[dpdk-dev] [PATCH v4 2/5] vhost: support selective datapath

[Wang, Zhihong]

> -----Original Message-----
> From: Maxime Coquelin [mailto:maxime.coquelin at redhat.com]
> Sent: Saturday, March 31, 2018 2:10 PM
> To: Wang, Zhihong <zhihong.wang at intel.com>; dev at dpdk.org
> Cc: Tan, Jianfeng <jianfeng.tan at intel.com>; Bie, Tiwei
> <tiwei.bie at intel.com>; yliu at fridaylinux.org; Liang, Cunming
> <cunming.liang at intel.com>; Wang, Xiao W <xiao.w.wang at intel.com>; Daly,
> Dan <dan.daly at intel.com>
> Subject: Re: [PATCH v4 2/5] vhost: support selective datapath
> 
> 
> 
> On 03/10/2018 11:01 AM, Zhihong Wang wrote:
> > This patch set introduces support for selective datapath in DPDK vhost-user
> > lib. vDPA stands for vhost Data Path Acceleration. The idea is to support
> > virtio ring compatible devices to serve virtio driver directly to enable
> > datapath acceleration.
> >
> > A set of device ops is defined for device specific operations:
> >
> >       a. queue_num_get: Called to get supported queue number of the
> device.
> >
> >       b. feature_get: Called to get supported features of the device.
> >
> >       c. protocol_feature_get: Called to get supported protocol features of
> >          the device.
> >
> >       d. dev_conf: Called to configure the actual device when the virtio
> >          device becomes ready.
> >
> >       e. dev_close: Called to close the actual device when the virtio device
> >          is stopped.
> >
> >       f. vring_state_set: Called to change the state of the vring in the
> >          actual device when vring state changes.
> >
> >       g. feature_set: Called to set the negotiated features to device.
> >
> >       h. migration_done: Called to allow the device to response to RARP
> >          sending.
> >
> >       i. get_vfio_group_fd: Called to get the VFIO group fd of the device.
> >
> >       j. get_vfio_device_fd: Called to get the VFIO device fd of the device.
> >
> >       k. get_notify_area: Called to get the notify area info of the queue.
> >
> > Signed-off-by: Zhihong Wang <zhihong.wang at intel.com>
> > ---
> > Changes in v4:
> >
> >   1. Remove the "engine" concept in the lib.
> >
> > ---
> > Changes in v2:
> >
> >   1. Add VFIO related vDPA device ops.
> >
> >   lib/librte_vhost/Makefile              |  4 +-
> >   lib/librte_vhost/rte_vdpa.h            | 94
> +++++++++++++++++++++++++++++++++
> >   lib/librte_vhost/rte_vhost_version.map |  6 +++
> >   lib/librte_vhost/vdpa.c                | 96
> ++++++++++++++++++++++++++++++++++
> >   4 files changed, 198 insertions(+), 2 deletions(-)
> >   create mode 100644 lib/librte_vhost/rte_vdpa.h
> >   create mode 100644 lib/librte_vhost/vdpa.c
> >
> > diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile
> > index 5d6c6abae..37044ac03 100644
> > --- a/lib/librte_vhost/Makefile
> > +++ b/lib/librte_vhost/Makefile
> > @@ -22,9 +22,9 @@ LDLIBS += -lrte_eal -lrte_mempool -lrte_mbuf -
> lrte_ethdev -lrte_net
> >
> >   # all source are stored in SRCS-y
> >   SRCS-$(CONFIG_RTE_LIBRTE_VHOST) := fd_man.c iotlb.c socket.c vhost.c
> \
> > -					vhost_user.c virtio_net.c
> > +					vhost_user.c virtio_net.c vdpa.c
> >
> >   # install includes
> > -SYMLINK-$(CONFIG_RTE_LIBRTE_VHOST)-include += rte_vhost.h
> > +SYMLINK-$(CONFIG_RTE_LIBRTE_VHOST)-include += rte_vhost.h
> rte_vdpa.h
> >
> >   include $(RTE_SDK)/mk/rte.lib.mk
> > diff --git a/lib/librte_vhost/rte_vdpa.h b/lib/librte_vhost/rte_vdpa.h
> > new file mode 100644
> > index 000000000..a4bbbd93d
> > --- /dev/null
> > +++ b/lib/librte_vhost/rte_vdpa.h
> > @@ -0,0 +1,94 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2018 Intel Corporation
> > + */
> > +
> > +#ifndef _RTE_VDPA_H_
> > +#define _RTE_VDPA_H_
> > +
> > +/**
> > + * @file
> > + *
> > + * Device specific vhost lib
> > + */
> > +
> > +#include <rte_pci.h>
> > +#include "rte_vhost.h"
> > +
> > +#define MAX_VDPA_NAME_LEN 128
> > +
> > +enum vdpa_addr_type {
> > +	PCI_ADDR,
> > +	VDPA_ADDR_MAX
> > +};
> > +
> > +struct rte_vdpa_dev_addr {
> > +	enum vdpa_addr_type type;
> > +	union {
> > +		uint8_t __dummy[64];
> > +		struct rte_pci_addr pci_addr;
> > +	};
> > +};
> > +
> > +/* Get capabilities of this device */
> > +typedef int (*vdpa_dev_queue_num_get_t)(int did, uint32_t
> *queue_num);
> > +typedef int (*vdpa_dev_feature_get_t)(int did, uint64_t *features);
> > +
> > +/* Driver configure/close the device */
> > +typedef int (*vdpa_dev_conf_t)(int vid);
> > +typedef int (*vdpa_dev_close_t)(int vid);
> > +
> > +/* Enable/disable this vring */
> > +typedef int (*vdpa_vring_state_set_t)(int vid, int vring, int state);
> > +
> > +/* Set features when changed */
> > +typedef int (*vdpa_feature_set_t)(int vid);
> > +
> > +/* Destination operations when migration done */
> > +typedef int (*vdpa_migration_done_t)(int vid);
> > +
> > +/* Get the vfio group fd */
> > +typedef int (*vdpa_get_vfio_group_fd_t)(int vid);
> > +
> > +/* Get the vfio device fd */
> > +typedef int (*vdpa_get_vfio_device_fd_t)(int vid);
> > +
> > +/* Get the notify area info of the queue */
> > +typedef int (*vdpa_get_notify_area_t)(int vid, int qid, uint64_t *offset,
> > +		uint64_t *size);
> > +/* Device ops */
> > +struct rte_vdpa_dev_ops {
> > +	vdpa_dev_queue_num_get_t  queue_num_get;
> > +	vdpa_dev_feature_get_t    feature_get;
> > +	vdpa_dev_feature_get_t    protocol_feature_get;
> > +	vdpa_dev_conf_t           dev_conf;
> > +	vdpa_dev_close_t          dev_close;
> > +	vdpa_vring_state_set_t    vring_state_set;
> > +	vdpa_feature_set_t        feature_set;
> > +	vdpa_migration_done_t     migration_done;
> > +	vdpa_get_vfio_group_fd_t  get_vfio_group_fd;
> > +	vdpa_get_vfio_device_fd_t get_vfio_device_fd;
> > +	vdpa_get_notify_area_t    get_notify_area;
> 
> Maybe you could reserve some room here to avoid breaking the ABI in the
> future if we need to add some optional ops.

Good suggestion.

> 
> > +};
> > +
> > +struct rte_vdpa_device {
> > +	struct rte_vdpa_dev_addr addr;
> > +	struct rte_vdpa_dev_ops *ops;
> > +} __rte_cache_aligned;
> > +
> > +extern struct rte_vdpa_device *vdpa_devices[];
> > +extern uint32_t vdpa_device_num;
> > +
> > +/* Register a vdpa device, return did if successful, -1 on failure */
> > +int __rte_experimental
> > +rte_vdpa_register_device(struct rte_vdpa_dev_addr *addr,
> > +		struct rte_vdpa_dev_ops *ops);
> > +
> > +/* Unregister a vdpa device, return -1 on failure */
> > +int __rte_experimental
> > +rte_vdpa_unregister_device(int did);
> > +
> > +/* Find did of a vdpa device, return -1 on failure */
> > +int __rte_experimental
> > +rte_vdpa_find_device_id(struct rte_vdpa_dev_addr *addr);
> > +
> > +#endif /* _RTE_VDPA_H_ */
> > diff --git a/lib/librte_vhost/rte_vhost_version.map
> b/lib/librte_vhost/rte_vhost_version.map
> > index df0103129..7bcffb490 100644
> > --- a/lib/librte_vhost/rte_vhost_version.map
> > +++ b/lib/librte_vhost/rte_vhost_version.map
> > @@ -59,3 +59,9 @@ DPDK_18.02 {
> >   	rte_vhost_vring_call;
> >
> >   } DPDK_17.08;
> > +
> > +EXPERIMENTAL {
> > +	rte_vdpa_register_device;
> > +	rte_vdpa_unregister_device;
> > +	rte_vdpa_find_device_id;
> 
> I think you need also to declare the new structs here,
> not only the new functions.

Ok.

> 
> > +} DPDK_18.02;
> > diff --git a/lib/librte_vhost/vdpa.c b/lib/librte_vhost/vdpa.c
> > new file mode 100644
> > index 000000000..0c950d45f
> > --- /dev/null
> > +++ b/lib/librte_vhost/vdpa.c
> > @@ -0,0 +1,96 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2018 Intel Corporation
> > + */
> > +
> > +/**
> > + * @file
> > + *
> > + * Device specific vhost lib
> > + */
> > +
> > +#include <stdbool.h>
> > +
> > +#include <rte_malloc.h>
> > +#include "rte_vdpa.h"
> > +#include "vhost.h"
> > +
> > +struct rte_vdpa_device *vdpa_devices[MAX_VHOST_DEVICE];
> > +uint32_t vdpa_device_num;
> > +
> > +static int is_same_vdpa_dev_addr(struct rte_vdpa_dev_addr *a,
> > +		struct rte_vdpa_dev_addr *b)
> > +{
> 
> Given the boolean nature of the function name, I would return 1 if same
> device, 0 if different.

Ok, will use bool.

> 
> > +	int ret = 0;
> > +
> > +	if (a->type != b->type)
> > +		return -1;
> > +
> > +	switch (a->type) {
> > +	case PCI_ADDR:
> > +		if (a->pci_addr.domain != b->pci_addr.domain ||
> > +				a->pci_addr.bus != b->pci_addr.bus ||
> > +				a->pci_addr.devid != b->pci_addr.devid ||
> > +				a->pci_addr.function != b->pci_addr.function)
> > +			ret = -1;
> > +		break;
> > +	default:
> > +		break;
> > +	}
> > +
> > +	return ret;
> > +}
> > +
> > +int rte_vdpa_register_device(struct rte_vdpa_dev_addr *addr,
> > +		struct rte_vdpa_dev_ops *ops)
> > +{
> > +	struct rte_vdpa_device *dev;
> > +	char device_name[MAX_VDPA_NAME_LEN];
> > +	int i;
> > +
> > +	if (vdpa_device_num >= MAX_VHOST_DEVICE)
> > +		return -1;
> > +
> > +	for (i = 0; i < MAX_VHOST_DEVICE; i++) {
> > +		if (vdpa_devices[i] == NULL)
> > +			break;
> You might want to check same device isn't being registering a second
> time, and return an error in that case.

Will do.

Thanks
-Zhihong

> 
> This is not a blocker though, and can be done in a dedicated patch.
> 
> > +	}
> > +
> > +	sprintf(device_name, "vdpa-dev-%d", i);
> > +	dev = rte_zmalloc(device_name, sizeof(struct rte_vdpa_device),
> > +			RTE_CACHE_LINE_SIZE);
> > +	if (!dev)
> > +		return -1;
> > +
> > +	memcpy(&dev->addr, addr, sizeof(struct rte_vdpa_dev_addr));
> > +	dev->ops = ops;
> > +	vdpa_devices[i] = dev;
> > +	vdpa_device_num++;
> > +
> > +	return i;
> > +}
> > +
> > +int rte_vdpa_unregister_device(int did)
> > +{
> > +	if (did < 0 || did >= MAX_VHOST_DEVICE || vdpa_devices[did] ==
> NULL)
> > +		return -1;
> > +
> > +	rte_free(vdpa_devices[did]);
> > +	vdpa_devices[did] = NULL;
> > +	vdpa_device_num--;
> > +
> > +	return did;
> > +}
> > +
> > +int rte_vdpa_find_device_id(struct rte_vdpa_dev_addr *addr)
> > +{
> > +	struct rte_vdpa_device *dev;
> > +	int i;
> > +
> > +	for (i = 0; i < MAX_VHOST_DEVICE; ++i) {
> > +		dev = vdpa_devices[i];
> > +		if (dev && is_same_vdpa_dev_addr(&dev->addr, addr) == 0)
> > +			return i;
> > +	}
> > +
> > +	return -1;
> > +}
> >