QuoteChange log for September 22, 2023 Vulkan 1.3.265 spec update:source: https://github.com/KhronosGroup/Vulkan-Docs/commit/4871ab9e57fb07f98bf016cb10a3088924976e29Github Issues
* Remove misplaced
ename:VK_PIPELINE_STAGE_2_ACCELERATION_STRUCTURE_COPY_BIT_KHR
description from elink:VkPipelineStageFlagBits context (public PR 2210).
* Swap NV pipeline stage with promoted KHR stage in XML `syncequivalent`
tag (public issue 2211).
* Remove redundant slink:VkPipelineShaderStageCreateInfo pname:stage VUs
(public issue 2222).
* Remove duplicate definition in apiext:VK_AMD_shader_enqueue proposal
document (public issue 2225).
Internal Issues
* Add a common copy image VU to require compressed image-to-image copies
to have formats with matching texel block extents (internal issue 3610).
* Add `page` field to `validusage.json` for eventual use with Antora
(internal issue 3617).
* Add missing VUs to flink:vkGetCalibratedTimestampsEXT (internal MR
5974).
* Attempt to clarify apiext:VK_KHR_fragment_shader_barycentric
pname:triStripVertexOrderIndependentOfProvokingVertex description
(internal MR 6048).
* Add a slink:VkGraphicsPipelineCreateInfo VU restricting
ename:VK_PIPELINE_CREATE_RETAIN_LINK_TIME_OPTIMIZATION_INFO_BIT_EXT
(internal MR 6119).
* Reword several places to clarify that code:PrimitiveId is reset to `0`
between each draw in a multi-draw command (internal issue 3564 ).
* Consolidate and simplify VUs in the <<copies, Copy Commands>> chapter
(internal MR 6124).
* Remove inconsistent `KHR` suffix in
slink:VkDeviceImageMemoryRequirements VUIDs (internal MR 6124).
QuoteRay tracing using the DirectX or Vulkan API consists of first building acceleration structures then tracing rays by traversing those acceleration structures. Up until now, Radeon Raytracing Analyzer (RRA) only provided mechanisms for inspecting acceleration structures but did not have a way to view traversal data about the rays that traversed these acceleration structures. With RRA 1.3 there is a new Ray tab that shows ray traversal statistics and as well as the ability to inspect each casted ray in 3D.
Quote- Addition of a RAY tab, to enable the ability to visualize dispatches and individual rays.
- Ray Dispatches pane added to visualize ray dispatches and allow selection of pixels within those dispatches.
- Ray Inspector pane added to visualize individual rays and their collisions with scene objects.
- Reset button added to TLAS/BLAS viewer panes and Inspector pane to reset the UI to its default settings.
- Allow viewer UI state to be persistent between RRA sessions via a checkbox in the general settings.
- Bugs/stability fixes.
QuoteOne of the most anticipated features introduced with this release is improved support for aliased resources on the Resource overview pane. Other notable enhancements include support for loading Radeon GPU Detective crash dumps, system memory info in the Device configuration pane and hours mode for units of time. In addition, the handling of named resources and filtering of implicit buffers has been made more reliable.
Quote- Aliased Resource size calculation improvements on Resource overview pane.
- Support for loading Radeon GPU Detective crash dump files.
- Resource naming and implicit buffer filtering made more reliable in DX12.
- Support for hours added (formatted as "H:MM:SS.clk_cycles" and included with the CTRL+T time-cycling shortcut keys).
- System memory type added to device configuration pane (e.g. DDR4).
- Bug/stability fixes.
Quote- Support for additional AMD RDNA 3 hardware
- The vertical scroll bars in the ISA disassembly view (in Instruction timing and Pipeline state panes) now indicate the location of search matches and instruction latency hotspot
- Added support for showing the ray tracing pipeline in the Pipeline state pane for profiles generated by the RADV driver (the bulk of this work was implemented by the RADV team, but this RGP release contains some corresponding changes to support this)
- The Output Merger (OM) Pipeline state pane now shows the Stencil reference value as part of the Depth/Stencil state
- The Output Merger (OM) Pipeline state pane now shows the correct value for "Alpha to coverage enable" on recent hardware
- The Details panel for an event will now show the API shader hashes for each shader in the associated pipeline
- Fix an issue with incorrect behavior in the Wavefront Histogram for long-running compute events
- Fix issues when running RGP on some OS/desktops using a Dark theme (#88)
- Added instruction timing support for HIP kernels with function calls
- Allow opening .rgp files larger than 2 GB in size
Bug/stability fixes