From 904f58c226620a58fc7df42f6c693a364a24a1e9 Mon Sep 17 00:00:00 2001 From: Hao Chen Date: Thu, 21 Dec 2023 18:43:11 +0100 Subject: [PATCH 1/2] vdpa: move iova tree to the shared struct MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Next patches will register the vhost_vdpa memory listener while the VM is migrating at the destination, so we can map the memory to the device before stopping the VM at the source. The main goal is to reduce the downtime. However, the destination QEMU is unaware of which vhost_vdpa device will register its memory_listener. If the source guest has CVQ enabled, it will be the CVQ device. Otherwise, it will be the first one. Move the iova tree to VhostVDPAShared so all vhost_vdpa can use it, rather than always in the first or last vhost_vdpa. Backport from QEMU official community: 5edb02e8004c Signed-off-by: Eugenio Pérez Acked-by: Jason Wang Message-Id: <20231221174322.3130442-3-eperezma@redhat.com> Tested-by: Lei Yang Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/virtio/vhost-vdpa.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c index b5fb89b98e..ca6aefe7b9 100644 --- a/hw/virtio/vhost-vdpa.c +++ b/hw/virtio/vhost-vdpa.c @@ -359,7 +359,7 @@ static void vhost_vdpa_listener_region_add(MemoryListener *listener, mem_region.size = int128_get64(llsize) - 1, mem_region.perm = IOMMU_ACCESS_FLAG(true, section->readonly), - r = vhost_iova_tree_map_alloc(v->iova_tree, &mem_region); + r = vhost_iova_tree_map_alloc(v->shared->iova_tree, &mem_region); if (unlikely(r != IOVA_OK)) { error_report("Can't allocate a mapping (%d)", r); goto fail; @@ -380,7 +380,7 @@ static void vhost_vdpa_listener_region_add(MemoryListener *listener, fail_map: if (v->shadow_data) { - vhost_iova_tree_remove(v->iova_tree, mem_region); + vhost_iova_tree_remove(v->shared->iova_tree, mem_region); } fail: @@ -442,13 +442,13 @@ static void vhost_vdpa_listener_region_del(MemoryListener *listener, .size = int128_get64(llsize) - 1, }; - result = vhost_iova_tree_find_iova(v->iova_tree, &mem_region); + result = vhost_iova_tree_find_iova(v->shared->iova_tree, &mem_region); if (!result) { /* The memory listener map wasn't mapped */ return; } iova = result->iova; - vhost_iova_tree_remove(v->iova_tree, *result); + vhost_iova_tree_remove(v->shared->iova_tree, *result); } vhost_vdpa_iotlb_batch_begin_once(v); /* @@ -1071,7 +1071,8 @@ static void vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr addr) const DMAMap needle = { .translated_addr = addr, }; - const DMAMap *result = vhost_iova_tree_find_iova(v->iova_tree, &needle); + const DMAMap *result = vhost_iova_tree_find_iova(v->shared->iova_tree, + &needle); hwaddr size; int r; @@ -1087,7 +1088,7 @@ static void vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr addr) return; } - vhost_iova_tree_remove(v->iova_tree, *result); + vhost_iova_tree_remove(v->shared->iova_tree, *result); } static void vhost_vdpa_svq_unmap_rings(struct vhost_dev *dev, @@ -1115,7 +1116,7 @@ static bool vhost_vdpa_svq_map_ring(struct vhost_vdpa *v, DMAMap *needle, { int r; - r = vhost_iova_tree_map_alloc(v->iova_tree, needle); + r = vhost_iova_tree_map_alloc(v->shared->iova_tree, needle); if (unlikely(r != IOVA_OK)) { error_setg(errp, "Cannot allocate iova (%d)", r); return false; @@ -1127,7 +1128,7 @@ static bool vhost_vdpa_svq_map_ring(struct vhost_vdpa *v, DMAMap *needle, needle->perm == IOMMU_RO); if (unlikely(r != 0)) { error_setg_errno(errp, -r, "Cannot map region to device"); - vhost_iova_tree_remove(v->iova_tree, *needle); + vhost_iova_tree_remove(v->shared->iova_tree, *needle); } return r == 0; @@ -1228,7 +1229,7 @@ static bool vhost_vdpa_svqs_start(struct vhost_dev *dev) goto err; } - vhost_svq_start(svq, dev->vdev, vq, v->iova_tree); + vhost_svq_start(svq, dev->vdev, vq, v->shared->iova_tree); ok = vhost_vdpa_svq_map_rings(dev, svq, &addr, &err); if (unlikely(!ok)) { goto err_map; -- Gitee From eb1f414d9dcc2a493158a8231c3b9de3a36fb3de Mon Sep 17 00:00:00 2001 From: Hao Chen Date: Thu, 21 Dec 2023 18:43:12 +0100 Subject: [PATCH 2/2] vdpa: move iova_range to vhost_vdpa_shared MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Next patches will register the vhost_vdpa memory listener while the VM is migrating at the destination, so we can map the memory to the device before stopping the VM at the source. The main goal is to reduce the downtime. However, the destination QEMU is unaware of which vhost_vdpa device will register its memory_listener. If the source guest has CVQ enabled, it will be the CVQ device. Otherwise, it will be the first one. Move the iova range to VhostVDPAShared so all vhost_vdpa can use it, rather than always in the first or last vhost_vdpa. Backport from QEMU official community: ae25ff41b723 Signed-off-by: Eugenio Pérez Acked-by: Jason Wang Message-Id: <20231221174322.3130442-4-eperezma@redhat.com> Tested-by: Lei Yang Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/virtio/vdpa-dev.c | 5 ++++- hw/virtio/vhost-vdpa.c | 16 ++++++++++------ 2 files changed, 14 insertions(+), 7 deletions(-) diff --git a/hw/virtio/vdpa-dev.c b/hw/virtio/vdpa-dev.c index 7ce8547419..1ffa3f2390 100644 --- a/hw/virtio/vdpa-dev.c +++ b/hw/virtio/vdpa-dev.c @@ -120,7 +120,8 @@ static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp) strerror(-ret)); goto free_vqs; } - v->vdpa.iova_range = iova_range; + v->vdpa.shared = g_new0(VhostVDPAShared, 1); + v->vdpa.shared->iova_range = iova_range; ret = vhost_dev_init(&v->dev, &v->vdpa, VHOST_BACKEND_TYPE_VDPA, 0, NULL); if (ret < 0) { @@ -185,6 +186,7 @@ vhost_cleanup: } free_vqs: g_free(vqs); + g_free(v->vdpa.shared); out: qemu_close(v->vhostfd); v->vhostfd = -1; @@ -212,6 +214,7 @@ static void vhost_vdpa_device_unrealize(DeviceState *dev) if (s->iommufd) { vhost_vdpa_detach_container(s); } + g_free(s->vdpa.shared); qemu_close(s->vhostfd); s->vhostfd = -1; } diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c index ca6aefe7b9..40bbc4927f 100644 --- a/hw/virtio/vhost-vdpa.c +++ b/hw/virtio/vhost-vdpa.c @@ -214,10 +214,10 @@ static void vhost_vdpa_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) RCU_READ_LOCK_GUARD(); /* check if RAM section out of device range */ llend = int128_add(int128_makes64(iotlb->addr_mask), int128_makes64(iova)); - if (int128_gt(llend, int128_make64(v->iova_range.last))) { + if (int128_gt(llend, int128_make64(v->shared->iova_range.last))) { error_report("RAM section out of device range (max=0x%" PRIx64 ", end addr=0x%" PRIx64 ")", - v->iova_range.last, int128_get64(llend)); + v->shared->iova_range.last, int128_get64(llend)); return; } @@ -317,8 +317,10 @@ static void vhost_vdpa_listener_region_add(MemoryListener *listener, int page_size = qemu_target_page_size(); int page_mask = -page_size; - if (vhost_vdpa_listener_skipped_section(section, v->iova_range.first, - v->iova_range.last, page_mask)) { + if (vhost_vdpa_listener_skipped_section(section, + v->shared->iova_range.first, + v->shared->iova_range.last, + page_mask)) { return; } if (memory_region_is_iommu(section->mr)) { @@ -404,8 +406,10 @@ static void vhost_vdpa_listener_region_del(MemoryListener *listener, int page_size = qemu_target_page_size(); int page_mask = -page_size; - if (vhost_vdpa_listener_skipped_section(section, v->iova_range.first, - v->iova_range.last, page_mask)) { + if (vhost_vdpa_listener_skipped_section(section, + v->shared->iova_range.first, + v->shared->iova_range.last, + page_mask)) { return; } if (memory_region_is_iommu(section->mr)) { -- Gitee