[dpdk-dev] [PATCH v7 04/11] vhost: add two new messages to support a shared buffer

[JinYu <jin.yu@intel.com>]

This patch introduces two new messages VHOST_USER_GET_INFLIGHT_FD
and VHOST_USER_SET_INFLIGHT_FD to support transferring a shared
buffer between qemu and backend.

Signed-off-by: Lin Li <lilin24@baidu.com>
Signed-off-by: Xun Ni <nixun@baidu.com>
Signed-off-by: Yu Zhang <zhangyu31@baidu.com>
Signed-off-by: Jin Yu <jin.yu@intel.com>
---
 lib/librte_vhost/vhost.h      |   7 +
 lib/librte_vhost/vhost_user.c | 270 +++++++++++++++++++++++++++++++++-
 lib/librte_vhost/vhost_user.h |   4 +-
 3 files changed, 279 insertions(+), 2 deletions(-)

diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
index 884befa85..d67ba849a 100644
--- a/lib/librte_vhost/vhost.h
+++ b/lib/librte_vhost/vhost.h
@@ -286,6 +286,12 @@ struct guest_page {
 	uint64_t size;
 };
 
+struct inflight_mem_info {
+	int		fd;
+	void		*addr;
+	uint64_t	size;
+};
+
 /**
  * Device structure contains all configuration information relating
  * to the device.
@@ -303,6 +309,7 @@ struct virtio_net {
 	uint32_t		nr_vring;
 	int			dequeue_zero_copy;
 	struct vhost_virtqueue	*virtqueue[VHOST_MAX_QUEUE_PAIRS * 2];
+	struct inflight_mem_info *inflight_info;
 #define IF_NAME_SZ (PATH_MAX > IFNAMSIZ ? PATH_MAX : IFNAMSIZ)
 	char			ifname[IF_NAME_SZ];
 	uint64_t		log_size;
diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c
index c9e29ece8..ad2d36215 100644
--- a/lib/librte_vhost/vhost_user.c
+++ b/lib/librte_vhost/vhost_user.c
@@ -37,6 +37,10 @@
 #ifdef RTE_LIBRTE_VHOST_POSTCOPY
 #include <linux/userfaultfd.h>
 #endif
+#ifdef F_ADD_SEALS /* if file sealing is supported, so is memfd */
+#include <linux/memfd.h>
+#define MEMFD_SUPPORTED
+#endif
 
 #include <rte_common.h>
 #include <rte_malloc.h>
@@ -49,6 +53,15 @@
 #define VIRTIO_MIN_MTU 68
 #define VIRTIO_MAX_MTU 65535
 
+#define INFLIGHT_ALIGNMENT	64
+#define INFLIGHT_VERSION	0xabcd
+#define VIRTQUEUE_MAX_SIZE	1024
+
+#define CLOEXEC		0x0001U
+
+#define ALIGN_DOWN(n, m) ((n) / (m) * (m))
+#define ALIGN_UP(n, m) ALIGN_DOWN((n) + (m) - 1, (m))
+
 static const char *vhost_message_str[VHOST_USER_MAX] = {
 	[VHOST_USER_NONE] = "VHOST_USER_NONE",
 	[VHOST_USER_GET_FEATURES] = "VHOST_USER_GET_FEATURES",
@@ -78,6 +91,8 @@ static const char *vhost_message_str[VHOST_USER_MAX] = {
 	[VHOST_USER_POSTCOPY_ADVISE]  = "VHOST_USER_POSTCOPY_ADVISE",
 	[VHOST_USER_POSTCOPY_LISTEN]  = "VHOST_USER_POSTCOPY_LISTEN",
 	[VHOST_USER_POSTCOPY_END]  = "VHOST_USER_POSTCOPY_END",
+	[VHOST_USER_GET_INFLIGHT_FD] = "VHOST_USER_GET_INFLIGHT_FD",
+	[VHOST_USER_SET_INFLIGHT_FD] = "VHOST_USER_SET_INFLIGHT_FD",
 };
 
 static int send_vhost_reply(int sockfd, struct VhostUserMsg *msg);
@@ -160,6 +175,21 @@ vhost_backend_cleanup(struct virtio_net *dev)
 		dev->log_addr = 0;
 	}
 
+	if (dev->inflight_info->addr) {
+		munmap(dev->inflight_info->addr, dev->inflight_info->size);
+		dev->inflight_info->addr = NULL;
+	}
+
+	if (dev->inflight_info->fd > 0) {
+		close(dev->inflight_info->fd);
+		dev->inflight_info->fd = -1;
+	}
+
+	if (dev->inflight_info) {
+		free(dev->inflight_info);
+		dev->inflight_info = NULL;
+	}
+
 	if (dev->slave_req_fd >= 0) {
 		close(dev->slave_req_fd);
 		dev->slave_req_fd = -1;
@@ -1165,6 +1195,243 @@ virtio_is_ready(struct virtio_net *dev)
 	return 1;
 }
 
+static int
+mem_create(const char *name, unsigned int flags)
+{
+#ifdef MEMFD_SUPPORTED
+	return memfd_create(name, flags);
+#else
+	return -1;
+#endif
+}
+
+static void *
+inflight_mem_alloc(const char *name, size_t size, int *fd)
+{
+	void *ptr;
+	int mfd = -1;
+	char fname[20] = "/tmp/memfd-XXXXXX";
+
+	*fd = -1;
+	mfd = mem_create(name, CLOEXEC);
+	if (mfd != -1) {
+		if (ftruncate(mfd, size) == -1) {
+			RTE_LOG(ERR, VHOST_CONFIG,
+				"ftruncate fail for alloc inflight buffer\n");
+			close(mfd);
+			return NULL;
+		}
+	} else {
+		mfd = mkstemp(fname);
+		unlink(fname);
+
+		if (mfd == -1) {
+			RTE_LOG(ERR, VHOST_CONFIG,
+				"mkstemp fail for alloc inflight buffer\n");
+			return NULL;
+		}
+
+		if (ftruncate(mfd, size) == -1) {
+			RTE_LOG(ERR, VHOST_CONFIG,
+				"ftruncate fail for alloc inflight buffer\n");
+			close(mfd);
+			return NULL;
+		}
+	}
+
+	ptr = mmap(0, size, PROT_READ | PROT_WRITE, MAP_SHARED, mfd, 0);
+	if (ptr == MAP_FAILED) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"mmap fail for alloc inflight buffer\n");
+		close(mfd);
+		return NULL;
+	}
+
+	*fd = mfd;
+	return ptr;
+}
+
+static uint32_t
+get_pervq_shm_size_split(uint16_t queue_size)
+{
+	return ALIGN_UP(sizeof(struct rte_vhost_inflight_desc_split) *
+			queue_size + sizeof(uint64_t) + sizeof(uint16_t) * 4,
+			INFLIGHT_ALIGNMENT);
+}
+
+static uint32_t
+get_pervq_shm_size_packed(uint16_t queue_size)
+{
+	return ALIGN_UP(sizeof(struct rte_vhost_inflight_desc_packed) *
+			queue_size + sizeof(uint64_t) + sizeof(uint16_t) * 6 +
+			sizeof(uint8_t) * 9, INFLIGHT_ALIGNMENT);
+}
+
+static int
+vhost_user_get_inflight_fd(struct virtio_net **pdev,
+				     VhostUserMsg *msg,
+				     int main_fd __rte_unused)
+{
+	int fd, i, j;
+	uint64_t pervq_inflight_size, mmap_size;
+	void *addr;
+	uint16_t num_queues, queue_size;
+	struct virtio_net *dev = *pdev;
+	struct rte_vhost_inflight_info_packed *inflight_packed = NULL;
+
+	if (msg->size != sizeof(msg->payload.inflight)) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Invalid get_inflight_fd message size is %d",
+			msg->size);
+		return RTE_VHOST_MSG_RESULT_ERR;
+	}
+
+	dev->inflight_info = calloc(1, sizeof(struct inflight_mem_info));
+	if (!dev->inflight_info) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Failed to alloc dev inflight area");
+		return RTE_VHOST_MSG_RESULT_ERR;
+	}
+
+	num_queues = msg->payload.inflight.num_queues;
+	queue_size = msg->payload.inflight.queue_size;
+
+	RTE_LOG(INFO, VHOST_CONFIG, "get_inflight_fd num_queues: %u\n",
+		msg->payload.inflight.num_queues);
+	RTE_LOG(INFO, VHOST_CONFIG, "get_inflight_fd queue_size: %u\n",
+		msg->payload.inflight.queue_size);
+
+	if (vq_is_packed(dev))
+		pervq_inflight_size = get_pervq_shm_size_packed(queue_size);
+	else
+		pervq_inflight_size = get_pervq_shm_size_split(queue_size);
+
+	mmap_size = num_queues * pervq_inflight_size;
+	addr = inflight_mem_alloc("vhost-inflight", mmap_size, &fd);
+	if (!addr) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Failed to alloc vhost inflight area");
+			msg->payload.inflight.mmap_size = 0;
+		return RTE_VHOST_MSG_RESULT_ERR;
+	}
+	memset(addr, 0, mmap_size);
+
+	dev->inflight_info->addr = addr;
+	dev->inflight_info->size = msg->payload.inflight.mmap_size = mmap_size;
+	dev->inflight_info->fd = msg->fds[0] = fd;
+	msg->payload.inflight.mmap_offset = 0;
+	msg->fd_num = 1;
+
+	if (vq_is_packed(dev)) {
+		for (i = 0; i < num_queues; i++) {
+			inflight_packed =
+				(struct rte_vhost_inflight_info_packed *)addr;
+			inflight_packed->used_wrap_counter = 1;
+			inflight_packed->old_used_wrap_counter = 1;
+			for (j = 0; j < queue_size; j++)
+				inflight_packed->desc[j].next = j + 1;
+			addr = (void *)((char *)addr + pervq_inflight_size);
+		}
+	}
+
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"send inflight mmap_size: %lu\n",
+		msg->payload.inflight.mmap_size);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"send inflight mmap_offset: %lu\n",
+		msg->payload.inflight.mmap_offset);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"send inflight fd: %d\n", msg->fds[0]);
+
+	return RTE_VHOST_MSG_RESULT_REPLY;
+}
+
+static int
+vhost_user_set_inflight_fd(struct virtio_net **pdev, VhostUserMsg *msg,
+		int main_fd __rte_unused)
+{
+	int fd, i;
+	uint64_t mmap_size, mmap_offset;
+	uint16_t num_queues, queue_size;
+	uint32_t pervq_inflight_size;
+	void *addr;
+	struct vhost_virtqueue *vq;
+	struct virtio_net *dev = *pdev;
+
+	fd = msg->fds[0];
+	if (msg->size != sizeof(msg->payload.inflight) || fd < 0) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Invalid set_inflight_fd message size is %d,fd is %d\n",
+			msg->size, fd);
+		return RTE_VHOST_MSG_RESULT_ERR;
+	}
+
+	mmap_size = msg->payload.inflight.mmap_size;
+	mmap_offset = msg->payload.inflight.mmap_offset;
+	num_queues = msg->payload.inflight.num_queues;
+	queue_size = msg->payload.inflight.queue_size;
+
+	if (vq_is_packed(dev))
+		pervq_inflight_size = get_pervq_shm_size_packed(queue_size);
+	else
+		pervq_inflight_size = get_pervq_shm_size_split(queue_size);
+
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd mmap_size: %lu\n", mmap_size);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd mmap_offset: %lu\n", mmap_offset);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd num_queues: %u\n", num_queues);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd queue_size: %u\n", queue_size);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd fd: %d\n", fd);
+	RTE_LOG(INFO, VHOST_CONFIG,
+		"set_inflight_fd pervq_inflight_size: %d\n",
+		pervq_inflight_size);
+
+	if (!dev->inflight_info) {
+		dev->inflight_info = calloc(1,
+					    sizeof(struct inflight_mem_info));
+		if (dev->inflight_info == NULL) {
+			RTE_LOG(ERR, VHOST_CONFIG,
+				"Failed to alloc dev inflight area");
+			return RTE_VHOST_MSG_RESULT_ERR;
+		}
+	}
+
+	if (dev->inflight_info->addr)
+		munmap(dev->inflight_info->addr, dev->inflight_info->size);
+
+	addr = mmap(0, mmap_size, PROT_READ | PROT_WRITE, MAP_SHARED,
+			fd, mmap_offset);
+	if (addr == MAP_FAILED) {
+		RTE_LOG(ERR, VHOST_CONFIG, "failed to mmap share memory.\n");
+		return RTE_VHOST_MSG_RESULT_ERR;
+	}
+
+	if (dev->inflight_info->fd)
+		close(dev->inflight_info->fd);
+
+	dev->inflight_info->fd = fd;
+	dev->inflight_info->addr = addr;
+	dev->inflight_info->size = mmap_size;
+
+	for (i = 0; i < num_queues; i++) {
+		vq = dev->virtqueue[i];
+		if (vq_is_packed(dev)) {
+			vq->inflight_packed = addr;
+			vq->inflight_packed->desc_num = queue_size;
+		} else {
+			vq->inflight_split = addr;
+			vq->inflight_split->desc_num = queue_size;
+		}
+		addr = (void *)((char *)addr + pervq_inflight_size);
+	}
+
+	return RTE_VHOST_MSG_RESULT_OK;
+}
+
 static int
 vhost_user_set_vring_call(struct virtio_net **pdev, struct VhostUserMsg *msg,
 			int main_fd __rte_unused)
@@ -1762,9 +2029,10 @@ static vhost_message_handler_t vhost_message_handlers[VHOST_USER_MAX] = {
 	[VHOST_USER_POSTCOPY_ADVISE] = vhost_user_set_postcopy_advise,
 	[VHOST_USER_POSTCOPY_LISTEN] = vhost_user_set_postcopy_listen,
 	[VHOST_USER_POSTCOPY_END] = vhost_user_postcopy_end,
+	[VHOST_USER_GET_INFLIGHT_FD] = vhost_user_get_inflight_fd,
+	[VHOST_USER_SET_INFLIGHT_FD] = vhost_user_set_inflight_fd,
 };
 
-
 /* return bytes# of read on success or negative val on failure. */
 static int
 read_vhost_message(int sockfd, struct VhostUserMsg *msg)
diff --git a/lib/librte_vhost/vhost_user.h b/lib/librte_vhost/vhost_user.h
index 17a1d7bca..6563f7315 100644
--- a/lib/librte_vhost/vhost_user.h
+++ b/lib/librte_vhost/vhost_user.h
@@ -54,7 +54,9 @@ typedef enum VhostUserRequest {
 	VHOST_USER_POSTCOPY_ADVISE = 28,
 	VHOST_USER_POSTCOPY_LISTEN = 29,
 	VHOST_USER_POSTCOPY_END = 30,
-	VHOST_USER_MAX = 31
+	VHOST_USER_GET_INFLIGHT_FD = 31,
+	VHOST_USER_SET_INFLIGHT_FD = 32,
+	VHOST_USER_MAX = 33
 } VhostUserRequest;
 
 typedef enum VhostUserSlaveRequest {
-- 
2.17.2