mirror of
https://github.com/brl/mutter.git
synced 2024-11-21 23:50:41 -05:00
cogl/matrix-stack: Use graphene APIs
This one is a bit tricky. The tl;dr; is that switching from right-hand multiplication to left-hand multiplication required applying the stack from left to root. This actually allowed simplifying the code a bit, since CoglMatrixEntry only stores a pointer to its parent, and that's all we need to know for left-hand multiplication. https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/1439
This commit is contained in:
parent
d03ffe1e97
commit
9e1004da0c
@ -238,10 +238,10 @@ cogl_matrix_stack_frustum (CoglMatrixStack *stack,
|
||||
_cogl_matrix_stack_push_replacement_entry (stack,
|
||||
COGL_MATRIX_OP_LOAD);
|
||||
|
||||
cogl_matrix_init_identity (&entry->matrix);
|
||||
cogl_matrix_frustum (&entry->matrix,
|
||||
left, right, bottom, top,
|
||||
z_near, z_far);
|
||||
graphene_matrix_init_frustum (&entry->matrix,
|
||||
left, right,
|
||||
bottom, top,
|
||||
z_near, z_far);
|
||||
}
|
||||
|
||||
void
|
||||
@ -256,9 +256,9 @@ cogl_matrix_stack_perspective (CoglMatrixStack *stack,
|
||||
entry =
|
||||
_cogl_matrix_stack_push_replacement_entry (stack,
|
||||
COGL_MATRIX_OP_LOAD);
|
||||
cogl_matrix_init_identity (&entry->matrix);
|
||||
cogl_matrix_perspective (&entry->matrix,
|
||||
fov_y, aspect, z_near, z_far);
|
||||
graphene_matrix_init_perspective (&entry->matrix,
|
||||
fov_y, aspect,
|
||||
z_near, z_far);
|
||||
}
|
||||
|
||||
void
|
||||
@ -275,9 +275,10 @@ cogl_matrix_stack_orthographic (CoglMatrixStack *stack,
|
||||
entry =
|
||||
_cogl_matrix_stack_push_replacement_entry (stack,
|
||||
COGL_MATRIX_OP_LOAD);
|
||||
cogl_matrix_init_identity (&entry->matrix);
|
||||
cogl_matrix_orthographic (&entry->matrix,
|
||||
x_1, y_1, x_2, y_2, near, far);
|
||||
graphene_matrix_init_ortho (&entry->matrix,
|
||||
x_1, x_2,
|
||||
y_2, y_1,
|
||||
near, far);
|
||||
}
|
||||
|
||||
void
|
||||
@ -372,44 +373,94 @@ graphene_matrix_t *
|
||||
cogl_matrix_entry_get (CoglMatrixEntry *entry,
|
||||
graphene_matrix_t *matrix)
|
||||
{
|
||||
int depth;
|
||||
CoglMatrixEntry *current;
|
||||
CoglMatrixEntry **children;
|
||||
int i;
|
||||
int depth;
|
||||
|
||||
for (depth = 0, current = entry;
|
||||
graphene_matrix_init_identity (matrix);
|
||||
|
||||
for (current = entry, depth = 0;
|
||||
current;
|
||||
current = current->parent, depth++)
|
||||
{
|
||||
switch (current->op)
|
||||
{
|
||||
case COGL_MATRIX_OP_TRANSLATE:
|
||||
{
|
||||
CoglMatrixEntryTranslate *translate =
|
||||
(CoglMatrixEntryTranslate *) current;
|
||||
graphene_matrix_translate (matrix, &translate->translate);
|
||||
break;
|
||||
}
|
||||
case COGL_MATRIX_OP_ROTATE:
|
||||
{
|
||||
CoglMatrixEntryRotate *rotate =
|
||||
(CoglMatrixEntryRotate *) current;
|
||||
graphene_matrix_rotate (matrix, rotate->angle, &rotate->axis);
|
||||
break;
|
||||
}
|
||||
case COGL_MATRIX_OP_ROTATE_EULER:
|
||||
{
|
||||
CoglMatrixEntryRotateEuler *rotate =
|
||||
(CoglMatrixEntryRotateEuler *) current;
|
||||
graphene_matrix_rotate_euler (matrix, &rotate->euler);
|
||||
break;
|
||||
}
|
||||
case COGL_MATRIX_OP_SCALE:
|
||||
{
|
||||
CoglMatrixEntryScale *scale =
|
||||
(CoglMatrixEntryScale *) current;
|
||||
graphene_matrix_scale (matrix, scale->x, scale->y, scale->z);
|
||||
break;
|
||||
}
|
||||
case COGL_MATRIX_OP_MULTIPLY:
|
||||
{
|
||||
CoglMatrixEntryMultiply *multiply =
|
||||
(CoglMatrixEntryMultiply *) current;
|
||||
graphene_matrix_multiply (matrix, &multiply->matrix, matrix);
|
||||
break;
|
||||
}
|
||||
|
||||
case COGL_MATRIX_OP_LOAD_IDENTITY:
|
||||
cogl_matrix_init_identity (matrix);
|
||||
goto initialized;
|
||||
goto applied;
|
||||
|
||||
case COGL_MATRIX_OP_LOAD:
|
||||
{
|
||||
CoglMatrixEntryLoad *load = (CoglMatrixEntryLoad *)current;
|
||||
_cogl_matrix_init_from_matrix_without_inverse (matrix,
|
||||
&load->matrix);
|
||||
goto initialized;
|
||||
CoglMatrixEntryLoad *load = (CoglMatrixEntryLoad *) current;
|
||||
graphene_matrix_multiply (matrix, &load->matrix, matrix);
|
||||
goto applied;
|
||||
}
|
||||
case COGL_MATRIX_OP_SAVE:
|
||||
{
|
||||
CoglMatrixEntrySave *save = (CoglMatrixEntrySave *)current;
|
||||
CoglMatrixEntrySave *save = (CoglMatrixEntrySave *) current;
|
||||
if (!save->cache_valid)
|
||||
{
|
||||
cogl_matrix_entry_get (current->parent, &save->cache);
|
||||
save->cache_valid = TRUE;
|
||||
}
|
||||
_cogl_matrix_init_from_matrix_without_inverse (matrix, &save->cache);
|
||||
goto initialized;
|
||||
graphene_matrix_multiply (matrix, &save->cache, matrix);
|
||||
goto applied;
|
||||
}
|
||||
default:
|
||||
continue;
|
||||
}
|
||||
}
|
||||
|
||||
initialized:
|
||||
applied:
|
||||
|
||||
#ifdef COGL_ENABLE_DEBUG
|
||||
if (!current)
|
||||
{
|
||||
g_warning ("Inconsistent matrix stack");
|
||||
return NULL;
|
||||
}
|
||||
|
||||
entry->composite_gets++;
|
||||
|
||||
if (COGL_DEBUG_ENABLED (COGL_DEBUG_PERFORMANCE) &&
|
||||
entry->composite_gets >= 2)
|
||||
{
|
||||
COGL_NOTE (PERFORMANCE,
|
||||
"Re-composing a matrix stack entry multiple times");
|
||||
}
|
||||
#endif
|
||||
|
||||
if (depth == 0)
|
||||
{
|
||||
@ -438,97 +489,6 @@ initialized:
|
||||
return NULL;
|
||||
}
|
||||
|
||||
#ifdef COGL_ENABLE_DEBUG
|
||||
if (!current)
|
||||
{
|
||||
g_warning ("Inconsistent matrix stack");
|
||||
return NULL;
|
||||
}
|
||||
|
||||
entry->composite_gets++;
|
||||
#endif
|
||||
|
||||
children = g_alloca (sizeof (CoglMatrixEntry) * depth);
|
||||
|
||||
/* We need walk the list of entries from the init/load/save entry
|
||||
* back towards the leaf node but the nodes don't link to their
|
||||
* children so we need to re-walk them here to add to a separate
|
||||
* array. */
|
||||
for (i = depth - 1, current = entry;
|
||||
i >= 0 && current;
|
||||
i--, current = current->parent)
|
||||
{
|
||||
children[i] = current;
|
||||
}
|
||||
|
||||
#ifdef COGL_ENABLE_DEBUG
|
||||
if (COGL_DEBUG_ENABLED (COGL_DEBUG_PERFORMANCE) &&
|
||||
entry->composite_gets >= 2)
|
||||
{
|
||||
COGL_NOTE (PERFORMANCE,
|
||||
"Re-composing a matrix stack entry multiple times");
|
||||
}
|
||||
#endif
|
||||
|
||||
for (i = 0; i < depth; i++)
|
||||
{
|
||||
switch (children[i]->op)
|
||||
{
|
||||
case COGL_MATRIX_OP_TRANSLATE:
|
||||
{
|
||||
CoglMatrixEntryTranslate *translate =
|
||||
(CoglMatrixEntryTranslate *)children[i];
|
||||
cogl_matrix_translate (matrix,
|
||||
translate->translate.x,
|
||||
translate->translate.y,
|
||||
translate->translate.z);
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_ROTATE:
|
||||
{
|
||||
CoglMatrixEntryRotate *rotate=
|
||||
(CoglMatrixEntryRotate *)children[i];
|
||||
cogl_matrix_rotate (matrix,
|
||||
rotate->angle,
|
||||
graphene_vec3_get_x (&rotate->axis),
|
||||
graphene_vec3_get_y (&rotate->axis),
|
||||
graphene_vec3_get_z (&rotate->axis));
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_ROTATE_EULER:
|
||||
{
|
||||
CoglMatrixEntryRotateEuler *rotate =
|
||||
(CoglMatrixEntryRotateEuler *)children[i];
|
||||
cogl_matrix_rotate_euler (matrix,
|
||||
&rotate->euler);
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_SCALE:
|
||||
{
|
||||
CoglMatrixEntryScale *scale =
|
||||
(CoglMatrixEntryScale *)children[i];
|
||||
cogl_matrix_scale (matrix,
|
||||
scale->x,
|
||||
scale->y,
|
||||
scale->z);
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_MULTIPLY:
|
||||
{
|
||||
CoglMatrixEntryMultiply *multiply =
|
||||
(CoglMatrixEntryMultiply *)children[i];
|
||||
cogl_matrix_multiply (matrix, matrix, &multiply->matrix);
|
||||
continue;
|
||||
}
|
||||
|
||||
case COGL_MATRIX_OP_LOAD_IDENTITY:
|
||||
case COGL_MATRIX_OP_LOAD:
|
||||
case COGL_MATRIX_OP_SAVE:
|
||||
g_warn_if_reached ();
|
||||
continue;
|
||||
}
|
||||
}
|
||||
|
||||
return NULL;
|
||||
}
|
||||
|
||||
@ -793,7 +753,7 @@ cogl_matrix_entry_equal (CoglMatrixEntry *entry0,
|
||||
{
|
||||
CoglMatrixEntryMultiply *mult0 = (CoglMatrixEntryMultiply *)entry0;
|
||||
CoglMatrixEntryMultiply *mult1 = (CoglMatrixEntryMultiply *)entry1;
|
||||
if (!cogl_matrix_equal (&mult0->matrix, &mult1->matrix))
|
||||
if (!graphene_matrix_equal (&mult0->matrix, &mult1->matrix))
|
||||
return FALSE;
|
||||
}
|
||||
break;
|
||||
@ -804,7 +764,7 @@ cogl_matrix_entry_equal (CoglMatrixEntry *entry0,
|
||||
/* There's no need to check any further since an
|
||||
* _OP_LOAD makes all the ancestors redundant as far as
|
||||
* the final matrix value is concerned. */
|
||||
return cogl_matrix_equal (&load0->matrix, &load1->matrix);
|
||||
return graphene_matrix_equal (&load0->matrix, &load1->matrix);
|
||||
}
|
||||
case COGL_MATRIX_OP_SAVE:
|
||||
/* We skip over saves above so we shouldn't see save entries */
|
||||
@ -890,14 +850,14 @@ cogl_debug_matrix_entry_print (CoglMatrixEntry *entry)
|
||||
{
|
||||
CoglMatrixEntryMultiply *mult = (CoglMatrixEntryMultiply *)entry;
|
||||
g_print (" MULT:\n");
|
||||
_cogl_matrix_prefix_print (" ", &mult->matrix);
|
||||
graphene_matrix_print (&mult->matrix);
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_LOAD:
|
||||
{
|
||||
CoglMatrixEntryLoad *load = (CoglMatrixEntryLoad *)entry;
|
||||
g_print (" LOAD:\n");
|
||||
_cogl_matrix_prefix_print (" ", &load->matrix);
|
||||
graphene_matrix_print (&load->matrix);
|
||||
continue;
|
||||
}
|
||||
case COGL_MATRIX_OP_SAVE:
|
||||
|
Loading…
Reference in New Issue
Block a user