Race scenario:

thread A			thread B

p9_write_work()                p9_fd_request()

if (list_empty
  (&m->unsent_req_list))
  ...

                               spin_lock(&client->lock);
                               req->status = REQ_STATUS_UNSENT;
                               list_add_tail(..., &m->unsent_req_list);
                               spin_unlock(&client->lock);
                               ....
                               if (n & POLLOUT &&
                               !test_and_set_bit(Wworksched, &m->wsched)
                               schedule_work(&m->wq);
                               --> not done because Wworksched is set

  clear_bit(Wworksched, &m->wsched);
  return;

--> nobody will take care of sending the new request.

This is not very likely to happen though, because p9_write_work()
being called with an empty unsent_req_list is not frequent.
But this also means that taking the lock earlier will not be costly.
Signed-off-by: NSimon Derr <simon.derr@bull.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Both modular callers of sock_map_fd() had been buggy; sctp one leaks
descriptor and file if copy_to_user() fails, 9p one shouldn't be
exposing file in the descriptor table at all.

Switch both to sock_alloc_file(), export it, unexport sock_map_fd() and
make it static.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

See previous commit about p9_read_work() for details.

This fixes a similar race between p9_write_work() and p9_poll_mux()
Signed-off-by: NSimon Derr <simon.derr@bull.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

At the end of p9_write_work() we want to test if there is still data to send.
This means:
- either the current request still has data to send (wsize != 0)
- or there are requests in the unsent queue
Signed-off-by: NSimon Derr <simon.derr@bull.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Race scenario between p9_read_work() and p9_poll_mux()

Data arrive, Rworksched is set, p9_read_work() is called.

thread A                                thread B

                                        p9_read_work()
                                                .
                                        reads data
                                                .
                                        checks if new data ready. No.
                                                .
                                        gets preempted
                                                .
More data arrive, p9_poll_mux() is called.      .
                                                .
                                                .
p9_poll_mux()                                   .
                                                .
if (!test_and_set_bit(Rworksched,               .
                      &m->wsched)) {            .
  schedule_work(&m->rq);                        .
}                                               .
                                                .
-> does not schedule work because               .
   Rworksched is set                            .
                                                .
                                        clear_bit(Rworksched, &m->wsched);
                                        return;

No work has been scheduled, and yet data are waiting.

Currently p9_read_work() checks if there is data to read,
and if not, it clears Rworksched.

I think it should clear Rworksched first, and then check if there is data to read.
Signed-off-by: NSimon Derr <simon.derr@bull.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

While working on a modified server I had the Linux clients crash
a few times. This lead me to find this:

Some error codes are directly extracted from the server replies.
A malformed server reply could contain an invalid error code, with a
very large value. If this value is then passed to ERR_PTR() it will
not be properly detected as an error code by IS_ERR() and as a result
the kernel will dereference an invalid pointer.

This patch tries to avoid this.
Signed-off-by: NSimon Derr <simon.derr@bull.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

flush[_delayed]_work_sync() are now spurious.  Mark them deprecated
and convert all users to flush[_delayed]_work().

If you're cc'd and wondering what's going on: Now all workqueues are
non-reentrant and the regular flushes guarantee that the work item is
not pending or running on any CPU on return, so there's no reason to
use the sync flushes at all and they're going away.

This patch doesn't make any functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Mattia Dongili <malattia@linux.it>
Cc: Kent Yoder <key@linux.vnet.ibm.com>
Cc: David Airlie <airlied@linux.ie>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Karsten Keil <isdn@linux-pingi.de>
Cc: Bryan Wu <bryan.wu@canonical.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-wireless@vger.kernel.org
Cc: Anton Vorontsov <cbou@mail.ru>
Cc: Sangbeom Kim <sbkim73@samsung.com>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Eric Van Hensbergen <ericvh@gmail.com>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Petr Vandrovec <petr@vandrovec.name>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Avi Kivity <avi@redhat.com> 

Fix incorrect start markers, wrapped summary lines, missing section
breaks, incorrect separators, and some name mismatches.
Signed-off-by: NBen Hutchings <bhutchings@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

I don't think we're actually likely to hit this limit but if we do
then the comparison should be done as size_t.  The original code
is equivalent to:
        len = strlen(sptr) % USHRT_MAX;
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The BUG_ON() in pack_sg_list() would get triggered only one time after we've
corrupted some memory by sg_set_buf() into an invalid sg buffer.

I'm still working on figuring out why I manage to trigger that bug...
Signed-off-by: NSasha Levin <levinsasha928@gmail.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

A recent commit that removed unnecessary casts of pointers
to the same type uncovered a missing __force cast.

Add it.

Reported by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adding casts of objects to the same type is unnecessary
and confusing for a human reader.

For example, this cast:

	int y;
	int *p = (int *)&y;

I used the coccinelle script below to find and remove these
unnecessary casts.  I manually removed the conversions this
script produces of casts with __force and __user.

@@
type T;
T *p;
@@

-	(T *)p
+	p
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a virtio_9p pci device is being removed, we should close down any
active channels and free up resources, we're not supposed to BUG() if there's
still an open channel since it's a valid case when removing the PCI device.

Otherwise, removing the PCI device with an open channel would cause the
following BUG():

[ 1184.671416] ------------[ cut here ]------------
[ 1184.672057] kernel BUG at net/9p/trans_virtio.c:618!
[ 1184.672057] invalid opcode: 0000 [#1] PREEMPT SMP
[ 1184.672057] CPU 3
[ 1184.672057] Pid: 5, comm: kworker/u:0 Tainted: G        W    3.4.0-rc2-next-20120413-sasha-dirty #76
[ 1184.672057] RIP: 0010:[<ffffffff825c9116>]  [<ffffffff825c9116>] p9_virtio_remove+0x16/0x90
[ 1184.672057] RSP: 0018:ffff88000d653ac0  EFLAGS: 00010202
[ 1184.672057] RAX: ffffffff836bfb40 RBX: ffff88000c9b2148 RCX: ffff88000d658978
[ 1184.672057] RDX: 0000000000000006 RSI: 0000000000000000 RDI: ffff880028868000
[ 1184.672057] RBP: ffff88000d653ad0 R08: 0000000000000000 R09: 0000000000000000
[ 1184.672057] R10: 0000000000000000 R11: 0000000000000001 R12: ffff880028868000
[ 1184.672057] R13: ffffffff835aa7c0 R14: ffff880041630000 R15: ffff88000d653da0
[ 1184.672057] FS:  0000000000000000(0000) GS:ffff880035a00000(0000) knlGS:0000000000000000
[ 1184.672057] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 1184.672057] CR2: 0000000001181000 CR3: 000000000eba1000 CR4: 00000000000406e0
[ 1184.672057] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
x000000000117a190 *[ 1184.672057] DR3: 00000000000000**
00 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 1184.672057] Process kworker/u:0 (pid: 5, threadinfo ffff88000d652000, task ffff88000d658000)
[ 1184.672057] Stack:
[ 1184.672057]  ffff880028868000 ffffffff836bfb40 ffff88000d653af0 ffffffff8193661b
[ 1184.672057]  ffff880028868008 ffffffff836bfb40 ffff88000d653b10 ffffffff81af1c81
[ 1184.672057]  ffff880028868068 ffff880028868008 ffff88000d653b30 ffffffff81af257a
[ 1184.795301] Call Trace:
[ 1184.795301]  [<ffffffff8193661b>] virtio_dev_remove+0x1b/0x60
[ 1184.795301]  [<ffffffff81af1c81>] __device_release_driver+0x81/0xd0
[ 1184.795301]  [<ffffffff81af257a>] device_release_driver+0x2a/0x40
[ 1184.795301]  [<ffffffff81af0d48>] bus_remove_device+0x138/0x150
[ 1184.795301]  [<ffffffff81aef08d>] device_del+0x14d/0x1b0
[ 1184.795301]  [<ffffffff81aef138>] device_unregister+0x48/0x60
[ 1184.795301]  [<ffffffff8193694d>] unregister_virtio_device+0xd/0x10
[ 1184.795301]  [<ffffffff8265fc74>] virtio_pci_remove+0x2a/0x6c
[ 1184.795301]  [<ffffffff818a95ad>] pci_device_remove+0x4d/0x110
[ 1184.795301]  [<ffffffff81af1c81>] __device_release_driver+0x81/0xd0
[ 1184.795301]  [<ffffffff81af257a>] device_release_driver+0x2a/0x40
[ 1184.795301]  [<ffffffff81af0d48>] bus_remove_device+0x138/0x150
[ 1184.795301]  [<ffffffff81aef08d>] device_del+0x14d/0x1b0
[ 1184.795301]  [<ffffffff81aef138>] device_unregister+0x48/0x60
[ 1184.795301]  [<ffffffff818a36fa>] pci_stop_bus_device+0x6a/0x90
[ 1184.795301]  [<ffffffff818a3791>] pci_stop_and_remove_bus_device+0x11/0x20
[ 1184.795301]  [<ffffffff818c21d9>] remove_callback+0x9/0x10
[ 1184.795301]  [<ffffffff81252d91>] sysfs_schedule_callback_work+0x21/0x60
[ 1184.795301]  [<ffffffff810cb1a1>] process_one_work+0x281/0x430
[ 1184.795301]  [<ffffffff810cb140>] ? process_one_work+0x220/0x430
[ 1184.795301]  [<ffffffff81252d70>] ? sysfs_read_file+0x1c0/0x1c0
[ 1184.795301]  [<ffffffff810cc613>] worker_thread+0x1f3/0x320
[ 1184.795301]  [<ffffffff810cc420>] ? manage_workers.clone.13+0x130/0x130
[ 1184.795301]  [<ffffffff810d30b2>] kthread+0xb2/0xc0
[ 1184.795301]  [<ffffffff826783f4>] kernel_thread_helper+0x4/0x10
[ 1184.795301]  [<ffffffff810deb18>] ? finish_task_switch+0x78/0xf0
[ 1184.795301]  [<ffffffff82676574>] ? retint_restore_args+0x13/0x13
[ 1184.795301]  [<ffffffff810d3000>] ? kthread_flush_work_fn+0x10/0x10
[ 1184.795301]  [<ffffffff826783f0>] ? gs_change+0x13/0x13
[ 1184.795301] Code: c1 9e 0a 00 48 83 c4 08 5b c9 c3 66 2e 0f 1f 84 00 00 00 00 00 55 48 89 e5 41 54 49 89 fc 53 48 8b 9f a8 04 00 00 80 3b 00 74 0a <0f> 0b 0f 1f 84 00 00 00 00 00 48 8b 87 88 04 00 00 ff 50 30 31
[ 1184.795301] RIP  [<ffffffff825c9116>] p9_virtio_remove+0x16/0x90
[ 1184.795301]  RSP <ffff88000d653ac0>
[ 1184.952618] ---[ end trace a307b3ed40206b4c ]---
Signed-off-by: NSasha Levin <levinsasha928@gmail.com>
Signed-off-by: NRusty Russell <rusty@rustcorp.com.au>

Use of "unsigned int" is preferred to bare "unsigned" in net tree.
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a Tclunk or Tremove request is flushed, the fid is not freed on the
server.

p9_client_clunk() should retry once on interrupt, then if interrupted
again, leak the fid for the duration of the connection.

p9_client_remove() should call p9_client_clunk() on interrupt
instead of unconditionally destroying the fid.
Signed-off-by: NJim Garlick <garlick@llnl.gov>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

When a signal is received while sending a Tflush, the client,
which has recursed into p9_client_rpc() while sending another request,
should wait for Rflush as long as the transport is still up.
Signed-off-by: NJim Garlick <garlick@llnl.gov>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Remove wrapper functions. This makes the allocation type explicit in
all callers; I used GPF_KERNEL where it seemed obvious, left it at
GFP_ATOMIC otherwise.
Signed-off-by: NRusty Russell <rusty@rustcorp.com.au>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

Reduce object size by deduplicating formats.

Use vsprintf extension %pV.
Rename P9_DPRINTK uses to p9_debug, align arguments.
Add function for _p9_debug and macro to add __func__.
Add missing "\n"s to p9_debug uses.
Remove embedded function names as p9_debug adds it.
Remove P9_EPRINTK macro and convert use to pr_<level>.
Add and use pr_fmt and pr_<level>.

$ size fs/9p/built-in.o*
   text	   data	    bss	    dec	    hex	filename
  62133	    984	  16000	  79117	  1350d	fs/9p/built-in.o.new
  67342	    984	  16928	  85254	  14d06	fs/9p/built-in.o.old
$ size net/9p/built-in.o*
   text	   data	    bss	    dec	    hex	filename
  88792	   4148	  22024	 114964	  1c114	net/9p/built-in.o.new
  94072	   4148	  23232	 121452	  1da6c	net/9p/built-in.o.old
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

This helps in more control over debugging.
root@qemu-img-64:~# ls /pass/123
ls: cannot access /pass/123: No such file or directory
root@qemu-img-64:~# cat /sys/kernel/debug/tracing/trace
# tracer: nop
#
#           TASK-PID    CPU#    TIMESTAMP  FUNCTION
#              | |       |          |         |
              ls-1536  [001]    70.928584: 9p_protocol_dump: clnt 18446612132784021504 P9_TWALK(tag = 1)
000: 16 00 00 00 6e 01 00 01 00 00 00 02 00 00 00 01
010: 00 03 00 31 32 33 00 00 00 ff ff ff ff 00 00 00

              ls-1536  [001]    70.928587: <stack trace>
 => trace_9p_protocol_dump
 => p9pdu_finalize
 => p9_client_rpc
 => p9_client_walk
 => v9fs_vfs_lookup
 => d_alloc_and_lookup
 => walk_component
 => path_lookupat
              ls-1536  [000]    70.929696: 9p_protocol_dump: clnt 18446612132784021504 P9_RLERROR(tag = 1)
000: 0b 00 00 00 07 01 00 02 00 00 00 4e 03 00 02 00
010: 00 00 00 00 03 00 02 00 00 00 00 00 ff 43 00 00

              ls-1536  [000]    70.929697: <stack trace>
 => trace_9p_protocol_dump
 => p9_client_rpc
 => p9_client_walk
 => v9fs_vfs_lookup
 => d_alloc_and_lookup
 => walk_component
 => path_lookupat
 => do_path_lookup
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Without this msize=4294967295 will result in a crash
Signed-off-by: NDan Carpenter <error27@gmail.com>
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Instead of saying all integer argument option should be listed in the beginning
move integer parsing to each option type.
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

We dereferenced "req->tc" and "req->rc" before checking for NULL.
Signed-off-by: NDan Carpenter <error27@gmail.com>
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

* remove lot of update to different data structure
* add a seperate callback for zero copy request.
* above makes non zero copy code path simpler
* remove conditionalizing TREAD/TREADDIR/TWRITE in the zero copy path
* Fix the dotu p9_check_errors with zero copy. Add sufficient doc around
* Add support for both in and output buffers in zero copy callback
* pin and unpin pages in the same context
* use helpers instead of defining page offset and rest of page ourself
* Fix mem leak in p9_check_errors
* Remove 'E' and 'F' in p9pdu_vwritef
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

With msize equal to 512K (PAGE_SIZE * VIRTQUEUE_NUM), we hit multiple
crashes. This patch fix those.
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

msize represents the maximum PDU size that includes P9_IOHDRSZ.

Signed-off-by: Venkateswararao Jujjuri "<jvrao@linux.vnet.ibm.com>
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

unlinkat - Remove a directory entry

size[4] Tunlinkat tag[2] dirfid[4] name[s] flag[4]
size[4] Runlinkat tag[2]

older Tremove have the below request format

size[4] Tremove tag[2] fid[4]

The remove message is used to remove a directory entry either file or directory
The remove opreation is actually a directory opertation and should ideally have
dirfid, if not we cannot represent the fid on server with anything other than
name. We will have to derive the directory name from fid in the Tremove request.

NOTE: The operation doesn't clunk the unlink fid.
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

renameat - change name of file or directory

size[4] Trenameat tag[2] olddirfid[4] oldname[s] newdirfid[4] newname[s]
size[4] Rrenameat tag[2]

older Trename have the below request format

size[4] Trename tag[2] fid[4] newdirfid[4] name[s]

The rename message is used to change the name of a file, possibly moving it
to a new directory. The rename opreation is actually a directory opertation
and should ideally have olddirfid, if not we cannot represent the fid on server
with anything other than name. We will have to derive the old directory name
from fid in the Trename request.
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Signed-off-by: NPrem Karat <prem.karat@linux.vnet.ibm.com>
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

free the fid even in case of failed clunk.
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Signed-off-by: Venkateswararao Jujjuri "<jvrao@linux.vnet.ibm.com>
Reviewed-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Signed-off-by: Venkateswararao Jujjuri "<jvrao@linux.vnet.ibm.com>
Reviewed-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Switch to generic kernel hexdump library and cleanup macros to
be more consistent with the way we do normal debug prints.
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

There was a BUG_ON to protect against a bad id which could be dealt with
more gracefully.
Reported-by: NNatalie Orlin <norlin@us.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

The RDMA CM currently infers the QP type from the port space selected
by the user.  In the future (eg with RDMA_PS_IB or XRC), there may not
be a 1-1 correspondence between port space and QP type.  For netlink
export of RDMA CM state, we want to export the QP type to userspace,
so it is cleaner to explicitly associate a QP type to an ID.

Modify rdma_create_id() to allow the user to specify the QP type, and
use it to make our selections of datagram versus connected mode.
Signed-off-by: NSean Hefty <sean.hefty@intel.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Teach 9p filesystem to work in container with non-default network namespace.
(Note: I also patched the unix domain socket code but don't have a test case
for that.  It's the same fix, I just don't have a server for it...)

To test, run diod server (http://code.google.com/p/diod):
  diod -n -f -L stderr -l 172.23.255.1:9999 -c /dev/null -e /root
and then mount like so:
  mount -t 9p -o port=9999,aname=/root,version=9p2000.L 172.23.255.1 /mnt

A container test environment is described at http://landley.net/lxcSigned-off-by: NRob Landley <rob@landley.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

We need to return -1 on error. Also handle error properly
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

The 9p client is currently undergoing regular regresssion and
stress testing as a by-product of the virtfs work.  I think its
finally time to take off the experimental tags from the well-tested
code paths.
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Typo fixes and minor cleanups for v9fs
Signed-off-by: NRob Landley <rob@landley.net>
Reviewed-by: NVenkateswararao Jujjuri (JV) <jvrao@linux.vnet.ibm.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

As on Jeopardy, my question is in the form of a patch:  Does this have
some special meaning, or is it an accident?  (I looked at other
filesystems but they didn't bother having doc entries for their
init/exit function that I could find.)
Signed-off-by: NRob Landley <rob@landley.net>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>

Probably just cut and pasted from the other parse_opts() implementations
in the 9p sources.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>