This patch validates feature dependencies during probe and fail the probing
if a dependency is missed. This fixes the issues of hitting BUG()
when qemu fails to advertise features correctly. One example is booting
guest with ctrl_vq=off through qemu.
Cc: Rusty Russell <[email protected]>
Cc: Michael S. Tsirkin <[email protected]>
Cc: Cornelia Huck <[email protected]>
Cc: Wanlong Gao <[email protected]>
Signed-off-by: Jason Wang <[email protected]>
---
Changes from V1:
- Drop VIRTIO_NET_F_*_UFO from the checklist, since it was disabled
---
drivers/net/virtio_net.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 91 insertions(+)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index ec2a8b4..b16a761 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -1673,6 +1673,93 @@ static const struct attribute_group virtio_net_mrg_rx_group = {
};
#endif
+static int virtnet_validate_features(struct virtio_device *dev,
+ unsigned int *table,
+ int table_size,
+ unsigned int feature)
+{
+ int i;
+
+ if (!virtio_has_feature(dev, feature)) {
+ for (i = 0; i < table_size; i++) {
+ unsigned int f = table[i];
+
+ if (virtio_has_feature(dev, f)) {
+ dev_err(&dev->dev,
+ "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x was not",
+ f, feature);
+ return -EINVAL;
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int virtnet_check_features(struct virtio_device *dev)
+{
+ unsigned int features_for_ctrl_vq[] = {
+ VIRTIO_NET_F_CTRL_RX,
+ VIRTIO_NET_F_CTRL_VLAN,
+ VIRTIO_NET_F_GUEST_ANNOUNCE,
+ VIRTIO_NET_F_MQ,
+ VIRTIO_NET_F_CTRL_MAC_ADDR
+ };
+ unsigned int features_for_guest_csum[] = {
+ VIRTIO_NET_F_GUEST_TSO4,
+ VIRTIO_NET_F_GUEST_TSO6,
+ VIRTIO_NET_F_GUEST_ECN,
+ };
+ unsigned int features_for_host_csum[] = {
+ VIRTIO_NET_F_HOST_TSO4,
+ VIRTIO_NET_F_HOST_TSO6,
+ VIRTIO_NET_F_HOST_ECN,
+ };
+ int err;
+
+ err = virtnet_validate_features(dev, features_for_ctrl_vq,
+ ARRAY_SIZE(features_for_ctrl_vq),
+ VIRTIO_NET_F_CTRL_VQ);
+ if (err)
+ return err;
+
+ err = virtnet_validate_features(dev, features_for_guest_csum,
+ ARRAY_SIZE(features_for_guest_csum),
+ VIRTIO_NET_F_GUEST_CSUM);
+ if (err)
+ return err;
+
+ err = virtnet_validate_features(dev, features_for_host_csum,
+ ARRAY_SIZE(features_for_host_csum),
+ VIRTIO_NET_F_CSUM);
+ if (err)
+ return err;
+
+ if (virtio_has_feature(dev, VIRTIO_NET_F_GUEST_ECN) &&
+ (!virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO4) ||
+ !virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO6))) {
+ dev_err(&dev->dev,
+ "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
+ VIRTIO_NET_F_GUEST_ECN,
+ VIRTIO_NET_F_GUEST_TSO4,
+ VIRTIO_NET_F_GUEST_TSO6);
+ return -EINVAL;
+ }
+
+ if (virtio_has_feature(dev, VIRTIO_NET_F_HOST_ECN) &&
+ (!virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO4) ||
+ !virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO6))) {
+ dev_err(&dev->dev,
+ "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
+ VIRTIO_NET_F_HOST_ECN,
+ VIRTIO_NET_F_HOST_TSO4,
+ VIRTIO_NET_F_HOST_TSO6);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
static int virtnet_probe(struct virtio_device *vdev)
{
int i, err;
@@ -1680,6 +1767,10 @@ static int virtnet_probe(struct virtio_device *vdev)
struct virtnet_info *vi;
u16 max_queue_pairs;
+ err = virtnet_check_features(vdev);
+ if (err)
+ return -EINVAL;
+
/* Find if host supports multiqueue virtio_net device */
err = virtio_cread_feature(vdev, VIRTIO_NET_F_MQ,
struct virtio_net_config,
--
1.9.1
On Wed, 19 Nov 2014 15:21:29 +0800
Jason Wang <[email protected]> wrote:
> This patch validates feature dependencies during probe and fail the probing
> if a dependency is missed. This fixes the issues of hitting BUG()
> when qemu fails to advertise features correctly. One example is booting
> guest with ctrl_vq=off through qemu.
>
> Cc: Rusty Russell <[email protected]>
> Cc: Michael S. Tsirkin <[email protected]>
> Cc: Cornelia Huck <[email protected]>
> Cc: Wanlong Gao <[email protected]>
> Signed-off-by: Jason Wang <[email protected]>
> ---
> Changes from V1:
> - Drop VIRTIO_NET_F_*_UFO from the checklist, since it was disabled
> ---
> drivers/net/virtio_net.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 91 insertions(+)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index ec2a8b4..b16a761 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1673,6 +1673,93 @@ static const struct attribute_group virtio_net_mrg_rx_group = {
> };
> #endif
>
> +static int virtnet_validate_features(struct virtio_device *dev,
> + unsigned int *table,
> + int table_size,
> + unsigned int feature)
> +{
> + int i;
> +
> + if (!virtio_has_feature(dev, feature)) {
Do an early return, get rid of one indentation level?
> + for (i = 0; i < table_size; i++) {
> + unsigned int f = table[i];
> +
> + if (virtio_has_feature(dev, f)) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x was not",
s/hyperviser/hypervisor/ (also below)
> + f, feature);
> + return -EINVAL;
> + }
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int virtnet_check_features(struct virtio_device *dev)
> +{
> + unsigned int features_for_ctrl_vq[] = {
> + VIRTIO_NET_F_CTRL_RX,
> + VIRTIO_NET_F_CTRL_VLAN,
> + VIRTIO_NET_F_GUEST_ANNOUNCE,
> + VIRTIO_NET_F_MQ,
> + VIRTIO_NET_F_CTRL_MAC_ADDR
> + };
> + unsigned int features_for_guest_csum[] = {
> + VIRTIO_NET_F_GUEST_TSO4,
> + VIRTIO_NET_F_GUEST_TSO6,
> + VIRTIO_NET_F_GUEST_ECN,
> + };
> + unsigned int features_for_host_csum[] = {
> + VIRTIO_NET_F_HOST_TSO4,
> + VIRTIO_NET_F_HOST_TSO6,
> + VIRTIO_NET_F_HOST_ECN,
> + };
I'm wondering whether it would be easier to read if you listed all
prereqs per feature instead of all features that depend on a feature?
It would still be hard to express the v4/v6 or conditions below in
tables, though.
Or call the arrays features_depending_on_foo?
> + int err;
> +
> + err = virtnet_validate_features(dev, features_for_ctrl_vq,
> + ARRAY_SIZE(features_for_ctrl_vq),
> + VIRTIO_NET_F_CTRL_VQ);
> + if (err)
> + return err;
If you already print a message that a user may use to fix their
hypervisor (or bug someone about it), would it make sense to check all
dependencies and print a full list of everything that is broken in the
advertised feature bits? I usually hate it if I fix one thing only to
hit the next bug when the program could have already told me about
everything I need to fix :)
> +
> + err = virtnet_validate_features(dev, features_for_guest_csum,
> + ARRAY_SIZE(features_for_guest_csum),
> + VIRTIO_NET_F_GUEST_CSUM);
> + if (err)
> + return err;
> +
> + err = virtnet_validate_features(dev, features_for_host_csum,
> + ARRAY_SIZE(features_for_host_csum),
> + VIRTIO_NET_F_CSUM);
> + if (err)
> + return err;
> +
> + if (virtio_has_feature(dev, VIRTIO_NET_F_GUEST_ECN) &&
> + (!virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO4) ||
> + !virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO6))) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
> + VIRTIO_NET_F_GUEST_ECN,
> + VIRTIO_NET_F_GUEST_TSO4,
> + VIRTIO_NET_F_GUEST_TSO6);
> + return -EINVAL;
> + }
> +
> + if (virtio_has_feature(dev, VIRTIO_NET_F_HOST_ECN) &&
> + (!virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO4) ||
> + !virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO6))) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
"Hypervisor bug: advertised feature <foo> but not <bar> or <baz>"
?
> + VIRTIO_NET_F_HOST_ECN,
> + VIRTIO_NET_F_HOST_TSO4,
> + VIRTIO_NET_F_HOST_TSO6);
> + return -EINVAL;
> + }
> +
> + return 0;
> +}
> +
> static int virtnet_probe(struct virtio_device *vdev)
> {
> int i, err;
On Wed, Nov 19, 2014 at 03:21:29PM +0800, Jason Wang wrote:
> This patch validates feature dependencies during probe and fail the probing
> if a dependency is missed. This fixes the issues of hitting BUG()
> when qemu fails to advertise features correctly. One example is booting
> guest with ctrl_vq=off through qemu.
>
> Cc: Rusty Russell <[email protected]>
> Cc: Michael S. Tsirkin <[email protected]>
> Cc: Cornelia Huck <[email protected]>
> Cc: Wanlong Gao <[email protected]>
> Signed-off-by: Jason Wang <[email protected]>
> ---
> Changes from V1:
> - Drop VIRTIO_NET_F_*_UFO from the checklist, since it was disabled
In this form, it's not 3.18 material anyway.
Let's just focus on fixing BUG that you see for now,
and for 3.19, let's fix UFO.
> ---
> drivers/net/virtio_net.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 91 insertions(+)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index ec2a8b4..b16a761 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1673,6 +1673,93 @@ static const struct attribute_group virtio_net_mrg_rx_group = {
> };
> #endif
>
> +static int virtnet_validate_features(struct virtio_device *dev,
> + unsigned int *table,
> + int table_size,
> + unsigned int feature)
> +{
> + int i;
> +
> + if (!virtio_has_feature(dev, feature)) {
> + for (i = 0; i < table_size; i++) {
> + unsigned int f = table[i];
> +
> + if (virtio_has_feature(dev, f)) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x was not",
> + f, feature);
> + return -EINVAL;
> + }
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int virtnet_check_features(struct virtio_device *dev)
> +{
> + unsigned int features_for_ctrl_vq[] = {
> + VIRTIO_NET_F_CTRL_RX,
> + VIRTIO_NET_F_CTRL_VLAN,
> + VIRTIO_NET_F_GUEST_ANNOUNCE,
> + VIRTIO_NET_F_MQ,
> + VIRTIO_NET_F_CTRL_MAC_ADDR
> + };
> + unsigned int features_for_guest_csum[] = {
> + VIRTIO_NET_F_GUEST_TSO4,
> + VIRTIO_NET_F_GUEST_TSO6,
> + VIRTIO_NET_F_GUEST_ECN,
> + };
> + unsigned int features_for_host_csum[] = {
> + VIRTIO_NET_F_HOST_TSO4,
> + VIRTIO_NET_F_HOST_TSO6,
> + VIRTIO_NET_F_HOST_ECN,
> + };
> + int err;
> +
> + err = virtnet_validate_features(dev, features_for_ctrl_vq,
> + ARRAY_SIZE(features_for_ctrl_vq),
> + VIRTIO_NET_F_CTRL_VQ);
> + if (err)
> + return err;
> +
> + err = virtnet_validate_features(dev, features_for_guest_csum,
> + ARRAY_SIZE(features_for_guest_csum),
> + VIRTIO_NET_F_GUEST_CSUM);
> + if (err)
> + return err;
> +
> + err = virtnet_validate_features(dev, features_for_host_csum,
> + ARRAY_SIZE(features_for_host_csum),
> + VIRTIO_NET_F_CSUM);
> + if (err)
> + return err;
> +
> + if (virtio_has_feature(dev, VIRTIO_NET_F_GUEST_ECN) &&
> + (!virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO4) ||
> + !virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO6))) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
> + VIRTIO_NET_F_GUEST_ECN,
> + VIRTIO_NET_F_GUEST_TSO4,
> + VIRTIO_NET_F_GUEST_TSO6);
> + return -EINVAL;
> + }
> +
> + if (virtio_has_feature(dev, VIRTIO_NET_F_HOST_ECN) &&
> + (!virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO4) ||
> + !virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO6))) {
> + dev_err(&dev->dev,
> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
> + VIRTIO_NET_F_HOST_ECN,
> + VIRTIO_NET_F_HOST_TSO4,
> + VIRTIO_NET_F_HOST_TSO6);
> + return -EINVAL;
> + }
> +
> + return 0;
> +}
> +
> static int virtnet_probe(struct virtio_device *vdev)
> {
> int i, err;
> @@ -1680,6 +1767,10 @@ static int virtnet_probe(struct virtio_device *vdev)
> struct virtnet_info *vi;
> u16 max_queue_pairs;
>
> + err = virtnet_check_features(vdev);
> + if (err)
> + return -EINVAL;
> +
> /* Find if host supports multiqueue virtio_net device */
> err = virtio_cread_feature(vdev, VIRTIO_NET_F_MQ,
> struct virtio_net_config,
> --
> 1.9.1
On 11/19/2014 04:54 PM, Cornelia Huck wrote:
> On Wed, 19 Nov 2014 15:21:29 +0800
> Jason Wang <[email protected]> wrote:
>
>> This patch validates feature dependencies during probe and fail the probing
>> if a dependency is missed. This fixes the issues of hitting BUG()
>> when qemu fails to advertise features correctly. One example is booting
>> guest with ctrl_vq=off through qemu.
>>
>> Cc: Rusty Russell <[email protected]>
>> Cc: Michael S. Tsirkin <[email protected]>
>> Cc: Cornelia Huck <[email protected]>
>> Cc: Wanlong Gao <[email protected]>
>> Signed-off-by: Jason Wang <[email protected]>
>> ---
>> Changes from V1:
>> - Drop VIRTIO_NET_F_*_UFO from the checklist, since it was disabled
>> ---
>> drivers/net/virtio_net.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 91 insertions(+)
>>
>> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
>> index ec2a8b4..b16a761 100644
>> --- a/drivers/net/virtio_net.c
>> +++ b/drivers/net/virtio_net.c
>> @@ -1673,6 +1673,93 @@ static const struct attribute_group virtio_net_mrg_rx_group = {
>> };
>> #endif
>>
>> +static int virtnet_validate_features(struct virtio_device *dev,
>> + unsigned int *table,
>> + int table_size,
>> + unsigned int feature)
>> +{
>> + int i;
>> +
>> + if (!virtio_has_feature(dev, feature)) {
> Do an early return, get rid of one indentation level?
This sounds good.
>> + for (i = 0; i < table_size; i++) {
>> + unsigned int f = table[i];
>> +
>> + if (virtio_has_feature(dev, f)) {
>> + dev_err(&dev->dev,
>> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x was not",
> s/hyperviser/hypervisor/ (also below)
Yes, thanks for the catching.
>
>> + f, feature);
>> + return -EINVAL;
>> + }
>> + }
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int virtnet_check_features(struct virtio_device *dev)
>> +{
>> + unsigned int features_for_ctrl_vq[] = {
>> + VIRTIO_NET_F_CTRL_RX,
>> + VIRTIO_NET_F_CTRL_VLAN,
>> + VIRTIO_NET_F_GUEST_ANNOUNCE,
>> + VIRTIO_NET_F_MQ,
>> + VIRTIO_NET_F_CTRL_MAC_ADDR
>> + };
>> + unsigned int features_for_guest_csum[] = {
>> + VIRTIO_NET_F_GUEST_TSO4,
>> + VIRTIO_NET_F_GUEST_TSO6,
>> + VIRTIO_NET_F_GUEST_ECN,
>> + };
>> + unsigned int features_for_host_csum[] = {
>> + VIRTIO_NET_F_HOST_TSO4,
>> + VIRTIO_NET_F_HOST_TSO6,
>> + VIRTIO_NET_F_HOST_ECN,
>> + };
> I'm wondering whether it would be easier to read if you listed all
> prereqs per feature instead of all features that depend on a feature?
> It would still be hard to express the v4/v6 or conditions below in
> tables, though.
For v4 and v6, we could use something like
unsigned int feature_for_host_tso6[] = {
VIRTIO_NET_F_HOST_ECN,
};
unsigned int feature_for_host_tso4[] = {
VIRTIO_NET_F_HOST_ECN,
}
To avoid the following open-coding for ECN. And probably we need another
device specific dependency table and let virtio core do this instead.
>
> Or call the arrays features_depending_on_foo?
Yes.
>
>> + int err;
>> +
>> + err = virtnet_validate_features(dev, features_for_ctrl_vq,
>> + ARRAY_SIZE(features_for_ctrl_vq),
>> + VIRTIO_NET_F_CTRL_VQ);
>> + if (err)
>> + return err;
> If you already print a message that a user may use to fix their
> hypervisor (or bug someone about it), would it make sense to check all
> dependencies and print a full list of everything that is broken in the
> advertised feature bits? I usually hate it if I fix one thing only to
> hit the next bug when the program could have already told me about
> everything I need to fix :)
>
Right this sounds good.
>> +
>> + err = virtnet_validate_features(dev, features_for_guest_csum,
>> + ARRAY_SIZE(features_for_guest_csum),
>> + VIRTIO_NET_F_GUEST_CSUM);
>> + if (err)
>> + return err;
>> +
>> + err = virtnet_validate_features(dev, features_for_host_csum,
>> + ARRAY_SIZE(features_for_host_csum),
>> + VIRTIO_NET_F_CSUM);
>> + if (err)
>> + return err;
>> +
>> + if (virtio_has_feature(dev, VIRTIO_NET_F_GUEST_ECN) &&
>> + (!virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO4) ||
>> + !virtio_has_feature(dev, VIRTIO_NET_F_GUEST_TSO6))) {
>> + dev_err(&dev->dev,
>> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
>> + VIRTIO_NET_F_GUEST_ECN,
>> + VIRTIO_NET_F_GUEST_TSO4,
>> + VIRTIO_NET_F_GUEST_TSO6);
>> + return -EINVAL;
>> + }
>> +
>> + if (virtio_has_feature(dev, VIRTIO_NET_F_HOST_ECN) &&
>> + (!virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO4) ||
>> + !virtio_has_feature(dev, VIRTIO_NET_F_HOST_TSO6))) {
>> + dev_err(&dev->dev,
>> + "buggy hyperviser: feature 0x%x was advertised but its dependency 0x%x or 0x%x was not",
> "Hypervisor bug: advertised feature <foo> but not <bar> or <baz>"
>
> ?
More compact, looks good. Thanks
From: Jason Wang <[email protected]>
Date: Wed, 19 Nov 2014 17:21:46 +0800
> More compact, looks good. Thanks
I am assuming there is therefore a V3 of this patch forthcoming.