Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Signed-off-by: NJens Axboe <axboe@suse.de>

cfq_exit_lock is protecting two things now:

- The per-ioc rbtree of cfq_io_contexts

- The per-cfqd linked list of cfq_io_contexts

The per-cfqd linked list can be protected by the queue lock, as it is (by
definition) per cfqd as the queue lock is.

The per-ioc rbtree is mainly used and updated by the process itself only.
The only outside use is the io priority changing. If we move the
priority changing to not browsing the rbtree, we can remove any locking
from the rbtree updates and lookup completely. Let the sys_ioprio syscall
just mark processes as having the iopriority changed and lazily update
the private cfq io contexts the next time io is queued, and we can
remove this locking as well.
Signed-off-by: NJens Axboe <axboe@suse.de>

o As per ELF specifications, it looks like that elf note "namesz" field
  contains the length of "name" including the size of null character.  And
  currently we are filling "namesz" without taking into the consideration
  the null character size.

o Kexec-tools performs this check deligently hence I ran into the issue
  while trying to open /proc/kcore in kexec-tools for some info.
Signed-off-by: NVivek Goyal <vgoyal@in.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This unlock/lock on a super-unlikely path isn't worth the kernel text.

Cc: Vadim Lobanov <vlobanov@speakeasy.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Perform a code cleanup against the expand_fdtable() and expand_files()
functions inside fs/file.c.  It aims to make the flow of code within these
functions simpler and easier to understand, via added comments and modest
refactoring.
Signed-off-by: NVadim Lobanov <vlobanov@speakeasy.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Add access control lists for tmpfs.
Signed-off-by: NAndreas Gruenbacher <agruen@suse.de>
Cc: Hugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The patches solve the following problem: We want to grant access to devices
based on who is logged in from where, etc.  This includes switching back and
forth between multiple user sessions, etc.

Using ACLs to define device access for logged-in users gives us all the
flexibility we need in order to fully solve the problem.

Device special files nowadays usually live on tmpfs, hence tmpfs ACLs.

Different distros have come up with solutions that solve the problem to
different degrees: SUSE uses a resource manager which tracks login sessions
and sets ACLs on device inodes as appropriate.  RedHat uses pam_console, which
changes the primary file ownership to the logged-in user.  Others use a set of
groups that users must be in in order to be granted the appropriate accesses.

The freedesktop.org project plans to implement a combination of a
console-tracker and a HAL-device-list based solution to grant access to
devices to users, and more distros will likely follow this approach.

These patches have first been posted here on 2 February 2005, and again
on 8 January 2006. We have been shipping them in SLES9 and SLES10 with
no problems reported.  The previous submission is archived here:

   http://lkml.org/lkml/2006/1/8/229
   http://lkml.org/lkml/2006/1/8/230
   http://lkml.org/lkml/2006/1/8/231

This patch:

Add some infrastructure for access control lists on in-memory
filesystems such as tmpfs.
Signed-off-by: NAndreas Gruenbacher <agruen@suse.de>
Cc: Hugh Dickins <hugh@veritas.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

POSIX states that poll() shall fail with EINVAL if nfds > OPEN_MAX.  In
this context, POSIX is referring to sysconf(OPEN_MAX), which is the value
of current->signal->rlim[RLIMIT_NOFILE].rlim_cur in the linux kernel, not
the compile-time constant which happens to also be named OPEN_MAX.  In the
current code, an application may poll up to max_fdset file descriptors,
even if this exceeds RLIMIT_NOFILE.  The current code also breaks
applications which poll more than max_fdset descriptors, which worked circa
2.4.18 when the check was against NR_OPEN, which is 1024*1024.  This patch
enforces the limit precisely as POSIX defines, even if RLIMIT_NOFILE has
been changed at run time with ulimit -n.

To elaborate on the rationale for this, there are three cases:

1) RLIMIT_NOFILE is at the default value of 1024

In this (default) case, the patch changes nothing.  Calls with nfds > 1024
fail with EINVAL both before and after the patch, and calls with nfds <=
1024 pass the check both before and after the patch, since 1024 is the
initial value of max_fdset.

2) RLIMIT_NOFILE has been raised above the default

In this case, poll() becomes more permissive, allowing polling up to
RLIMIT_NOFILE file descriptors even if less than 1024 have been opened.
The patch won't introduce new errors here.  If an application somehow
depends on poll() failing when it polls with duplicate or invalid file
descriptors, it's already broken, since this is already allowed below 1024,
and will also work above 1024 if enough file descriptors have been open at
some point to cause max_fdset to have been increased above nfds.

3) RLIMIT_NOFILE has been lowered below the default

In this case, the system administrator or the user has gone out of their
way to protect the system from inefficient (or malicious) applications
wasting kernel memory.  The current code allows polling up to 1024 file
descriptors even if RLIMIT_NOFILE is much lower, which is not what the user
or administrator intended.  Well-written applications which only poll
valid, unique file descriptors will never notice the difference, because
they'll hit the limit on open() first.  If an application gets broken
because of the patch in this case, then it was already poorly/maliciously
designed, and allowing it to work in the past was a violation of POSIX and
a DoS risk on low-resource systems.

With this patch, poll() will permit exactly what POSIX suggests, no more,
no less, and for any run-time value set with ulimit -n, not just 256 or
1024.  There are existing apps which which poll a large number of file
descriptors, some of which may be invalid, and if those numbers stradle
1024, they currently fail with or without the patch in -mm, though they
worked fine under 2.4.18.
Signed-off-by: NChris Snook <csnook@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Replace current->fs by fs helper variable to reduce some indirection
overhead and (at least at the moment, before the current_thread_info() %gs
PDA improvement is available) get rid of more costly current references.
Reduces fs/namei.o from 37786 to 37082 Bytes (704 Bytes saved).

[akpm@osdl.org: cleanup]
Signed-off-by: NAndreas Mohr <andi@lisas.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Everyone passes valid pointer there.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

If register_filesystem() fails mux workqueue must be killed.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: Eric Van Hensbergen <ericvh@gmail.com>
Cc: Ron Minnich <rminnich@lanl.gov>
Cc: Latchesar Ionkov <lucho@ionkov.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

It always returns 0, so relying on it is useless.  The only caller isn't
checking return value.  In general, un-, de-, -free functions should return
void.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

If register_filesystem() fails, vxfs_inode cache must be destroyed.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Two lines -- two bugs. :-(
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

At the beginning of the routine, "copied" is set to 0, but it is no good
because in lines 805 and 812 it is set to other values.  Finally, the
routine returns as if it copied 12 (=ENOMEM) bytes less than it actually
did.
Signed-off-by: NFrederik Deweerdt <frederik.deweerdt@gmail.com>
Acked-by: NEric Biederman <ebiederm@xmission.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

In the case below we are locking the whole disk not a partition.  This
change simply brings the code in line with the piece above where when we
are the 'first' opener, and we are a partition.
Signed-off-by: NJason Baron <jbaron@redhat.com>
Acked-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

During testing I've found that the mount pending flag can be left set at
exit from autofs4_lookup after a failed mount request.  This shouldn't be
allowed to happen and causes incorrect error returns.
Signed-off-by: NIan Kent <raven@themaw.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The check for an empty directory in the autofs4_follow_link method fails
occassionally due to old dentrys.  We had the same problem
autofs4_revalidate ages ago.  I thought we wouldn't need this in
autofs4_follow_link, silly me.
Signed-off-by: NIan Kent <raven@themaw.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Add some documentation comments for the cdev interface.
Signed-off-by: NJonathan Corbet <corbet@lwn.net>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Acked-by: N"Randy.Dunlap" <rdunlap@xenotime.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

There was an I/O error that prevented reading the last partial block of
large files in an ISO9660 filesystem.  The error was generated when a file
comprised more than one section and had a size that was not an exact
multiple of the filesystem block size.  This patch removes the check (and
failure) for reading into the last partial block (and possibly beyond) for
multiple-section files.

It worked in my testing to prevent reading beyond the end of the section;
my first patch just incremented the sect_size block count for a partial
block and continued doing the check.  But there is a commment in the source
code about reading beyond the end of the file to fill a page cache.
Failing to access beyond the section would prevent reading beyond the end
of the file.
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Dquot passes the tty to tty_write_message without locking
Signed-off-by: NJan Kara <jack@suse.cz>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

do_each_thread() is rcu-safe, and all tasks which use this ->mm must sleep
in wait_for_completion(&mm->core_done) at this point, so we can use RCU
locks.

Also, remove unneeded INIT_LIST_HEAD(new) before list_add(new, head).
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Acked-By: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

do_each_thread() is rcu-safe, and all tasks which use this ->mm must sleep
in wait_for_completion(&mm->core_done) at this point, so we can use RCU
locks.

Also, remove unneeded INIT_LIST_HEAD(new) before list_add(new, head).
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The problem is that close() syscalls can call a file system's flush
handler, which in turn might sleep interruptibly and ultimately pass back
an -ERESTARTSYS return value.  This happens for files backed by an
interruptible NFS mount under nfs_file_flush() when a large file has just
been written and nfs_wait_bit_interruptible() detects that there is a
signal pending.

I have a test case where the "strace" command is used to attach to a
process sleeping in such a close().  Since the SIGSTOP is forced onto the
victim process (removing it from the thread's "blocked" mask in
force_sig_info()), the RPC wait is interrupted and the close() is
terminated early.

But the file table entry has already been cleared before the flush handler
was called.  Thus, when the syscall is restarted, the file descriptor
appears closed and an EBADF error is returned (which is wrong).  What's
worse, there is the hypothetical case where another thread of a
multi-threaded application might have reused the file descriptor, in which
case that file would be mistakenly closed.

The bottom line is that close() syscalls are not restartable, and thus
-ERESTARTSYS return values should be mapped to -EINTR.  This is consistent
with the close(2) manual page.  The fix is below.
Signed-off-by: NErnie Petrides <petrides@redhat.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The code in __register_chrdev_region checks that if the driver wishing to
register has the same major as an existing driver the new minor range is
strictly less than the existing minor range.  However, it does not also
check that the new minor range is strictly greater than the existing minor
range.  That is, if driver X has registered with major=x and minor=0-3,
__register_chrdev_region will allow driver Y to register with major=x and
minor=1-4.
Signed-off-by: NAmos Waterland <apw@us.ibm.com>
Cc: Linas Vepstas <linas@austin.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Fixed race on put_files_struct on exec with proc.  Restoring files on
current on error path may lead to proc having a pointer to already kfree-d
files_struct.

->files changing at exit.c and khtread.c are safe as exit_files() makes all
things under lock.

Found during OpenVZ stress testing.

[akpm@osdl.org: add export]
Signed-off-by: NPavel Emelianov <xemul@openvz.org>
Signed-off-by: NKirill Korotaev <dev@openvz.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Fat is commonly used on removable media.  Mounting with -o flush tells the
FS to write things to disk as quickly as possible.  It is like -o sync, but
much faster (and not as safe).
Signed-off-by: NChris Mason <mason@suse.com>
Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

[assuming BSD security levels are deleted]
The only user of i_security, f_security, s_security fields is SELinux,
however, quite a few security modules are trying to get into kernel.
So, wrap them under CONFIG_SECURITY. Adding config option for each
security field is likely an overkill.

Following Stephen Smalley's suggestion, i_security initialization is
moved to security_inode_alloc() to not clutter core code with ifdefs
and make alloc_inode() codepath tiny little bit smaller and faster.

The user of (highly greppable) struct fown_struct::security field is
still to be found. I've checked every "fown_struct" and every "f_owner"
occurence. Additionally it's removal doesn't break i386 allmodconfig
build.

struct inode, struct file, struct super_block, struct fown_struct
become smaller.

P.S. Combined with two reiserfs inode shrinking patches sent to
linux-fsdevel, I can finally suck 12 reiserfs inodes into one page.

		/proc/slabinfo

	-ext2_inode_cache	388	10
	+ext2_inode_cache	384	10
	-inode_cache		280	14
	+inode_cache		276	14
	-proc_inode_cache	296	13
	+proc_inode_cache	292	13
	-reiser_inode_cache	336	11
	+reiser_inode_cache	332	12 <=
	-shmem_inode_cache	372	10
	+shmem_inode_cache	368	10
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Shrink reiserfs inode more (by 8 bytes) for ACL non-users:

	-reiser_inode_cache     344     11
	+reiser_inode_cache     336     11
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: <reiserfs-dev@namesys.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Shrink reiserfs inode by 12 bytes for xattr non-users (me).

	-reiser_inode_cache     356     11
	+reiser_inode_cache     344     11
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: <reiserfs-dev@namesys.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

ReiserFS does periodic cleanup of old transactions in order to limit the
length of time a journal replay may take after a crash.  Sometimes, writing
metadata from an old (already committed) transaction may require committing
a newer transaction, which also requires writing all data=ordered buffers.
This can cause very long stalls on journal_begin.

This patch makes sure new transactions will not need to be committed before
trying a periodic reclaim of an old transaction.  It is low risk because if
a bad decision is made, it just means a slightly longer journal replay
after a crash.
Signed-off-by: NChris Mason <mason@suse.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

make sure that reiserfs_fsync only triggers barriers when mounted with -o
barrier=flush
Signed-off-by: NChris Mason <mason@suse.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

There's a bug where a UDF_PART_FLAG_READ_ONLY udf partition gets mounted
read-write, then subsequent problems happen; files seem to be able to be
removed, but file creation results in EIO or worse, oops.

EIO is coming from udf_new_block(), which returns EIO if the right flags
aren't set; only UDF_PART_FLAG_READ_ONLY is set in this case.  We probably
s hould not have gotten this far...

Attached patch seems to fix it - and includes a printk to alert the user
that their "rw" mount request has been converted to "ro."

Here's the testcase I used:

[root@magnesium ~]# mkisofs -R -J -udf -o testiso /tmp/
...
Total translation table size: 0
Total rockridge attributes bytes: 342923
Total directory bytes: 382312
Path table size(bytes): 104
Max brk space used 103000
105059 extents written (205 MB)

[root@magnesium ~]# mount -o loop testiso /mnt/test/
[root@magnesium ~]# ls /mnt/test/fsfile
/mnt/test/fsfile
[root@magnesium ~]# rm /mnt/test/fsfile
[root@magnesium ~]# ls /mnt/test/fsfile
ls: /mnt/test/fsfile: No such file or directory
[root@magnesium ~]# touch /mnt/test/fsfile
touch: cannot touch `/mnt/test/fsfile': Input/output error
[root@magnesium tmp]# grep udf /proc/mounts
/dev/loop1 /mnt/test udf rw 0 0

Force readonly mounts of UDF partitions marked as read-only.
Signed-off-by: NEric Sandeen <sandeen@sandeen.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The on-disk data structures from AIX are not known, also the filesystem
layout is not known.  There is a msdos partition signature at the end of
the first block, and the kernel recognizes 3 small (and overlapping)
partitions.  But they are not usable.  Maybe the firmware uses it to find
the bootloader for AIX, but AIX boots also if the first block is cleared.

This is the content of the partition table:
 # dd if=/dev/sdb count=$(( 4 * 16 )) bs=1 skip=$(( 0x1be )) | xxd
0000000: 0000 0000 0000 0000 0000 0000 0000 0000  ................
0000010: 80ff ffff 41ff ffff 1b11 0000 381b 0000  ....A.......8...
0000020: 00ff ffff 41ff ffff 0211 0000 1900 0000  ....A...........
0000030: 80ff ffff 41ff ffff 1b11 0000 381b 0000  ....A.......8...

Handle the whole disk as empty disk.

This fixes also YaST which compares the output from parted (and formerly
fdisk) with /proc/partitions.  fdisk recognizes the AIX label since a long
time, SuSE has a patch for parted to handle the disk label as unknown.

dmesg will look like this:
 sda: [AIX]  unknown partition table

Tested on an IBM B50 with AIX V4.3.3.
Signed-off-by: NOlaf Hering <olh@suse.de>
Cc: Albert Cahalan <acahalan@gmail.com>
Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

In the "operation does permission checking" model used by fuse, chdir
permission is not checked, since there's no chdir method.

For this case set a lookup flag, which will be passed to ->permission(), so
fuse can distinguish it from permission checks for other operations.
Signed-off-by: NMiklos Szeredi <miklos@szeredi.hu>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Some filesystems may want to report different values depending on the path
within the filesystem, i.e.  one mount is actually several filesystems.  This
can be the case for a network filesystem exported by an unprivileged server
(e.g.  sshfs).

This is now possible, thanks to David Howells "VFS: Permit filesystem to
perform statfs with a known root dentry" patch.

This change is backward compatible, so no need to change interface version.
Signed-off-by: NMiklos Szeredi <miklos@szeredi.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Files supported by fs/proc/base.c, i.e.  /proc/<pid>/*, are not capable of
meeting the validity checks in ELF load_elf_*() handling because they have
no mmap handler which is required by ELF.  In order to stop a.out
executables being used as part of an exploit attack against /proc-related
vulnerabilities, we make a.out executables depend on ->mmap() existing.
Signed-off-by: NEugene Teo <eteo@redhat.com>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

grab_super gets called with sb_lock held, and releases it.  Add a lock
annotation to this function so that sparse can check callers for lock
pairing, and so that sparse will not complain about this function since it
intentionally uses the lock in this manner.
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>