We need a full barrier after writing out event index, using
virt_store_mb there seems better than open-coding. As usual, we need a
wrapper to account for strong barriers.
It's tempting to use this in vhost as well, for that, we'll
need a variant of smp_store_mb that works on __user pointers.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
virtio ring uses smp_wmb on SMP and wmb on !SMP,
the reason for the later being that it might be
talking to another kernel on the same SMP machine.
This is exactly what virt_xxx barriers do,
so switch to these instead of homegrown ifdef hacks.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Alexander Duyck <alexander.duyck@gmail.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
This reverts commit 9e1a27ea42.
While that commit optimizes !CONFIG_SMP, it mixes
up DMA and SMP concepts, making the code hard
to figure out.
A better way to optimize this is with the new __smp_XXX
barriers.
As a first step, go back to full rmb/wmb barriers
for !SMP.
We switch to __smp_XXX barriers in the next patch.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Alexander Duyck <alexander.duyck@gmail.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
This change makes it so that instead of using smp_wmb/rmb which varies
depending on the kernel configuration we can can use dma_wmb/rmb which for
most architectures should be equal to or slightly more strict than
smp_wmb/rmb.
The advantage to this is that these barriers are available to uniprocessor
builds as well so the performance should improve under such a
configuration.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Currently a host kick error is silently ignored and not reflected in
the virtqueue of a particular virtio device.
Changing the notify API for guest->host notification seems to be one
prerequisite in order to be able to handle such errors in the context
where the kick is triggered.
This patch changes the notify API. The notify function must return a
bool return value. It returns false if the host notification failed.
Signed-off-by: Heinz Graalfs <graalfs@linux.vnet.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
virtio_ring.h uses mb() and friends, make
it pull in asm/barrier.h itself, not rely
on other headers to do it.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Michael Kerrisk <mtk.manpages@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
Instead of storing the queue index in transport-specific virtio structs,
this patch moves them to vring_virtqueue and introduces an helper to get
the value. This lets drivers simplify their management and tracing of
virtqueues.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We were cheating with our barriers; using the smp ones rather than the
real device ones. That was fine, until rpmsg came along, which is
used to talk to a real device (a non-SMP CPU).
Unfortunately, just putting back the real barriers (reverting
d57ed95d) causes a performance regression on virtio-pci. In
particular, Amos reports netbench's TCP_RR over virtio_net CPU
utilization increased up to 35% while throughput went down by up to
14%.
By comparison, this branch is in the noise.
Reference: https://lkml.org/lkml/2011/12/11/22
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Based on the layout description in the comments, take account of
the *_event_idx in functions vring_init and vring_size.
Signed-off-by: Wang Sheng-Hui <shhuiw@gmail.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
With the new used_event and avail_event and features, both
host and guest need similar logic to check whether events are
enabled, so it helps to put the common code in the header.
Note that Xen has similar logic for notification hold-off
in include/xen/interface/io/ring.h with req_event and req_prod
corresponding to event_idx + 1 and new_idx respectively.
+1 comes from the fact that req_event and req_prod in Xen start at 1,
while event index in virtio starts at 0.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Define a new feature bit for the guest and host to utilize
an event index (like Xen) instead if a flag bit to enable/disable
interrupts and kicks.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
It's unclear to me if it's important, but it's obviously causing my
technical colleages some headaches and I'd hate such imprecision to
slow virtio adoption.
I've emailed this to all non-trivial contributors for approval, too.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Acked-by: Ryan Harper <ryanh@us.ibm.com>
Acked-by: Anthony Liguori <aliguori@us.ibm.com>
Acked-by: Eric Van Hensbergen <ericvh@gmail.com>
Acked-by: john cooper <john.cooper@redhat.com>
Acked-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Fernando Luis Vazquez Cao <fernando@oss.ntt.co.jp>
I've been doing this for years, and akpm picked me up on it about 12
months ago. lguest partly serves as example code, so let's do it Right.
Also, remove two unused fields in struct vblk_info in the example launcher.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Ingo Molnar <mingo@redhat.com>
Add a new feature flag for indirect ring entries. These are ring
entries which point to a table of buffer descriptors.
The idea here is to increase the ring capacity by allowing a larger
effective ring size whereby the ring size dictates the number of
requests that may be outstanding, rather than the size of those
requests.
This should be most effective in the case of block I/O where we can
potentially benefit by concurrently dispatching a large number of
large requests. Even in the simple case of single segment block
requests, this results in a threefold increase in ring capacity.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Add a linked list of all virtqueues for a virtio device: this helps for
debugging and is also needed for upcoming interface change.
Also, add a "name" field for clearer debug messages.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This allows each virtio user to hand in the alignment appropriate to
their virtio_ring structures.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
It's really the alignment desired for consumer/producer separation;
historically this x86 pagesize, but with PowerPC it'll still be x86
pagesize. And in theory lguest could choose a different value.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
To prepare for virtio_ring transport feature bits, hook in a call in
all the users to manipulate them. This currently just clears all the
bits, since it doesn't understand any features.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The other side (host) can set the NO_NOTIFY flag as an optimization,
to say "no need to kick me when you add things". Make it clear that
this is advisory only; especially that we should always notify when
the ring is full.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Using unsigned int resulted in silent truncation of the upper 32-bit
on x86_64 resulting in an OOPS since the ring was being initialized
wrong.
Please reconsider my previous patch to just use PAGE_ALIGN(). Open
coding this sort of stuff, no matter how simple it seems, is just
asking for this sort of trouble.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
It seems that virtio_net wants to disable callbacks (interrupts) before
calling netif_rx_schedule(), so we can't use the return value to do so.
Rename "restart" to "cb_enable" and introduce "cb_disable" hook: callback
now returns void, rather than a boolean.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The virtio descriptor rings of size N-1 were nicely set up to be
aligned to an N-byte boundary. But as Anthony Liguori points out, the
free-running indices used by virtio require that the sizes be a power
of 2, otherwise we get problems on wrap (demonstrated with lguest).
So we replace the clever "2^n-1" scheme with a simple "align to page
boundary" scheme: this means that all virtio rings take at least two
pages, but it's safer than guessing cache alignment.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This patch fixes a typo in vring_init(). This happens to work today in lguest
because the sizeof(struct vring_desc) is 16 and struct vring contains 3
pointers and an unsigned int so on 32-bit
sizeof(struct vring_desc) == sizeof(struct vring). However, this is no longer
true on 64-bit where the bug is exposed.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
These helper routines supply most of the virtqueue_ops for hypervisors
which want to use a ring for virtio. Unlike the previous lguest
implementation:
1) The rings are variable sized (2^n-1 elements).
2) They have an unfortunate limit of 65535 bytes per sg element.
3) The page numbers are always 64 bit (PAE anyone?)
4) They no longer place used[] on a separate page, just a separate
cacheline.
5) We do a modulo on a variable. We could be tricky if we cared.
6) Interrupts and notifies are suppressed using flags within the rings.
Users need only get the ring pages and provide a notify hook (KVM
wants the guest to allocate the rings, lguest does it sanely).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Dor Laor <dor.laor@qumranet.com>
Michael S. Tsirkin