NVIDIA/open-gpu-kernel-modules
1
0
Fork 0
mirror of https://github.com/NVIDIA/open-gpu-kernel-modules.git synced 2026-02-26 09:53:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

575.51.02

Browse Source
This commit is contained in:
Bernhard Stoeckner
2025-04-17 19:35:38 +02:00
parent e8113f665d
commit 4159579888
1142 changed files with 309085 additions and 272273 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
kernel-open/nvidia-uvm/uvm_mmu.c
View File

@@ -134,7 +134,7 @@ static NV_STATUS phys_mem_allocate_sysmem(uvm_page_tree_t *tree, NvLength size,
// Check for fake GPUs from the unit test
if (tree->gpu->parent->pci_dev)
status = uvm_parent_gpu_map_cpu_pages(tree->gpu->parent, out->handle.page, UVM_PAGE_ALIGN_UP(size), &dma_addr);
status = uvm_gpu_map_cpu_pages(tree->gpu, out->handle.page, UVM_PAGE_ALIGN_UP(size), &dma_addr);
else
dma_addr = page_to_phys(out->handle.page);
@@ -703,7 +703,7 @@ static NV_STATUS write_gpu_state_cpu(uvm_page_tree_t *tree,
// See the comments in write_gpu_state_gpu()
tree->gpu->parent->host_hal->tlb_invalidate_all(&push,
uvm_page_tree_pdb(tree)->addr,
uvm_page_tree_pdb_address(tree),
invalidate_depth,
UVM_MEMBAR_NONE);
page_tree_end(tree, &push);
@@ -787,7 +787,7 @@ static NV_STATUS write_gpu_state_gpu(uvm_page_tree_t *tree,
// Upgrades don't have to flush out accesses, so no membar is needed on the
// TLB invalidate.
tree->gpu->parent->host_hal->tlb_invalidate_all(&push,
uvm_page_tree_pdb(tree)->addr,
uvm_page_tree_pdb_address(tree),
invalidate_depth,
UVM_MEMBAR_NONE);
@@ -876,7 +876,10 @@ static NV_STATUS page_tree_ats_init(uvm_page_tree_t *tree)
if (!page_tree_ats_init_required(tree))
return NV_OK;
page_size = mmu_biggest_page_size(tree, UVM_APERTURE_VID);
if (tree->gpu->mem_info.size)
page_size = mmu_biggest_page_size(tree, UVM_APERTURE_VID);
else
page_size = mmu_biggest_page_size(tree, UVM_APERTURE_SYS);
uvm_cpu_get_unaddressable_range(&max_va_lower, &min_va_upper);
@@ -899,19 +902,17 @@ static NV_STATUS page_tree_ats_init(uvm_page_tree_t *tree)
UVM_ASSERT(tree->no_ats_ranges[0].entry_count == 1);
if (uvm_platform_uses_canonical_form_address()) {
// Upper half
status = uvm_page_tree_get_ptes(tree,
page_size,
min_va_upper - page_size,
page_size,
UVM_PMM_ALLOC_FLAGS_EVICT,
&tree->no_ats_ranges[1]);
if (status != NV_OK)
return status;
// Upper half
status = uvm_page_tree_get_ptes(tree,
page_size,
min_va_upper - page_size,
page_size,
UVM_PMM_ALLOC_FLAGS_EVICT,
&tree->no_ats_ranges[1]);
if (status != NV_OK)
return status;
UVM_ASSERT(tree->no_ats_ranges[1].entry_count == 1);
}
UVM_ASSERT(tree->no_ats_ranges[1].entry_count == 1);
return NV_OK;
}
@@ -1056,6 +1057,9 @@ error:
// sysmem | - || <disallowed>
// default | - || <disallowed>
//
// - If a GPU has no vidmem (broken fb or integrated GPU) then the page table
// must be in sysmem.
//
static void page_tree_set_location(uvm_page_tree_t *tree, uvm_aperture_t location)
{
UVM_ASSERT(tree->gpu != NULL);
@@ -1071,6 +1075,9 @@ static void page_tree_set_location(uvm_page_tree_t *tree, uvm_aperture_t locatio
if (uvm_parent_gpu_is_virt_mode_sriov_heavy(tree->gpu->parent) || g_uvm_global.conf_computing_enabled)
UVM_ASSERT(location == UVM_APERTURE_VID);
if (!tree->gpu->mem_info.size)
UVM_ASSERT(location == UVM_APERTURE_SYS);
}
if (location == UVM_APERTURE_DEFAULT) {
@@ -1177,7 +1184,7 @@ void uvm_page_tree_deinit(uvm_page_tree_t *tree)
// error. We can't perform the unmap, so just leave things in place for
// debug.
if (status == NV_OK) {
tree->gpu->parent->host_hal->tlb_invalidate_all(&push, uvm_page_tree_pdb(tree)->addr, 0, UVM_MEMBAR_NONE);
tree->gpu->parent->host_hal->tlb_invalidate_all(&push, uvm_page_tree_pdb_address(tree), 0, UVM_MEMBAR_NONE);
page_tree_end(tree, &push);
page_tree_tracker_overwrite_with_push(tree, &push);
}
@@ -1303,7 +1310,7 @@ void uvm_page_tree_put_ptes_async(uvm_page_tree_t *tree, uvm_page_table_range_t
uvm_hal_wfi_membar(&push, membar_after_pde_clears);
tree->gpu->parent->host_hal->tlb_invalidate_all(&push,
uvm_page_tree_pdb(tree)->addr,
uvm_page_tree_pdb_address(tree),
invalidate_depth,
membar_after_invalidate);
@@ -2964,7 +2971,7 @@ NV_STATUS uvm_test_invalidate_tlb(UVM_TEST_INVALIDATE_TLB_PARAMS *params, struct
"Pushing test invalidate, GPU %s",
uvm_gpu_name(gpu));
if (status == NV_OK)
gpu->parent->host_hal->tlb_invalidate_test(&push, uvm_page_tree_pdb(&gpu_va_space->page_tables)->addr, params);
gpu->parent->host_hal->tlb_invalidate_test(&push, uvm_page_tree_pdb_address(&gpu_va_space->page_tables), params);
unlock_exit:
// Wait for the invalidation to be performed