GNUBoot
/
coreboot-kgpe-d16
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

cbgfx: use bilinear interpolation to scale bitmap 

This change replaces the current scaling algorithm (nearest neighbor) used
for bitmap rendering with the bilinear interpolation, which has much better
reproduction.

BUG=none
BRANCH=tot
TEST=Tested on Samus

Change-Id: I02520883debb7db40ffc19d4480244e0acabc818
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Original-Commit-Id: 764b383c1763a022728f2b2d9fb90e27c9e32e94
Original-Change-Id: I0ddd184343428904d04d8a76fe18a885529c7d3d
Original-Signed-off-by: Daisuke Nojiri <dnojiri@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/302195
Original-Reviewed-by: Randall Spangler <rspangler@chromium.org>
Reviewed-on: http://review.coreboot.org/11926
Tested-by: build bot (Jenkins)
This commit is contained in:
Daisuke Nojiri 2015-09-24 12:42:51 -07:00 committed by Patrick Georgi
parent bed16d9706
commit f86515244a
1 changed files with 44 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
payloads/libpayload/drivers/video/graphics.c
View File

@ -243,11 +243,20 @@ static int check_bound(const struct vector *image,
return within_box(&p, &bound) < 0;
}
static uint32_t bli(uint32_t q00, uint32_t q10, uint32_t q01, uint32_t q11,
struct fraction *tx, struct fraction *ty)
{
uint32_t r0 = (tx->nume * q10 + (tx->deno - tx->nume) * q00) / tx->deno;
uint32_t r1 = (tx->nume * q11 + (tx->deno - tx->nume) * q01) / tx->deno;
uint32_t p = (ty->nume * r1 + (ty->deno - ty->nume) * r0) / ty->deno;
return p;
}
static int draw_bitmap_v3(const struct vector *top_left,
const struct scale *scale,
const struct vector *image,
const struct bitmap_header_v3 *header,
const struct bitmap_palette_element_v3 *palette,
const struct bitmap_palette_element_v3 *pal,
const uint8_t *pixel_array)
{
const int bpp = header->bits_per_pixel;
@ -292,26 +301,50 @@ static int draw_bitmap_v3(const struct vector *top_left,
dir = -1;
}
/*
* Plot pixels scaled by the nearest neighbor interpolation. We scan
* Plot pixels scaled by the bilinear interpolation. We scan
* over the image on canvas (using d) and find the corresponding pixel
* in the bitmap data (using s).
*/
struct vector s, d;
struct vector s0, s1, d;
struct fraction tx, ty;
for (d.y = 0; d.y < image->height; d.y++, p.y += dir) {
s.y = d.y * scale->y.deno / scale->y.nume;
const uint8_t *data = pixel_array + s.y * y_stride;
s0.y = d.y * scale->y.deno / scale->y.nume;
s1.y = s0.y;
if (s0.y + 1 < ABS(header->height))
s1.y++;
ty.deno = scale->y.nume;
ty.nume = (d.y * scale->y.deno) % scale->y.nume;
const uint8_t *data0 = pixel_array + s0.y * y_stride;
const uint8_t *data1 = pixel_array + s1.y * y_stride;
p.x = top_left->x;
for (d.x = 0; d.x < image->width; d.x++, p.x++) {
s.x = d.x * scale->x.deno / scale->x.nume;
uint8_t index = data[s.x];
if (index >= header->colors_used) {
s0.x = d.x * scale->x.deno / scale->x.nume;
s1.x = s0.x;
if (s1.x + 1 < header->width)
s1.x++;
tx.deno = scale->x.nume;
tx.nume = (d.x * scale->x.deno) % scale->x.nume;
uint8_t c00 = data0[s0.x];
uint8_t c10 = data0[s1.x];
uint8_t c01 = data1[s0.x];
uint8_t c11 = data1[s1.x];
if (c00 >= header->colors_used
|| c10 >= header->colors_used
|| c01 >= header->colors_used
|| c11 >= header->colors_used) {
LOG("Color index exceeds palette boundary\n");
return CBGFX_ERROR_BITMAP_DATA;
}
const struct rgb_color rgb = {
.red = palette[index].red,
.green = palette[index].green,
.blue = palette[index].blue,
.red = bli(pal[c00].red, pal[c10].red,
pal[c01].red, pal[c11].red,
&tx, &ty),
.green = bli(pal[c00].green, pal[c10].green,
pal[c01].green, pal[c11].green,
&tx, &ty),
.blue = bli(pal[c00].blue, pal[c10].blue,
pal[c01].blue, pal[c11].blue,
&tx, &ty),
};
set_pixel(&p, calculate_color(&rgb));
}