nick/WickedEngine
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix out of bounds crash in SRV and UAV subresource vectors (#1250)

Browse Source 
* Fix out of bounds crash in SRV and UAV subresource vectors

* Remove redundant type casting
This commit is contained in:
Stanislav Denisov
2025-10-22 13:15:21 +02:00
committed by GitHub
parent 301c3b9517
commit 0bdd7a2a78
2 changed files with 38 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files
WickedEngine/wiGraphicsDevice_DX12.cpp
+10 -6
View File
@@ -1009,7 +1009,7 @@ namespace dx12_internal
constexpr TextureDesc _ConvertTextureDesc_Inv(const D3D12_RESOURCE_DESC& desc)
{
TextureDesc retVal;
switch (desc.Dimension)
{
case D3D12_RESOURCE_DIMENSION_TEXTURE1D:
@@ -4311,7 +4311,7 @@ std::mutex queue_locker;
{
internal_state->desc.Flags |= D3D12_RAYTRACING_ACCELERATION_STRUCTURE_BUILD_FLAG_MINIMIZE_MEMORY;
}
switch (desc->type)
{
@@ -4345,7 +4345,7 @@ std::mutex queue_locker;
}
else if (x.type == RaytracingAccelerationStructureDesc::BottomLevel::Geometry::Type::PROCEDURAL_AABBS)
{
geometry.Type = D3D12_RAYTRACING_GEOMETRY_TYPE_PROCEDURAL_PRIMITIVE_AABBS;
geometry.Type = D3D12_RAYTRACING_GEOMETRY_TYPE_PROCEDURAL_PRIMITIVE_AABBS;
geometry.AABBs.AABBs.StartAddress = to_internal(&x.aabbs.aabb_buffer)->gpu_address +
(D3D12_GPU_VIRTUAL_ADDRESS)x.aabbs.offset;
geometry.AABBs.AABBs.StrideInBytes = (UINT64)x.aabbs.stride;
@@ -5134,7 +5134,7 @@ std::mutex queue_locker;
if (resource == nullptr || !resource->IsValid())
return -1;
auto internal_state = to_internal(resource);
const auto internal_state = to_internal(resource);
switch (type)
{
@@ -5146,6 +5146,8 @@ std::mutex queue_locker;
}
else
{
if (subresource >= (int)internal_state->subresources_srv.size())
return -1;
return internal_state->subresources_srv[subresource].index;
}
break;
@@ -5156,6 +5158,8 @@ std::mutex queue_locker;
}
else
{
if (subresource >= (int)internal_state->subresources_uav.size())
return -1;
return internal_state->subresources_uav[subresource].index;
}
break;
@@ -5200,7 +5204,7 @@ std::mutex queue_locker;
const void* identifier = internal_state->stateObjectProperties->GetShaderIdentifier(internal_state->group_strings[group_index].c_str());
std::memcpy(dest, identifier, D3D12_SHADER_IDENTIFIER_SIZE_IN_BYTES);
}
void GraphicsDevice_DX12::SetName(GPUResource* pResource, const char* name) const
{
wchar_t text[256];
@@ -7241,7 +7245,7 @@ std::mutex queue_locker;
if (x.type == RaytracingAccelerationStructureDesc::BottomLevel::Geometry::Type::TRIANGLES)
{
geometry.Triangles.VertexBuffer.StartAddress = to_internal(&x.triangles.vertex_buffer)->gpu_address +
geometry.Triangles.VertexBuffer.StartAddress = to_internal(&x.triangles.vertex_buffer)->gpu_address +
(D3D12_GPU_VIRTUAL_ADDRESS)x.triangles.vertex_byte_offset;
geometry.Triangles.IndexBuffer = to_internal(&x.triangles.index_buffer)->gpu_address +
(D3D12_GPU_VIRTUAL_ADDRESS)x.triangles.index_offset * (x.triangles.index_format == IndexBufferFormat::UINT16 ? sizeof(uint16_t) : sizeof(uint32_t));
WickedEngine/wiGraphicsDevice_Vulkan.cpp
+28 -20
View File
@@ -617,7 +617,7 @@ namespace vulkan_internal
}
VKAPI_ATTR VkBool32 VKAPI_CALL debugUtilsMessengerCallback(
VkDebugUtilsMessageSeverityFlagBitsEXT message_severity,
VkDebugUtilsMessageSeverityFlagBitsEXT message_severity,
VkDebugUtilsMessageTypeFlagsEXT message_type,
const VkDebugUtilsMessengerCallbackDataEXT* callback_data,
void* user_data)
@@ -2412,7 +2412,7 @@ using namespace vulkan_internal;
}
}
#endif // _WIN32
if (validationMode != ValidationMode::Disabled)
{
// Determine the optimal validation layers to enable that are necessary for useful debugging
@@ -2954,7 +2954,7 @@ using namespace vulkan_internal;
queueFamiliesVideo[i].sType = VK_STRUCTURE_TYPE_QUEUE_FAMILY_VIDEO_PROPERTIES_KHR;
}
vkGetPhysicalDeviceQueueFamilyProperties2(physicalDevice, &queueFamilyCount, queueFamilies.data());
// Query base queue families:
for (uint32_t i = 0; i < queueFamilyCount; ++i)
{
@@ -3286,7 +3286,7 @@ using namespace vulkan_internal;
allocInfo.preferredFlags = VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
vulkan_check(vmaCreateBuffer(allocationhandler->allocator, &bufferInfo, &allocInfo, &nullBuffer, &nullBufferAllocation, nullptr));
VkBufferViewCreateInfo viewInfo = {};
viewInfo.sType = VK_STRUCTURE_TYPE_BUFFER_VIEW_CREATE_INFO;
viewInfo.format = VK_FORMAT_R32G32B32A32_SFLOAT;
@@ -3739,14 +3739,14 @@ using namespace vulkan_internal;
size_t size{};
vulkan_check(vkGetPipelineCacheData(device, pipelineCache, &size, nullptr));
// Get data of pipeline cache
// Get data of pipeline cache
wi::vector<uint8_t> data(size);
vulkan_check(vkGetPipelineCacheData(device, pipelineCache, &size, data.data()));
// Write pipeline cache data to a file in binary format
wi::helper::FileWrite(get_shader_cache_path(), data.data(), size);
// Destroy Vulkan pipeline cache
// Destroy Vulkan pipeline cache
vkDestroyPipelineCache(device, pipelineCache, nullptr);
pipelineCache = VK_NULL_HANDLE;
}
@@ -4131,7 +4131,7 @@ using namespace vulkan_internal;
1,
&copyRegion
);
copyAllocator.submit(cmd);
}
}
@@ -5511,7 +5511,7 @@ using namespace vulkan_internal;
default:
break;
}
}
VkPipelineLayoutCreateInfo pipelineLayoutInfo = {};
@@ -6978,56 +6978,64 @@ using namespace vulkan_internal;
case SubresourceType::SRV:
if (resource->IsBuffer())
{
auto internal_state = to_internal((const GPUBuffer*)resource);
const auto internal_state = to_internal((const GPUBuffer*)resource);
if (subresource < 0)
{
return internal_state->srv.index;
}
else
{
if (subresource >= (int)internal_state->subresources_srv.size())
return -1;
return internal_state->subresources_srv[subresource].index;
}
}
else if(resource->IsTexture())
else if (resource->IsTexture())
{
auto internal_state = to_internal((const Texture*)resource);
const auto internal_state = to_internal((const Texture*)resource);
if (subresource < 0)
{
return internal_state->srv.index;
}
else
{
if (subresource >= (int)internal_state->subresources_srv.size())
return -1;
return internal_state->subresources_srv[subresource].index;
}
}
else if (resource->IsAccelerationStructure())
{
auto internal_state = to_internal((const RaytracingAccelerationStructure*)resource);
const auto internal_state = to_internal((const RaytracingAccelerationStructure*)resource);
return internal_state->index;
}
break;
case SubresourceType::UAV:
if (resource->IsBuffer())
{
auto internal_state = to_internal((const GPUBuffer*)resource);
const auto internal_state = to_internal((const GPUBuffer*)resource);
if (subresource < 0)
{
return internal_state->uav.index;
}
else
{
if (subresource >= (int)internal_state->subresources_uav.size())
return -1;
return internal_state->subresources_uav[subresource].index;
}
}
else if (resource->IsTexture())
{
auto internal_state = to_internal((const Texture*)resource);
const auto internal_state = to_internal((const Texture*)resource);
if (subresource < 0)
{
return internal_state->uav.index;
}
else
{
if (subresource >= (int)internal_state->subresources_uav.size())
return -1;
return internal_state->subresources_uav[subresource].index;
}
}
@@ -7096,7 +7104,7 @@ using namespace vulkan_internal;
{
vulkan_check(vkGetRayTracingShaderGroupHandlesKHR(device, to_internal(rtpso)->pipeline, group_index, 1, SHADER_IDENTIFIER_SIZE, dest));
}
void GraphicsDevice_Vulkan::SetName(GPUResource* pResource, const char* name) const
{
if (!debugUtils || pResource == nullptr || !pResource->IsValid())
@@ -7839,7 +7847,7 @@ using namespace vulkan_internal;
assert(0);
}
commandlist.prev_swapchains.push_back(*swapchain);
VkRenderingInfo info = {};
info.sType = VK_STRUCTURE_TYPE_RENDERING_INFO;
info.renderArea.offset.x = 0;
@@ -8078,7 +8086,7 @@ using namespace vulkan_internal;
barrier.newLayout = _ConvertImageLayout(image.layout);
assert(barrier.newLayout != VK_IMAGE_LAYOUT_UNDEFINED);
barrier.srcStageMask = _ConvertPipelineStage(image.layout_before);
barrier.dstStageMask = _ConvertPipelineStage(image.layout);
barrier.srcAccessMask = _ParseResourceState(image.layout_before);
@@ -9232,7 +9240,7 @@ using namespace vulkan_internal;
raygen.deviceAddress += desc->ray_generation.offset;
raygen.size = desc->ray_generation.size;
raygen.stride = raygen.size; // raygen specifically must be size == stride
VkStridedDeviceAddressRegionKHR miss = {};
miss.deviceAddress = desc->miss.buffer ? to_internal(desc->miss.buffer)->address : 0;
miss.deviceAddress += desc->miss.offset;
@@ -9257,8 +9265,8 @@ using namespace vulkan_internal;
&miss,
&hitgroup,
&callable,
desc->width,
desc->height,
desc->width,
desc->height,
desc->depth
);
}