[v2] vhost: enable reply-ack systematically
Checks
Commit Message
As announced during v20.05 release cycle, this
patch makes reply-ack protocol feature to be enabled
unconditionally.
This protocol feature makes the communication between the
master and the slave more robust, avoiding for example
possible undefined behaviour with VHOST_USER_SET_MEM_TABLE.
Also, reply-ack support will be required for upcoming
VHOST_USER_SET_STATUS request.
Note that this protocol feature was disabled by default
because Qemu version 2.7.0 to 2.9.0 had a bug causing a
deadlock when reply-ack was negotiated and multiqueue
enabled. These Qemu version are now very old and no more
maintained, so we can resonnably consider we no more
support them.
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
---
v2:
- Fix commit message
- Remove mention to the workaround in the doc
doc/guides/prog_guide/vhost_lib.rst | 5 -----
lib/librte_vhost/vhost_user.c | 9 ---------
2 files changed, 14 deletions(-)
Comments
> -----Original Message-----
> From: dev <dev-bounces@dpdk.org> On Behalf Of Maxime Coquelin
> Sent: Thursday, May 28, 2020 5:04 PM
> To: dev@dpdk.org; Ye, Xiaolong <xiaolong.ye@intel.com>;
> amorenoz@redhat.com; david.marchand@redhat.com; thomas@monjalon.net
> Cc: Maxime Coquelin <maxime.coquelin@redhat.com>
> Subject: [dpdk-dev] [PATCH v2] vhost: enable reply-ack systematically
>
> As announced during v20.05 release cycle, this patch makes reply-ack protocol
> feature to be enabled unconditionally.
>
> This protocol feature makes the communication between the master and the
> slave more robust, avoiding for example possible undefined behaviour with
> VHOST_USER_SET_MEM_TABLE.
>
> Also, reply-ack support will be required for upcoming VHOST_USER_SET_STATUS
> request.
>
> Note that this protocol feature was disabled by default because Qemu version
> 2.7.0 to 2.9.0 had a bug causing a deadlock when reply-ack was negotiated and
> multiqueue enabled. These Qemu version are now very old and no more
> maintained, so we can resonnably consider we no more support them.
>
> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
> ---
>
> v2:
> - Fix commit message
> - Remove mention to the workaround in the doc
>
> doc/guides/prog_guide/vhost_lib.rst | 5 -----
> lib/librte_vhost/vhost_user.c | 9 ---------
> 2 files changed, 14 deletions(-)
>
> diff --git a/doc/guides/prog_guide/vhost_lib.rst
> b/doc/guides/prog_guide/vhost_lib.rst
> index 07e40e3c5e..db921f922b 100644
> --- a/doc/guides/prog_guide/vhost_lib.rst
> +++ b/doc/guides/prog_guide/vhost_lib.rst
> @@ -104,11 +104,6 @@ The following is an overview of some key Vhost API
> functions:
> from accessing memory the virtio device isn't allowed to, when the feature
> is negotiated and an IOMMU device is declared.
>
> - However, this feature enables vhost-user's reply-ack protocol feature,
> - which implementation is buggy in Qemu v2.7.0-v2.9.0 when doing multiqueue.
> - Enabling this flag with these Qemu version results in Qemu being blocked
> - when multiple queue pairs are declared.
> -
> - ``RTE_VHOST_USER_POSTCOPY_SUPPORT``
>
> Postcopy live-migration support will be enabled when this flag is set.
> diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c index
> 84bebad792..ea9cd107b9 100644
> --- a/lib/librte_vhost/vhost_user.c
> +++ b/lib/librte_vhost/vhost_user.c
> @@ -2012,15 +2012,6 @@ vhost_user_get_protocol_features(struct virtio_net
> **pdev,
> rte_vhost_driver_get_features(dev->ifname, &features);
> rte_vhost_driver_get_protocol_features(dev->ifname,
> &protocol_features);
>
> - /*
> - * REPLY_ACK protocol feature is only mandatory for now
> - * for IOMMU feature. If IOMMU is explicitly disabled by the
> - * application, disable also REPLY_ACK feature for older buggy
> - * Qemu versions (from v2.7.0 to v2.9.0).
> - */
> - if (!(features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)))
> - protocol_features &= ~(1ULL <<
> VHOST_USER_PROTOCOL_F_REPLY_ACK);
> -
> msg->payload.u64 = protocol_features;
> msg->size = sizeof(msg->payload.u64);
> msg->fd_num = 0;
> --
> 2.26.2
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
On 5/28/20 11:03 AM, Maxime Coquelin wrote:
> As announced during v20.05 release cycle, this
> patch makes reply-ack protocol feature to be enabled
> unconditionally.
>
> This protocol feature makes the communication between the
> master and the slave more robust, avoiding for example
> possible undefined behaviour with VHOST_USER_SET_MEM_TABLE.
>
> Also, reply-ack support will be required for upcoming
> VHOST_USER_SET_STATUS request.
>
> Note that this protocol feature was disabled by default
> because Qemu version 2.7.0 to 2.9.0 had a bug causing a
> deadlock when reply-ack was negotiated and multiqueue
> enabled. These Qemu version are now very old and no more
> maintained, so we can resonnably consider we no more
> support them.
>
> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
> ---
Applied to dpdk-next-virtio/master
Thanks,
Maxime
@@ -104,11 +104,6 @@ The following is an overview of some key Vhost API functions:
from accessing memory the virtio device isn't allowed to, when the feature
is negotiated and an IOMMU device is declared.
- However, this feature enables vhost-user's reply-ack protocol feature,
- which implementation is buggy in Qemu v2.7.0-v2.9.0 when doing multiqueue.
- Enabling this flag with these Qemu version results in Qemu being blocked
- when multiple queue pairs are declared.
-
- ``RTE_VHOST_USER_POSTCOPY_SUPPORT``
Postcopy live-migration support will be enabled when this flag is set.
@@ -2012,15 +2012,6 @@ vhost_user_get_protocol_features(struct virtio_net **pdev,
rte_vhost_driver_get_features(dev->ifname, &features);
rte_vhost_driver_get_protocol_features(dev->ifname, &protocol_features);
- /*
- * REPLY_ACK protocol feature is only mandatory for now
- * for IOMMU feature. If IOMMU is explicitly disabled by the
- * application, disable also REPLY_ACK feature for older buggy
- * Qemu versions (from v2.7.0 to v2.9.0).
- */
- if (!(features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)))
- protocol_features &= ~(1ULL << VHOST_USER_PROTOCOL_F_REPLY_ACK);
-
msg->payload.u64 = protocol_features;
msg->size = sizeof(msg->payload.u64);
msg->fd_num = 0;