2005-04-17 00:20:36 +02:00
|
|
|
/*
|
|
|
|
* Network block device - make block devices work over TCP
|
|
|
|
*
|
|
|
|
* Note that you can not swap over this thing, yet. Seems to work but
|
|
|
|
* deadlocks sometimes - you can not swap over TCP in general.
|
|
|
|
*
|
2010-07-18 14:27:13 +02:00
|
|
|
* Copyright 1997-2000, 2008 Pavel Machek <pavel@ucw.cz>
|
2005-04-17 00:20:36 +02:00
|
|
|
* Parts copyright 2001 Steven Whitehouse <steve@chygwyn.com>
|
|
|
|
*
|
2006-06-25 14:47:42 +02:00
|
|
|
* This file is released under GPLv2 or later.
|
2005-04-17 00:20:36 +02:00
|
|
|
*
|
2006-06-25 14:47:42 +02:00
|
|
|
* (part of code stolen from loop.c)
|
2005-04-17 00:20:36 +02:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/major.h>
|
|
|
|
|
|
|
|
#include <linux/blkdev.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/bio.h>
|
|
|
|
#include <linux/stat.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/file.h>
|
|
|
|
#include <linux/ioctl.h>
|
2010-06-02 14:28:52 +02:00
|
|
|
#include <linux/mutex.h>
|
2006-01-06 09:09:47 +01:00
|
|
|
#include <linux/compiler.h>
|
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/kernel.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
#include <net/sock.h>
|
2007-11-13 03:10:39 +01:00
|
|
|
#include <linux/net.h>
|
2008-04-29 10:02:46 +02:00
|
|
|
#include <linux/kthread.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
#include <asm/uaccess.h>
|
|
|
|
#include <asm/types.h>
|
|
|
|
|
|
|
|
#include <linux/nbd.h>
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
#define NBD_MAGIC 0x68797548
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
#ifdef NDEBUG
|
|
|
|
#define dprintk(flags, fmt...)
|
|
|
|
#else /* NDEBUG */
|
|
|
|
#define dprintk(flags, fmt...) do { \
|
|
|
|
if (debugflags & (flags)) printk(KERN_DEBUG fmt); \
|
|
|
|
} while (0)
|
|
|
|
#define DBG_IOCTL 0x0004
|
|
|
|
#define DBG_INIT 0x0010
|
|
|
|
#define DBG_EXIT 0x0020
|
|
|
|
#define DBG_BLKDEV 0x0100
|
|
|
|
#define DBG_RX 0x0200
|
|
|
|
#define DBG_TX 0x0400
|
|
|
|
static unsigned int debugflags;
|
|
|
|
#endif /* NDEBUG */
|
|
|
|
|
2006-07-01 13:36:36 +02:00
|
|
|
static unsigned int nbds_max = 16;
|
2008-02-08 13:21:51 +01:00
|
|
|
static struct nbd_device *nbd_dev;
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
static int max_part;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Use just one lock (or at most 1 per NIC). Two arguments for this:
|
|
|
|
* 1. Each NIC is essentially a synchronization point for all servers
|
|
|
|
* accessed through that NIC so there's no need to have more locks
|
|
|
|
* than NICs anyway.
|
|
|
|
* 2. More locks lead to more "Dirty cache line bouncing" which will slow
|
|
|
|
* down each lock to the point where they're actually slower than just
|
|
|
|
* a single lock.
|
|
|
|
* Thanks go to Jens Axboe and Al Viro for their LKML emails explaining this!
|
|
|
|
*/
|
|
|
|
static DEFINE_SPINLOCK(nbd_lock);
|
|
|
|
|
|
|
|
#ifndef NDEBUG
|
|
|
|
static const char *ioctl_cmd_to_ascii(int cmd)
|
|
|
|
{
|
|
|
|
switch (cmd) {
|
|
|
|
case NBD_SET_SOCK: return "set-sock";
|
|
|
|
case NBD_SET_BLKSIZE: return "set-blksize";
|
|
|
|
case NBD_SET_SIZE: return "set-size";
|
2012-10-05 02:16:15 +02:00
|
|
|
case NBD_SET_TIMEOUT: return "set-timeout";
|
|
|
|
case NBD_SET_FLAGS: return "set-flags";
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_DO_IT: return "do-it";
|
|
|
|
case NBD_CLEAR_SOCK: return "clear-sock";
|
|
|
|
case NBD_CLEAR_QUE: return "clear-que";
|
|
|
|
case NBD_PRINT_DEBUG: return "print-debug";
|
|
|
|
case NBD_SET_SIZE_BLOCKS: return "set-size-blocks";
|
|
|
|
case NBD_DISCONNECT: return "disconnect";
|
|
|
|
case BLKROSET: return "set-read-only";
|
|
|
|
case BLKFLSBUF: return "flush-buffer-cache";
|
|
|
|
}
|
|
|
|
return "unknown";
|
|
|
|
}
|
|
|
|
|
|
|
|
static const char *nbdcmd_to_ascii(int cmd)
|
|
|
|
{
|
|
|
|
switch (cmd) {
|
|
|
|
case NBD_CMD_READ: return "read";
|
|
|
|
case NBD_CMD_WRITE: return "write";
|
|
|
|
case NBD_CMD_DISC: return "disconnect";
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
case NBD_CMD_FLUSH: return "flush";
|
2012-10-05 02:16:18 +02:00
|
|
|
case NBD_CMD_TRIM: return "trim/discard";
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
return "invalid";
|
|
|
|
}
|
|
|
|
#endif /* NDEBUG */
|
|
|
|
|
|
|
|
static void nbd_end_request(struct request *req)
|
|
|
|
{
|
2007-12-11 23:44:06 +01:00
|
|
|
int error = req->errors ? -EIO : 0;
|
2007-07-24 09:28:11 +02:00
|
|
|
struct request_queue *q = req->q;
|
2005-04-17 00:20:36 +02:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
dprintk(DBG_BLKDEV, "%s: request %p: %s\n", req->rq_disk->disk_name,
|
2007-12-11 23:44:06 +01:00
|
|
|
req, error ? "failed" : "done");
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
spin_lock_irqsave(q->queue_lock, flags);
|
2009-05-07 15:24:45 +02:00
|
|
|
__blk_end_request_all(req, error);
|
2005-04-17 00:20:36 +02:00
|
|
|
spin_unlock_irqrestore(q->queue_lock, flags);
|
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static void sock_shutdown(struct nbd_device *nbd, int lock)
|
2007-10-17 08:27:37 +02:00
|
|
|
{
|
|
|
|
/* Forcibly shutdown the socket causing all listeners
|
|
|
|
* to error
|
|
|
|
*
|
|
|
|
* FIXME: This code is duplicated from sys_shutdown, but
|
|
|
|
* there should be a more generic interface rather than
|
|
|
|
* calling socket ops directly here */
|
|
|
|
if (lock)
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_lock(&nbd->tx_lock);
|
|
|
|
if (nbd->sock) {
|
|
|
|
dev_warn(disk_to_dev(nbd->disk), "shutting down socket\n");
|
|
|
|
kernel_sock_shutdown(nbd->sock, SHUT_RDWR);
|
|
|
|
nbd->sock = NULL;
|
2007-10-17 08:27:37 +02:00
|
|
|
}
|
|
|
|
if (lock)
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_unlock(&nbd->tx_lock);
|
2007-10-17 08:27:37 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static void nbd_xmit_timeout(unsigned long arg)
|
|
|
|
{
|
|
|
|
struct task_struct *task = (struct task_struct *)arg;
|
|
|
|
|
|
|
|
printk(KERN_WARNING "nbd: killing hung xmit (%s, pid: %d)\n",
|
|
|
|
task->comm, task->pid);
|
|
|
|
force_sig(SIGKILL, task);
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/*
|
|
|
|
* Send or receive packet.
|
|
|
|
*/
|
2012-03-28 23:42:51 +02:00
|
|
|
static int sock_xmit(struct nbd_device *nbd, int send, void *buf, int size,
|
2005-04-17 00:20:36 +02:00
|
|
|
int msg_flags)
|
|
|
|
{
|
2012-03-28 23:42:51 +02:00
|
|
|
struct socket *sock = nbd->sock;
|
2005-04-17 00:20:36 +02:00
|
|
|
int result;
|
|
|
|
struct msghdr msg;
|
|
|
|
struct kvec iov;
|
2007-07-16 08:41:32 +02:00
|
|
|
sigset_t blocked, oldset;
|
2012-08-01 01:44:32 +02:00
|
|
|
unsigned long pflags = current->flags;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-04-02 22:04:47 +02:00
|
|
|
if (unlikely(!sock)) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Attempted %s on closed socket in sock_xmit\n",
|
|
|
|
(send ? "send" : "recv"));
|
2008-04-02 22:04:47 +02:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/* Allow interception of SIGKILL only
|
|
|
|
* Don't allow other signals to interrupt the transmission */
|
2007-07-16 08:41:32 +02:00
|
|
|
siginitsetinv(&blocked, sigmask(SIGKILL));
|
|
|
|
sigprocmask(SIG_SETMASK, &blocked, &oldset);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-08-01 01:44:32 +02:00
|
|
|
current->flags |= PF_MEMALLOC;
|
2005-04-17 00:20:36 +02:00
|
|
|
do {
|
2012-08-01 01:44:32 +02:00
|
|
|
sock->sk->sk_allocation = GFP_NOIO | __GFP_MEMALLOC;
|
2005-04-17 00:20:36 +02:00
|
|
|
iov.iov_base = buf;
|
|
|
|
iov.iov_len = size;
|
|
|
|
msg.msg_name = NULL;
|
|
|
|
msg.msg_namelen = 0;
|
|
|
|
msg.msg_control = NULL;
|
|
|
|
msg.msg_controllen = 0;
|
|
|
|
msg.msg_flags = msg_flags | MSG_NOSIGNAL;
|
|
|
|
|
2007-10-17 08:27:37 +02:00
|
|
|
if (send) {
|
|
|
|
struct timer_list ti;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
if (nbd->xmit_timeout) {
|
2007-10-17 08:27:37 +02:00
|
|
|
init_timer(&ti);
|
|
|
|
ti.function = nbd_xmit_timeout;
|
|
|
|
ti.data = (unsigned long)current;
|
2012-03-28 23:42:51 +02:00
|
|
|
ti.expires = jiffies + nbd->xmit_timeout;
|
2007-10-17 08:27:37 +02:00
|
|
|
add_timer(&ti);
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
result = kernel_sendmsg(sock, &msg, &iov, 1, size);
|
2012-03-28 23:42:51 +02:00
|
|
|
if (nbd->xmit_timeout)
|
2007-10-17 08:27:37 +02:00
|
|
|
del_timer_sync(&ti);
|
|
|
|
} else
|
2011-05-28 14:44:46 +02:00
|
|
|
result = kernel_recvmsg(sock, &msg, &iov, 1, size,
|
|
|
|
msg.msg_flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (signal_pending(current)) {
|
|
|
|
siginfo_t info;
|
|
|
|
printk(KERN_WARNING "nbd (pid %d: %s) got signal %d\n",
|
2007-10-19 08:40:40 +02:00
|
|
|
task_pid_nr(current), current->comm,
|
2007-07-16 08:41:32 +02:00
|
|
|
dequeue_signal_lock(current, ¤t->blocked, &info));
|
2005-04-17 00:20:36 +02:00
|
|
|
result = -EINTR;
|
2012-03-28 23:42:51 +02:00
|
|
|
sock_shutdown(nbd, !send);
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (result <= 0) {
|
|
|
|
if (result == 0)
|
|
|
|
result = -EPIPE; /* short read */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
size -= result;
|
|
|
|
buf += result;
|
|
|
|
} while (size > 0);
|
|
|
|
|
2007-07-16 08:41:32 +02:00
|
|
|
sigprocmask(SIG_SETMASK, &oldset, NULL);
|
2012-08-01 01:44:32 +02:00
|
|
|
tsk_restore_flags(current, pflags, PF_MEMALLOC);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static inline int sock_send_bvec(struct nbd_device *nbd, struct bio_vec *bvec,
|
2005-04-17 00:20:36 +02:00
|
|
|
int flags)
|
|
|
|
{
|
|
|
|
int result;
|
|
|
|
void *kaddr = kmap(bvec->bv_page);
|
2012-03-28 23:42:51 +02:00
|
|
|
result = sock_xmit(nbd, 1, kaddr + bvec->bv_offset,
|
|
|
|
bvec->bv_len, flags);
|
2005-04-17 00:20:36 +02:00
|
|
|
kunmap(bvec->bv_page);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2007-10-17 08:27:37 +02:00
|
|
|
/* always call with the tx_lock held */
|
2012-03-28 23:42:51 +02:00
|
|
|
static int nbd_send_req(struct nbd_device *nbd, struct request *req)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2007-09-25 12:35:59 +02:00
|
|
|
int result, flags;
|
2005-04-17 00:20:36 +02:00
|
|
|
struct nbd_request request;
|
2009-05-07 15:24:45 +02:00
|
|
|
unsigned long size = blk_rq_bytes(req);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2014-06-06 23:38:30 +02:00
|
|
|
memset(&request, 0, sizeof(request));
|
2005-04-17 00:20:36 +02:00
|
|
|
request.magic = htonl(NBD_REQUEST_MAGIC);
|
|
|
|
request.type = htonl(nbd_cmd(req));
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
|
2014-06-06 23:38:30 +02:00
|
|
|
if (nbd_cmd(req) != NBD_CMD_FLUSH && nbd_cmd(req) != NBD_CMD_DISC) {
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
request.from = cpu_to_be64((u64)blk_rq_pos(req) << 9);
|
|
|
|
request.len = htonl(size);
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
memcpy(request.handle, &req, sizeof(req));
|
|
|
|
|
2009-05-07 15:24:39 +02:00
|
|
|
dprintk(DBG_TX, "%s: request %p: sending control (%s@%llu,%uB)\n",
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->disk->disk_name, req,
|
2005-04-17 00:20:36 +02:00
|
|
|
nbdcmd_to_ascii(nbd_cmd(req)),
|
2009-05-07 15:24:39 +02:00
|
|
|
(unsigned long long)blk_rq_pos(req) << 9,
|
2009-05-07 15:24:45 +02:00
|
|
|
blk_rq_bytes(req));
|
2012-03-28 23:42:51 +02:00
|
|
|
result = sock_xmit(nbd, 1, &request, sizeof(request),
|
2007-10-17 08:27:37 +02:00
|
|
|
(nbd_cmd(req) == NBD_CMD_WRITE) ? MSG_MORE : 0);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (result <= 0) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Send control failed (result %d)\n", result);
|
2005-04-17 00:20:36 +02:00
|
|
|
goto error_out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (nbd_cmd(req) == NBD_CMD_WRITE) {
|
2007-09-25 12:35:59 +02:00
|
|
|
struct req_iterator iter;
|
2013-11-24 02:19:00 +01:00
|
|
|
struct bio_vec bvec;
|
2005-04-17 00:20:36 +02:00
|
|
|
/*
|
|
|
|
* we are really probing at internals to determine
|
|
|
|
* whether to set MSG_MORE or not...
|
|
|
|
*/
|
2007-09-25 12:35:59 +02:00
|
|
|
rq_for_each_segment(bvec, req, iter) {
|
2007-08-16 13:43:12 +02:00
|
|
|
flags = 0;
|
2013-08-07 23:26:21 +02:00
|
|
|
if (!rq_iter_last(bvec, iter))
|
2007-08-16 13:43:12 +02:00
|
|
|
flags = MSG_MORE;
|
|
|
|
dprintk(DBG_TX, "%s: request %p: sending %d bytes data\n",
|
2013-11-24 02:19:00 +01:00
|
|
|
nbd->disk->disk_name, req, bvec.bv_len);
|
|
|
|
result = sock_send_bvec(nbd, &bvec, flags);
|
2007-08-16 13:43:12 +02:00
|
|
|
if (result <= 0) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Send data failed (result %d)\n",
|
|
|
|
result);
|
2007-08-16 13:43:12 +02:00
|
|
|
goto error_out;
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
error_out:
|
2009-04-03 01:58:42 +02:00
|
|
|
return -EIO;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static struct request *nbd_find_request(struct nbd_device *nbd,
|
2007-10-17 08:26:14 +02:00
|
|
|
struct request *xreq)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2007-10-17 08:26:14 +02:00
|
|
|
struct request *req, *tmp;
|
2006-01-06 09:09:47 +01:00
|
|
|
int err;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
err = wait_event_interruptible(nbd->active_wq, nbd->active_req != xreq);
|
2006-01-06 09:09:47 +01:00
|
|
|
if (unlikely(err))
|
|
|
|
goto out;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_lock(&nbd->queue_lock);
|
|
|
|
list_for_each_entry_safe(req, tmp, &nbd->queue_head, queuelist) {
|
2005-04-17 00:20:36 +02:00
|
|
|
if (req != xreq)
|
|
|
|
continue;
|
|
|
|
list_del_init(&req->queuelist);
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_unlock(&nbd->queue_lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
return req;
|
|
|
|
}
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_unlock(&nbd->queue_lock);
|
2006-01-06 09:09:47 +01:00
|
|
|
|
|
|
|
err = -ENOENT;
|
|
|
|
|
|
|
|
out:
|
|
|
|
return ERR_PTR(err);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static inline int sock_recv_bvec(struct nbd_device *nbd, struct bio_vec *bvec)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
int result;
|
|
|
|
void *kaddr = kmap(bvec->bv_page);
|
2012-03-28 23:42:51 +02:00
|
|
|
result = sock_xmit(nbd, 0, kaddr + bvec->bv_offset, bvec->bv_len,
|
2005-04-17 00:20:36 +02:00
|
|
|
MSG_WAITALL);
|
|
|
|
kunmap(bvec->bv_page);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* NULL returned = something went wrong, inform userspace */
|
2012-03-28 23:42:51 +02:00
|
|
|
static struct request *nbd_read_stat(struct nbd_device *nbd)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
int result;
|
|
|
|
struct nbd_reply reply;
|
|
|
|
struct request *req;
|
|
|
|
|
|
|
|
reply.magic = 0;
|
2012-03-28 23:42:51 +02:00
|
|
|
result = sock_xmit(nbd, 0, &reply, sizeof(reply), MSG_WAITALL);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (result <= 0) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Receive control failed (result %d)\n", result);
|
2005-04-17 00:20:36 +02:00
|
|
|
goto harderror;
|
|
|
|
}
|
2006-07-30 12:03:31 +02:00
|
|
|
|
|
|
|
if (ntohl(reply.magic) != NBD_REPLY_MAGIC) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk), "Wrong magic (0x%lx)\n",
|
2006-07-30 12:03:31 +02:00
|
|
|
(unsigned long)ntohl(reply.magic));
|
|
|
|
result = -EPROTO;
|
|
|
|
goto harderror;
|
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
req = nbd_find_request(nbd, *(struct request **)reply.handle);
|
2008-04-29 10:03:09 +02:00
|
|
|
if (IS_ERR(req)) {
|
2006-01-06 09:09:47 +01:00
|
|
|
result = PTR_ERR(req);
|
|
|
|
if (result != -ENOENT)
|
|
|
|
goto harderror;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk), "Unexpected reply (%p)\n",
|
2011-08-19 14:48:22 +02:00
|
|
|
reply.handle);
|
2005-04-17 00:20:36 +02:00
|
|
|
result = -EBADR;
|
|
|
|
goto harderror;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ntohl(reply.error)) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk), "Other side returned error (%d)\n",
|
2011-08-19 14:48:22 +02:00
|
|
|
ntohl(reply.error));
|
2005-04-17 00:20:36 +02:00
|
|
|
req->errors++;
|
|
|
|
return req;
|
|
|
|
}
|
|
|
|
|
|
|
|
dprintk(DBG_RX, "%s: request %p: got reply\n",
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->disk->disk_name, req);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (nbd_cmd(req) == NBD_CMD_READ) {
|
2007-09-25 12:35:59 +02:00
|
|
|
struct req_iterator iter;
|
2013-11-24 02:19:00 +01:00
|
|
|
struct bio_vec bvec;
|
2007-09-25 12:35:59 +02:00
|
|
|
|
|
|
|
rq_for_each_segment(bvec, req, iter) {
|
2013-11-24 02:19:00 +01:00
|
|
|
result = sock_recv_bvec(nbd, &bvec);
|
2007-08-16 13:43:12 +02:00
|
|
|
if (result <= 0) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk), "Receive data failed (result %d)\n",
|
2011-08-19 14:48:22 +02:00
|
|
|
result);
|
2007-08-16 13:43:12 +02:00
|
|
|
req->errors++;
|
|
|
|
return req;
|
|
|
|
}
|
|
|
|
dprintk(DBG_RX, "%s: request %p: got %d bytes data\n",
|
2013-11-24 02:19:00 +01:00
|
|
|
nbd->disk->disk_name, req, bvec.bv_len);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return req;
|
|
|
|
harderror:
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->harderror = result;
|
2005-04-17 00:20:36 +02:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2007-05-21 22:08:01 +02:00
|
|
|
static ssize_t pid_show(struct device *dev,
|
|
|
|
struct device_attribute *attr, char *buf)
|
2006-12-07 05:40:53 +01:00
|
|
|
{
|
2007-05-21 22:08:01 +02:00
|
|
|
struct gendisk *disk = dev_to_disk(dev);
|
|
|
|
|
|
|
|
return sprintf(buf, "%ld\n",
|
2006-12-07 05:40:53 +01:00
|
|
|
(long) ((struct nbd_device *)disk->private_data)->pid);
|
|
|
|
}
|
|
|
|
|
2007-05-21 22:08:01 +02:00
|
|
|
static struct device_attribute pid_attr = {
|
2008-10-19 05:28:50 +02:00
|
|
|
.attr = { .name = "pid", .mode = S_IRUGO},
|
2006-12-07 05:40:53 +01:00
|
|
|
.show = pid_show,
|
|
|
|
};
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static int nbd_do_it(struct nbd_device *nbd)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct request *req;
|
2007-05-09 11:33:36 +02:00
|
|
|
int ret;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
BUG_ON(nbd->magic != NBD_MAGIC);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-08-01 01:44:32 +02:00
|
|
|
sk_set_memalloc(nbd->sock->sk);
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->pid = task_pid_nr(current);
|
|
|
|
ret = device_create_file(disk_to_dev(nbd->disk), &pid_attr);
|
2007-05-09 11:33:36 +02:00
|
|
|
if (ret) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk), "device_create_file failed!\n");
|
|
|
|
nbd->pid = 0;
|
2007-05-09 11:33:36 +02:00
|
|
|
return ret;
|
|
|
|
}
|
2006-12-07 05:40:53 +01:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
while ((req = nbd_read_stat(nbd)) != NULL)
|
2005-04-17 00:20:36 +02:00
|
|
|
nbd_end_request(req);
|
2006-12-07 05:40:53 +01:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
device_remove_file(disk_to_dev(nbd->disk), &pid_attr);
|
|
|
|
nbd->pid = 0;
|
2007-05-09 11:33:36 +02:00
|
|
|
return 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static void nbd_clear_que(struct nbd_device *nbd)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct request *req;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
BUG_ON(nbd->magic != NBD_MAGIC);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2006-01-06 09:09:47 +01:00
|
|
|
/*
|
2012-03-28 23:42:51 +02:00
|
|
|
* Because we have set nbd->sock to NULL under the tx_lock, all
|
2006-01-06 09:09:47 +01:00
|
|
|
* modifications to the list must have completed by now. For
|
|
|
|
* the same reason, the active_req must be NULL.
|
|
|
|
*
|
|
|
|
* As a consequence, we don't need to take the spin lock while
|
|
|
|
* purging the list here.
|
|
|
|
*/
|
2012-03-28 23:42:51 +02:00
|
|
|
BUG_ON(nbd->sock);
|
|
|
|
BUG_ON(nbd->active_req);
|
2006-01-06 09:09:47 +01:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
while (!list_empty(&nbd->queue_head)) {
|
|
|
|
req = list_entry(nbd->queue_head.next, struct request,
|
2006-01-06 09:09:47 +01:00
|
|
|
queuelist);
|
|
|
|
list_del_init(&req->queuelist);
|
|
|
|
req->errors++;
|
|
|
|
nbd_end_request(req);
|
|
|
|
}
|
2012-09-17 23:09:02 +02:00
|
|
|
|
|
|
|
while (!list_empty(&nbd->waiting_queue)) {
|
|
|
|
req = list_entry(nbd->waiting_queue.next, struct request,
|
|
|
|
queuelist);
|
|
|
|
list_del_init(&req->queuelist);
|
|
|
|
req->errors++;
|
|
|
|
nbd_end_request(req);
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2007-10-17 08:27:37 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static void nbd_handle_req(struct nbd_device *nbd, struct request *req)
|
2008-04-29 10:02:46 +02:00
|
|
|
{
|
2010-08-07 18:17:56 +02:00
|
|
|
if (req->cmd_type != REQ_TYPE_FS)
|
2008-04-29 10:02:46 +02:00
|
|
|
goto error_out;
|
|
|
|
|
|
|
|
nbd_cmd(req) = NBD_CMD_READ;
|
|
|
|
if (rq_data_dir(req) == WRITE) {
|
2012-10-05 02:16:18 +02:00
|
|
|
if ((req->cmd_flags & REQ_DISCARD)) {
|
|
|
|
WARN_ON(!(nbd->flags & NBD_FLAG_SEND_TRIM));
|
|
|
|
nbd_cmd(req) = NBD_CMD_TRIM;
|
|
|
|
} else
|
|
|
|
nbd_cmd(req) = NBD_CMD_WRITE;
|
2012-10-05 02:16:15 +02:00
|
|
|
if (nbd->flags & NBD_FLAG_READ_ONLY) {
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Write on read-only\n");
|
2008-04-29 10:02:46 +02:00
|
|
|
goto error_out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
if (req->cmd_flags & REQ_FLUSH) {
|
|
|
|
BUG_ON(unlikely(blk_rq_sectors(req)));
|
|
|
|
nbd_cmd(req) = NBD_CMD_FLUSH;
|
|
|
|
}
|
|
|
|
|
2008-04-29 10:02:46 +02:00
|
|
|
req->errors = 0;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_lock(&nbd->tx_lock);
|
|
|
|
if (unlikely(!nbd->sock)) {
|
|
|
|
mutex_unlock(&nbd->tx_lock);
|
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Attempted send on closed socket\n");
|
2009-04-03 01:58:42 +02:00
|
|
|
goto error_out;
|
2008-04-29 10:02:46 +02:00
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->active_req = req;
|
2008-04-29 10:02:46 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
if (nbd_send_req(nbd, req) != 0) {
|
|
|
|
dev_err(disk_to_dev(nbd->disk), "Request send failed\n");
|
2008-04-29 10:02:46 +02:00
|
|
|
req->errors++;
|
|
|
|
nbd_end_request(req);
|
|
|
|
} else {
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_lock(&nbd->queue_lock);
|
2012-07-31 08:47:13 +02:00
|
|
|
list_add_tail(&req->queuelist, &nbd->queue_head);
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_unlock(&nbd->queue_lock);
|
2008-04-29 10:02:46 +02:00
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->active_req = NULL;
|
|
|
|
mutex_unlock(&nbd->tx_lock);
|
|
|
|
wake_up_all(&nbd->active_wq);
|
2008-04-29 10:02:46 +02:00
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
error_out:
|
|
|
|
req->errors++;
|
|
|
|
nbd_end_request(req);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int nbd_thread(void *data)
|
|
|
|
{
|
2012-03-28 23:42:51 +02:00
|
|
|
struct nbd_device *nbd = data;
|
2008-04-29 10:02:46 +02:00
|
|
|
struct request *req;
|
|
|
|
|
2014-03-11 11:09:12 +01:00
|
|
|
set_user_nice(current, MIN_NICE);
|
2012-03-28 23:42:51 +02:00
|
|
|
while (!kthread_should_stop() || !list_empty(&nbd->waiting_queue)) {
|
2008-04-29 10:02:46 +02:00
|
|
|
/* wait for something to do */
|
2012-03-28 23:42:51 +02:00
|
|
|
wait_event_interruptible(nbd->waiting_wq,
|
2008-04-29 10:02:46 +02:00
|
|
|
kthread_should_stop() ||
|
2012-03-28 23:42:51 +02:00
|
|
|
!list_empty(&nbd->waiting_queue));
|
2008-04-29 10:02:46 +02:00
|
|
|
|
|
|
|
/* extract request */
|
2012-03-28 23:42:51 +02:00
|
|
|
if (list_empty(&nbd->waiting_queue))
|
2008-04-29 10:02:46 +02:00
|
|
|
continue;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_lock_irq(&nbd->queue_lock);
|
|
|
|
req = list_entry(nbd->waiting_queue.next, struct request,
|
2008-04-29 10:02:46 +02:00
|
|
|
queuelist);
|
|
|
|
list_del_init(&req->queuelist);
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_unlock_irq(&nbd->queue_lock);
|
2008-04-29 10:02:46 +02:00
|
|
|
|
|
|
|
/* handle request */
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd_handle_req(nbd, req);
|
2008-04-29 10:02:46 +02:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/*
|
|
|
|
* We always wait for result of write, for now. It would be nice to make it optional
|
|
|
|
* in future
|
2012-03-28 23:42:51 +02:00
|
|
|
* if ((rq_data_dir(req) == WRITE) && (nbd->flags & NBD_WRITE_NOCHK))
|
2005-04-17 00:20:36 +02:00
|
|
|
* { printk( "Warning: Ignoring result!\n"); nbd_end_request( req ); }
|
|
|
|
*/
|
|
|
|
|
2009-04-03 01:58:42 +02:00
|
|
|
static void do_nbd_request(struct request_queue *q)
|
2013-02-28 02:05:28 +01:00
|
|
|
__releases(q->queue_lock) __acquires(q->queue_lock)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct request *req;
|
|
|
|
|
2009-05-08 04:54:16 +02:00
|
|
|
while ((req = blk_fetch_request(q)) != NULL) {
|
2012-03-28 23:42:51 +02:00
|
|
|
struct nbd_device *nbd;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-04-29 10:02:46 +02:00
|
|
|
spin_unlock_irq(q->queue_lock);
|
|
|
|
|
2006-08-10 08:44:47 +02:00
|
|
|
dprintk(DBG_BLKDEV, "%s: request %p: dequeued (flags=%x)\n",
|
|
|
|
req->rq_disk->disk_name, req, req->cmd_type);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd = req->rq_disk->private_data;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
BUG_ON(nbd->magic != NBD_MAGIC);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
if (unlikely(!nbd->sock)) {
|
|
|
|
dev_err(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:22 +02:00
|
|
|
"Attempted send on closed socket\n");
|
2009-02-11 22:04:45 +01:00
|
|
|
req->errors++;
|
|
|
|
nbd_end_request(req);
|
|
|
|
spin_lock_irq(q->queue_lock);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
spin_lock_irq(&nbd->queue_lock);
|
|
|
|
list_add_tail(&req->queuelist, &nbd->waiting_queue);
|
|
|
|
spin_unlock_irq(&nbd->queue_lock);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
wake_up(&nbd->waiting_wq);
|
2006-01-06 09:09:47 +01:00
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
spin_lock_irq(q->queue_lock);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-04-03 01:58:41 +02:00
|
|
|
/* Must be called with tx_lock held */
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
|
2009-04-03 01:58:41 +02:00
|
|
|
unsigned int cmd, unsigned long arg)
|
|
|
|
{
|
2005-04-17 00:20:36 +02:00
|
|
|
switch (cmd) {
|
2009-04-03 01:58:41 +02:00
|
|
|
case NBD_DISCONNECT: {
|
|
|
|
struct request sreq;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_info(disk_to_dev(nbd->disk), "NBD_DISCONNECT\n");
|
2013-02-28 02:05:25 +01:00
|
|
|
if (!nbd->sock)
|
|
|
|
return -EINVAL;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2013-02-28 02:05:25 +01:00
|
|
|
mutex_unlock(&nbd->tx_lock);
|
|
|
|
fsync_bdev(bdev);
|
|
|
|
mutex_lock(&nbd->tx_lock);
|
2008-04-29 09:54:37 +02:00
|
|
|
blk_rq_init(NULL, &sreq);
|
2006-08-10 08:44:47 +02:00
|
|
|
sreq.cmd_type = REQ_TYPE_SPECIAL;
|
2005-04-17 00:20:36 +02:00
|
|
|
nbd_cmd(&sreq) = NBD_CMD_DISC;
|
2013-02-28 02:05:25 +01:00
|
|
|
|
|
|
|
/* Check again after getting mutex back. */
|
2012-03-28 23:42:51 +02:00
|
|
|
if (!nbd->sock)
|
2005-04-17 00:20:36 +02:00
|
|
|
return -EINVAL;
|
2013-02-28 02:05:25 +01:00
|
|
|
|
2013-07-04 00:09:04 +02:00
|
|
|
nbd->disconnect = 1;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd_send_req(nbd, &sreq);
|
2013-07-04 00:09:04 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2009-04-03 01:58:41 +02:00
|
|
|
case NBD_CLEAR_SOCK: {
|
2014-03-06 02:41:36 +01:00
|
|
|
struct socket *sock = nbd->sock;
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->sock = NULL;
|
|
|
|
nbd_clear_que(nbd);
|
|
|
|
BUG_ON(!list_empty(&nbd->queue_head));
|
2012-09-17 23:09:02 +02:00
|
|
|
BUG_ON(!list_empty(&nbd->waiting_queue));
|
2013-02-28 02:05:25 +01:00
|
|
|
kill_bdev(bdev);
|
2014-03-06 02:41:36 +01:00
|
|
|
if (sock)
|
|
|
|
sockfd_put(sock);
|
2009-04-03 01:58:41 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
case NBD_SET_SOCK: {
|
2014-03-06 02:41:36 +01:00
|
|
|
struct socket *sock;
|
|
|
|
int err;
|
|
|
|
if (nbd->sock)
|
2005-04-17 00:20:36 +02:00
|
|
|
return -EBUSY;
|
2014-03-06 02:41:36 +01:00
|
|
|
sock = sockfd_lookup(arg, &err);
|
|
|
|
if (sock) {
|
|
|
|
nbd->sock = sock;
|
|
|
|
if (max_part > 0)
|
|
|
|
bdev->bd_invalidated = 1;
|
|
|
|
nbd->disconnect = 0; /* we're connected now */
|
|
|
|
return 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2009-04-03 01:58:41 +02:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_SET_BLKSIZE:
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->blksize = arg;
|
|
|
|
nbd->bytesize &= ~(nbd->blksize-1);
|
|
|
|
bdev->bd_inode->i_size = nbd->bytesize;
|
|
|
|
set_blocksize(bdev, nbd->blksize);
|
|
|
|
set_capacity(nbd->disk, nbd->bytesize >> 9);
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_SET_SIZE:
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->bytesize = arg & ~(nbd->blksize-1);
|
|
|
|
bdev->bd_inode->i_size = nbd->bytesize;
|
|
|
|
set_blocksize(bdev, nbd->blksize);
|
|
|
|
set_capacity(nbd->disk, nbd->bytesize >> 9);
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2007-10-17 08:27:37 +02:00
|
|
|
case NBD_SET_TIMEOUT:
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->xmit_timeout = arg * HZ;
|
2007-10-17 08:27:37 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2012-10-05 02:16:15 +02:00
|
|
|
case NBD_SET_FLAGS:
|
|
|
|
nbd->flags = arg;
|
|
|
|
return 0;
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_SET_SIZE_BLOCKS:
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->bytesize = ((u64) arg) * nbd->blksize;
|
|
|
|
bdev->bd_inode->i_size = nbd->bytesize;
|
|
|
|
set_blocksize(bdev, nbd->blksize);
|
|
|
|
set_capacity(nbd->disk, nbd->bytesize >> 9);
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
|
|
|
case NBD_DO_IT: {
|
|
|
|
struct task_struct *thread;
|
2014-03-06 02:41:36 +01:00
|
|
|
struct socket *sock;
|
2009-04-03 01:58:41 +02:00
|
|
|
int error;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
if (nbd->pid)
|
2009-01-15 22:51:03 +01:00
|
|
|
return -EBUSY;
|
2014-03-06 02:41:36 +01:00
|
|
|
if (!nbd->sock)
|
2005-04-17 00:20:36 +02:00
|
|
|
return -EINVAL;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_unlock(&nbd->tx_lock);
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2013-02-28 02:05:26 +01:00
|
|
|
if (nbd->flags & NBD_FLAG_READ_ONLY)
|
|
|
|
set_device_ro(bdev, true);
|
2012-10-05 02:16:18 +02:00
|
|
|
if (nbd->flags & NBD_FLAG_SEND_TRIM)
|
|
|
|
queue_flag_set_unlocked(QUEUE_FLAG_DISCARD,
|
|
|
|
nbd->disk->queue);
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
if (nbd->flags & NBD_FLAG_SEND_FLUSH)
|
|
|
|
blk_queue_flush(nbd->disk->queue, REQ_FLUSH);
|
|
|
|
else
|
|
|
|
blk_queue_flush(nbd->disk->queue, 0);
|
2012-10-05 02:16:18 +02:00
|
|
|
|
2013-07-04 00:01:14 +02:00
|
|
|
thread = kthread_create(nbd_thread, nbd, "%s",
|
|
|
|
nbd->disk->disk_name);
|
2009-04-03 01:58:41 +02:00
|
|
|
if (IS_ERR(thread)) {
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_lock(&nbd->tx_lock);
|
2008-04-29 10:02:46 +02:00
|
|
|
return PTR_ERR(thread);
|
2009-04-03 01:58:41 +02:00
|
|
|
}
|
2008-04-29 10:02:46 +02:00
|
|
|
wake_up_process(thread);
|
2012-03-28 23:42:51 +02:00
|
|
|
error = nbd_do_it(nbd);
|
2008-04-29 10:02:46 +02:00
|
|
|
kthread_stop(thread);
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_lock(&nbd->tx_lock);
|
2007-05-09 11:33:36 +02:00
|
|
|
if (error)
|
|
|
|
return error;
|
2012-03-28 23:42:51 +02:00
|
|
|
sock_shutdown(nbd, 0);
|
2014-03-06 02:41:36 +01:00
|
|
|
sock = nbd->sock;
|
|
|
|
nbd->sock = NULL;
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd_clear_que(nbd);
|
|
|
|
dev_warn(disk_to_dev(nbd->disk), "queue cleared\n");
|
2013-02-28 02:05:25 +01:00
|
|
|
kill_bdev(bdev);
|
2012-10-05 02:16:18 +02:00
|
|
|
queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, nbd->disk->queue);
|
2013-02-28 02:05:26 +01:00
|
|
|
set_device_ro(bdev, false);
|
2014-03-06 02:41:36 +01:00
|
|
|
if (sock)
|
|
|
|
sockfd_put(sock);
|
nbd: support FLUSH requests
Currently, the NBD device does not accept flush requests from the Linux
block layer. If the NBD server opened the target with neither O_SYNC nor
O_DSYNC, however, the device will be effectively backed by a writeback
cache. Without issuing flushes properly, operation of the NBD device will
not be safe against power losses.
The NBD protocol has support for both a cache flush command and a FUA
command flag; the server will also pass a flag to note its support for
these features. This patch adds support for the cache flush command and
flag. In the kernel, we receive the flags via the NBD_SET_FLAGS ioctl,
and map NBD_FLAG_SEND_FLUSH to the argument of blk_queue_flush. When the
flag is active the block layer will send REQ_FLUSH requests, which we
translate to NBD_CMD_FLUSH commands.
FUA support is not included in this patch because all free software
servers implement it with a full fdatasync; thus it has no advantage over
supporting flush only. Because I [Paolo] cannot really benchmark it in a
realistic scenario, I cannot tell if it is a good idea or not. It is also
not clear if it is valid for an NBD server to support FUA but not flush.
The Linux block layer gives a warning for this combination, the NBD
protocol documentation says nothing about it.
The patch also fixes a small problem in the handling of flags: nbd->flags
must be cleared at the end of NBD_DO_IT, but the driver was not doing
that. The bug manifests itself as follows. Suppose you two different
client/server pairs to start the NBD device. Suppose also that the first
client supports NBD_SET_FLAGS, and the first server sends
NBD_FLAG_SEND_FLUSH; the second pair instead does neither of these two
things. Before this patch, the second invocation of NBD_DO_IT will use a
stale value of nbd->flags, and the second server will issue an error every
time it receives an NBD_CMD_FLUSH command.
This bug is pre-existing, but it becomes much more important after this
patch; flush failures make the device pretty much unusable, unlike
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Acked-by: Paul Clements <Paul.Clements@steeleye.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 02:05:23 +01:00
|
|
|
nbd->flags = 0;
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->bytesize = 0;
|
2008-03-02 15:33:33 +01:00
|
|
|
bdev->bd_inode->i_size = 0;
|
2012-03-28 23:42:51 +02:00
|
|
|
set_capacity(nbd->disk, 0);
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
if (max_part > 0)
|
2008-03-02 15:33:33 +01:00
|
|
|
ioctl_by_bdev(bdev, BLKRRPART, 0);
|
2013-07-04 00:09:04 +02:00
|
|
|
if (nbd->disconnect) /* user requested, ignore socket errors */
|
|
|
|
return 0;
|
2012-03-28 23:42:51 +02:00
|
|
|
return nbd->harderror;
|
2009-04-03 01:58:41 +02:00
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_CLEAR_QUE:
|
2006-01-06 09:09:47 +01:00
|
|
|
/*
|
|
|
|
* This is for compatibility only. The queue is always cleared
|
|
|
|
* by NBD_DO_IT or NBD_CLEAR_SOCK.
|
|
|
|
*/
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
case NBD_PRINT_DEBUG:
|
2012-03-28 23:42:51 +02:00
|
|
|
dev_info(disk_to_dev(nbd->disk),
|
2011-08-19 14:48:28 +02:00
|
|
|
"next = %p, prev = %p, head = %p\n",
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->queue_head.next, nbd->queue_head.prev,
|
|
|
|
&nbd->queue_head);
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
|
|
|
}
|
2009-04-03 01:58:41 +02:00
|
|
|
return -ENOTTY;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int nbd_ioctl(struct block_device *bdev, fmode_t mode,
|
|
|
|
unsigned int cmd, unsigned long arg)
|
|
|
|
{
|
2012-03-28 23:42:51 +02:00
|
|
|
struct nbd_device *nbd = bdev->bd_disk->private_data;
|
2009-04-03 01:58:41 +02:00
|
|
|
int error;
|
|
|
|
|
|
|
|
if (!capable(CAP_SYS_ADMIN))
|
|
|
|
return -EPERM;
|
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
BUG_ON(nbd->magic != NBD_MAGIC);
|
2009-04-03 01:58:41 +02:00
|
|
|
|
|
|
|
/* Anyone capable of this syscall can do *real bad* things */
|
|
|
|
dprintk(DBG_IOCTL, "%s: nbd_ioctl cmd=%s(0x%x) arg=%lu\n",
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd->disk->disk_name, ioctl_cmd_to_ascii(cmd), cmd, arg);
|
2009-04-03 01:58:41 +02:00
|
|
|
|
2012-03-28 23:42:51 +02:00
|
|
|
mutex_lock(&nbd->tx_lock);
|
|
|
|
error = __nbd_ioctl(bdev, nbd, cmd, arg);
|
|
|
|
mutex_unlock(&nbd->tx_lock);
|
2009-04-03 01:58:41 +02:00
|
|
|
|
|
|
|
return error;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2009-09-22 02:01:13 +02:00
|
|
|
static const struct block_device_operations nbd_fops =
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
.owner = THIS_MODULE,
|
2010-07-08 10:18:46 +02:00
|
|
|
.ioctl = nbd_ioctl,
|
2005-04-17 00:20:36 +02:00
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* And here should be modules and kernel interface
|
|
|
|
* (Just smiley confuses emacs :-)
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int __init nbd_init(void)
|
|
|
|
{
|
|
|
|
int err = -ENOMEM;
|
|
|
|
int i;
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
int part_shift;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2006-03-25 12:07:04 +01:00
|
|
|
BUILD_BUG_ON(sizeof(struct nbd_request) != 28);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
if (max_part < 0) {
|
2011-08-19 14:48:28 +02:00
|
|
|
printk(KERN_ERR "nbd: max_part must be >= 0\n");
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
part_shift = 0;
|
2011-05-28 14:44:46 +02:00
|
|
|
if (max_part > 0) {
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
part_shift = fls(max_part);
|
|
|
|
|
2011-05-28 14:44:46 +02:00
|
|
|
/*
|
|
|
|
* Adjust max_part according to part_shift as it is exported
|
|
|
|
* to user space so that user can know the max number of
|
|
|
|
* partition kernel should be able to manage.
|
|
|
|
*
|
|
|
|
* Note that -1 is required because partition 0 is reserved
|
|
|
|
* for the whole disk.
|
|
|
|
*/
|
|
|
|
max_part = (1UL << part_shift) - 1;
|
|
|
|
}
|
|
|
|
|
2011-05-28 14:44:46 +02:00
|
|
|
if ((1UL << part_shift) > DISK_MAX_PARTS)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (nbds_max > 1UL << (MINORBITS - part_shift))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2015-01-27 13:38:22 +01:00
|
|
|
nbd_dev = kcalloc(nbds_max, sizeof(*nbd_dev), GFP_KERNEL);
|
|
|
|
if (!nbd_dev)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2005-05-01 17:59:07 +02:00
|
|
|
for (i = 0; i < nbds_max; i++) {
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
struct gendisk *disk = alloc_disk(1 << part_shift);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!disk)
|
|
|
|
goto out;
|
|
|
|
nbd_dev[i].disk = disk;
|
|
|
|
/*
|
|
|
|
* The new linux 2.5 block layer implementation requires
|
|
|
|
* every gendisk to have its very own request_queue struct.
|
|
|
|
* These structs are big so we dynamically allocate them.
|
|
|
|
*/
|
|
|
|
disk->queue = blk_init_queue(do_nbd_request, &nbd_lock);
|
|
|
|
if (!disk->queue) {
|
|
|
|
put_disk(disk);
|
|
|
|
goto out;
|
|
|
|
}
|
2008-10-31 10:06:37 +01:00
|
|
|
/*
|
|
|
|
* Tell the block layer that we are not a rotational device
|
|
|
|
*/
|
|
|
|
queue_flag_set_unlocked(QUEUE_FLAG_NONROT, disk->queue);
|
2014-10-04 18:55:32 +02:00
|
|
|
queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, disk->queue);
|
2012-10-05 02:16:18 +02:00
|
|
|
disk->queue->limits.discard_granularity = 512;
|
|
|
|
disk->queue->limits.max_discard_sectors = UINT_MAX;
|
|
|
|
disk->queue->limits.discard_zeroes_data = 0;
|
2013-05-01 00:28:28 +02:00
|
|
|
blk_queue_max_hw_sectors(disk->queue, 65536);
|
|
|
|
disk->queue->limits.max_sectors = 256;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (register_blkdev(NBD_MAJOR, "nbd")) {
|
|
|
|
err = -EIO;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
printk(KERN_INFO "nbd: registered device at major %d\n", NBD_MAJOR);
|
|
|
|
dprintk(DBG_INIT, "nbd: debugflags=0x%x\n", debugflags);
|
|
|
|
|
2005-05-01 17:59:07 +02:00
|
|
|
for (i = 0; i < nbds_max; i++) {
|
2005-04-17 00:20:36 +02:00
|
|
|
struct gendisk *disk = nbd_dev[i].disk;
|
2012-03-28 23:42:51 +02:00
|
|
|
nbd_dev[i].magic = NBD_MAGIC;
|
2008-04-29 10:02:46 +02:00
|
|
|
INIT_LIST_HEAD(&nbd_dev[i].waiting_queue);
|
2005-04-17 00:20:36 +02:00
|
|
|
spin_lock_init(&nbd_dev[i].queue_lock);
|
|
|
|
INIT_LIST_HEAD(&nbd_dev[i].queue_head);
|
2006-03-23 12:00:38 +01:00
|
|
|
mutex_init(&nbd_dev[i].tx_lock);
|
2006-01-06 09:09:47 +01:00
|
|
|
init_waitqueue_head(&nbd_dev[i].active_wq);
|
2008-04-29 10:02:46 +02:00
|
|
|
init_waitqueue_head(&nbd_dev[i].waiting_wq);
|
2005-04-17 00:20:36 +02:00
|
|
|
nbd_dev[i].blksize = 1024;
|
2007-10-17 08:27:36 +02:00
|
|
|
nbd_dev[i].bytesize = 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
disk->major = NBD_MAJOR;
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
disk->first_minor = i << part_shift;
|
2005-04-17 00:20:36 +02:00
|
|
|
disk->fops = &nbd_fops;
|
|
|
|
disk->private_data = &nbd_dev[i];
|
|
|
|
sprintf(disk->disk_name, "nbd%d", i);
|
2007-10-17 08:27:36 +02:00
|
|
|
set_capacity(disk, 0);
|
2005-04-17 00:20:36 +02:00
|
|
|
add_disk(disk);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
out:
|
|
|
|
while (i--) {
|
|
|
|
blk_cleanup_queue(nbd_dev[i].disk->queue);
|
|
|
|
put_disk(nbd_dev[i].disk);
|
|
|
|
}
|
2008-08-20 23:09:07 +02:00
|
|
|
kfree(nbd_dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __exit nbd_cleanup(void)
|
|
|
|
{
|
|
|
|
int i;
|
2005-05-01 17:59:07 +02:00
|
|
|
for (i = 0; i < nbds_max; i++) {
|
2005-04-17 00:20:36 +02:00
|
|
|
struct gendisk *disk = nbd_dev[i].disk;
|
2005-05-01 17:59:07 +02:00
|
|
|
nbd_dev[i].magic = 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
if (disk) {
|
|
|
|
del_gendisk(disk);
|
|
|
|
blk_cleanup_queue(disk->queue);
|
|
|
|
put_disk(disk);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
unregister_blkdev(NBD_MAJOR, "nbd");
|
2008-08-20 23:09:07 +02:00
|
|
|
kfree(nbd_dev);
|
2005-04-17 00:20:36 +02:00
|
|
|
printk(KERN_INFO "nbd: unregistered device at major %d\n", NBD_MAJOR);
|
|
|
|
}
|
|
|
|
|
|
|
|
module_init(nbd_init);
|
|
|
|
module_exit(nbd_cleanup);
|
|
|
|
|
|
|
|
MODULE_DESCRIPTION("Network Block Device");
|
|
|
|
MODULE_LICENSE("GPL");
|
|
|
|
|
2005-05-01 17:59:07 +02:00
|
|
|
module_param(nbds_max, int, 0444);
|
NBD: add partition support
Permit the use of partitions with network block devices (NBD).
A new parameter is introduced to define how many partition we want to be able
to manage per network block device. This parameter is "max_part".
For instance, to manage 63 partitions / loop device, we will do:
[on the server side]
# nbd-server 1234 /dev/sdb
[on the client side]
# modprobe nbd max_part=63
# ls -l /dev/nbd*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:14 /dev/nbd0
brw-rw---- 1 root disk 43, 64 2008-03-25 11:11 /dev/nbd1
brw-rw---- 1 root disk 43, 640 2008-03-25 11:11 /dev/nbd10
brw-rw---- 1 root disk 43, 704 2008-03-25 11:11 /dev/nbd11
brw-rw---- 1 root disk 43, 768 2008-03-25 11:11 /dev/nbd12
brw-rw---- 1 root disk 43, 832 2008-03-25 11:11 /dev/nbd13
brw-rw---- 1 root disk 43, 896 2008-03-25 11:11 /dev/nbd14
brw-rw---- 1 root disk 43, 960 2008-03-25 11:11 /dev/nbd15
brw-rw---- 1 root disk 43, 128 2008-03-25 11:11 /dev/nbd2
brw-rw---- 1 root disk 43, 192 2008-03-25 11:11 /dev/nbd3
brw-rw---- 1 root disk 43, 256 2008-03-25 11:11 /dev/nbd4
brw-rw---- 1 root disk 43, 320 2008-03-25 11:11 /dev/nbd5
brw-rw---- 1 root disk 43, 384 2008-03-25 11:11 /dev/nbd6
brw-rw---- 1 root disk 43, 448 2008-03-25 11:11 /dev/nbd7
brw-rw---- 1 root disk 43, 512 2008-03-25 11:11 /dev/nbd8
brw-rw---- 1 root disk 43, 576 2008-03-25 11:11 /dev/nbd9
# nbd-client localhost 1234 /dev/nbd0
Negotiation: ..size = 80418240KB
bs=1024, sz=80418240
-------NOTE, RFC: partition table is not automatically read.
The driver sets bdev->bd_invalidated to 1 to force the read of the partition
table of the device, but this is done only on an open of the device.
So we have to do a "touch /dev/nbdX" or something like that.
It can't be done from the nbd-client or nbd driver because at this
level we can't ask to read the partition table and to serve the request
at the same time (-> deadlock)
If someone has a better idea, I'm open to any suggestion.
-------NOTE, RFC
# fdisk -l /dev/nbd0
Disk /dev/nbd0: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/nbd0p1 * 1 9965 80043831 83 Linux
/dev/nbd0p2 9966 10011 369495 5 Extended
/dev/nbd0p5 9966 10011 369463+ 82 Linux swap / Solaris
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
brw-rw---- 1 root disk 43, 1 2008-03-25 11:16 /dev/nbd0p1
brw-rw---- 1 root disk 43, 2 2008-03-25 11:16 /dev/nbd0p2
brw-rw---- 1 root disk 43, 5 2008-03-25 11:16 /dev/nbd0p5
# mount /dev/nbd0p1 /mnt
# ls /mnt
bin dev initrd lost+found opt sbin sys var
boot etc initrd.img media proc selinux tmp vmlinuz
cdrom home lib mnt root srv usr
# umount /mnt
# nbd-client -d /dev/nbd0
# ls -l /dev/nbd0*
brw-rw---- 1 root disk 43, 0 2008-03-25 11:16 /dev/nbd0
-------NOTE
On "nbd-client -d", we can do an iocl(BLKRRPART) to update partition table:
as the size of the device is 0, we don't have to serve the partition manager
request (-> no deadlock).
-------NOTE
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-29 10:02:51 +02:00
|
|
|
MODULE_PARM_DESC(nbds_max, "number of network block devices to initialize (default: 16)");
|
|
|
|
module_param(max_part, int, 0444);
|
|
|
|
MODULE_PARM_DESC(max_part, "number of partitions per device (default: 0)");
|
2005-04-17 00:20:36 +02:00
|
|
|
#ifndef NDEBUG
|
|
|
|
module_param(debugflags, int, 0644);
|
|
|
|
MODULE_PARM_DESC(debugflags, "flags for controlling debug output");
|
|
|
|
#endif
|