flodavid/suyu
1
0
Fork 0
forked from suyu/suyu
Code Pull requests Activity

Merge pull request #1630 from bunnei/fix-mapbufferex

Browse source 
memory_manager: Do not MapBufferEx over already in use memory.
This commit is contained in:
bunnei 2018-11-07 00:14:36 -08:00 committed by GitHub
commit 81ff9e2473
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 52 additions and 31 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

75
src/video_core/memory_manager.cpp
View file

@ -4,18 +4,21 @@
#include "common/alignment.h"
#include "common/assert.h"
#include "common/logging/log.h"
#include "video_core/memory_manager.h"
namespace Tegra {
GPUVAddr MemoryManager::AllocateSpace(u64 size, u64 align) {
std::optional<GPUVAddr> gpu_addr = FindFreeBlock(size, align);
ASSERT(gpu_addr);
const std::optional<GPUVAddr> gpu_addr{FindFreeBlock(0, size, align, PageStatus::Unmapped)};
for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
VAddr& slot = PageSlot(*gpu_addr + offset);
ASSERT_MSG(gpu_addr, "unable to find available GPU memory");
for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
VAddr& slot{PageSlot(*gpu_addr + offset)};
ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
slot = static_cast<u64>(PageStatus::Allocated);
}
@ -23,10 +26,11 @@ GPUVAddr MemoryManager::AllocateSpace(u64 size, u64 align) {
}
GPUVAddr MemoryManager::AllocateSpace(GPUVAddr gpu_addr, u64 size, u64 align) {
for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
VAddr& slot = PageSlot(gpu_addr + offset);
for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
VAddr& slot{PageSlot(gpu_addr + offset)};
ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
slot = static_cast<u64>(PageStatus::Allocated);
}
@ -34,17 +38,19 @@ GPUVAddr MemoryManager::AllocateSpace(GPUVAddr gpu_addr, u64 size, u64 align) {
}
GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, u64 size) {
std::optional<GPUVAddr> gpu_addr = FindFreeBlock(size, PAGE_SIZE);
ASSERT(gpu_addr);
const std::optional<GPUVAddr> gpu_addr{FindFreeBlock(0, size, PAGE_SIZE, PageStatus::Unmapped)};
for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
VAddr& slot = PageSlot(*gpu_addr + offset);
ASSERT_MSG(gpu_addr, "unable to find available GPU memory");
for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
VAddr& slot{PageSlot(*gpu_addr + offset)};
ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
slot = cpu_addr + offset;
}
MappedRegion region{cpu_addr, *gpu_addr, size};
const MappedRegion region{cpu_addr, *gpu_addr, size};
mapped_regions.push_back(region);
return *gpu_addr;
@ -53,14 +59,31 @@ GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, u64 size) {
GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, GPUVAddr gpu_addr, u64 size) {
ASSERT((gpu_addr & PAGE_MASK) == 0);
for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
VAddr& slot = PageSlot(gpu_addr + offset);
if (PageSlot(gpu_addr) != static_cast<u64>(PageStatus::Allocated)) {
// Page has been already mapped. In this case, we must find a new area of memory to use that
// is different than the specified one. Super Mario Odyssey hits this scenario when changing
// areas, but we do not want to overwrite the old pages.
// TODO(bunnei): We need to write a hardware test to confirm this behavior.
LOG_ERROR(HW_GPU, "attempting to map addr 0x{:016X}, which is not available!", gpu_addr);
const std::optional<GPUVAddr> new_gpu_addr{
FindFreeBlock(gpu_addr, size, PAGE_SIZE, PageStatus::Allocated)};
ASSERT_MSG(new_gpu_addr, "unable to find available GPU memory");
gpu_addr = *new_gpu_addr;
}
for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
VAddr& slot{PageSlot(gpu_addr + offset)};
ASSERT(slot == static_cast<u64>(PageStatus::Allocated));
slot = cpu_addr + offset;
}
MappedRegion region{cpu_addr, gpu_addr, size};
const MappedRegion region{cpu_addr, gpu_addr, size};
mapped_regions.push_back(region);
return gpu_addr;
@ -69,11 +92,12 @@ GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, GPUVAddr gpu_addr, u64 size)
GPUVAddr MemoryManager::UnmapBuffer(GPUVAddr gpu_addr, u64 size) {
ASSERT((gpu_addr & PAGE_MASK) == 0);
for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
VAddr& slot = PageSlot(gpu_addr + offset);
for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
VAddr& slot{PageSlot(gpu_addr + offset)};
ASSERT(slot != static_cast<u64>(PageStatus::Allocated) &&
slot != static_cast<u64>(PageStatus::Unmapped));
slot = static_cast<u64>(PageStatus::Unmapped);
}
@ -97,13 +121,14 @@ GPUVAddr MemoryManager::GetRegionEnd(GPUVAddr region_start) const {
return {};
}
std::optional<GPUVAddr> MemoryManager::FindFreeBlock(u64 size, u64 align) {
GPUVAddr gpu_addr = 0;
u64 free_space = 0;
std::optional<GPUVAddr> MemoryManager::FindFreeBlock(GPUVAddr region_start, u64 size, u64 align,
PageStatus status) {
GPUVAddr gpu_addr{region_start};
u64 free_space{};
align = (align + PAGE_MASK) & ~PAGE_MASK;
while (gpu_addr + free_space < MAX_ADDRESS) {
if (!IsPageMapped(gpu_addr + free_space)) {
if (PageSlot(gpu_addr + free_space) == static_cast<u64>(status)) {
free_space += PAGE_SIZE;
if (free_space >= size) {
return gpu_addr;
@ -119,7 +144,7 @@ std::optional<GPUVAddr> MemoryManager::FindFreeBlock(u64 size, u64 align) {
}
std::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr gpu_addr) {
VAddr base_addr = PageSlot(gpu_addr);
const VAddr base_addr{PageSlot(gpu_addr)};
if (base_addr == static_cast<u64>(PageStatus::Allocated) ||
base_addr == static_cast<u64>(PageStatus::Unmapped)) {
@ -133,19 +158,15 @@ std::vector<GPUVAddr> MemoryManager::CpuToGpuAddress(VAddr cpu_addr) const {
std::vector<GPUVAddr> results;
for (const auto& region : mapped_regions) {
if (cpu_addr >= region.cpu_addr && cpu_addr < (region.cpu_addr + region.size)) {
u64 offset = cpu_addr - region.cpu_addr;
const u64 offset{cpu_addr - region.cpu_addr};
results.push_back(region.gpu_addr + offset);
}
}
return results;
}
bool MemoryManager::IsPageMapped(GPUVAddr gpu_addr) {
return PageSlot(gpu_addr) != static_cast<u64>(PageStatus::Unmapped);
}
VAddr& MemoryManager::PageSlot(GPUVAddr gpu_addr) {
auto& block = page_table[(gpu_addr >> (PAGE_BITS + PAGE_TABLE_BITS)) & PAGE_TABLE_MASK];
auto& block{page_table[(gpu_addr >> (PAGE_BITS + PAGE_TABLE_BITS)) & PAGE_TABLE_MASK]};
if (!block) {
block = std::make_unique<PageBlock>();
block->fill(static_cast<VAddr>(PageStatus::Unmapped));

8
src/video_core/memory_manager.h
View file

@ -34,15 +34,15 @@ public:
static constexpr u64 PAGE_MASK = PAGE_SIZE - 1;
private:
std::optional<GPUVAddr> FindFreeBlock(u64 size, u64 align = 1);
bool IsPageMapped(GPUVAddr gpu_addr);
VAddr& PageSlot(GPUVAddr gpu_addr);
enum class PageStatus : u64 {
Unmapped = 0xFFFFFFFFFFFFFFFFULL,
Allocated = 0xFFFFFFFFFFFFFFFEULL,
};
std::optional<GPUVAddr> FindFreeBlock(GPUVAddr region_start, u64 size, u64 align,
PageStatus status);
VAddr& PageSlot(GPUVAddr gpu_addr);
static constexpr u64 MAX_ADDRESS{0x10000000000ULL};
static constexpr u64 PAGE_TABLE_BITS{10};
static constexpr u64 PAGE_TABLE_SIZE{1 << PAGE_TABLE_BITS};