Merge pull request #88 from yushli/master

Update quat.h

.travis.yml

@@ -57,3 +57,6 @@ after_success:
         --gcov-options '\-lp'
         --verbose;
     fi
+
+after_failure:
+  - cat ./test-suite.log

configure.ac

@@ -7,7 +7,7 @@
 #*****************************************************************************
 
 AC_PREREQ([2.69])
-AC_INIT([cglm], [0.5.3], [info@recp.me])
+AC_INIT([cglm], [0.5.4], [info@recp.me])
 AM_INIT_AUTOMAKE([-Wall -Werror foreign subdir-objects])
 
 AC_CONFIG_MACRO_DIR([m4])

docs/source/build.rst

@@ -1,9 +1,7 @@
-Building cglm
+Build cglm
 ================================
 
-| **cglm** does not have external dependencies except for unit testing.
+| **cglm** does not have external dependencies except for unit testing. When you pulled **cglm** repo with submodules all dependencies will be pulled too. `build-deps.sh` will pull all dependencies/submodules and build for you.
-| When you pulled cglm repo with submodules all dependencies will be pulled too.
-| `build-deps.sh` will pull all dependencies/submodules and build for you.
 
 External dependencies:
   * cmocka - for unit testing
@@ -12,7 +10,8 @@ External dependencies:
 If you only need to inline versions, you don't need to build **cglm**, you don't need to link it to your program.
 Just import cglm to your project as dependency / external lib by copy-paste then use it as usual
 
-**Unix (Autotools):**
+Unix (Autotools):
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 .. code-block:: bash
   :linenos:
@@ -26,11 +25,12 @@ Just import cglm to your project as dependency / external lib by copy-paste then
   $ [sudo] make install   # install to system (optional)
 
 **make** will build cglm to **.libs** sub folder in project folder.
-If you don't want to install cglm to your system's folder you can get static and dynamic libs in this folder.
+If you don't want to install **cglm** to your system's folder you can get static and dynamic libs in this folder.
 
-**Build dependencies (windows):**
+Windows (MSBuild):
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Windows related build files, project files are located in win folder,
+Windows related build files, project files are located in `win` folder,
 make sure you are inside in cglm/win folder.
 
 Code Analysis are enabled, it may take awhile to build.
@@ -50,3 +50,19 @@ then try to build with *devenv*:
   $ devenv cglm.sln /Build Release
 
 Currently tests are not available on Windows.
+
+Documentation (Sphinx):
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+**cglm** uses sphinx framework for documentation, it allows lot of formats for documentation. To see all options see sphinx build page:
+
+https://www.sphinx-doc.org/en/master/man/sphinx-build.html
+
+Example build:
+
+.. code-block:: bash
+  :linenos:
+
+  $ cd cglm/docs
+  $ sphinx-build source build
+

docs/source/conf.py

@@ -62,9 +62,9 @@ author = u'Recep Aslantas'
 # built documents.
 #
 # The short X.Y version.
-version = u'0.5.3'
+version = u'0.5.4'
 # The full version, including alpha/beta/rc tags.
-release = u'0.5.3'
+release = u'0.5.4'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
@@ -90,7 +90,7 @@ todo_include_todos = False
 # The theme to use for HTML and HTML Help pages.  See the documentation for
 # a list of builtin themes.
 #
-html_theme = 'alabaster'
+html_theme = 'sphinx_rtd_theme'
 
 # Theme options are theme-specific and customize the look and feel of a theme
 # further.  For a list of options available for each theme, see the
@@ -99,13 +99,13 @@ html_theme = 'alabaster'
 # html_theme_options = {}
 
 html_theme_options = {
-    'github_banner': 'true',
+    # 'github_banner': 'true',
-    'github_button': 'true',
+    # 'github_button': 'true',
-    'github_user': 'recp',
+    # 'github_user': 'recp',
-    'github_repo': 'cglm',
+    # 'github_repo': 'cglm',
-    'travis_button': 'true',
+    # 'travis_button': 'true',
-    'show_related': 'true',
+    # 'show_related': 'true',
-    'fixed_sidebar': 'true'
+    # 'fixed_sidebar': 'true'
 }
 
 # Add any paths that contain custom static files (such as style sheets) here,

docs/source/features.rst

@@ -0,0 +1,23 @@
+Features
+================================================================================
+
+* general purpose matrix operations (mat4, mat3)
+* chain matrix multiplication (square only)
+* general purpose vector operations (cross, dot, rotate, proj, angle...)
+* affine transforms
+* matrix decomposition (extract rotation, scaling factor)
+* optimized affine transform matrices (mul, rigid-body inverse)
+* camera (lookat)
+* projections (ortho, perspective)
+* quaternions
+* euler angles / yaw-pitch-roll to matrix
+* extract euler angles
+* inline or pre-compiled function call
+* frustum (extract view frustum planes, corners...)
+* bounding box (AABB in Frustum (culling), crop, merge...)
+* bounding sphere
+* project, unproject
+* easing functions
+* curves
+* curve interpolation helpers (SMC, deCasteljau...)
+* and other...

docs/source/getting_started.rst

@@ -9,23 +9,26 @@ Types:
 .. code-block:: c
   :linenos:
 
-   typedef float vec3[3];
+  typedef float                   vec2[2];
-   typedef int  ivec3[3];
+  typedef float                   vec3[3];
-   typedef CGLM_ALIGN(16) float vec4[4];
+  typedef int                    ivec3[3];
+  typedef CGLM_ALIGN_IF(16) float vec4[4];
+  typedef vec4                    versor;
+  typedef vec3                    mat3[3];
 
-   typedef vec3 mat3[3];
+  #ifdef __AVX__
-   typedef vec4 mat4[4];
+  typedef CGLM_ALIGN_IF(32) vec4  mat4[4];
-
+  #else
-   typedef vec4 versor;
+  typedef CGLM_ALIGN_IF(16) vec4  mat4[4];
+  #endif
 
 As you can see types don't store extra informations in favor of space.
 You can send these values e.g. matrix to OpenGL directly without casting or calling a function like *value_ptr*
 
-Alignment is Required:
+Alignment Is Required:
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-**vec4** and **mat4** requires 16 byte alignment because vec4 and mat4 operations are
+**vec4** and **mat4** requires 16 (32 for **mat4** if AVX is enabled) byte alignment because **vec4** and **mat4** operations are vectorized by SIMD instructions (SSE/AVX/NEON).
-vectorized by SIMD instructions (SSE/AVX).
 
 **UPDATE:**
   By starting v0.4.5 cglm provides an option to disable alignment requirement, it is enabled as default
@@ -37,10 +40,9 @@ vectorized by SIMD instructions (SSE/AVX).
 Allocations:
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 *cglm* doesn't alloc any memory on heap. So it doesn't provide any allocator.
-You must allocate memory yourself. You should alloc memory for out parameters too if you pass pointer of memory location.
+You must allocate memory yourself. You should alloc memory for out parameters too if you pass pointer of memory location. When allocating memory, don't forget that **vec4** and **mat4** require alignment.
-When allocating memory don't forget that **vec4** and **mat4** requires alignment.
 
-**NOTE:** Unaligned vec4 and unaligned mat4 operations will be supported in the future. Check todo list.
+**NOTE:** Unaligned **vec4** and unaligned **mat4** operations will be supported in the future. Check todo list.
 Because you may want to multiply a CGLM matrix with external matrix.
 There is no guarantee that non-CGLM matrix is aligned. Unaligned types will have *u* prefix e.g. **umat4**
 

docs/source/index.rst

@@ -3,7 +3,7 @@
    You can adapt this file completely to your liking, but it should at least
    contain the root `toctree` directive.
 
-Welcome to cglm's documentation!
+cglm Documentation
 ================================
 
 **cglm** is optimized 3D math library written in C99 (compatible with C89).
@@ -14,33 +14,36 @@ is considered to be supported as optional.
 
 Also currently only **float** type is supported for most operations.
 
-**Features**
-
-* general purpose matrix operations (mat4, mat3)
-* chain matrix multiplication (square only)
-* general purpose vector operations (cross, dot, rotate, proj, angle...)
-* affine transforms
-* matrix decomposition (extract rotation, scaling factor)
-* optimized affine transform matrices (mul, rigid-body inverse)
-* camera (lookat)
-* projections (ortho, perspective)
-* quaternions
-* euler angles / yaw-pitch-roll to matrix
-* extract euler angles
-* inline or pre-compiled function call
-* frustum (extract view frustum planes, corners...)
-* bounding box (AABB in Frustum (culling), crop, merge...)
-
-
 .. toctree::
-   :maxdepth: 1
+   :maxdepth: 2
-   :caption: Table Of Contents:
+   :caption: Getting Started:
 
+   features
    build
    getting_started
+
+.. toctree::
+   :maxdepth: 2
+   :caption: How To:
+
    opengl
+
+.. toctree::
+   :maxdepth: 2
+   :caption: API:
+
    api
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Options:
+
    opt
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Troubleshooting:
+
    troubleshooting
 
 Indices and tables

docs/source/opengl.rst

@@ -45,7 +45,7 @@ array of matrices:
 
 in this way, passing aray of matrices is same
 
-Passing / Uniforming Vectors to OpenGL:¶
+Passing / Uniforming Vectors to OpenGL:
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 You don't need to do extra thing when passing cglm vectors to OpengL or other APIs.

include/cglm/affine-mat.h

@@ -152,7 +152,7 @@ glm_inv_tr(mat4 mat) {
   glm_inv_tr_sse2(mat);
 #else
   CGLM_ALIGN_MAT mat3 r;
-  CGLM_ALIGN(16) vec3 t;
+  CGLM_ALIGN(8)  vec3 t;
 
   /* rotate */
   glm_mat4_pick3t(mat, r);

include/cglm/bezier.h

@@ -8,6 +8,8 @@
 #ifndef cglm_bezier_h
 #define cglm_bezier_h
 
+#include "common.h"
+
 #define GLM_BEZIER_MAT_INIT  {{-1.0f,  3.0f, -3.0f,  1.0f},                   \
                               { 3.0f, -6.0f,  3.0f,  0.0f},                   \
                               {-3.0f,  3.0f,  0.0f,  0.0f},                   \

include/cglm/common.h

@@ -11,8 +11,10 @@
 #define _USE_MATH_DEFINES /* for windows */
 
 #include <stdint.h>
+#include <stddef.h>
 #include <math.h>
 #include <float.h>
+#include <stdbool.h>
 
 #if defined(_MSC_VER)
 #  ifdef CGLM_DLL

include/cglm/project.h

@@ -8,6 +8,7 @@
 #ifndef cglm_project_h
 #define cglm_project_h
 
+#include "common.h"
 #include "vec3.h"
 #include "vec4.h"
 #include "mat4.h"

include/cglm/quat.h

@@ -233,7 +233,7 @@ glm_quat_normalize_to(versor q, versor dest) {
   dot = glm_vec4_norm2(q);
 
   if (dot <= 0.0f) {
-    glm_quat_identity(q);
+    glm_quat_identity(dest);
     return;
   }
 

include/cglm/util.h

@@ -19,7 +19,6 @@
 #define cglm_util_h
 
 #include "common.h"
-#include <stdbool.h>
 
 #define GLM_MIN(X, Y) (((X) < (Y)) ? (X) : (Y))
 #define GLM_MAX(X, Y) (((X) > (Y)) ? (X) : (Y))

include/cglm/vec3-ext.h

@@ -31,9 +31,6 @@
 
 #include "common.h"
 #include "util.h"
-#include <stdbool.h>
-#include <math.h>
-#include <float.h>
 
 /*!
  * @brief fill a vector with specified value

include/cglm/vec3.h

@@ -428,8 +428,8 @@ glm_vec3_maxadd(vec3 a, vec3 b, vec3 dest) {
  *
  * it applies += operator so dest must be initialized
  *
- * @param[in]  a    vector
+ * @param[in]  a    vector 1
- * @param[in]  b    scalar
+ * @param[in]  b    vector 2
  * @param[out] dest dest += min(a, b)
  */
 CGLM_INLINE

include/cglm/vec4-ext.h

@@ -31,9 +31,6 @@
 
 #include "common.h"
 #include "vec3-ext.h"
-#include <stdbool.h>
-#include <math.h>
-#include <float.h>
 
 /*!
  * @brief fill a vector with specified value

include/cglm/vec4.h

@@ -576,8 +576,8 @@ glm_vec4_maxadd(vec4 a, vec4 b, vec4 dest) {
  *
  * it applies += operator so dest must be initialized
  *
- * @param[in]  a    vector
+ * @param[in]  a    vector 1
- * @param[in]  b    scalar
+ * @param[in]  b    vector 2
  * @param[out] dest dest += min(a, b)
  */
 CGLM_INLINE
@@ -690,18 +690,9 @@ CGLM_INLINE
 float
 glm_vec4_distance(vec4 a, vec4 b) {
 #if defined( __SSE__ ) || defined( __SSE2__ )
-  __m128 x0;
+  return glmm_norm(_mm_sub_ps(glmm_load(b), glmm_load(a)));
-  x0 = _mm_sub_ps(glmm_load(b), glmm_load(a));
-  x0 = _mm_mul_ps(x0, x0);
-  x0 = _mm_add_ps(x0, glmm_shuff1(x0, 1, 0, 3, 2));
-  return _mm_cvtss_f32(_mm_sqrt_ss(_mm_add_ss(x0,
-                                              glmm_shuff1(x0, 0, 1, 0, 1))));
 #elif defined(CGLM_NEON_FP)
-  float32x4_t v0;
+  return glmm_norm(vsubq_f32(glmm_load(a), glmm_load(b)));
-  float32_t   r;
-  v0 = vsubq_f32(vld1q_f32(a), vld1q_f32(b));
-  r  = vaddvq_f32(vmulq_f32(v0, v0));
-  return sqrtf(r);
 #else
   return sqrtf(glm_pow2(b[0] - a[0])
              + glm_pow2(b[1] - a[1])

include/cglm/version.h

@@ -10,6 +10,6 @@
 
 #define CGLM_VERSION_MAJOR 0
 #define CGLM_VERSION_MINOR 5
-#define CGLM_VERSION_PATCH 3
+#define CGLM_VERSION_PATCH 4
 
 #endif /* cglm_version_h */

post-build.sh

@@ -8,12 +8,17 @@
 
 cd $(dirname "$0")
 
-mkdir -p .libs
+mkdir -p "$(pwd)/.libs"
+
+libmocka_folder=$(pwd)/test/lib/cmocka/build/src/
 
 if [ "$(uname)" = "Darwin" ]; then
-  ln -sf $(pwd)/test/lib/cmocka/build/src/libcmocka.0.dylib \
+  libcmocka=libcmocka.0.dylib
-      .libs/libcmocka.0.dylib;
 else
-  ln -sf $(pwd)/test/lib/cmocka/build/src/libcmocka.so.0 \
+  libcmocka=libcmocka.so.0
-      .libs/libcmocka.so.0;
+fi
+
+libcmocka_path="$libmocka_folder$libcmocka"
+if [ -f "$libcmocka_path" ]; then
+  ln -sf "$libcmocka_path" "$(pwd)/.libs/$libcmocka";
 fi

test/src/test_bezier.c

@@ -51,8 +51,8 @@ test_bezier(void **state) {
   Bs_plain = test_bezier_plain(s, p0, c0, c1, p1);
 
   assert_true(glm_eq(Bs,  Bs_plain));
-  assert_true(glm_eq(smc, Bs_plain));
+  test_assert_eqf(smc, Bs_plain);
-  assert_true(glm_eq(Bs,  smc));
+  test_assert_eqf(Bs,  smc);
 
   /* test cubic hermite */
   smc      = glm_smc(s, GLM_HERMITE_MAT, (vec4){p0, p1, c0, c1});

test/src/test_common.c

@@ -5,6 +5,7 @@
 
 #include "test_common.h"
 #include <stdlib.h>
+#include <math.h>
 
 #define m 4
 #define n 4

test/src/test_vec4.c

@@ -93,6 +93,13 @@ test_vec4(void **state) {
     /* 3. test SIMD norm2 */
     test_rand_vec4(v);
     test_assert_eqf(test_vec4_norm2(v), glm_vec4_norm2(v));
+
+    /* 4. test SSE/SIMD distance */
+    test_rand_vec4(v1);
+    test_rand_vec4(v2);
+    d1 = glm_vec4_distance(v1, v2);
+    d2 = sqrtf(powf(v1[0]-v2[0], 2.0f) + pow(v1[1]-v2[1], 2.0f) + pow(v1[2]-v2[2], 2.0f) + pow(v1[3]-v2[3], 2.0f));
+    assert_true(fabsf(d1 - d2) <= 0.000009);
   }
 
   /* test zero */