* [Qemu-devel] [PATCH v3 0/2] vhost-user: Back SET/GET_CONFIG with a protocol feature @ 2018-03-29 7:52 Maxime Coquelin 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init Maxime Coquelin 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature Maxime Coquelin 0 siblings, 2 replies; 5+ messages in thread From: Maxime Coquelin @ 2018-03-29 7:52 UTC (permalink / raw) To: mst, changpeng.liu, marcandre.lureau, qemu-devel; +Cc: Maxime Coquelin V3 removes the assert of vhost_dev's vhost_ops, as it is not yet set at that time. While reviewing DPDK series adding support to VHOST_USER_SET_CONFIG and VHOST_USER_GET_CONFIG request, I found that it was not backed with a dedicated protocol feature. This series addresses this by adding a new protocol feature bit, and by only negotiating it if the device supports it, as suggested by Michael. Indeed, if the feature is supported by other type of devices in the future, it would confuse the backends as it couldn't know whether the device really support it or not. To know whether the vhost device support config feature, the trick is to check whether it implemented the config_ops. That's the reason why the first patch moves setting the config ops in vhost-user-blk befoire calling vhost_user_init(). The series targets v2.12 release, else we may have to disable these requests in this release. *NOTE*: The series has only been tested as I don't have the environment to try it. Changpeng, can you please test it? Thanks, Maxime Changes since v2: ================= - Remove assert (Changpeng) - Fix typo in comment Changes since v1: ================= - Fail vhost-user init if device implements config feature but the backend doesn't. (mst) Maxime Coquelin (2): vhost-user-blk: set config ops before vhost-user init vhost-user: back SET/GET_CONFIG requests with a protocol feature docs/interop/vhost-user.txt | 21 ++++++++++++--------- hw/block/vhost-user-blk.c | 4 ++-- hw/virtio/vhost-user.c | 22 ++++++++++++++++++++++ hw/virtio/vhost.c | 1 - 4 files changed, 36 insertions(+), 12 deletions(-) -- 2.14.3 ^ permalink raw reply [flat|nested] 5+ messages in thread
* [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init 2018-03-29 7:52 [Qemu-devel] [PATCH v3 0/2] vhost-user: Back SET/GET_CONFIG with a protocol feature Maxime Coquelin @ 2018-03-29 7:52 ` Maxime Coquelin 2018-04-08 8:53 ` Liu, Changpeng 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature Maxime Coquelin 1 sibling, 1 reply; 5+ messages in thread From: Maxime Coquelin @ 2018-03-29 7:52 UTC (permalink / raw) To: mst, changpeng.liu, marcandre.lureau, qemu-devel; +Cc: Maxime Coquelin As soon as vhost-user init is done, the backend may send VHOST_USER_SLAVE_CONFIG_CHANGE_MSG, so let's set the notification callback before it. Also, it will be used to know whether the device supports the config feature to advertize it or not. Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> --- hw/block/vhost-user-blk.c | 4 ++-- hw/virtio/vhost.c | 1 - 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c index f840f07dfe..262baca432 100644 --- a/hw/block/vhost-user-blk.c +++ b/hw/block/vhost-user-blk.c @@ -259,6 +259,8 @@ static void vhost_user_blk_device_realize(DeviceState *dev, Error **errp) s->dev.vq_index = 0; s->dev.backend_features = 0; + vhost_dev_set_config_notifier(&s->dev, &blk_ops); + ret = vhost_dev_init(&s->dev, &s->chardev, VHOST_BACKEND_TYPE_USER, 0); if (ret < 0) { error_setg(errp, "vhost-user-blk: vhost initialization failed: %s", @@ -277,8 +279,6 @@ static void vhost_user_blk_device_realize(DeviceState *dev, Error **errp) s->blkcfg.num_queues = s->num_queues; } - vhost_dev_set_config_notifier(&s->dev, &blk_ops); - return; vhost_err: diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c index 250f886acb..b6c314e350 100644 --- a/hw/virtio/vhost.c +++ b/hw/virtio/vhost.c @@ -1451,7 +1451,6 @@ int vhost_dev_set_config(struct vhost_dev *hdev, const uint8_t *data, void vhost_dev_set_config_notifier(struct vhost_dev *hdev, const VhostDevConfigOps *ops) { - assert(hdev->vhost_ops); hdev->config_ops = ops; } -- 2.14.3 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init Maxime Coquelin @ 2018-04-08 8:53 ` Liu, Changpeng 0 siblings, 0 replies; 5+ messages in thread From: Liu, Changpeng @ 2018-04-08 8:53 UTC (permalink / raw) To: Maxime Coquelin, mst@redhat.com, marcandre.lureau@redhat.com, qemu-devel@nongnu.org > -----Original Message----- > From: Maxime Coquelin [mailto:maxime.coquelin@redhat.com] > Sent: Thursday, March 29, 2018 3:53 PM > To: mst@redhat.com; Liu, Changpeng <changpeng.liu@intel.com>; > marcandre.lureau@redhat.com; qemu-devel@nongnu.org > Cc: Maxime Coquelin <maxime.coquelin@redhat.com> > Subject: [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init > > As soon as vhost-user init is done, the backend may send > VHOST_USER_SLAVE_CONFIG_CHANGE_MSG, so let's set the > notification callback before it. > > Also, it will be used to know whether the device supports > the config feature to advertize it or not. > > Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Acked-by: Changpeng Liu <changpeng.liu@intel.com> > --- > hw/block/vhost-user-blk.c | 4 ++-- > hw/virtio/vhost.c | 1 - > 2 files changed, 2 insertions(+), 3 deletions(-) > > diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c > index f840f07dfe..262baca432 100644 > --- a/hw/block/vhost-user-blk.c > +++ b/hw/block/vhost-user-blk.c > @@ -259,6 +259,8 @@ static void vhost_user_blk_device_realize(DeviceState > *dev, Error **errp) > s->dev.vq_index = 0; > s->dev.backend_features = 0; > > + vhost_dev_set_config_notifier(&s->dev, &blk_ops); > + > ret = vhost_dev_init(&s->dev, &s->chardev, VHOST_BACKEND_TYPE_USER, 0); > if (ret < 0) { > error_setg(errp, "vhost-user-blk: vhost initialization failed: %s", > @@ -277,8 +279,6 @@ static void vhost_user_blk_device_realize(DeviceState > *dev, Error **errp) > s->blkcfg.num_queues = s->num_queues; > } > > - vhost_dev_set_config_notifier(&s->dev, &blk_ops); > - > return; > > vhost_err: > diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c > index 250f886acb..b6c314e350 100644 > --- a/hw/virtio/vhost.c > +++ b/hw/virtio/vhost.c > @@ -1451,7 +1451,6 @@ int vhost_dev_set_config(struct vhost_dev *hdev, > const uint8_t *data, > void vhost_dev_set_config_notifier(struct vhost_dev *hdev, > const VhostDevConfigOps *ops) > { > - assert(hdev->vhost_ops); > hdev->config_ops = ops; > } > > -- > 2.14.3 ^ permalink raw reply [flat|nested] 5+ messages in thread
* [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature 2018-03-29 7:52 [Qemu-devel] [PATCH v3 0/2] vhost-user: Back SET/GET_CONFIG with a protocol feature Maxime Coquelin 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init Maxime Coquelin @ 2018-03-29 7:52 ` Maxime Coquelin 2018-04-08 8:54 ` Liu, Changpeng 1 sibling, 1 reply; 5+ messages in thread From: Maxime Coquelin @ 2018-03-29 7:52 UTC (permalink / raw) To: mst, changpeng.liu, marcandre.lureau, qemu-devel; +Cc: Maxime Coquelin Without a dedicated protocol feature, QEMU cannot know whether the backend can handle VHOST_USER_SET_CONFIG and VHOST_USER_GET_CONFIG messages. This patch adds a protocol feature that is only advertised by QEMU if the device implements the config ops. Vhost user init fails if the device support the feature but the backend doesn't. The backend should only send VHOST_USER_SLAVE_CONFIG_CHANGE_MSG requests if the protocol feature has been negotiated. Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> --- docs/interop/vhost-user.txt | 21 ++++++++++++--------- hw/virtio/vhost-user.c | 22 ++++++++++++++++++++++ 2 files changed, 34 insertions(+), 9 deletions(-) diff --git a/docs/interop/vhost-user.txt b/docs/interop/vhost-user.txt index c058c407df..534caab18a 100644 --- a/docs/interop/vhost-user.txt +++ b/docs/interop/vhost-user.txt @@ -379,6 +379,7 @@ Protocol features #define VHOST_USER_PROTOCOL_F_CROSS_ENDIAN 6 #define VHOST_USER_PROTOCOL_F_CRYPTO_SESSION 7 #define VHOST_USER_PROTOCOL_F_PAGEFAULT 8 +#define VHOST_USER_PROTOCOL_F_CONFIG 9 Master message types -------------------- @@ -664,7 +665,8 @@ Master message types Master payload: virtio device config space Slave payload: virtio device config space - Submitted by the vhost-user master to fetch the contents of the virtio + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, this message is + submitted by the vhost-user master to fetch the contents of the virtio device configuration space, vhost-user slave's payload size MUST match master's request, vhost-user slave uses zero length of payload to indicate an error to vhost-user master. The vhost-user master may @@ -677,7 +679,8 @@ Master message types Master payload: virtio device config space Slave payload: N/A - Submitted by the vhost-user master when the Guest changes the virtio + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, this message is + submitted by the vhost-user master when the Guest changes the virtio device configuration space and also can be used for live migration on the destination host. The vhost-user slave must check the flags field, and slaves MUST NOT accept SET_CONFIG for read-only @@ -766,13 +769,13 @@ Slave message types Slave payload: N/A Master payload: N/A - Vhost-user slave sends such messages to notify that the virtio device's - configuration space has changed, for those host devices which can support - such feature, host driver can send VHOST_USER_GET_CONFIG message to slave - to get the latest content. If VHOST_USER_PROTOCOL_F_REPLY_ACK is - negotiated, and slave set the VHOST_USER_NEED_REPLY flag, master must - respond with zero when operation is successfully completed, or non-zero - otherwise. + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, vhost-user slave sends + such messages to notify that the virtio device's configuration space has + changed, for those host devices which can support such feature, host + driver can send VHOST_USER_GET_CONFIG message to slave to get the latest + content. If VHOST_USER_PROTOCOL_F_REPLY_ACK is negotiated, and slave set + the VHOST_USER_NEED_REPLY flag, master must respond with zero when + operation is successfully completed, or non-zero otherwise. VHOST_USER_PROTOCOL_F_REPLY_ACK: ------------------------------- diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c index 44aea5c0a8..38da8692bb 100644 --- a/hw/virtio/vhost-user.c +++ b/hw/virtio/vhost-user.c @@ -46,6 +46,7 @@ enum VhostUserProtocolFeature { VHOST_USER_PROTOCOL_F_CROSS_ENDIAN = 6, VHOST_USER_PROTOCOL_F_CRYPTO_SESSION = 7, VHOST_USER_PROTOCOL_F_PAGEFAULT = 8, + VHOST_USER_PROTOCOL_F_CONFIG = 9, VHOST_USER_PROTOCOL_F_MAX }; @@ -1211,6 +1212,17 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque) dev->protocol_features = protocol_features & VHOST_USER_PROTOCOL_FEATURE_MASK; + + if (!dev->config_ops || !dev->config_ops->vhost_dev_config_notifier) { + /* Don't acknowledge CONFIG feature if device doesn't support it */ + dev->protocol_features &= ~(1ULL << VHOST_USER_PROTOCOL_F_CONFIG); + } else if (!(protocol_features & + (1ULL << VHOST_USER_PROTOCOL_F_CONFIG))) { + error_report("Device expects VHOST_USER_PROTOCOL_F_CONFIG " + "but backend does not support it."); + return -1; + } + err = vhost_user_set_protocol_features(dev, dev->protocol_features); if (err < 0) { return err; @@ -1405,6 +1417,11 @@ static int vhost_user_get_config(struct vhost_dev *dev, uint8_t *config, .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + config_len, }; + if (!virtio_has_feature(dev->protocol_features, + VHOST_USER_PROTOCOL_F_CONFIG)) { + return -1; + } + if (config_len > VHOST_USER_MAX_CONFIG_SIZE) { return -1; } @@ -1448,6 +1465,11 @@ static int vhost_user_set_config(struct vhost_dev *dev, const uint8_t *data, .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + size, }; + if (!virtio_has_feature(dev->protocol_features, + VHOST_USER_PROTOCOL_F_CONFIG)) { + return -1; + } + if (reply_supported) { msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK; } -- 2.14.3 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature Maxime Coquelin @ 2018-04-08 8:54 ` Liu, Changpeng 0 siblings, 0 replies; 5+ messages in thread From: Liu, Changpeng @ 2018-04-08 8:54 UTC (permalink / raw) To: Maxime Coquelin, mst@redhat.com, marcandre.lureau@redhat.com, qemu-devel@nongnu.org > -----Original Message----- > From: Maxime Coquelin [mailto:maxime.coquelin@redhat.com] > Sent: Thursday, March 29, 2018 3:53 PM > To: mst@redhat.com; Liu, Changpeng <changpeng.liu@intel.com>; > marcandre.lureau@redhat.com; qemu-devel@nongnu.org > Cc: Maxime Coquelin <maxime.coquelin@redhat.com> > Subject: [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a > protocol feature > > Without a dedicated protocol feature, QEMU cannot know whether > the backend can handle VHOST_USER_SET_CONFIG and > VHOST_USER_GET_CONFIG messages. > > This patch adds a protocol feature that is only advertised by > QEMU if the device implements the config ops. Vhost user init > fails if the device support the feature but the backend doesn't. > > The backend should only send VHOST_USER_SLAVE_CONFIG_CHANGE_MSG > requests if the protocol feature has been negotiated. > > Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Acked-by: Changpeng Liu <changpeng.liu@intel.com> > --- > docs/interop/vhost-user.txt | 21 ++++++++++++--------- > hw/virtio/vhost-user.c | 22 ++++++++++++++++++++++ > 2 files changed, 34 insertions(+), 9 deletions(-) > > diff --git a/docs/interop/vhost-user.txt b/docs/interop/vhost-user.txt > index c058c407df..534caab18a 100644 > --- a/docs/interop/vhost-user.txt > +++ b/docs/interop/vhost-user.txt > @@ -379,6 +379,7 @@ Protocol features > #define VHOST_USER_PROTOCOL_F_CROSS_ENDIAN 6 > #define VHOST_USER_PROTOCOL_F_CRYPTO_SESSION 7 > #define VHOST_USER_PROTOCOL_F_PAGEFAULT 8 > +#define VHOST_USER_PROTOCOL_F_CONFIG 9 > > Master message types > -------------------- > @@ -664,7 +665,8 @@ Master message types > Master payload: virtio device config space > Slave payload: virtio device config space > > - Submitted by the vhost-user master to fetch the contents of the virtio > + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, this message is > + submitted by the vhost-user master to fetch the contents of the virtio > device configuration space, vhost-user slave's payload size MUST match > master's request, vhost-user slave uses zero length of payload to > indicate an error to vhost-user master. The vhost-user master may > @@ -677,7 +679,8 @@ Master message types > Master payload: virtio device config space > Slave payload: N/A > > - Submitted by the vhost-user master when the Guest changes the virtio > + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, this message is > + submitted by the vhost-user master when the Guest changes the virtio > device configuration space and also can be used for live migration > on the destination host. The vhost-user slave must check the flags > field, and slaves MUST NOT accept SET_CONFIG for read-only > @@ -766,13 +769,13 @@ Slave message types > Slave payload: N/A > Master payload: N/A > > - Vhost-user slave sends such messages to notify that the virtio device's > - configuration space has changed, for those host devices which can support > - such feature, host driver can send VHOST_USER_GET_CONFIG message to > slave > - to get the latest content. If VHOST_USER_PROTOCOL_F_REPLY_ACK is > - negotiated, and slave set the VHOST_USER_NEED_REPLY flag, master must > - respond with zero when operation is successfully completed, or non-zero > - otherwise. > + When VHOST_USER_PROTOCOL_F_CONFIG is negotiated, vhost-user slave > sends > + such messages to notify that the virtio device's configuration space has > + changed, for those host devices which can support such feature, host > + driver can send VHOST_USER_GET_CONFIG message to slave to get the latest > + content. If VHOST_USER_PROTOCOL_F_REPLY_ACK is negotiated, and slave > set > + the VHOST_USER_NEED_REPLY flag, master must respond with zero when > + operation is successfully completed, or non-zero otherwise. > > VHOST_USER_PROTOCOL_F_REPLY_ACK: > ------------------------------- > diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c > index 44aea5c0a8..38da8692bb 100644 > --- a/hw/virtio/vhost-user.c > +++ b/hw/virtio/vhost-user.c > @@ -46,6 +46,7 @@ enum VhostUserProtocolFeature { > VHOST_USER_PROTOCOL_F_CROSS_ENDIAN = 6, > VHOST_USER_PROTOCOL_F_CRYPTO_SESSION = 7, > VHOST_USER_PROTOCOL_F_PAGEFAULT = 8, > + VHOST_USER_PROTOCOL_F_CONFIG = 9, > VHOST_USER_PROTOCOL_F_MAX > }; > > @@ -1211,6 +1212,17 @@ static int vhost_user_init(struct vhost_dev *dev, void > *opaque) > > dev->protocol_features = > protocol_features & VHOST_USER_PROTOCOL_FEATURE_MASK; > + > + if (!dev->config_ops || !dev->config_ops->vhost_dev_config_notifier) { > + /* Don't acknowledge CONFIG feature if device doesn't support it */ > + dev->protocol_features &= ~(1ULL << > VHOST_USER_PROTOCOL_F_CONFIG); > + } else if (!(protocol_features & > + (1ULL << VHOST_USER_PROTOCOL_F_CONFIG))) { > + error_report("Device expects VHOST_USER_PROTOCOL_F_CONFIG " > + "but backend does not support it."); > + return -1; > + } > + > err = vhost_user_set_protocol_features(dev, dev->protocol_features); > if (err < 0) { > return err; > @@ -1405,6 +1417,11 @@ static int vhost_user_get_config(struct vhost_dev > *dev, uint8_t *config, > .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + config_len, > }; > > + if (!virtio_has_feature(dev->protocol_features, > + VHOST_USER_PROTOCOL_F_CONFIG)) { > + return -1; > + } > + > if (config_len > VHOST_USER_MAX_CONFIG_SIZE) { > return -1; > } > @@ -1448,6 +1465,11 @@ static int vhost_user_set_config(struct vhost_dev > *dev, const uint8_t *data, > .hdr.size = VHOST_USER_CONFIG_HDR_SIZE + size, > }; > > + if (!virtio_has_feature(dev->protocol_features, > + VHOST_USER_PROTOCOL_F_CONFIG)) { > + return -1; > + } > + > if (reply_supported) { > msg.hdr.flags |= VHOST_USER_NEED_REPLY_MASK; > } > -- > 2.14.3 ^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2018-04-08 8:54 UTC | newest] Thread overview: 5+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2018-03-29 7:52 [Qemu-devel] [PATCH v3 0/2] vhost-user: Back SET/GET_CONFIG with a protocol feature Maxime Coquelin 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 1/2] vhost-user-blk: set config ops before vhost-user init Maxime Coquelin 2018-04-08 8:53 ` Liu, Changpeng 2018-03-29 7:52 ` [Qemu-devel] [PATCH v3 2/2] vhost-user: back SET/GET_CONFIG requests with a protocol feature Maxime Coquelin 2018-04-08 8:54 ` Liu, Changpeng
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).