On Sat, Oct 11, 2014 at 03:16:44PM +0800, Jason Wang wrote:
Below should be useful for some experiments Jason is doing.
I thought I'd send it out for early review/feedback.
event idx feature allows us to defer interrupts until
a specific # of descriptors were used.
Sometimes it might be
On restore, virtio pci does the following:
+ set features
+ init vqs etc - device can be used at this point!
+ set ACKNOWLEDGE,DRIVER and DRIVER_OK status bits
This is in violation of the virtio spec, which
requires the following order:
- ACKNOWLEDGE
- DRIVER
- init vqs
- DRIVER_OK
This
Defer config changed notifications that arrive during
probe/scan/freeze/restore.
This will allow drivers to set DRIVER_OK earlier, without worrying about
racing with config change interrupts.
This change will also benefit old hypervisors (before 2009)
that send interrupts without checking
Rusty, please review this, and consider for this merge window.
This fixes the following virtio spec compliance issues:
1. on restore, drivers use device before setting ACKNOWLEDGE and DRIVER bits
2. on probe, drivers aren't prepared to handle config interrupts
arriving before probe returns
3.
config_mutex served two purposes: prevent multiple concurrent config
change handlers, and synchronize access to config_enable flag.
Since commit dbf2576e37da0fcc7aacbfbb9fd5d3de7888a3c1
workqueue: make all workqueues non-reentrant
all workqueues are non-reentrant, and config_enable
is now
Replace duplicated code in all transports with a single wrapper in
virtio.c.
The only functional change is in virtio_mmio.c: if a buggy device sends
us an interrupt before driver is set, we previously returned IRQ_NONE,
now we return IRQ_HANDLED.
As this must not happen in practice, this does
Now that virtio core ensures config changes don't
arrive during probing, drop config_enable flag
in virtio blk.
On removal, flush is now sufficient to guarantee that
no change work is queued.
This help simplify the driver, and will allow
setting DRIVER_OK earlier without losing config
change
This is in preparation to extending config changed event handling
in core.
Wrapping these in an API also seems to make for a cleaner code.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
Reviewed-by: Cornelia Huck cornelia.h...@de.ibm.com
---
include/linux/virtio.h | 6 +
config_mutex served two purposes: prevent multiple concurrent config
change handlers, and synchronize access to config_enable flag.
Since commit dbf2576e37da0fcc7aacbfbb9fd5d3de7888a3c1
workqueue: make all workqueues non-reentrant
all workqueues are non-reentrant, and config_enable
is now
goto done;
done:
return;
is ugly, it was put there to make diff review easier.
replace by open-coded return.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
Acked-by: Cornelia Huck cornelia.h...@de.ibm.com
---
drivers/net/virtio_net.c | 6 ++
1 file changed, 2
virtio spec 0.9.X requires DRIVER_OK to be set before
VQs are used, but some drivers use VQs before probe
function returns.
Since DRIVER_OK is set after probe, this violates the spec.
Even though under virtio 1.0 transitional devices support this
behaviour, we want to make it possible for those
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after probe returns, virtio block violated this
rule by calling add_disk, which causes the VQ to be used directly within
probe.
To fix, call virtio_enable_vqs_early before using VQs.
Signed-off-by: Michael
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after probe returns, but virtio 9p device
adds self to channel list within probe, at which point VQ can be
used in violation of the spec.
To fix, call virtio_enable_vqs_early before using VQs.
In the extremely unlikely event that driver initialization fails after
RX buffers are added, virtio net frees RX buffers while VQs are
still active, potentially causing device to use a freed buffer.
To fix, reset device first - same as we do on device removal.
Signed-off-by: Michael S. Tsirkin
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after probe returns, virtio net violated this
rule by using receive VQs within probe.
To fix, call virtio_enable_vqs_early before using VQs.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
Reviewed-by:
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after probe returns, virtio console violated this
rule by adding inbufs, which causes the VQ to be used directly within
probe.
To fix, call virtio_enable_vqs_early before using VQs.
Signed-off-by: Michael
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after resume returns, virtio console violated this
rule by adding inbufs, which causes the VQ to be used directly within
restore.
To fix, call virtio_enable_vqs_early before using VQs.
Signed-off-by:
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after restore returns, virtio net violated this
rule by using receive VQs within restore.
To fix, call virtio_enable_vqs_early before using VQs.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after restore returns, virtio scsi violated
this rule on restore by kicking event vq within restore.
To fix, call virtio_enable_vqs_early before using event queue.
Signed-off-by: Michael S. Tsirkin
We currently kick event within virtscsi_init,
before host is fully initialized.
This can in theory confuse guest if device
consumes the buffers immediately.
To fix, move virtscsi_kick_event_all out to scan/restore.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after restore returns, virtio block violated
this rule on restore by restarting queues, which might in theory
cause the VQ to be used directly within restore.
To fix, call virtio_enable_vqs_early before
We cancel event work on device removal, but an interrupt
could trigger immediately after this, and queue it
again.
To fix, set a flag.
Loosely based on patch by Paolo Bonzini
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
From: Paolo Bonzini pbonz...@redhat.com
Michael S. Tsirkin noticed a race condition:
we reset device on freeze, but system WQ is still
running so it might try adding bufs to a VQ meanwhile.
To fix, switch to handling events from the freezable WQ.
Reported-by: Michael S. Tsirkin m...@redhat.com
virtio spec requires drivers to set DRIVER_OK before using VQs.
This is set automatically after resume returns, virtio balloon
violated this rule by adding bufs, which causes the VQ to be used
directly within restore.
To fix, call virtio_enable_vqs_early before using VQ.
Signed-off-by: Michael
Code like
vi-vq = NULL;
kfree(vi)
does not make sense.
Clean it up, use goto error labels for cleanup.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
drivers/char/hw_random/virtio-rng.c | 15 +--
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git
Enable VQs early like we do for restore.
This makes it possible to drop the scan callback,
moving scanning into the probe function, and making
code simpler.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
drivers/scsi/virtio_scsi.c | 23 +++
1 file changed, 7
26 matches
Mail list logo
