Lerking/gtk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

gpu: Don't use a vec2 for scales

Browse Source 
This is wasteful (since it gets stored as 4 highly-aligned floats),
and all we ever do is unpack and repack that simf4.
This commit is contained in:
Matthias Clasen
2024-01-20 16:26:06 -05:00
parent 761995ce3c
commit 1355367360
8 changed files with 78 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
gsk/gpu/gskgpuglobalsop.c
View File

@@ -12,6 +12,8 @@
#include "gskvulkandescriptorsprivate.h"
#endif
#include "gskrectprivate.h"
typedef struct _GskGpuGlobalsOp GskGpuGlobalsOp;
struct _GskGpuGlobalsOp
@@ -87,7 +89,7 @@ static const GskGpuOpClass GSK_GPU_GLOBALS_OP_CLASS = {
void
gsk_gpu_globals_op (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
const graphene_matrix_t *mvp,
const GskRoundedRect *clip)
{
@@ -97,5 +99,5 @@ gsk_gpu_globals_op (GskGpuFrame *frame,
graphene_matrix_to_float (mvp, self->instance.mvp);
gsk_rounded_rect_to_float (clip, graphene_point_zero (), self->instance.clip);
graphene_vec2_to_float (scale, self->instance.scale);
gsk_point_to_float (scale, self->instance.scale);
}

2
gsk/gpu/gskgpuglobalsopprivate.h
View File

@@ -17,7 +17,7 @@ struct _GskGpuGlobalsInstance
};
void gsk_gpu_globals_op (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
const graphene_matrix_t *mvp,
const GskRoundedRect *clip);

99
gsk/gpu/gskgpunodeprocessor.c
View File

@@ -104,12 +104,12 @@ struct _GskGpuNodeProcessor
GskGpuFrame *frame;
GskGpuDescriptors *desc;
cairo_rectangle_int_t scissor;
GskGpuClip clip;
GskGpuBlend blend;
graphene_point_t offset;
graphene_matrix_t projection;
graphene_vec2_t scale;
graphene_point_t scale;
GskTransform *modelview;
GskGpuClip clip;
graphene_matrix_t projection;
float opacity;
GskGpuGlobals pending_globals;
@@ -130,7 +130,7 @@ struct _GskGpuPatternWriter
graphene_rect_t bounds;
graphene_point_t offset;
graphene_vec2_t scale;
graphene_point_t scale;
guint stack;
PatternBuffer buffer;
@@ -188,9 +188,8 @@ gsk_gpu_node_processor_init (GskGpuNodeProcessor *self,
self->modelview = NULL;
gsk_gpu_image_get_projection_matrix (target, &self->projection);
graphene_vec2_init (&self->scale,
width / viewport->size.width,
height / viewport->size.height);
self->scale = GRAPHENE_POINT_INIT (width / viewport->size.width,
height / viewport->size.height);
self->offset = GRAPHENE_POINT_INIT (-viewport->origin.x,
-viewport->origin.y);
self->opacity = 1.0;
@@ -306,14 +305,14 @@ gsk_gpu_node_processor_add_images (GskGpuNodeProcessor *self,
static void
rect_round_to_pixels (const graphene_rect_t *src,
const graphene_vec2_t *pixel_scale,
const graphene_point_t *pixel_scale,
const graphene_point_t *pixel_offset,
graphene_rect_t *dest)
{
float x, y, xscale, yscale, inv_xscale, inv_yscale;
xscale = graphene_vec2_get_x (pixel_scale);
yscale = graphene_vec2_get_y (pixel_scale);
xscale = pixel_scale->x;
yscale = pixel_scale->y;
inv_xscale = 1.0f / xscale;
inv_yscale = 1.0f / yscale;
@@ -327,19 +326,19 @@ rect_round_to_pixels (const graphene_rect_t *src,
}
static GskGpuImage *
gsk_gpu_node_processor_init_draw (GskGpuNodeProcessor *self,
GskGpuFrame *frame,
GdkMemoryDepth depth,
const graphene_vec2_t *scale,
const graphene_rect_t *viewport)
gsk_gpu_node_processor_init_draw (GskGpuNodeProcessor *self,
GskGpuFrame *frame,
GdkMemoryDepth depth,
const graphene_point_t *scale,
const graphene_rect_t *viewport)
{
GskGpuImage *image;
cairo_rectangle_int_t area;
area.x = 0;
area.y = 0;
area.width = ceilf (graphene_vec2_get_x (scale) * viewport->size.width);
area.height = ceilf (graphene_vec2_get_y (scale) * viewport->size.height);
area.width = ceilf (scale->x * viewport->size.width);
area.height = ceilf (scale->y * viewport->size.height);
image = gsk_gpu_device_create_offscreen_image (gsk_gpu_frame_get_device (frame),
FALSE,
@@ -398,7 +397,7 @@ gsk_gpu_node_processor_process (GskGpuFrame *frame,
static void
gsk_gpu_pattern_writer_init (GskGpuPatternWriter *self,
GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
const graphene_point_t *offset,
const graphene_rect_t *bounds)
{
@@ -623,8 +622,8 @@ gsk_gpu_node_processor_rect_is_integer (GskGpuNodeProcessor *self,
cairo_rectangle_int_t *int_rect)
{
graphene_rect_t transformed_rect;
float scale_x = graphene_vec2_get_x (&self->scale);
float scale_y = graphene_vec2_get_y (&self->scale);
float scale_x = self->scale.x;
float scale_y = self->scale.y;
switch (gsk_transform_get_category (self->modelview))
{
@@ -758,7 +757,7 @@ gsk_gpu_node_processor_image_op (GskGpuNodeProcessor *self,
static GskGpuImage *
gsk_gpu_get_node_as_image (GskGpuFrame *frame,
const graphene_rect_t *clip_bounds,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
GskRenderNode *node,
graphene_rect_t *out_bounds)
{
@@ -994,8 +993,8 @@ gsk_gpu_node_processor_blur_op (GskGpuNodeProcessor *self,
if (!gsk_rect_intersection (rect, &clip_rect, &intermediate_rect))
return;
width = ceilf (graphene_vec2_get_x (&self->scale) * intermediate_rect.size.width);
height = ceilf (graphene_vec2_get_y (&self->scale) * intermediate_rect.size.height);
width = ceilf (self->scale.x * intermediate_rect.size.width);
height = ceilf (self->scale.y * intermediate_rect.size.height);
intermediate = gsk_gpu_device_create_offscreen_image (gsk_gpu_frame_get_device (self->frame),
FALSE,
@@ -1535,7 +1534,7 @@ gsk_gpu_node_processor_add_transform_node (GskGpuNodeProcessor *self,
GskRenderNode *child;
GskTransform *transform;
graphene_point_t old_offset;
graphene_vec2_t old_scale;
graphene_point_t old_scale;
GskTransform *old_modelview;
GskGpuClip old_clip;
@@ -1568,13 +1567,13 @@ gsk_gpu_node_processor_add_transform_node (GskGpuNodeProcessor *self,
gsk_transform_to_affine (transform, &scale_x, &scale_y, &dx, &dy);
gsk_gpu_clip_scale (&self->clip, &old_clip, scale_x, scale_y);
self->offset.x = (self->offset.x + dx) / scale_x;
self->offset.y = (self->offset.y + dy) / scale_y;
graphene_vec2_init (&self->scale, fabs (scale_x), fabs (scale_y));
graphene_vec2_multiply (&self->scale, &old_scale, &self->scale);
self->offset = GRAPHENE_POINT_INIT ((self->offset.x + dx) / scale_x,
(self->offset.y + dy) / scale_y);
self->scale = GRAPHENE_POINT_INIT (fabsf (scale_x) * old_scale.x,
fabsf (scale_y) * old_scale.y);
self->modelview = gsk_transform_scale (self->modelview,
scale_x / fabs (scale_x),
scale_y / fabs (scale_y));
scale_x / fabsf (scale_x),
scale_y / fabsf (scale_y));
}
break;
@@ -1631,14 +1630,13 @@ gsk_gpu_node_processor_add_transform_node (GskGpuNodeProcessor *self,
old_modelview = gsk_transform_ref (self->modelview);
self->modelview = gsk_transform_scale (self->modelview,
graphene_vec2_get_x (&self->scale),
graphene_vec2_get_y (&self->scale));
self->scale.x, self->scale.y);
self->modelview = gsk_transform_transform (self->modelview, clip_transform);
gsk_transform_unref (clip_transform);
extract_scale_from_transform (self->modelview, &scale_x, &scale_y);
old_pixels = graphene_vec2_get_x (&old_scale) * graphene_vec2_get_y (&old_scale) *
old_pixels = old_scale.x * old_scale.y *
old_clip.rect.bounds.size.width * old_clip.rect.bounds.size.height;
new_pixels = scale_x * scale_y * self->clip.rect.bounds.size.width * self->clip.rect.bounds.size.height;
if (new_pixels > 2 * old_pixels)
@@ -1649,7 +1647,7 @@ gsk_gpu_node_processor_add_transform_node (GskGpuNodeProcessor *self,
}
self->modelview = gsk_transform_scale (self->modelview, 1 / scale_x, 1 / scale_y);
graphene_vec2_init (&self->scale, scale_x, scale_y);
self->scale = GRAPHENE_POINT_INIT (scale_x, scale_y);
self->offset = *graphene_point_zero ();
}
break;
@@ -1678,7 +1676,7 @@ gsk_gpu_node_processor_create_transform_pattern (GskGpuPatternWriter *self,
GskRenderNode *child;
GskTransform *transform;
graphene_point_t old_offset;
graphene_vec2_t old_scale;
graphene_point_t old_scale;
graphene_rect_t old_bounds;
gboolean result;
@@ -1721,8 +1719,7 @@ gsk_gpu_node_processor_create_transform_pattern (GskGpuPatternWriter *self,
self->bounds.size.width *= inv_sx;
self->bounds.size.height *= inv_sy;
self->offset = GRAPHENE_POINT_INIT (0, 0);
graphene_vec2_init (&self->scale, fabs (sx), fabs (sy));
graphene_vec2_multiply (&self->scale, &old_scale, &self->scale);
self->scale = GRAPHENE_POINT_INIT (fabsf (sx) * old_scale.x, fabsf (sy) * old_scale.y);
}
break;
@@ -1744,9 +1741,9 @@ gsk_gpu_node_processor_create_transform_pattern (GskGpuPatternWriter *self,
gsk_gpu_pattern_writer_append_uint (self, GSK_GPU_PATTERN_POSITION_POP);
gsk_gpu_pattern_writer_pop_stack (self);
self->scale = old_scale;
self->bounds = old_bounds;
self->offset = old_offset;
self->scale = old_scale;
self->bounds = old_bounds;
self->offset = old_offset;
return result;
}
@@ -1806,8 +1803,8 @@ gsk_gpu_node_processor_add_color_node (GskGpuNodeProcessor *self,
return;
}
scale_x = graphene_vec2_get_x (&self->scale);
scale_y = graphene_vec2_get_y (&self->scale);
scale_x = self->scale.x;
scale_y = self->scale.y;
clipped = GRAPHENE_RECT_INIT (int_clipped.x / scale_x, int_clipped.y / scale_y,
int_clipped.width / scale_x, int_clipped.height / scale_y);
shader_clip = gsk_gpu_clip_get_shader_clip (&self->clip, graphene_point_zero(), &clipped);
@@ -1934,8 +1931,8 @@ gsk_gpu_node_processor_add_texture_node (GskGpuNodeProcessor *self,
}
if (gsk_gpu_frame_should_optimize (self->frame, GSK_GPU_OPTIMIZE_MIPMAP) &&
(gdk_texture_get_width (texture) > 2 * node->bounds.size.width * graphene_vec2_get_x (&self->scale) ||
gdk_texture_get_height (texture) > 2 * node->bounds.size.height * graphene_vec2_get_y (&self->scale)))
(gdk_texture_get_width (texture) > 2 * node->bounds.size.width * self->scale.x ||
gdk_texture_get_height (texture) > 2 * node->bounds.size.height * self->scale.y))
{
guint32 descriptor;
@@ -2001,8 +1998,8 @@ gsk_gpu_node_processor_create_texture_pattern (GskGpuPatternWriter *self,
}
if (gsk_gpu_frame_should_optimize (self->frame, GSK_GPU_OPTIMIZE_MIPMAP) &&
(gdk_texture_get_width (texture) > 2 * node->bounds.size.width * graphene_vec2_get_x (&self->scale) ||
gdk_texture_get_height (texture) > 2 * node->bounds.size.height * graphene_vec2_get_y (&self->scale)))
(gdk_texture_get_width (texture) > 2 * node->bounds.size.width * self->scale.x ||
gdk_texture_get_height (texture) > 2 * node->bounds.size.height * self->scale.y))
{
image = gsk_gpu_node_processor_ensure_image (self->frame,
image,
@@ -2045,8 +2042,8 @@ gsk_gpu_node_processor_add_texture_scale_node (GskGpuNodeProcessor *self,
guint32 descriptor;
gboolean need_mipmap, need_offscreen;
need_offscreen = self->modelview != NULL ||
!graphene_vec2_equal (&self->scale, graphene_vec2_one ());
need_offscreen = self->modelview != NULL ||
(self->scale.x != 1 || self->scale.y != 1);
if (need_offscreen)
{
GskGpuImage *offscreen;
@@ -2056,7 +2053,7 @@ gsk_gpu_node_processor_add_texture_scale_node (GskGpuNodeProcessor *self,
return;
gsk_rect_round_larger (&clip_bounds);
offscreen = gsk_gpu_render_pass_op_offscreen (self->frame,
graphene_vec2_one (),
&GRAPHENE_POINT_INIT (1, 1),
&clip_bounds,
node);
descriptor = gsk_gpu_node_processor_add_image (self, offscreen, GSK_GPU_SAMPLER_DEFAULT);
@@ -2996,7 +2993,7 @@ gsk_gpu_node_processor_add_glyph_node (GskGpuNodeProcessor *self,
offset.x += self->offset.x;
offset.y += self->offset.y;
scale = MAX (graphene_vec2_get_x (&self->scale), graphene_vec2_get_y (&self->scale));
scale = MAX (self->scale.x, self->scale.y);
inv_scale = 1.f / scale;
for (i = 0; i < num_glyphs; i++)
@@ -3067,7 +3064,7 @@ gsk_gpu_node_processor_create_glyph_pattern (GskGpuPatternWriter *self,
offset.x += self->offset.x;
offset.y += self->offset.y;
scale = MAX (graphene_vec2_get_x (&self->scale), graphene_vec2_get_y (&self->scale));
scale = MAX (self->scale.x, self->scale.y);
inv_scale = 1.f / scale;
gsk_gpu_pattern_writer_append_uint (self, GSK_GPU_PATTERN_GLYPHS);

12
gsk/gpu/gskgpurenderpassop.c
View File

@@ -335,16 +335,16 @@ gsk_gpu_render_pass_end_op (GskGpuFrame *frame,
}
GskGpuImage *
gsk_gpu_render_pass_op_offscreen (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_rect_t *viewport,
GskRenderNode *node)
gsk_gpu_render_pass_op_offscreen (GskGpuFrame *frame,
const graphene_point_t *scale,
const graphene_rect_t *viewport,
GskRenderNode *node)
{
GskGpuImage *image;
int width, height;
width = ceil (graphene_vec2_get_x (scale) * viewport->size.width);
height = ceil (graphene_vec2_get_y (scale) * viewport->size.height);
width = ceil (scale->x * viewport->size.width);
height = ceil (scale->y * viewport->size.height);
image = gsk_gpu_device_create_offscreen_image (gsk_gpu_frame_get_device (frame),
FALSE,

2
gsk/gpu/gskgpurenderpassopprivate.h
View File

@@ -24,7 +24,7 @@ void gsk_gpu_render_pass_end_op (GskGpuF
GskRenderPassType pass_type);
GskGpuImage * gsk_gpu_render_pass_op_offscreen (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
const graphene_rect_t *viewport,
GskRenderNode *node);

16
gsk/gpu/gskgpuuploadop.c
View File

@@ -451,12 +451,12 @@ static const GskGpuOpClass GSK_GPU_UPLOAD_CAIRO_OP_CLASS = {
};
GskGpuImage *
gsk_gpu_upload_cairo_op (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_rect_t *viewport,
GskGpuCairoFunc func,
gpointer user_data,
GDestroyNotify user_destroy)
gsk_gpu_upload_cairo_op (GskGpuFrame *frame,
const graphene_point_t *scale,
const graphene_rect_t *viewport,
GskGpuCairoFunc func,
gpointer user_data,
GDestroyNotify user_destroy)
{
GskGpuUploadCairoOp *self;
@@ -465,8 +465,8 @@ gsk_gpu_upload_cairo_op (GskGpuFrame *frame,
self->image = gsk_gpu_device_create_upload_image (gsk_gpu_frame_get_device (frame),
FALSE,
GDK_MEMORY_DEFAULT,
ceil (graphene_vec2_get_x (scale) * viewport->size.width),
ceil (graphene_vec2_get_y (scale) * viewport->size.height));
ceil (scale->x * viewport->size.width),
ceil (scale->y * viewport->size.height));
self->viewport = *viewport;
self->func = func;
self->user_data = user_data;

2
gsk/gpu/gskgpuuploadopprivate.h
View File

@@ -14,7 +14,7 @@ GskGpuImage * gsk_gpu_upload_texture_op_try (GskGpuF
GdkTexture *texture);
GskGpuImage * gsk_gpu_upload_cairo_op (GskGpuFrame *frame,
const graphene_vec2_t *scale,
const graphene_point_t *scale,
const graphene_rect_t *viewport,
GskGpuCairoFunc func,
gpointer user_data,

9
gsk/gskrectprivate.h
View File

@@ -122,6 +122,14 @@ gsk_gpu_rect_to_float (const graphene_rect_t *rect,
values[3] = rect->size.height;
}
static inline void
gsk_point_to_float (const graphene_point_t *point,
float values[2])
{
values[0] = point->x;
values[1] = point->y;
}
static inline void
gsk_rect_round_larger (graphene_rect_t *rect)
{
@@ -149,3 +157,4 @@ gsk_rect_scale (const graphene_rect_t *r,
res->size.width = r->size.width * sx;
res->size.height = r->size.height * sy;
}