enable single instance for f8xf16 gemm_universal in profiler

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn.hpp

@@ -40,7 +40,7 @@ using device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances = std::tuple<
         //#########################|        |        |        | Type|  Type|  Type|    Type|     Type| Elementwise| Elementwise| Elementwise|Specialization|  Size| Block| Block| Block|    |    | XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|                               Pipeline|                     Pipeline|
         //#########################|        |        |        |     |      |      |        |         |   Operation|   Operation|   Operation|              |      |      |      |      |    |    |    |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|                              Scheduler|                     Verision|
         //#########################|        |        |        |     |      |      |        |         |            |            |            |              |      |      |      |      |    |    |    |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |                                       |                             |
-        
+        /*
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,  F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   128,   128,    64,    16,   4,  32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,             16,              16,          0,    S<16, 16, 1>,    S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 32, 1, 8>,               8,  BlockGemmPipelineScheduler::Intrawave, BlockGemmPipelineVersion::v4>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,  F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   256,   128,   128,    16,   8,  32,   32,    4,    2,     S<8, 32, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,             16,              16,          0,    S<16, 16, 1>,    S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,          0,          1,           1,                   S<1, 32, 1, 8>,               8,  BlockGemmPipelineScheduler::Intrawave, BlockGemmPipelineVersion::v3>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,  F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   192,   256,    64,    16,   8,  32,   32,    3,    4,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,             16,              16,          0,    S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,          0,          1,           1,                   S<1, 32, 1, 8>,               8,  BlockGemmPipelineScheduler::Intrawave, BlockGemmPipelineVersion::v3>,
@@ -50,6 +50,7 @@ using device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances = std::tuple<
         // We prefer following instance, however, existing compiler bug cause it failed to generate sanity code.
         // DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,  F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   128,   128,    64,    16,   4,  32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,             16,              16,          0,    S<16, 16, 1>,    S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 32, 1, 8>,               8,  BlockGemmPipelineScheduler::Intrawave, BlockGemmPipelineVersion::v5>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,  F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   128,   128,    64,    16,   4,  32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,             16,              16,          0,    S<16, 16, 1>,    S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 32, 1, 8>,               8,  BlockGemmPipelineScheduler::Interwave, BlockGemmPipelineVersion::v1>
+        */
     // clang-format on
     >;
 
@@ -62,6 +63,7 @@ using device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances = std::tuple<
         //#########################|        |        |        |     |      |      |        |         |            |            |            |              |      |      |      |      |    |    |    |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |                  |                             |
 
         // Latency friendly
+        /*
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,    32,   16,   128,  16,   2,  16,   16,    1,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<64, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              2,          0,          1,           1,                   S<1, 16, 1, 8>,               2,  BlkGemmPipeSched, BlockGemmPipelineVersion::v1>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,    64,    16,   16,   128,  16,   4,  16,   16,    1,    1,     S<8, 8, 1>,      S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<32, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 16, 1, 4>,               4,  BlkGemmPipeSched, BlockGemmPipelineVersion::v1>,
         // Memory friendly
@@ -69,10 +71,13 @@ using device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances = std::tuple<
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   256,   256,   16,   128,  16,   2,  16,   16,    4,    1,     S<8, 32, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<64, 4, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,          0,          1,           1,                   S<1, 32, 1, 8>,               2,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,   128,   32,   128,  16,   4,  32,   32,    2,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<32, 4, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 16, 1, 8>,               4,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,   128,   16,   128,  16,   2,  16,   16,    4,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<64, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              2,          0,          1,           1,                   S<1, 16, 1, 8>,               2,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
-        DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,    64,   32,   128,  16,   4,  32,   32,    1,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<32, 4, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 16, 1, 8>,               4,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
+        */
+        DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,    64,   32,   128,  16,   4,  32,   32,    1,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<32, 4, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 16, 1, 8>,               4,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>
+        /*
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,    64,   16,   128,  16,   2,  16,   16,    2,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<64, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              2,          0,          1,           1,                   S<1, 16, 1, 8>,               2,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,   128,    32,   16,   128,  16,   2,  16,   16,    1,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<64, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              2,          0,          1,           1,                   S<1, 16, 1, 8>,               2,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>,
         DeviceGemm_Xdl_CShuffleV3<  Row,     Row,     Row,     F8,    F16,  F16,   F32,     F16,      PassThrough, PassThrough, PassThrough,       GemmSpec,    64,    16,   16,   128,  16,   4,  16,   16,    1,    1,     S<8, 8, 1>,      S<1, 0, 2>,    S<1, 0, 2>,               2,              16,             16,          0,    S<32, 2, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              4,          0,          1,           1,                   S<1, 16, 1, 4>,               4,  BlkGemmPipeSched, BlockGemmPipelineVersion::v2>
+        */
     // clang-format on
     >;
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_default_instance.cpp

@@ -13,8 +13,9 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_default_instances(
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmDefault>{});
+    // add_device_operation_instances(
+    //     instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmDefault>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_kpadding_instance.cpp

@@ -13,8 +13,9 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_kpadding_instances(
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmKPadding>{});
+    // add_device_operation_instances(
+    //     instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmKPadding>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_mnkpadding_instance.cpp

@@ -13,8 +13,9 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_mnkpadding_instances
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmMNKPadding>{});
+    // add_device_operation_instances(
+    //     instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmMNKPadding>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_mnpadding_instance.cpp

@@ -13,8 +13,9 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_mnpadding_instances(
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmMNPadding>{});
+    // add_device_operation_instances(
+    //     instances, device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_comp_instances<GemmMNPadding>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v1_default_instance.cpp

@@ -13,9 +13,10 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v1_default_instances(
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances,
-        device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Intrawave, GemmDefault>{});
+    // add_device_operation_instances(
+    //     instances,
+    //     device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Intrawave, GemmDefault>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v1_kpadding_instance.cpp

@@ -13,9 +13,10 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v1_kpadding_instances
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances,
-        device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Intrawave, GemmKPadding>{});
+    //  add_device_operation_instances(
+    //      instances,
+    //      device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Intrawave, GemmKPadding>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_default_instance.cpp

@@ -13,9 +13,10 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_default_instances(
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances,
-        device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmDefault>{});
+    // add_device_operation_instances(
+    //     instances,
+    //     device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmDefault>{});
+    ignore = instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_kpadding_instance.cpp

@@ -13,9 +13,10 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_kpadding_instances
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances,
-        device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmKPadding>{});
+    // add_device_operation_instances(
+    //     instances,
+    //     device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmKPadding>{});
+    ignore= instances;
 }
 
 } // namespace instance

library/src/tensor_operation_instance/gpu/gemm_universal/device_gemm_xdl_universal_f8_f16_f16/device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_mnkpadding_instance.cpp

@@ -16,9 +16,10 @@ void add_device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_v2_mnkpadding_instanc
         DeviceGemmV2<Row, Row, Row, F8, F16, F16, PassThrough, PassThrough, PassThrough>>>&
         instances)
 {
-    add_device_operation_instances(
-        instances,
-        device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmMNKPadding>{});
+    // add_device_operation_instances(
+    //     instances,
+    //     device_gemm_xdl_universal_f8_f16_f16_mk_kn_mn_mem_instances<Interwave, GemmMNKPadding>{});
+    ignore = instances;
 }
 
 } // namespace instance