Skip to contentSkip to conversation

add generic vDPA device support

PATCH Superseded Longpeng(Mike) via 3 participants 12 messages May 10, 2022 – May 12, 2022 Arm/Amlogic Meson SoC support

v4 of 11 Superseded Latest is v11 →

lore
Share & tools
lore
#0 lore raw

v4 cover add generic vDPA device support

From: Longpeng <[email protected]>

Hi guys,

With the generic vDPA device, QEMU won't need to touch the device types any more, such like vfio.

We can use the generic vDPA device as follow:
-device vhost-vdpa-device-pci,vhostdev=/dev/vhost-vdpa-X

I've done some simple tests on Huawei's offloading card (net, 0.95).

Changes v3 -> v4:
v3: https://www.mail-archive.com/[email protected]/msg877015.html
- reorganize the series [Stefano]
- fix some typos [Stefano]
- fix logical error in vhost_vdpa_device_realize [Stefano]

Changes v2 -> v3
Patch 4 & 5:
- only call vdpa ioctls in vdpa-dev.c [Stefano, Longpeng]
- s/VQS_NUM/VQS_COUNT [Stefano]
- check both vdpa_dev_fd and vdpa_dev [Stefano]
Patch 6:
- move all steps into vhost_vdpa_device_unrealize. [Stefano]

Changes RFC -> v2
Patch 1:
- rename 'pdev_id' to 'trans_devid' [Michael]
- only use transitional device id for the devices listed in the spec [Michael]
- use macros to make the id_info table clearer [Longpeng]
- add some modern devices in the id_info table [Longpeng]
Patch 2:
- remove the GET_VECTORS_NUM command [Jason]
Patch 4:
- expose vdpa_dev_fd as a QOM preperty [Stefan]
- introduce vhost_vdpa_device_get_u32 as a common function to make the code clearer [Stefan]
- fix the misleading description of 'dc->desc' [Stefano]
Patch 5:
- check returned number of virtqueues [Stefan]
Patch 6:
- init s->num_queues [Stefano]
- free s->dev.vqs [Stefano]

Longpeng (Mike) (4):
linux-headers: Update headers to Linux 5.18-rc6
virtio: get class_id and pci device id by the virtio id
vdpa: add vdpa-dev support
vdpa: add vdpa-dev-pci support

 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   2 +
 hw/virtio/vdpa-dev-pci.c     | 101 +++++++++
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 hw/virtio/virtio-pci.c       |  77 +++++++
 hw/virtio/virtio-pci.h       |   5 +
 include/hw/virtio/vdpa-dev.h |  43 ++++
 linux-headers/linux/vhost.h  |   7 +
 8 files changed, 625 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev-pci.c
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
Signature · 1 line
2.23.0
#1 lore raw

PATCH v4 1/4 linux-headers: Update headers to Linux 5.18-rc6

From: Longpeng <[email protected]>

Update headers to 5.18-rc6. I need latest vhost changes.

---
 linux-headers/linux/vhost.h | 7 +++++++
 1 file changed, 7 insertions(+)
linux-headers/linux/vhost.h +7 -0
index c998860d7b..5d99e7c242 100644
@@ -150,4 +150,11 @@
 /* Get the valid iova range */
 #define VHOST_VDPA_GET_IOVA_RANGE	_IOR(VHOST_VIRTIO, 0x78, \
 					     struct vhost_vdpa_iova_range)
+
+/* Get the config size */
+#define VHOST_VDPA_GET_CONFIG_SIZE	_IOR(VHOST_VIRTIO, 0x79, __u32)
+
+/* Get the count of all virtqueues */
+#define VHOST_VDPA_GET_VQS_COUNT	_IOR(VHOST_VIRTIO, 0x80, __u32)
+
 #endif
Signature · 1 line
2.23.0
#2 lore raw

PATCH v4 2/4 virtio: get class_id and pci device id by the virtio id

In reply to Longpeng(Mike) via

From: Longpeng <[email protected]>

Add helpers to get the "Transitional PCI Device ID" and "class_id" of the device specified by the "Virtio Device ID".

These helpers will be used to build the generic vDPA device later.

---
 hw/virtio/virtio-pci.c | 77 ++++++++++++++++++++++++++++++++++++++++++
 hw/virtio/virtio-pci.h |  5 +++
 2 files changed, 82 insertions(+)
hw/virtio/virtio-pci.c +77 -0
index 7cf1231c1c..fdfa205cee 100644
@@ -19,6 +19,7 @@
 
 #include "exec/memop.h"
 #include "standard-headers/linux/virtio_pci.h"
+#include "standard-headers/linux/virtio_ids.h"
 #include "hw/boards.h"
 #include "hw/virtio/virtio.h"
 #include "migration/qemu-file-types.h"
@@ -212,6 +213,79 @@ static int virtio_pci_load_queue(DeviceState *d, int n, QEMUFile *f)
     return 0;
 }
 
+typedef struct VirtIOPCIIDInfo {
+    /* virtio id */
+    uint16_t vdev_id;
+    /* pci device id for the transitional device */
+    uint16_t trans_devid;
+    uint16_t class_id;
+} VirtIOPCIIDInfo;
+
+#define VIRTIO_TRANS_DEV_ID_INFO(name, class)       \
+    {                                               \
+        .vdev_id = VIRTIO_ID_##name,                \
+        .trans_devid = PCI_DEVICE_ID_VIRTIO_##name, \
+        .class_id = class,                          \
+    }
+
+#define VIRTIO_MODERN_DEV_ID_NFO(name, class)       \
+    {                                               \
+        .vdev_id = VIRTIO_ID_##name,                \
+        .class_id = class,                          \
+    }
+
+static const VirtIOPCIIDInfo virtio_pci_id_info[] = {
+    /* Non-transitional devices */
+    VIRTIO_MODERN_DEV_ID_NFO(CRYPTO,    PCI_CLASS_OTHERS),
+    VIRTIO_MODERN_DEV_ID_NFO(FS,        PCI_CLASS_STORAGE_OTHER),
+    /* Transitional devices */
+    VIRTIO_TRANS_DEV_ID_INFO(NET,       PCI_CLASS_NETWORK_ETHERNET),
+    VIRTIO_TRANS_DEV_ID_INFO(BLOCK,     PCI_CLASS_STORAGE_SCSI),
+    VIRTIO_TRANS_DEV_ID_INFO(CONSOLE,   PCI_CLASS_COMMUNICATION_OTHER),
+    VIRTIO_TRANS_DEV_ID_INFO(SCSI,      PCI_CLASS_STORAGE_SCSI),
+    VIRTIO_TRANS_DEV_ID_INFO(9P,        PCI_BASE_CLASS_NETWORK),
+    VIRTIO_TRANS_DEV_ID_INFO(BALLOON,   PCI_CLASS_OTHERS),
+    VIRTIO_TRANS_DEV_ID_INFO(RNG,       PCI_CLASS_OTHERS),
+};
+
+static const VirtIOPCIIDInfo *virtio_pci_get_id_info(uint16_t vdev_id)
+{
+    const VirtIOPCIIDInfo *info = NULL;
+    int i;
+
+    for (i = 0; i < ARRAY_SIZE(virtio_pci_id_info); i++) {
+        if (virtio_pci_id_info[i].vdev_id == vdev_id) {
+            info = &virtio_pci_id_info[i];
+            break;
+        }
+    }
+
+    if (!info) {
+        /* The device id is invalid or not added to the id_info yet. */
+        error_report("Invalid virtio device(id %u)", vdev_id);
+        abort();
+    }
+
+    return info;
+}
+
+/*
+ * Get the Transitional Device ID for the specific device, return
+ * zero if the device is non-transitional.
+ */
+uint16_t virtio_pci_get_trans_devid(uint16_t device_id)
+{
+    return virtio_pci_get_id_info(device_id)->trans_devid;
+}
+
+/*
+ * Get the Class ID for the specific device.
+ */
+uint16_t virtio_pci_get_class_id(uint16_t device_id)
+{
+    return virtio_pci_get_id_info(device_id)->class_id;
+}
+
 static bool virtio_pci_ioeventfd_enabled(DeviceState *d)
 {
     VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
@@ -1675,6 +1749,9 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
          * is set to PCI_SUBVENDOR_ID_REDHAT_QUMRANET by default.
          */
         pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
+        if (proxy->trans_devid) {
+            pci_config_set_device_id(config, proxy->trans_devid);
+        }
     } else {
         /* pure virtio-1.0 */
         pci_set_word(config + PCI_VENDOR_ID,
hw/virtio/virtio-pci.h +5 -0
index 2446dcd9ae..f08665cd1b 100644
@@ -146,6 +146,8 @@ struct VirtIOPCIProxy {
     bool disable_modern;
     bool ignore_backend_features;
     OnOffAuto disable_legacy;
+    /* Transitional device id */
+    uint16_t trans_devid;
     uint32_t class_code;
     uint32_t nvectors;
     uint32_t dfselect;
@@ -158,6 +160,9 @@ struct VirtIOPCIProxy {
     VirtioBusState bus;
 };
 
+uint16_t virtio_pci_get_trans_devid(uint16_t device_id);
+uint16_t virtio_pci_get_class_id(uint16_t device_id);
+
 static inline bool virtio_pci_modern(VirtIOPCIProxy *proxy)
 {
     return !proxy->disable_modern;
Signature · 1 line
2.23.0
#3 lore raw

PATCH v4 3/4 vdpa: add vdpa-dev support

In reply to Longpeng(Mike) via

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   1 +
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 include/hw/virtio/vdpa-dev.h |  43 ++++
 4 files changed, 434 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
hw/virtio/Kconfig +5 -0
index c144d42f9b..2723283382 100644
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
     bool
     default y
     depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
 virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
 virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
 virtio_pci_ss = ss.source_set()
 virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +385 -0

385 lines not on this page · raw message

include/hw/virtio/vdpa-dev.h new file +43 -0

43 lines not on this page · raw message

Signature · 1 line
2.23.0
#4 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 20 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   1 +
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 include/hw/virtio/vdpa-dev.h |  43 ++++
 4 files changed, 434 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
hw/virtio/Kconfig +0 -0
index c144d42f9b..2723283382 100644

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 7 lines
hw/virtio/Kconfig +4 -0
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
     bool
     default y
     depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 80 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
 virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
 virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
 virtio_pci_ss = ss.source_set()
 virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +61 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;

hw/virtio/vdpa-dev.c line 71

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 250 lines
hw/virtio/vdpa-dev.c new file +250 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {

hw/virtio/vdpa-dev.c line 331

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

Thanks

Quoted 106 lines
hw/virtio/vdpa-dev.c new file +54 -0
+    DEFINE_PROP_UINT16("queue-size", VhostVdpaDevice, queue_size, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static const VMStateDescription vmstate_vhost_vdpa_device = {
+    .name = "vhost-vdpa-device",
+    .unmigratable = 1,
+    .minimum_version_id = 1,
+    .version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_VIRTIO_DEVICE,
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static void vhost_vdpa_device_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    device_class_set_props(dc, vhost_vdpa_device_properties);
+    dc->desc = "VDPA-based generic device assignment";
+    dc->vmsd = &vmstate_vhost_vdpa_device;
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    vdc->realize = vhost_vdpa_device_realize;
+    vdc->unrealize = vhost_vdpa_device_unrealize;
+    vdc->get_config = vhost_vdpa_device_get_config;
+    vdc->set_config = vhost_vdpa_device_set_config;
+    vdc->get_features = vhost_vdpa_device_get_features;
+    vdc->set_status = vhost_vdpa_device_set_status;
+}
+
+static void vhost_vdpa_device_instance_init(Object *obj)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(obj);
+
+    device_add_bootindex_property(obj, &s->bootindex, "bootindex",
+                                  NULL, DEVICE(obj));
+}
+
+static const TypeInfo vhost_vdpa_device_info = {
+    .name = TYPE_VHOST_VDPA_DEVICE,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VhostVdpaDevice),
+    .class_init = vhost_vdpa_device_class_init,
+    .instance_init = vhost_vdpa_device_instance_init,
+};
+
+static void register_vhost_vdpa_device_type(void)
+{
+    type_register_static(&vhost_vdpa_device_info);
+}
+
+type_init(register_vhost_vdpa_device_type);
include/hw/virtio/vdpa-dev.h new file +43 -1
new file mode 100644
index 0000000000..4dbf98195c
@@ -0,0 +1,43 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk.h" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#ifndef _VHOST_VDPA_DEVICE_H
+#define _VHOST_VDPA_DEVICE_H
+
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/vhost-vdpa.h"
+#include "qom/object.h"
+
+
+#define TYPE_VHOST_VDPA_DEVICE "vhost-vdpa-device"
+OBJECT_DECLARE_SIMPLE_TYPE(VhostVdpaDevice, VHOST_VDPA_DEVICE)
+
+struct VhostVdpaDevice {
+    VirtIODevice parent_obj;
+    char *vhostdev;
+    int vhostfd;
+    int32_t bootindex;
+    uint32_t vdev_id;
+    uint32_t num_queues;
+    struct vhost_dev dev;
+    struct vhost_vdpa vdpa;
+    VirtQueue **virtqs;
+    uint8_t *config;
+    int config_size;
+    uint16_t queue_size;
+    bool started;
+    int (*post_init)(VhostVdpaDevice *v, Error **errp);
+};
+
+#endif
--

2.23.0

#5 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 16 lines · Jason Wang
Quoted 15 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
  hw/virtio/Kconfig            |   5 +
  hw/virtio/meson.build        |   1 +
  hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
  include/hw/virtio/vdpa-dev.h |  43 ++++
  4 files changed, 434 insertions(+)

create mode 100644 hw/virtio/vdpa-dev.c
create mode 100644 include/hw/virtio/vdpa-dev.h

Quoted 15 lines

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 12 lines
hw/virtio/Kconfig +4 -0
index c144d42f9b..2723283382 100644
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
      bool
      default y
      depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

There is currently no plan. Maybe we can try to support it in the next step.

Quoted 85 lines
Quoted 85 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
  virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
  virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
  virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
  virtio_pci_ss = ss.source_set()
  virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +66 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
Quoted 9 lines

hw/virtio/vdpa-dev.c line 71

Quoted 5 lines
hw/virtio/vdpa-dev.c new file +5 -0
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

I think the error messages already set in *errp, right?

Quoted 254 lines
Quoted 254 lines
hw/virtio/vdpa-dev.c new file +254 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
Quoted 9 lines

hw/virtio/vdpa-dev.c line 331

Quoted 6 lines
hw/virtio/vdpa-dev.c new file +6 -0
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

It seems "vhostdev" is enough. I prefer to remove "vhostfd" if there is no objection.

Thanks.

Quoted 110 lines

Thanks

Quoted 104 lines
hw/virtio/vdpa-dev.c new file +54 -0
+    DEFINE_PROP_UINT16("queue-size", VhostVdpaDevice, queue_size, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static const VMStateDescription vmstate_vhost_vdpa_device = {
+    .name = "vhost-vdpa-device",
+    .unmigratable = 1,
+    .minimum_version_id = 1,
+    .version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_VIRTIO_DEVICE,
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static void vhost_vdpa_device_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    device_class_set_props(dc, vhost_vdpa_device_properties);
+    dc->desc = "VDPA-based generic device assignment";
+    dc->vmsd = &vmstate_vhost_vdpa_device;
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    vdc->realize = vhost_vdpa_device_realize;
+    vdc->unrealize = vhost_vdpa_device_unrealize;
+    vdc->get_config = vhost_vdpa_device_get_config;
+    vdc->set_config = vhost_vdpa_device_set_config;
+    vdc->get_features = vhost_vdpa_device_get_features;
+    vdc->set_status = vhost_vdpa_device_set_status;
+}
+
+static void vhost_vdpa_device_instance_init(Object *obj)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(obj);
+
+    device_add_bootindex_property(obj, &s->bootindex, "bootindex",
+                                  NULL, DEVICE(obj));
+}
+
+static const TypeInfo vhost_vdpa_device_info = {
+    .name = TYPE_VHOST_VDPA_DEVICE,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VhostVdpaDevice),
+    .class_init = vhost_vdpa_device_class_init,
+    .instance_init = vhost_vdpa_device_instance_init,
+};
+
+static void register_vhost_vdpa_device_type(void)
+{
+    type_register_static(&vhost_vdpa_device_info);
+}
+
+type_init(register_vhost_vdpa_device_type);
include/hw/virtio/vdpa-dev.h new file +43 -1
new file mode 100644
index 0000000000..4dbf98195c
@@ -0,0 +1,43 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk.h" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#ifndef _VHOST_VDPA_DEVICE_H
+#define _VHOST_VDPA_DEVICE_H
+
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/vhost-vdpa.h"
+#include "qom/object.h"
+
+
+#define TYPE_VHOST_VDPA_DEVICE "vhost-vdpa-device"
+OBJECT_DECLARE_SIMPLE_TYPE(VhostVdpaDevice, VHOST_VDPA_DEVICE)
+
+struct VhostVdpaDevice {
+    VirtIODevice parent_obj;
+    char *vhostdev;
+    int vhostfd;
+    int32_t bootindex;
+    uint32_t vdev_id;
+    uint32_t num_queues;
+    struct vhost_dev dev;
+    struct vhost_vdpa vdpa;
+    VirtQueue **virtqs;
+    uint8_t *config;
+    int config_size;
+    uint16_t queue_size;
+    bool started;
+    int (*post_init)(VhostVdpaDevice *v, Error **errp);
+};
+
+#endif
--
Quoted 2 lines

2.23.0

.

#6 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 35 lines · Cloud Infrastructure Service Product Dept.)
Quoted 16 lines · Jason Wang
Quoted 15 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
  hw/virtio/Kconfig            |   5 +
  hw/virtio/meson.build        |   1 +
  hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
  include/hw/virtio/vdpa-dev.h |  43 ++++
  4 files changed, 434 insertions(+)

create mode 100644 hw/virtio/vdpa-dev.c
create mode 100644 include/hw/virtio/vdpa-dev.h

Quoted 15 lines

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 12 lines
hw/virtio/Kconfig +4 -0
index c144d42f9b..2723283382 100644
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
      bool
      default y
      depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 1 line

There is currently no plan. Maybe we can try to support it in the next step.

That would be better, allowing MMIO on top of vhost-vDPA is considered to be one of the important advantages.

Quoted 95 lines
Quoted 86 lines
Quoted 86 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
  virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
  virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
  virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
  virtio_pci_ss = ss.source_set()
  virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +66 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
Quoted 9 lines

hw/virtio/vdpa-dev.c line 71

Quoted 5 lines
hw/virtio/vdpa-dev.c new file +5 -0
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 1 line

I think the error messages already set in *errp, right?

Oh right.

Quoted 264 lines
Quoted 255 lines
Quoted 255 lines
hw/virtio/vdpa-dev.c new file +254 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
Quoted 9 lines

hw/virtio/vdpa-dev.c line 331

Quoted 6 lines
hw/virtio/vdpa-dev.c new file +6 -0
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

Quoted 2 lines

It seems "vhostdev" is enough. I prefer to remove "vhostfd" if there is no objection.

Yes, it is the way the current vhost-vDPA networking backend did.

Thanks

Quoted 114 lines

Thanks.

Quoted 110 lines

Thanks

Quoted 104 lines
hw/virtio/vdpa-dev.c new file +54 -0
+    DEFINE_PROP_UINT16("queue-size", VhostVdpaDevice, queue_size, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static const VMStateDescription vmstate_vhost_vdpa_device = {
+    .name = "vhost-vdpa-device",
+    .unmigratable = 1,
+    .minimum_version_id = 1,
+    .version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_VIRTIO_DEVICE,
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static void vhost_vdpa_device_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    device_class_set_props(dc, vhost_vdpa_device_properties);
+    dc->desc = "VDPA-based generic device assignment";
+    dc->vmsd = &vmstate_vhost_vdpa_device;
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    vdc->realize = vhost_vdpa_device_realize;
+    vdc->unrealize = vhost_vdpa_device_unrealize;
+    vdc->get_config = vhost_vdpa_device_get_config;
+    vdc->set_config = vhost_vdpa_device_set_config;
+    vdc->get_features = vhost_vdpa_device_get_features;
+    vdc->set_status = vhost_vdpa_device_set_status;
+}
+
+static void vhost_vdpa_device_instance_init(Object *obj)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(obj);
+
+    device_add_bootindex_property(obj, &s->bootindex, "bootindex",
+                                  NULL, DEVICE(obj));
+}
+
+static const TypeInfo vhost_vdpa_device_info = {
+    .name = TYPE_VHOST_VDPA_DEVICE,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VhostVdpaDevice),
+    .class_init = vhost_vdpa_device_class_init,
+    .instance_init = vhost_vdpa_device_instance_init,
+};
+
+static void register_vhost_vdpa_device_type(void)
+{
+    type_register_static(&vhost_vdpa_device_info);
+}
+
+type_init(register_vhost_vdpa_device_type);
include/hw/virtio/vdpa-dev.h new file +43 -1
new file mode 100644
index 0000000000..4dbf98195c
@@ -0,0 +1,43 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk.h" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#ifndef _VHOST_VDPA_DEVICE_H
+#define _VHOST_VDPA_DEVICE_H
+
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/vhost-vdpa.h"
+#include "qom/object.h"
+
+
+#define TYPE_VHOST_VDPA_DEVICE "vhost-vdpa-device"
+OBJECT_DECLARE_SIMPLE_TYPE(VhostVdpaDevice, VHOST_VDPA_DEVICE)
+
+struct VhostVdpaDevice {
+    VirtIODevice parent_obj;
+    char *vhostdev;
+    int vhostfd;
+    int32_t bootindex;
+    uint32_t vdev_id;
+    uint32_t num_queues;
+    struct vhost_dev dev;
+    struct vhost_vdpa vdpa;
+    VirtQueue **virtqs;
+    uint8_t *config;
+    int config_size;
+    uint16_t queue_size;
+    bool started;
+    int (*post_init)(VhostVdpaDevice *v, Error **errp);
+};
+
+#endif
--
Quoted 2 lines

2.23.0

.

#7 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 37 lines · Jason Wang
Quoted 35 lines · Cloud Infrastructure Service Product Dept.)
Quoted 31 lines · Jason Wang
Quoted 20 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
   hw/virtio/Kconfig            |   5 +
   hw/virtio/meson.build        |   1 +
   hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
   include/hw/virtio/vdpa-dev.h |  43 ++++
   4 files changed, 434 insertions(+)

create mode 100644 hw/virtio/vdpa-dev.c create mode 100644 include/hw/virtio/vdpa-dev.h

hw/virtio/Kconfig +0 -0
index c144d42f9b..2723283382 100644

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 7 lines
hw/virtio/Kconfig +4 -0
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
       bool
       default y
       depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 5 lines
Quoted 1 line

There is currently no plan. Maybe we can try to support it in the next step.

That would be better, allowing MMIO on top of vhost-vDPA is considered to be one of the important advantages.

OK, I've done some simple tests, and it works fine. I'll send v5 later.

Quoted 486 lines
Quoted 95 lines
Quoted 86 lines
Quoted 86 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
   virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
   virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
   virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
   virtio_pci_ss = ss.source_set()
   virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +66 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
Quoted 9 lines

hw/virtio/vdpa-dev.c line 71

Quoted 5 lines
hw/virtio/vdpa-dev.c new file +5 -0
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 1 line

I think the error messages already set in *errp, right?

Oh right.

Quoted 264 lines
Quoted 255 lines
Quoted 255 lines
hw/virtio/vdpa-dev.c new file +254 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
Quoted 9 lines

hw/virtio/vdpa-dev.c line 331

Quoted 6 lines
hw/virtio/vdpa-dev.c new file +6 -0
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

Quoted 2 lines

It seems "vhostdev" is enough. I prefer to remove "vhostfd" if there is no objection.

Yes, it is the way the current vhost-vDPA networking backend did.

Thanks

Quoted 114 lines

Thanks.

Quoted 110 lines

Thanks

Quoted 104 lines
hw/virtio/vdpa-dev.c new file +54 -0
+    DEFINE_PROP_UINT16("queue-size", VhostVdpaDevice, queue_size, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static const VMStateDescription vmstate_vhost_vdpa_device = {
+    .name = "vhost-vdpa-device",
+    .unmigratable = 1,
+    .minimum_version_id = 1,
+    .version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_VIRTIO_DEVICE,
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static void vhost_vdpa_device_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    device_class_set_props(dc, vhost_vdpa_device_properties);
+    dc->desc = "VDPA-based generic device assignment";
+    dc->vmsd = &vmstate_vhost_vdpa_device;
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    vdc->realize = vhost_vdpa_device_realize;
+    vdc->unrealize = vhost_vdpa_device_unrealize;
+    vdc->get_config = vhost_vdpa_device_get_config;
+    vdc->set_config = vhost_vdpa_device_set_config;
+    vdc->get_features = vhost_vdpa_device_get_features;
+    vdc->set_status = vhost_vdpa_device_set_status;
+}
+
+static void vhost_vdpa_device_instance_init(Object *obj)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(obj);
+
+    device_add_bootindex_property(obj, &s->bootindex, "bootindex",
+                                  NULL, DEVICE(obj));
+}
+
+static const TypeInfo vhost_vdpa_device_info = {
+    .name = TYPE_VHOST_VDPA_DEVICE,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VhostVdpaDevice),
+    .class_init = vhost_vdpa_device_class_init,
+    .instance_init = vhost_vdpa_device_instance_init,
+};
+
+static void register_vhost_vdpa_device_type(void)
+{
+    type_register_static(&vhost_vdpa_device_info);
+}
+
+type_init(register_vhost_vdpa_device_type);
include/hw/virtio/vdpa-dev.h new file +43 -1
new file mode 100644
index 0000000000..4dbf98195c
@@ -0,0 +1,43 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk.h" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#ifndef _VHOST_VDPA_DEVICE_H
+#define _VHOST_VDPA_DEVICE_H
+
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/vhost-vdpa.h"
+#include "qom/object.h"
+
+
+#define TYPE_VHOST_VDPA_DEVICE "vhost-vdpa-device"
+OBJECT_DECLARE_SIMPLE_TYPE(VhostVdpaDevice, VHOST_VDPA_DEVICE)
+
+struct VhostVdpaDevice {
+    VirtIODevice parent_obj;
+    char *vhostdev;
+    int vhostfd;
+    int32_t bootindex;
+    uint32_t vdev_id;
+    uint32_t num_queues;
+    struct vhost_dev dev;
+    struct vhost_vdpa vdpa;
+    VirtQueue **virtqs;
+    uint8_t *config;
+    int config_size;
+    uint16_t queue_size;
+    bool started;
+    int (*post_init)(VhostVdpaDevice *v, Error **errp);
+};
+
+#endif
--
Quoted 2 lines

2.23.0

.

.

#8 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

In reply to Jason Wang
Quoted 379 lines · Jason Wang
Quoted 20 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   1 +
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 include/hw/virtio/vdpa-dev.h |  43 ++++
 4 files changed, 434 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
hw/virtio/Kconfig +0 -0
index c144d42f9b..2723283382 100644

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 7 lines
hw/virtio/Kconfig +4 -0
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
     bool
     default y
     depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 80 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
 virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
 virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
 virtio_pci_ss = ss.source_set()
 virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +61 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;

hw/virtio/vdpa-dev.c line 71

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 254 lines
hw/virtio/vdpa-dev.c new file +254 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
Quoted 8 lines

hw/virtio/vdpa-dev.c line 331

Quoted 6 lines
hw/virtio/vdpa-dev.c new file +6 -0
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

IIRC for other vhost devices (e.g. vhost-vsock) the management layer (e.g. libvirt) opens the device (because I think QEMU has less permissions) and then passes the fd to QEMU, could this be useful for this scenario as well?

Thanks,
Stefano

#9 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 391 lines · Stefano Garzarella
Quoted 380 lines · Jason Wang
Quoted 20 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   1 +
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 include/hw/virtio/vdpa-dev.h |  43 ++++
 4 files changed, 434 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
hw/virtio/Kconfig +0 -0
index c144d42f9b..2723283382 100644

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 7 lines
hw/virtio/Kconfig +4 -0
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
     bool
     default y
     depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 80 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
 virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
 virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
 virtio_pci_ss = ss.source_set()
 virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +61 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;

hw/virtio/vdpa-dev.c line 71

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 255 lines
hw/virtio/vdpa-dev.c new file +255 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
+    }
Quoted 8 lines

hw/virtio/vdpa-dev.c line 331

Quoted 5 lines
hw/virtio/vdpa-dev.c new file +5 -0
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

Quoted 4 lines

IIRC for other vhost devices (e.g. vhost-vsock) the management layer (e.g. libvirt) opens the device (because I think QEMU has less permissions) and then passes the fd to QEMU, could this be useful for this scenario as well?

Yes, it's the way current vhost-vDPA did (the fd is passing via SCM_RIGHTS).

Thanks

Quoted 4 lines

Thanks,
Stefano

#10 lore raw

Re: [PATCH v4 3/4] vdpa: add vdpa-dev support

Quoted 392 lines · Jason Wang
Quoted 391 lines · Stefano Garzarella
Quoted 388 lines · Jason Wang
Quoted 20 lines · Longpeng(Mike)

From: Longpeng <[email protected]>

Supports vdpa-dev.

---
 hw/virtio/Kconfig            |   5 +
 hw/virtio/meson.build        |   1 +
 hw/virtio/vdpa-dev.c         | 385 +++++++++++++++++++++++++++++++++++
 include/hw/virtio/vdpa-dev.h |  43 ++++
 4 files changed, 434 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev.c
 create mode 100644 include/hw/virtio/vdpa-dev.h
hw/virtio/Kconfig +0 -0
index c144d42f9b..2723283382 100644

hw/virtio/Kconfig config VHOST_USER_RNG · line 74

Quoted 7 lines
hw/virtio/Kconfig +4 -0
@@ -68,3 +68,8 @@ config VHOST_USER_RNG
     bool
     default y
     depends on VIRTIO && VHOST_USER
+
+config VHOST_VDPA_DEV
+    bool
+    default y if VIRTIO_PCI

Do we have the plan to add VIRTIO_MMIO support?

Quoted 80 lines
hw/virtio/Kconfig +1 -0
+    depends on VIRTIO && VHOST_VDPA && LINUX
hw/virtio/meson.build +1 -0
index 67dc77e00f..8f6f86db71 100644
@@ -29,6 +29,7 @@ virtio_ss.add(when: 'CONFIG_VHOST_USER_I2C', if_true: files('vhost-user-i2c.c'))
 virtio_ss.add(when: ['CONFIG_VIRTIO_PCI', 'CONFIG_VHOST_USER_I2C'], if_true: files('vhost-user-i2c-pci.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER_RNG', if_true: files('vhost-user-rng.c'))
 virtio_ss.add(when: ['CONFIG_VHOST_USER_RNG', 'CONFIG_VIRTIO_PCI'], if_true: files('vhost-user-rng-pci.c'))
+virtio_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev.c'))
 
 virtio_pci_ss = ss.source_set()
 virtio_pci_ss.add(when: 'CONFIG_VHOST_VSOCK', if_true: files('vhost-vsock-pci.c'))
hw/virtio/vdpa-dev.c new file +61 -0
new file mode 100644
index 0000000000..543b5b4b81
@@ -0,0 +1,385 @@
+/*
+ * Vhost Vdpa Device
+ *
+ * Copyright (c) Huawei Technologies Co., Ltd. 2022. All Rights Reserved.
+ *
+ * Authors:
+ *   Longpeng <[email protected]>
+ *
+ * Largely based on the "vhost-user-blk-pci.c" and "vhost-user-blk.c" implemented by:
+ *   Changpeng Liu <[email protected]>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+#include "qemu/osdep.h"
+#include <sys/ioctl.h>
+#include <linux/vhost.h>
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "qemu/cutils.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "hw/qdev-properties-system.h"
+#include "hw/virtio/vhost.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/vdpa-dev.h"
+#include "sysemu/sysemu.h"
+#include "sysemu/runstate.h"
+
+static void
+vhost_vdpa_device_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+{
+    /* Nothing to do */
+}
+
+static uint32_t
+vhost_vdpa_device_get_u32(int fd, unsigned long int cmd, Error **errp)
+{
+    uint32_t val = (uint32_t)-1;
+
+    if (ioctl(fd, cmd, &val) < 0) {
+        error_setg(errp, "vhost-vdpa-device: cmd 0x%lx failed: %s",
+                   cmd, strerror(errno));
+    }
+
+    return val;
+}
+
+static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *v = VHOST_VDPA_DEVICE(vdev);
+    uint16_t max_queue_size;
+    struct vhost_virtqueue *vqs;
+    int i, ret;
+
+    if (!v->vhostdev && v->vhostfd == -1) {
+        error_setg(errp, "both vhostdev and vhostfd are missing");
+        return;

hw/virtio/vdpa-dev.c line 71

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+    }
+
+    if (v->vhostdev && v->vhostfd != -1) {
+        error_setg(errp, "both vhostdev and vhostfd are set");
+        return;
+    }
+
+    if (v->vhostfd == -1) {
+        v->vhostfd = qemu_open(v->vhostdev, O_RDWR, errp);
+        if (*errp) {

Is this better to set error messages for all the possible failures during realization?

Quoted 250 lines
hw/virtio/vdpa-dev.c new file +250 -0
+            return;
+        }
+    }
+    v->vdpa.device_fd = v->vhostfd;
+
+    v->vdev_id = vhost_vdpa_device_get_u32(v->vhostfd,
+                                           VHOST_VDPA_GET_DEVICE_ID, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    max_queue_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_VRING_NUM, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (v->queue_size > max_queue_size) {
+        error_setg(errp, "vhost-vdpa-device: invalid queue_size: %u (max:%u)",
+                   v->queue_size, max_queue_size);
+        goto out;
+    } else if (!v->queue_size) {
+        v->queue_size = max_queue_size;
+    }
+
+    v->num_queues = vhost_vdpa_device_get_u32(v->vhostfd,
+                                              VHOST_VDPA_GET_VQS_COUNT, errp);
+    if (*errp) {
+        goto out;
+    }
+
+    if (!v->num_queues || v->num_queues > VIRTIO_QUEUE_MAX) {
+        error_setg(errp, "invalid number of virtqueues: %u (max:%u)",
+                   v->num_queues, VIRTIO_QUEUE_MAX);
+        goto out;
+    }
+
+    v->dev.nvqs = v->num_queues;
+    vqs = g_new0(struct vhost_virtqueue, v->dev.nvqs);
+    v->dev.vqs = vqs;
+    v->dev.vq_index = 0;
+    v->dev.vq_index_end = v->dev.nvqs;
+    v->dev.backend_features = 0;
+    v->started = false;
+
+    ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: vhost initialization failed: %s",
+                   strerror(-ret));
+        goto free_vqs;
+    }
+
+    v->config_size = vhost_vdpa_device_get_u32(v->vhostfd,
+                                               VHOST_VDPA_GET_CONFIG_SIZE, errp);
+    if (*errp) {
+        goto vhost_cleanup;
+    }
+
+    if (v->post_init && v->post_init(v, errp) < 0) {
+        goto free_virtio;
+    }
+
+    v->config = g_malloc0(v->config_size);
+
+    ret = vhost_dev_get_config(&v->dev, v->config, v->config_size, NULL);
+    if (ret < 0) {
+        error_setg(errp, "vhost-vdpa-device: get config failed");
+        goto free_config;
+    }
+
+    virtio_init(vdev, "vhost-vdpa", v->vdev_id, v->config_size);
+
+    v->virtqs = g_new0(VirtQueue *, v->dev.nvqs);
+    for (i = 0; i < v->dev.nvqs; i++) {
+        v->virtqs[i] = virtio_add_queue(vdev, v->queue_size,
+                                        vhost_vdpa_device_dummy_handle_output);
+    }
+
+    return;
+
+free_virtio:
+    for (i = 0; i < v->num_queues; i++) {
+        virtio_delete_queue(v->virtqs[i]);
+    }
+    g_free(v->virtqs);
+    virtio_cleanup(vdev);
+free_config:
+    g_free(v->config);
+vhost_cleanup:
+    vhost_dev_cleanup(&v->dev);
+free_vqs:
+    g_free(vqs);
+out:
+    qemu_close(v->vhostfd);
+    v->vhostfd = -1;
+}
+
+static void vhost_vdpa_device_unrealize(DeviceState *dev)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int i;
+
+    virtio_set_status(vdev, 0);
+
+    for (i = 0; i < s->num_queues; i++) {
+        virtio_delete_queue(s->virtqs[i]);
+    }
+    g_free(s->virtqs);
+    virtio_cleanup(vdev);
+
+    g_free(s->config);
+    g_free(s->dev.vqs);
+    vhost_dev_cleanup(&s->dev);
+    qemu_close(s->vhostfd);
+    s->vhostfd = -1;
+}
+
+static void
+vhost_vdpa_device_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+
+    memcpy(config, s->config, s->config_size);
+}
+
+static void
+vhost_vdpa_device_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    int ret;
+
+    ret = vhost_dev_set_config(&s->dev, s->config, 0, s->config_size,
+                               VHOST_SET_CONFIG_TYPE_MASTER);
+    if (ret) {
+        error_report("set device config space failed");
+        return;
+    }
+}
+
+static uint64_t vhost_vdpa_device_get_features(VirtIODevice *vdev,
+                                               uint64_t features,
+                                               Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    uint64_t backend_features = s->dev.features;
+
+    if (!virtio_has_feature(features, VIRTIO_F_IOMMU_PLATFORM)) {
+        virtio_clear_feature(&backend_features, VIRTIO_F_IOMMU_PLATFORM);
+    }
+
+    return backend_features;
+}
+
+static int vhost_vdpa_device_start(VirtIODevice *vdev, Error **errp)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int i, ret;
+
+    if (!k->set_guest_notifiers) {
+        error_setg(errp, "binding does not support guest notifiers");
+        return -ENOSYS;
+    }
+
+    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error enabling host notifiers");
+        return ret;
+    }
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error binding guest notifier");
+        goto err_host_notifiers;
+    }
+
+    s->dev.acked_features = vdev->guest_features;
+
+    ret = vhost_dev_start(&s->dev, vdev);
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Error starting vhost");
+        goto err_guest_notifiers;
+    }
+    s->started = true;
+
+    /*
+     * guest_notifier_mask/pending not used yet, so just unmask
+     * everything here. virtio-pci will do the right thing by
+     * enabling/disabling irqfd.
+     */
+    for (i = 0; i < s->dev.nvqs; i++) {
+        vhost_virtqueue_mask(&s->dev, vdev, i, false);
+    }
+
+    return ret;
+
+err_guest_notifiers:
+    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+err_host_notifiers:
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+    return ret;
+}
+
+static void vhost_vdpa_device_stop(VirtIODevice *vdev)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
+    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
+    int ret;
+
+    if (!s->started) {
+        return;
+    }
+    s->started = false;
+
+    if (!k->set_guest_notifiers) {
+        return;
+    }
+
+    vhost_dev_stop(&s->dev, vdev);
+
+    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
+    if (ret < 0) {
+        error_report("vhost guest notifier cleanup failed: %d", ret);
+        return;
+    }
+
+    vhost_dev_disable_notifiers(&s->dev, vdev);
+}
+
+static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
+    bool should_start = virtio_device_started(vdev, status);
+    Error *local_err = NULL;
+    int ret;
+
+    if (!vdev->vm_running) {
+        should_start = false;
+    }
+
+    if (s->started == should_start) {
+        return;
+    }
+
+    if (should_start) {
+        ret = vhost_vdpa_device_start(vdev, &local_err);
+        if (ret < 0) {

hw/virtio/vdpa-dev.c line 331

Quoted 10 lines
hw/virtio/vdpa-dev.c new file +10 -0
+            error_reportf_err(local_err, "vhost-vdpa-device: start failed: ");
+        }
+    } else {
+        vhost_vdpa_device_stop(vdev);
+    }
+}
+
+static Property vhost_vdpa_device_properties[] = {
+    DEFINE_PROP_STRING("vhostdev", VhostVdpaDevice, vhostdev),
+    DEFINE_PROP_INT32("vhostfd", VhostVdpaDevice, vhostfd, -1),

This is probably not needed since we can "abuse" /dev/fd/X for vhostdev.

Quoted 6 lines
Quoted 4 lines

IIRC for other vhost devices (e.g. vhost-vsock) the management layer (e.g. libvirt) opens the device (because I think QEMU has less permissions) and then passes the fd to QEMU, could this be useful for this scenario as well?

Yes, it's the way current vhost-vDPA did (the fd is passing via SCM_RIGHTS).

Okay, I looked at qemu_open() and I see what you mean. Basically libvirt can pass the fd directly with vhostdev=/dev/fdset/X instead of having a dedicated property.

I did not know that, I agree to remove "vhostfd" then!

Thanks for the clarification,
Stefano

#11 lore raw

PATCH v4 4/4 vdpa: add vdpa-dev-pci support

In reply to Longpeng(Mike) via

From: Longpeng <[email protected]>

Supports vdpa-dev-pci, we can use the device as follow:

-device vhost-vdpa-device-pci,vhostdev=/dev/vhost-vdpa-X

---
 hw/virtio/meson.build    |   1 +
 hw/virtio/vdpa-dev-pci.c | 101 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 102 insertions(+)
 create mode 100644 hw/virtio/vdpa-dev-pci.c
hw/virtio/meson.build +1 -0
index 8f6f86db71..c2da69616f 100644
@@ -50,6 +50,7 @@ virtio_pci_ss.add(when: 'CONFIG_VIRTIO_SERIAL', if_true: files('virtio-serial-pc
 virtio_pci_ss.add(when: 'CONFIG_VIRTIO_PMEM', if_true: files('virtio-pmem-pci.c'))
 virtio_pci_ss.add(when: 'CONFIG_VIRTIO_IOMMU', if_true: files('virtio-iommu-pci.c'))
 virtio_pci_ss.add(when: 'CONFIG_VIRTIO_MEM', if_true: files('virtio-mem-pci.c'))
+virtio_pci_ss.add(when: 'CONFIG_VHOST_VDPA_DEV', if_true: files('vdpa-dev-pci.c'))
 
 virtio_ss.add_all(when: 'CONFIG_VIRTIO_PCI', if_true: virtio_pci_ss)
 
hw/virtio/vdpa-dev-pci.c new file +101 -0

101 lines not on this page · raw message

Signature · 1 line
2.23.0