glutils: fix matrix operations everywhere

- correct the matrix multiplication
- Use column-major matrices
  - reverse order of matrix multiplications

https://bugzilla.gnome.org/show_bug.cgi?id=785980

ext/gl/gstglimagesink.c

@@ -2237,10 +2237,14 @@ gst_glimage_sink_on_draw (GstGLImageSink * gl_sink)
           gst_buffer_get_video_affine_transformation_meta
           (gl_sink->stored_buffer[0]);
 
-      gst_gl_get_affine_transformation_meta_as_ndc_ext (af_meta, matrix);
+      if (gl_sink->transform_matrix) {
+        gfloat tmp[16];
 
-      if (gl_sink->transform_matrix)
-        gst_gl_multiply_matrix4 (gl_sink->transform_matrix, matrix, matrix);
+        gst_gl_get_affine_transformation_meta_as_ndc_ext (af_meta, tmp);
+        gst_gl_multiply_matrix4 (tmp, gl_sink->transform_matrix, matrix);
+      } else {
+        gst_gl_get_affine_transformation_meta_as_ndc_ext (af_meta, matrix);
+      }
 
       gst_gl_shader_set_uniform_matrix_4fv (gl_sink->redisplay_shader,
           "u_transformation", 1, FALSE, matrix);

ext/gl/gstgltransformation.c

@@ -770,20 +770,6 @@ gst_gl_transformation_gl_start (GstGLBaseFilter * base_filter)
   return TRUE;
 }
 
-static const gfloat from_ndc_matrix[] = {
-  0.5f, 0.0f, 0.0, 0.5f,
-  0.0f, 0.5f, 0.0, 0.5f,
-  0.0f, 0.0f, 0.5, 0.5f,
-  0.0f, 0.0f, 0.0, 1.0f,
-};
-
-static const gfloat to_ndc_matrix[] = {
-  2.0f, 0.0f, 0.0, -1.0f,
-  0.0f, 2.0f, 0.0, -1.0f,
-  0.0f, 0.0f, 2.0, -1.0f,
-  0.0f, 0.0f, 0.0, 1.0f,
-};
-
 static GstFlowReturn
 gst_gl_transformation_prepare_output_buffer (GstBaseTransform * trans,
     GstBuffer * inbuf, GstBuffer ** outbuf)
@@ -794,7 +780,8 @@ gst_gl_transformation_prepare_output_buffer (GstBaseTransform * trans,
   if (transformation->downstream_supports_affine_meta &&
       gst_video_info_is_equal (&filter->in_info, &filter->out_info)) {
     GstVideoAffineTransformationMeta *af_meta;
-    graphene_matrix_t upstream_matrix, from_ndc, to_ndc, tmp, tmp2, inv_aspect;
+    graphene_matrix_t upstream_matrix, tmp, tmp2, inv_aspect, yflip;
+    float upstream[16], downstream[16];
 
     *outbuf = gst_buffer_make_writable (inbuf);
 
@@ -805,19 +792,23 @@ gst_gl_transformation_prepare_output_buffer (GstBaseTransform * trans,
     GST_LOG_OBJECT (trans, "applying transformation to existing affine "
         "transformation meta");
 
+    gst_gl_get_affine_transformation_meta_as_ndc_ext (af_meta, upstream);
+
     /* apply the transformation to the existing affine meta */
-    graphene_matrix_init_from_float (&from_ndc, from_ndc_matrix);
-    graphene_matrix_init_from_float (&to_ndc, to_ndc_matrix);
-    graphene_matrix_init_from_float (&upstream_matrix, af_meta->matrix);
+    graphene_matrix_init_from_float (&upstream_matrix, upstream);
+    graphene_matrix_init_scale (&inv_aspect, transformation->aspect, -1., 1.);
+    graphene_matrix_init_scale (&yflip, 1., -1., 1.);
 
-    graphene_matrix_init_scale (&inv_aspect, transformation->aspect, 1., 1.);
+    /* invert the aspect effects */
+    graphene_matrix_multiply (&upstream_matrix, &inv_aspect, &tmp2);
+    /* apply the transformation */
+    graphene_matrix_multiply (&tmp2, &transformation->mvp_matrix, &tmp);
+    /* and undo yflip */
+    graphene_matrix_multiply (&tmp, &yflip, &tmp2);
 
-    graphene_matrix_multiply (&from_ndc, &upstream_matrix, &tmp);
-    graphene_matrix_multiply (&tmp, &transformation->mvp_matrix, &tmp2);
-    graphene_matrix_multiply (&tmp2, &inv_aspect, &tmp);
-    graphene_matrix_multiply (&tmp, &to_ndc, &tmp2);
+    graphene_matrix_to_float (&tmp2, downstream);
+    gst_gl_set_affine_transformation_meta_from_ndc_ext (af_meta, downstream);
 
-    graphene_matrix_to_float (&tmp2, af_meta->matrix);
     return GST_FLOW_OK;
   }
 

ext/gl/gstglutils.c

@@ -108,24 +108,24 @@ gst_gl_context_gen_shader (GstGLContext * context, const gchar * vert_src,
 }
 
 static const gfloat identity_matrix[] = {
-  1.0f, 0.0f, 0.0, 0.0f,
-  0.0f, 1.0f, 0.0, 0.0f,
-  0.0f, 0.0f, 1.0, 0.0f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  1.0, 0.0, 0.0, 0.0,
+  0.0, 1.0, 0.0, 0.0,
+  0.0, 0.0, 1.0, 0.0,
+  0.0, 0.0, 0.0, 1.0,
 };
 
 static const gfloat from_ndc_matrix[] = {
-  0.5f, 0.0f, 0.0, 0.5f,
-  0.0f, 0.5f, 0.0, 0.5f,
-  0.0f, 0.0f, 0.5, 0.5f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  0.5, 0.0, 0.0, 0.0,
+  0.0, 0.5, 0.0, 0.0,
+  0.0, 0.0, 0.5, 0.0,
+  0.5, 0.5, 0.5, 1.0,
 };
 
 static const gfloat to_ndc_matrix[] = {
-  2.0f, 0.0f, 0.0, -1.0f,
-  0.0f, 2.0f, 0.0, -1.0f,
-  0.0f, 0.0f, 2.0, -1.0f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  2.0, 0.0, 0.0, 0.0,
+  0.0, 2.0, 0.0, 0.0,
+  0.0, 0.0, 2.0, 0.0,
+  -1.0, -1.0, -1.0, 1.0,
 };
 
 void
@@ -137,10 +137,10 @@ gst_gl_multiply_matrix4 (const gfloat * a, const gfloat * b, gfloat * result)
   if (!a || !b || !result)
     return;
 
-  for (i = 0; i < 4; i++) {
-    for (j = 0; j < 4; j++) {
+  for (i = 0; i < 4; i++) {     /* column */
+    for (j = 0; j < 4; j++) {   /* row */
       for (k = 0; k < 4; k++) {
-        tmp[i + (j * 4)] += a[i + (k * 4)] * b[k + (j * 4)];
+        tmp[j + (i * 4)] += a[k + (i * 4)] * b[j + (k * 4)];
       }
     }
   }
@@ -159,9 +159,20 @@ void gst_gl_get_affine_transformation_meta_as_ndc_ext
       matrix[i] = identity_matrix[i];
     }
   } else {
-    gfloat tmp[16] = { 0.0f };
+    float tmp[16];
 
-    gst_gl_multiply_matrix4 (from_ndc_matrix, meta->matrix, tmp);
-    gst_gl_multiply_matrix4 (tmp, to_ndc_matrix, matrix);
+    gst_gl_multiply_matrix4 (to_ndc_matrix, meta->matrix, tmp);
+    gst_gl_multiply_matrix4 (tmp, from_ndc_matrix, matrix);
   }
 }
+
+void gst_gl_set_affine_transformation_meta_from_ndc_ext
+    (GstVideoAffineTransformationMeta * meta, const gfloat * matrix)
+{
+  float tmp[16];
+
+  g_return_if_fail (meta != NULL);
+
+  gst_gl_multiply_matrix4 (from_ndc_matrix, matrix, tmp);
+  gst_gl_multiply_matrix4 (tmp, to_ndc_matrix, meta->matrix);
+}

ext/gl/gstglutils.h

@@ -31,6 +31,7 @@ gboolean gst_gl_context_gen_shader (GstGLContext * context,
 void gst_gl_multiply_matrix4 (const gfloat * a, const gfloat * b, gfloat * result);
 void gst_gl_get_affine_transformation_meta_as_ndc_ext (GstVideoAffineTransformationMeta *
     meta, gfloat * matrix);
+void gst_gl_set_affine_transformation_meta_from_ndc_ext (GstVideoAffineTransformationMeta * meta, const gfloat * matrix);
 
 G_END_DECLS
 

gst-libs/gst/gl/gstglutils.c

@@ -779,32 +779,35 @@ gst_gl_value_set_texture_target_from_mask (GValue * value,
       ret = TRUE;
     }
 
-    g_value_unset(&item);
+    g_value_unset (&item);
     return ret;
   }
 }
 
 static const gfloat identity_matrix[] = {
-  1.0f, 0.0f, 0.0, 0.0f,
-  0.0f, 1.0f, 0.0, 0.0f,
-  0.0f, 0.0f, 1.0, 0.0f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  1.0, 0.0, 0.0, 0.0,
+  0.0, 1.0, 0.0, 0.0,
+  0.0, 0.0, 1.0, 0.0,
+  0.0, 0.0, 0.0, 1.0,
 };
 
 static const gfloat from_ndc_matrix[] = {
-  0.5f, 0.0f, 0.0, 0.5f,
-  0.0f, 0.5f, 0.0, 0.5f,
-  0.0f, 0.0f, 0.5, 0.5f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  0.5, 0.0, 0.0, 0.0,
+  0.0, 0.5, 0.0, 0.0,
+  0.0, 0.0, 0.5, 0.0,
+  0.5, 0.5, 0.5, 1.0,
 };
 
 static const gfloat to_ndc_matrix[] = {
-  2.0f, 0.0f, 0.0, -1.0f,
-  0.0f, 2.0f, 0.0, -1.0f,
-  0.0f, 0.0f, 2.0, -1.0f,
-  0.0f, 0.0f, 0.0, 1.0f,
+  2.0, 0.0, 0.0, 0.0,
+  0.0, 2.0, 0.0, 0.0,
+  0.0, 0.0, 2.0, 0.0,
+  -1.0, -1.0, -1.0, 1.0,
 };
 
+/* multiplies two 4x4 matrices, @a X @b, and stores the result in @result
+ * https://en.wikipedia.org/wiki/Matrix_multiplication
+ */
 static void
 gst_gl_multiply_matrix4 (const gfloat * a, const gfloat * b, gfloat * result)
 {
@@ -813,11 +816,10 @@ gst_gl_multiply_matrix4 (const gfloat * a, const gfloat * b, gfloat * result)
 
   if (!a || !b || !result)
     return;
-
-  for (i = 0; i < 4; i++) {
-    for (j = 0; j < 4; j++) {
+  for (i = 0; i < 4; i++) {     /* column */
+    for (j = 0; j < 4; j++) {   /* row */
       for (k = 0; k < 4; k++) {
-        tmp[i + (j * 4)] += a[i + (k * 4)] * b[k + (j * 4)];
+        tmp[j + (i * 4)] += a[k + (i * 4)] * b[j + (k * 4)];
       }
     }
   }
@@ -826,6 +828,19 @@ gst_gl_multiply_matrix4 (const gfloat * a, const gfloat * b, gfloat * result)
     result[i] = tmp[i];
 }
 
+/*
+ * gst_gl_get_affine_transformation_meta_as_ndc:
+ * @meta: (nullable): a #GstVideoAffineTransformationMeta
+ * @matrix: (out): result of the 4x4 matrix
+ *
+ * Retrieves the stored 4x4 affine transformation matrix stored in @meta in
+ * NDC coordinates. if @meta is NULL, an identity matrix is returned.
+ *
+ * NDC is a left-handed coordinate sytem
+ * - x - [-1, 1] - +ve X moves right
+ * - y - [-1, 1] - +ve Y moves up
+ * - z - [-1, 1] - +ve Z moves into
+ */
 void
 gst_gl_get_affine_transformation_meta_as_ndc (GstVideoAffineTransformationMeta *
     meta, gfloat * matrix)
@@ -837,9 +852,22 @@ gst_gl_get_affine_transformation_meta_as_ndc (GstVideoAffineTransformationMeta *
       matrix[i] = identity_matrix[i];
     }
   } else {
-    gfloat tmp[16] = { 0.0f };
+    float tmp[16];
 
-    gst_gl_multiply_matrix4 (from_ndc_matrix, meta->matrix, tmp);
-    gst_gl_multiply_matrix4 (tmp, to_ndc_matrix, matrix);
+    /* change of basis multiplications */
+    gst_gl_multiply_matrix4 (to_ndc_matrix, meta->matrix, tmp);
+    gst_gl_multiply_matrix4 (tmp, from_ndc_matrix, matrix);
   }
 }
+
+void gst_gl_set_affine_transformation_meta_from_ndc
+    (GstVideoAffineTransformationMeta * meta, const gfloat * matrix)
+{
+  float tmp[16];
+
+  g_return_if_fail (meta != NULL);
+
+  /* change of basis multiplications */
+  gst_gl_multiply_matrix4 (from_ndc_matrix, matrix, tmp);
+  gst_gl_multiply_matrix4 (tmp, to_ndc_matrix, meta->matrix);
+}

gst-libs/gst/gl/gstglutils_private.h

@@ -27,6 +27,7 @@ G_BEGIN_DECLS
 
 G_GNUC_INTERNAL gboolean gst_gl_run_query (GstElement * element, GstQuery * query, GstPadDirection direction);
 G_GNUC_INTERNAL void gst_gl_get_affine_transformation_meta_as_ndc (GstVideoAffineTransformationMeta * meta, gfloat * matrix);
+G_GNUC_INTERNAL void gst_gl_set_affine_transformation_meta_from_ndc (GstVideoAffineTransformationMeta * meta, const gfloat * matrix);
 
 G_END_DECLS
 

tests/check/Makefile.am

@@ -209,6 +209,7 @@ check_gl=libs/gstglcontext \
     libs/gstglmemory \
     libs/gstglupload \
     libs/gstglcolorconvert \
+    libs/gstglmatrix \
     libs/gstglquery \
     libs/gstglsl \
     libs/gstglheaders \
@@ -588,6 +589,16 @@ libs_gstglcolorconvert_LDADD = \
 	$(GST_PLUGINS_BASE_LIBS) $(GST_VIDEO_LIBS) $(GST_BASE_LIBS) \
 	$(GST_LIBS) $(LDADD)
 
+libs_gstglmatrix_CFLAGS = \
+	$(GST_PLUGINS_BAD_CFLAGS) $(GST_PLUGINS_BASE_CFLAGS) \
+	-DGST_USE_UNSTABLE_API \
+	$(GST_BASE_CFLAGS) $(GST_CFLAGS) $(AM_CFLAGS)
+
+libs_gstglmatrix_LDADD = \
+	$(top_builddir)/gst-libs/gst/gl/libgstgl-@GST_API_VERSION@.la \
+	$(GST_PLUGINS_BASE_LIBS) $(GST_VIDEO_LIBS) $(GST_BASE_LIBS) \
+	$(GST_LIBS) $(LDADD)
+
 libs_gstglquery_LDADD = \
 	$(top_builddir)/gst-libs/gst/gl/libgstgl-@GST_API_VERSION@.la \
 	$(GST_PLUGINS_BASE_LIBS) $(GST_BASE_LIBS) $(GST_LIBS) $(LDADD)

tests/check/libs/.gitignore

@@ -13,3 +13,4 @@ gstglcolorconvert
 gstglsl
 gstglquery
 gstglheaders
+gstglmatrix

tests/check/libs/gstglmatrix.c

@@ -0,0 +1,262 @@
+/* GStreamer
+ *
+ * Copyright (C) 2014 Matthew Waters <ystreet00@gmail.com>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Library General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Library General Public License for more details.
+ *
+ * You should have received a copy of the GNU Library General Public
+ * License along with this library; if not, write to the
+ * Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
+ * Boston, MA 02110-1301, USA.
+ */
+
+#ifdef HAVE_CONFIG_H
+#  include "config.h"
+#endif
+
+#include <gst/gl/gstglutils.c>
+#undef GST_CAT_DEFAULT
+#include <gst/check/gstcheck.h>
+
+#define EPSILON 0.0001f
+#define FEQ(a,b) (fabs(a-b) < EPSILON)
+
+static void
+debug_matrix (const float *m)
+{
+  int i;
+  for (i = 0; i < 4; i++) {
+    GST_DEBUG ("%10.4f %10.4f %10.4f %10.4f", m[i * 4 + 0], m[i * 4 + 1],
+        m[i * 4 + 2], m[i * 4 + 3]);
+  }
+}
+
+GST_START_TEST (test_matrix_multiply)
+{
+  /* A * B == C */
+  const float A[] = {
+    1., 1., 2., 5.,
+    0., 3., 0., 1.,
+    2., 0., 3., 1.,
+    3., 2., 1., 0.,
+  };
+
+  const float B[] = {
+    3., 1., 0., 2.,
+    1., 0., 3., 2.,
+    0., 1., 2., 3.,
+    3., 2., 1., 0.,
+  };
+
+  const float C[] = {
+    19., 13., 12., 10.,
+    6., 2., 10., 6.,
+    9., 7., 7., 13.,
+    11., 4., 8., 13.,
+  };
+
+  float res[16];
+  int i;
+
+  gst_gl_multiply_matrix4 (A, B, res);
+  GST_DEBUG ("result");
+  debug_matrix (res);
+
+  for (i = 0; i < G_N_ELEMENTS (res); i++) {
+    fail_unless (FEQ (res[i], C[i]), "value %f at index %u does not match "
+        "expected value %f", res[i], i, C[i]);
+  }
+}
+
+GST_END_TEST;
+
+GST_START_TEST (test_matrix_ndc)
+{
+  GstBuffer *buffer = gst_buffer_new ();
+  GstVideoAffineTransformationMeta *aff_meta;
+  float res[16];
+  int i;
+
+  const float m[] = {
+    1., 0., 0., 0.,
+    0., 1., 0., 0.,
+    0., 0., 1., 0.,
+    0., 0., 0., 1.,
+  };
+
+  const float n[] = {
+    4., 6., 4., 9.,
+    1., 5., 8., 2.,
+    9., 3., 5., 8.,
+    3., 7., 9., 1.,
+  };
+
+  aff_meta = gst_buffer_add_video_affine_transformation_meta (buffer);
+
+  /* test default identity matrix */
+  gst_gl_get_affine_transformation_meta_as_ndc (aff_meta, res);
+  GST_DEBUG ("result");
+  debug_matrix (res);
+
+  for (i = 0; i < G_N_ELEMENTS (res); i++) {
+    fail_unless (FEQ (res[i], m[i]), "value %f at index %u does not match "
+        "expected value %f", res[i], i, m[i]);
+  }
+
+  /* test setting and receiving the same values */
+  gst_gl_set_affine_transformation_meta_from_ndc (aff_meta, n);
+  gst_gl_get_affine_transformation_meta_as_ndc (aff_meta, res);
+
+  GST_DEBUG ("result");
+  debug_matrix (res);
+
+  for (i = 0; i < G_N_ELEMENTS (res); i++) {
+    fail_unless (FEQ (res[i], n[i]), "value %f at index %u does not match "
+        "expected value %f", res[i], i, n[i]);
+  }
+
+  gst_buffer_unref (buffer);
+}
+
+GST_END_TEST;
+#if 0
+static void
+transpose_matrix4 (float *m, float *res)
+{
+  int i, j;
+
+  for (i = 0; i < 4; i++) {
+    for (j = 0; j < 4; j++) {
+      int idx = i + (j * 4);
+      int swapped_idx = j + (i * 4);
+
+      if (i == j)
+        fail_unless (idx == swapped_idx);
+
+      res[swapped_idx] = m[idx];
+    }
+  }
+}
+
+static float
+dot4 (float *v1, float *v2)
+{
+  GST_TRACE ("%.4f * %.4f + %.4f * %.4f + %.4f * %.4f + %.4f * %.4f",
+      v1[0], v2[0], v1[1], v2[1], v1[2], v2[2], v1[3], v2[3]);
+  return v1[0] * v2[0] + v1[1] * v2[1] + v1[2] * v2[2] + v1[3] * v2[3];
+}
+
+/* m * v */
+static void
+_matrix_mult_vertex4 (float *m, float *v, float *res)
+{
+  res[0] = dot4 (&m[0], v);
+  res[1] = dot4 (&m[4], v);
+  res[2] = dot4 (&m[8], v);
+  res[3] = dot4 (&m[12], v);
+}
+
+/* v * m */
+static void
+_vertex_mult_matrix4 (float *v, float *m, float *res)
+{
+  float tmp[16] = { 0., };
+
+  transpose_matrix4 (m, tmp);
+  _matrix_mult_vertex4 (tmp, v, res);
+}
+
+GST_START_TEST (test_matrix_vertex_identity)
+{
+  float identity[] = {
+    1., 0., 0., 0.,
+    0., 1., 0., 0.,
+    0., 0., 1., 0.,
+    0., 0., 0., 1.,
+  };
+
+  float v[] = { 1., 1., 1., 1. };
+  float res[4] = { 0., };
+  int i;
+
+  _vertex_mult_matrix4 (v, identity, res);
+  GST_DEBUG ("vertex: %.4f %.4f %.4f %.4f", v[0], v[1], v[2], v[3]);
+  GST_DEBUG ("result: %.4f %.4f %.4f %.4f", res[0], res[1], res[2], res[3]);
+
+  for (i = 0; i < 4; i++) {
+    fail_unless (FEQ (res[i], v[i]), "value %f at index %u does not match "
+        "expected value %f", res[i], i, v[i]);
+  }
+
+  _matrix_mult_vertex4 (identity, v, res);
+  GST_DEBUG ("vertex: %.4f %.4f %.4f %.4f", v[0], v[1], v[2], v[3]);
+  GST_DEBUG ("result: %.4f %.4f %.4f %.4f", res[0], res[1], res[2], res[3]);
+
+  for (i = 0; i < 4; i++) {
+    fail_unless (FEQ (res[i], v[i]), "value %f at index %u does not match "
+        "expected value %f", res[i], i, v[i]);
+  }
+}
+
+GST_END_TEST;
+
+GST_START_TEST (test_matrix_vertex_scale)
+{
+  float scale[] = {
+    1.5, 0., 0., 0.,
+    0., 2.5, 0., 0.,
+    0., 0., 3., 0.,
+    0., 0., 0., 1.,
+  };
+
+  float v[] = { 1., 1., 1., 1. };
+  float expected[] = { 1.5, 2.5, 3., 1. };
+  float res[4] = { 0., };
+  int i;
+
+  _vertex_mult_matrix4 (v, scale, res);
+  GST_DEBUG ("vertex: %.4f %.4f %.4f %.4f", v[0], v[1], v[2], v[3]);
+  GST_DEBUG ("result: %.4f %.4f %.4f %.4f", res[0], res[1], res[2], res[3]);
+
+  for (i = 0; i < 4; i++) {
+    fail_unless (FEQ (res[i], expected[i]),
+        "value %f at index %u does not match " "expected value %f", res[i], i,
+        expected[i]);
+  }
+
+  _matrix_mult_vertex4 (scale, v, res);
+  GST_DEBUG ("vertex: %.4f %.4f %.4f %.4f", v[0], v[1], v[2], v[3]);
+  GST_DEBUG ("result: %.4f %.4f %.4f %.4f", res[0], res[1], res[2], res[3]);
+
+  for (i = 0; i < 4; i++) {
+    fail_unless (FEQ (res[i], expected[i]),
+        "value %f at index %u does not match " "expected value %f", res[i], i,
+        expected[i]);
+  }
+}
+
+GST_END_TEST;
+#endif
+
+static Suite *
+gst_gl_upload_suite (void)
+{
+  Suite *s = suite_create ("GstGLMatrix");
+  TCase *tc_chain = tcase_create ("matrix");
+
+  suite_add_tcase (s, tc_chain);
+  tcase_add_test (tc_chain, test_matrix_multiply);
+  tcase_add_test (tc_chain, test_matrix_ndc);
+
+  return s;
+}
+
+GST_CHECK_MAIN (gst_gl_upload);