KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ilo: add BLT-based blitting methods to ilo_blitter 

Port BLT code in ilo_blit.c to BLT-based blitting methods of ilo_blitter.  Add
BLT-based clears.  The latter is verifed with util_clear(), but it is not in
use yet.
This commit is contained in:
Chia-I Wu 2013-06-17 13:59:51 +08:00
parent b4b3a5c6dc
commit ebfd7a61c0
4 changed files with 847 additions and 495 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/gallium/drivers/ilo/Makefile.sources
View File

@ -6,6 +6,7 @@ C_SOURCES := \
ilo_3d_pipeline_gen7.c \
ilo_blit.c \
ilo_blitter.c \
ilo_blitter_blt.c \
ilo_blitter_pipe.c \
ilo_context.c \
ilo_cp.c \

514
src/gallium/drivers/ilo/ilo_blit.c
View File

@ -25,7 +25,6 @@
* Chia-I Wu <olv@lunarg.com>
*/
#include "util/u_pack_color.h"
#include "util/u_surface.h"
#include "intel_reg.h"
@ -36,424 +35,6 @@
#include "ilo_screen.h"
#include "ilo_blit.h"
/*
* From the Sandy Bridge PRM, volume 1 part 5, page 7:
*
* "The BLT engine is capable of transferring very large quantities of
* graphics data. Any graphics data read from and written to the
* destination is permitted to represent a number of pixels that occupies
* up to 65,536 scan lines and up to 32,768 bytes per scan line at the
* destination. The maximum number of pixels that may be represented per
* scan line's worth of graphics data depends on the color depth."
*/
static const int gen6_max_bytes_per_scanline = 32768;
static const int gen6_max_scanlines = 65536;
static void
ilo_blit_own_blt_ring(struct ilo_context *ilo)
{
ilo_cp_set_ring(ilo->cp, ILO_CP_RING_BLT);
ilo_cp_set_owner(ilo->cp, NULL, 0);
}
static void
gen6_MI_FLUSH_DW(struct ilo_context *ilo)
{
const uint8_t cmd_len = 4;
struct ilo_cp *cp = ilo->cp;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, MI_FLUSH_DW | (cmd_len - 2));
ilo_cp_write(cp, 0);
ilo_cp_write(cp, 0);
ilo_cp_write(cp, 0);
ilo_cp_end(cp);
}
static void
gen6_MI_LOAD_REGISTER_IMM(struct ilo_context *ilo, uint32_t reg, uint32_t val)
{
const uint8_t cmd_len = 3;
struct ilo_cp *cp = ilo->cp;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, MI_LOAD_REGISTER_IMM | (cmd_len - 2));
ilo_cp_write(cp, reg);
ilo_cp_write(cp, val);
ilo_cp_end(cp);
}
static void
gen6_XY_COLOR_BLT(struct ilo_context *ilo, struct intel_bo *dst_bo,
uint32_t dst_offset, int16_t dst_pitch,
enum intel_tiling_mode dst_tiling,
int16_t x1, int16_t y1, int16_t x2, int16_t y2,
uint32_t color,
uint8_t rop, int cpp, bool write_alpha)
{
const uint8_t cmd_len = 6;
struct ilo_cp *cp = ilo->cp;
int dst_align, dst_pitch_shift;
uint32_t dw0, dw1;
dw0 = XY_COLOR_BLT_CMD | (cmd_len - 2);
if (dst_tiling == INTEL_TILING_NONE) {
dst_align = 4;
dst_pitch_shift = 0;
}
else {
dw0 |= XY_DST_TILED;
dst_align = (dst_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
dst_pitch_shift = 2;
}
assert(cpp == 4 || cpp == 2 || cpp == 1);
assert((x2 - x1) * cpp < gen6_max_bytes_per_scanline);
assert(y2 - y1 < gen6_max_scanlines);
assert(dst_offset % dst_align == 0 && dst_pitch % dst_align == 0);
dw1 = rop << 16 |
dst_pitch >> dst_pitch_shift;
switch (cpp) {
case 4:
dw0 |= XY_BLT_WRITE_RGB;
if (write_alpha)
dw0 |= XY_BLT_WRITE_ALPHA;
dw1 |= BR13_8888;
break;
case 2:
dw1 |= BR13_565;
break;
case 1:
dw1 |= BR13_8;
break;
}
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, y1 << 16 | x1);
ilo_cp_write(cp, y2 << 16 | x2);
ilo_cp_write_bo(cp, dst_offset, dst_bo,
INTEL_DOMAIN_RENDER, INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, color);
ilo_cp_end(cp);
}
static void
gen6_SRC_COPY_BLT(struct ilo_context *ilo, struct intel_bo *dst_bo,
uint32_t dst_offset, int16_t dst_pitch,
uint16_t width, uint16_t height,
struct intel_bo *src_bo,
uint32_t src_offset, int16_t src_pitch,
uint8_t rop, int cpp, bool write_alpha, bool dir_rtl)
{
const uint8_t cmd_len = 6;
struct ilo_cp *cp = ilo->cp;
uint32_t dw0, dw1;
assert(cpp == 4 || cpp == 2 || cpp == 1);
assert(width < gen6_max_bytes_per_scanline);
assert(height < gen6_max_scanlines);
/* offsets are naturally aligned and pitches are dword-aligned */
assert(dst_offset % cpp == 0 && dst_pitch % 4 == 0);
assert(src_offset % cpp == 0 && src_pitch % 4 == 0);
#ifndef SRC_COPY_BLT_CMD
#define SRC_COPY_BLT_CMD (CMD_2D | (0x43 << 22))
#endif
dw0 = SRC_COPY_BLT_CMD | (cmd_len - 2);
dw1 = rop << 16 | dst_pitch;
if (dir_rtl)
dw1 |= 1 << 30;
switch (cpp) {
case 4:
dw0 |= XY_BLT_WRITE_RGB;
if (write_alpha)
dw0 |= XY_BLT_WRITE_ALPHA;
dw1 |= BR13_8888;
break;
case 2:
dw1 |= BR13_565;
break;
case 1:
dw1 |= BR13_8;
break;
}
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, height << 16 | width);
ilo_cp_write_bo(cp, dst_offset, dst_bo, INTEL_DOMAIN_RENDER,
INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, src_pitch);
ilo_cp_write_bo(cp, src_offset, src_bo, INTEL_DOMAIN_RENDER, 0);
ilo_cp_end(cp);
}
static void
gen6_XY_SRC_COPY_BLT(struct ilo_context *ilo, struct intel_bo *dst_bo,
uint32_t dst_offset, int16_t dst_pitch,
enum intel_tiling_mode dst_tiling,
int16_t x1, int16_t y1, int16_t x2, int16_t y2,
struct intel_bo *src_bo,
uint32_t src_offset, int16_t src_pitch,
enum intel_tiling_mode src_tiling,
int16_t src_x, int16_t src_y,
uint8_t rop, int cpp, bool write_alpha)
{
const uint8_t cmd_len = 8;
struct ilo_cp *cp = ilo->cp;
int dst_align, dst_pitch_shift;
int src_align, src_pitch_shift;
uint32_t dw0, dw1;
dw0 = XY_SRC_COPY_BLT_CMD | (cmd_len - 2);
if (dst_tiling == INTEL_TILING_NONE) {
dst_align = 4;
dst_pitch_shift = 0;
}
else {
dw0 |= XY_DST_TILED;
dst_align = (dst_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
dst_pitch_shift = 2;
}
if (src_tiling == INTEL_TILING_NONE) {
src_align = 4;
src_pitch_shift = 0;
}
else {
dw0 |= XY_SRC_TILED;
src_align = (src_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
src_pitch_shift = 2;
}
assert(cpp == 4 || cpp == 2 || cpp == 1);
assert((x2 - x1) * cpp < gen6_max_bytes_per_scanline);
assert(y2 - y1 < gen6_max_scanlines);
assert(dst_offset % dst_align == 0 && dst_pitch % dst_align == 0);
assert(src_offset % src_align == 0 && src_pitch % src_align == 0);
dw1 = rop << 16 |
dst_pitch >> dst_pitch_shift;
switch (cpp) {
case 4:
dw0 |= XY_BLT_WRITE_RGB;
if (write_alpha)
dw0 |= XY_BLT_WRITE_ALPHA;
dw1 |= BR13_8888;
break;
case 2:
dw1 |= BR13_565;
break;
case 1:
dw1 |= BR13_8;
break;
}
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, y1 << 16 | x1);
ilo_cp_write(cp, y2 << 16 | x2);
ilo_cp_write_bo(cp, dst_offset, dst_bo, INTEL_DOMAIN_RENDER,
INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, src_y << 16 | src_x);
ilo_cp_write(cp, src_pitch >> src_pitch_shift);
ilo_cp_write_bo(cp, src_offset, src_bo, INTEL_DOMAIN_RENDER, 0);
ilo_cp_end(cp);
}
static bool
tex_copy_region(struct ilo_context *ilo,
struct ilo_texture *dst,
unsigned dst_level,
unsigned dst_x, unsigned dst_y, unsigned dst_z,
struct ilo_texture *src,
unsigned src_level,
const struct pipe_box *src_box)
{
const struct util_format_description *desc =
util_format_description(dst->bo_format);
const unsigned max_extent = 32767; /* INT16_MAX */
const uint8_t rop = 0xcc; /* SRCCOPY */
struct intel_bo *aper_check[3];
uint32_t swctrl;
int cpp, xscale, slice;
/* no W-tiling support */
if (dst->separate_s8 || src->separate_s8)
return false;
if (dst->bo_stride > max_extent || src->bo_stride > max_extent)
return false;
cpp = desc->block.bits / 8;
xscale = 1;
/* accommodate for larger cpp */
if (cpp > 4) {
if (cpp % 2 == 1)
return false;
cpp = (cpp % 4 == 0) ? 4 : 2;
xscale = (desc->block.bits / 8) / cpp;
}
ilo_blit_own_blt_ring(ilo);
/* make room if necessary */
aper_check[0] = ilo->cp->bo;
aper_check[1] = dst->bo;
aper_check[2] = src->bo;
if (intel_winsys_check_aperture_space(ilo->winsys, aper_check, 3))
ilo_cp_flush(ilo->cp);
swctrl = 0x0;
if (dst->tiling == INTEL_TILING_Y) {
swctrl |= BCS_SWCTRL_DST_Y << 16 |
BCS_SWCTRL_DST_Y;
}
if (src->tiling == INTEL_TILING_Y) {
swctrl |= BCS_SWCTRL_SRC_Y << 16 |
BCS_SWCTRL_SRC_Y;
}
if (swctrl) {
/*
* Most clients expect BLT engine to be stateless. If we have to set
* BCS_SWCTRL to a non-default value, we have to set it back in the same
* batch buffer.
*/
if (ilo_cp_space(ilo->cp) < (4 + 3) * 2 + src_box->depth * 8)
ilo_cp_flush(ilo->cp);
ilo_cp_assert_no_implicit_flush(ilo->cp, true);
/*
* From the Ivy Bridge PRM, volume 1 part 4, page 133:
*
* "SW is required to flush the HW before changing the polarity of
* this bit (Tile Y Destination/Source)."
*/
gen6_MI_FLUSH_DW(ilo);
gen6_MI_LOAD_REGISTER_IMM(ilo, BCS_SWCTRL, swctrl);
swctrl &= ~(BCS_SWCTRL_DST_Y | BCS_SWCTRL_SRC_Y);
}
for (slice = 0; slice < src_box->depth; slice++) {
const struct ilo_texture_slice *dst_slice =
&dst->slice_offsets[dst_level][dst_z + slice];
const struct ilo_texture_slice *src_slice =
&src->slice_offsets[src_level][src_box->z + slice];
unsigned x1, y1, x2, y2, src_x, src_y;
x1 = (dst_slice->x + dst_x) * xscale;
y1 = dst_slice->y + dst_y;
x2 = (x1 + src_box->width) * xscale;
y2 = y1 + src_box->height;
src_x = (src_slice->x + src_box->x) * xscale;
src_y = src_slice->y + src_box->y;
x1 /= desc->block.width;
y1 /= desc->block.height;
x2 = (x2 + desc->block.width - 1) / desc->block.width;
y2 = (y2 + desc->block.height - 1) / desc->block.height;
src_x /= desc->block.width;
src_y /= desc->block.height;
if (x2 > max_extent || y2 > max_extent ||
src_x > max_extent || src_y > max_extent ||
(x2 - x1) * cpp > gen6_max_bytes_per_scanline)
break;
gen6_XY_SRC_COPY_BLT(ilo,
dst->bo, 0, dst->bo_stride, dst->tiling,
x1, y1, x2, y2,
src->bo, 0, src->bo_stride, src->tiling,
src_x, src_y, rop, cpp, true);
}
if (swctrl) {
gen6_MI_FLUSH_DW(ilo);
gen6_MI_LOAD_REGISTER_IMM(ilo, BCS_SWCTRL, swctrl);
ilo_cp_assert_no_implicit_flush(ilo->cp, false);
}
return (slice == src_box->depth);
}
static bool
buf_copy_region(struct ilo_context *ilo,
struct ilo_buffer *dst, unsigned dst_offset,
struct ilo_buffer *src, unsigned src_offset,
unsigned size)
{
const uint8_t rop = 0xcc; /* SRCCOPY */
unsigned offset = 0;
struct intel_bo *aper_check[3];
ilo_blit_own_blt_ring(ilo);
/* make room if necessary */
aper_check[0] = ilo->cp->bo;
aper_check[1] = dst->bo;
aper_check[2] = src->bo;
if (intel_winsys_check_aperture_space(ilo->winsys, aper_check, 3))
ilo_cp_flush(ilo->cp);
while (size) {
unsigned width, height;
int16_t pitch;
width = size;
height = 1;
pitch = 0;
if (width > gen6_max_bytes_per_scanline) {
/* less than INT16_MAX and dword-aligned */
pitch = 32764;
width = pitch;
height = size / width;
if (height > gen6_max_scanlines)
height = gen6_max_scanlines;
}
gen6_SRC_COPY_BLT(ilo,
dst->bo, dst_offset + offset, pitch,
width, height,
src->bo, src_offset + offset, pitch,
rop, 1, true, false);
offset += pitch * height;
size -= width * height;
}
return true;
}
static void
ilo_resource_copy_region(struct pipe_context *pipe,
struct pipe_resource *dst,
@ -462,73 +43,16 @@ ilo_resource_copy_region(struct pipe_context *pipe,
struct pipe_resource *src,
unsigned src_level,
const struct pipe_box *src_box)
{
bool success;
if (dst->target != PIPE_BUFFER && src->target != PIPE_BUFFER) {
success = tex_copy_region(ilo_context(pipe),
ilo_texture(dst), dst_level, dstx, dsty, dstz,
ilo_texture(src), src_level, src_box);
}
else if (dst->target == PIPE_BUFFER && src->target == PIPE_BUFFER) {
const unsigned dst_offset = dstx;
const unsigned src_offset = src_box->x;
const unsigned size = src_box->width;
assert(dst_level == 0 && dsty == 0 && dstz == 0);
assert(src_level == 0 &&
src_box->y == 0 &&
src_box->z == 0 &&
src_box->height == 1 &&
src_box->depth == 1);
success = buf_copy_region(ilo_context(pipe),
ilo_buffer(dst), dst_offset, ilo_buffer(src), src_offset, size);
}
else {
success = false;
}
if (!success) {
util_resource_copy_region(pipe, dst, dst_level,
dstx, dsty, dstz, src, src_level, src_box);
}
}
static bool
blitter_xy_color_blt(struct pipe_context *pipe,
struct pipe_resource *res,
int16_t x1, int16_t y1,
int16_t x2, int16_t y2,
uint32_t color)
{
struct ilo_context *ilo = ilo_context(pipe);
struct ilo_texture *tex = ilo_texture(res);
const int cpp = util_format_get_blocksize(tex->bo_format);
const uint8_t rop = 0xf0; /* PATCOPY */
struct intel_bo *aper_check[2];
/* how to support Y-tiling? */
if (tex->tiling == INTEL_TILING_Y)
return false;
if (ilo_blitter_blt_copy_resource(ilo->blitter,
dst, dst_level, dstx, dsty, dstz,
src, src_level, src_box))
return;
/* nothing to clear */
if (x1 >= x2 || y1 >= y2)
return true;
ilo_blit_own_blt_ring(ilo);
/* make room if necessary */
aper_check[0] = ilo->cp->bo;
aper_check[1] = tex->bo;
if (intel_winsys_check_aperture_space(ilo->winsys, aper_check, 2))
ilo_cp_flush(ilo->cp);
gen6_XY_COLOR_BLT(ilo,
tex->bo, 0, tex->bo_stride, tex->tiling,
x1, y1, x2, y2, color, rop, cpp, true);
return true;
util_resource_copy_region(&ilo->base, dst, dst_level,
dstx, dsty, dstz, src, src_level, src_box);
}
static void
@ -551,7 +75,6 @@ ilo_clear_render_target(struct pipe_context *pipe,
unsigned width, unsigned height)
{
struct ilo_context *ilo = ilo_context(pipe);
union util_color packed;
if (!width || !height || dstx >= dst->width || dsty >= dst->height)
return;
@ -561,13 +84,8 @@ ilo_clear_render_target(struct pipe_context *pipe,
if (dsty + height > dst->height)
height = dst->height - dsty;
util_pack_color(color->f, dst->format, &packed);
/* try HW blit first */
if (blitter_xy_color_blt(pipe, dst->texture,
dstx, dsty,
dstx + width, dsty + height,
packed.ui))
if (ilo_blitter_blt_clear_rt(ilo->blitter,
dst, color, dstx, dsty, width, height))
return;
ilo_blitter_pipe_clear_rt(ilo->blitter,
@ -585,11 +103,17 @@ ilo_clear_depth_stencil(struct pipe_context *pipe,
{
struct ilo_context *ilo = ilo_context(pipe);
/*
* The PRM claims that HW blit supports Y-tiling since GEN6, but it does
* not tell us how to program it. Since depth buffers are always Y-tiled,
* HW blit will not work.
*/
if (!width || !height || dstx >= dst->width || dsty >= dst->height)
return;
if (dstx + width > dst->width)
width = dst->width - dstx;
if (dsty + height > dst->height)
height = dst->height - dsty;
if (ilo_blitter_blt_clear_zs(ilo->blitter,
dst, clear_flags, depth, stencil, dstx, dsty, width, height))
return;
ilo_blitter_pipe_clear_zs(ilo->blitter,
dst, clear_flags, depth, stencil, dstx, dsty, width, height);

22
src/gallium/drivers/ilo/ilo_blitter.h
View File

@ -77,4 +77,26 @@ ilo_blitter_pipe_clear_fb(struct ilo_blitter *blitter,
const union pipe_color_union *color,
double depth, unsigned stencil);
bool
ilo_blitter_blt_copy_resource(struct ilo_blitter *blitter,
struct pipe_resource *dst, unsigned dst_level,
unsigned dst_x, unsigned dst_y, unsigned dst_z,
struct pipe_resource *src, unsigned src_level,
const struct pipe_box *src_box);
bool
ilo_blitter_blt_clear_rt(struct ilo_blitter *blitter,
struct pipe_surface *rt,
const union pipe_color_union *color,
unsigned x, unsigned y,
unsigned width, unsigned height);
bool
ilo_blitter_blt_clear_zs(struct ilo_blitter *blitter,
struct pipe_surface *zs,
unsigned clear_flags,
double depth, unsigned stencil,
unsigned x, unsigned y,
unsigned width, unsigned height);
#endif /* ILO_BLITTER_H */

805
src/gallium/drivers/ilo/ilo_blitter_blt.c Normal file
View File

@ -0,0 +1,805 @@
/*
* Mesa 3-D graphics library
*
* Copyright (C) 2013 LunarG, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
*
* Authors:
* Chia-I Wu <olv@lunarg.com>
*/
#include "util/u_pack_color.h"
#include "intel_reg.h"
#include "ilo_context.h"
#include "ilo_cp.h"
#include "ilo_resource.h"
#include "ilo_blitter.h"
#ifndef COLOR_BLT_CMD
#define COLOR_BLT_CMD (CMD_2D | (0x40 << 22))
#endif
#ifndef SRC_COPY_BLT_CMD
#define SRC_COPY_BLT_CMD (CMD_2D | (0x43 << 22))
#endif
enum gen6_blt_mask {
GEN6_BLT_MASK_8,
GEN6_BLT_MASK_16,
GEN6_BLT_MASK_32,
GEN6_BLT_MASK_32_LO,
GEN6_BLT_MASK_32_HI,
};
/*
* From the Sandy Bridge PRM, volume 1 part 5, page 7:
*
* "The BLT engine is capable of transferring very large quantities of
* graphics data. Any graphics data read from and written to the
* destination is permitted to represent a number of pixels that occupies
* up to 65,536 scan lines and up to 32,768 bytes per scan line at the
* destination. The maximum number of pixels that may be represented per
* scan line's worth of graphics data depends on the color depth."
*/
static const int gen6_max_bytes_per_scanline = 32768;
static const int gen6_max_scanlines = 65536;
static void
gen6_emit_MI_FLUSH_DW(struct ilo_dev_info *dev, struct ilo_cp *cp)
{
const uint8_t cmd_len = 4;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, MI_FLUSH_DW | (cmd_len - 2));
ilo_cp_write(cp, 0);
ilo_cp_write(cp, 0);
ilo_cp_write(cp, 0);
ilo_cp_end(cp);
}
static void
gen6_emit_MI_LOAD_REGISTER_IMM(struct ilo_dev_info *dev,
uint32_t reg, uint32_t val,
struct ilo_cp *cp)
{
const uint8_t cmd_len = 3;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, MI_LOAD_REGISTER_IMM | (cmd_len - 2));
ilo_cp_write(cp, reg);
ilo_cp_write(cp, val);
ilo_cp_end(cp);
}
static uint32_t
gen6_translate_blt_value_mask(enum gen6_blt_mask value_mask)
{
switch (value_mask) {
case GEN6_BLT_MASK_8: return BR13_8;
case GEN6_BLT_MASK_16: return BR13_565;
default: return BR13_8888;
}
}
static uint32_t
gen6_translate_blt_write_mask(enum gen6_blt_mask write_mask)
{
switch (write_mask) {
case GEN6_BLT_MASK_32: return XY_BLT_WRITE_RGB |
XY_BLT_WRITE_ALPHA;
case GEN6_BLT_MASK_32_LO: return XY_BLT_WRITE_RGB;
case GEN6_BLT_MASK_32_HI: return XY_BLT_WRITE_ALPHA;
default: return 0;
}
}
static uint32_t
gen6_translate_blt_cpp(enum gen6_blt_mask mask)
{
switch (mask) {
case GEN6_BLT_MASK_8: return 1;
case GEN6_BLT_MASK_16: return 2;
default: return 4;
}
}
static void
gen6_emit_COLOR_BLT(struct ilo_dev_info *dev,
struct intel_bo *dst_bo,
int16_t dst_pitch, uint32_t dst_offset,
uint16_t width, uint16_t height,
uint32_t pattern, uint8_t rop,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask,
struct ilo_cp *cp)
{
const uint8_t cmd_len = 5;
const int cpp = gen6_translate_blt_cpp(value_mask);
uint32_t dw0, dw1;
dw0 = COLOR_BLT_CMD |
gen6_translate_blt_write_mask(write_mask) |
(cmd_len - 2);
assert(width < gen6_max_bytes_per_scanline);
assert(height < gen6_max_scanlines);
/* offsets are naturally aligned and pitches are dword-aligned */
assert(dst_offset % cpp == 0 && dst_pitch % 4 == 0);
dw1 = rop << 16 |
gen6_translate_blt_value_mask(value_mask) |
dst_pitch;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, height << 16 | width);
ilo_cp_write_bo(cp, dst_offset, dst_bo, INTEL_DOMAIN_RENDER,
INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, pattern);
ilo_cp_end(cp);
}
static void
gen6_emit_XY_COLOR_BLT(struct ilo_dev_info *dev,
struct intel_bo *dst_bo,
enum intel_tiling_mode dst_tiling,
int16_t dst_pitch, uint32_t dst_offset,
int16_t x1, int16_t y1, int16_t x2, int16_t y2,
uint32_t pattern, uint8_t rop,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask,
struct ilo_cp *cp)
{
const uint8_t cmd_len = 6;
const int cpp = gen6_translate_blt_cpp(value_mask);
int dst_align, dst_pitch_shift;
uint32_t dw0, dw1;
dw0 = XY_COLOR_BLT_CMD |
gen6_translate_blt_write_mask(write_mask) |
(cmd_len - 2);
if (dst_tiling == INTEL_TILING_NONE) {
dst_align = 4;
dst_pitch_shift = 0;
}
else {
dw0 |= XY_DST_TILED;
dst_align = (dst_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
dst_pitch_shift = 2;
}
assert((x2 - x1) * cpp < gen6_max_bytes_per_scanline);
assert(y2 - y1 < gen6_max_scanlines);
assert(dst_offset % dst_align == 0 && dst_pitch % dst_align == 0);
dw1 = rop << 16 |
gen6_translate_blt_value_mask(value_mask) |
dst_pitch >> dst_pitch_shift;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, y1 << 16 | x1);
ilo_cp_write(cp, y2 << 16 | x2);
ilo_cp_write_bo(cp, dst_offset, dst_bo,
INTEL_DOMAIN_RENDER, INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, pattern);
ilo_cp_end(cp);
}
static void
gen6_emit_SRC_COPY_BLT(struct ilo_dev_info *dev,
struct intel_bo *dst_bo,
int16_t dst_pitch, uint32_t dst_offset,
uint16_t width, uint16_t height,
struct intel_bo *src_bo,
int16_t src_pitch, uint32_t src_offset,
bool dir_rtl, uint8_t rop,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask,
struct ilo_cp *cp)
{
const uint8_t cmd_len = 6;
const int cpp = gen6_translate_blt_cpp(value_mask);
uint32_t dw0, dw1;
dw0 = SRC_COPY_BLT_CMD |
gen6_translate_blt_write_mask(write_mask) |
(cmd_len - 2);
assert(width < gen6_max_bytes_per_scanline);
assert(height < gen6_max_scanlines);
/* offsets are naturally aligned and pitches are dword-aligned */
assert(dst_offset % cpp == 0 && dst_pitch % 4 == 0);
assert(src_offset % cpp == 0 && src_pitch % 4 == 0);
dw1 = rop << 16 |
gen6_translate_blt_value_mask(value_mask) |
dst_pitch;
if (dir_rtl)
dw1 |= 1 << 30;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, height << 16 | width);
ilo_cp_write_bo(cp, dst_offset, dst_bo, INTEL_DOMAIN_RENDER,
INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, src_pitch);
ilo_cp_write_bo(cp, src_offset, src_bo, INTEL_DOMAIN_RENDER, 0);
ilo_cp_end(cp);
}
static void
gen6_emit_XY_SRC_COPY_BLT(struct ilo_dev_info *dev,
struct intel_bo *dst_bo,
enum intel_tiling_mode dst_tiling,
int16_t dst_pitch, uint32_t dst_offset,
int16_t x1, int16_t y1, int16_t x2, int16_t y2,
struct intel_bo *src_bo,
enum intel_tiling_mode src_tiling,
int16_t src_pitch, uint32_t src_offset,
int16_t src_x, int16_t src_y, uint8_t rop,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask,
struct ilo_cp *cp)
{
const uint8_t cmd_len = 8;
const int cpp = gen6_translate_blt_cpp(value_mask);
int dst_align, dst_pitch_shift;
int src_align, src_pitch_shift;
uint32_t dw0, dw1;
dw0 = XY_SRC_COPY_BLT_CMD |
gen6_translate_blt_write_mask(write_mask) |
(cmd_len - 2);
if (dst_tiling == INTEL_TILING_NONE) {
dst_align = 4;
dst_pitch_shift = 0;
}
else {
dw0 |= XY_DST_TILED;
dst_align = (dst_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
dst_pitch_shift = 2;
}
if (src_tiling == INTEL_TILING_NONE) {
src_align = 4;
src_pitch_shift = 0;
}
else {
dw0 |= XY_SRC_TILED;
src_align = (src_tiling == INTEL_TILING_Y) ? 128 : 512;
/* in dwords when tiled */
src_pitch_shift = 2;
}
assert((x2 - x1) * cpp < gen6_max_bytes_per_scanline);
assert(y2 - y1 < gen6_max_scanlines);
assert(dst_offset % dst_align == 0 && dst_pitch % dst_align == 0);
assert(src_offset % src_align == 0 && src_pitch % src_align == 0);
dw1 = rop << 16 |
gen6_translate_blt_value_mask(value_mask) |
dst_pitch >> dst_pitch_shift;
ilo_cp_begin(cp, cmd_len);
ilo_cp_write(cp, dw0);
ilo_cp_write(cp, dw1);
ilo_cp_write(cp, y1 << 16 | x1);
ilo_cp_write(cp, y2 << 16 | x2);
ilo_cp_write_bo(cp, dst_offset, dst_bo, INTEL_DOMAIN_RENDER,
INTEL_DOMAIN_RENDER);
ilo_cp_write(cp, src_y << 16 | src_x);
ilo_cp_write(cp, src_pitch >> src_pitch_shift);
ilo_cp_write_bo(cp, src_offset, src_bo, INTEL_DOMAIN_RENDER, 0);
ilo_cp_end(cp);
}
static uint32_t
ilo_blitter_blt_begin(struct ilo_blitter *blitter, int max_cmd_size,
struct intel_bo *dst, enum intel_tiling_mode dst_tiling,
struct intel_bo *src, enum intel_tiling_mode src_tiling)
{
struct ilo_context *ilo = blitter->ilo;
struct intel_bo *aper_check[3];
int count;
uint32_t swctrl;
/* change ring */
ilo_cp_set_ring(ilo->cp, ILO_CP_RING_BLT);
ilo_cp_set_owner(ilo->cp, NULL, 0);
/* check aperture space */
aper_check[0] = ilo->cp->bo;
aper_check[1] = dst;
count = 2;
if (src) {
aper_check[2] = src;
count++;
}
if (intel_winsys_check_aperture_space(ilo->winsys, aper_check, count))
ilo_cp_flush(ilo->cp);
/* set BCS_SWCTRL */
swctrl = 0x0;
if (dst_tiling == INTEL_TILING_Y) {
swctrl |= BCS_SWCTRL_DST_Y << 16 |
BCS_SWCTRL_DST_Y;
}
if (src && src_tiling == INTEL_TILING_Y) {
swctrl |= BCS_SWCTRL_SRC_Y << 16 |
BCS_SWCTRL_SRC_Y;
}
if (swctrl) {
/*
* Most clients expect BLT engine to be stateless. If we have to set
* BCS_SWCTRL to a non-default value, we have to set it back in the same
* batch buffer.
*/
if (ilo_cp_space(ilo->cp) < (4 + 3) * 2 + max_cmd_size)
ilo_cp_flush(ilo->cp);
ilo_cp_assert_no_implicit_flush(ilo->cp, true);
/*
* From the Ivy Bridge PRM, volume 1 part 4, page 133:
*
* "SW is required to flush the HW before changing the polarity of
* this bit (Tile Y Destination/Source)."
*/
gen6_emit_MI_FLUSH_DW(ilo->dev, ilo->cp);
gen6_emit_MI_LOAD_REGISTER_IMM(ilo->dev, BCS_SWCTRL, swctrl, ilo->cp);
swctrl &= ~(BCS_SWCTRL_DST_Y | BCS_SWCTRL_SRC_Y);
}
return swctrl;
}
static void
ilo_blitter_blt_end(struct ilo_blitter *blitter, uint32_t swctrl)
{
struct ilo_context *ilo = blitter->ilo;
/* set BCS_SWCTRL back */
if (swctrl) {
gen6_emit_MI_FLUSH_DW(ilo->dev, ilo->cp);
gen6_emit_MI_LOAD_REGISTER_IMM(ilo->dev, BCS_SWCTRL, swctrl, ilo->cp);
ilo_cp_assert_no_implicit_flush(ilo->cp, false);
}
}
static bool
buf_clear_region(struct ilo_blitter *blitter,
struct ilo_buffer *dst,
unsigned dst_offset, unsigned dst_size,
uint32_t val,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask)
{
const uint8_t rop = 0xf0; /* PATCOPY */
const int cpp = gen6_translate_blt_cpp(value_mask);
struct ilo_context *ilo = blitter->ilo;
unsigned offset = 0;
if (dst_offset % cpp || dst_size % cpp)
return false;
ilo_blitter_blt_begin(blitter, 0,
dst->bo, INTEL_TILING_NONE, NULL, INTEL_TILING_NONE);
while (dst_size) {
unsigned width, height;
int16_t pitch;
width = dst_size;
height = 1;
pitch = 0;
if (width > gen6_max_bytes_per_scanline) {
/* less than INT16_MAX and dword-aligned */
pitch = 32764;
width = pitch;
height = dst_size / width;
if (height > gen6_max_scanlines)
height = gen6_max_scanlines;
}
gen6_emit_COLOR_BLT(ilo->dev, dst->bo, pitch, dst_offset + offset,
width, height, val, rop, value_mask, write_mask, ilo->cp);
offset += pitch * height;
dst_size -= width * height;
}
ilo_blitter_blt_end(blitter, 0);
return true;
}
static bool
buf_copy_region(struct ilo_blitter *blitter,
struct ilo_buffer *dst, unsigned dst_offset,
struct ilo_buffer *src, unsigned src_offset,
unsigned size)
{
const uint8_t rop = 0xcc; /* SRCCOPY */
struct ilo_context *ilo = blitter->ilo;
unsigned offset = 0;
ilo_blitter_blt_begin(blitter, 0,
dst->bo, INTEL_TILING_NONE, src->bo, INTEL_TILING_NONE);
while (size) {
unsigned width, height;
int16_t pitch;
width = size;
height = 1;
pitch = 0;
if (width > gen6_max_bytes_per_scanline) {
/* less than INT16_MAX and dword-aligned */
pitch = 32764;
width = pitch;
height = size / width;
if (height > gen6_max_scanlines)
height = gen6_max_scanlines;
}
gen6_emit_SRC_COPY_BLT(ilo->dev,
dst->bo, pitch, dst_offset + offset,
width, height,
src->bo, pitch, src_offset + offset,
false, rop, GEN6_BLT_MASK_8, GEN6_BLT_MASK_8,
ilo->cp);
offset += pitch * height;
size -= width * height;
}
ilo_blitter_blt_end(blitter, 0);
return true;
}
static bool
tex_clear_region(struct ilo_blitter *blitter,
struct ilo_texture *dst, unsigned dst_level,
const struct pipe_box *dst_box,
uint32_t val,
enum gen6_blt_mask value_mask,
enum gen6_blt_mask write_mask)
{
const int cpp = gen6_translate_blt_cpp(value_mask);
const unsigned max_extent = 32767; /* INT16_MAX */
const uint8_t rop = 0xf0; /* PATCOPY */
struct ilo_context *ilo = blitter->ilo;
uint32_t swctrl;
int slice;
/* no W-tiling support */
if (dst->separate_s8)
return false;
if (dst->bo_stride > max_extent)
return false;
swctrl = ilo_blitter_blt_begin(blitter, dst_box->depth * 6,
dst->bo, dst->tiling, NULL, INTEL_TILING_NONE);
for (slice = 0; slice < dst_box->depth; slice++) {
const struct ilo_texture_slice *dst_slice =
&dst->slice_offsets[dst_level][dst_box->z + slice];
unsigned x1, y1, x2, y2;
x1 = dst_slice->x + dst_box->x;
y1 = dst_slice->y + dst_box->y;
x2 = x1 + dst_box->width;
y2 = y1 + dst_box->height;
if (x2 > max_extent || y2 > max_extent ||
(x2 - x1) * cpp > gen6_max_bytes_per_scanline)
break;
gen6_emit_XY_COLOR_BLT(ilo->dev,
dst->bo, dst->tiling, dst->bo_stride, 0,
x1, y1, x2, y2, val, rop, value_mask, write_mask,
ilo->cp);
}
ilo_blitter_blt_end(blitter, swctrl);
return (slice == dst_box->depth);
}
static bool
tex_copy_region(struct ilo_blitter *blitter,
struct ilo_texture *dst,
unsigned dst_level,
unsigned dst_x, unsigned dst_y, unsigned dst_z,
struct ilo_texture *src,
unsigned src_level,
const struct pipe_box *src_box)
{
const struct util_format_description *desc =
util_format_description(dst->bo_format);
const unsigned max_extent = 32767; /* INT16_MAX */
const uint8_t rop = 0xcc; /* SRCCOPY */
struct ilo_context *ilo = blitter->ilo;
enum gen6_blt_mask mask;
uint32_t swctrl;
int cpp, xscale, slice;
/* no W-tiling support */
if (dst->separate_s8 || src->separate_s8)
return false;
if (dst->bo_stride > max_extent || src->bo_stride > max_extent)
return false;
cpp = desc->block.bits / 8;
xscale = 1;
/* accommodate for larger cpp */
if (cpp > 4) {
if (cpp % 2 == 1)
return false;
cpp = (cpp % 4 == 0) ? 4 : 2;
xscale = (desc->block.bits / 8) / cpp;
}
switch (cpp) {
case 1:
mask = GEN6_BLT_MASK_8;
break;
case 2:
mask = GEN6_BLT_MASK_16;
break;
case 4:
mask = GEN6_BLT_MASK_32;
break;
default:
return false;
break;
}
swctrl = ilo_blitter_blt_begin(blitter, src_box->depth * 8,
dst->bo, dst->tiling, src->bo, src->tiling);
for (slice = 0; slice < src_box->depth; slice++) {
const struct ilo_texture_slice *dst_slice =
&dst->slice_offsets[dst_level][dst_z + slice];
const struct ilo_texture_slice *src_slice =
&src->slice_offsets[src_level][src_box->z + slice];
unsigned x1, y1, x2, y2, src_x, src_y;
x1 = (dst_slice->x + dst_x) * xscale;
y1 = dst_slice->y + dst_y;
x2 = (x1 + src_box->width) * xscale;
y2 = y1 + src_box->height;
src_x = (src_slice->x + src_box->x) * xscale;
src_y = src_slice->y + src_box->y;
/* in blocks */
x1 /= desc->block.width;
y1 /= desc->block.height;
x2 = (x2 + desc->block.width - 1) / desc->block.width;
y2 = (y2 + desc->block.height - 1) / desc->block.height;
src_x /= desc->block.width;
src_y /= desc->block.height;
if (x2 > max_extent || y2 > max_extent ||
src_x > max_extent || src_y > max_extent ||
(x2 - x1) * cpp > gen6_max_bytes_per_scanline)
break;
gen6_emit_XY_SRC_COPY_BLT(ilo->dev,
dst->bo, dst->tiling, dst->bo_stride, 0,
x1, y1, x2, y2,
src->bo, src->tiling, src->bo_stride, 0,
src_x, src_y, rop, mask, mask,
ilo->cp);
}
ilo_blitter_blt_end(blitter, swctrl);
return (slice == src_box->depth);
}
bool
ilo_blitter_blt_copy_resource(struct ilo_blitter *blitter,
struct pipe_resource *dst, unsigned dst_level,
unsigned dst_x, unsigned dst_y, unsigned dst_z,
struct pipe_resource *src, unsigned src_level,
const struct pipe_box *src_box)
{
bool success;
if (dst->target == PIPE_BUFFER && src->target == PIPE_BUFFER) {
const unsigned dst_offset = dst_x;
const unsigned src_offset = src_box->x;
const unsigned size = src_box->width;
assert(dst_level == 0 && dst_y == 0 && dst_z == 0);
assert(src_level == 0 &&
src_box->y == 0 &&
src_box->z == 0 &&
src_box->height == 1 &&
src_box->depth == 1);
success = buf_copy_region(blitter,
ilo_buffer(dst), dst_offset, ilo_buffer(src), src_offset, size);
}
else if (dst->target != PIPE_BUFFER && src->target != PIPE_BUFFER) {
success = tex_copy_region(blitter,
ilo_texture(dst), dst_level, dst_x, dst_y, dst_z,
ilo_texture(src), src_level, src_box);
}
else {
success = false;
}
return success;
}
bool
ilo_blitter_blt_clear_rt(struct ilo_blitter *blitter,
struct pipe_surface *rt,
const union pipe_color_union *color,
unsigned x, unsigned y,
unsigned width, unsigned height)
{
const int cpp = util_format_get_blocksize(rt->format);
enum gen6_blt_mask mask;
union util_color packed;
bool success;
switch (cpp) {
case 1:
mask = GEN6_BLT_MASK_8;
break;
case 2:
mask = GEN6_BLT_MASK_16;
break;
case 4:
mask = GEN6_BLT_MASK_32;
break;
default:
return false;
break;
}
if (util_format_is_pure_integer(rt->format) ||
util_format_is_compressed(rt->format))
return false;
util_pack_color(color->f, rt->format, &packed);
if (rt->texture->target == PIPE_BUFFER) {
unsigned offset, end, size;
assert(y == 0 && height == 1);
offset = (rt->u.buf.first_element + x) * cpp;
end = (rt->u.buf.last_element + 1) * cpp;
size = width * cpp;
if (offset + size > end)
size = end - offset;
success = buf_clear_region(blitter, ilo_buffer(rt->texture),
offset, size, packed.ui, mask, mask);
}
else {
struct pipe_box box;
u_box_3d(x, y, rt->u.tex.first_layer, width, height,
rt->u.tex.last_layer - rt->u.tex.first_layer + 1, &box);
success = tex_clear_region(blitter, ilo_texture(rt->texture),
rt->u.tex.level, &box, packed.ui, mask, mask);
}
return success;
}
bool
ilo_blitter_blt_clear_zs(struct ilo_blitter *blitter,
struct pipe_surface *zs,
unsigned clear_flags,
double depth, unsigned stencil,
unsigned x, unsigned y,
unsigned width, unsigned height)
{
enum gen6_blt_mask value_mask, write_mask;
struct pipe_box box;
uint32_t val;
switch (zs->format) {
case PIPE_FORMAT_Z16_UNORM:
if (!(clear_flags & PIPE_CLEAR_DEPTH))
return true;
value_mask = GEN6_BLT_MASK_16;
write_mask = GEN6_BLT_MASK_16;
break;
case PIPE_FORMAT_Z32_FLOAT:
if (!(clear_flags & PIPE_CLEAR_DEPTH))
return true;
value_mask = GEN6_BLT_MASK_32;
write_mask = GEN6_BLT_MASK_32;
break;
case PIPE_FORMAT_Z24X8_UNORM:
if (!(clear_flags & PIPE_CLEAR_DEPTH))
return true;
value_mask = GEN6_BLT_MASK_32;
write_mask = GEN6_BLT_MASK_32_LO;
break;
case PIPE_FORMAT_Z24_UNORM_S8_UINT:
if (!(clear_flags & PIPE_CLEAR_DEPTHSTENCIL))
return true;
value_mask = GEN6_BLT_MASK_32;
if ((clear_flags & PIPE_CLEAR_DEPTHSTENCIL) == PIPE_CLEAR_DEPTHSTENCIL)
write_mask = GEN6_BLT_MASK_32;
else if (clear_flags & PIPE_CLEAR_DEPTH)
write_mask = GEN6_BLT_MASK_32_LO;
else
write_mask = GEN6_BLT_MASK_32_HI;
break;
default:
return false;
break;
}
val = util_pack_z_stencil(zs->format, depth, stencil);
u_box_3d(x, y, zs->u.tex.first_layer, width, height,
zs->u.tex.last_layer - zs->u.tex.first_layer + 1, &box);
assert(zs->texture->target != PIPE_BUFFER);
return tex_clear_region(blitter, ilo_texture(zs->texture),
zs->u.tex.level, &box, val, value_mask, write_mask);
}