[dpdk-dev,v3,2/7] vhost: get guest/host physical address mappings
Commit Message
So that we can convert a guest physical address to host physical
address, which will be used in later Tx zero copy implementation.
MAP_POPULATE is set while mmaping guest memory regions, to make
sure the page tables are setup and then rte_mem_virt2phy() could
yield proper physical address.
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
---
v2: - use MAP_POPULATE option to make sure the page table will
be already setup while getting the phys address
- do a simple merge if the last 2 pages are continuous
- dump guest pages only in debug mode
---
lib/librte_vhost/vhost.h | 30 +++++++++++++
lib/librte_vhost/vhost_user.c | 100 +++++++++++++++++++++++++++++++++++++++++-
2 files changed, 128 insertions(+), 2 deletions(-)
Comments
在 2016/10/9 15:27, Yuanhan Liu 写道:
> + dev->nr_guest_pages = 0;
> + if (!dev->guest_pages) {
> + dev->max_guest_pages = 8;
> + dev->guest_pages = malloc(dev->max_guest_pages *
> + sizeof(struct guest_page));
> + }
> +
when to free guest_pages ?
在 2016/10/9 15:27, Yuanhan Liu 写道:
> +static void
> +add_guest_pages(struct virtio_net *dev, struct virtio_memory_region *reg,
> + uint64_t page_size)
> +{
> + uint64_t reg_size = reg->size;
> + uint64_t host_user_addr = reg->host_user_addr;
> + uint64_t guest_phys_addr = reg->guest_phys_addr;
> + uint64_t host_phys_addr;
> + uint64_t size;
> +
> + host_phys_addr = rte_mem_virt2phy((void *)(uintptr_t)host_user_addr);
> + size = page_size - (guest_phys_addr & (page_size - 1));
> + size = RTE_MIN(size, reg_size);
Have you use 1G hugepage to create a VM with 25G memory?
When I try, vhost crashed, use below code fixed:
- size = page_size - (guest_phys_addr & (page_size - 1));
- size = RTE_MIN(size, reg_size);
+ size = reg_size % page_size;
@@ -114,6 +114,12 @@ struct vhost_virtqueue {
#define VIRTIO_F_VERSION_1 32
#endif
+struct guest_page {
+ uint64_t guest_phys_addr;
+ uint64_t host_phys_addr;
+ uint64_t size;
+};
+
/**
* Device structure contains all configuration information relating
* to the device.
@@ -137,6 +143,10 @@ struct virtio_net {
uint64_t log_addr;
struct ether_addr mac;
+ uint32_t nr_guest_pages;
+ uint32_t max_guest_pages;
+ struct guest_page *guest_pages;
+
} __rte_cache_aligned;
/**
@@ -217,6 +227,26 @@ gpa_to_vva(struct virtio_net *dev, uint64_t gpa)
return 0;
}
+/* Convert guest physical address to host physical address */
+static inline phys_addr_t __attribute__((always_inline))
+gpa_to_hpa(struct virtio_net *dev, uint64_t gpa, uint64_t size)
+{
+ uint32_t i;
+ struct guest_page *page;
+
+ for (i = 0; i < dev->nr_guest_pages; i++) {
+ page = &dev->guest_pages[i];
+
+ if (gpa >= page->guest_phys_addr &&
+ gpa + size < page->guest_phys_addr + page->size) {
+ return gpa - page->guest_phys_addr +
+ page->host_phys_addr;
+ }
+ }
+
+ return 0;
+}
+
struct virtio_net_device_ops const *notify_ops;
struct virtio_net *get_device(int vid);
@@ -372,6 +372,91 @@ vhost_user_set_vring_base(struct virtio_net *dev,
return 0;
}
+static void
+add_one_guest_page(struct virtio_net *dev, uint64_t guest_phys_addr,
+ uint64_t host_phys_addr, uint64_t size)
+{
+ struct guest_page *page, *last_page;
+
+ if (dev->nr_guest_pages == dev->max_guest_pages) {
+ dev->max_guest_pages *= 2;
+ dev->guest_pages = realloc(dev->guest_pages,
+ dev->max_guest_pages * sizeof(*page));
+ }
+
+ if (dev->nr_guest_pages > 0) {
+ last_page = &dev->guest_pages[dev->nr_guest_pages - 1];
+ /* merge if the two pages are continuous */
+ if (host_phys_addr == last_page->host_phys_addr +
+ last_page->size) {
+ last_page->size += size;
+ return;
+ }
+ }
+
+ page = &dev->guest_pages[dev->nr_guest_pages++];
+ page->guest_phys_addr = guest_phys_addr;
+ page->host_phys_addr = host_phys_addr;
+ page->size = size;
+}
+
+static void
+add_guest_pages(struct virtio_net *dev, struct virtio_memory_region *reg,
+ uint64_t page_size)
+{
+ uint64_t reg_size = reg->size;
+ uint64_t host_user_addr = reg->host_user_addr;
+ uint64_t guest_phys_addr = reg->guest_phys_addr;
+ uint64_t host_phys_addr;
+ uint64_t size;
+
+ host_phys_addr = rte_mem_virt2phy((void *)(uintptr_t)host_user_addr);
+ size = page_size - (guest_phys_addr & (page_size - 1));
+ size = RTE_MIN(size, reg_size);
+
+ add_one_guest_page(dev, guest_phys_addr, host_phys_addr, size);
+ host_user_addr += size;
+ guest_phys_addr += size;
+ reg_size -= size;
+
+ while (reg_size > 0) {
+ host_phys_addr = rte_mem_virt2phy((void *)(uintptr_t)
+ host_user_addr);
+ add_one_guest_page(dev, guest_phys_addr, host_phys_addr,
+ page_size);
+
+ host_user_addr += page_size;
+ guest_phys_addr += page_size;
+ reg_size -= page_size;
+ }
+}
+
+#ifdef RTE_LIBRTE_VHOST_DEBUG
+/* TODO: enable it only in debug mode? */
+static void
+dump_guest_pages(struct virtio_net *dev)
+{
+ uint32_t i;
+ struct guest_page *page;
+
+ for (i = 0; i < dev->nr_guest_pages; i++) {
+ page = &dev->guest_pages[i];
+
+ RTE_LOG(INFO, VHOST_CONFIG,
+ "guest physical page region %u\n"
+ "\t guest_phys_addr: %" PRIx64 "\n"
+ "\t host_phys_addr : %" PRIx64 "\n"
+ "\t size : %" PRIx64 "\n",
+ i,
+ page->guest_phys_addr,
+ page->host_phys_addr,
+ page->size);
+ }
+}
+#else
+#define dump_guest_pages(dev)
+#endif
+
static int
vhost_user_set_mem_table(struct virtio_net *dev, struct VhostUserMsg *pmsg)
{
@@ -396,6 +481,13 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct VhostUserMsg *pmsg)
dev->mem = NULL;
}
+ dev->nr_guest_pages = 0;
+ if (!dev->guest_pages) {
+ dev->max_guest_pages = 8;
+ dev->guest_pages = malloc(dev->max_guest_pages *
+ sizeof(struct guest_page));
+ }
+
dev->mem = rte_zmalloc("vhost-mem-table", sizeof(struct virtio_memory) +
sizeof(struct virtio_memory_region) * memory.nregions, 0);
if (dev->mem == NULL) {
@@ -434,8 +526,8 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct VhostUserMsg *pmsg)
}
mmap_size = RTE_ALIGN_CEIL(mmap_size, alignment);
- mmap_addr = mmap(NULL, mmap_size,
- PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
+ mmap_addr = mmap(NULL, mmap_size, PROT_READ | PROT_WRITE,
+ MAP_SHARED | MAP_POPULATE, fd, 0);
if (mmap_addr == MAP_FAILED) {
RTE_LOG(ERR, VHOST_CONFIG,
@@ -448,6 +540,8 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct VhostUserMsg *pmsg)
reg->host_user_addr = (uint64_t)(uintptr_t)mmap_addr +
mmap_offset;
+ add_guest_pages(dev, reg, alignment);
+
RTE_LOG(INFO, VHOST_CONFIG,
"guest memory region %u, size: 0x%" PRIx64 "\n"
"\t guest physical addr: 0x%" PRIx64 "\n"
@@ -467,6 +561,8 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct VhostUserMsg *pmsg)
mmap_offset);
}
+ dump_guest_pages(dev);
+
return 0;
err_mmap: