Currently, when vhost_init_used() fails the sock refcnt and ubufs were
leaked. Correct this by calling vhost_init_used() before assign ubufs and
restore the oldsock when it fails.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Zero copy TX has been around for a while now.
We seem to be down to eliminating theoretical bugs
and performance tuning at this point:
it's probably time to enable it by default so that
most users get the benefit.

Keep the flag around meanwhile so users can experiment
with disabling this if they experience regressions.
I expect that we will remove it in the future.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

For short packets zerocopy mode adds overhead
of managing heads which isn't necessary: we
could simly update used ring directly
same as with zerocopy disabled.

Things seem to run a bit faster if we detect
and bypass head management when zcopy isn't used.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

When memory map changes, we need to flush outstanding
DMAs as they might in theory reference old memory addresses.
To do this simply stop initiating new DMAs
and wait for ubufs ref count to drop to 0.
Afterwards reset the count back to 1.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

vring changes already do a flush internally where appropriate, so we do
not need a second flush.

It's currently not very expensive but a follow-up patch makes flush more
heavy-weight, so remove the extra flush here to avoid regressing
performance if call or kick fds are changed on data path.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

These packet counters are used to drive the zercopy
selection heuristic so nothing too bad happens if they are off a bit -
and they are also reset once in a while.
But it's cleaner to clear them when backend is set so that
we start in a known state.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It seems that to avoid deadlocks it is enough to poll vq before
 we are going to use the last buffer.  This is faster than
c70aa540.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Even when vhost-net is in zero-copy transmit mode,
net core might still decide to copy the skb later
which is somewhat slower than a copy in user
context: data copy overhead is added to the cost of
page pin/unpin. The result is that enabling tx zero copy
option leads to higher CPU utilization for guest to guest
and guest to host traffic.

To fix this, suppress zero copy tx after a given number of
packets triggered late data copy. Re-enable periodically
to detect workload changes.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Zerocopy handling code is vhost-net specific.
Move it from vhost.c/vhost.h out to net.c
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Better document macros for DMA tracking. Add an
explicit one for DMA in progress instead of
relying on user supplying len != 1.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We copy head count to a 16 bit field, this works by chance on LE but on
BE guest gets 0. Fix it up.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Tested-by: NAlexander Graf <agraf@suse.de>
Cc: stable@vger.kernel.org
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In order for other vhost devices to use the VHOST_FEATURES bits the
vhost-net specific bits need to be moved to their own VHOST_NET_FEATURES
constant.

(Asias: Update drivers/vhost/test.c to use VHOST_NET_FEATURES)
Signed-off-by: NStefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Cc: Zhi Yong Wu <wuzhy@cn.ibm.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Asias He <asias@redhat.com>
Signed-off-by: NNicholas A. Bellinger <nab@risingtidesystems.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Take vlan header length into account, when vlan id is stored as
vlan_tci. Otherwise tagged packets coming from macvtap will be
truncated.
Signed-off-by: NBasil Gor <basil.gor@gmail.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a packet were fully copied in zerocopy, we don't wait for the DMA done to
mark the done flag, so after the packet were passed to lower device, we need to
add used and signal guest immediately.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Currently, we restart tx polling unconditionally when sendmsg()
fails. This would cause unnecessary wakeups of vhost wokers and waste
cpu utlization when evil userspace(guest driver) is able to hit EFAULT or
EINVAL.

The polling is only needed when the socket send buffer were exceeded or not
enough memory. So fix this by restarting polling only when sendmsg() returns
EAGAIN/ENOBUFS.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

When we want to disable vhost_net backend while there's a tx work, a possible
NULL pointer defernece may happen we we try to deference the vq->bufs after
vhost_net_set_backend() assign a NULL to it.

As suggested by Michael, fix this by checking the vq->bufs instead of
vhost_sock_zcopy().
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

The skb struct ubuf_info callback gets passed struct ubuf_info
itself, not the arg value as the field name and the function signature
seem to imply. Rename the arg field to ctx to match usage,
add documentation and change the callback argument type
to make usage clear and to have compiler check correctness.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We shouldn't hold any locks on release path. Pass a flag to
vhost_dev_cleanup to use the lockdep info correctly.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Tested-by: NSasha Levin <levinsasha928@gmail.com>

By adding some module aliases, programs (or users) won't have to explicitly
call modprobe. Vhost-net will always be available if built into the kernel.
It does require assigning a permanent minor number for depmod to work.

Also:
  - use C99 style initialization.
  - add missing entry in documentation for loop-control
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Acked-By: NKay Sievers <kay.sievers@vrfy.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The meth for calculating the # of outstanding buffers gives
incorrect results when vq->upend_idx wraps around zero.
Fix that.
Signed-off-by: NShirley Ma <xma@us.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

On backend change, we flushed out outstanding skbs
but forgot to update the used ring, so that
done entries were left in the ubuf_info ring.
As a result we lose heads or complete incorrect ones,
crashing the guest or leaking memory.
Fix by updating the used ring.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Move the used ring initialization after backend was set. This
makes it possible to disable the backend and tweak the used ring,
then restart. This will also make it possible to log the used ring
write correctly.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

>From: Shirley Ma <mashirle@us.ibm.com>

This adds experimental zero copy support in vhost-net,
disabled by default. To enable, set
experimental_zcopytx module option to 1.

This patch maintains the outstanding userspace buffers in the
sequence it is delivered to vhost. The outstanding userspace buffers
will be marked as done once the lower device buffers DMA has finished.
This is monitored through last reference of kfree_skb callback. Two
buffer indices are used for this purpose.

The vhost-net device passes the userspace buffers info to lower device
skb through message control. DMA done status check and guest
notification are handled by handle_tx: in the worst case is all buffers
in the vq are in pending/done status, so we need to notify guest to
release DMA done buffers first before we get any new buffers from the
vq.

One known problem is that if the guest stops submitting
buffers, buffers might never get used until some
further action, e.g. device reset. This does not
seem to affect linux guests.
Signed-off-by: NShirley <xma@us.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Support the new event index feature. When acked,
utilize it to reduce the # of interrupts sent to the guest.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NRusty Russell <rusty@rustcorp.com.au>

Use of skb_queue_empty(&sock->sk->sk_receive_queue)
without taking the sk_receive_queue.lock is unsafe
or useless. Take it out.
Reported-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

vhost takes a sock lock to try and prevent
the skb from being pulled from the receive queue
after skb_peek.  However this is not the right lock to use for that,
sk_receive_queue.lock is. Fix that up.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Codes duplication were found between the handling of mergeable and big
buffers, so this patch tries to unify them. This could be easily done
by adding a quota to the get_rx_bufs() which is used to limit the
number of buffers it returns (for mergeable buffer, the quota is
simply UIO_MAXIOV, for big buffers, the quota is just 1), and then the
previous handle_rx_mergeable() could be resued also for big buffers.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

No need to check the support of mergeable buffer inside the recevie
loop as the whole handle_rx()_xx is in the read critical region.  So
this patch move it ahead of the receiving loop.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Minor cleanup of vhost.c and net.c to match coding style.
Signed-off-by: NKrishna Kumar <krkumar2@in.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

When built with rcu checks enabled, vhost triggers
bogus warnings as vhost features are read without
dev->mutex sometimes, and private pointer is read
with our kind of rcu where work serves as a
read side critical section.

Fixing it properly is not trivial.
Disable the warnings by stubbing out the checks for now.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Incorrect rcu check was used as rcu isn't done
under mutex here. Force check to 1 for now,
to stop it from complaining.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Move use/unuse mm to vhost.c which makes it possible to batch these
operations.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

All file_operations should get a .llseek operation so we can make
nonseekable_open the default for future file operations without a
.llseek pointer.

The three cases that we can automatically detect are no_llseek, seq_lseek
and default_llseek. For cases where we can we can automatically prove that
the file offset is always ignored, we use noop_llseek, which maintains
the current behavior of not returning an error from a seek.

New drivers should normally not use noop_llseek but instead use no_llseek
and call nonseekable_open at open time.  Existing drivers can be converted
to do the same when the maintainer knows for certain that no user code
relies on calling seek on the device file.

The generated code is often incorrectly indented and right now contains
comments that clarify for each added line why a specific variant was
chosen. In the version that gets submitted upstream, the comments will
be gone and I will manually fix the indentation, because there does not
seem to be a way to do that using coccinelle.

Some amount of new code is currently sitting in linux-next that should get
the same modifications, which I will do at the end of the merge window.

Many thanks to Julia Lawall for helping me learn to write a semantic
patch that does all this.

===== begin semantic patch =====
// This adds an llseek= method to all file operations,
// as a preparation for making no_llseek the default.
//
// The rules are
// - use no_llseek explicitly if we do nonseekable_open
// - use seq_lseek for sequential files
// - use default_llseek if we know we access f_pos
// - use noop_llseek if we know we don't access f_pos,
//   but we still want to allow users to call lseek
//
@ open1 exists @
identifier nested_open;
@@
nested_open(...)
{
<+...
nonseekable_open(...)
...+>
}

@ open exists@
identifier open_f;
identifier i, f;
identifier open1.nested_open;
@@
int open_f(struct inode *i, struct file *f)
{
<+...
(
nonseekable_open(...)
|
nested_open(...)
)
...+>
}

@ read disable optional_qualifier exists @
identifier read_f;
identifier f, p, s, off;
type ssize_t, size_t, loff_t;
expression E;
identifier func;
@@
ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off)
{
<+...
(
   *off = E
|
   *off += E
|
   func(..., off, ...)
|
   E = *off
)
...+>
}

@ read_no_fpos disable optional_qualifier exists @
identifier read_f;
identifier f, p, s, off;
type ssize_t, size_t, loff_t;
@@
ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off)
{
... when != off
}

@ write @
identifier write_f;
identifier f, p, s, off;
type ssize_t, size_t, loff_t;
expression E;
identifier func;
@@
ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off)
{
<+...
(
  *off = E
|
  *off += E
|
  func(..., off, ...)
|
  E = *off
)
...+>
}

@ write_no_fpos @
identifier write_f;
identifier f, p, s, off;
type ssize_t, size_t, loff_t;
@@
ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off)
{
... when != off
}

@ fops0 @
identifier fops;
@@
struct file_operations fops = {
 ...
};

@ has_llseek depends on fops0 @
identifier fops0.fops;
identifier llseek_f;
@@
struct file_operations fops = {
...
 .llseek = llseek_f,
...
};

@ has_read depends on fops0 @
identifier fops0.fops;
identifier read_f;
@@
struct file_operations fops = {
...
 .read = read_f,
...
};

@ has_write depends on fops0 @
identifier fops0.fops;
identifier write_f;
@@
struct file_operations fops = {
...
 .write = write_f,
...
};

@ has_open depends on fops0 @
identifier fops0.fops;
identifier open_f;
@@
struct file_operations fops = {
...
 .open = open_f,
...
};

// use no_llseek if we call nonseekable_open
////////////////////////////////////////////
@ nonseekable1 depends on !has_llseek && has_open @
identifier fops0.fops;
identifier nso ~= "nonseekable_open";
@@
struct file_operations fops = {
...  .open = nso, ...
+.llseek = no_llseek, /* nonseekable */
};

@ nonseekable2 depends on !has_llseek @
identifier fops0.fops;
identifier open.open_f;
@@
struct file_operations fops = {
...  .open = open_f, ...
+.llseek = no_llseek, /* open uses nonseekable */
};

// use seq_lseek for sequential files
/////////////////////////////////////
@ seq depends on !has_llseek @
identifier fops0.fops;
identifier sr ~= "seq_read";
@@
struct file_operations fops = {
...  .read = sr, ...
+.llseek = seq_lseek, /* we have seq_read */
};

// use default_llseek if there is a readdir
///////////////////////////////////////////
@ fops1 depends on !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier readdir_e;
@@
// any other fop is used that changes pos
struct file_operations fops = {
... .readdir = readdir_e, ...
+.llseek = default_llseek, /* readdir is present */
};

// use default_llseek if at least one of read/write touches f_pos
/////////////////////////////////////////////////////////////////
@ fops2 depends on !fops1 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier read.read_f;
@@
// read fops use offset
struct file_operations fops = {
... .read = read_f, ...
+.llseek = default_llseek, /* read accesses f_pos */
};

@ fops3 depends on !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier write.write_f;
@@
// write fops use offset
struct file_operations fops = {
... .write = write_f, ...
+	.llseek = default_llseek, /* write accesses f_pos */
};

// Use noop_llseek if neither read nor write accesses f_pos
///////////////////////////////////////////////////////////

@ fops4 depends on !fops1 && !fops2 && !fops3 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier read_no_fpos.read_f;
identifier write_no_fpos.write_f;
@@
// write fops use offset
struct file_operations fops = {
...
 .write = write_f,
 .read = read_f,
...
+.llseek = noop_llseek, /* read and write both use no f_pos */
};

@ depends on has_write && !has_read && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier write_no_fpos.write_f;
@@
struct file_operations fops = {
... .write = write_f, ...
+.llseek = noop_llseek, /* write uses no f_pos */
};

@ depends on has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
identifier read_no_fpos.read_f;
@@
struct file_operations fops = {
... .read = read_f, ...
+.llseek = noop_llseek, /* read uses no f_pos */
};

@ depends on !has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @
identifier fops0.fops;
@@
struct file_operations fops = {
...
+.llseek = noop_llseek, /* no read or write fn */
};
===== End semantic patch =====
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Cc: Julia Lawall <julia@diku.dk>
Cc: Christoph Hellwig <hch@infradead.org>

Qemu supports up to UIO_MAXIOV s/g so we have to match that because guest
drivers may rely on this.

Allocate indirect and log arrays dynamically to avoid using too much contigious
memory and make the length of hdr array to match the header length since each
iovec entry has a least one byte.

Test with copying large files w/ and w/o migration in both linux and windows
guests.
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

In mergeable buffer case, we use headcount, log_num
and seg as indexes in same-size arrays, and
we know that headcount <= seg and
log_num equals either 0 or seg.

Therefore, the right thing to do is range-check seg,
not headcount as we do now: these will be different
if guest chains s/g descriptors (this does not
happen now, but we can not trust the guest).

Long term, we should add BUG_ON checks to verify
two other indexes are what we think they should be.
Reported-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Also add rcu_dereference_protected() for code paths where locks are held.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>

This adds support for mergeable buffers in vhost-net: this is needed
for older guests without indirect buffer support, as well
as for zero copy with some devices.

Includes changes by Michael S. Tsirkin to make the
patch as low risk as possible (i.e., close to no changes
when feature is disabled).
Signed-off-by: NDavid Stevens <dlstevens@us.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

Replace vhost_workqueue with per-vhost kthread.  Other than callback
argument change from struct work_struct * to struct vhost_work *,
there's no visible change to vhost_poll_*() interface.

This conversion is to make each vhost use a dedicated kthread so that
resource control via cgroup can be applied.

Partially based on Sridhar Samudrala's patch.

* Updated to use sub structure vhost_work instead of directly using
  vhost_poll at Michael's suggestion.

* Added flusher wake_up() optimization at Michael's suggestion.

Changes by MST:
* Converted atomics/barrier use to a spinlock.
* Create thread on SET_OWNER
* Fix flushing
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Cc: Sridhar Samudrala <samudrala.sridhar@gmail.com>

Reported by Stephen Rothwell.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>