From 6e9e91be051b507e95912664825cb1a1b203d1ff Mon Sep 17 00:00:00 2001
From: Recep Aslantas <info@recp.me>
Date: Sat, 22 Jul 2023 12:00:23 +0300
Subject: [PATCH] add _mul for non-square matrices

---
 include/cglm/call/mat2x3.h |  4 ++++
 include/cglm/call/mat2x4.h |  4 ++++
 include/cglm/call/mat3x2.h |  4 ++++
 include/cglm/call/mat3x4.h |  4 ++++
 include/cglm/call/mat4x2.h |  4 ++++
 include/cglm/call/mat4x3.h |  4 ++++
 include/cglm/mat2x3.h      | 30 +++++++++++++++++++++++++
 include/cglm/mat2x4.h      | 32 +++++++++++++++++++++++++++
 include/cglm/mat3x2.h      | 36 ++++++++++++++++++++++++++++++
 include/cglm/mat3x4.h      | 38 ++++++++++++++++++++++++++++++++
 include/cglm/mat4x2.h      | 45 ++++++++++++++++++++++++++++++++++++++
 include/cglm/mat4x3.h      | 45 ++++++++++++++++++++++++++++++++++++++
 src/mat2x3.c               |  6 +++++
 src/mat2x4.c               |  6 +++++
 src/mat3x2.c               |  6 +++++
 src/mat3x4.c               |  6 +++++
 src/mat4x2.c               |  6 +++++
 src/mat4x3.c               |  6 +++++
 18 files changed, 286 insertions(+)

diff --git a/include/cglm/call/mat2x3.h b/include/cglm/call/mat2x3.h
index c7f0946..4e106e2 100644
--- a/include/cglm/call/mat2x3.h
+++ b/include/cglm/call/mat2x3.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat2x3_make(float * __restrict src, mat2x3 dest);
 
+CGLM_EXPORT
+void
+glmc_mat2x3_mul(mat2x3 m1, mat3x2 m2, mat2 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/call/mat2x4.h b/include/cglm/call/mat2x4.h
index c2b6495..ead2b9c 100644
--- a/include/cglm/call/mat2x4.h
+++ b/include/cglm/call/mat2x4.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat2x4_make(float * __restrict src, mat2x4 dest);
 
+CGLM_EXPORT
+void
+glmc_mat2x4_mul(mat2x4 m1, mat4x2 m2, mat2 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/call/mat3x2.h b/include/cglm/call/mat3x2.h
index 491151b..78dcab4 100644
--- a/include/cglm/call/mat3x2.h
+++ b/include/cglm/call/mat3x2.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat3x2_make(float * __restrict src, mat3x2 dest);
 
+CGLM_EXPORT
+void
+glmc_mat3x2_mul(mat3x2 m1, mat2x3 m2, mat3 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/call/mat3x4.h b/include/cglm/call/mat3x4.h
index 714d744..bb4c20b 100644
--- a/include/cglm/call/mat3x4.h
+++ b/include/cglm/call/mat3x4.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat3x4_make(float * __restrict src, mat3x4 dest);
 
+CGLM_EXPORT
+void
+glmc_mat3x4_mul(mat3x4 m1, mat4x3 m2, mat3 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/call/mat4x2.h b/include/cglm/call/mat4x2.h
index 3710745..05875b5 100644
--- a/include/cglm/call/mat4x2.h
+++ b/include/cglm/call/mat4x2.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat4x2_make(float * __restrict src, mat4x2 dest);
 
+CGLM_EXPORT
+void
+glmc_mat4x2_mul(mat4x2 m1, mat2x4 m2, mat4 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/call/mat4x3.h b/include/cglm/call/mat4x3.h
index 7790aa8..a0b1af6 100644
--- a/include/cglm/call/mat4x3.h
+++ b/include/cglm/call/mat4x3.h
@@ -25,6 +25,10 @@ CGLM_EXPORT
 void
 glmc_mat4x3_make(float * __restrict src, mat4x3 dest);
 
+CGLM_EXPORT
+void
+glmc_mat4x3_mul(mat4x3 m1, mat3x4 m2, mat4 dest);
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/include/cglm/mat2x3.h b/include/cglm/mat2x3.h
index 8ae65ea..fa507f9 100644
--- a/include/cglm/mat2x3.h
+++ b/include/cglm/mat2x3.h
@@ -72,4 +72,34 @@ glm_mat2x3_make(float * __restrict src, mat2x3 dest) {
   dest[1][2] = src[5];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat2x3_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat2x3_mul(mat2x3 m1, mat3x2 m2, mat2 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1],
+        a10 = m1[1][0], a11 = m1[1][1],
+        a20 = m1[2][0], a21 = m1[2][1],
+
+        b00 = m2[0][0], b01 = m2[0][1],
+        b10 = m2[1][0], b11 = m2[1][1],
+        b20 = m2[2][0], b21 = m2[2][1];
+
+  dest[0][0] = a00 * b00 + a10 * b01 + a20 * b20;
+  dest[0][1] = a00 * b10 + a10 * b11 + a20 * b21;
+  dest[1][0] = a01 * b00 + a11 * b01 + a21 * b20;
+  dest[1][1] = a01 * b10 + a11 * b11 + a21 * b21;
+}
+
 #endif
diff --git a/include/cglm/mat2x4.h b/include/cglm/mat2x4.h
index 4b6faa1..b8a1e62 100644
--- a/include/cglm/mat2x4.h
+++ b/include/cglm/mat2x4.h
@@ -70,4 +70,36 @@ glm_mat2x4_make(float * __restrict src, mat2x4 dest) {
   dest[1][3] = src[7];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat2x4_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat2x4_mul(mat2x4 m1, mat4x2 m2, mat2 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1],
+        a10 = m1[1][0], a11 = m1[1][1],
+        a20 = m1[2][0], a21 = m1[2][1],
+        a30 = m1[3][0], a31 = m1[3][1],
+
+        b00 = m2[0][0], b01 = m2[0][1],
+        b10 = m2[1][0], b11 = m2[1][1],
+        b20 = m2[2][0], b21 = m2[2][1],
+        b30 = m2[3][0], b31 = m2[3][1];
+
+  dest[0][0] = a00 * b00 + a10 * b01 + a20 * b20 + a30 * b30;
+  dest[0][1] = a00 * b10 + a10 * b11 + a20 * b21 + a30 * b31;
+  dest[1][0] = a01 * b00 + a11 * b01 + a21 * b20 + a31 * b30;
+  dest[1][1] = a01 * b10 + a11 * b11 + a21 * b21 + a31 * b31;
+}
+
 #endif
diff --git a/include/cglm/mat3x2.h b/include/cglm/mat3x2.h
index 0b10286..1ef742d 100644
--- a/include/cglm/mat3x2.h
+++ b/include/cglm/mat3x2.h
@@ -74,4 +74,40 @@ glm_mat3x2_make(float * __restrict src, mat3x2 dest) {
   dest[2][1] = src[5];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat3x2_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat3x2_mul(mat3x2 m1, mat2x3 m2, mat3 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1],
+        a10 = m1[1][0], a11 = m1[1][1],
+        a20 = m1[2][0], a21 = m1[2][1],
+
+        b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2],
+        b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2];
+
+  dest[0][0] = a00 * b00 + a10 * b10;
+  dest[0][1] = a00 * b01 + a10 * b11;
+  dest[0][2] = a00 * b02 + a10 * b12;
+
+  dest[1][0] = a01 * b00 + a11 * b10;
+  dest[1][1] = a01 * b01 + a11 * b11;
+  dest[1][2] = a01 * b02 + a11 * b12;
+
+  dest[2][0] = a20 * b00 + a21 * b10;
+  dest[2][1] = a20 * b01 + a21 * b11;
+  dest[2][2] = a20 * b02 + a21 * b12;
+}
+
 #endif
diff --git a/include/cglm/mat3x4.h b/include/cglm/mat3x4.h
index a22ca4c..ff963e7 100644
--- a/include/cglm/mat3x4.h
+++ b/include/cglm/mat3x4.h
@@ -77,4 +77,42 @@ glm_mat3x4_make(float * __restrict src, mat3x4 dest) {
   dest[2][3] = src[11];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat3x4_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat3x4_mul(mat3x4 m1, mat4x3 m2, mat3 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2], a03 = m1[0][3],
+        a10 = m1[1][0], a11 = m1[1][1], a12 = m1[1][2], a13 = m1[1][3],
+        a20 = m1[2][0], a21 = m1[2][1], a22 = m1[2][2], a23 = m1[2][3],
+
+        b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2],
+        b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2],
+        b20 = m2[2][0], b21 = m2[2][1], b22 = m2[2][2],
+        b30 = m2[3][0], b31 = m2[3][1], b32 = m2[3][2];
+
+  dest[0][0] = a00 * b00 + a01 * b10 + a02 * b20 + a03 * b30;
+  dest[0][1] = a00 * b01 + a01 * b11 + a02 * b21 + a03 * b31;
+  dest[0][2] = a00 * b02 + a01 * b12 + a02 * b22 + a03 * b32;
+
+  dest[1][0] = a10 * b00 + a11 * b10 + a12 * b20 + a13 * b30;
+  dest[1][1] = a10 * b01 + a11 * b11 + a12 * b21 + a13 * b31;
+  dest[1][2] = a10 * b02 + a11 * b12 + a12 * b22 + a13 * b32;
+
+  dest[2][0] = a20 * b00 + a21 * b10 + a22 * b20 + a23 * b30;
+  dest[2][1] = a20 * b01 + a21 * b11 + a22 * b21 + a23 * b31;
+  dest[2][2] = a20 * b02 + a21 * b12 + a22 * b22 + a23 * b32;
+}
+
 #endif
diff --git a/include/cglm/mat4x2.h b/include/cglm/mat4x2.h
index b5074d3..5a5aca1 100644
--- a/include/cglm/mat4x2.h
+++ b/include/cglm/mat4x2.h
@@ -80,4 +80,49 @@ glm_mat4x2_make(float * __restrict src, mat4x2 dest) {
   dest[3][1] = src[7];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat4x2_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat4x2_mul(mat4x2 m1, mat2x4 m2, mat4 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1],
+        a10 = m1[1][0], a11 = m1[1][1],
+        a20 = m1[2][0], a21 = m1[2][1],
+        a30 = m1[3][0], a31 = m1[3][1],
+
+        b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2], b03 = m2[0][3],
+        b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2], b13 = m2[1][3];
+
+  dest[0][0] = a00 * b00 + a10 * b10;
+  dest[0][1] = a00 * b01 + a10 * b11;
+  dest[0][2] = a00 * b02 + a10 * b12;
+  dest[0][3] = a00 * b03 + a10 * b13;
+
+  dest[1][0] = a01 * b00 + a11 * b10;
+  dest[1][1] = a01 * b01 + a11 * b11;
+  dest[1][2] = a01 * b02 + a11 * b12;
+  dest[1][3] = a01 * b03 + a11 * b13;
+
+  dest[2][0] = a20 * b00 + a21 * b10;
+  dest[2][1] = a20 * b01 + a21 * b11;
+  dest[2][2] = a20 * b02 + a21 * b12;
+  dest[2][3] = a20 * b03 + a21 * b13;
+
+  dest[3][0] = a30 * b00 + a31 * b10;
+  dest[3][1] = a30 * b01 + a31 * b11;
+  dest[3][2] = a30 * b02 + a31 * b12;
+  dest[3][3] = a30 * b03 + a31 * b13;
+}
+
 #endif
diff --git a/include/cglm/mat4x3.h b/include/cglm/mat4x3.h
index b185d75..7dc0e19 100644
--- a/include/cglm/mat4x3.h
+++ b/include/cglm/mat4x3.h
@@ -85,4 +85,49 @@ glm_mat4x3_make(float * __restrict src, mat4x3 dest) {
   dest[3][2] = src[11];
 }
 
+/*!
+ * @brief multiply m1 and m2 to dest
+ *
+ * m1, m2 and dest matrices can be same matrix, it is possible to write this:
+ *
+ * @code
+ * glm_mat4x3_mul(m, m, m);
+ * @endcode
+ *
+ * @param[in]  m1   left matrix
+ * @param[in]  m2   right matrix
+ * @param[out] dest destination matrix
+ */
+CGLM_INLINE
+void
+glm_mat4x3_mul(mat4x3 m1, mat3x4 m2, mat4 dest) {
+  float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2],
+        a10 = m1[1][0], a11 = m1[1][1], a12 = m1[1][2],
+        a20 = m1[2][0], a21 = m1[2][1], a22 = m1[2][2],
+        a30 = m1[3][0], a31 = m1[3][1], a32 = m1[3][2],
+
+        b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2], b03 = m2[0][3],
+        b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2], b13 = m2[1][3],
+        b20 = m2[2][0], b21 = m2[2][1], b22 = m2[2][2], b23 = m2[2][3];
+
+  dest[0][0] = a00 * b00 + a01 * b10 + a02 * b20;
+  dest[0][1] = a00 * b01 + a01 * b11 + a02 * b21;
+  dest[0][2] = a00 * b02 + a01 * b12 + a02 * b22;
+  dest[0][3] = a00 * b03 + a01 * b13 + a02 * b23;
+
+  dest[1][0] = a10 * b00 + a11 * b10 + a12 * b20;
+  dest[1][1] = a10 * b01 + a11 * b11 + a12 * b21;
+  dest[1][2] = a10 * b02 + a11 * b12 + a12 * b22;
+  dest[1][3] = a10 * b03 + a11 * b13 + a12 * b23;
+
+  dest[2][0] = a20 * b00 + a21 * b10 + a22 * b20;
+  dest[2][1] = a20 * b01 + a21 * b11 + a22 * b21;
+  dest[2][2] = a20 * b02 + a21 * b12 + a22 * b22;
+  dest[2][3] = a20 * b03 + a21 * b13 + a22 * b23;
+
+  dest[3][0] = a30 * b00 + a31 * b10 + a32 * b20;
+  dest[3][1] = a30 * b01 + a31 * b11 + a32 * b21;
+  dest[3][2] = a30 * b02 + a31 * b12 + a32 * b22;
+  dest[3][3] = a30 * b03 + a31 * b13 + a32 * b23;
+}
 #endif
diff --git a/src/mat2x3.c b/src/mat2x3.c
index ab32870..8dff83f 100644
--- a/src/mat2x3.c
+++ b/src/mat2x3.c
@@ -25,3 +25,9 @@ void
 glmc_mat2x3_make(float * __restrict src, mat2x3 dest) {
   glm_mat2x3_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat2x3_mul(mat2x3 m1, mat3x2 m2, mat2 dest) {
+  glm_mat2x3_mul(m1, m2, dest);
+}
diff --git a/src/mat2x4.c b/src/mat2x4.c
index ad3ec0d..1b93520 100644
--- a/src/mat2x4.c
+++ b/src/mat2x4.c
@@ -25,3 +25,9 @@ void
 glmc_mat2x4_make(float * __restrict src, mat2x4 dest) {
   glm_mat2x4_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat2x4_mul(mat2x4 m1, mat4x2 m2, mat2 dest) {
+  glm_mat2x4_mul(m1, m2, dest);
+}
diff --git a/src/mat3x2.c b/src/mat3x2.c
index ac59ebf..377b411 100644
--- a/src/mat3x2.c
+++ b/src/mat3x2.c
@@ -25,3 +25,9 @@ void
 glmc_mat3x2_make(float * __restrict src, mat3x2 dest) {
   glm_mat3x2_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat3x2_mul(mat3x2 m1, mat2x3 m2, mat3 dest) {
+  glm_mat3x2_mul(m1, m2, dest);
+}
diff --git a/src/mat3x4.c b/src/mat3x4.c
index 64be161..f5f5f23 100644
--- a/src/mat3x4.c
+++ b/src/mat3x4.c
@@ -25,3 +25,9 @@ void
 glmc_mat3x4_make(float * __restrict src, mat3x4 dest) {
   glm_mat3x4_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat3x4_mul(mat3x4 m1, mat4x3 m2, mat3 dest) {
+  glm_mat3x4_mul(m1, m2, dest);
+}
diff --git a/src/mat4x2.c b/src/mat4x2.c
index 622d1e2..0c0bbb7 100644
--- a/src/mat4x2.c
+++ b/src/mat4x2.c
@@ -25,3 +25,9 @@ void
 glmc_mat4x2_make(float * __restrict src, mat4x2 dest) {
   glm_mat4x2_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat4x2_mul(mat4x2 m1, mat2x4 m2, mat4 dest) {
+  glm_mat4x2_mul(m1, m2, dest);
+}
diff --git a/src/mat4x3.c b/src/mat4x3.c
index 60a14fe..4d94b07 100644
--- a/src/mat4x3.c
+++ b/src/mat4x3.c
@@ -25,3 +25,9 @@ void
 glmc_mat4x3_make(float * __restrict src, mat4x3 dest) {
   glm_mat4x3_make(src, dest);
 }
+
+CGLM_EXPORT
+void
+glmc_mat4x3_mul(mat4x3 m1, mat3x4 m2, mat4 dest) {
+  glm_mat4x3_mul(m1, m2, dest);
+}