_fd(files, fd);
-	spin_unlock(&files->file_lock);
-	retval = filp_close(filp, files);
-
+	retval = __close_fd(proc->files, fd);
 	/* can't restart close syscall because file table entry was cleared */
 	if (unlikely(retval == -ERESTARTSYS ||
 		     retval == -ERESTARTNOINTR ||
@@ -493,10 +409,6 @@ static long task_close_fd(struct binder_proc *proc, unsigned int fd)
 		retval = -EINTR;
 
 	return retval;
-
-out_unlock:
-	spin_unlock(&files->file_lock);
-	return -EBADF;
 }
 
 static void binder_set_nice(long nice)
@@ -2793,6 +2705,9 @@ static int binder_mmap(struct file *filp, struct vm_area_struct *vma)
 	const char *failure_string;
 	struct binder_buffer *buffer;
 
+	if (proc->tsk != current)
+		return -EINVAL;
+
 	if ((vma->vm_end - vma->vm_start) > SZ_4M)
 		vma->vm_end = vma->vm_start + SZ_4M;
 
@@ -2857,7 +2772,7 @@ static int binder_mmap(struct file *filp, struct vm_area_struct *vma)
 	binder_insert_free_buffer(proc, buffer);
 	proc->free_async_space = proc->buffer_size / 2;
 	barrier();
-	proc->files = get_files_struct(proc->tsk);
+	proc->files = get_files_struct(current);
 	proc->vma = vma;
 	proc->vma_vm_mm = vma->vm_mm;
 

drivers/staging/omapdrm/omap_gem.c

@@ -592,9 +592,8 @@ int omap_gem_mmap_obj(struct drm_gem_object *obj,
 		 * in particular in the case of mmap'd dmabufs)
 		 */
 		fput(vma->vm_file);
-		get_file(obj->filp);
 		vma->vm_pgoff = 0;
-		vma->vm_file  = obj->filp;
+		vma->vm_file  = get_file(obj->filp);
 
 		vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
 	}

drivers/tty/tty_io.c

@@ -1166,10 +1166,8 @@ ssize_t redirected_tty_write(struct file *file, const char __user *buf,
 	struct file *p = NULL;
 
 	spin_lock(&redirect_lock);
-	if (redirect) {
-		get_file(redirect);
-		p = redirect;
-	}
+	if (redirect)
+		p = get_file(redirect);
 	spin_unlock(&redirect_lock);
 
 	if (p) {
@@ -2264,8 +2262,7 @@ static int tioccons(struct file *file)
 		spin_unlock(&redirect_lock);
 		return -EBUSY;
 	}
-	get_file(file);
-	redirect = file;
+	redirect = get_file(file);
 	spin_unlock(&redirect_lock);
 	return 0;
 }
@@ -2809,6 +2806,13 @@ static long tty_compat_ioctl(struct file *file, unsigned int cmd,
 }
 #endif
 
+static int this_tty(const void *t, struct file *file, unsigned fd)
+{
+	if (likely(file->f_op->read != tty_read))
+		return 0;
+	return file_tty(file) != t ? 0 : fd + 1;
+}
+	
 /*
  * This implements the "Secure Attention Key" ---  the idea is to
  * prevent trojan horses by killing all processes associated with this
@@ -2836,8 +2840,6 @@ void __do_SAK(struct tty_struct *tty)
 	struct task_struct *g, *p;
 	struct pid *session;
 	int		i;
-	struct file	*filp;
-	struct fdtable *fdt;
 
 	if (!tty)
 		return;
@@ -2867,27 +2869,12 @@ void __do_SAK(struct tty_struct *tty)
 			continue;
 		}
 		task_lock(p);
-		if (p->files) {
-			/*
-			 * We don't take a ref to the file, so we must
-			 * hold ->file_lock instead.
-			 */
-			spin_lock(&p->files->file_lock);
-			fdt = files_fdtable(p->files);
-			for (i = 0; i < fdt->max_fds; i++) {
-				filp = fcheck_files(p->files, i);
-				if (!filp)
-					continue;
-				if (filp->f_op->read == tty_read &&
-				    file_tty(filp) == tty) {
-					printk(KERN_NOTICE "SAK: killed process %d"
-					    " (%s): fd#%d opened to the tty\n",
-					    task_pid_nr(p), p->comm, i);
-					force_sig(SIGKILL, p);
-					break;
-				}
-			}
-			spin_unlock(&p->files->file_lock);
+		i = iterate_fd(p->files, 0, this_tty, tty);
+		if (i != 0) {
+			printk(KERN_NOTICE "SAK: killed process %d"
+			    " (%s): fd#%d opened to the tty\n",
+				    task_pid_nr(p), p->comm, i - 1);
+			force_sig(SIGKILL, p);
 		}
 		task_unlock(p);
 	} while_each_thread(g, p);