Merge branch 'main' into Lokesh9106-patch-1

Author: metascroy

.github/workflows/add-unanswered-to-project.yml

@@ -20,31 +20,32 @@ jobs:
 
             // List of authors to exclude
             const excludedAuthors = new Set([
-              "nil-is-all", "cbilgin", "kimishpatel", "psiddh", "digantdesai", "SS-JIA", "ahmtox", "mcr229", "shoumikhin",
-              "manuelcandales", "metascroy", "cccclai", "rohansjoshi", "kirklandsign", "abhinaykukkadapu", "JacobSzwejbka",
-              "Conarnar", "lucylq", "larryliu0820", "BujSet", "Gasoonjia", "Juntian777", "guangy10", "jackzhxng",
-              "GregoryComer", "leafs1", "swolchok", "mergennachin", "tarun292", "byjlw", "jathu", "Jack-Khuu", "georgehong",
+              "nil-is-all", "tanvirislam-meta", "cbilgin", "kimishpatel", "psiddh", "digantdesai", "SS-JIA", "ahmtox", "mcr229", 
+              "shoumikhin", "manuelcandales", "metascroy", "cccclai", "rohansjoshi", "kirklandsign", "abhinaykukkadapu", 
+              "JacobSzwejbka", "Conarnar", "lucylq", "larryliu0820", "BujSet", "Gasoonjia", "Juntian777", "guangy10", "jackzhxng",
+              "GregoryComer", "leafs1", "swolchok", "mergennachin", "tarun292", "byjlw", "jathu", "Jack-Khuu", "georgehong", 
               "zhenyan-zhang-meta", "silverguo", "harishs88ss", "AlannaBurke", "dbort", "huydhn", "mcremon-meta", "trivedivivek", 
-              "angelayi", "helunwencser", "hsharma35", "zhxchen17", "iseeyuan", "svekars", "nathanaelsee", "dulinriley", "jerryzh168",
-              "cmodi-meta", "bigfootjon", "sxu", "ydwu4", "Riandy", "tugsbayasgalan", "bsoyluoglu", "yangw-dev", "YIWENX14",
-              "namanahuja", "yushangdi", "limintang", "pianpwk", "viveknayakatmeta", "andreanicastro", "JakeStevens",
+              "angelayi", "helunwencser", "hsharma35", "zhxchen17", "iseeyuan", "svekars", "nathanaelsee", "dulinriley", 
+              "jerryzh168", "cmodi-meta", "bigfootjon", "sxu", "ydwu4", "Riandy", "tugsbayasgalan", "bsoyluoglu", "yangw-dev", 
+              "YIWENX14", "namanahuja", "yushangdi", "limintang", "pianpwk", "viveknayakatmeta", "andreanicastro", "JakeStevens",
               "gmagogsfm", "zonglinpeng", "eigen-k", "derekxu", "salilsdesai", "skrtskrtfb", "pssrawat", "r-barnes",
               "kalpit-meta-1", "Will-MingLun-Li", "KapJI", "piyengar", "j-bahr", "BoyuanFeng", "fgasperij", "DariusHolmgren",
               "sammarden-meta", "kushrast", "meta-emilian", "Rittzz", "jeanschmidt", "copyrightly", "mikekgfb", "vmpuri",
-              "zonglinpengmeta", "maggiemoss", "aorenste", "hoangminhle98", "Solumin", "meyering", "rchen152",
-              "AishwaryaSivaraman", "migeed-z", "ebgraham", "Esteb37", "nausicaasnow", "Camyll", "ezyang", "huiyujie",
-              "dltn", "cjhopman", "blackm00n", "agunapal", "SamGondelman", "Ninja91", "ivayloen", "DrJessop", "rodrigos01meta",
-              "akrieger", "cmt0", "yiming0416", "ethansfng", "ThomasJannaud", "nirvanagth", "marcinkwiatkowski", "3l1",
-              "omerjerk", "nitish2112", "yipjustin", "ejnguyen", "andrewor14", "phaiting", "mgiordy", "LeeOHzzZ", "adicatana",
-              "Polyomino", "ezrilow", "navsud", "YifanShenSZ", "RdoubleA", "Olivia-liu", "Abhi-hpp", "Vysarat", "azad-meta",
-              "pytorchbot", "pytorchmergebot", "pytorchupdatebot", "facebook-github-bot", "app/dependabot", "Erik-Lundell",
-              "zingo", "AdrianLundell", "oscarandersson8218", "per", "Sebastian-Larsson", "SaoirseARM", "robell", "mansnils",
-              "martinlsm", "freddan80", "YufengShi-dudu", "tom-arm", "perheld", "Jerry-Ge", "gggekov", "fumchin", "wwwind",
+              "zonglinpengmeta", "maggiemoss", "aorenste", "hoangminhle98", "Solumin", "meyering", "rchen152", "AishwaryaSivaraman", 
+              "migeed-z", "ebgraham", "Esteb37", "nausicaasnow", "Camyll", "ezyang", "huiyujie", "dltn", "cjhopman", "blackm00n", 
+              "agunapal", "SamGondelman", "Ninja91", "ivayloen", "DrJessop", "rodrigos01meta", "akrieger", "cmt0", "yiming0416", 
+              "ethansfng", "ThomasJannaud", "nirvanagth", "marcinkwiatkowski", "3l1", "omerjerk", "nitish2112", "yipjustin", 
+              "ejnguyen", "andrewor14", "phaiting", "mgiordy", "LeeOHzzZ", "adicatana", "Polyomino", "ezrilow", "navsud", 
+              "michaelmaitland", "RahulC7", "seyeong-han", "YifanShenSZ", "RdoubleA", "Olivia-liu", "Abhi-hpp", "Vysarat", 
+              "azad-meta", "junpi", "pytorchbot", "pytorchmergebot", "pytorchupdatebot", "facebook-github-bot", "app/dependabot", 
+              "Erik-Lundell", "zingo", "AdrianLundell", "oscarandersson8218", "per", "Sebastian-Larsson", "SaoirseARM", "robell", 
+              "mansnils", "martinlsm", "freddan80", "YufengShi-dudu", "tom-arm", "perheld", "Jerry-Ge", "gggekov", "fumchin", "wwwind",
               "benkli01", "Tessil", "maddun01", "Michiel-Olieslagers", "armwaheed", "agrima1304", "emmakujala", "annietllnd", 
-              "haowhsu-quic", "shewu-quic", "winskuo-quic", "chunit-quic", "DannyYuyang-quic", "chuntl", "thchenqti", 
-              "jethroqti", "cymbalrush", "DenisVieriu97", "billmguo", "StrycekSimon", "jirioc", "robert-kalmar", "skywall", 
-              "MartinPavella", "roman-janik-nxp", "novak-vaclav ", "neuropilot-captain", "dijopaul", "cad-rlc", "cad-audio", 
-              "ynimmaga", "daniil-lyakhov", "emmanuel-ferdman", "cavusmustafa", "Jiseong-oh", "alexdean08"
+              "MatthiasHertel80", "AlexTawseArm", "jmahbs", "haowhsu-quic", "shewu-quic", "winskuo-quic", "chunit-quic", 
+              "DannyYuyang-quic", "chuntl", "thchenqti", "jethroqti", "chenweng-quic", "cymbalrush", "DenisVieriu97", "billmguo", 
+              "StrycekSimon", "jirioc", "robert-kalmar", "skywall", "MartinPavella", "roman-janik-nxp", "novak-vaclav ", 
+              "neuropilot-captain", "dijopaul", "cad-rlc", "cad-audio", "ynimmaga", "daniil-lyakhov", "emmanuel-ferdman", 
+              "cavusmustafa", "Jiseong-oh", "alexdean08"
             ]);
 
             async function addItem(contentId, type, number) {

.github/workflows/cuda.yml

@@ -24,7 +24,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        cuda-version: ["12.6", "12.8", "13.0"]
+        cuda-version: ["12.6", "12.8", "12.9", "13.0"]
 
     name: test-executorch-cuda-build-${{ matrix.cuda-version }}
     uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main

backends/cadence/fusion_g3/operators/op_add.cpp

@@ -162,7 +162,7 @@ Tensor& add_out(
     float alpha_val;
     torch::executor::native::utils::extract_scalar(alpha, &alpha_val);
 
-    if ((a.numel() == 1) && (alpha_val == 1.0)) {
+    if ((a.numel() == 1) && (alpha_val == 1.0f)) {
       XT_KERNEL_CHECK(
           ctx,
           out,

backends/cadence/hifi/kernels/kernels.cpp

@@ -39,8 +39,8 @@ void* allocate_temp_memory(KernelRuntimeContext& ctx, size_t size) {
 template <typename T>
 __attribute__((always_inline)) T
 quantize(const float x, float scale, int32_t zero_point) {
-  constexpr float min_val = std::numeric_limits<T>::min();
-  constexpr float max_val = std::numeric_limits<T>::max();
+  constexpr float min_val = static_cast<float>(std::numeric_limits<T>::min());
+  constexpr float max_val = static_cast<float>(std::numeric_limits<T>::max());
   float tmp = roundf(x * scale + zero_point);
   return std::max(std::min(tmp, max_val), min_val);
 }
@@ -56,8 +56,8 @@ void quantize(
   xtfloatx2 scale_vec = (xtfloatx2)scale;
   xtfloatx2 zero_vec = XT_FLOAT_SX2(zero_point, 0);
 
-  constexpr float min_val = std::numeric_limits<T>::min();
-  constexpr float max_val = std::numeric_limits<T>::max();
+  constexpr float min_val = static_cast<float>(std::numeric_limits<T>::min());
+  constexpr float max_val = static_cast<float>(std::numeric_limits<T>::max());
 
   const xtfloatx2* __restrict__ p0 = (const xtfloatx2* __restrict__)x;
   ae_valign va0 = XT_LASX2PP(p0);

backends/cadence/hifi/operators/op_quantized_relu_out.cpp

@@ -6,17 +6,18 @@
  * LICENSE file in the root directory of this source tree.
  */
 
+#include <executorch/backends/cadence/common/xt_macros.h>
 #include <executorch/backends/cadence/hifi/kernels/kernels.h>
 #include <executorch/runtime/kernel/kernel_includes.h>
 
-using executorch::aten::ScalarType;
-using executorch::aten::Tensor;
-using torch::executor::KernelRuntimeContext;
-
 namespace impl {
 namespace HiFi {
 namespace native {
 
+using ::executorch::aten::ScalarType;
+using ::executorch::aten::Tensor;
+using ::executorch::runtime::KernelRuntimeContext;
+
 void quantized_relu_per_tensor_out(
     KernelRuntimeContext& ctx,
     const Tensor& input,
@@ -34,7 +35,10 @@ void quantized_relu_per_tensor_out(
     const uint8_t* p_in = input.const_data_ptr<uint8_t>();
     uint8_t* p_out = output.mutable_data_ptr<uint8_t>();
 
-    WORD32 ret_val = xa_nn_vec_relu_asym8u_asym8u(
+    XT_KERNEL_CHECK(
+        ctx,
+        ,
+        xa_nn_vec_relu_asym8u_asym8u,
         p_out,
         p_in,
         _in_zero_point,
@@ -45,15 +49,16 @@ void quantized_relu_per_tensor_out(
         255,
         input.numel());
 
-    ET_CHECK_MSG(ret_val == 0, "An internal error occured");
-
   } else if (input.scalar_type() == executorch::aten::ScalarType::Char) {
-    const int8_t _in_zero_point = static_cast<int8_t>(in_zero_point);
-    const int8_t _out_zero_point = static_cast<int8_t>(out_zero_point);
+    const int _in_zero_point = static_cast<int>(in_zero_point);
+    const int _out_zero_point = static_cast<int>(out_zero_point);
     const int8_t* p_in = input.const_data_ptr<int8_t>();
     int8_t* p_out = output.mutable_data_ptr<int8_t>();
 
-    WORD32 ret_val = xa_nn_vec_relu_asym8s_asym8s(
+    XT_KERNEL_CHECK(
+        ctx,
+        ,
+        xa_nn_vec_relu_asym8s_asym8s,
         p_out,
         p_in,
         _in_zero_point,
@@ -64,8 +69,6 @@ void quantized_relu_per_tensor_out(
         127,
         input.numel());
 
-    ET_CHECK_MSG(ret_val == 0, "An internal error occured");
-
   } else {
     ET_CHECK_MSG(
         false,

backends/cadence/hifi/operators/tests/test_op_quantized_relu_out.cpp

@@ -57,14 +57,14 @@ class HiFiQuantizedReluTest : public OperatorTest {
 
 TEST_F(HiFiQuantizedReluTest, MultiDimensionalTest) {
   TensorFactory<ScalarType::Char> tf_chars;
+  TensorFactory<ScalarType::Int> tf_ints;
   const std::vector<int32_t> sizes{2, 3, 5, 6};
   Tensor quantized_input = tf_chars.full(sizes, -128);
   Tensor quantized_output = tf_chars.full(sizes, 100);
   Tensor in_zero_point = tf_chars.full({1}, 127);
   int64_t out_zero_point = -128;
-  Tensor out_multiplier =
-      TensorFactory<ScalarType::Int>().full({1}, 1077952640);
-  Tensor out_shift = TensorFactory<ScalarType::Int>().full({1}, 5);
+  Tensor out_multiplier = tf_ints.full({1}, 1077952640);
+  Tensor out_shift = tf_ints.full({1}, 5);
 
   quantized_relu_out(
       quantized_input,
@@ -80,14 +80,14 @@ TEST_F(HiFiQuantizedReluTest, MultiDimensionalTest) {
 
 TEST_F(HiFiQuantizedReluTest, OneDimensionalTest) {
   TensorFactory<ScalarType::Char> tf_chars;
+  TensorFactory<ScalarType::Int> tf_ints;
   const std::vector<int32_t> sizes{56};
   Tensor quantized_input = tf_chars.full(sizes, -128);
   Tensor quantized_output = tf_chars.full(sizes, 100);
   Tensor in_zero_point = tf_chars.full({1}, 127);
   int64_t out_zero_point = -128;
-  Tensor out_multiplier =
-      TensorFactory<ScalarType::Int>().full({1}, 1077952640);
-  Tensor out_shift = TensorFactory<ScalarType::Int>().full({1}, 5);
+  Tensor out_multiplier = tf_ints.full({1}, 1077952640);
+  Tensor out_shift = tf_ints.full({1}, 5);
 
   quantized_relu_out(
       quantized_input,

backends/cadence/hifi/third-party/nnlib/targets.bzl

@@ -13,6 +13,10 @@ def define_common_targets():
             "@EXECUTORCH_CLIENTS",
         ],
         compatible_with = ["ovr_config//cpu:xtensa"],
+        compiler_flags = [
+            "-Wno-pointer-sign",
+            "-Wno-incompatible-pointer-types-discards-qualifiers",
+        ],
         deps = [
             "fbsource//third-party/nnlib-hifi4/xa_nnlib:libxa_nnlib",
         ],

backends/cadence/hifi/third-party/nnlib/xa_nn_elm_atan2_f32.c

@@ -21,7 +21,7 @@
 ******************************************************************************/
 #include <float.h>
 
-#include "../include/NatureDSP_Signal_math.h"
+#include "NatureDSP_Signal_math.h"
 #include "NatureDSP_types.h"
 #include "xa_nn_common.h"
 

backends/cadence/hifi/third-party/nnlib/xa_nn_elm_pow_f32.c

@@ -20,7 +20,7 @@
 
 ******************************************************************************/
 
-#include "../include/NatureDSP_Signal_math.h"
+#include "NatureDSP_Signal_math.h"
 #include "NatureDSP_types.h"
 #include "xa_nn_common.h"
 

backends/cadence/hifi/third-party/nnlib/xa_nn_elm_where_f32xf32_f32.c

@@ -117,6 +117,7 @@ WORD32 xa_nn_elm_where_f32xf32_f32(FLOAT32 * __restrict__ p_out,
         XT_MOVF_S(a, a2, s);
         XT_SSI(a, (xtfloat *)out, 0);
     }
+    return 0;
 }
 
 static void internal_elm_where_broadcast_f32xf32_f32(FLOAT32 * __restrict__ p_out,