Linux Aarch64 support

.github/workflows/ci.yaml

@@ -3,19 +3,50 @@ name: CI
 on:
   push:
     branches:
-    - master
+      - master
   pull_request:
+    branches:
+      - master
 
 jobs:
   build:
+    name: Linux x86_64
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout
+      uses: actions/checkout@v4
+
+    - name: Build
+      run:  make -j
+
+    - name: Check
+      run: |
+        file hifiasm | grep x86-64
+        ./hifiasm
+
+  build-aarch64:
+    name: Linux aarch64
     runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        compiler: [gcc]
 
     steps:
-    - name: Checkout minimap2
-      uses: actions/checkout@v2
+    - name: Checkout
+      uses: actions/checkout@v4
 
-    - name: Compile with ${{ matrix.compiler }}
-      run:  make CC=${{ matrix.compiler }}
+    - name: Build
+      uses: uraimo/run-on-arch-action@v2
+      with:
+        arch: aarch64
+        distro: ubuntu20.04
+        githubToken: ${{ github.token }}
+        dockerRunArgs: |
+          --volume "${PWD}:/hifiasm"
+        install: |
+          apt-get update -q -y
+          apt-get install -q -y make gcc-10 g++-10 zlib1g-dev file
+        run: |
+          cd /hifiasm
+          make CC=gcc-10 CXX=g++-10 ARCH_FLAGS="-march=native" -j
+          file hifiasm | grep aarch64
+          ./hifiasm
+

Levenshtein_distance.h

@@ -3,10 +3,14 @@
 
 #define __STDC_LIMIT_MACROS
 #include <stdint.h>
+#ifdef __ARM_NEON
+#include "sse2neon.h"
+#else
 #include "emmintrin.h"
 #include "nmmintrin.h"
 #include "smmintrin.h"
 #include <immintrin.h>
+#endif // ~__ARM_NEON
 #include <string.h>
 #include <stdlib.h>
 #include <stdio.h>

Makefile

@@ -1,6 +1,7 @@
 CXX=		g++
 CC=			gcc
-CXXFLAGS=	-g -O3 -msse4.2 -mpopcnt -fomit-frame-pointer -Wall
+ARCH_FLAGS=-msse4.2 -mpopcnt
+CXXFLAGS=	-g -O3 $(ARCH_FLAGS) -fomit-frame-pointer -Wall
 CFLAGS=		$(CXXFLAGS)
 CPPFLAGS=
 INCLUDES=

ksw2_extz2_sse.c

@@ -3,7 +3,11 @@
 #include "ksw2.h"
 
 #ifdef __SSE2__
+#ifdef __ARM_NEON
+#include "sse2neon.h"
+#else
 #include <emmintrin.h>
+#endif // ~__ARM_NEON
 
 #ifdef KSW_SSE2_ONLY
 #undef __SSE4_1__