Backport of 6.4.2 for cherry-pick list (#714)

Auto-submit by Jenkins

Author: ywang103-amd

.github/workflows/formatting.yml

@@ -13,7 +13,7 @@ concurrency:
 
 jobs:
   python:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-22.04
 
     steps:
     - name: Checkout
@@ -35,7 +35,7 @@ jobs:
       uses: isort/isort-action@master
 
   cmake:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-22.04
 
     steps:
     - uses: actions/checkout@v4
@@ -58,7 +58,7 @@ jobs:
         fi
 
   python-bytecode:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-22.04
 
     steps:
     - uses: actions/checkout@v4

CHANGELOG.md

@@ -2,6 +2,21 @@
 
 Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.amd.com/projects/rocprofiler-compute/en/latest/](https://rocm.docs.amd.com/projects/rocprofiler-compute/en/latest/).
 
+## ROCm Compute Profiler 3.2.0 for ROCm 6.4.2
+
+### Added
+
+* Add FP8 metrics' support for MI300
+* Add additional datatype for roofline: FP8, FP16, BF16, FP32, FP64, I8, I32, I64 (dependent on gpu architecture)
+* Add datatype selection option for roofline profiling: --roofline-data-type / -R option (Default is FP32)
+* Change dependency from rocm-smi to amd-smi
+
+### Changed
+
+
+### Resolved issues
+* Fixed a crash related to Agent ID caused by the new format of the rocprofv3  output CSV file
+
 ## ROCm Compute Profiler 3.1.0 for ROCm 6.4.0
 
 ### Added

VERSION

@@ -1 +1 @@
-3.1.0
+3.2.0

docs/data/profile/sample-roof-plot.jpg

@@ -75,6 +75,11 @@ application's profiling data:
 
 #. Memory Chart Analysis
 #. Empirical Roofline Analysis
+
+   Use ``--roofline-data-type`` option to specify which data type(s) you would like plotted on the roofline PDFs in the standalone analysis GUI.
+   Datatypes can be stacked- for example, "--roofline-data-type FP32 FP64 I32" would display one PDF with FP32 and FP64 stacked, and one PDF with INT32.
+   Default roofline datatype plotted is FP32.
+
 #. Top Stats (Top Kernel Statistics)
 #. System Info
 #. System Speed-of-Light

docs/how-to/analyze/standalone-gui.rst

@@ -398,6 +398,9 @@ Roofline options
    Allows you to specify a device ID to collect performance data from when
    running a roofline benchmark on your system.
 
+``--roofline-data-type <datatype>``
+   Allows you to specify the data types that you want plotted in the roofline PDF output(s). Selecting more than one data type will overlay the results onto the same plot. Default data type: FP32
+
 To distinguish different kernels in your ``.pdf`` roofline plot use
 ``--kernel-names``. This will give each kernel a unique marker identifiable from
 the plot's key.
@@ -431,8 +434,7 @@ successfully.
 
    $ ls workloads/vcopy/MI200/
    total 48
-   -rw-r--r-- 1 auser agroup 13331 Mar  1 16:05 empirRoof_gpu-0_fp32_fp64.pdf
-   -rw-r--r-- 1 auser agroup 13136 Mar  1 16:05 empirRoof_gpu-0_int8_fp16.pdf
+   -rw-r--r-- 1 auser agroup 13331 Mar  1 16:05 empirRoof_gpu-0_FP32.pdf
    drwxr-xr-x 1 auser agroup     0 Mar  1 16:03 perfmon
    -rw-r--r-- 1 auser agroup  1101 Mar  1 16:03 pmc_perf.csv
    -rw-r--r-- 1 auser agroup  1715 Mar  1 16:05 roofline.csv
@@ -441,11 +443,9 @@ successfully.
 
 .. note::
 
-   ROCm Compute Profiler generates two roofline outputs to organize results and reduce
-   clutter. One chart plots FP32/FP64 performance while the other plots I8/FP16
-   performance.
+   ROCm Compute Profiler currently captures roofline profiling for all data types, and you can reduce the clutter in the PDF outputs by filtering the data type(s). Selecting multiple data types will overlay the results into the same PDF. To generate results in separate PDFs for each data type from the same workload run, you can re-run the profiling command with each data type as long as the ``roofline.csv`` file still exists in the workload folder.
 
-The following image is a sample ``empirRoof_gpu-0_int8_fp16.pdf`` roofline
+The following image is a sample ``empirRoof_gpu-0_FP32.pdf`` roofline
 plot.
 
 .. image:: ../../data/profile/sample-roof-plot.jpg

docs/how-to/profile/mode.rst

@@ -231,7 +231,7 @@ The following table lists ROCm Compute Profiler's basic operations, their
 
    * - :ref:`Standalone roofline analysis <standalone-roofline>`
      - ``profile``
-     - ``--name``, ``--roof-only``, ``-- <profile_cmd>``
+     - ``--name``, ``--roof-only``, ``--roofline-data-type <data_type>``, ``-- <profile_cmd>``
 
    * - :ref:`Import a workload to database <grafana-gui-import>`
      - ``database``

docs/how-to/use.rst

@@ -154,7 +154,7 @@ def omniarg_parser(
         default=False,
         action="store_true",
         help=argparse.SUPPRESS,
-        #help="\t\t\tKokkos trace, traces Kokkos API calls.",
+        # help="\t\t\tKokkos trace, traces Kokkos API calls.",
     )
     profile_group.add_argument(
         "-k",
@@ -316,6 +316,19 @@ def omniarg_parser(
         action="store_true",
         help="\t\t\tInclude kernel names in roofline plot.",
     )
+
+    roofline_group.add_argument(
+        "-R",
+        "--roofline-data-type",
+        required=False,
+        choices=["FP8", "FP16", "BF16", "FP32", "FP64", "I8", "I32", "I64"],
+        metavar="",
+        nargs="+",
+        type=str,
+        default=["FP32"],
+        help="\t\t\tChoose datatypes to view roofline PDFs for: (DEFAULT: FP32)\n\t\t\t   FP8\n\t\t\t   FP16\n\t\t\t   BF16\n\t\t\t   FP32\n\t\t\t   FP64\n\t\t\t   I8\n\t\t\t  I32\n\t\t\t I64\n\t\t\t ",
+    )
+
     # roofline_group.add_argument('-w', '--workgroups', required=False, default=-1, type=int, help="\t\t\tNumber of kernel workgroups (DEFAULT: 1024)")
     # roofline_group.add_argument('--wsize', required=False, default=-1, type=int, help="\t\t\tWorkgroup size (DEFAULT: 256)")
     # roofline_group.add_argument('--dataset', required=False, default = -1, type=int, help="\t\t\tDataset size (DEFAULT: 536M)")
@@ -510,6 +523,19 @@ def omniarg_parser(
         const=8050,
         help="\t\tActivate a GUI to interate with rocprofiler-compute metrics.\n\t\tOptionally, specify port to launch application (DEFAULT: 8050)",
     )
+
+    analyze_group.add_argument(
+        "-R",
+        "--roofline-data-type",
+        required=False,
+        choices=["FP8", "FP16", "BF16", "FP32", "FP64", "I8", "I32", "I64"],
+        metavar="",
+        nargs="+",
+        type=str,
+        default=["FP32"],
+        help="\t\t\tChoose datatypes to view roofline PDFs for: (DEFAULT: FP32)\n\t\t\t   FP8\n\t\t\t   FP16\n\t\t\t   BF16\n\t\t\t   FP32\n\t\t\t   FP64\n\t\t\t   I8\n\t\t\t  I32\n\t\t\t I64\n\t\t\t ",
+    )
+
     analyze_advanced_group.add_argument(
         "--random-port",
         action="store_true",

src/argparser.py

@@ -61,6 +61,8 @@ def __init__(self, args, supported_archs):
         # define any elements which will have full width
         self.__full_width_elements = {1801}
 
+        self.__roofline_data_type = args.roofline_data_type
+
     @demarcate
     def build_layout(self, input_filters, arch_configs):
         """
@@ -180,6 +182,7 @@ def generate_from_filter(
                         "mem_level": "ALL",
                         "include_kernel_names": False,
                         "is_standalone": False,
+                        "roofline_data_type": self.__roofline_data_type,
                     }
                 )
                 roof_obj = self.get_socs()[self.arch].roofline_obj

src/rocprof_compute_analyze/analysis_webui.py

@@ -32,6 +32,12 @@ Panel Config:
             peak: (((($max_sclk * $cu_per_gpu) * 64) * 2) / 1000)
             pop: None # No perf counter
             tips:
+          MFMA FLOPs (F8):
+            value: None # No HW module
+            unit: GFLOP
+            peak: None # No HW module
+            pop: None # No HW module
+            tips:
           MFMA FLOPs (BF16):
             value: None # No perf counter
             unit: GFLOPs