drm/nv50-nvc0/vm: support unsnooped system memory
v2 (Emil Velikov <emil.l.velikov@gmail.com>):
- Fixed a regression on certain nv50 IGP due to not passing the correct
target type to nv50_vm_addr()
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Tested-by: Johannes Obermayr <johannesobermayr@gmx.de>
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 63e77fc..3a69a4a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -1759,6 +1759,7 @@
#define NV_MEM_ACCESS_RW (NV_MEM_ACCESS_RO | NV_MEM_ACCESS_WO)
#define NV_MEM_ACCESS_SYS 4
#define NV_MEM_ACCESS_VM 8
+#define NV_MEM_ACCESS_NOSNOOP 16
#define NV_MEM_TARGET_VRAM 0
#define NV_MEM_TARGET_PCI 1
diff --git a/drivers/gpu/drm/nouveau/nv50_vm.c b/drivers/gpu/drm/nouveau/nv50_vm.c
index 6f38cea..44fbac9 100644
--- a/drivers/gpu/drm/nouveau/nv50_vm.c
+++ b/drivers/gpu/drm/nouveau/nv50_vm.c
@@ -57,27 +57,15 @@
}
static inline u64
-nv50_vm_addr(struct nouveau_vma *vma, u64 phys, u32 memtype, u32 target)
+vm_addr(struct nouveau_vma *vma, u64 phys, u32 memtype, u32 target)
{
- struct drm_nouveau_private *dev_priv = vma->vm->dev->dev_private;
-
phys |= 1; /* present */
phys |= (u64)memtype << 40;
-
- /* IGPs don't have real VRAM, re-target to stolen system memory */
- if (target == 0 && dev_priv->vram_sys_base) {
- phys += dev_priv->vram_sys_base;
- target = 3;
- }
-
phys |= target << 4;
-
if (vma->access & NV_MEM_ACCESS_SYS)
phys |= (1 << 6);
-
if (!(vma->access & NV_MEM_ACCESS_WO))
phys |= (1 << 3);
-
return phys;
}
@@ -85,11 +73,19 @@
nv50_vm_map(struct nouveau_vma *vma, struct nouveau_gpuobj *pgt,
struct nouveau_mem *mem, u32 pte, u32 cnt, u64 phys, u64 delta)
{
+ struct drm_nouveau_private *dev_priv = vma->vm->dev->dev_private;
u32 comp = (mem->memtype & 0x180) >> 7;
- u32 block;
+ u32 block, target;
int i;
- phys = nv50_vm_addr(vma, phys, mem->memtype, 0);
+ /* IGPs don't have real VRAM, re-target to stolen system memory */
+ target = 0;
+ if (dev_priv->vram_sys_base) {
+ phys += dev_priv->vram_sys_base;
+ target = 3;
+ }
+
+ phys = vm_addr(vma, phys, mem->memtype, target);
pte <<= 3;
cnt <<= 3;
@@ -125,9 +121,10 @@
nv50_vm_map_sg(struct nouveau_vma *vma, struct nouveau_gpuobj *pgt,
struct nouveau_mem *mem, u32 pte, u32 cnt, dma_addr_t *list)
{
+ u32 target = (vma->access & NV_MEM_ACCESS_NOSNOOP) ? 3 : 2;
pte <<= 3;
while (cnt--) {
- u64 phys = nv50_vm_addr(vma, (u64)*list++, mem->memtype, 2);
+ u64 phys = vm_addr(vma, (u64)*list++, mem->memtype, target);
nv_wo32(pgt, pte + 0, lower_32_bits(phys));
nv_wo32(pgt, pte + 4, upper_32_bits(phys));
pte += 8;
diff --git a/drivers/gpu/drm/nouveau/nvc0_vm.c b/drivers/gpu/drm/nouveau/nvc0_vm.c
index 9e35294..30d2bd5 100644
--- a/drivers/gpu/drm/nouveau/nvc0_vm.c
+++ b/drivers/gpu/drm/nouveau/nvc0_vm.c
@@ -77,9 +77,11 @@
nvc0_vm_map_sg(struct nouveau_vma *vma, struct nouveau_gpuobj *pgt,
struct nouveau_mem *mem, u32 pte, u32 cnt, dma_addr_t *list)
{
+ u32 target = (vma->access & NV_MEM_ACCESS_NOSNOOP) ? 7 : 5;
+
pte <<= 3;
while (cnt--) {
- u64 phys = nvc0_vm_addr(vma, *list++, mem->memtype, 5);
+ u64 phys = nvc0_vm_addr(vma, *list++, mem->memtype, target);
nv_wo32(pgt, pte + 0, lower_32_bits(phys));
nv_wo32(pgt, pte + 4, upper_32_bits(phys));
pte += 8;
