Commit 4080fc10 authored by Stefano Garzarella's avatar Stefano Garzarella Committed by Michael S. Tsirkin

vdpa_sim: use iova module to allocate IOVA addresses

The identical mapping used until now created issues when mapping
different virtual pages with the same physical address.
To solve this issue, we can use the iova module, to handle the IOVA
allocation.
For simplicity we use an IOVA allocator with byte granularity.

We add two new functions, vdpasim_map_range() and vdpasim_unmap_range(),
to handle the IOVA allocation and the registration into the IOMMU/IOTLB.
These functions are used by dma_map_ops callbacks.
Acked-by: default avatarJason Wang <jasowang@redhat.com>
Signed-off-by: default avatarStefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20210315163450.254396-2-sgarzare@redhat.comSigned-off-by: default avatarMichael S. Tsirkin <mst@redhat.com>
parent e8ef6124
...@@ -14,6 +14,7 @@ config VDPA_SIM ...@@ -14,6 +14,7 @@ config VDPA_SIM
depends on RUNTIME_TESTING_MENU && HAS_DMA depends on RUNTIME_TESTING_MENU && HAS_DMA
select DMA_OPS select DMA_OPS
select VHOST_RING select VHOST_RING
select IOMMU_IOVA
help help
Enable this module to support vDPA device simulators. These devices Enable this module to support vDPA device simulators. These devices
are used for testing, prototyping and development of vDPA. are used for testing, prototyping and development of vDPA.
......
...@@ -17,6 +17,7 @@ ...@@ -17,6 +17,7 @@
#include <linux/vringh.h> #include <linux/vringh.h>
#include <linux/vdpa.h> #include <linux/vdpa.h>
#include <linux/vhost_iotlb.h> #include <linux/vhost_iotlb.h>
#include <linux/iova.h>
#include "vdpa_sim.h" #include "vdpa_sim.h"
...@@ -128,30 +129,57 @@ static int dir_to_perm(enum dma_data_direction dir) ...@@ -128,30 +129,57 @@ static int dir_to_perm(enum dma_data_direction dir)
return perm; return perm;
} }
static dma_addr_t vdpasim_map_range(struct vdpasim *vdpasim, phys_addr_t paddr,
size_t size, unsigned int perm)
{
struct iova *iova;
dma_addr_t dma_addr;
int ret;
/* We set the limit_pfn to the maximum (ULONG_MAX - 1) */
iova = alloc_iova(&vdpasim->iova, size, ULONG_MAX - 1, true);
if (!iova)
return DMA_MAPPING_ERROR;
dma_addr = iova_dma_addr(&vdpasim->iova, iova);
spin_lock(&vdpasim->iommu_lock);
ret = vhost_iotlb_add_range(vdpasim->iommu, (u64)dma_addr,
(u64)dma_addr + size - 1, (u64)paddr, perm);
spin_unlock(&vdpasim->iommu_lock);
if (ret) {
__free_iova(&vdpasim->iova, iova);
return DMA_MAPPING_ERROR;
}
return dma_addr;
}
static void vdpasim_unmap_range(struct vdpasim *vdpasim, dma_addr_t dma_addr,
size_t size)
{
spin_lock(&vdpasim->iommu_lock);
vhost_iotlb_del_range(vdpasim->iommu, (u64)dma_addr,
(u64)dma_addr + size - 1);
spin_unlock(&vdpasim->iommu_lock);
free_iova(&vdpasim->iova, iova_pfn(&vdpasim->iova, dma_addr));
}
static dma_addr_t vdpasim_map_page(struct device *dev, struct page *page, static dma_addr_t vdpasim_map_page(struct device *dev, struct page *page,
unsigned long offset, size_t size, unsigned long offset, size_t size,
enum dma_data_direction dir, enum dma_data_direction dir,
unsigned long attrs) unsigned long attrs)
{ {
struct vdpasim *vdpasim = dev_to_sim(dev); struct vdpasim *vdpasim = dev_to_sim(dev);
struct vhost_iotlb *iommu = vdpasim->iommu; phys_addr_t paddr = page_to_phys(page) + offset;
u64 pa = (page_to_pfn(page) << PAGE_SHIFT) + offset; int perm = dir_to_perm(dir);
int ret, perm = dir_to_perm(dir);
if (perm < 0) if (perm < 0)
return DMA_MAPPING_ERROR; return DMA_MAPPING_ERROR;
/* For simplicity, use identical mapping to avoid e.g iova return vdpasim_map_range(vdpasim, paddr, size, perm);
* allocator.
*/
spin_lock(&vdpasim->iommu_lock);
ret = vhost_iotlb_add_range(iommu, pa, pa + size - 1,
pa, dir_to_perm(dir));
spin_unlock(&vdpasim->iommu_lock);
if (ret)
return DMA_MAPPING_ERROR;
return (dma_addr_t)(pa);
} }
static void vdpasim_unmap_page(struct device *dev, dma_addr_t dma_addr, static void vdpasim_unmap_page(struct device *dev, dma_addr_t dma_addr,
...@@ -159,12 +187,8 @@ static void vdpasim_unmap_page(struct device *dev, dma_addr_t dma_addr, ...@@ -159,12 +187,8 @@ static void vdpasim_unmap_page(struct device *dev, dma_addr_t dma_addr,
unsigned long attrs) unsigned long attrs)
{ {
struct vdpasim *vdpasim = dev_to_sim(dev); struct vdpasim *vdpasim = dev_to_sim(dev);
struct vhost_iotlb *iommu = vdpasim->iommu;
spin_lock(&vdpasim->iommu_lock); vdpasim_unmap_range(vdpasim, dma_addr, size);
vhost_iotlb_del_range(iommu, (u64)dma_addr,
(u64)dma_addr + size - 1);
spin_unlock(&vdpasim->iommu_lock);
} }
static void *vdpasim_alloc_coherent(struct device *dev, size_t size, static void *vdpasim_alloc_coherent(struct device *dev, size_t size,
...@@ -172,27 +196,22 @@ static void *vdpasim_alloc_coherent(struct device *dev, size_t size, ...@@ -172,27 +196,22 @@ static void *vdpasim_alloc_coherent(struct device *dev, size_t size,
unsigned long attrs) unsigned long attrs)
{ {
struct vdpasim *vdpasim = dev_to_sim(dev); struct vdpasim *vdpasim = dev_to_sim(dev);
struct vhost_iotlb *iommu = vdpasim->iommu; phys_addr_t paddr;
void *addr = kmalloc(size, flag); void *addr;
int ret;
spin_lock(&vdpasim->iommu_lock); addr = kmalloc(size, flag);
if (!addr) { if (!addr) {
*dma_addr = DMA_MAPPING_ERROR; *dma_addr = DMA_MAPPING_ERROR;
} else { return NULL;
u64 pa = virt_to_phys(addr); }
ret = vhost_iotlb_add_range(iommu, (u64)pa, paddr = virt_to_phys(addr);
(u64)pa + size - 1,
pa, VHOST_MAP_RW); *dma_addr = vdpasim_map_range(vdpasim, paddr, size, VHOST_MAP_RW);
if (ret) { if (*dma_addr == DMA_MAPPING_ERROR) {
*dma_addr = DMA_MAPPING_ERROR;
kfree(addr); kfree(addr);
addr = NULL; return NULL;
} else
*dma_addr = (dma_addr_t)pa;
} }
spin_unlock(&vdpasim->iommu_lock);
return addr; return addr;
} }
...@@ -202,14 +221,10 @@ static void vdpasim_free_coherent(struct device *dev, size_t size, ...@@ -202,14 +221,10 @@ static void vdpasim_free_coherent(struct device *dev, size_t size,
unsigned long attrs) unsigned long attrs)
{ {
struct vdpasim *vdpasim = dev_to_sim(dev); struct vdpasim *vdpasim = dev_to_sim(dev);
struct vhost_iotlb *iommu = vdpasim->iommu;
spin_lock(&vdpasim->iommu_lock); vdpasim_unmap_range(vdpasim, dma_addr, size);
vhost_iotlb_del_range(iommu, (u64)dma_addr,
(u64)dma_addr + size - 1);
spin_unlock(&vdpasim->iommu_lock);
kfree(phys_to_virt((uintptr_t)dma_addr)); kfree(vaddr);
} }
static const struct dma_map_ops vdpasim_dma_ops = { static const struct dma_map_ops vdpasim_dma_ops = {
...@@ -271,6 +286,13 @@ struct vdpasim *vdpasim_create(struct vdpasim_dev_attr *dev_attr) ...@@ -271,6 +286,13 @@ struct vdpasim *vdpasim_create(struct vdpasim_dev_attr *dev_attr)
for (i = 0; i < dev_attr->nvqs; i++) for (i = 0; i < dev_attr->nvqs; i++)
vringh_set_iotlb(&vdpasim->vqs[i].vring, vdpasim->iommu); vringh_set_iotlb(&vdpasim->vqs[i].vring, vdpasim->iommu);
ret = iova_cache_get();
if (ret)
goto err_iommu;
/* For simplicity we use an IOVA allocator with byte granularity */
init_iova_domain(&vdpasim->iova, 1, 0);
vdpasim->vdpa.dma_dev = dev; vdpasim->vdpa.dma_dev = dev;
return vdpasim; return vdpasim;
...@@ -541,6 +563,8 @@ static void vdpasim_free(struct vdpa_device *vdpa) ...@@ -541,6 +563,8 @@ static void vdpasim_free(struct vdpa_device *vdpa)
struct vdpasim *vdpasim = vdpa_to_sim(vdpa); struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
cancel_work_sync(&vdpasim->work); cancel_work_sync(&vdpasim->work);
put_iova_domain(&vdpasim->iova);
iova_cache_put();
kvfree(vdpasim->buffer); kvfree(vdpasim->buffer);
if (vdpasim->iommu) if (vdpasim->iommu)
vhost_iotlb_free(vdpasim->iommu); vhost_iotlb_free(vdpasim->iommu);
......
...@@ -6,6 +6,7 @@ ...@@ -6,6 +6,7 @@
#ifndef _VDPA_SIM_H #ifndef _VDPA_SIM_H
#define _VDPA_SIM_H #define _VDPA_SIM_H
#include <linux/iova.h>
#include <linux/vringh.h> #include <linux/vringh.h>
#include <linux/vdpa.h> #include <linux/vdpa.h>
#include <linux/virtio_byteorder.h> #include <linux/virtio_byteorder.h>
...@@ -57,6 +58,7 @@ struct vdpasim { ...@@ -57,6 +58,7 @@ struct vdpasim {
/* virtio config according to device type */ /* virtio config according to device type */
void *config; void *config;
struct vhost_iotlb *iommu; struct vhost_iotlb *iommu;
struct iova_domain iova;
void *buffer; void *buffer;
u32 status; u32 status;
u32 generation; u32 generation;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment