Merge branch 'master' into fix-simd-signmask

wasm: replace usage of -0.f to 0x80000000

.github/workflows/cmake-wasm.yml

@@ -0,0 +1,97 @@
+name: CMake WebAssembly
+
+on:
+  push:
+    branches: [ "master" ]
+  pull_request:
+    branches: [ "master" ]
+
+env:
+  wasmtime_version: v7.0.0
+  wasmer_version: v3.1.1
+
+jobs:
+  build_wasi_sdk:
+    strategy:
+      matrix:
+        BUILD_TYPE: [Release, Debug, RelWithDebInfo, MinSizeRel]
+        C_FLAGS: ['', '-msimd128']
+        wasi_sdk_version: [19, 20]
+    # The CMake configure and build commands are platform agnostic and should work equally well on Windows or Mac.
+    # You can convert this to a matrix build if you need cross-platform coverage.
+    # See: https://docs.github.com/en/free-pro-team@latest/actions/learn-github-actions/managing-complex-workflows#using-a-build-matrix
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v3
+
+    - name: Downloading wasi-sdk
+      run: |
+        cd ${{github.workspace}}
+        wget --no-verbose https://github.com/WebAssembly/wasi-sdk/releases/download/wasi-sdk-${{matrix.wasi_sdk_version}}/wasi-sdk-${{matrix.wasi_sdk_version}}.0-linux.tar.gz
+        tar xf wasi-sdk-${{matrix.wasi_sdk_version}}.0-linux.tar.gz
+
+    - name: Configure CMake
+      # Configure CMake in a 'build' subdirectory. `CMAKE_BUILD_TYPE` is only required if you are using a single-configuration generator such as make.
+      # See https://cmake.org/cmake/help/latest/variable/CMAKE_BUILD_TYPE.html?highlight=cmake_build_type
+      run: cmake -B ${{github.workspace}}/build -DCMAKE_BUILD_TYPE=${{matrix.BUILD_TYPE}} -DCMAKE_C_FLAGS="${{matrix.C_FLAGS}}" -DCMAKE_TOOLCHAIN_FILE=${{github.workspace}}/wasi-sdk-${{matrix.wasi_sdk_version}}.0/share/cmake/wasi-sdk.cmake -DWASI_SDK_PREFIX=${{github.workspace}}/wasi-sdk-${{matrix.wasi_sdk_version}}.0 -DCGLM_USE_TEST=ON
+
+    - name: Build
+      # Build your program with the given configuration
+      run: cmake --build ${{github.workspace}}/build --config ${{matrix.BUILD_TYPE}}
+
+    - name: Test with wasmtime
+      run: |
+        cd ${{github.workspace}}
+        ls -lh ${{github.workspace}}/build/
+        wget --no-verbose https://github.com/bytecodealliance/wasmtime/releases/download/${{env.wasmtime_version}}/wasmtime-${{env.wasmtime_version}}-x86_64-linux.tar.xz
+        tar xf wasmtime-${{env.wasmtime_version}}-x86_64-linux.tar.xz
+        ./wasmtime-${{env.wasmtime_version}}-x86_64-linux/wasmtime run --wasm-features simd ${{github.workspace}}/build/tests
+
+    - name: Test with wasmer
+      run: |
+        cd ${{github.workspace}}
+        mkdir wasmer
+        cd wasmer
+        wget --no-verbose https://github.com/wasmerio/wasmer/releases/download/${{env.wasmer_version}}/wasmer-linux-amd64.tar.gz
+        tar xf wasmer-linux-amd64.tar.gz
+        ./bin/wasmer run --enable-simd ${{github.workspace}}/build/tests
+
+  build_emsdk:
+    strategy:
+      matrix:
+        BUILD_TYPE: [Release, Debug, RelWithDebInfo, MinSizeRel]
+        C_FLAGS: ['', '-msimd128', '-msse -msse2 -msimd128', '-msse -msse2 -msse3 -msse4 -msimd128']
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v3
+    - name: Setup emsdk
+      uses: mymindstorm/setup-emsdk@v11
+
+    - name: Verify emsdk
+      run: emcc -v
+
+    - name: Configure CMake
+      # Configure CMake in a 'build' subdirectory. `CMAKE_BUILD_TYPE` is only required if you are using a single-configuration generator such as make.
+      # See https://cmake.org/cmake/help/latest/variable/CMAKE_BUILD_TYPE.html?highlight=cmake_build_type
+      run: emcmake cmake -B ${{github.workspace}}/build -DCMAKE_BUILD_TYPE=${{matrix.BUILD_TYPE}} -DCMAKE_C_FLAGS="${{matrix.C_FLAGS}}" -DCMAKE_EXE_LINKER_FLAGS="-s STANDALONE_WASM" -DCGLM_STATIC=ON -DCGLM_USE_TEST=ON
+
+    - name: Build
+      # Build your program with the given configuration
+      run: cmake --build ${{github.workspace}}/build --config ${{matrix.BUILD_TYPE}}
+
+    - name: Test with wasmtime
+      run: |
+        cd ${{github.workspace}}
+        ls -lh ${{github.workspace}}/build/
+        wget --no-verbose https://github.com/bytecodealliance/wasmtime/releases/download/${{env.wasmtime_version}}/wasmtime-${{env.wasmtime_version}}-x86_64-linux.tar.xz
+        tar xf wasmtime-${{env.wasmtime_version}}-x86_64-linux.tar.xz
+        ./wasmtime-${{env.wasmtime_version}}-x86_64-linux/wasmtime run --wasm-features simd ${{github.workspace}}/build/tests.wasm
+    - name: Test with wasmer
+      run: |
+        cd ${{github.workspace}}
+        mkdir wasmer
+        cd wasmer
+        wget --no-verbose https://github.com/wasmerio/wasmer/releases/download/${{env.wasmer_version}}/wasmer-linux-amd64.tar.gz
+        tar xf wasmer-linux-amd64.tar.gz
+        ./bin/wasmer run --enable-simd ${{github.workspace}}/build/tests.wasm

CMakeLists.txt

@@ -16,6 +16,11 @@ option(CGLM_STATIC "Static build" OFF)
 option(CGLM_USE_C99 "" OFF)
 option(CGLM_USE_TEST "Enable Tests" OFF)
 
+if(CMAKE_SYSTEM_NAME STREQUAL WASI)
+  set(CGLM_STATIC ON CACHE BOOL "Static option" FORCE)
+  set(CGLM_SHARED OFF CACHE BOOL "Shared option" FORCE)
+endif()
+
 if(NOT CGLM_STATIC AND CGLM_SHARED)
   set(CGLM_BUILD SHARED)
 else(CGLM_STATIC)

Makefile.am

@@ -243,11 +243,6 @@ test_tests_SOURCES=\
     test/runner.c \
     test/src/test_common.c \
     test/src/tests.c \
-    test/src/test_cam.c \
-    test/src/test_cam_lh_zo.c \
-    test/src/test_cam_rh_zo.c \
-    test/src/test_cam_lh_no.c \
-    test/src/test_cam_rh_no.c \
     test/src/test_clamp.c \
     test/src/test_euler.c \
     test/src/test_bezier.c \

include/cglm/affine-mat.h

@@ -30,6 +30,10 @@
 #  include "simd/neon/affine.h"
 #endif
 
+#ifdef CGLM_SIMD_WASM
+#  include "simd/wasm/affine.h"
+#endif
+
 /*!
  * @brief this is similar to glm_mat4_mul but specialized to affine transform
  *
@@ -49,7 +53,9 @@
 CGLM_INLINE
 void
 glm_mul(mat4 m1, mat4 m2, mat4 dest) {
-#ifdef __AVX__
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mul_wasm(m1, m2, dest);
+#elif defined(__AVX__)
   glm_mul_avx(m1, m2, dest);
 #elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mul_sse2(m1, m2, dest);
@@ -107,7 +113,9 @@ glm_mul(mat4 m1, mat4 m2, mat4 dest) {
 CGLM_INLINE
 void
 glm_mul_rot(mat4 m1, mat4 m2, mat4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mul_rot_wasm(m1, m2, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mul_rot_sse2(m1, m2, dest);
 #elif defined(CGLM_NEON_FP)
   glm_mul_rot_neon(m1, m2, dest);
@@ -156,7 +164,9 @@ glm_mul_rot(mat4 m1, mat4 m2, mat4 dest) {
 CGLM_INLINE
 void
 glm_inv_tr(mat4 mat) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_inv_tr_wasm(mat);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_inv_tr_sse2(mat);
 #elif defined(CGLM_NEON_FP)
   glm_inv_tr_neon(mat);

include/cglm/io.h

@@ -68,7 +68,10 @@
 CGLM_INLINE
 void
 glm_arch_print_name(FILE* __restrict ostream) {
-#ifdef CGLM_SIMD_x86
+#if defined(CGLM_SIMD_WASM)
+  fprintf(ostream, CGLM_PRINT_COLOR "\ncglm arch: wasm SIMD128"
+    "\n\n" CGLM_PRINT_COLOR_RESET);
+#elif defined(CGLM_SIMD_x86)
   fprintf(ostream, CGLM_PRINT_COLOR "\ncglm arch: x86 SSE*" 
 #ifdef __AVX__
     " AVX"

include/cglm/mat2.h

@@ -44,6 +44,10 @@
 #  include "simd/neon/mat2.h"
 #endif
 
+#ifdef CGLM_SIMD_WASM
+#  include "simd/wasm/mat2.h"
+#endif
+
 #define GLM_MAT2_IDENTITY_INIT  {{1.0f, 0.0f}, {0.0f, 1.0f}}
 #define GLM_MAT2_ZERO_INIT      {{0.0f, 0.0f}, {0.0f, 0.0f}}
 
@@ -132,7 +136,9 @@ glm_mat2_zero(mat2 mat) {
 CGLM_INLINE
 void
 glm_mat2_mul(mat2 m1, mat2 m2, mat2 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat2_mul_wasm(m1, m2, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat2_mul_sse2(m1, m2, dest);
 #elif defined(CGLM_NEON_FP)
   glm_mat2_mul_neon(m1, m2, dest);
@@ -160,7 +166,9 @@ glm_mat2_mul(mat2 m1, mat2 m2, mat2 dest) {
 CGLM_INLINE
 void
 glm_mat2_transpose_to(mat2 m, mat2 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat2_transp_wasm(m, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat2_transp_sse2(m, dest);
 #else
   dest[0][0] = m[0][0];
@@ -222,7 +230,10 @@ glm_mat2_trace(mat2 m) {
 CGLM_INLINE
 void
 glm_mat2_scale(mat2 m, float s) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(m[0], wasm_f32x4_mul(wasm_v128_load(m[0]),
+                                  wasm_f32x4_splat(s)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(m[0], _mm_mul_ps(_mm_loadu_ps(m[0]), _mm_set1_ps(s)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(m[0], vmulq_f32(vld1q_f32(m[0]), vdupq_n_f32(s)));

include/cglm/mat3.h

@@ -42,6 +42,10 @@
 #  include "simd/sse2/mat3.h"
 #endif
 
+#ifdef CGLM_SIMD_WASM
+#  include "simd/wasm/mat3.h"
+#endif
+
 #define GLM_MAT3_IDENTITY_INIT  {{1.0f, 0.0f, 0.0f},                          \
                                  {0.0f, 1.0f, 0.0f},                          \
                                  {0.0f, 0.0f, 1.0f}}
@@ -148,7 +152,9 @@ glm_mat3_zero(mat3 mat) {
 CGLM_INLINE
 void
 glm_mat3_mul(mat3 m1, mat3 m2, mat3 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat3_mul_wasm(m1, m2, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat3_mul_sse2(m1, m2, dest);
 #else
   float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2],

include/cglm/mat4.h

@@ -64,6 +64,10 @@
 #  include "simd/neon/mat4.h"
 #endif
 
+#ifdef CGLM_SIMD_WASM
+#  include "simd/wasm/mat4.h"
+#endif
+
 #ifdef DEBUG
 # include <assert.h>
 #endif
@@ -121,7 +125,12 @@ glm_mat4_ucopy(mat4 mat, mat4 dest) {
 CGLM_INLINE
 void
 glm_mat4_copy(mat4 mat, mat4 dest) {
-#ifdef __AVX__
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest[0], glmm_load(mat[0]));
+  glmm_store(dest[1], glmm_load(mat[1]));
+  glmm_store(dest[2], glmm_load(mat[2]));
+  glmm_store(dest[3], glmm_load(mat[3]));
+#elif defined(__AVX__)
   glmm_store256(dest[0], glmm_load256(mat[0]));
   glmm_store256(dest[2], glmm_load256(mat[2]));
 #elif defined( __SSE__ ) || defined( __SSE2__ )
@@ -187,7 +196,14 @@ glm_mat4_identity_array(mat4 * __restrict mat, size_t count) {
 CGLM_INLINE
 void
 glm_mat4_zero(mat4 mat) {
-#ifdef __AVX__
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_128 x0;
+  x0 = wasm_f32x4_const_splat(0.f);
+  glmm_store(mat[0], x0);
+  glmm_store(mat[1], x0);
+  glmm_store(mat[2], x0);
+  glmm_store(mat[3], x0);
+#elif defined(__AVX__)
   __m256 y0;
   y0 = _mm256_setzero_ps();
   glmm_store256(mat[0], y0);
@@ -297,7 +313,9 @@ glm_mat4_ins3(mat3 mat, mat4 dest) {
 CGLM_INLINE
 void
 glm_mat4_mul(mat4 m1, mat4 m2, mat4 dest) {
-#ifdef __AVX__
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_mul_wasm(m1, m2, dest);
+#elif defined(__AVX__)
   glm_mat4_mul_avx(m1, m2, dest);
 #elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_mul_sse2(m1, m2, dest);
@@ -377,7 +395,9 @@ glm_mat4_mulN(mat4 * __restrict matrices[], uint32_t len, mat4 dest) {
 CGLM_INLINE
 void
 glm_mat4_mulv(mat4 m, vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_mulv_wasm(m, v, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_mulv_sse2(m, v, dest);
 #elif defined(CGLM_NEON_FP)
   glm_mat4_mulv_neon(m, v, dest);
@@ -497,7 +517,9 @@ glm_mat4_mulv3(mat4 m, vec3 v, float last, vec3 dest) {
 CGLM_INLINE
 void
 glm_mat4_transpose_to(mat4 m, mat4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_transp_wasm(m, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_transp_sse2(m, dest);
 #elif defined(CGLM_NEON_FP)
   glm_mat4_transp_neon(m, dest);
@@ -521,7 +543,9 @@ glm_mat4_transpose_to(mat4 m, mat4 dest) {
 CGLM_INLINE
 void
 glm_mat4_transpose(mat4 m) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_transp_wasm(m, m);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_transp_sse2(m, m);
 #elif defined(CGLM_NEON_FP)
   glm_mat4_transp_neon(m, m);
@@ -560,7 +584,9 @@ glm_mat4_scale_p(mat4 m, float s) {
 CGLM_INLINE
 void
 glm_mat4_scale(mat4 m, float s) {
-#ifdef __AVX__
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_scale_wasm(m, s);
+#elif defined(__AVX__)
   glm_mat4_scale_avx(m, s);
 #elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_scale_sse2(m, s);
@@ -581,7 +607,9 @@ glm_mat4_scale(mat4 m, float s) {
 CGLM_INLINE
 float
 glm_mat4_det(mat4 mat) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  return glm_mat4_det_wasm(mat);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   return glm_mat4_det_sse2(mat);
 #elif defined(CGLM_NEON_FP)
   return glm_mat4_det_neon(mat);
@@ -679,7 +707,9 @@ glm_mat4_inv(mat4 mat, mat4 dest) {
 CGLM_INLINE
 void
 glm_mat4_inv_fast(mat4 mat, mat4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_mat4_inv_fast_wasm(mat, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_mat4_inv_fast_sse2(mat, dest);
 #else
   glm_mat4_inv(mat, dest);

include/cglm/quat.h

@@ -70,6 +70,10 @@
 #  include "simd/neon/quat.h"
 #endif
 
+#ifdef CGLM_SIMD_WASM
+#  include "simd/wasm/quat.h"
+#endif
+
 CGLM_INLINE void glm_quat_normalize(versor q);
 
 /*
@@ -238,7 +242,22 @@ glm_quat_norm(versor q) {
 CGLM_INLINE
 void
 glm_quat_normalize_to(versor q, versor dest) {
-#if defined( __SSE2__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_128 xdot, x0;
+  float  dot;
+
+  x0   = glmm_load(q);
+  xdot = glmm_vdot(x0, x0);
+  /* dot  = _mm_cvtss_f32(xdot); */
+  dot  = wasm_f32x4_extract_lane(xdot, 0);
+
+  if (dot <= 0.0f) {
+    glm_quat_identity(dest);
+    return;
+  }
+
+  glmm_store(dest, wasm_f32x4_div(x0, wasm_f32x4_sqrt(xdot)));
+#elif defined( __SSE2__ ) || defined( __SSE2__ )
   __m128 xdot, x0;
   float  dot;
 
@@ -438,7 +457,9 @@ glm_quat_mul(versor p, versor q, versor dest) {
     + (a1 d2 + b1 c2 − c1 b2 + d1 a2)k
        a1 a2 − b1 b2 − c1 c2 − d1 d2
    */
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glm_quat_mul_wasm(p, q, dest);
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glm_quat_mul_sse2(p, q, dest);
 #elif defined(CGLM_NEON_FP)
   glm_quat_mul_neon(p, q, dest);

include/cglm/simd/arm.h

@@ -52,6 +52,10 @@ glmm_float32x4_init(float x, float y, float z, float w) {
 #  define glmm_float32x4_init(x, y, z, w) { x, y, z, w }
 #endif
 
+#define glmm_float32x4_SIGNMASK_PNPN glmm_float32x4_init( 0.f, -0.f,  0.f, -0.f)
+#define glmm_float32x4_SIGNMASK_NPNP glmm_float32x4_init(-0.f,  0.f, -0.f,  0.f)
+#define glmm_float32x4_SIGNMASK_NPPN glmm_float32x4_init(-0.f,  0.f,  0.f, -0.f)
+
 static inline
 float32x4_t
 glmm_abs(float32x4_t v) {

include/cglm/simd/intrin.h

@@ -99,13 +99,20 @@
 #  endif
 #endif
 
-#if defined(CGLM_SIMD_x86) || defined(CGLM_SIMD_ARM)
+/* WebAssembly */
+#if defined(__wasm__) && defined(__wasm_simd128__)
+#  ifndef CGLM_SIMD_WASM
+#    define CGLM_SIMD_WASM
+#  endif
+#endif
+
+#if defined(CGLM_SIMD_x86) || defined(CGLM_SIMD_ARM) || defined(CGLM_SIMD_WASM)
 #  ifndef CGLM_SIMD
 #    define CGLM_SIMD
 #  endif
 #endif
 
-#if defined(CGLM_SIMD_x86)
+#if defined(CGLM_SIMD_x86) && !defined(CGLM_SIMD_WASM)
 #  include "x86.h"
 #endif
 
@@ -113,4 +120,8 @@
 #  include "arm.h"
 #endif
 
+#if defined(CGLM_SIMD_WASM)
+#  include "wasm.h"
+#endif
+
 #endif /* cglm_intrin_h */

include/cglm/simd/neon/mat4.h

@@ -108,7 +108,7 @@ glm_mat4_det_neon(mat4 mat) {
   float32x4_t   r0, r1, r2, r3, x0, x1, x2;
   float32x2_t   ij, op, mn, kl, nn, mm, jj, ii, gh, ef, t12, t34;
   float32x4x2_t a1;
-  float32x4_t   x3 = glmm_float32x4_init(0.f, -0.f, 0.f, -0.f);
+  float32x4_t   x3 = glmm_float32x4_SIGNMASK_PNPN;
 
   /* 127 <- 0, [square] det(A) = det(At) */
   r0 = glmm_load(mat[0]);              /* d c b a */
@@ -181,7 +181,7 @@ glm_mat4_inv_neon(mat4 mat, mat4 dest) {
                 x0, x1, x2, x3, x4, x5, x6, x7, x8;
   float32x4x2_t a1;
   float32x2_t   lp, ko, hg, jn, im, fe, ae, bf, cg, dh;
-  float32x4_t   x9 = glmm_float32x4_init(-0.f, 0.f, -0.f, 0.f);
+  float32x4_t   x9 = glmm_float32x4_SIGNMASK_NPNP;
 
   x8 = vrev64q_f32(x9);
 

include/cglm/simd/neon/quat.h

@@ -23,7 +23,8 @@ glm_quat_mul_neon(versor p, versor q, versor dest) {
    */
 
   glmm_128 xp, xq, xqr, r, x, y, z, s2, s3;
-  glmm_128 s1 = glmm_float32x4_init(-0.f, 0.f, 0.f, -0.f);
+  glmm_128 s1 = glmm_float32x4_SIGNMASK_NPPN;
+
   float32x2_t   qh, ql;
   
   xp  = glmm_load(p); /* 3 2 1 0 */

include/cglm/simd/sse2/affine.h

@@ -98,7 +98,7 @@ glm_inv_tr_sse2(mat4 mat) {
   x2 = glmm_shuff1(r3, 0, 0, 0, 0);
   x3 = glmm_shuff1(r3, 1, 1, 1, 1);
   x4 = glmm_shuff1(r3, 2, 2, 2, 2);
-  x5 = _mm_set1_ps(-0.f);
+  x5 = glmm_float32x4_SIGNMASK_NEG;
 
   x0 = glmm_fmadd(r0, x2, glmm_fmadd(r1, x3, _mm_mul_ps(r2, x4)));
   x0 = _mm_xor_ps(x0, x5);

include/cglm/simd/sse2/mat4.h

@@ -153,7 +153,7 @@ glm_mat4_det_sse2(mat4 mat) {
                   _mm_shuffle_ps(x0, x1, _MM_SHUFFLE(2, 2, 3, 1)),
                   x2);
   
-  x2 = _mm_xor_ps(x2, _mm_set_ps(-0.f, 0.f, -0.f, 0.f));
+  x2 = _mm_xor_ps(x2, glmm_float32x4_SIGNMASK_NPNP);
   
   return glmm_hadd(_mm_mul_ps(x2, r0));
 }
@@ -166,7 +166,8 @@ glm_mat4_inv_fast_sse2(mat4 mat, mat4 dest) {
          t0, t1, t2, t3, t4, t5,
          x0, x1, x2, x3, x4, x5, x6, x7, x8, x9;
 
-  x8 = _mm_set_ps(-0.f, 0.f, -0.f, 0.f);
+  /* x8 = _mm_set_ps(-0.f, 0.f, -0.f, 0.f); */
+  x8 = glmm_float32x4_SIGNMASK_NPNP;
   x9 = glmm_shuff1(x8, 2, 1, 2, 1);
 
   /* 127 <- 0 */
@@ -302,7 +303,8 @@ glm_mat4_inv_sse2(mat4 mat, mat4 dest) {
          t0, t1, t2, t3, t4, t5,
          x0, x1, x2, x3, x4, x5, x6, x7, x8, x9;
 
-  x8 = _mm_set_ps(-0.f, 0.f, -0.f, 0.f);
+  /* x8 = _mm_set_ps(-0.f, 0.f, -0.f, 0.f); */
+  x8 = glmm_float32x4_SIGNMASK_NPNP;
   x9 = glmm_shuff1(x8, 2, 1, 2, 1);
 
   /* 127 <- 0 */

include/cglm/simd/sse2/quat.h

@@ -26,7 +26,7 @@ glm_quat_mul_sse2(versor p, versor q, versor dest) {
 
   xp = glmm_load(p); /* 3 2 1 0 */
   xq = glmm_load(q);
-  x1 = _mm_set_ps(-0.f, 0.f, -0.f, 0.f); /* TODO: _mm_set1_ss() + shuff ? */
+  x1 = glmm_float32x4_SIGNMASK_NPNP; /* TODO: _mm_set1_ss() + shuff ? */
   r  = _mm_mul_ps(glmm_splat_w(xp), xq);
 
   x2 = _mm_unpackhi_ps(x1, x1);

include/cglm/simd/wasm.h

@@ -0,0 +1,198 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_simd_wasm_h
+#define cglm_simd_wasm_h
+#include "intrin.h"
+#ifdef CGLM_SIMD_WASM
+#include <wasm_simd128.h>
+
+#define glmm_load(p)      wasm_v128_load(p)
+#define glmm_store(p, a)  wasm_v128_store(p, (a))
+
+#define glmm_set1(x) wasm_f32x4_splat(x)
+#define glmm_128     v128_t
+
+#define glmm_shuff1(xmm, z, y, x, w) wasm_i32x4_shuffle(xmm, xmm, w, x, y, z)
+
+#define glmm_splat(x, lane) glmm_shuff1(x, lane, lane, lane, lane)
+
+#define glmm_splat_x(x) glmm_splat(x, 0)
+#define glmm_splat_y(x) glmm_splat(x, 1)
+#define glmm_splat_z(x) glmm_splat(x, 2)
+#define glmm_splat_w(x) glmm_splat(x, 3)
+
+#define GLMM_NEGZEROf 0x80000000 /*  0x80000000 ---> -0.0f  */
+
+/* _mm_set_ps(X, Y, Z, W); */
+#define GLMM__SIGNMASKf(X, Y, Z, W) wasm_i32x4_const(X, Y, Z, W)
+
+#define glmm_float32x4_SIGNMASK_PNPN GLMM__SIGNMASKf(0, GLMM_NEGZEROf, 0, GLMM_NEGZEROf)
+#define glmm_float32x4_SIGNMASK_NPNP GLMM__SIGNMASKf(GLMM_NEGZEROf, 0, GLMM_NEGZEROf, 0)
+#define glmm_float32x4_SIGNMASK_NPPN GLMM__SIGNMASKf(GLMM_NEGZEROf, 0, 0, GLMM_NEGZEROf)
+#define glmm_float32x4_SIGNMASK_NEG wasm_i32x4_const_splat(GLMM_NEGZEROf)
+
+static inline
+glmm_128
+glmm_abs(glmm_128 x) {
+  return wasm_f32x4_abs(x);
+}
+
+static inline
+glmm_128
+glmm_vhadd(glmm_128 v) {
+  glmm_128 x0;
+  x0 = wasm_f32x4_add(v,  glmm_shuff1(v, 0, 1, 2, 3));
+  x0 = wasm_f32x4_add(x0, glmm_shuff1(x0, 1, 0, 0, 1));
+  return x0;
+}
+
+static inline
+glmm_128
+glmm_vhadds(glmm_128 v) {
+  glmm_128 shuf, sums;
+  shuf = glmm_shuff1(v, 2, 3, 0, 1);
+  sums = wasm_f32x4_add(v, shuf);
+  /* shuf = _mm_movehl_ps(shuf, sums); */
+  shuf = wasm_i32x4_shuffle(shuf, sums, 6, 7, 2, 3);
+  sums = wasm_i32x4_shuffle(sums, wasm_f32x4_add(sums, shuf), 4, 1, 2, 3);
+  return sums;
+}
+
+static inline
+float
+glmm_hadd(glmm_128 v) {
+  return wasm_f32x4_extract_lane(glmm_vhadds(v), 0);
+}
+
+static inline
+glmm_128
+glmm_vhmin(glmm_128 v) {
+  glmm_128 x0, x1, x2;
+  x0 = glmm_shuff1(v, 2, 3, 2, 3);     /* [2, 3, 2, 3] */
+  x1 = wasm_f32x4_pmin(x0, v);   /* [0|2, 1|3, 2|2, 3|3] */
+  x2 = glmm_splat(x1, 1);              /* [1|3, 1|3, 1|3, 1|3] */
+  return wasm_f32x4_pmin(x1, x2);
+}
+
+static inline
+float
+glmm_hmin(glmm_128 v) {
+  return wasm_f32x4_extract_lane(glmm_vhmin(v), 0);
+}
+
+static inline
+glmm_128
+glmm_vhmax(glmm_128 v) {
+  glmm_128 x0, x1, x2;
+  x0 = glmm_shuff1(v, 2, 3, 2, 3);     /* [2, 3, 2, 3] */
+  x1 = wasm_f32x4_pmax(x0, v);   /* [0|2, 1|3, 2|2, 3|3] */
+  x2 = glmm_splat(x1, 1);              /* [1|3, 1|3, 1|3, 1|3] */
+  /* _mm_max_ss */
+  return wasm_i32x4_shuffle(x1, wasm_f32x4_pmax(x1, x2), 4, 1, 2, 3);
+}
+
+static inline
+float
+glmm_hmax(glmm_128 v) {
+  return wasm_f32x4_extract_lane(glmm_vhmax(v), 0);
+}
+
+static inline
+glmm_128
+glmm_vdots(glmm_128 a, glmm_128 b) {
+  return glmm_vhadds(wasm_f32x4_mul(a, b));
+}
+
+static inline
+glmm_128
+glmm_vdot(glmm_128 a, glmm_128 b) {
+  glmm_128 x0;
+  x0 = wasm_f32x4_mul(a, b);
+  x0 = wasm_f32x4_add(x0, glmm_shuff1(x0, 1, 0, 3, 2));
+  return wasm_f32x4_add(x0, glmm_shuff1(x0, 0, 1, 0, 1));
+}
+
+static inline
+float
+glmm_dot(glmm_128 a, glmm_128 b) {
+  return wasm_f32x4_extract_lane(glmm_vdots(a, b), 0);
+}
+
+static inline
+float
+glmm_norm(glmm_128 a) {
+  glmm_128 x0;
+  x0 = glmm_vhadds(wasm_f32x4_mul(a, a));
+  return wasm_f32x4_extract_lane(
+          wasm_i32x4_shuffle(x0, wasm_f32x4_sqrt(x0),4, 1, 2, 3), 0);
+}
+
+static inline
+float
+glmm_norm2(glmm_128 a) {
+  return wasm_f32x4_extract_lane(glmm_vhadds(wasm_f32x4_mul(a, a)), 0);
+}
+
+static inline
+float
+glmm_norm_one(glmm_128 a) {
+  return wasm_f32x4_extract_lane(glmm_vhadds(glmm_abs(a)), 0);
+}
+
+static inline
+float
+glmm_norm_inf(glmm_128 a) {
+  return wasm_f32x4_extract_lane(glmm_vhmax(glmm_abs(a)), 0);
+}
+
+static inline
+glmm_128
+glmm_load3(float v[3]) {
+  glmm_128 xy = wasm_v128_load64_zero(v);
+  return wasm_f32x4_replace_lane(xy, 2, v[2]);
+}
+
+static inline
+void
+glmm_store3(float v[3], glmm_128 vx) {
+  wasm_v128_store64_lane(v, vx, 0);
+  wasm_v128_store32_lane(&v[2], vx, 2);
+}
+
+static inline
+glmm_128
+glmm_div(glmm_128 a, glmm_128 b) {
+  return wasm_f32x4_div(a, b);
+}
+
+static inline
+glmm_128
+glmm_fmadd(glmm_128 a, glmm_128 b, glmm_128 c) {
+  return wasm_f32x4_add(c, wasm_f32x4_mul(a, b));
+}
+
+static inline
+glmm_128
+glmm_fnmadd(glmm_128 a, glmm_128 b, glmm_128 c) {
+  return wasm_f32x4_sub(c, wasm_f32x4_mul(a, b));
+}
+
+static inline
+glmm_128
+glmm_fmsub(glmm_128 a, glmm_128 b, glmm_128 c) {
+  return wasm_f32x4_sub(wasm_f32x4_mul(a, b), c);
+}
+
+static inline
+glmm_128
+glmm_fnmsub(glmm_128 a, glmm_128 b, glmm_128 c) {
+  return wasm_f32x4_neg(wasm_f32x4_add(wasm_f32x4_mul(a, b), c));
+}
+
+#endif
+#endif /* cglm_simd_wasm_h */

include/cglm/simd/wasm/affine.h

@@ -0,0 +1,127 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_affine_mat_wasm_h
+#define cglm_affine_mat_wasm_h
+#if defined(__wasm__) && defined(__wasm_simd128__)
+
+#include "../../common.h"
+#include "../intrin.h"
+
+CGLM_INLINE
+void
+glm_mul_wasm(mat4 m1, mat4 m2, mat4 dest) {
+  /* D = R * L (Column-Major) */
+  glmm_128 l, r0, r1, r2, r3, v0, v1, v2, v3;
+
+  l  = glmm_load(m1[0]);
+  r0 = glmm_load(m2[0]);
+  r1 = glmm_load(m2[1]);
+  r2 = glmm_load(m2[2]);
+  r3 = glmm_load(m2[3]);
+
+  v0 = wasm_f32x4_mul(glmm_splat_x(r0), l);
+  v1 = wasm_f32x4_mul(glmm_splat_x(r1), l);
+  v2 = wasm_f32x4_mul(glmm_splat_x(r2), l);
+  v3 = wasm_f32x4_mul(glmm_splat_x(r3), l);
+
+  l  = glmm_load(m1[1]);
+  v0 = glmm_fmadd(glmm_splat_y(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_y(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_y(r2), l, v2);
+  v3 = glmm_fmadd(glmm_splat_y(r3), l, v3);
+
+  l  = glmm_load(m1[2]);
+  v0 = glmm_fmadd(glmm_splat_z(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_z(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_z(r2), l, v2);
+  v3 = glmm_fmadd(glmm_splat_z(r3), l, v3);
+
+  l  = glmm_load(m1[3]);
+  v3 = glmm_fmadd(glmm_splat_w(r3), l, v3);
+
+  glmm_store(dest[0], v0);
+  glmm_store(dest[1], v1);
+  glmm_store(dest[2], v2);
+  glmm_store(dest[3], v3);
+}
+
+CGLM_INLINE
+void
+glm_mul_rot_wasm(mat4 m1, mat4 m2, mat4 dest) {
+  /* D = R * L (Column-Major) */
+
+  glmm_128 l, r0, r1, r2, v0, v1, v2;
+
+  l  = glmm_load(m1[0]);
+  r0 = glmm_load(m2[0]);
+  r1 = glmm_load(m2[1]);
+  r2 = glmm_load(m2[2]);
+
+  v0 = wasm_f32x4_mul(glmm_splat_x(r0), l);
+  v1 = wasm_f32x4_mul(glmm_splat_x(r1), l);
+  v2 = wasm_f32x4_mul(glmm_splat_x(r2), l);
+
+  l  = glmm_load(m1[1]);
+  v0 = glmm_fmadd(glmm_splat_y(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_y(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_y(r2), l, v2);
+
+  l  = glmm_load(m1[2]);
+  v0 = glmm_fmadd(glmm_splat_z(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_z(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_z(r2), l, v2);
+
+  glmm_store(dest[0], v0);
+  glmm_store(dest[1], v1);
+  glmm_store(dest[2], v2);
+  glmm_store(dest[3], glmm_load(m1[3]));
+}
+
+CGLM_INLINE
+void
+glm_inv_tr_wasm(mat4 mat) {
+  glmm_128 r0, r1, r2, r3, x0, x1, x2, x3, x4, x5;
+
+  r0 = glmm_load(mat[0]);
+  r1 = glmm_load(mat[1]);
+  r2 = glmm_load(mat[2]);
+  r3 = glmm_load(mat[3]);
+  x1 = wasm_f32x4_const(0.0f, 0.0f, 0.0f, 1.0f);
+
+  /* _MM_TRANSPOSE4_PS(r0, r1, r2, x1); */
+  x2 = wasm_i32x4_shuffle(r0, r1, 0, 4, 1, 5);
+  x3 = wasm_i32x4_shuffle(r0, r1, 2, 6, 3, 7);
+  x4 = wasm_i32x4_shuffle(r2, x1, 0, 4, 1, 5);
+  x5 = wasm_i32x4_shuffle(r2, x1, 2, 6, 3, 7);
+  /* r0 = _mm_movelh_ps(x2, x4); */
+  r0 = wasm_i32x4_shuffle(x2, x4, 0, 1, 4, 5);
+  /* r1 = _mm_movehl_ps(x4, x2); */
+  r1 = wasm_i32x4_shuffle(x4, x2, 6, 7, 2, 3);
+  /* r2 = _mm_movelh_ps(x3, x5); */
+  r2 = wasm_i32x4_shuffle(x3, x5, 0, 1, 4, 5);
+  /* x1 = _mm_movehl_ps(x5, x3); */
+  x1 = wasm_i32x4_shuffle(x5, x3, 6, 7, 2, 3);
+
+  x2 = glmm_shuff1(r3, 0, 0, 0, 0);
+  x3 = glmm_shuff1(r3, 1, 1, 1, 1);
+  x4 = glmm_shuff1(r3, 2, 2, 2, 2);
+
+  x0 = glmm_fmadd(r0, x2,
+                  glmm_fmadd(r1, x3, wasm_f32x4_mul(r2, x4)));
+  x0 = wasm_f32x4_neg(x0);
+
+  x0 = wasm_f32x4_add(x0, x1);
+
+  glmm_store(mat[0], r0);
+  glmm_store(mat[1], r1);
+  glmm_store(mat[2], r2);
+  glmm_store(mat[3], x0);
+}
+
+#endif
+#endif /* cglm_affine_mat_wasm_h */

include/cglm/simd/wasm/mat2.h

@@ -0,0 +1,50 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_mat2_wasm_h
+#define cglm_mat2_wasm_h
+#if defined(__wasm__) && defined(__wasm_simd128__)
+
+#include "../../common.h"
+#include "../intrin.h"
+
+CGLM_INLINE
+void
+glm_mat2_mul_wasm(mat2 m1, mat2 m2, mat2 dest) {
+  glmm_128 x0, x1, x2, x3, x4;
+
+  x1 = glmm_load(m1[0]); /* d c b a */
+  x2 = glmm_load(m2[0]); /* h g f e */
+
+  x3 = glmm_shuff1(x2, 2, 2, 0, 0);
+  x4 = glmm_shuff1(x2, 3, 3, 1, 1);
+  /* x0 = _mm_movelh_ps(x1, x1); */
+  x0 = wasm_i32x4_shuffle(x1, x1, 0, 1, 4, 5);
+  /* x2 = _mm_movehl_ps(x1, x1); */
+  x2 = wasm_i32x4_shuffle(x1, x1, 6, 7, 2, 3);
+
+  /*
+   dest[0][0] = a * e + c * f;
+   dest[0][1] = b * e + d * f;
+   dest[1][0] = a * g + c * h;
+   dest[1][1] = b * g + d * h;
+   */
+  x0 = glmm_fmadd(x0, x3, wasm_f32x4_mul(x2, x4));
+
+  glmm_store(dest[0], x0);
+}
+
+CGLM_INLINE
+void
+glm_mat2_transp_wasm(mat2 m, mat2 dest) {
+  /* d c b a */
+  /* d b c a */
+  glmm_store(dest[0], glmm_shuff1(glmm_load(m[0]), 3, 1, 2, 0));
+}
+
+#endif
+#endif /* cglm_mat2_wasm_h */

include/cglm/simd/wasm/mat3.h

@@ -0,0 +1,85 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_mat3_wasm_h
+#define cglm_mat3_wasm_h
+#if defined(__wasm__) && defined(__wasm_simd128__)
+
+#include "../../common.h"
+#include "../intrin.h"
+
+CGLM_INLINE
+void
+glm_mat3_mul_wasm(mat3 m1, mat3 m2, mat3 dest) {
+  glmm_128 l0, l1, l2, r0, r1, r2, x0, x1, x2, x3, x4, x5, x6, x7, x8, x9;
+  
+  l0 = wasm_v128_load(m1[0]);
+  l1 = wasm_v128_load(&m1[1][1]);
+
+  r0 = wasm_v128_load(m2[0]);
+  r1 = wasm_v128_load(&m2[1][1]);
+
+  x8 = glmm_shuff1(l0, 0, 2, 1, 0);                     /* a00 a02 a01 a00 */
+  x1 = glmm_shuff1(r0, 3, 0, 0, 0);                     /* b10 b00 b00 b00 */
+  x2 = wasm_i32x4_shuffle(l0, l1, 3, 3, 4, 5);          /* a12 a11 a10 a10 */
+  x3 = wasm_i32x4_shuffle(r0, r1, 1, 3, 4, 6);          /* b20 b11 b10 b01 */
+  x0 = wasm_f32x4_mul(x8, x1);
+
+  x6 = glmm_shuff1(l0, 1, 0, 2, 1);                     /* a01 a00 a02 a01 */
+  x7 = glmm_shuff1(x3, 3, 3, 1, 1);                     /* b20 b20 b10 b10 */
+  l2 = wasm_v128_load32_zero(&m1[2][2]);
+  r2 = wasm_v128_load32_zero(&m2[2][2]);
+  x1 = wasm_f32x4_mul(x6, x7);
+  l2 = glmm_shuff1(l2, 0, 0, 1, 0);                     /* a22 a22 0.f a22 */
+  r2 = glmm_shuff1(r2, 0, 0, 1, 0);                     /* b22 b22 0.f b22 */
+
+  x4 = glmm_shuff1(x2, 0, 3, 2, 0);                     /* a10 a12 a11 a10 */
+  x5 = glmm_shuff1(x2, 2, 0, 3, 2);                     /* a11 a10 a12 a11 */
+  x6 = glmm_shuff1(x3, 2, 0, 0, 0);                     /* b11 b01 b01 b01 */
+  x2 = glmm_shuff1(r1, 3, 3, 0, 0);                     /* b21 b21 b11 b11 */
+
+  /* x8 = _mm_unpackhi_ps(x8, x4); */
+  /* x9 = _mm_unpackhi_ps(x7, x2); */
+  x8 = wasm_i32x4_shuffle(x8, x4, 2, 6, 3, 7);          /* a10 a00 a12 a02 */
+  x9 = wasm_i32x4_shuffle(x7, x2, 2, 6, 3, 7);          /* b21 b20 b21 b20 */
+
+  x0 = glmm_fmadd(x4, x6, x0);
+  x1 = glmm_fmadd(x5, x2, x1);
+
+  /* x2 = _mm_movehl_ps(l2, l1); */
+  x2 = wasm_i32x4_shuffle(l2, l1, 6, 7, 2, 3);          /* a22 a22 a21 a20 */
+  x3 = glmm_shuff1(x2, 0, 2, 1, 0);                     /* a20 a22 a21 a20 */
+  x2 = glmm_shuff1(x2, 1, 0, 2, 1);                     /* a21 a20 a22 a21 */
+  x4 = wasm_i32x4_shuffle(r0, r1, 2, 2, 5, 5);          /* b12 b12 b02 b02 */
+  
+  x5 = glmm_shuff1(x4, 3, 0, 0, 0);                     /* b12 b02 b02 b02 */
+  /* x4 = _mm_movehl_ps(r2, x4); */
+  x4 = wasm_i32x4_shuffle(r2, x4, 6, 7, 2, 3);          /* b22 b22 b12 b12 */
+  x0 = glmm_fmadd(x3, x5, x0);
+  x1 = glmm_fmadd(x2, x4, x1);
+
+  /*
+   Dot Product : dest[2][2] =  a02 * b20 +
+                               a12 * b21 +
+                               a22 * b22 +
+                               0   * 00                                    */
+  /* x2 = _mm_movelh_ps(x8, l2); */
+  /* x3 = _mm_movelh_ps(x9, r2); */
+  x2 = wasm_i32x4_shuffle(x8, l2, 0, 1, 4, 5);           /* 0.f a22 a12 a02 */
+  x3 = wasm_i32x4_shuffle(x9, r2, 0, 1, 4, 5);           /* 0.f b22 b21 b20 */
+  x2 = glmm_vdots(x2, x3);
+
+  /* _mm_storeu_ps(&dest[0][0], x0); */
+  wasm_v128_store(&dest[0][0], x0);
+  /* _mm_storeu_ps(&dest[1][1], x1); */
+  wasm_v128_store(&dest[1][1], x1);
+  /* _mm_store_ss (&dest[2][2], x2); */
+  wasm_v128_store32_lane(&dest[2][2], x2, 0);
+}
+
+#endif
+#endif /* cglm_mat3_wasm_h */

include/cglm/simd/wasm/mat4.h

@@ -0,0 +1,454 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_mat_wasm_h
+#define cglm_mat_wasm_h
+#if defined(__wasm__) && defined(__wasm_simd128__)
+
+#include "../../common.h"
+#include "../intrin.h"
+
+#define glm_mat4_inv_precise_wasm(mat, dest) glm_mat4_inv_wasm(mat, dest)
+
+CGLM_INLINE
+void
+glm_mat4_scale_wasm(mat4 m, float s) {
+  glmm_128 x0;
+  x0 = wasm_f32x4_splat(s);
+
+  glmm_store(m[0], wasm_f32x4_mul(glmm_load(m[0]), x0));
+  glmm_store(m[1], wasm_f32x4_mul(glmm_load(m[1]), x0));
+  glmm_store(m[2], wasm_f32x4_mul(glmm_load(m[2]), x0));
+  glmm_store(m[3], wasm_f32x4_mul(glmm_load(m[3]), x0));
+}
+
+CGLM_INLINE
+void
+glm_mat4_transp_wasm(mat4 m, mat4 dest) {
+  glmm_128 r0, r1, r2, r3, tmp0, tmp1, tmp2, tmp3;
+
+  r0 = glmm_load(m[0]);
+  r1 = glmm_load(m[1]);
+  r2 = glmm_load(m[2]);
+  r3 = glmm_load(m[3]);
+
+  /* _MM_TRANSPOSE4_PS(r0, r1, r2, r3); */
+  tmp0 = wasm_i32x4_shuffle(r0, r1, 0, 4, 1, 5);
+  tmp1 = wasm_i32x4_shuffle(r0, r1, 2, 6, 3, 7);
+  tmp2 = wasm_i32x4_shuffle(r2, r3, 0, 4, 1, 5);
+  tmp3 = wasm_i32x4_shuffle(r2, r3, 2, 6, 3, 7);
+  /* r0 = _mm_movelh_ps(tmp0, tmp2); */
+  r0 = wasm_i32x4_shuffle(tmp0, tmp2, 0, 1, 4, 5);
+  /* r1 = _mm_movehl_ps(tmp2, tmp0); */
+  r1 = wasm_i32x4_shuffle(tmp2, tmp0, 6, 7, 2, 3);
+  /* r2 = _mm_movelh_ps(tmp1, tmp3); */
+  r2 = wasm_i32x4_shuffle(tmp1, tmp3, 0, 1, 4, 5);
+  /* r3 = _mm_movehl_ps(tmp3, tmp1); */
+  r3 = wasm_i32x4_shuffle(tmp3, tmp1, 6, 7, 2, 3);
+
+  glmm_store(dest[0], r0);
+  glmm_store(dest[1], r1);
+  glmm_store(dest[2], r2);
+  glmm_store(dest[3], r3);
+}
+
+CGLM_INLINE
+void
+glm_mat4_mul_wasm(mat4 m1, mat4 m2, mat4 dest) {
+  /* D = R * L (Column-Major) */
+
+  glmm_128 l, r0, r1, r2, r3, v0, v1, v2, v3;
+
+  l  = glmm_load(m1[0]);
+  r0 = glmm_load(m2[0]);
+  r1 = glmm_load(m2[1]);
+  r2 = glmm_load(m2[2]);
+  r3 = glmm_load(m2[3]);
+
+  v0 = wasm_f32x4_mul(glmm_splat_x(r0), l);
+  v1 = wasm_f32x4_mul(glmm_splat_x(r1), l);
+  v2 = wasm_f32x4_mul(glmm_splat_x(r2), l);
+  v3 = wasm_f32x4_mul(glmm_splat_x(r3), l);
+
+  l  = glmm_load(m1[1]);
+  v0 = glmm_fmadd(glmm_splat_y(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_y(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_y(r2), l, v2);
+  v3 = glmm_fmadd(glmm_splat_y(r3), l, v3);
+
+  l  = glmm_load(m1[2]);
+  v0 = glmm_fmadd(glmm_splat_z(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_z(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_z(r2), l, v2);
+  v3 = glmm_fmadd(glmm_splat_z(r3), l, v3);
+
+  l  = glmm_load(m1[3]);
+  v0 = glmm_fmadd(glmm_splat_w(r0), l, v0);
+  v1 = glmm_fmadd(glmm_splat_w(r1), l, v1);
+  v2 = glmm_fmadd(glmm_splat_w(r2), l, v2);
+  v3 = glmm_fmadd(glmm_splat_w(r3), l, v3);
+
+  glmm_store(dest[0], v0);
+  glmm_store(dest[1], v1);
+  glmm_store(dest[2], v2);
+  glmm_store(dest[3], v3);
+}
+
+CGLM_INLINE
+void
+glm_mat4_mulv_wasm(mat4 m, vec4 v, vec4 dest) {
+  glmm_128 x0, x1, m0, m1, m2, m3, v0, v1, v2, v3;
+
+  m0 = glmm_load(m[0]);
+  m1 = glmm_load(m[1]);
+  m2 = glmm_load(m[2]);
+  m3 = glmm_load(m[3]);
+
+  x0 = glmm_load(v);
+  v0 = glmm_splat_x(x0);
+  v1 = glmm_splat_y(x0);
+  v2 = glmm_splat_z(x0);
+  v3 = glmm_splat_w(x0);
+
+  x1 = wasm_f32x4_mul(m3, v3);
+  x1 = glmm_fmadd(m2, v2, x1);
+  x1 = glmm_fmadd(m1, v1, x1);
+  x1 = glmm_fmadd(m0, v0, x1);
+
+  glmm_store(dest, x1);
+}
+
+CGLM_INLINE
+float
+glm_mat4_det_wasm(mat4 mat) {
+  glmm_128 r0, r1, r2, r3, x0, x1, x2;
+
+  /* 127 <- 0, [square] det(A) = det(At) */
+  r0 = glmm_load(mat[0]); /* d c b a */
+  r1 = glmm_load(mat[1]); /* h g f e */
+  r2 = glmm_load(mat[2]); /* l k j i */
+  r3 = glmm_load(mat[3]); /* p o n m */
+
+  /*
+   t[1] = j * p - n * l;
+   t[2] = j * o - n * k;
+   t[3] = i * p - m * l;
+   t[4] = i * o - m * k;
+   */
+  x0 = glmm_fnmadd(glmm_shuff1(r3, 0, 0, 1, 1), glmm_shuff1(r2, 2, 3, 2, 3),
+                   wasm_f32x4_mul(glmm_shuff1(r2, 0, 0, 1, 1),
+                              glmm_shuff1(r3, 2, 3, 2, 3)));
+  /*
+   t[0] = k * p - o * l;
+   t[0] = k * p - o * l;
+   t[5] = i * n - m * j;
+   t[5] = i * n - m * j;
+   */
+  x1 = glmm_fnmadd(glmm_shuff1(r3, 0, 0, 2, 2), glmm_shuff1(r2, 1, 1, 3, 3),
+                   wasm_f32x4_mul(glmm_shuff1(r2, 0, 0, 2, 2),
+                              glmm_shuff1(r3, 1, 1, 3, 3)));
+
+  /*
+     a * (f * t[0] - g * t[1] + h * t[2])
+   - b * (e * t[0] - g * t[3] + h * t[4])
+   + c * (e * t[1] - f * t[3] + h * t[5])
+   - d * (e * t[2] - f * t[4] + g * t[5])
+   */
+  x2 = glmm_fnmadd(glmm_shuff1(r1, 1, 1, 2, 2), glmm_shuff1(x0, 3, 2, 2, 0),
+                   wasm_f32x4_mul(glmm_shuff1(r1, 0, 0, 0, 1),
+                              wasm_i32x4_shuffle(x1, x0, 0, 0, 4, 5)));
+  x2 = glmm_fmadd(glmm_shuff1(r1, 2, 3, 3, 3),
+                  wasm_i32x4_shuffle(x0, x1, 1, 3, 6, 6),
+                  x2);
+  /* x2 = wasm_v128_xor(x2, wasm_f32x4_const(0.f, -0.f, 0.f, -0.f)); */
+  x2 = wasm_v128_xor(x2, glmm_float32x4_SIGNMASK_PNPN);
+  
+  return glmm_hadd(wasm_f32x4_mul(x2, r0));
+}
+
+CGLM_INLINE
+void
+glm_mat4_inv_fast_wasm(mat4 mat, mat4 dest) {
+  glmm_128 r0, r1, r2, r3,
+         v0, v1, v2, v3,
+         t0, t1, t2, t3, t4, t5,
+         x0, x1, x2, x3, x4, x5, x6, x7, x8, x9;
+
+  /* x8 = wasm_f32x4_const(0.f, -0.f, 0.f, -0.f); */
+  x8 = glmm_float32x4_SIGNMASK_PNPN;
+  x9 = glmm_shuff1(x8, 2, 1, 2, 1);
+
+  /* 127 <- 0 */
+  r0 = glmm_load(mat[0]); /* d c b a */
+  r1 = glmm_load(mat[1]); /* h g f e */
+  r2 = glmm_load(mat[2]); /* l k j i */
+  r3 = glmm_load(mat[3]); /* p o n m */
+  /* x0 = _mm_movehl_ps(r3, r2); */
+  x0 = wasm_i32x4_shuffle(r3, r2, 6, 7, 2, 3);           /* p o l k */
+  /* x3 = _mm_movelh_ps(r2, r3); */
+  x3 = wasm_i32x4_shuffle(r2, r3, 0, 1, 4, 5);           /* n m j i */
+  x1 = glmm_shuff1(x0, 1, 3, 3 ,3);                      /* l p p p */
+  x2 = glmm_shuff1(x0, 0, 2, 2, 2);                      /* k o o o */
+  x4 = glmm_shuff1(x3, 1, 3, 3, 3);                      /* j n n n */
+  x7 = glmm_shuff1(x3, 0, 2, 2, 2);                      /* i m m m */
+
+  x6 = wasm_i32x4_shuffle(r2, r1, 0, 0, 4, 4);           /* e e i i */
+  x5 = wasm_i32x4_shuffle(r2, r1, 1, 1, 5, 5);           /* f f j j */
+  x3 = wasm_i32x4_shuffle(r2, r1, 2, 2, 6, 6);           /* g g k k */
+  x0 = wasm_i32x4_shuffle(r2, r1, 3, 3, 7, 7);           /* h h l l */
+  
+  t0 = wasm_f32x4_mul(x3, x1);
+  t1 = wasm_f32x4_mul(x5, x1);
+  t2 = wasm_f32x4_mul(x5, x2);
+  t3 = wasm_f32x4_mul(x6, x1);
+  t4 = wasm_f32x4_mul(x6, x2);
+  t5 = wasm_f32x4_mul(x6, x4);
+  
+  /* t1[0] = k * p - o * l;
+     t1[0] = k * p - o * l;
+     t2[0] = g * p - o * h;
+     t3[0] = g * l - k * h; */
+  t0 = glmm_fnmadd(x2, x0, t0);
+  
+  /* t1[1] = j * p - n * l;
+     t1[1] = j * p - n * l;
+     t2[1] = f * p - n * h;
+     t3[1] = f * l - j * h; */
+   t1 = glmm_fnmadd(x4, x0, t1);
+  
+  /* t1[2] = j * o - n * k
+     t1[2] = j * o - n * k;
+     t2[2] = f * o - n * g;
+     t3[2] = f * k - j * g; */
+  t2 = glmm_fnmadd(x4, x3, t2);
+  
+  /* t1[3] = i * p - m * l;
+     t1[3] = i * p - m * l;
+     t2[3] = e * p - m * h;
+     t3[3] = e * l - i * h; */
+  t3 = glmm_fnmadd(x7, x0, t3);
+  
+  /* t1[4] = i * o - m * k;
+     t1[4] = i * o - m * k;
+     t2[4] = e * o - m * g;
+     t3[4] = e * k - i * g; */
+  t4 = glmm_fnmadd(x7, x3, t4);
+  
+  /* t1[5] = i * n - m * j;
+     t1[5] = i * n - m * j;
+     t2[5] = e * n - m * f;
+     t3[5] = e * j - i * f; */
+  t5 = glmm_fnmadd(x7, x5, t5);
+  /* x4 = _mm_movelh_ps(r0, r1); */
+  x4 = wasm_i32x4_shuffle(r0, r1, 0, 1, 4, 5);           /* f e b a */
+  /* x5 = _mm_movehl_ps(r1, r0); */
+  x5 = wasm_i32x4_shuffle(r1, r0, 6, 7, 2, 3);           /* h g d c */
+  
+  x0 = glmm_shuff1(x4, 0, 0, 0, 2);                      /* a a a e */
+  x1 = glmm_shuff1(x4, 1, 1, 1, 3);                      /* b b b f */
+  x2 = glmm_shuff1(x5, 0, 0, 0, 2);                      /* c c c g */
+  x3 = glmm_shuff1(x5, 1, 1, 1, 3);                      /* d d d h */
+  
+  v2 = wasm_f32x4_mul(x0, t1);
+  v1 = wasm_f32x4_mul(x0, t0);
+  v3 = wasm_f32x4_mul(x0, t2);
+  v0 = wasm_f32x4_mul(x1, t0);
+  
+  v2 = glmm_fnmadd(x1, t3, v2);
+  v3 = glmm_fnmadd(x1, t4, v3);
+  v0 = glmm_fnmadd(x2, t1, v0);
+  v1 = glmm_fnmadd(x2, t3, v1);
+  
+  v3 = glmm_fmadd(x2, t5, v3);
+  v0 = glmm_fmadd(x3, t2, v0);
+  v2 = glmm_fmadd(x3, t5, v2);
+  v1 = glmm_fmadd(x3, t4, v1);
+
+  /*
+   dest[0][0] =  f * t1[0] - g * t1[1] + h * t1[2];
+   dest[0][1] =-(b * t1[0] - c * t1[1] + d * t1[2]);
+   dest[0][2] =  b * t2[0] - c * t2[1] + d * t2[2];
+   dest[0][3] =-(b * t3[0] - c * t3[1] + d * t3[2]); */
+  v0 = wasm_v128_xor(v0, x8);
+  
+  /*
+   dest[2][0] =  e * t1[1] - f * t1[3] + h * t1[5];
+   dest[2][1] =-(a * t1[1] - b * t1[3] + d * t1[5]);
+   dest[2][2] =  a * t2[1] - b * t2[3] + d * t2[5];
+   dest[2][3] =-(a * t3[1] - b * t3[3] + d * t3[5]);*/
+  v2 = wasm_v128_xor(v2, x8);
+
+  /*
+   dest[1][0] =-(e * t1[0] - g * t1[3] + h * t1[4]);
+   dest[1][1] =  a * t1[0] - c * t1[3] + d * t1[4];
+   dest[1][2] =-(a * t2[0] - c * t2[3] + d * t2[4]);
+   dest[1][3] =  a * t3[0] - c * t3[3] + d * t3[4]; */
+  v1 = wasm_v128_xor(v1, x9);
+
+  /*
+   dest[3][0] =-(e * t1[2] - f * t1[4] + g * t1[5]);
+   dest[3][1] =  a * t1[2] - b * t1[4] + c * t1[5];
+   dest[3][2] =-(a * t2[2] - b * t2[4] + c * t2[5]);
+   dest[3][3] =  a * t3[2] - b * t3[4] + c * t3[5]; */
+  v3 = wasm_v128_xor(v3, x9);
+
+  /* determinant */
+  x0 = wasm_i32x4_shuffle(v0, v1, 0, 0, 4, 4);
+  x1 = wasm_i32x4_shuffle(v2, v3, 0, 0, 4, 4);
+  x0 = wasm_i32x4_shuffle(x0, x1, 0, 2, 4, 6);
+
+  /* x0 = _mm_rcp_ps(glmm_vhadd(wasm_f32x4_mul(x0, r0))); */
+  x0 = wasm_f32x4_div(wasm_f32x4_const_splat(1.0f),
+                      glmm_vhadd(wasm_f32x4_mul(x0, r0)));
+
+  glmm_store(dest[0], wasm_f32x4_mul(v0, x0));
+  glmm_store(dest[1], wasm_f32x4_mul(v1, x0));
+  glmm_store(dest[2], wasm_f32x4_mul(v2, x0));
+  glmm_store(dest[3], wasm_f32x4_mul(v3, x0));
+}
+
+CGLM_INLINE
+void
+glm_mat4_inv_wasm(mat4 mat, mat4 dest) {
+  glmm_128 r0, r1, r2, r3,
+         v0, v1, v2, v3,
+         t0, t1, t2, t3, t4, t5,
+         x0, x1, x2, x3, x4, x5, x6, x7, x8, x9;
+
+  /* x8 = wasm_f32x4_const(0.f, -0.f, 0.f, -0.f); */
+  x8 = glmm_float32x4_SIGNMASK_PNPN;
+  x9 = glmm_shuff1(x8, 2, 1, 2, 1);
+
+  /* 127 <- 0 */
+  r0 = glmm_load(mat[0]); /* d c b a */
+  r1 = glmm_load(mat[1]); /* h g f e */
+  r2 = glmm_load(mat[2]); /* l k j i */
+  r3 = glmm_load(mat[3]); /* p o n m */
+  /* x0 = _mm_movehl_ps(r3, r2); */
+  x0 = wasm_i32x4_shuffle(r3, r2, 6, 7, 2, 3);           /* p o l k */
+  /* x3 = _mm_movelh_ps(r2, r3); */
+  x3 = wasm_i32x4_shuffle(r2, r3, 0, 1, 4, 5);           /* n m j i */
+  x1 = glmm_shuff1(x0, 1, 3, 3 ,3);                      /* l p p p */
+  x2 = glmm_shuff1(x0, 0, 2, 2, 2);                      /* k o o o */
+  x4 = glmm_shuff1(x3, 1, 3, 3, 3);                      /* j n n n */
+  x7 = glmm_shuff1(x3, 0, 2, 2, 2);                      /* i m m m */
+
+  x6 = wasm_i32x4_shuffle(r2, r1, 0, 0, 4, 4);           /* e e i i */
+  x5 = wasm_i32x4_shuffle(r2, r1, 1, 1, 5, 5);           /* f f j j */
+  x3 = wasm_i32x4_shuffle(r2, r1, 2, 2, 6, 6);           /* g g k k */
+  x0 = wasm_i32x4_shuffle(r2, r1, 3, 3, 7, 7);           /* h h l l */
+  
+  t0 = wasm_f32x4_mul(x3, x1);
+  t1 = wasm_f32x4_mul(x5, x1);
+  t2 = wasm_f32x4_mul(x5, x2);
+  t3 = wasm_f32x4_mul(x6, x1);
+  t4 = wasm_f32x4_mul(x6, x2);
+  t5 = wasm_f32x4_mul(x6, x4);
+  
+  /* t1[0] = k * p - o * l;
+     t1[0] = k * p - o * l;
+     t2[0] = g * p - o * h;
+     t3[0] = g * l - k * h; */
+  t0 = glmm_fnmadd(x2, x0, t0);
+  
+  /* t1[1] = j * p - n * l;
+     t1[1] = j * p - n * l;
+     t2[1] = f * p - n * h;
+     t3[1] = f * l - j * h; */
+   t1 = glmm_fnmadd(x4, x0, t1);
+  
+  /* t1[2] = j * o - n * k
+     t1[2] = j * o - n * k;
+     t2[2] = f * o - n * g;
+     t3[2] = f * k - j * g; */
+  t2 = glmm_fnmadd(x4, x3, t2);
+  
+  /* t1[3] = i * p - m * l;
+     t1[3] = i * p - m * l;
+     t2[3] = e * p - m * h;
+     t3[3] = e * l - i * h; */
+  t3 = glmm_fnmadd(x7, x0, t3);
+  
+  /* t1[4] = i * o - m * k;
+     t1[4] = i * o - m * k;
+     t2[4] = e * o - m * g;
+     t3[4] = e * k - i * g; */
+  t4 = glmm_fnmadd(x7, x3, t4);
+  
+  /* t1[5] = i * n - m * j;
+     t1[5] = i * n - m * j;
+     t2[5] = e * n - m * f;
+     t3[5] = e * j - i * f; */
+  t5 = glmm_fnmadd(x7, x5, t5);
+  /* x4 = _mm_movelh_ps(r0, r1); */
+  x4 = wasm_i32x4_shuffle(r0, r1, 0, 1, 4, 5);           /* f e b a */
+  /* x5 = _mm_movehl_ps(r1, r0); */
+  x5 = wasm_i32x4_shuffle(r1, r0, 6, 7, 2, 3);           /* h g d c */
+  
+  x0 = glmm_shuff1(x4, 0, 0, 0, 2);                      /* a a a e */
+  x1 = glmm_shuff1(x4, 1, 1, 1, 3);                      /* b b b f */
+  x2 = glmm_shuff1(x5, 0, 0, 0, 2);                      /* c c c g */
+  x3 = glmm_shuff1(x5, 1, 1, 1, 3);                      /* d d d h */
+  
+  v2 = wasm_f32x4_mul(x0, t1);
+  v1 = wasm_f32x4_mul(x0, t0);
+  v3 = wasm_f32x4_mul(x0, t2);
+  v0 = wasm_f32x4_mul(x1, t0);
+  
+  v2 = glmm_fnmadd(x1, t3, v2);
+  v3 = glmm_fnmadd(x1, t4, v3);
+  v0 = glmm_fnmadd(x2, t1, v0);
+  v1 = glmm_fnmadd(x2, t3, v1);
+  
+  v3 = glmm_fmadd(x2, t5, v3);
+  v0 = glmm_fmadd(x3, t2, v0);
+  v2 = glmm_fmadd(x3, t5, v2);
+  v1 = glmm_fmadd(x3, t4, v1);
+
+  /*
+   dest[0][0] =  f * t1[0] - g * t1[1] + h * t1[2];
+   dest[0][1] =-(b * t1[0] - c * t1[1] + d * t1[2]);
+   dest[0][2] =  b * t2[0] - c * t2[1] + d * t2[2];
+   dest[0][3] =-(b * t3[0] - c * t3[1] + d * t3[2]); */
+  v0 = wasm_v128_xor(v0, x8);
+  
+  /*
+   dest[2][0] =  e * t1[1] - f * t1[3] + h * t1[5];
+   dest[2][1] =-(a * t1[1] - b * t1[3] + d * t1[5]);
+   dest[2][2] =  a * t2[1] - b * t2[3] + d * t2[5];
+   dest[2][3] =-(a * t3[1] - b * t3[3] + d * t3[5]);*/
+  v2 = wasm_v128_xor(v2, x8);
+
+  /*
+   dest[1][0] =-(e * t1[0] - g * t1[3] + h * t1[4]);
+   dest[1][1] =  a * t1[0] - c * t1[3] + d * t1[4];
+   dest[1][2] =-(a * t2[0] - c * t2[3] + d * t2[4]);
+   dest[1][3] =  a * t3[0] - c * t3[3] + d * t3[4]; */
+  v1 = wasm_v128_xor(v1, x9);
+
+  /*
+   dest[3][0] =-(e * t1[2] - f * t1[4] + g * t1[5]);
+   dest[3][1] =  a * t1[2] - b * t1[4] + c * t1[5];
+   dest[3][2] =-(a * t2[2] - b * t2[4] + c * t2[5]);
+   dest[3][3] =  a * t3[2] - b * t3[4] + c * t3[5]; */
+  v3 = wasm_v128_xor(v3, x9);
+
+  /* determinant */
+  x0 = wasm_i32x4_shuffle(v0, v1, 0, 0, 4, 4);
+  x1 = wasm_i32x4_shuffle(v2, v3, 0, 0, 4, 4);
+  x0 = wasm_i32x4_shuffle(x0, x1, 0, 2, 4, 6);
+
+  x0 = wasm_f32x4_div(wasm_f32x4_splat(1.0f), glmm_vhadd(wasm_f32x4_mul(x0, r0)));
+
+  glmm_store(dest[0], wasm_f32x4_mul(v0, x0));
+  glmm_store(dest[1], wasm_f32x4_mul(v1, x0));
+  glmm_store(dest[2], wasm_f32x4_mul(v2, x0));
+  glmm_store(dest[3], wasm_f32x4_mul(v3, x0));
+}
+
+#endif
+#endif /* cglm_mat_wasm_h */

include/cglm/simd/wasm/quat.h

@@ -0,0 +1,55 @@
+/*
+ * Copyright (c), Recep Aslantas.
+ *
+ * MIT License (MIT), http://opensource.org/licenses/MIT
+ * Full license can be found in the LICENSE file
+ */
+
+#ifndef cglm_quat_wasm_h
+#define cglm_quat_wasm_h
+#if defined(__wasm__) && defined(__wasm_simd128__)
+
+#include "../../common.h"
+#include "../intrin.h"
+
+CGLM_INLINE
+void
+glm_quat_mul_wasm(versor p, versor q, versor dest) {
+  /*
+   + (a1 b2 + b1 a2 + c1 d2 − d1 c2)i
+   + (a1 c2 − b1 d2 + c1 a2 + d1 b2)j
+   + (a1 d2 + b1 c2 − c1 b2 + d1 a2)k
+     a1 a2 − b1 b2 − c1 c2 − d1 d2
+   */
+
+  glmm_128 xp, xq, x1, x2, x3, r, x, y, z;
+
+  xp = glmm_load(p); /* 3 2 1 0 */
+  xq = glmm_load(q);
+  /* x1 = wasm_f32x4_const(0.f, -0.f, 0.f, -0.f); */
+  x1 = glmm_float32x4_SIGNMASK_PNPN; /* TODO: _mm_set1_ss() + shuff ? */
+  r  = wasm_f32x4_mul(glmm_splat_w(xp), xq);
+  /* x2 = _mm_unpackhi_ps(x1, x1); */
+  x2 = wasm_i32x4_shuffle(x1, x1, 2, 6, 3, 7);
+  x3 = glmm_shuff1(x1, 3, 2, 0, 1);
+  x  = glmm_splat_x(xp);
+  y  = glmm_splat_y(xp);
+  z  = glmm_splat_z(xp);
+
+  x  = wasm_v128_xor(x, x1);
+  y  = wasm_v128_xor(y, x2);
+  z  = wasm_v128_xor(z, x3);
+  
+  x1 = glmm_shuff1(xq, 0, 1, 2, 3);
+  x2 = glmm_shuff1(xq, 1, 0, 3, 2);
+  x3 = glmm_shuff1(xq, 2, 3, 0, 1);
+  
+  r  = glmm_fmadd(x, x1, r);
+  r  = glmm_fmadd(y, x2, r);
+  r  = glmm_fmadd(z, x3, r);
+
+  glmm_store(dest, r);
+}
+
+#endif
+#endif /* cglm_quat_wasm_h */

include/cglm/simd/x86.h

@@ -54,10 +54,23 @@
 #  endif
 #endif
 
+#define GLMM_NEGZEROf 0x80000000 /*  0x80000000 ---> -0.0f  */
+
+#define GLMM__SIGNMASKf(X, Y, Z, W)                                           \
+   _mm_castsi128_ps(_mm_set_epi32(X, Y, Z, W))
+  /* _mm_set_ps(X, Y, Z, W); */
+
+#define glmm_float32x4_SIGNMASK_PNPN GLMM__SIGNMASKf(0, GLMM_NEGZEROf, 0, GLMM_NEGZEROf)
+#define glmm_float32x4_SIGNMASK_NPNP GLMM__SIGNMASKf(GLMM_NEGZEROf, 0, GLMM_NEGZEROf, 0)
+#define glmm_float32x4_SIGNMASK_NPPN GLMM__SIGNMASKf(GLMM_NEGZEROf, 0, 0, GLMM_NEGZEROf)
+
+#define glmm_float32x4_SIGNMASK_NEG _mm_castsi128_ps(_mm_set1_epi32(0x80000000)) /* _mm_set1_ps(-0.0f) */
+#define glmm_float32x8_SIGNMASK_NEG _mm256_castsi256_ps(_mm256_set1_epi32(GLMM_NEGZEROf))
+
 static inline
 __m128
 glmm_abs(__m128 x) {
-  return _mm_andnot_ps(_mm_set1_ps(-0.0f), x);
+  return _mm_andnot_ps(glmm_float32x4_SIGNMASK_NEG, x);
 }
 
 static inline
@@ -256,7 +269,8 @@ glmm_fnmsub(__m128 a, __m128 b, __m128 c) {
 #ifdef __FMA__
   return _mm_fnmsub_ps(a, b, c);
 #else
-  return _mm_xor_ps(_mm_add_ps(_mm_mul_ps(a, b), c), _mm_set1_ps(-0.0f));
+  return _mm_xor_ps(_mm_add_ps(_mm_mul_ps(a, b), c),
+                    glmm_float32x4_SIGNMASK_NEG);
 #endif
 }
 
@@ -298,7 +312,7 @@ glmm256_fnmsub(__m256 a, __m256 b, __m256 c) {
   return _mm256_fmsub_ps(a, b, c);
 #else
   return _mm256_xor_ps(_mm256_sub_ps(_mm256_mul_ps(a, b), c),
-                       _mm256_set1_ps(-0.0f));
+                       glmm_float32x8_SIGNMASK_NEG);
 #endif
 }
 #endif

include/cglm/struct/box.h

@@ -31,7 +31,7 @@ glms_aabb_(transform)(vec3s box[2], mat4s m, vec3s dest[2]) {
   vec3 rawBox[2];
   vec3 rawDest[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_aabb_transform(rawBox, m.raw, rawDest);
   glms_vec3_pack(dest, rawDest, 2);
 }
@@ -53,10 +53,10 @@ glms_aabb_(merge)(vec3s box1[2], vec3s box2[2], vec3s dest[2]) {
   vec3 rawBox2[2];
   vec3 rawDest[2];
 
-  glms_vec3_unpack(rawBox1, box1, 2);
+  glms_vec3_(unpack)(rawBox1, box1, 2);
-  glms_vec3_unpack(rawBox2, box2, 2);
+  glms_vec3_(unpack)(rawBox2, box2, 2);
   glm_aabb_merge(rawBox1, rawBox2, rawDest);
-  glms_vec3_pack(dest, rawDest, 2);
+  glms_vec3_(pack)(dest, rawDest, 2);
 }
 
 /*!
@@ -77,10 +77,10 @@ glms_aabb_(crop)(vec3s box[2], vec3s cropBox[2], vec3s dest[2]) {
   vec3 rawCropBox[2];
   vec3 rawDest[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
-  glms_vec3_unpack(rawCropBox, cropBox, 2);
+  glms_vec3_(unpack)(rawCropBox, cropBox, 2);
   glm_aabb_crop(rawBox, rawCropBox, rawDest);
-  glms_vec3_pack(dest, rawDest, 2);
+  glms_vec3_(pack)(dest, rawDest, 2);
 }
 
 /*!
@@ -101,8 +101,8 @@ glms_aabb_(crop_until)(vec3s box[2],
                        vec3s cropBox[2],
                        vec3s clampBox[2],
                        vec3s dest[2]) {
-  glms_aabb_crop(box, cropBox, dest);
+  glms_aabb_(crop)(box, cropBox, dest);
-  glms_aabb_merge(clampBox, dest, dest);
+  glms_aabb_(merge)(clampBox, dest, dest);
 }
 
 /*!
@@ -125,8 +125,8 @@ glms_aabb_(frustum)(vec3s box[2], vec4s planes[6]) {
   vec3 rawBox[2];
   vec4 rawPlanes[6];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
-  glms_vec4_unpack(rawPlanes, planes, 6);
+  glms_vec4_(unpack)(rawPlanes, planes, 6);
   return glm_aabb_frustum(rawBox, rawPlanes);
 }
 
@@ -138,8 +138,8 @@ glms_aabb_(frustum)(vec3s box[2], vec4s planes[6]) {
 CGLM_INLINE
 void
 glms_aabb_(invalidate)(vec3s box[2]) {
-  box[0] = glms_vec3_broadcast(FLT_MAX);
+  box[0] = glms_vec3_(broadcast)(FLT_MAX);
-  box[1] = glms_vec3_broadcast(-FLT_MAX);
+  box[1] = glms_vec3_(broadcast)(-FLT_MAX);
 }
 
 /*!
@@ -151,7 +151,7 @@ CGLM_INLINE
 bool
 glms_aabb_(isvalid)(vec3s box[2]) {
   vec3 rawBox[2];
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   return glm_aabb_isvalid(rawBox);
 }
 
@@ -174,7 +174,7 @@ glms_aabb_(size)(vec3s box[2]) {
 CGLM_INLINE
 float
 glms_aabb_(radius)(vec3s box[2]) {
-  return glms_aabb_size(box) * 0.5f;
+  return glms_aabb_(size)(box) * 0.5f;
 }
 
 /*!
@@ -186,7 +186,7 @@ glms_aabb_(radius)(vec3s box[2]) {
 CGLM_INLINE
 vec3s
 glms_aabb_(center)(vec3s box[2]) {
-  return glms_vec3_center(box[0], box[1]);
+  return glms_vec3_(center)(box[0], box[1]);
 }
 
 /*!
@@ -201,8 +201,8 @@ glms_aabb_(aabb)(vec3s box[2], vec3s other[2]) {
   vec3 rawBox[2];
   vec3 rawOther[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
-  glms_vec3_unpack(rawOther, other, 2);
+  glms_vec3_(unpack)(rawOther, other, 2);
   return glm_aabb_aabb(rawBox, rawOther);
 }
 
@@ -220,7 +220,7 @@ bool
 glms_aabb_(sphere)(vec3s box[2], vec4s s) {
   vec3 rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   return glm_aabb_sphere(rawBox, s.raw);
 }
 
@@ -235,7 +235,7 @@ bool
 glms_aabb_(point)(vec3s box[2], vec3s point) {
   vec3 rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   return glm_aabb_point(rawBox, point.raw);
 }
 
@@ -251,8 +251,8 @@ glms_aabb_(contains)(vec3s box[2], vec3s other[2]) {
   vec3 rawBox[2];
   vec3 rawOther[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
-  glms_vec3_unpack(rawOther, other, 2);
+  glms_vec3_(unpack)(rawOther, other, 2);
   return glm_aabb_contains(rawBox, rawOther);
 }
 

include/cglm/struct/clipspace/ortho_lh_no.h

@@ -64,7 +64,7 @@ glms_ortho_aabb_lh_no(vec3s box[2]) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_lh_no(rawBox, dest.raw);
 
   return dest;
@@ -87,7 +87,7 @@ glms_ortho_aabb_p_lh_no(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_p_lh_no(rawBox, padding, dest.raw);
 
   return dest;
@@ -110,7 +110,7 @@ glms_ortho_aabb_pz_lh_no(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_pz_lh_no(rawBox, padding, dest.raw);
 
   return dest;

include/cglm/struct/clipspace/ortho_lh_zo.h

@@ -64,7 +64,7 @@ glms_ortho_aabb_lh_zo(vec3s box[2]) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_lh_zo(rawBox, dest.raw);
 
   return dest;
@@ -87,7 +87,7 @@ glms_ortho_aabb_p_lh_zo(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_p_lh_zo(rawBox, padding, dest.raw);
 
   return dest;
@@ -110,7 +110,7 @@ glms_ortho_aabb_pz_lh_zo(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_pz_lh_zo(rawBox, padding, dest.raw);
 
   return dest;

include/cglm/struct/clipspace/ortho_rh_no.h

@@ -64,7 +64,7 @@ glms_ortho_aabb_rh_no(vec3s box[2]) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_rh_no(rawBox, dest.raw);
 
   return dest;
@@ -87,7 +87,7 @@ glms_ortho_aabb_p_rh_no(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_p_rh_no(rawBox, padding, dest.raw);
 
   return dest;
@@ -110,7 +110,7 @@ glms_ortho_aabb_pz_rh_no(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_pz_rh_no(rawBox, padding, dest.raw);
 
   return dest;

include/cglm/struct/clipspace/ortho_rh_zo.h

@@ -64,7 +64,7 @@ glms_ortho_aabb_rh_zo(vec3s box[2]) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_rh_zo(rawBox, dest.raw);
 
   return dest;
@@ -87,7 +87,7 @@ glms_ortho_aabb_p_rh_zo(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_p_rh_zo(rawBox, padding, dest.raw);
 
   return dest;
@@ -110,7 +110,7 @@ glms_ortho_aabb_pz_rh_zo(vec3s box[2], float padding) {
   mat4s dest;
   vec3  rawBox[2];
 
-  glms_vec3_unpack(rawBox, box, 2);
+  glms_vec3_(unpack)(rawBox, box, 2);
   glm_ortho_aabb_pz_rh_zo(rawBox, padding, dest.raw);
 
   return dest;

include/cglm/struct/frustum.h

@@ -57,7 +57,7 @@ void
 glms_frustum_planes(mat4s m, vec4s dest[6]) {
   vec4 rawDest[6];
   glm_frustum_planes(m.raw, rawDest);
-  glms_vec4_pack(dest, rawDest, 6);
+  glms_vec4_(pack)(dest, rawDest, 6);
 }
 
 /*!
@@ -88,7 +88,7 @@ void
 glms_frustum_corners(mat4s invMat, vec4s dest[8]) {
   vec4 rawDest[8];
   glm_frustum_corners(invMat.raw, rawDest);
-  glms_vec4_pack(dest, rawDest, 8);
+  glms_vec4_(pack)(dest, rawDest, 8);
 }
 
 /*!
@@ -103,7 +103,7 @@ glms_frustum_center(vec4s corners[8]) {
   vec4 rawCorners[8];
   vec4s r;
 
-  glms_vec4_unpack(rawCorners, corners, 8);
+  glms_vec4_(unpack)(rawCorners, corners, 8);
   glm_frustum_center(rawCorners, r.raw);
   return r;
 }
@@ -121,9 +121,9 @@ glms_frustum_box(vec4s corners[8], mat4s m, vec3s box[2]) {
   vec4 rawCorners[8];
   vec3 rawBox[2];
 
-  glms_vec4_unpack(rawCorners, corners, 8);
+  glms_vec4_(unpack)(rawCorners, corners, 8);
   glm_frustum_box(rawCorners, m.raw, rawBox);
-  glms_vec3_pack(box, rawBox, 2);
+  glms_vec3_(pack)(box, rawBox, 2);
 }
 
 /*!
@@ -147,9 +147,9 @@ glms_frustum_corners_at(vec4s corners[8],
   vec4 rawCorners[8];
   vec4 rawPlaneCorners[4];
 
-  glms_vec4_unpack(rawCorners, corners, 8);
+  glms_vec4_(unpack)(rawCorners, corners, 8);
   glm_frustum_corners_at(rawCorners, splitDist, farDist, rawPlaneCorners);
-  glms_vec4_pack(planeCorners, rawPlaneCorners, 8);
+  glms_vec4_(pack)(planeCorners, rawPlaneCorners, 8);
 }
 
 #endif /* cglms_frustums_h */

include/cglm/struct/io.h

@@ -75,7 +75,7 @@ glms_aabb_print(vec3s                   bbox[2],
                 FILE       * __restrict ostream) {
   vec3 rawBbox[2];
 
-  glms_vec3_unpack(rawBbox, bbox, 2);
+  glms_vec3_(unpack)(rawBbox, bbox, 2);
   glm_aabb_print(rawBbox, tag, ostream);
 }
 

include/cglm/struct/mat4.h

@@ -240,7 +240,7 @@ glms_mat4_(mulN)(mat4s * __restrict matrices[], uint32_t len) {
   size_t i;
 
   for (i = 0; i < len; i++) {
-    r = glms_mat4_mul(r, *matrices[i]);
+    r = glms_mat4_(mul)(r, *matrices[i]);
   }
   
   return r;

include/cglm/vec4-ext.h

@@ -45,7 +45,9 @@
 CGLM_INLINE
 void
 glm_vec4_broadcast(float val, vec4 d) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(d, wasm_f32x4_splat(val));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(d, _mm_set1_ps(val));
 #else
   d[0] = d[1] = d[2] = d[3] = val;
@@ -61,7 +63,9 @@ glm_vec4_broadcast(float val, vec4 d) {
 CGLM_INLINE
 void
 glm_vec4_fill(vec4 v, float val) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(v, wasm_f32x4_splat(val));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(v, _mm_set1_ps(val));
 #else
   v[0] = v[1] = v[2] = v[3] = val;
@@ -247,7 +251,9 @@ glm_vec4_sign(vec4 v, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_abs(vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, glmm_abs(glmm_load(v)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, glmm_abs(glmm_load(v)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vabsq_f32(vld1q_f32(v)));
@@ -284,7 +290,9 @@ glm_vec4_fract(vec4 v, vec4 dest) {
 CGLM_INLINE
 float
 glm_vec4_hadd(vec4 v) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  return glmm_hadd(glmm_load(v));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   return glmm_hadd(glmm_load(v));
 #else
   return v[0] + v[1] + v[2] + v[3];
@@ -300,7 +308,9 @@ glm_vec4_hadd(vec4 v) {
 CGLM_INLINE
 void
 glm_vec4_sqrt(vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_sqrt(glmm_load(v)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_sqrt_ps(glmm_load(v)));
 #else
   dest[0] = sqrtf(v[0]);

include/cglm/vec4.h

@@ -137,7 +137,9 @@ glm_vec4_copy3(vec4 a, vec3 dest) {
 CGLM_INLINE
 void
 glm_vec4_copy(vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, glmm_load(v));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, glmm_load(v));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vld1q_f32(v));
@@ -160,10 +162,15 @@ glm_vec4_copy(vec4 v, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_ucopy(vec4 v, vec4 dest) {
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  /* note here wasm v128.load/v128.store support unaligned loads and stores */
+  wasm_v128_store(dest, wasm_v128_load(v));
+#else
   dest[0] = v[0];
   dest[1] = v[1];
   dest[2] = v[2];
   dest[3] = v[3];
+#endif
 }
 
 /*!
@@ -174,7 +181,9 @@ glm_vec4_ucopy(vec4 v, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_zero(vec4 v) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(v, wasm_f32x4_const_splat(0.f));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(v, _mm_setzero_ps());
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(v, vdupq_n_f32(0.0f));
@@ -194,7 +203,9 @@ glm_vec4_zero(vec4 v) {
 CGLM_INLINE
 void
 glm_vec4_one(vec4 v) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(v, wasm_f32x4_const_splat(1.0f));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(v, _mm_set1_ps(1.0f));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(v, vdupq_n_f32(1.0f));
@@ -320,7 +331,9 @@ glm_vec4_norm_inf(vec4 v) {
 CGLM_INLINE
 void
 glm_vec4_add(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vaddq_f32(vld1q_f32(a), vld1q_f32(b)));
@@ -342,7 +355,9 @@ glm_vec4_add(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_adds(vec4 v, float s, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(glmm_load(v), wasm_f32x4_splat(s)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(v), _mm_set1_ps(s)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vaddq_f32(vld1q_f32(v), vdupq_n_f32(s)));
@@ -364,7 +379,9 @@ glm_vec4_adds(vec4 v, float s, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_sub(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_sub(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_sub_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vsubq_f32(vld1q_f32(a), vld1q_f32(b)));
@@ -386,7 +403,9 @@ glm_vec4_sub(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_subs(vec4 v, float s, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_sub(glmm_load(v), wasm_f32x4_splat(s)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_sub_ps(glmm_load(v), _mm_set1_ps(s)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vsubq_f32(vld1q_f32(v), vdupq_n_f32(s)));
@@ -408,7 +427,9 @@ glm_vec4_subs(vec4 v, float s, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_mul(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_mul(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_mul_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vmulq_f32(vld1q_f32(a), vld1q_f32(b)));
@@ -430,7 +451,9 @@ glm_vec4_mul(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_scale(vec4 v, float s, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_mul(glmm_load(v), wasm_f32x4_splat(s)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_mul_ps(glmm_load(v), _mm_set1_ps(s)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vmulq_f32(vld1q_f32(v), vdupq_n_f32(s)));
@@ -493,7 +516,9 @@ glm_vec4_div(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_divs(vec4 v, float s, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_div(glmm_load(v), wasm_f32x4_splat(s)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_div_ps(glmm_load(v), _mm_set1_ps(s)));
 #else
   glm_vec4_scale(v, 1.0f / s, dest);
@@ -512,7 +537,11 @@ glm_vec4_divs(vec4 v, float s, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_addadd(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(
+          glmm_load(dest),
+          wasm_f32x4_add(glmm_load(a), glmm_load(b))));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(dest),
                               _mm_add_ps(glmm_load(a),
                                          glmm_load(b))));
@@ -540,7 +569,11 @@ glm_vec4_addadd(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_subadd(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(
+          glmm_load(dest),
+          wasm_f32x4_sub(glmm_load(a), glmm_load(b))));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(dest),
                               _mm_sub_ps(glmm_load(a),
                                          glmm_load(b))));
@@ -612,7 +645,11 @@ glm_vec4_muladds(vec4 a, float s, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_maxadd(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(
+          glmm_load(dest),
+          wasm_f32x4_max(glmm_load(a), glmm_load(b))));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(dest),
                               _mm_max_ps(glmm_load(a),
                                          glmm_load(b))));
@@ -640,7 +677,11 @@ glm_vec4_maxadd(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_minadd(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_add(
+          glmm_load(dest),
+          wasm_f32x4_min(glmm_load(a), glmm_load(b))));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_add_ps(glmm_load(dest),
                               _mm_min_ps(glmm_load(a),
                                          glmm_load(b))));
@@ -665,8 +706,10 @@ glm_vec4_minadd(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_negate_to(vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
-  glmm_store(dest, _mm_xor_ps(glmm_load(v), _mm_set1_ps(-0.0f)));
+  glmm_store(dest, wasm_f32x4_neg(glmm_load(v)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
+  glmm_store(dest, _mm_xor_ps(glmm_load(v), glmm_float32x4_SIGNMASK_NEG));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vnegq_f32(vld1q_f32(v)));
 #else
@@ -697,7 +740,22 @@ glm_vec4_negate(vec4 v) {
 CGLM_INLINE
 void
 glm_vec4_normalize_to(vec4 v, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_128 xdot, x0;
+  float  dot;
+
+  x0   = glmm_load(v);
+  xdot = glmm_vdot(x0, x0);
+  /* dot  = _mm_cvtss_f32(xdot); */
+  dot  = wasm_f32x4_extract_lane(xdot, 0);
+
+  if (dot == 0.0f) {
+    glmm_store(dest, wasm_f32x4_const_splat(0.f));
+    return;
+  }
+
+  glmm_store(dest, wasm_f32x4_div(x0, wasm_f32x4_sqrt(xdot)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   __m128 xdot, x0;
   float  dot;
 
@@ -746,7 +804,9 @@ glm_vec4_normalize(vec4 v) {
 CGLM_INLINE
 float
 glm_vec4_distance(vec4 a, vec4 b) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  return glmm_norm(wasm_f32x4_sub(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   return glmm_norm(_mm_sub_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   return glmm_norm(vsubq_f32(glmm_load(a), glmm_load(b)));
@@ -768,7 +828,9 @@ glm_vec4_distance(vec4 a, vec4 b) {
 CGLM_INLINE
 float
 glm_vec4_distance2(vec4 a, vec4 b) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  return glmm_norm2(wasm_f32x4_sub(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   return glmm_norm2(_mm_sub_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   return glmm_norm2(vsubq_f32(glmm_load(a), glmm_load(b)));
@@ -790,7 +852,9 @@ glm_vec4_distance2(vec4 a, vec4 b) {
 CGLM_INLINE
 void
 glm_vec4_maxv(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_max(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_max_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vmaxq_f32(vld1q_f32(a), vld1q_f32(b)));
@@ -812,7 +876,9 @@ glm_vec4_maxv(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_minv(vec4 a, vec4 b, vec4 dest) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(dest, wasm_f32x4_min(glmm_load(a), glmm_load(b)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(dest, _mm_min_ps(glmm_load(a), glmm_load(b)));
 #elif defined(CGLM_NEON_FP)
   vst1q_f32(dest, vminq_f32(vld1q_f32(a), vld1q_f32(b)));
@@ -834,7 +900,11 @@ glm_vec4_minv(vec4 a, vec4 b, vec4 dest) {
 CGLM_INLINE
 void
 glm_vec4_clamp(vec4 v, float minVal, float maxVal) {
-#if defined( __SSE__ ) || defined( __SSE2__ )
+#if defined(__wasm__) && defined(__wasm_simd128__)
+  glmm_store(v, wasm_f32x4_min(
+          wasm_f32x4_max(glmm_load(v), wasm_f32x4_splat(minVal)),
+          wasm_f32x4_splat(maxVal)));
+#elif defined( __SSE__ ) || defined( __SSE2__ )
   glmm_store(v, _mm_min_ps(_mm_max_ps(glmm_load(v), _mm_set1_ps(minVal)),
                            _mm_set1_ps(maxVal)));
 #elif defined(CGLM_NEON_FP)

meson.build

@@ -108,11 +108,6 @@ if get_option('build_tests') == true
 test_src = files(
     'test/runner.c',
     'test/src/test_bezier.c',
-    'test/src/test_cam.c',
-    'test/src/test_cam_lh_no.c',
-    'test/src/test_cam_lh_zo.c',
-    'test/src/test_cam_rh_no.c',
-    'test/src/test_cam_rh_zo.c',
     'test/src/test_clamp.c',
     'test/src/test_common.c',
     'test/src/test_euler.c',

test/CMakeLists.txt

@@ -5,11 +5,6 @@ set(TESTFILES
   runner.c
   src/test_euler.c
   src/test_bezier.c
-  src/test_cam.c
-  src/test_cam_lh_zo.c
-  src/test_cam_rh_zo.c
-  src/test_cam_lh_no.c
-  src/test_cam_rh_no.c
   src/test_struct.c
   src/test_clamp.c
   src/test_common.c
@@ -22,6 +17,11 @@ set(TEST_RUNNER_PARAMS "")
 add_executable(${TEST_MAIN} ${TESTFILES})
 target_compile_definitions(${TEST_MAIN} PRIVATE CGLM_DEFINE_PRINTS=1)
 
+if(CMAKE_SYSTEM_NAME STREQUAL WASI)
+  target_compile_definitions(${TEST_MAIN} PRIVATE _WASI_EMULATED_PROCESS_CLOCKS=1)
+  target_link_options(${TEST_MAIN} PRIVATE "-lwasi-emulated-process-clocks")
+endif()
+
 if(NOT MSVC)
   target_link_libraries(${TEST_MAIN} PRIVATE m)
 endif()

test/src/test_cam.c

@@ -1,55 +0,0 @@
-/*
- * Copyright (c), Recep Aslantas.
- *
- * MIT License (MIT), http://opensource.org/licenses/MIT
- * Full license can be found in the LICENSE file
- */
-
-#include "test_common.h"
-
-TEST_IMPL(camera_lookat) {
-  mat4 view1, view2;
-  vec3 center,
-       eye = {0.024f, 14.6f, 67.04f},
-       dir = {0.0f, 0.0f, -1.0f},
-       up  = {0.0f, 1.0f, 0.0f};
-
-  glm_vec3_add(eye, dir, center);
-  glm_lookat(eye, center, up, view1);
-
-  glm_look(eye, dir, up, view2);
-
-  ASSERTIFY(test_assert_mat4_eq(view1, view2))
-
-  TEST_SUCCESS
-}
-
-TEST_IMPL(camera_decomp) {
-  mat4  proj, proj2;
-  vec4  sizes;
-  float aspect, fovy, nearZ, farZ;
-
-  aspect  = 0.782f;
-  fovy    = glm_rad(49.984f);
-  nearZ   = 0.1f;
-  farZ    = 100.0f;
-
-  glm_perspective(fovy, aspect, nearZ, farZ, proj);
-  ASSERT(fabsf(aspect  - glm_persp_aspect(proj)) < GLM_FLT_EPSILON)
-  ASSERT(fabsf(fovy    - glm_persp_fovy(proj))   < GLM_FLT_EPSILON)
-  ASSERT(fabsf(49.984f - glm_deg(glm_persp_fovy(proj))) < GLM_FLT_EPSILON)
-
-  glm_persp_sizes(proj, fovy, sizes);
-
-  glm_frustum(-sizes[0] * 0.5f,
-               sizes[0] * 0.5f,
-              -sizes[1] * 0.5f,
-               sizes[1] * 0.5f,
-               nearZ,
-               farZ,
-               proj2);
-
-  ASSERTIFY(test_assert_mat4_eq(proj, proj2))
-
-  TEST_SUCCESS
-}

test/src/test_cam.h

@@ -66,3 +66,50 @@ TEST_IMPL(GLM_PREFIX, frustum) {
 
   TEST_SUCCESS
 }
+
+TEST_IMPL(GLM_PREFIX, camera_lookat) {
+  mat4 view1, view2;
+  vec3 center,
+       eye = {0.024f, 14.6f, 67.04f},
+       dir = {0.0f, 0.0f, -1.0f},
+       up  = {0.0f, 1.0f, 0.0f};
+
+  glm_vec3_add(eye, dir, center);
+  glm_lookat(eye, center, up, view1);
+
+  glm_look(eye, dir, up, view2);
+
+  ASSERTIFY(test_assert_mat4_eq(view1, view2))
+
+  TEST_SUCCESS
+}
+
+TEST_IMPL(GLM_PREFIX, camera_decomp) {
+  mat4  proj, proj2;
+  vec4  sizes;
+  float aspect, fovy, nearZ, farZ;
+
+  aspect  = 0.782f;
+  fovy    = glm_rad(49.984f);
+  nearZ   = 0.1f;
+  farZ    = 100.0f;
+
+  glm_perspective(fovy, aspect, nearZ, farZ, proj);
+  ASSERT(fabsf(aspect  - glm_persp_aspect(proj)) < GLM_FLT_EPSILON)
+  ASSERT(fabsf(fovy    - glm_persp_fovy(proj))   < GLM_FLT_EPSILON)
+  ASSERT(fabsf(49.984f - glm_deg(glm_persp_fovy(proj))) < GLM_FLT_EPSILON)
+
+  glm_persp_sizes(proj, fovy, sizes);
+
+  glm_frustum(-sizes[0] * 0.5f,
+               sizes[0] * 0.5f,
+              -sizes[1] * 0.5f,
+               sizes[1] * 0.5f,
+               nearZ,
+               farZ,
+               proj2);
+
+  ASSERTIFY(test_assert_mat4_eq(proj, proj2))
+
+  TEST_SUCCESS
+}

test/src/test_cam_lh_no.h

@@ -6,15 +6,17 @@
  */
 
 #include "test_common.h"
+#include "../../include/cglm/clipspace/persp_lh_no.h"
+#include "../../include/cglm/call/clipspace/persp_lh_no.h"
 
-TEST_IMPL(perspective_lh_no) {
+TEST_IMPL(GLM_PREFIX, perspective_lh_no) {
   mat4 dst;
   const float fovy     = glm_rad(45.0f);
   const float aspect   = 640/480.0f;
   const float zNearVal = 0.1f;
   const float zFarVal  = 100.0f;
 
-  glm_perspective_lh_no(fovy, aspect, zNearVal, zFarVal, dst);
+  GLM(perspective_lh_no)(fovy, aspect, zNearVal, zFarVal, dst);
 
   /* Sanity mk. I: longhand version */
   ASSERT(test_eq(dst[0][0],  1.0f / (tanf(fovy / 2) * aspect)))

test/src/test_cam_lh_zo.h

@@ -6,15 +6,17 @@
  */
 
 #include "test_common.h"
+#include "../../include/cglm/clipspace/persp_lh_zo.h"
+#include "../../include/cglm/call/clipspace/persp_lh_zo.h"
 
-TEST_IMPL(perspective_lh_zo) {
+TEST_IMPL(GLM_PREFIX, perspective_lh_zo) {
   mat4 dst;
   const float fovy     = glm_rad(45.0f);
   const float aspect   = 640/480.0f;
   const float zNearVal = 0.1f;
   const float zFarVal  = 100.0f;
 
-  glm_perspective_lh_zo(fovy, aspect, zNearVal, zFarVal, dst);
+  GLM(perspective_lh_zo)(fovy, aspect, zNearVal, zFarVal, dst);
 
   /* Sanity mk. I: longhand version */
   ASSERT(test_eq(dst[0][0],  1.0f / (tanf(fovy / 2) * aspect)))

test/src/test_cam_rh_no.h

@@ -6,15 +6,17 @@
  */
 
 #include "test_common.h"
+#include "../../include/cglm/clipspace/persp_rh_no.h"
+#include "../../include/cglm/call/clipspace/persp_rh_no.h"
 
-TEST_IMPL(perspective_rh_no) {
+TEST_IMPL(GLM_PREFIX, perspective_rh_no) {
   mat4 dst;
   const float fovy     = glm_rad(45.0f);
   const float aspect   = 640/480.0f;
   const float zNearVal = 0.1f;
   const float zFarVal  = 100.0f;
 
-  glm_perspective_rh_no(fovy, aspect, zNearVal, zFarVal, dst);
+  GLM(perspective_rh_no)(fovy, aspect, zNearVal, zFarVal, dst);
 
   /* Sanity mk. I: longhand version */
   ASSERT(test_eq(dst[0][0],  1.0f / (tanf(fovy / 2) * aspect)))

test/src/test_cam_rh_zo.h

@@ -6,15 +6,17 @@
  */
 
 #include "test_common.h"
+#include "../../include/cglm/clipspace/persp_rh_zo.h"
+#include "../../include/cglm/call/clipspace/persp_rh_zo.h"
 
-TEST_IMPL(perspective_rh_zo) {
+TEST_IMPL(GLM_PREFIX, perspective_rh_zo) {
   mat4 dst;
   const float fovy     = glm_rad(45.0f);
   const float aspect   = 640/480.0f;
   const float zNearVal = 0.1f;
   const float zFarVal  = 100.0f;
 
-  glm_perspective_rh_zo(fovy, aspect, zNearVal, zFarVal, dst);
+  GLM(perspective_rh_zo)(fovy, aspect, zNearVal, zFarVal, dst);
 
   /* Sanity mk. I: longhand version */
   ASSERT(test_eq(dst[0][0],  1 / (tanf(fovy / 2) * aspect)))

test/src/test_common.h

@@ -10,6 +10,10 @@
 
 #include "../include/common.h"
 
+#if !defined(_WIN32) && !defined(_MSC_VER)
+#  pragma GCC diagnostic ignored "-Wstrict-prototypes"
+#endif
+
 void
 test_rand_mat4(mat4 dest);
 

test/src/tests.c

@@ -28,7 +28,11 @@
 #include "test_affine2d.h"
 #include "test_affine_mat.h"
 #include "test_ray.h"
-#include "test_camera.h"
+#include "test_cam.h"
+#include "test_cam_lh_no.h"
+#include "test_cam_lh_zo.h"
+#include "test_cam_rh_no.h"
+#include "test_cam_rh_zo.h"
 
 #undef GLM
 #undef GLM_PREFIX
@@ -55,7 +59,11 @@
 #include "test_affine2d.h"
 #include "test_affine_mat.h"
 #include "test_ray.h"
-#include "test_camera.h"
+#include "test_cam.h"
+#include "test_cam_lh_no.h"
+#include "test_cam_lh_zo.h"
+#include "test_cam_rh_no.h"
+#include "test_cam_rh_zo.h"
 
 #undef GLM
 #undef GLM_PREFIX

test/tests.h

@@ -223,12 +223,19 @@ TEST_DECLARE(glmc_mat2_swap_row)
 TEST_DECLARE(glmc_mat2_rmc)
 
 /* camera (incl [LR]H cross [NZ]O) */
-TEST_DECLARE(perspective_lh_zo)
+TEST_DECLARE(glm_perspective_lh_zo)
-TEST_DECLARE(perspective_rh_zo)
+TEST_DECLARE(glm_perspective_rh_zo)
-TEST_DECLARE(perspective_lh_no)
+TEST_DECLARE(glm_perspective_lh_no)
-TEST_DECLARE(perspective_rh_no)
+TEST_DECLARE(glm_perspective_rh_no)
-TEST_DECLARE(camera_lookat)
+TEST_DECLARE(glm_camera_lookat)
-TEST_DECLARE(camera_decomp)
+TEST_DECLARE(glm_camera_decomp)
+
+TEST_DECLARE(glmc_perspective_lh_zo)
+TEST_DECLARE(glmc_perspective_rh_zo)
+TEST_DECLARE(glmc_perspective_lh_no)
+TEST_DECLARE(glmc_perspective_rh_no)
+TEST_DECLARE(glmc_camera_lookat)
+TEST_DECLARE(glmc_camera_decomp)
 
 TEST_DECLARE(glm_frustum)
 
@@ -1057,12 +1064,19 @@ TEST_LIST {
   TEST_ENTRY(glmc_mat2_rmc)
 
   /* camera (incl [LR]H cross [NZ]O) */
-  TEST_ENTRY(perspective_lh_zo)
+  TEST_ENTRY(glm_perspective_lh_zo)
-  TEST_ENTRY(perspective_rh_zo)
+  TEST_ENTRY(glm_perspective_rh_zo)
-  TEST_ENTRY(perspective_lh_no)
+  TEST_ENTRY(glm_perspective_lh_no)
-  TEST_ENTRY(perspective_rh_no)
+  TEST_ENTRY(glm_perspective_rh_no)
-  TEST_ENTRY(camera_lookat)
+  TEST_ENTRY(glm_camera_lookat)
-  TEST_ENTRY(camera_decomp)
+  TEST_ENTRY(glm_camera_decomp)
+
+  TEST_ENTRY(glmc_perspective_lh_zo)
+  TEST_ENTRY(glmc_perspective_rh_zo)
+  TEST_ENTRY(glmc_perspective_lh_no)
+  TEST_ENTRY(glmc_perspective_rh_no)
+  TEST_ENTRY(glmc_camera_lookat)
+  TEST_ENTRY(glmc_camera_decomp)
   
   TEST_ENTRY(glm_frustum)
   

win/cglm-test.vcxproj

@@ -46,11 +46,6 @@
     <ClCompile Include="..\test\runner.c" />
     <ClCompile Include="..\test\src\tests.c" />
     <ClCompile Include="..\test\src\test_bezier.c" />
-    <ClCompile Include="..\test\src\test_cam.c" />
-    <ClCompile Include="..\test\src\test_cam_lh_no.c" />
-    <ClCompile Include="..\test\src\test_cam_lh_zo.c" />
-    <ClCompile Include="..\test\src\test_cam_rh_no.c" />
-    <ClCompile Include="..\test\src\test_cam_rh_zo.c" />
     <ClCompile Include="..\test\src\test_clamp.c" />
     <ClCompile Include="..\test\src\test_common.c" />
     <ClCompile Include="..\test\src\test_euler.c" />
@@ -61,7 +56,11 @@
     <ClInclude Include="..\test\src\test_affine.h" />
     <ClInclude Include="..\test\src\test_affine2d.h" />
     <ClInclude Include="..\test\src\test_affine_mat.h" />
-    <ClInclude Include="..\test\src\test_camera.h" />
+    <ClInclude Include="..\test\src\test_cam.h" />
+    <ClInclude Include="..\test\src\test_cam_lh_no.h" />
+    <ClInclude Include="..\test\src\test_cam_lh_zo.h" />
+    <ClInclude Include="..\test\src\test_cam_rh_no.h" />
+    <ClInclude Include="..\test\src\test_cam_rh_zo.h" />
     <ClInclude Include="..\test\src\test_common.h" />
     <ClInclude Include="..\test\src\test_ivec2.h" />
     <ClInclude Include="..\test\src\test_ivec3.h" />
@@ -72,10 +71,10 @@
     <ClInclude Include="..\test\src\test_plane.h" />
     <ClInclude Include="..\test\src\test_project.h" />
     <ClInclude Include="..\test\src\test_quat.h" />
+    <ClInclude Include="..\test\src\test_ray.h" />
     <ClInclude Include="..\test\src\test_vec2.h" />
     <ClInclude Include="..\test\src\test_vec3.h" />
     <ClInclude Include="..\test\src\test_vec4.h" />
-    <ClInclude Include="..\test\src\test_ray.h" />
     <ClInclude Include="..\test\tests.h" />
   </ItemGroup>
   <ItemGroup>

win/cglm-test.vcxproj.filters

@@ -17,9 +17,6 @@
     <ClCompile Include="..\test\src\test_bezier.c">
       <Filter>src</Filter>
     </ClCompile>
-    <ClCompile Include="..\test\src\test_cam.c">
-      <Filter>src</Filter>
-    </ClCompile>
     <ClCompile Include="..\test\src\test_clamp.c">
       <Filter>src</Filter>
     </ClCompile>
@@ -35,18 +32,6 @@
     <ClCompile Include="..\test\src\tests.c">
       <Filter>src</Filter>
     </ClCompile>
-    <ClCompile Include="..\test\src\test_cam_lh_no.c">
-      <Filter>src</Filter>
-    </ClCompile>
-    <ClCompile Include="..\test\src\test_cam_lh_zo.c">
-      <Filter>src</Filter>
-    </ClCompile>
-    <ClCompile Include="..\test\src\test_cam_rh_no.c">
-      <Filter>src</Filter>
-    </ClCompile>
-    <ClCompile Include="..\test\src\test_cam_rh_zo.c">
-      <Filter>src</Filter>
-    </ClCompile>
   </ItemGroup>
   <ItemGroup>
     <ClInclude Include="..\test\tests.h">
@@ -58,46 +43,28 @@
     <ClInclude Include="..\test\include\common.h">
       <Filter>include</Filter>
     </ClInclude>
-    <ClInclude Include="..\test\src\test_mat3.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_mat4.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_project.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_quat.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_vec3.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_vec4.h">
-      <Filter>src</Filter>
-    </ClInclude>
     <ClInclude Include="..\test\src\test_affine.h">
       <Filter>src</Filter>
     </ClInclude>
     <ClInclude Include="..\test\src\test_affine_mat.h">
       <Filter>src</Filter>
     </ClInclude>
-    <ClInclude Include="..\test\src\test_mat2.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_plane.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_vec2.h">
-      <Filter>src</Filter>
-    </ClInclude>
-    <ClInclude Include="..\test\src\test_ray.h">
-      <Filter>src</Filter>
-    </ClInclude>
     <ClInclude Include="..\test\src\test_affine2d.h">
       <Filter>src</Filter>
     </ClInclude>
-    <ClInclude Include="..\test\src\test_camera.h">
+    <ClInclude Include="..\test\src\test_cam.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_cam_lh_no.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_cam_lh_zo.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_cam_rh_no.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_cam_rh_zo.h">
       <Filter>src</Filter>
     </ClInclude>
     <ClInclude Include="..\test\src\test_ivec2.h">
@@ -109,5 +76,35 @@
     <ClInclude Include="..\test\src\test_ivec4.h">
       <Filter>src</Filter>
     </ClInclude>
+    <ClInclude Include="..\test\src\test_mat2.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_mat3.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_mat4.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_plane.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_project.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_quat.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_ray.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_vec2.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_vec3.h">
+      <Filter>src</Filter>
+    </ClInclude>
+    <ClInclude Include="..\test\src\test_vec4.h">
+      <Filter>src</Filter>
+    </ClInclude>
   </ItemGroup>
 </Project>