From 8604ffcbf04f8f4f3f55a9e46e5ff948b2ed4290 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 16 Aug 2018 12:01:03 +0200 Subject: drm/amdgpu: fix VM clearing for the root PD MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We need to figure out the address after validating the BO, not before. Signed-off-by: Christian König Reviewed-by: Felix Kuehling Reviewed-by: Junwei Zhang Reviewed-by: Huang Rui Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index ece0ac703e27..e40ca8676418 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -369,7 +369,6 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, uint64_t addr; int r; - addr = amdgpu_bo_gpu_offset(bo); entries = amdgpu_bo_size(bo) / 8; if (pte_support_ats) { @@ -401,6 +400,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, if (r) goto error; + addr = amdgpu_bo_gpu_offset(bo); if (ats_entries) { uint64_t ats_value; -- cgit v1.2.2 From fca5d959972c18839f0306f19df2f121623447dd Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Tue, 21 Aug 2018 17:14:32 -0400 Subject: drm/amdgpu: Adjust the VM size based on system memory size v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Set the VM size based on system memory size between the ASIC-specific limits given by min_vm_size and max_bits. GFXv9 GPUs will keep their default VM size of 256TB (48 bit). Only older GPUs will adjust VM size depending on system memory size. This makes more VM space available for ROCm applications on GFXv8 GPUs that want to map all available VRAM and system memory in their SVM address space. v2: * Clarify comment * Round up memory size before >> 30 * Round up automatic vm_size to power of two Signed-off-by: Felix Kuehling Acked-by: Junwei Zhang Reviewed-by: Huang Rui Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 32 ++++++++++++++++++++++++++++---- 1 file changed, 28 insertions(+), 4 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index e40ca8676418..d174d50e3bd3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -2483,28 +2483,52 @@ static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size) * amdgpu_vm_adjust_size - adjust vm size, block size and fragment size * * @adev: amdgpu_device pointer - * @vm_size: the default vm size if it's set auto + * @min_vm_size: the minimum vm size in GB if it's set auto * @fragment_size_default: Default PTE fragment size * @max_level: max VMPT level * @max_bits: max address space size in bits * */ -void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size, +void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t min_vm_size, uint32_t fragment_size_default, unsigned max_level, unsigned max_bits) { + unsigned int max_size = 1 << (max_bits - 30); + unsigned int vm_size; uint64_t tmp; /* adjust vm size first */ if (amdgpu_vm_size != -1) { - unsigned max_size = 1 << (max_bits - 30); - vm_size = amdgpu_vm_size; if (vm_size > max_size) { dev_warn(adev->dev, "VM size (%d) too large, max is %u GB\n", amdgpu_vm_size, max_size); vm_size = max_size; } + } else { + struct sysinfo si; + unsigned int phys_ram_gb; + + /* Optimal VM size depends on the amount of physical + * RAM available. Underlying requirements and + * assumptions: + * + * - Need to map system memory and VRAM from all GPUs + * - VRAM from other GPUs not known here + * - Assume VRAM <= system memory + * - On GFX8 and older, VM space can be segmented for + * different MTYPEs + * - Need to allow room for fragmentation, guard pages etc. + * + * This adds up to a rough guess of system memory x3. + * Round up to power of two to maximize the available + * VM size with the given page table size. + */ + si_meminfo(&si); + phys_ram_gb = ((uint64_t)si.totalram * si.mem_unit + + (1 << 30) - 1) >> 30; + vm_size = roundup_pow_of_two( + min(max(phys_ram_gb * 3, min_vm_size), max_size)); } adev->vm_manager.max_pfn = (uint64_t)vm_size << 18; -- cgit v1.2.2 From 6ddd9769db4fc11a98bd7e58be1764e47fdb8384 Mon Sep 17 00:00:00 2001 From: Emily Deng Date: Tue, 28 Aug 2018 20:52:40 +0800 Subject: drm/amdgpu: Need to set moved to true when evict bo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the VMC page fault when the running sequence is as below: 1.amdgpu_gem_create_ioctl 2.ttm_bo_swapout->amdgpu_vm_bo_invalidate, as not called amdgpu_vm_bo_base_init, so won't called list_add_tail(&base->bo_list, &bo->va). Even the bo was evicted, it won't set the bo_base->moved. 3.drm_gem_open_ioctl->amdgpu_vm_bo_base_init, here only called list_move_tail(&base->vm_status, &vm->evicted), but not set the bo_base->moved. 4.amdgpu_vm_bo_map->amdgpu_vm_bo_insert_map, as the bo_base->moved is not set true, the function amdgpu_vm_bo_insert_map will call list_move(&bo_va->base.vm_status, &vm->moved) 5.amdgpu_cs_ioctl won't validate the swapout bo, as it is only in the moved list, not in the evict list. So VMC page fault occurs. Signed-off-by: Emily Deng Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index d174d50e3bd3..b17771dd5ce7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -172,6 +172,7 @@ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, * is validated on next vm use to avoid fault. * */ list_move_tail(&base->vm_status, &vm->evicted); + base->moved = true; } /** -- cgit v1.2.2