Add the rt linux 4.1.3-rt3 as base
[kvmfornfv.git] / kernel / fs / compat.c
diff --git a/kernel/fs/compat.c b/kernel/fs/compat.c
new file mode 100644 (file)
index 0000000..6fd272d
--- /dev/null
@@ -0,0 +1,1481 @@
+/*
+ *  linux/fs/compat.c
+ *
+ *  Kernel compatibililty routines for e.g. 32 bit syscall support
+ *  on 64 bit kernels.
+ *
+ *  Copyright (C) 2002       Stephen Rothwell, IBM Corporation
+ *  Copyright (C) 1997-2000  Jakub Jelinek  (jakub@redhat.com)
+ *  Copyright (C) 1998       Eddie C. Dost  (ecd@skynet.be)
+ *  Copyright (C) 2001,2002  Andi Kleen, SuSE Labs 
+ *  Copyright (C) 2003       Pavel Machek (pavel@ucw.cz)
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License version 2 as
+ *  published by the Free Software Foundation.
+ */
+
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/linkage.h>
+#include <linux/compat.h>
+#include <linux/errno.h>
+#include <linux/time.h>
+#include <linux/fs.h>
+#include <linux/fcntl.h>
+#include <linux/namei.h>
+#include <linux/file.h>
+#include <linux/fdtable.h>
+#include <linux/vfs.h>
+#include <linux/ioctl.h>
+#include <linux/init.h>
+#include <linux/ncp_mount.h>
+#include <linux/nfs4_mount.h>
+#include <linux/syscalls.h>
+#include <linux/ctype.h>
+#include <linux/dirent.h>
+#include <linux/fsnotify.h>
+#include <linux/highuid.h>
+#include <linux/personality.h>
+#include <linux/rwsem.h>
+#include <linux/tsacct_kern.h>
+#include <linux/security.h>
+#include <linux/highmem.h>
+#include <linux/signal.h>
+#include <linux/poll.h>
+#include <linux/mm.h>
+#include <linux/fs_struct.h>
+#include <linux/slab.h>
+#include <linux/pagemap.h>
+#include <linux/aio.h>
+
+#include <asm/uaccess.h>
+#include <asm/mmu_context.h>
+#include <asm/ioctls.h>
+#include "internal.h"
+
+int compat_log = 1;
+
+int compat_printk(const char *fmt, ...)
+{
+       va_list ap;
+       int ret;
+       if (!compat_log)
+               return 0;
+       va_start(ap, fmt);
+       ret = vprintk(fmt, ap);
+       va_end(ap);
+       return ret;
+}
+
+/*
+ * Not all architectures have sys_utime, so implement this in terms
+ * of sys_utimes.
+ */
+COMPAT_SYSCALL_DEFINE2(utime, const char __user *, filename,
+                      struct compat_utimbuf __user *, t)
+{
+       struct timespec tv[2];
+
+       if (t) {
+               if (get_user(tv[0].tv_sec, &t->actime) ||
+                   get_user(tv[1].tv_sec, &t->modtime))
+                       return -EFAULT;
+               tv[0].tv_nsec = 0;
+               tv[1].tv_nsec = 0;
+       }
+       return do_utimes(AT_FDCWD, filename, t ? tv : NULL, 0);
+}
+
+COMPAT_SYSCALL_DEFINE4(utimensat, unsigned int, dfd, const char __user *, filename, struct compat_timespec __user *, t, int, flags)
+{
+       struct timespec tv[2];
+
+       if  (t) {
+               if (compat_get_timespec(&tv[0], &t[0]) ||
+                   compat_get_timespec(&tv[1], &t[1]))
+                       return -EFAULT;
+
+               if (tv[0].tv_nsec == UTIME_OMIT && tv[1].tv_nsec == UTIME_OMIT)
+                       return 0;
+       }
+       return do_utimes(dfd, filename, t ? tv : NULL, flags);
+}
+
+COMPAT_SYSCALL_DEFINE3(futimesat, unsigned int, dfd, const char __user *, filename, struct compat_timeval __user *, t)
+{
+       struct timespec tv[2];
+
+       if (t) {
+               if (get_user(tv[0].tv_sec, &t[0].tv_sec) ||
+                   get_user(tv[0].tv_nsec, &t[0].tv_usec) ||
+                   get_user(tv[1].tv_sec, &t[1].tv_sec) ||
+                   get_user(tv[1].tv_nsec, &t[1].tv_usec))
+                       return -EFAULT;
+               if (tv[0].tv_nsec >= 1000000 || tv[0].tv_nsec < 0 ||
+                   tv[1].tv_nsec >= 1000000 || tv[1].tv_nsec < 0)
+                       return -EINVAL;
+               tv[0].tv_nsec *= 1000;
+               tv[1].tv_nsec *= 1000;
+       }
+       return do_utimes(dfd, filename, t ? tv : NULL, 0);
+}
+
+COMPAT_SYSCALL_DEFINE2(utimes, const char __user *, filename, struct compat_timeval __user *, t)
+{
+       return compat_sys_futimesat(AT_FDCWD, filename, t);
+}
+
+static int cp_compat_stat(struct kstat *stat, struct compat_stat __user *ubuf)
+{
+       struct compat_stat tmp;
+
+       if (!old_valid_dev(stat->dev) || !old_valid_dev(stat->rdev))
+               return -EOVERFLOW;
+
+       memset(&tmp, 0, sizeof(tmp));
+       tmp.st_dev = old_encode_dev(stat->dev);
+       tmp.st_ino = stat->ino;
+       if (sizeof(tmp.st_ino) < sizeof(stat->ino) && tmp.st_ino != stat->ino)
+               return -EOVERFLOW;
+       tmp.st_mode = stat->mode;
+       tmp.st_nlink = stat->nlink;
+       if (tmp.st_nlink != stat->nlink)
+               return -EOVERFLOW;
+       SET_UID(tmp.st_uid, from_kuid_munged(current_user_ns(), stat->uid));
+       SET_GID(tmp.st_gid, from_kgid_munged(current_user_ns(), stat->gid));
+       tmp.st_rdev = old_encode_dev(stat->rdev);
+       if ((u64) stat->size > MAX_NON_LFS)
+               return -EOVERFLOW;
+       tmp.st_size = stat->size;
+       tmp.st_atime = stat->atime.tv_sec;
+       tmp.st_atime_nsec = stat->atime.tv_nsec;
+       tmp.st_mtime = stat->mtime.tv_sec;
+       tmp.st_mtime_nsec = stat->mtime.tv_nsec;
+       tmp.st_ctime = stat->ctime.tv_sec;
+       tmp.st_ctime_nsec = stat->ctime.tv_nsec;
+       tmp.st_blocks = stat->blocks;
+       tmp.st_blksize = stat->blksize;
+       return copy_to_user(ubuf, &tmp, sizeof(tmp)) ? -EFAULT : 0;
+}
+
+COMPAT_SYSCALL_DEFINE2(newstat, const char __user *, filename,
+                      struct compat_stat __user *, statbuf)
+{
+       struct kstat stat;
+       int error;
+
+       error = vfs_stat(filename, &stat);
+       if (error)
+               return error;
+       return cp_compat_stat(&stat, statbuf);
+}
+
+COMPAT_SYSCALL_DEFINE2(newlstat, const char __user *, filename,
+                      struct compat_stat __user *, statbuf)
+{
+       struct kstat stat;
+       int error;
+
+       error = vfs_lstat(filename, &stat);
+       if (error)
+               return error;
+       return cp_compat_stat(&stat, statbuf);
+}
+
+#ifndef __ARCH_WANT_STAT64
+COMPAT_SYSCALL_DEFINE4(newfstatat, unsigned int, dfd,
+                      const char __user *, filename,
+                      struct compat_stat __user *, statbuf, int, flag)
+{
+       struct kstat stat;
+       int error;
+
+       error = vfs_fstatat(dfd, filename, &stat, flag);
+       if (error)
+               return error;
+       return cp_compat_stat(&stat, statbuf);
+}
+#endif
+
+COMPAT_SYSCALL_DEFINE2(newfstat, unsigned int, fd,
+                      struct compat_stat __user *, statbuf)
+{
+       struct kstat stat;
+       int error = vfs_fstat(fd, &stat);
+
+       if (!error)
+               error = cp_compat_stat(&stat, statbuf);
+       return error;
+}
+
+static int put_compat_statfs(struct compat_statfs __user *ubuf, struct kstatfs *kbuf)
+{
+       
+       if (sizeof ubuf->f_blocks == 4) {
+               if ((kbuf->f_blocks | kbuf->f_bfree | kbuf->f_bavail |
+                    kbuf->f_bsize | kbuf->f_frsize) & 0xffffffff00000000ULL)
+                       return -EOVERFLOW;
+               /* f_files and f_ffree may be -1; it's okay
+                * to stuff that into 32 bits */
+               if (kbuf->f_files != 0xffffffffffffffffULL
+                && (kbuf->f_files & 0xffffffff00000000ULL))
+                       return -EOVERFLOW;
+               if (kbuf->f_ffree != 0xffffffffffffffffULL
+                && (kbuf->f_ffree & 0xffffffff00000000ULL))
+                       return -EOVERFLOW;
+       }
+       if (!access_ok(VERIFY_WRITE, ubuf, sizeof(*ubuf)) ||
+           __put_user(kbuf->f_type, &ubuf->f_type) ||
+           __put_user(kbuf->f_bsize, &ubuf->f_bsize) ||
+           __put_user(kbuf->f_blocks, &ubuf->f_blocks) ||
+           __put_user(kbuf->f_bfree, &ubuf->f_bfree) ||
+           __put_user(kbuf->f_bavail, &ubuf->f_bavail) ||
+           __put_user(kbuf->f_files, &ubuf->f_files) ||
+           __put_user(kbuf->f_ffree, &ubuf->f_ffree) ||
+           __put_user(kbuf->f_namelen, &ubuf->f_namelen) ||
+           __put_user(kbuf->f_fsid.val[0], &ubuf->f_fsid.val[0]) ||
+           __put_user(kbuf->f_fsid.val[1], &ubuf->f_fsid.val[1]) ||
+           __put_user(kbuf->f_frsize, &ubuf->f_frsize) ||
+           __put_user(kbuf->f_flags, &ubuf->f_flags) ||
+           __clear_user(ubuf->f_spare, sizeof(ubuf->f_spare)))
+               return -EFAULT;
+       return 0;
+}
+
+/*
+ * The following statfs calls are copies of code from fs/statfs.c and
+ * should be checked against those from time to time
+ */
+COMPAT_SYSCALL_DEFINE2(statfs, const char __user *, pathname, struct compat_statfs __user *, buf)
+{
+       struct kstatfs tmp;
+       int error = user_statfs(pathname, &tmp);
+       if (!error)
+               error = put_compat_statfs(buf, &tmp);
+       return error;
+}
+
+COMPAT_SYSCALL_DEFINE2(fstatfs, unsigned int, fd, struct compat_statfs __user *, buf)
+{
+       struct kstatfs tmp;
+       int error = fd_statfs(fd, &tmp);
+       if (!error)
+               error = put_compat_statfs(buf, &tmp);
+       return error;
+}
+
+static int put_compat_statfs64(struct compat_statfs64 __user *ubuf, struct kstatfs *kbuf)
+{
+       if (sizeof ubuf->f_blocks == 4) {
+               if ((kbuf->f_blocks | kbuf->f_bfree | kbuf->f_bavail |
+                    kbuf->f_bsize | kbuf->f_frsize) & 0xffffffff00000000ULL)
+                       return -EOVERFLOW;
+               /* f_files and f_ffree may be -1; it's okay
+                * to stuff that into 32 bits */
+               if (kbuf->f_files != 0xffffffffffffffffULL
+                && (kbuf->f_files & 0xffffffff00000000ULL))
+                       return -EOVERFLOW;
+               if (kbuf->f_ffree != 0xffffffffffffffffULL
+                && (kbuf->f_ffree & 0xffffffff00000000ULL))
+                       return -EOVERFLOW;
+       }
+       if (!access_ok(VERIFY_WRITE, ubuf, sizeof(*ubuf)) ||
+           __put_user(kbuf->f_type, &ubuf->f_type) ||
+           __put_user(kbuf->f_bsize, &ubuf->f_bsize) ||
+           __put_user(kbuf->f_blocks, &ubuf->f_blocks) ||
+           __put_user(kbuf->f_bfree, &ubuf->f_bfree) ||
+           __put_user(kbuf->f_bavail, &ubuf->f_bavail) ||
+           __put_user(kbuf->f_files, &ubuf->f_files) ||
+           __put_user(kbuf->f_ffree, &ubuf->f_ffree) ||
+           __put_user(kbuf->f_namelen, &ubuf->f_namelen) ||
+           __put_user(kbuf->f_fsid.val[0], &ubuf->f_fsid.val[0]) ||
+           __put_user(kbuf->f_fsid.val[1], &ubuf->f_fsid.val[1]) ||
+           __put_user(kbuf->f_frsize, &ubuf->f_frsize) ||
+           __put_user(kbuf->f_flags, &ubuf->f_flags) ||
+           __clear_user(ubuf->f_spare, sizeof(ubuf->f_spare)))
+               return -EFAULT;
+       return 0;
+}
+
+COMPAT_SYSCALL_DEFINE3(statfs64, const char __user *, pathname, compat_size_t, sz, struct compat_statfs64 __user *, buf)
+{
+       struct kstatfs tmp;
+       int error;
+
+       if (sz != sizeof(*buf))
+               return -EINVAL;
+
+       error = user_statfs(pathname, &tmp);
+       if (!error)
+               error = put_compat_statfs64(buf, &tmp);
+       return error;
+}
+
+COMPAT_SYSCALL_DEFINE3(fstatfs64, unsigned int, fd, compat_size_t, sz, struct compat_statfs64 __user *, buf)
+{
+       struct kstatfs tmp;
+       int error;
+
+       if (sz != sizeof(*buf))
+               return -EINVAL;
+
+       error = fd_statfs(fd, &tmp);
+       if (!error)
+               error = put_compat_statfs64(buf, &tmp);
+       return error;
+}
+
+/*
+ * This is a copy of sys_ustat, just dealing with a structure layout.
+ * Given how simple this syscall is that apporach is more maintainable
+ * than the various conversion hacks.
+ */
+COMPAT_SYSCALL_DEFINE2(ustat, unsigned, dev, struct compat_ustat __user *, u)
+{
+       struct compat_ustat tmp;
+       struct kstatfs sbuf;
+       int err = vfs_ustat(new_decode_dev(dev), &sbuf);
+       if (err)
+               return err;
+
+       memset(&tmp, 0, sizeof(struct compat_ustat));
+       tmp.f_tfree = sbuf.f_bfree;
+       tmp.f_tinode = sbuf.f_ffree;
+       if (copy_to_user(u, &tmp, sizeof(struct compat_ustat)))
+               return -EFAULT;
+       return 0;
+}
+
+static int get_compat_flock(struct flock *kfl, struct compat_flock __user *ufl)
+{
+       if (!access_ok(VERIFY_READ, ufl, sizeof(*ufl)) ||
+           __get_user(kfl->l_type, &ufl->l_type) ||
+           __get_user(kfl->l_whence, &ufl->l_whence) ||
+           __get_user(kfl->l_start, &ufl->l_start) ||
+           __get_user(kfl->l_len, &ufl->l_len) ||
+           __get_user(kfl->l_pid, &ufl->l_pid))
+               return -EFAULT;
+       return 0;
+}
+
+static int put_compat_flock(struct flock *kfl, struct compat_flock __user *ufl)
+{
+       if (!access_ok(VERIFY_WRITE, ufl, sizeof(*ufl)) ||
+           __put_user(kfl->l_type, &ufl->l_type) ||
+           __put_user(kfl->l_whence, &ufl->l_whence) ||
+           __put_user(kfl->l_start, &ufl->l_start) ||
+           __put_user(kfl->l_len, &ufl->l_len) ||
+           __put_user(kfl->l_pid, &ufl->l_pid))
+               return -EFAULT;
+       return 0;
+}
+
+#ifndef HAVE_ARCH_GET_COMPAT_FLOCK64
+static int get_compat_flock64(struct flock *kfl, struct compat_flock64 __user *ufl)
+{
+       if (!access_ok(VERIFY_READ, ufl, sizeof(*ufl)) ||
+           __get_user(kfl->l_type, &ufl->l_type) ||
+           __get_user(kfl->l_whence, &ufl->l_whence) ||
+           __get_user(kfl->l_start, &ufl->l_start) ||
+           __get_user(kfl->l_len, &ufl->l_len) ||
+           __get_user(kfl->l_pid, &ufl->l_pid))
+               return -EFAULT;
+       return 0;
+}
+#endif
+
+#ifndef HAVE_ARCH_PUT_COMPAT_FLOCK64
+static int put_compat_flock64(struct flock *kfl, struct compat_flock64 __user *ufl)
+{
+       if (!access_ok(VERIFY_WRITE, ufl, sizeof(*ufl)) ||
+           __put_user(kfl->l_type, &ufl->l_type) ||
+           __put_user(kfl->l_whence, &ufl->l_whence) ||
+           __put_user(kfl->l_start, &ufl->l_start) ||
+           __put_user(kfl->l_len, &ufl->l_len) ||
+           __put_user(kfl->l_pid, &ufl->l_pid))
+               return -EFAULT;
+       return 0;
+}
+#endif
+
+static unsigned int
+convert_fcntl_cmd(unsigned int cmd)
+{
+       switch (cmd) {
+       case F_GETLK64:
+               return F_GETLK;
+       case F_SETLK64:
+               return F_SETLK;
+       case F_SETLKW64:
+               return F_SETLKW;
+       }
+
+       return cmd;
+}
+
+COMPAT_SYSCALL_DEFINE3(fcntl64, unsigned int, fd, unsigned int, cmd,
+                      compat_ulong_t, arg)
+{
+       mm_segment_t old_fs;
+       struct flock f;
+       long ret;
+       unsigned int conv_cmd;
+
+       switch (cmd) {
+       case F_GETLK:
+       case F_SETLK:
+       case F_SETLKW:
+               ret = get_compat_flock(&f, compat_ptr(arg));
+               if (ret != 0)
+                       break;
+               old_fs = get_fs();
+               set_fs(KERNEL_DS);
+               ret = sys_fcntl(fd, cmd, (unsigned long)&f);
+               set_fs(old_fs);
+               if (cmd == F_GETLK && ret == 0) {
+                       /* GETLK was successful and we need to return the data...
+                        * but it needs to fit in the compat structure.
+                        * l_start shouldn't be too big, unless the original
+                        * start + end is greater than COMPAT_OFF_T_MAX, in which
+                        * case the app was asking for trouble, so we return
+                        * -EOVERFLOW in that case.
+                        * l_len could be too big, in which case we just truncate it,
+                        * and only allow the app to see that part of the conflicting
+                        * lock that might make sense to it anyway
+                        */
+
+                       if (f.l_start > COMPAT_OFF_T_MAX)
+                               ret = -EOVERFLOW;
+                       if (f.l_len > COMPAT_OFF_T_MAX)
+                               f.l_len = COMPAT_OFF_T_MAX;
+                       if (ret == 0)
+                               ret = put_compat_flock(&f, compat_ptr(arg));
+               }
+               break;
+
+       case F_GETLK64:
+       case F_SETLK64:
+       case F_SETLKW64:
+       case F_OFD_GETLK:
+       case F_OFD_SETLK:
+       case F_OFD_SETLKW:
+               ret = get_compat_flock64(&f, compat_ptr(arg));
+               if (ret != 0)
+                       break;
+               old_fs = get_fs();
+               set_fs(KERNEL_DS);
+               conv_cmd = convert_fcntl_cmd(cmd);
+               ret = sys_fcntl(fd, conv_cmd, (unsigned long)&f);
+               set_fs(old_fs);
+               if ((conv_cmd == F_GETLK || conv_cmd == F_OFD_GETLK) && ret == 0) {
+                       /* need to return lock information - see above for commentary */
+                       if (f.l_start > COMPAT_LOFF_T_MAX)
+                               ret = -EOVERFLOW;
+                       if (f.l_len > COMPAT_LOFF_T_MAX)
+                               f.l_len = COMPAT_LOFF_T_MAX;
+                       if (ret == 0)
+                               ret = put_compat_flock64(&f, compat_ptr(arg));
+               }
+               break;
+
+       default:
+               ret = sys_fcntl(fd, cmd, arg);
+               break;
+       }
+       return ret;
+}
+
+COMPAT_SYSCALL_DEFINE3(fcntl, unsigned int, fd, unsigned int, cmd,
+                      compat_ulong_t, arg)
+{
+       switch (cmd) {
+       case F_GETLK64:
+       case F_SETLK64:
+       case F_SETLKW64:
+       case F_OFD_GETLK:
+       case F_OFD_SETLK:
+       case F_OFD_SETLKW:
+               return -EINVAL;
+       }
+       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)
+
+ssize_t compat_rw_copy_check_uvector(int type,
+               const struct compat_iovec __user *uvector, unsigned long nr_segs,
+               unsigned long fast_segs, struct iovec *fast_pointer,
+               struct iovec **ret_pointer)
+{
+       compat_ssize_t tot_len;
+       struct iovec *iov = *ret_pointer = fast_pointer;
+       ssize_t ret = 0;
+       int seg;
+
+       /*
+        * SuS says "The readv() function *may* fail if the iovcnt argument
+        * was less than or equal to 0, or greater than {IOV_MAX}.  Linux has
+        * traditionally returned zero for zero segments, so...
+        */
+       if (nr_segs == 0)
+               goto out;
+
+       ret = -EINVAL;
+       if (nr_segs > UIO_MAXIOV || nr_segs < 0)
+               goto out;
+       if (nr_segs > fast_segs) {
+               ret = -ENOMEM;
+               iov = kmalloc(nr_segs*sizeof(struct iovec), GFP_KERNEL);
+               if (iov == NULL)
+                       goto out;
+       }
+       *ret_pointer = iov;
+
+       ret = -EFAULT;
+       if (!access_ok(VERIFY_READ, uvector, nr_segs*sizeof(*uvector)))
+               goto out;
+
+       /*
+        * Single unix specification:
+        * We should -EINVAL if an element length is not >= 0 and fitting an
+        * ssize_t.
+        *
+        * In Linux, the total length is limited to MAX_RW_COUNT, there is
+        * no overflow possibility.
+        */
+       tot_len = 0;
+       ret = -EINVAL;
+       for (seg = 0; seg < nr_segs; seg++) {
+               compat_uptr_t buf;
+               compat_ssize_t len;
+
+               if (__get_user(len, &uvector->iov_len) ||
+                  __get_user(buf, &uvector->iov_base)) {
+                       ret = -EFAULT;
+                       goto out;
+               }
+               if (len < 0)    /* size_t not fitting in compat_ssize_t .. */
+                       goto out;
+               if (type >= 0 &&
+                   !access_ok(vrfy_dir(type), compat_ptr(buf), len)) {
+                       ret = -EFAULT;
+                       goto out;
+               }
+               if (len > MAX_RW_COUNT - tot_len)
+                       len = MAX_RW_COUNT - tot_len;
+               tot_len += len;
+               iov->iov_base = compat_ptr(buf);
+               iov->iov_len = (compat_size_t) len;
+               uvector++;
+               iov++;
+       }
+       ret = tot_len;
+
+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;
+       __compat_uid_t mounted_uid;
+       compat_pid_t wdog_pid;
+       unsigned char mounted_vol[NCP_VOLNAME_LEN + 1];
+       compat_uint_t time_out;
+       compat_uint_t retry_count;
+       compat_uint_t flags;
+       __compat_uid_t uid;
+       __compat_gid_t gid;
+       compat_mode_t file_mode;
+       compat_mode_t dir_mode;
+};
+
+struct compat_ncp_mount_data_v4 {
+       compat_int_t version;
+       compat_ulong_t flags;
+       compat_ulong_t mounted_uid;
+       compat_long_t wdog_pid;
+       compat_uint_t ncp_fd;
+       compat_uint_t time_out;
+       compat_uint_t retry_count;
+       compat_ulong_t uid;
+       compat_ulong_t gid;
+       compat_ulong_t file_mode;
+       compat_ulong_t dir_mode;
+};
+
+static void *do_ncp_super_data_conv(void *raw_data)
+{
+       int version = *(unsigned int *)raw_data;
+
+       if (version == 3) {
+               struct compat_ncp_mount_data *c_n = raw_data;
+               struct ncp_mount_data *n = raw_data;
+
+               n->dir_mode = c_n->dir_mode;
+               n->file_mode = c_n->file_mode;
+               n->gid = c_n->gid;
+               n->uid = c_n->uid;
+               memmove (n->mounted_vol, c_n->mounted_vol, (sizeof (c_n->mounted_vol) + 3 * sizeof (unsigned int)));
+               n->wdog_pid = c_n->wdog_pid;
+               n->mounted_uid = c_n->mounted_uid;
+       } else if (version == 4) {
+               struct compat_ncp_mount_data_v4 *c_n = raw_data;
+               struct ncp_mount_data_v4 *n = raw_data;
+
+               n->dir_mode = c_n->dir_mode;
+               n->file_mode = c_n->file_mode;
+               n->gid = c_n->gid;
+               n->uid = c_n->uid;
+               n->retry_count = c_n->retry_count;
+               n->time_out = c_n->time_out;
+               n->ncp_fd = c_n->ncp_fd;
+               n->wdog_pid = c_n->wdog_pid;
+               n->mounted_uid = c_n->mounted_uid;
+               n->flags = c_n->flags;
+       } else if (version != 5) {
+               return NULL;
+       }
+
+       return raw_data;
+}
+
+
+struct compat_nfs_string {
+       compat_uint_t len;
+       compat_uptr_t data;
+};
+
+static inline void compat_nfs_string(struct nfs_string *dst,
+                                    struct compat_nfs_string *src)
+{
+       dst->data = compat_ptr(src->data);
+       dst->len = src->len;
+}
+
+struct compat_nfs4_mount_data_v1 {
+       compat_int_t version;
+       compat_int_t flags;
+       compat_int_t rsize;
+       compat_int_t wsize;
+       compat_int_t timeo;
+       compat_int_t retrans;
+       compat_int_t acregmin;
+       compat_int_t acregmax;
+       compat_int_t acdirmin;
+       compat_int_t acdirmax;
+       struct compat_nfs_string client_addr;
+       struct compat_nfs_string mnt_path;
+       struct compat_nfs_string hostname;
+       compat_uint_t host_addrlen;
+       compat_uptr_t host_addr;
+       compat_int_t proto;
+       compat_int_t auth_flavourlen;
+       compat_uptr_t auth_flavours;
+};
+
+static int do_nfs4_super_data_conv(void *raw_data)
+{
+       int version = *(compat_uint_t *) raw_data;
+
+       if (version == 1) {
+               struct compat_nfs4_mount_data_v1 *raw = raw_data;
+               struct nfs4_mount_data *real = raw_data;
+
+               /* copy the fields backwards */
+               real->auth_flavours = compat_ptr(raw->auth_flavours);
+               real->auth_flavourlen = raw->auth_flavourlen;
+               real->proto = raw->proto;
+               real->host_addr = compat_ptr(raw->host_addr);
+               real->host_addrlen = raw->host_addrlen;
+               compat_nfs_string(&real->hostname, &raw->hostname);
+               compat_nfs_string(&real->mnt_path, &raw->mnt_path);
+               compat_nfs_string(&real->client_addr, &raw->client_addr);
+               real->acdirmax = raw->acdirmax;
+               real->acdirmin = raw->acdirmin;
+               real->acregmax = raw->acregmax;
+               real->acregmin = raw->acregmin;
+               real->retrans = raw->retrans;
+               real->timeo = raw->timeo;
+               real->wsize = raw->wsize;
+               real->rsize = raw->rsize;
+               real->flags = raw->flags;
+               real->version = raw->version;
+       }
+
+       return 0;
+}
+
+#define NCPFS_NAME      "ncpfs"
+#define NFS4_NAME      "nfs4"
+
+COMPAT_SYSCALL_DEFINE5(mount, const char __user *, dev_name,
+                      const char __user *, dir_name,
+                      const char __user *, type, compat_ulong_t, flags,
+                      const void __user *, data)
+{
+       char *kernel_type;
+       unsigned long data_page;
+       char *kernel_dev;
+       int retval;
+
+       kernel_type = copy_mount_string(type);
+       retval = PTR_ERR(kernel_type);
+       if (IS_ERR(kernel_type))
+               goto out;
+
+       kernel_dev = copy_mount_string(dev_name);
+       retval = PTR_ERR(kernel_dev);
+       if (IS_ERR(kernel_dev))
+               goto out1;
+
+       retval = copy_mount_options(data, &data_page);
+       if (retval < 0)
+               goto out2;
+
+       retval = -EINVAL;
+
+       if (kernel_type && data_page) {
+               if (!strcmp(kernel_type, NCPFS_NAME)) {
+                       do_ncp_super_data_conv((void *)data_page);
+               } else if (!strcmp(kernel_type, NFS4_NAME)) {
+                       if (do_nfs4_super_data_conv((void *) data_page))
+                               goto out3;
+               }
+       }
+
+       retval = do_mount(kernel_dev, dir_name, kernel_type,
+                       flags, (void*)data_page);
+
+ out3:
+       free_page(data_page);
+ out2:
+       kfree(kernel_dev);
+ out1:
+       kfree(kernel_type);
+ out:
+       return retval;
+}
+
+struct compat_old_linux_dirent {
+       compat_ulong_t  d_ino;
+       compat_ulong_t  d_offset;
+       unsigned short  d_namlen;
+       char            d_name[1];
+};
+
+struct compat_readdir_callback {
+       struct dir_context ctx;
+       struct compat_old_linux_dirent __user *dirent;
+       int result;
+};
+
+static int compat_fillonedir(struct dir_context *ctx, const char *name,
+                            int namlen, loff_t offset, u64 ino,
+                            unsigned int d_type)
+{
+       struct compat_readdir_callback *buf =
+               container_of(ctx, struct compat_readdir_callback, ctx);
+       struct compat_old_linux_dirent __user *dirent;
+       compat_ulong_t d_ino;
+
+       if (buf->result)
+               return -EINVAL;
+       d_ino = ino;
+       if (sizeof(d_ino) < sizeof(ino) && d_ino != ino) {
+               buf->result = -EOVERFLOW;
+               return -EOVERFLOW;
+       }
+       buf->result++;
+       dirent = buf->dirent;
+       if (!access_ok(VERIFY_WRITE, dirent,
+                       (unsigned long)(dirent->d_name + namlen + 1) -
+                               (unsigned long)dirent))
+               goto efault;
+       if (    __put_user(d_ino, &dirent->d_ino) ||
+               __put_user(offset, &dirent->d_offset) ||
+               __put_user(namlen, &dirent->d_namlen) ||
+               __copy_to_user(dirent->d_name, name, namlen) ||
+               __put_user(0, dirent->d_name + namlen))
+               goto efault;
+       return 0;
+efault:
+       buf->result = -EFAULT;
+       return -EFAULT;
+}
+
+COMPAT_SYSCALL_DEFINE3(old_readdir, unsigned int, fd,
+               struct compat_old_linux_dirent __user *, dirent, unsigned int, count)
+{
+       int error;
+       struct fd f = fdget(fd);
+       struct compat_readdir_callback buf = {
+               .ctx.actor = compat_fillonedir,
+               .dirent = dirent
+       };
+
+       if (!f.file)
+               return -EBADF;
+
+       error = iterate_dir(f.file, &buf.ctx);
+       if (buf.result)
+               error = buf.result;
+
+       fdput(f);
+       return error;
+}
+
+struct compat_linux_dirent {
+       compat_ulong_t  d_ino;
+       compat_ulong_t  d_off;
+       unsigned short  d_reclen;
+       char            d_name[1];
+};
+
+struct compat_getdents_callback {
+       struct dir_context ctx;
+       struct compat_linux_dirent __user *current_dir;
+       struct compat_linux_dirent __user *previous;
+       int count;
+       int error;
+};
+
+static int compat_filldir(struct dir_context *ctx, const char *name, int namlen,
+               loff_t offset, u64 ino, unsigned int d_type)
+{
+       struct compat_linux_dirent __user * dirent;
+       struct compat_getdents_callback *buf =
+               container_of(ctx, struct compat_getdents_callback, ctx);
+       compat_ulong_t d_ino;
+       int reclen = ALIGN(offsetof(struct compat_linux_dirent, d_name) +
+               namlen + 2, sizeof(compat_long_t));
+
+       buf->error = -EINVAL;   /* only used if we fail.. */
+       if (reclen > buf->count)
+               return -EINVAL;
+       d_ino = ino;
+       if (sizeof(d_ino) < sizeof(ino) && d_ino != ino) {
+               buf->error = -EOVERFLOW;
+               return -EOVERFLOW;
+       }
+       dirent = buf->previous;
+       if (dirent) {
+               if (__put_user(offset, &dirent->d_off))
+                       goto efault;
+       }
+       dirent = buf->current_dir;
+       if (__put_user(d_ino, &dirent->d_ino))
+               goto efault;
+       if (__put_user(reclen, &dirent->d_reclen))
+               goto efault;
+       if (copy_to_user(dirent->d_name, name, namlen))
+               goto efault;
+       if (__put_user(0, dirent->d_name + namlen))
+               goto efault;
+       if (__put_user(d_type, (char  __user *) dirent + reclen - 1))
+               goto efault;
+       buf->previous = dirent;
+       dirent = (void __user *)dirent + reclen;
+       buf->current_dir = dirent;
+       buf->count -= reclen;
+       return 0;
+efault:
+       buf->error = -EFAULT;
+       return -EFAULT;
+}
+
+COMPAT_SYSCALL_DEFINE3(getdents, unsigned int, fd,
+               struct compat_linux_dirent __user *, dirent, unsigned int, count)
+{
+       struct fd f;
+       struct compat_linux_dirent __user * lastdirent;
+       struct compat_getdents_callback buf = {
+               .ctx.actor = compat_filldir,
+               .current_dir = dirent,
+               .count = count
+       };
+       int error;
+
+       if (!access_ok(VERIFY_WRITE, dirent, count))
+               return -EFAULT;
+
+       f = fdget(fd);
+       if (!f.file)
+               return -EBADF;
+
+       error = iterate_dir(f.file, &buf.ctx);
+       if (error >= 0)
+               error = buf.error;
+       lastdirent = buf.previous;
+       if (lastdirent) {
+               if (put_user(buf.ctx.pos, &lastdirent->d_off))
+                       error = -EFAULT;
+               else
+                       error = count - buf.count;
+       }
+       fdput(f);
+       return error;
+}
+
+#ifdef __ARCH_WANT_COMPAT_SYS_GETDENTS64
+
+struct compat_getdents_callback64 {
+       struct dir_context ctx;
+       struct linux_dirent64 __user *current_dir;
+       struct linux_dirent64 __user *previous;
+       int count;
+       int error;
+};
+
+static int compat_filldir64(struct dir_context *ctx, const char *name,
+                           int namlen, loff_t offset, u64 ino,
+                           unsigned int d_type)
+{
+       struct linux_dirent64 __user *dirent;
+       struct compat_getdents_callback64 *buf =
+               container_of(ctx, struct compat_getdents_callback64, ctx);
+       int reclen = ALIGN(offsetof(struct linux_dirent64, d_name) + namlen + 1,
+               sizeof(u64));
+       u64 off;
+
+       buf->error = -EINVAL;   /* only used if we fail.. */
+       if (reclen > buf->count)
+               return -EINVAL;
+       dirent = buf->previous;
+
+       if (dirent) {
+               if (__put_user_unaligned(offset, &dirent->d_off))
+                       goto efault;
+       }
+       dirent = buf->current_dir;
+       if (__put_user_unaligned(ino, &dirent->d_ino))
+               goto efault;
+       off = 0;
+       if (__put_user_unaligned(off, &dirent->d_off))
+               goto efault;
+       if (__put_user(reclen, &dirent->d_reclen))
+               goto efault;
+       if (__put_user(d_type, &dirent->d_type))
+               goto efault;
+       if (copy_to_user(dirent->d_name, name, namlen))
+               goto efault;
+       if (__put_user(0, dirent->d_name + namlen))
+               goto efault;
+       buf->previous = dirent;
+       dirent = (void __user *)dirent + reclen;
+       buf->current_dir = dirent;
+       buf->count -= reclen;
+       return 0;
+efault:
+       buf->error = -EFAULT;
+       return -EFAULT;
+}
+
+COMPAT_SYSCALL_DEFINE3(getdents64, unsigned int, fd,
+               struct linux_dirent64 __user *, dirent, unsigned int, count)
+{
+       struct fd f;
+       struct linux_dirent64 __user * lastdirent;
+       struct compat_getdents_callback64 buf = {
+               .ctx.actor = compat_filldir64,
+               .current_dir = dirent,
+               .count = count
+       };
+       int error;
+
+       if (!access_ok(VERIFY_WRITE, dirent, count))
+               return -EFAULT;
+
+       f = fdget(fd);
+       if (!f.file)
+               return -EBADF;
+
+       error = iterate_dir(f.file, &buf.ctx);
+       if (error >= 0)
+               error = buf.error;
+       lastdirent = buf.previous;
+       if (lastdirent) {
+               typeof(lastdirent->d_off) d_off = buf.ctx.pos;
+               if (__put_user_unaligned(d_off, &lastdirent->d_off))
+                       error = -EFAULT;
+               else
+                       error = count - buf.count;
+       }
+       fdput(f);
+       return error;
+}
+#endif /* __ARCH_WANT_COMPAT_SYS_GETDENTS64 */
+
+/*
+ * Exactly like fs/open.c:sys_open(), except that it doesn't set the
+ * O_LARGEFILE flag.
+ */
+COMPAT_SYSCALL_DEFINE3(open, const char __user *, filename, int, flags, umode_t, mode)
+{
+       return do_sys_open(AT_FDCWD, filename, flags, mode);
+}
+
+/*
+ * Exactly like fs/open.c:sys_openat(), except that it doesn't set the
+ * O_LARGEFILE flag.
+ */
+COMPAT_SYSCALL_DEFINE4(openat, int, dfd, const char __user *, filename, int, flags, umode_t, mode)
+{
+       return do_sys_open(dfd, filename, flags, mode);
+}
+
+#define __COMPAT_NFDBITS       (8 * sizeof(compat_ulong_t))
+
+static int poll_select_copy_remaining(struct timespec *end_time, void __user *p,
+                                     int timeval, int ret)
+{
+       struct timespec ts;
+
+       if (!p)
+               return ret;
+
+       if (current->personality & STICKY_TIMEOUTS)
+               goto sticky;
+
+       /* No update for zero timeout */
+       if (!end_time->tv_sec && !end_time->tv_nsec)
+               return ret;
+
+       ktime_get_ts(&ts);
+       ts = timespec_sub(*end_time, ts);
+       if (ts.tv_sec < 0)
+               ts.tv_sec = ts.tv_nsec = 0;
+
+       if (timeval) {
+               struct compat_timeval rtv;
+
+               rtv.tv_sec = ts.tv_sec;
+               rtv.tv_usec = ts.tv_nsec / NSEC_PER_USEC;
+
+               if (!copy_to_user(p, &rtv, sizeof(rtv)))
+                       return ret;
+       } else {
+               struct compat_timespec rts;
+
+               rts.tv_sec = ts.tv_sec;
+               rts.tv_nsec = ts.tv_nsec;
+
+               if (!copy_to_user(p, &rts, sizeof(rts)))
+                       return ret;
+       }
+       /*
+        * If an application puts its timeval in read-only memory, we
+        * don't want the Linux-specific update to the timeval to
+        * cause a fault after the select has completed
+        * successfully. However, because we're not updating the
+        * timeval, we can't restart the system call.
+        */
+
+sticky:
+       if (ret == -ERESTARTNOHAND)
+               ret = -EINTR;
+       return ret;
+}
+
+/*
+ * Ooo, nasty.  We need here to frob 32-bit unsigned longs to
+ * 64-bit unsigned longs.
+ */
+static
+int compat_get_fd_set(unsigned long nr, compat_ulong_t __user *ufdset,
+                       unsigned long *fdset)
+{
+       nr = DIV_ROUND_UP(nr, __COMPAT_NFDBITS);
+       if (ufdset) {
+               unsigned long odd;
+
+               if (!access_ok(VERIFY_WRITE, ufdset, nr*sizeof(compat_ulong_t)))
+                       return -EFAULT;
+
+               odd = nr & 1UL;
+               nr &= ~1UL;
+               while (nr) {
+                       unsigned long h, l;
+                       if (__get_user(l, ufdset) || __get_user(h, ufdset+1))
+                               return -EFAULT;
+                       ufdset += 2;
+                       *fdset++ = h << 32 | l;
+                       nr -= 2;
+               }
+               if (odd && __get_user(*fdset, ufdset))
+                       return -EFAULT;
+       } else {
+               /* Tricky, must clear full unsigned long in the
+                * kernel fdset at the end, this makes sure that
+                * actually happens.
+                */
+               memset(fdset, 0, ((nr + 1) & ~1)*sizeof(compat_ulong_t));
+       }
+       return 0;
+}
+
+static
+int compat_set_fd_set(unsigned long nr, compat_ulong_t __user *ufdset,
+                     unsigned long *fdset)
+{
+       unsigned long odd;
+       nr = DIV_ROUND_UP(nr, __COMPAT_NFDBITS);
+
+       if (!ufdset)
+               return 0;
+
+       odd = nr & 1UL;
+       nr &= ~1UL;
+       while (nr) {
+               unsigned long h, l;
+               l = *fdset++;
+               h = l >> 32;
+               if (__put_user(l, ufdset) || __put_user(h, ufdset+1))
+                       return -EFAULT;
+               ufdset += 2;
+               nr -= 2;
+       }
+       if (odd && __put_user(*fdset, ufdset))
+               return -EFAULT;
+       return 0;
+}
+
+
+/*
+ * This is a virtual copy of sys_select from fs/select.c and probably
+ * should be compared to it from time to time
+ */
+
+/*
+ * We can actually return ERESTARTSYS instead of EINTR, but I'd
+ * like to be certain this leads to no problems. So I return
+ * EINTR just for safety.
+ *
+ * Update: ERESTARTSYS breaks at least the xview clock binary, so
+ * I'm trying ERESTARTNOHAND which restart only when you want to.
+ */
+int compat_core_sys_select(int n, compat_ulong_t __user *inp,
+       compat_ulong_t __user *outp, compat_ulong_t __user *exp,
+       struct timespec *end_time)
+{
+       fd_set_bits fds;
+       void *bits;
+       int size, max_fds, ret = -EINVAL;
+       struct fdtable *fdt;
+       long stack_fds[SELECT_STACK_ALLOC/sizeof(long)];
+
+       if (n < 0)
+               goto out_nofds;
+
+       /* max_fds can increase, so grab it once to avoid race */
+       rcu_read_lock();
+       fdt = files_fdtable(current->files);
+       max_fds = fdt->max_fds;
+       rcu_read_unlock();
+       if (n > max_fds)
+               n = max_fds;
+
+       /*
+        * We need 6 bitmaps (in/out/ex for both incoming and outgoing),
+        * since we used fdset we need to allocate memory in units of
+        * long-words.
+        */
+       size = FDS_BYTES(n);
+       bits = stack_fds;
+       if (size > sizeof(stack_fds) / 6) {
+               bits = kmalloc(6 * size, GFP_KERNEL);
+               ret = -ENOMEM;
+               if (!bits)
+                       goto out_nofds;
+       }
+       fds.in      = (unsigned long *)  bits;
+       fds.out     = (unsigned long *) (bits +   size);
+       fds.ex      = (unsigned long *) (bits + 2*size);
+       fds.res_in  = (unsigned long *) (bits + 3*size);
+       fds.res_out = (unsigned long *) (bits + 4*size);
+       fds.res_ex  = (unsigned long *) (bits + 5*size);
+
+       if ((ret = compat_get_fd_set(n, inp, fds.in)) ||
+           (ret = compat_get_fd_set(n, outp, fds.out)) ||
+           (ret = compat_get_fd_set(n, exp, fds.ex)))
+               goto out;
+       zero_fd_set(n, fds.res_in);
+       zero_fd_set(n, fds.res_out);
+       zero_fd_set(n, fds.res_ex);
+
+       ret = do_select(n, &fds, end_time);
+
+       if (ret < 0)
+               goto out;
+       if (!ret) {
+               ret = -ERESTARTNOHAND;
+               if (signal_pending(current))
+                       goto out;
+               ret = 0;
+       }
+
+       if (compat_set_fd_set(n, inp, fds.res_in) ||
+           compat_set_fd_set(n, outp, fds.res_out) ||
+           compat_set_fd_set(n, exp, fds.res_ex))
+               ret = -EFAULT;
+out:
+       if (bits != stack_fds)
+               kfree(bits);
+out_nofds:
+       return ret;
+}
+
+COMPAT_SYSCALL_DEFINE5(select, int, n, compat_ulong_t __user *, inp,
+       compat_ulong_t __user *, outp, compat_ulong_t __user *, exp,
+       struct compat_timeval __user *, tvp)
+{
+       struct timespec end_time, *to = NULL;
+       struct compat_timeval tv;
+       int ret;
+
+       if (tvp) {
+               if (copy_from_user(&tv, tvp, sizeof(tv)))
+                       return -EFAULT;
+
+               to = &end_time;
+               if (poll_select_set_timeout(to,
+                               tv.tv_sec + (tv.tv_usec / USEC_PER_SEC),
+                               (tv.tv_usec % USEC_PER_SEC) * NSEC_PER_USEC))
+                       return -EINVAL;
+       }
+
+       ret = compat_core_sys_select(n, inp, outp, exp, to);
+       ret = poll_select_copy_remaining(&end_time, tvp, 1, ret);
+
+       return ret;
+}
+
+struct compat_sel_arg_struct {
+       compat_ulong_t n;
+       compat_uptr_t inp;
+       compat_uptr_t outp;
+       compat_uptr_t exp;
+       compat_uptr_t tvp;
+};
+
+COMPAT_SYSCALL_DEFINE1(old_select, struct compat_sel_arg_struct __user *, arg)
+{
+       struct compat_sel_arg_struct a;
+
+       if (copy_from_user(&a, arg, sizeof(a)))
+               return -EFAULT;
+       return compat_sys_select(a.n, compat_ptr(a.inp), compat_ptr(a.outp),
+                                compat_ptr(a.exp), compat_ptr(a.tvp));
+}
+
+static long do_compat_pselect(int n, compat_ulong_t __user *inp,
+       compat_ulong_t __user *outp, compat_ulong_t __user *exp,
+       struct compat_timespec __user *tsp, compat_sigset_t __user *sigmask,
+       compat_size_t sigsetsize)
+{
+       compat_sigset_t ss32;
+       sigset_t ksigmask, sigsaved;
+       struct compat_timespec ts;
+       struct timespec end_time, *to = NULL;
+       int ret;
+
+       if (tsp) {
+               if (copy_from_user(&ts, tsp, sizeof(ts)))
+                       return -EFAULT;
+
+               to = &end_time;
+               if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
+                       return -EINVAL;
+       }
+
+       if (sigmask) {
+               if (sigsetsize != sizeof(compat_sigset_t))
+                       return -EINVAL;
+               if (copy_from_user(&ss32, sigmask, sizeof(ss32)))
+                       return -EFAULT;
+               sigset_from_compat(&ksigmask, &ss32);
+
+               sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
+               sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
+       }
+
+       ret = compat_core_sys_select(n, inp, outp, exp, to);
+       ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
+
+       if (ret == -ERESTARTNOHAND) {
+               /*
+                * Don't restore the signal mask yet. Let do_signal() deliver
+                * the signal on the way back to userspace, before the signal
+                * mask is restored.
+                */
+               if (sigmask) {
+                       memcpy(&current->saved_sigmask, &sigsaved,
+                                       sizeof(sigsaved));
+                       set_restore_sigmask();
+               }
+       } else if (sigmask)
+               sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+       return ret;
+}
+
+COMPAT_SYSCALL_DEFINE6(pselect6, int, n, compat_ulong_t __user *, inp,
+       compat_ulong_t __user *, outp, compat_ulong_t __user *, exp,
+       struct compat_timespec __user *, tsp, void __user *, sig)
+{
+       compat_size_t sigsetsize = 0;
+       compat_uptr_t up = 0;
+
+       if (sig) {
+               if (!access_ok(VERIFY_READ, sig,
+                               sizeof(compat_uptr_t)+sizeof(compat_size_t)) ||
+                       __get_user(up, (compat_uptr_t __user *)sig) ||
+                       __get_user(sigsetsize,
+                               (compat_size_t __user *)(sig+sizeof(up))))
+                       return -EFAULT;
+       }
+       return do_compat_pselect(n, inp, outp, exp, tsp, compat_ptr(up),
+                                sigsetsize);
+}
+
+COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
+       unsigned int,  nfds, struct compat_timespec __user *, tsp,
+       const compat_sigset_t __user *, sigmask, compat_size_t, sigsetsize)
+{
+       compat_sigset_t ss32;
+       sigset_t ksigmask, sigsaved;
+       struct compat_timespec ts;
+       struct timespec end_time, *to = NULL;
+       int ret;
+
+       if (tsp) {
+               if (copy_from_user(&ts, tsp, sizeof(ts)))
+                       return -EFAULT;
+
+               to = &end_time;
+               if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
+                       return -EINVAL;
+       }
+
+       if (sigmask) {
+               if (sigsetsize != sizeof(compat_sigset_t))
+                       return -EINVAL;
+               if (copy_from_user(&ss32, sigmask, sizeof(ss32)))
+                       return -EFAULT;
+               sigset_from_compat(&ksigmask, &ss32);
+
+               sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
+               sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
+       }
+
+       ret = do_sys_poll(ufds, nfds, to);
+
+       /* We can restart this syscall, usually */
+       if (ret == -EINTR) {
+               /*
+                * Don't restore the signal mask yet. Let do_signal() deliver
+                * the signal on the way back to userspace, before the signal
+                * mask is restored.
+                */
+               if (sigmask) {
+                       memcpy(&current->saved_sigmask, &sigsaved,
+                               sizeof(sigsaved));
+                       set_restore_sigmask();
+               }
+               ret = -ERESTARTNOHAND;
+       } else if (sigmask)
+               sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+       ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
+
+       return ret;
+}
+
+#ifdef CONFIG_FHANDLE
+/*
+ * Exactly like fs/open.c:sys_open_by_handle_at(), except that it
+ * doesn't set the O_LARGEFILE flag.
+ */
+COMPAT_SYSCALL_DEFINE3(open_by_handle_at, int, mountdirfd,
+                            struct file_handle __user *, handle, int, flags)
+{
+       return do_handle_open(mountdirfd, handle, flags);
+}
+#endif