Documentation updates for V1.14

Author: chesik-amd

README.md

@@ -17,7 +17,7 @@ In order to use the latest features of RGP, it is strongly recommended that user
    * Radeon GPU Profiler (RGP)
 3. To capture a profile from a game, run the Radeon Developer Panel and follow the instructions in the Help. Help can be found in the following locations:
    * Help web pages exist in the "docs" sub directory
-   * Help web pages can be accessed from the **Help** button in the Developer Panel
+   * Help web pages can be accessed from the **Help** button in the Radeon Developer Panel
    * Help web pages can be accessed from the Welcome screen in the Radeon GPU Profiler, or from the **Help** menu
    * The documentation is hosted publicly at:
      * https://radeon-developer-panel.readthedocs.io/en/latest/
@@ -29,6 +29,7 @@ In order to use the latest features of RGP, it is strongly recommended that user
  * Vulkan
 
 ### Supported RDNA and GCN hardware
+* AMD Radeon RX 7000 series
 * AMD Radeon RX 6000 series
 * AMD Radeon RX 5000 series
 * AMD Radeon VII
@@ -41,16 +42,16 @@ In order to use the latest features of RGP, it is strongly recommended that user
 ### Supported Operating Systems
 * Windows® 10
 * Windows® 11
-* Ubuntu 18.04 LTS (Vulkan only)
-* Ubuntu 20.04 LTS (Vulkan only)
 * Ubuntu 22.04 LTS (Vulkan only)
 
 
 ## Compute APIs, RDNA and GCN hardware, and operating systems
 ### Supported APIs
 * OpenCL™
+* HIP
 
 ### Supported RDNA and GCN hardware
+* AMD Radeon RX 7000 series
 * AMD Radeon RX 6000 series
 * AMD Radeon RX 5000 series
 * AMD Radeon VII

Release_Notes.txt

@@ -68,7 +68,7 @@ The table shows the following information:
    the Wavefront occupancy, Event timing, Instruction timing or Pipeline
    state panes and view the event represented by the selected row in these
    panes, as well as in the side panels. The user can also see the parent
-   command buffer in the Frame Summary pane or navigate to the Render/depth
+   command buffer in the Frame summary pane or navigate to the Render/depth
    targets view and view the event in the timeline.
 
    Below is a screenshot of what the right-click context menu looks like:
@@ -102,23 +102,24 @@ The following Layout Transition columns are shown in the Barriers table:
 See `https://gpuopen.com/dcc-overview/ <https://gpuopen.com/dcc-overview/>`_ for more information
 on what may cause a **DCC Decompress** or what "clear" values can be used to skip **Fast Clear Eliminates**.
 
-Barriers and OpenCL
-~~~~~~~~~~~~~~~~~~~
+Barriers and OpenCL/HIP
+~~~~~~~~~~~~~~~~~~~~~~~
 
-Barriers for OpenCL profiles provide visibility into how the driver scheduled
+Barriers for OpenCL or HIP profiles provide visibility into how the driver scheduled
 dispatches to the GPU and dependencies between kernel dispatches. These barriers
 are the same synchronization primitives used by DirectX12 and Vulkan that are described above.
 
-The barriers shown in an OpenCL profile correspond to the barriers
-inserted by the OpenCL driver for one of the following reasons.
-
-#. **Queue Profiling** - The application has enabled profiling CL_QUEUE_PROFILING_ENABLE property
-   when creating a command queue. This causes barriers to be inserted so that timestamps can be recorded.
+The barriers shown in an OpenCL or HIP profile correspond to the barriers
+inserted by the OpenCL or HIP driver for one of the following reasons.
 
 #. **Data Dependencies** - There are data dependencies between subsequent dispatches. For
    example, reading the results of a previous kernel dispatch. This causes barriers to be inserted
    so that caches can be invalidated.
 
+#. **Queue Profiling** - (OpenCL-specific) The application has enabled profiling CL_QUEUE_PROFILING_ENABLE
+   property when creating a command queue. This causes barriers to be inserted so that timestamps can be
+   recorded.
+
 OpenCL command queues process dispatches one after another and it is common for a
 subsequent kernel dispatch to use the results of a previous kernel dispatch. For this reason, it
 can be expected that an RGP profile will have a large number of barriers.
@@ -132,7 +133,9 @@ As we see, the time taken due to barriers is typically very small since inter-di
 .. image:: media_rgp/rgp_barriers_opencl_2.png
 
 
-It should be noted that the meaning of barriers in RGP for OpenCL is different from OpenCL's synchronization
-APIs and is not related to the OpenCL synchronization APIs based on cl_event or cl_barrier.
-For this reason, the barriers seen in OpenCL profiles are known as cmdBarrier() which is not a part of the OpenCL API.
-For OpenCL profiles, RGP does not presently show OpenCL events or host synchronization.
+It should be noted that the meaning of barriers in RGP for OpenCL/HIP is different from
+OpenCL or HIP built-in synchronization APIs. For example, barriers that appear in an
+OpenCL RGP profile are not related to the OpenCL synchronization APIs based on cl_event
+or cl_barrier. For this reason, the barriers seen in OpenCL/HIP profiles are displayed
+as **CmdBarrier()** which is not a part of the OpenCL or HIP API. For these profiles,
+RGP does not currently show API-specific events or host synchronization.

docs/source/barriers.rst

@@ -54,9 +54,9 @@
 # built documents.
 #
 # The short X.Y version.
-version = u'1.13.1'
+version = u'1.14.0'
 # The full version, including alpha/beta/rc tags.
-release = u'1.13.1'
+release = u'1.14.0'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

docs/source/conf.py

@@ -127,12 +127,11 @@ of the vertices shaded by the vertex shader.
 Users may also obtain information about an event's parent command buffer
 by right-clicking on an event. This will bring up a context menu which
 contains a menu item to find the event's parent command buffer. Selecting
-this menu item will navigate to the Frame Summary and set focus on the selected
-event's parent command buffer. Once here, users can obtain valuable system-level
-insight about the surrounding context for the event in question.
+this menu item will navigate to the Frame summary pane and set focus on the
+selected event's parent command buffer. Once here, users can obtain valuable
+system-level insight about the surrounding context for the event in question.
 
-Compute dispatches for both graphics APIs and OpenCL have a simpler structure.
-A sample compute event is shown below.
+Compute dispatches have a simpler structure. A sample compute event is shown below.
 
 .. image:: media_rgp/rgp_compute_event.png
 

docs/source/event_timing.rst

@@ -2,9 +2,9 @@ The Radeon™ GPU Profiler
 ========================
 
 The Radeon GPU Profiler is a performance tool that can be used by
-developers to optimize DirectX®12, Vulkan® and OpenCL™ applications for AMD RDNA™
-and GCN hardware. It is part of a suite of tools comprised of the following
-software:
+developers to optimize DirectX®12, Vulkan®, OpenCL™ and HIP applications for
+AMD RDNA™ and GCN hardware. It is part of a suite of tools comprised of the
+following software:
 
 -  **Radeon Developer Mode Driver** - This is shipped as part of the AMD
    public Adrenalin driver and supports the developer mode features required for
@@ -17,7 +17,7 @@ software:
 
 -  **Radeon Developer Panel (RDP)** - A GUI application that allows the
    developer to configure driver settings and generate profiler data
-   from DirectX12, Vulkan and OpenCL applications.
+   from DirectX12, Vulkan, OpenCL and HIP applications.
 
 -  **Radeon GPU Profiler (RGP)** - A GUI tool used to visualize and
    analyze the profile data.
@@ -34,7 +34,7 @@ software:
    provides unique visualizations of queue synchronization using fences
    and semaphores, asynchronous compute, and barrier timings. Currently,
    it supports the explicit graphics APIs (DirectX12 and Vulkan), compute
-   APIs (OpenCL) and will NOT work with older graphics APIs such as
+   APIs (OpenCL and HIP) and will NOT work with older graphics APIs such as
    DirectX11 or OpenGL.
 
 Graphics APIs, RDNA and GCN hardware, and operating systems
@@ -48,6 +48,8 @@ Graphics APIs, RDNA and GCN hardware, and operating systems
 
 \ **Supported RDNA and GCN hardware**
 
+-  AMD Radeon RX 7000 series
+
 -  AMD Radeon RX 6000 series
 
 -  AMD Radeon RX 5000 series
@@ -70,10 +72,6 @@ Graphics APIs, RDNA and GCN hardware, and operating systems
 
 -  Windows® 11
 
--  Ubuntu 18.04 LTS (Vulkan only)
-
--  Ubuntu 20.04 LTS (Vulkan only)
-
 -  Ubuntu 22.04 LTS (Vulkan only)
 
 Compute APIs, RDNA and GCN hardware, and operating systems
@@ -83,8 +81,12 @@ Compute APIs, RDNA and GCN hardware, and operating systems
 
 -  OpenCL
 
+-  HIP
+
 \ **Supported RDNA and GCN hardware**
 
+-  AMD Radeon RX 7000 series
+
 -  AMD Radeon RX 6000 series
 
 -  AMD Radeon RX 5000 series
@@ -183,19 +185,19 @@ the profile data are within the **Overview** and **Events** sections.
 2. **Overview**
 
    a. **Frame Summary** - Contains a summary of the structure of the
-      graphics frame. This overview section is not available for OpenCL profiles.
+      graphics frame. This overview section is not available for OpenCL or HIP profiles.
 
-   b. **Profile Summary** - Contains a summary of the structure of the OpenCL profile.
+   b. **Profile Summary** - Contains a summary of the structure of the OpenCL of HIP profile.
 
    c. **Barriers** - Details of the barrier usage in the profile.
 
    d. **Context rolls** - Details of the hardware context register usage.
-      This overview section is not available for OpenCL profiles.
+      This overview section is not available for OpenCL or HIP profiles.
 
    e. **Most expensive events** - List of the most expensive events.
 
    f. **Render/depth targets** - Overview of render targets used throughout
-      the graphics frame. This overview section is not available for OpenCL profiles.
+      the graphics frame. This overview section is not available for OpenCL or HIP profiles.
 
    g. **Pipelines** - Details of the pipeline usage in the profile.
 
@@ -597,7 +599,8 @@ User Debug Markers
 ==================
 
 User markers can help application developers to correlate the data seen
-in RGP with their application behavior. User Markers are presently not supported for OpenCL.
+in RGP with their application behavior. User Markers are currently
+not supported for OpenCL or HIP.
 
 DirectX12 User Markers
 ----------------------