rvandegrift
/
efl
forked from enlightenment/efl
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ector_software_rasterizer: Implement mask blending function. 

Summary:
Add Mask blend function(Add, Substract, Intersect, Difference)
this blending function only use mask blending case.

Test Plan: N/A

Reviewers: Hermet

Reviewed By: Hermet

Subscribers: cedric, kimcinoo, #reviewers, smohanty, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8509
This commit is contained in:
junsu choi 2019-04-11 19:12:40 +09:00 committed by Hermet Park
parent 04a41a3712
commit 83d5ea1a42
1 changed files with 163 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

173
src/lib/ector/software/ector_software_rasterizer.c
View File

@ -11,8 +11,23 @@
#include "draw.h"
//FIXME: This enum add temporarily to help understanding of additional code
//related to masking in prepare_mask.
//This needs to be formally declared through the eo class.
typedef enum _EFL_CANVAS_VG_NODE_BLEND_TYPE
{
EFL_CANVAS_VG_NODE_BLEND_TYPE_NONE = 0,
EFL_CANVAS_VG_NODE_BLEND_TYPE_ALPHA,
EFL_CANVAS_VG_NODE_BLEND_TYPE_ALPHA_INV,
EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_ADD,
EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_SUBSTRACT,
EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_INTERSECT,
EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_DIFFERENCE
}EFL_CANVAS_VG_NODE_BLEND_TYPE;
//
static void
_blend_color_argb(int count, const SW_FT_Span *spans, void *user_data)
_blend_argb(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
uint32_t color, *buffer, *target;
@ -34,7 +49,7 @@ _blend_color_argb(int count, const SW_FT_Span *spans, void *user_data)
}
static void
_blend_color_argb_with_maskA(int count, const SW_FT_Span *spans, void *user_data)
_blend_alpha(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
const int pix_stride = sd->raster_buffer->stride / 4;
@ -77,7 +92,7 @@ _blend_color_argb_with_maskA(int count, const SW_FT_Span *spans, void *user_data
}
static void
_blend_color_argb_with_maskInvA(int count, const SW_FT_Span *spans, void *user_data)
_blend_alpha_inv(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
const int pix_stride = sd->raster_buffer->stride / 4;
@ -120,6 +135,122 @@ _blend_color_argb_with_maskInvA(int count, const SW_FT_Span *spans, void *user_d
}
}
static void
_blend_mask_add(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
Ector_Software_Buffer_Base_Data *mask = sd->mask;
uint32_t color = DRAW_MUL4_SYM(sd->color, sd->mul_col);
RGBA_Comp_Func_Solid comp_func = efl_draw_func_solid_span_get(sd->op, color);
uint32_t *mbuffer = mask->pixels.u32;
while (count--)
{
uint32_t *ttarget = alloca(sizeof(uint32_t) * spans->len);
memset(ttarget, 0x00, sizeof(uint32_t) * spans->len);
uint32_t *mtarget = mbuffer + ((mask->generic->w * spans->y) + spans->x);
comp_func(ttarget, spans->len, color, spans->coverage);
for (int i = 0; i < spans->len; i++)
{
double adst = A_VAL(&mtarget[i]) == 0 ? 0 : (double)(A_VAL(&mtarget[i])) / (double)255;
double asrc = A_VAL(&ttarget[i]) == 0 ? 0 : (double)(A_VAL(&ttarget[i])) / (double)255;
uint32_t aout = (int)(((adst * (1 - asrc)) + asrc) * 255);
mtarget[i] = (aout<<24) + (0x00FFFFFF & mtarget[i]);
}
++spans;
}
}
static void
_blend_mask_sub(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
Ector_Software_Buffer_Base_Data *mask = sd->mask;
uint32_t color = DRAW_MUL4_SYM(sd->color, sd->mul_col);
RGBA_Comp_Func_Solid comp_func = efl_draw_func_solid_span_get(sd->op, color);
uint32_t *mtarget = mask->pixels.u32;
int tsize = sd->raster_buffer->generic->w * sd->raster_buffer->generic->h;
uint32_t *tbuffer = alloca(sizeof(uint32_t) * tsize);
memset(tbuffer, 0x00, sizeof(uint32_t) * tsize);
while (count--)
{
uint32_t *ttarget = tbuffer + ((mask->generic->w * spans->y) + spans->x);
comp_func(ttarget, spans->len, color, spans->coverage);
++spans;
}
for(int i = 0; i < tsize; i++)
{
double adst = A_VAL(&mtarget[i]) == 0 ? 0 : (double)(A_VAL(&mtarget[i])) / (double)255;
double asrc = A_VAL(&tbuffer[i]) == 0 ? 0 : (double)(A_VAL(&tbuffer[i])) / (double)255;
uint32_t aout = (int)((adst * (1 - asrc)) * 255);
mtarget[i] = (aout<<24) + (0x00FFFFFF & mtarget[i]);
}
}
static void
_blend_mask_ins(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
Ector_Software_Buffer_Base_Data *mask = sd->mask;
uint32_t color = DRAW_MUL4_SYM(sd->color, sd->mul_col);
RGBA_Comp_Func_Solid comp_func = efl_draw_func_solid_span_get(sd->op, color);
uint32_t *mtarget = mask->pixels.u32;
int tsize = sd->raster_buffer->generic->w * sd->raster_buffer->generic->h;
uint32_t *tbuffer = alloca(sizeof(uint32_t) * tsize);
memset(tbuffer, 0x00, sizeof(uint32_t) * tsize);
while (count--)
{
uint32_t *ttarget = tbuffer + ((mask->generic->w * spans->y) + spans->x);
comp_func(ttarget, spans->len, color, spans->coverage);
++spans;
}
for(int i = 0; i < tsize; i++)
{
double adst = A_VAL(&mtarget[i]) == 0 ? 0 : (double)(A_VAL(&mtarget[i])) / (double)255;
double asrc = A_VAL(&tbuffer[i]) == 0 ? 0 : (double)(A_VAL(&tbuffer[i])) / (double)255;
uint32_t aout = (int)((adst * asrc) * 255);
mtarget[i] = (aout<<24) + (0x00FFFFFF & mtarget[i]);
}
}
static void
_blend_mask_diff(int count, const SW_FT_Span *spans, void *user_data)
{
Span_Data *sd = user_data;
Ector_Software_Buffer_Base_Data *mask = sd->mask;
uint32_t color = DRAW_MUL4_SYM(sd->color, sd->mul_col);
RGBA_Comp_Func_Solid comp_func = efl_draw_func_solid_span_get(sd->op, color);
uint32_t *mbuffer = mask->pixels.u32;
while (count--)
{
uint32_t *ttarget = alloca(sizeof(uint32_t) * spans->len);
memset(ttarget, 0x00, sizeof(uint32_t) * spans->len);
uint32_t *mtarget = mbuffer + ((mask->generic->w * spans->y) + spans->x);
comp_func(ttarget, spans->len, color, spans->coverage);
for (int i = 0; i < spans->len; i++)
{
double adst = A_VAL(&mtarget[i]) == 0 ? 0 : (double)(A_VAL(&mtarget[i])) / (double)255;
double asrc = A_VAL(&ttarget[i]) == 0 ? 0 : (double)(A_VAL(&ttarget[i])) / (double)255;
uint32_t aout = (int)((((1 - adst) * asrc) + ((1 - asrc) * adst)) * 255);
mtarget[i] = (aout<<24) + (0x00FFFFFF & mtarget[i]);
}
++spans;
}
}
#define BLEND_GRADIENT_BUFFER_SIZE 2048
typedef void (*src_fetch) (unsigned int *buffer, Span_Data *data, int y, int x, int length);
@ -362,14 +493,32 @@ _adjust_span_fill_methods(Span_Data *spdata)
case Solid:
{
if (spdata->mask)
{
if (spdata->mask_op == 2)
spdata->unclipped_blend = &_blend_color_argb_with_maskInvA;
else
spdata->unclipped_blend = &_blend_color_argb_with_maskA;
}
{
switch (spdata->mask_op)
{
default:
case EFL_CANVAS_VG_NODE_BLEND_TYPE_ALPHA:
spdata->unclipped_blend = &_blend_alpha;
break;
case EFL_CANVAS_VG_NODE_BLEND_TYPE_ALPHA_INV:
spdata->unclipped_blend = &_blend_alpha_inv;
break;
case EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_ADD:
spdata->unclipped_blend = &_blend_mask_add;
break;
case EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_SUBSTRACT:
spdata->unclipped_blend = &_blend_mask_sub;
break;
case EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_INTERSECT:
spdata->unclipped_blend = &_blend_mask_ins;
break;
case EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_DIFFERENCE:
spdata->unclipped_blend = &_blend_mask_diff;
break;
}
}
else
spdata->unclipped_blend = &_blend_color_argb;
spdata->unclipped_blend = &_blend_argb;
}
break;
case LinearGradient:
@ -378,6 +527,10 @@ _adjust_span_fill_methods(Span_Data *spdata)
break;
}
//FIXME: Mask and mask case is not use clipping.
if (spdata->mask_op >= EFL_CANVAS_VG_NODE_BLEND_TYPE_MASK_ADD)
spdata->clip.enabled = EINA_FALSE;
// Clipping Function
if (spdata->clip.enabled)
{