Added the SVE-optimized implementation for copyBits

bolt/common/base/BitUtil.h

@@ -38,6 +38,10 @@
 #include <cstring>
 #include <string>
 
+#if defined(__ARM_FEATURE_SVE) && defined(__aarch64__)
+#include <arm_sve.h>
+#endif
+
 #ifdef __BMI2__
 #include <x86intrin.h>
 #endif
@@ -933,6 +937,32 @@ inline void copyBits(
     uint64_t targetOffset,
     uint64_t numBits) {
   uint64_t i = 0;
+#if defined(__ARM_FEATURE_SVE) && defined(__aarch64__) && SVE_BITS == 256
+  if ((sourceOffset & 7) == (targetOffset & 7)) {
+    const uint8_t* constSrcBase =
+        reinterpret_cast<const uint8_t*>(source) + (sourceOffset >> 3);
+    uint8_t* srcBase = const_cast<uint8_t*>(constSrcBase);
+    uint8_t* dstBase = reinterpret_cast<uint8_t*>(target) + (targetOffset >> 3);
+    uint64_t bitStartOffset = 0;
+    if ((sourceOffset & 7) != 0) {
+      bitStartOffset = std::min<uint64_t>(8 - (sourceOffset & 7), numBits);
+      auto headBits =
+          detail::loadBits<uint8_t>(source, sourceOffset, bitStartOffset);
+      detail::storeBits<uint8_t>(
+          target, targetOffset, headBits, bitStartOffset);
+      i += bitStartOffset;
+      srcBase++;
+      dstBase++;
+    }
+    svbool_t pg = svptrue_b8();
+    while (i + SVE_BITS <= numBits) {
+      svuint8_t curBitsCopy =
+          svld1_u8(pg, srcBase + ((i - bitStartOffset) >> 3));
+      svst1_u8(pg, dstBase + ((i - bitStartOffset) >> 3), curBitsCopy);
+      i += SVE_BITS;
+    }
+  }
+#endif
   for (; i + 64 <= numBits; i += 64) {
     uint64_t word = detail::loadBits<uint64_t>(source, i + sourceOffset, 64);
     detail::storeBits<uint64_t>(target, targetOffset + i, word, 64);

conanfile.py

@@ -18,6 +18,7 @@
 import pydot
 import platform
 import re
+import subprocess
 from conan import ConanFile
 from conan.tools import files, scm
 from conan.tools.cmake import CMake, CMakeDeps, CMakeToolchain, cmake_layout
@@ -505,14 +506,27 @@ def generate(self):
             tc.cache_variables["CMAKE_CXX_FLAGS"] = flags
             tc.cache_variables["CMAKE_C_FLAGS"] = flags
 
+        #Check SVE
+        sve_supported = False
+        result = subprocess.run(['lscpu'], capture_output=True, text=True, timeout=3)
+        if result.returncode == 0 and 'sve' in result.stdout.lower():
+            sve_supported = True
+
         if str(self.settings.arch) in ["armv8", "arm"]:
-            # Support CRC & NEON on ARMv8
-            flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv8.3-a"
+            if sve_supported:
+                # Support CRC & NEON & SVE on ARMv8
+                flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv8.3-a+sve -msve-vector-bits=256 -DSVE_BITS=256"
+            else:
+                # Support CRC & NEON on ARMv8
+                flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv8.3-a"
             tc.cache_variables["CMAKE_CXX_FLAGS"] = flags
             tc.cache_variables["CMAKE_C_FLAGS"] = flags
         elif str(self.settings.arch) in ["armv9"]:
             # gcc 12+ https://www.phoronix.com/news/GCC-12-ARMv9-march-armv9-a
-            flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv9-a"
+            if sve_supported:
+                flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv9-a+sve -msve-vector-bits=256 -DSVE_BITS=256"
+            else:
+                flags = f"{self.BOLT_GLOABL_FLAGS} -march=armv9-a"
             tc.variables["CMAKE_C_FLAGS"] = flags
             tc.variables["CMAKE_CXX_FLAGS"] = flags
         if (