On 9/27/2022 12:36 PM, Jason Wang wrote:
On Tue, Sep 27, 2022 at 11:09 AM Zhu Lingshan <lingshan.zhu@xxxxxxxxx> wrote:
This commit adds a new vDPA netlink attribution
VDPA_ATTR_VDPA_DEV_SUPPORTED_FEATURES. Userspace can query
features of vDPA devices through this new attr.
This commit invokes vdpa_config_ops.get_config()
rather than vdpa_get_config_unlocked() to read
the device config spcae, so no races in
vdpa_set_features_unlocked()
Userspace tool iproute2 example:
$ vdpa dev config show vdpa0
vdpa0: mac 00:e8:ca:11:be:05 link up link_announce false max_vq_pairs 4 mtu 1500
negotiated_features MRG_RXBUF CTRL_VQ MQ VERSION_1 ACCESS_PLATFORM
dev_features MTU MAC MRG_RXBUF CTRL_VQ MQ ANY_LAYOUT VERSION_1 ACCESS_PLATFORM
Signed-off-by: Zhu Lingshan <lingshan.zhu@xxxxxxxxx>
---
drivers/vdpa/vdpa.c | 17 ++++++++++++-----
include/uapi/linux/vdpa.h | 4 ++++
2 files changed, 16 insertions(+), 5 deletions(-)
diff --git a/drivers/vdpa/vdpa.c b/drivers/vdpa/vdpa.c
index c06c02704461..2035700d6fc8 100644
--- a/drivers/vdpa/vdpa.c
+++ b/drivers/vdpa/vdpa.c
@@ -491,6 +491,7 @@ static int vdpa_mgmtdev_fill(const struct vdpa_mgmt_dev *mdev, struct sk_buff *m
err = -EMSGSIZE;
goto msg_err;
}
+
Nit: Unnecessary changes.
oh, yes!
if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_SUPPORTED_FEATURES,
mdev->supported_features, VDPA_ATTR_PAD)) {
err = -EMSGSIZE;
@@ -815,10 +816,10 @@ static int vdpa_dev_net_mq_config_fill(struct vdpa_device *vdev,
static int vdpa_dev_net_config_fill(struct vdpa_device *vdev, struct sk_buff *msg)
{
struct virtio_net_config config = {};
- u64 features;
+ u64 features_device, features_driver;
u16 val_u16;
- vdpa_get_config_unlocked(vdev, 0, &config, sizeof(config));
+ vdev->config->get_config(vdev, 0, &config, sizeof(config));
if (nla_put(msg, VDPA_ATTR_DEV_NET_CFG_MACADDR, sizeof(config.mac),
config.mac))
@@ -832,12 +833,18 @@ static int vdpa_dev_net_config_fill(struct vdpa_device *vdev, struct sk_buff *ms
if (nla_put_u16(msg, VDPA_ATTR_DEV_NET_CFG_MTU, val_u16))
return -EMSGSIZE;
- features = vdev->config->get_driver_features(vdev);
- if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_NEGOTIATED_FEATURES, features,
+ features_driver = vdev->config->get_driver_features(vdev);
+ if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_NEGOTIATED_FEATURES, features_driver,
+ VDPA_ATTR_PAD))
+ return -EMSGSIZE;
It looks to me that those parts were removed in patch 2. I wonder if
it's better to reorder the patch to let patch 2 come first?
I am not sure, if we apply patch 2(check FEATURES_OK in
vdpa_dev_config_fill and report driver features there) first,
we need to remove driver_features in vdpa_dev_net_config_fill(),
however, we still not introduce and initialize device features(which is
in patch 1, introduce device_features) yet,
so the last line "return vdpa_dev_net_mq_config_fill()" which needs
features as its parameter may not work,
filling NULL looks worse than keep features_driver here.
Thanks
Zhu Lingshan
Thanks
+
+ features_device = vdev->config->get_device_features(vdev);
+
+ if (nla_put_u64_64bit(msg, VDPA_ATTR_VDPA_DEV_SUPPORTED_FEATURES, features_device,
VDPA_ATTR_PAD))
return -EMSGSIZE;
- return vdpa_dev_net_mq_config_fill(vdev, msg, features, &config);
+ return vdpa_dev_net_mq_config_fill(vdev, msg, features_driver, &config);
}
static int
diff --git a/include/uapi/linux/vdpa.h b/include/uapi/linux/vdpa.h
index 25c55cab3d7c..07474183fdb3 100644
--- a/include/uapi/linux/vdpa.h
+++ b/include/uapi/linux/vdpa.h
@@ -46,12 +46,16 @@ enum vdpa_attr {
VDPA_ATTR_DEV_NEGOTIATED_FEATURES, /* u64 */
VDPA_ATTR_DEV_MGMTDEV_MAX_VQS, /* u32 */
+ /* virtio features that are supported by the vDPA management device */
VDPA_ATTR_DEV_SUPPORTED_FEATURES, /* u64 */
VDPA_ATTR_DEV_QUEUE_INDEX, /* u32 */
VDPA_ATTR_DEV_VENDOR_ATTR_NAME, /* string */
VDPA_ATTR_DEV_VENDOR_ATTR_VALUE, /* u64 */
+ /* virtio features that are supported by the vDPA device */
+ VDPA_ATTR_VDPA_DEV_SUPPORTED_FEATURES, /* u64 */
+
/* new attributes must be added above here */
VDPA_ATTR_MAX,
};
--
2.31.1
