Missing acct_update_integrals() and update_mem_hiwater() calls
compared to it's native counterpart.
Signed-off-by: David S. Miller <davem@davemloft.net>
Noted by David Miller:
"The bug is that free_fd_array() takes a "num" argument, but when
calling it from __free_fdtable() we're instead passing in the size in
bytes (ie. "num * sizeof(struct file *)")."
Yes it is a bug. I think I messed it up while merging newer
changes with an older version where I was using size in bytes
to optimize.
Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Pavel Emelianov and Kirill Korotaev observe that fs and arch users of
security_vm_enough_memory tend to forget to vm_unacct_memory when a
failure occurs further down (typically in setup_arg_pages variants).
These are all users of insert_vm_struct, and that reservation will only
be unaccounted on exit if the vma is marked VM_ACCOUNT: which in some
cases it is (hidden inside VM_STACK_FLAGS) and in some cases it isn't.
So x86_64 32-bit and ppc64 vDSO ELFs have been leaking memory into
Committed_AS each time they're run. But don't add VM_ACCOUNT to them,
it's inappropriate to reserve against the very unlikely case that gdb
be used to COW a vDSO page - we ought to do something about that in
do_wp_page, but there are yet other inconsistencies to be resolved.
The safe and economical way to fix this is to let insert_vm_struct do
the security_vm_enough_memory check when it finds VM_ACCOUNT is set.
And the MIPS irix_brk has been calling security_vm_enough_memory before
calling do_brk which repeats it, doubly accounting and so also leaking.
Remove that, and all the fs and arch calls to security_vm_enough_memory:
give it a less misleading name later on.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-Off-By: Kirill Korotaev <dev@sw.ru>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It turns out that the BUG_ON() in fs/exec.c: de_thread() is unreliable
and can trigger due to the test itself being racy.
de_thread() does
while (atomic_read(&sig->count) > count) {
}
.....
.....
BUG_ON(!thread_group_empty(current));
but release_task does
write_lock_irq(&tasklist_lock)
__exit_signal
(this is where atomic_dec(&sig->count) is run)
__exit_sighand
__unhash_process
takes write lock on tasklist_lock
remove itself out of PIDTYPE_TGID list
write_unlock_irq(&tasklist_lock)
so there's a clear (although small) window between the
atomic_dec(&sig->count) and the actual PIDTYPE_TGID unhashing of the
thread.
And actually there is no need for all threads to have exited at this
point, so we simply kill the BUG_ON.
Big thanks to Marc Lehmann who provided the test-case.
Fixes Bug 5170 (http://bugme.osdl.org/show_bug.cgi?id=5170)
Signed-off-by: Alexander Nyberg <alexn@telia.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Andrew Morton <akpm@osdl.org>
Cc: Ingo Molnar <mingo@elte.hu>
Acked-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We could try to unlock the state lock here without having first locked it.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In the case of a lock which introduces a new lockowner, the openowner's
sequence id should be incremented, even when the operation fails, if the
error is a sequence-id-mutating error. The current code fails to do that
in some cases. Fix this by using the same sequence-id-incrementing
mechanism that all other such operations use.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It seems more natural to move the setting of the replay_owner into the
relevant procedure instead of doing it in nfsv4_proc_compound.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Demote some printk's that look like they could be triggered by non-buggy
clients to dprintk's. (For example, stale clientid's are normal
occurrences on reboot, and on a server with a lot of clients these messages
could become annoying.)
Also remove some redundant dprintk's (e.g. no need for both STALE_CLIENTID
and its callers to do dprintks).
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
reiserfs should use mark_inode_dirty during reiserfs_file_write and
reiserfs_commit_write. This makes sure the inode is properly flagged as
dirty, which is used during O_SYNC to decide when to trigger log commits.
This patch also removes the O_SYNC check from reiserfs_commit_write, since
that gets dealt with properly at higher layers once we start using
mark_inode_dirty.
Thanks to Hifumi Hisashi <hifumi.hisashi@lab.ntt.co.jp> for catching this.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When open(O_CREAT) is called and the error, ENFILE, is returned, the file
may be created anyway. This is counter intuitive, against the SUS V3
specification, and may cause applications to misbehave if they are not
coded correctly to handle this semantic. The SUS V3 specification
explicitly states "No files shall be created or modified if the function
returns -1.".
The error, ENFILE, is used to indicate the system wide open file table is
full and no more file structs can be allocated.
This is due to an ordering problem. The entry in the directory is created
before the file struct is allocated. If the allocation for the file struct
fails, then the system call must return an error, but the directory entry
was already created and can not be safely removed.
The solution to this situation is relatively easy. The file struct should
be allocated before the directory entry is created. If the allocation
fails, then the error can be returned directly. If the creation of the
directory entry fails, then the file struct can be easily freed.
Signed-off-by: Peter Staubach <staubach@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
an octal number to conform to how chmod(1) works, too. Thanks to
Giuseppe Bilotta and Horst von Brand for pointing out the errors of
my ways.
Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
Clash due to new delete_inode behavior (the filesystem now needs to do
the truncate_inode_pages() call itself).
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Use schedule_timeout_{,un}interruptible() instead of
set_current_state()/schedule_timeout() to reduce kernel size. Also use helper
functions to convert between human time units and jiffies rather than constant
HZ division to avoid rounding errors.
Signed-off-by: Nishanth Aravamudan <nacc@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Every file should #include the header with the prototypes of the global
functions it is offering.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
First one is list_for_each_entry (thanks maks), second 2 list_for_each_safe.
Signed-off-by: Maximilian Attems <janitor@sternwelten.at>
Signed-off-by: Domen Puncer <domen@coderock.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Make code more readable with list_for_each_entry.
Signed-off-by: Maximilian Attems <janitor@sternwelten.at>
Signed-off-by: Domen Puncer <domen@coderock.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Use list_for_each_entry to make code more readable.
Signed-off-by: Maximilian Attems <janitor@sternwelten.at>
Signed-off-by: Domen Puncer <domen@coderock.org>
Cc: <jffs-dev@axis.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch implements a task state bit (TASK_NONINTERACTIVE), which can be
used by blocking points to mark the task's wait as "non-interactive". This
does not mean the task will be considered a CPU-hog - the wait will simply
not have an effect on the waiting task's priority - positive or negative
alike. Right now only pipe_wait() will make use of it, because it's a
common source of not-so-interactive waits (kernel compilation jobs, etc.).
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch (written by me and also containing many suggestions of Arjan van
de Ven) does a major cleanup of the spinlock code. It does the following
things:
- consolidates and enhances the spinlock/rwlock debugging code
- simplifies the asm/spinlock.h files
- encapsulates the raw spinlock type and moves generic spinlock
features (such as ->break_lock) into the generic code.
- cleans up the spinlock code hierarchy to get rid of the spaghetti.
Most notably there's now only a single variant of the debugging code,
located in lib/spinlock_debug.c. (previously we had one SMP debugging
variant per architecture, plus a separate generic one for UP builds)
Also, i've enhanced the rwlock debugging facility, it will now track
write-owners. There is new spinlock-owner/CPU-tracking on SMP builds too.
All locks have lockup detection now, which will work for both soft and hard
spin/rwlock lockups.
The arch-level include files now only contain the minimally necessary
subset of the spinlock code - all the rest that can be generalized now
lives in the generic headers:
include/asm-i386/spinlock_types.h | 16
include/asm-x86_64/spinlock_types.h | 16
I have also split up the various spinlock variants into separate files,
making it easier to see which does what. The new layout is:
SMP | UP
----------------------------|-----------------------------------
asm/spinlock_types_smp.h | linux/spinlock_types_up.h
linux/spinlock_types.h | linux/spinlock_types.h
asm/spinlock_smp.h | linux/spinlock_up.h
linux/spinlock_api_smp.h | linux/spinlock_api_up.h
linux/spinlock.h | linux/spinlock.h
/*
* here's the role of the various spinlock/rwlock related include files:
*
* on SMP builds:
*
* asm/spinlock_types.h: contains the raw_spinlock_t/raw_rwlock_t and the
* initializers
*
* linux/spinlock_types.h:
* defines the generic type and initializers
*
* asm/spinlock.h: contains the __raw_spin_*()/etc. lowlevel
* implementations, mostly inline assembly code
*
* (also included on UP-debug builds:)
*
* linux/spinlock_api_smp.h:
* contains the prototypes for the _spin_*() APIs.
*
* linux/spinlock.h: builds the final spin_*() APIs.
*
* on UP builds:
*
* linux/spinlock_type_up.h:
* contains the generic, simplified UP spinlock type.
* (which is an empty structure on non-debug builds)
*
* linux/spinlock_types.h:
* defines the generic type and initializers
*
* linux/spinlock_up.h:
* contains the __raw_spin_*()/etc. version of UP
* builds. (which are NOPs on non-debug, non-preempt
* builds)
*
* (included on UP-non-debug builds:)
*
* linux/spinlock_api_up.h:
* builds the _spin_*() APIs.
*
* linux/spinlock.h: builds the final spin_*() APIs.
*/
All SMP and UP architectures are converted by this patch.
arm, i386, ia64, ppc, ppc64, s390/s390x, x64 was build-tested via
crosscompilers. m32r, mips, sh, sparc, have not been tested yet, but should
be mostly fine.
From: Grant Grundler <grundler@parisc-linux.org>
Booted and lightly tested on a500-44 (64-bit, SMP kernel, dual CPU).
Builds 32-bit SMP kernel (not booted or tested). I did not try to build
non-SMP kernels. That should be trivial to fix up later if necessary.
I converted bit ops atomic_hash lock to raw_spinlock_t. Doing so avoids
some ugly nesting of linux/*.h and asm/*.h files. Those particular locks
are well tested and contained entirely inside arch specific code. I do NOT
expect any new issues to arise with them.
If someone does ever need to use debug/metrics with them, then they will
need to unravel this hairball between spinlocks, atomic ops, and bit ops
that exist only because parisc has exactly one atomic instruction: LDCW
(load and clear word).
From: "Luck, Tony" <tony.luck@intel.com>
ia64 fix
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjanv@infradead.org>
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Cc: Matthew Wilcox <willy@debian.org>
Signed-off-by: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Mikael Pettersson <mikpe@csd.uu.se>
Signed-off-by: Benoit Boissinot <benoit.boissinot@ens-lyon.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Andrew lost this in patch reject resolution, and never noticed, since
the compat code isn't in use on x86.
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This converts the final 20 DEFINE_SPINLOCK holdouts. (another 580 places
are already using DEFINE_SPINLOCK). Build tested on x86.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes ability to interrupt and restart operations while there
hasn't been any side-effect.
The reason: applications. There are some apps it seems that generate
signals at a fast rate. This means, that if the operation cannot make
enough progress between two signals, it will be restarted for ever. This
bug actually manifested itself with 'krusader' trying to open a file for
writing under sshfs. Thanks to Eduard Czimbalmos for the report.
The problem can be solved just by making open() uninterruptible, because in
this case it was the truncate operation that slowed down the progress. But
it's better to solve this by simply not allowing interrupts at all (except
SIGKILL), because applications don't expect file operations to be
interruptible anyway. As an added bonus the code is simplified somewhat.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds a new FSYNCDIR request, which is sent when fsync is called
on directories. This operation is available in libfuse 2.3-pre1 or
greater.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Don't change mtime/ctime/atime to local time on read/write. Rather invalidate
file attributes, so next stat() will force a GETATTR call. Bug reported by
Ben Grimm.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Make data caching behavior selectable on a per-open basis instead of
per-mount. Compatibility for the old mount options 'kernel_cache' and
'direct_io' is retained in the userspace library (version 2.4.0-pre1 or
later).
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes a long lasting "hack" in FUSE, which used a separate
channel (a file descriptor refering to a disk-file) to transfer directory
contents from userspace to the kernel.
The patch adds three new operations (OPENDIR, READDIR, RELEASEDIR), which
have semantics and implementation exactly maching the respective file
operations (OPEN, READ, RELEASE).
This simplifies the directory reading code. Also disk space is not
necessary, which can be important in embedded systems.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds support for the "direct_io" mount option of FUSE.
When this mount option is specified, the page cache is bypassed for
read and write operations. This is useful for example, if the
filesystem doesn't know the size of files before reading them, or when
any kind of caching is harmful.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Check for the presence of all mandatory mount options.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch tightens the check for allowing processes to access non-privileged
mounts. The rational is that the filesystem implementation can control the
behavior or get otherwise unavailable information of the filesystem user. If
the filesystem user process has the same uid, gid, and is not suid or sgid
application, then access is safe. Otherwise access is not allowed unless the
"allow_other" mount option is given (for which policy is controlled by the
userspace mount utility).
Thanks to everyone linux-fsdevel, especially Martin Mares who helped uncover
problems with the previous approach.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds readpages support to FUSE.
With the help of the readpages() operation multiple reads are bundled
together and sent as a single request to userspace. This can improve
reading performace.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds the extended attribute operations to FUSE.
The following operations are added:
o getxattr
o setxattr
o listxattr
o removexattr
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds miscellaneous mount options to the FUSE filesystem.
The following mount options are added:
o default_permissions: check permissions with generic_permission()
o allow_other: allow other users to access files
o allow_root: allow root to access files
o kernel_cache: don't invalidate page cache on open
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds the file operations of FUSE.
The following operations are added:
o open
o flush
o release
o fsync
o readpage
o commit_write
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds the write filesystem operations of FUSE.
The following operations are added:
o setattr
o symlink
o mknod
o mkdir
o create
o unlink
o rmdir
o rename
o link
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds the read-only filesystem operations of FUSE.
This contains the following files:
o dir.c
- directory, symlink and file-inode operations
The following operations are added:
o lookup
o getattr
o readlink
o follow_link
o directory open
o readdir
o directory release
o permission
o dentry revalidate
o statfs
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This adds the FUSE device handling functions.
This contains the following files:
o dev.c
- fuse device operations (read, write, release, poll)
- registers misc device
- support for sending requests to userspace
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds FUSE core.
This contains the following files:
o inode.c
- superblock operations (alloc_inode, destroy_inode, read_inode,
clear_inode, put_super, show_options)
- registers FUSE filesystem
o fuse_i.h
- private header file
Requirements
============
The most important difference between orinary filesystems and FUSE is
the fact, that the filesystem data/metadata is provided by a userspace
process run with the privileges of the mount "owner" instead of the
kernel, or some remote entity usually running with elevated
privileges.
The security implication of this is that a non-privileged user must
not be able to use this capability to compromise the system. Obvious
requirements arising from this are:
- mount owner should not be able to get elevated privileges with the
help of the mounted filesystem
- mount owner should not be able to induce undesired behavior in
other users' or the super user's processes
- mount owner should not get illegitimate access to information from
other users' and the super user's processes
These are currently ensured with the following constraints:
1) mount is only allowed to directory or file which the mount owner
can modify without limitation (write access + no sticky bit for
directories)
2) nosuid,nodev mount options are forced
3) any process running with fsuid different from the owner is denied
all access to the filesystem
1) and 2) are ensured by the "fusermount" mount utility which is a
setuid root application doing the actual mount operation.
3) is ensured by a check in the permission() method in kernel
I started thinking about doing 3) in a different way because Christoph
H. made a big deal out of it, saying that FUSE is unacceptable into
mainline in this form.
The suggested use of private namespaces would be OK, but in their
current form have many limitations that make their use impractical (as
discussed in this thread).
Suggested improvements that would address these limitations:
- implement shared subtrees
- allow a process to join an existing namespace (make namespaces
first-class objects)
- implement the namespace creation/joining in a PAM module
With all that in place the check of owner against current->fsuid may
be removed from the FUSE kernel module, without compromising the
security requirements.
Suid programs still interesting questions, since they get access even
to the private namespace causing some information leak (exact
order/timing of filesystem operations performed), giving some
ptrace-like capabilities to unprivileged users. BTW this problem is
not strictly limited to the namespace approach, since suid programs
setting fsuid and accessing users' files will succeed with the current
approach too.
Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch attempts to do a better job of cleaning up after detecting errors
on the transport. This should also improve error reporting on broken
connections to servers.
Signed-off-by: Latchesar Ionkov <lucho@ionkov.net>
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
LANL reported some issues with random crashes during mount of legacy protocol
servers (9P2000 versus 9P2000.u) -- crash was always happening in readlink
(which should never happen in legacy mode). Added some sanity conditionals to
the get_inode code which should prevent the errors LANL was seeing. Code
tested benign through regression.
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Cleanup code in v9fs vfs_inode as suggested by Alexey Dobriyan. Did some
major revamping of the v9fs setattr code to remove unnecessary allocations and
clean up some dead-code.
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Change magic error numbers to system defined constants in v9fs error.h As
suggested by Jan-Benedict Glaw.
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This part of the patch contains debug and other misc routines.
Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
David S. Miller