alex/vello
1
0
Fork 0
mirror of https://github.com/italicsjenga/vello.git synced 2024-10-17 23:11:30 +11:00
Code Issues Packages Projects Releases Wiki Activity
vello/piet-gpu/shader/coarse.comp
Raph Levien e73049fe98 First cut at split blend stack 
Split the blend stack into register and memory segments. Do blending in registers up to that size, then spill to memory if needed.

This version may regress performance on Pixel 4, as it uses common memory for the blend stack, rather than keeping that memory read-only in fine rasterization, and using a separate buffer for blend stack. This needs investigation. It's possible we'll want to have single common memory as a config option, as it pools allocations and decreases the probability of failure.

Also a flaw in this version: there is no checking of memory overflow.

For understanding code history: this commit largely reverts #77, but there were some intervening changes to blending, and this commit also implements the split so some of the stack is in registers.

Closes #156
2022-05-16 11:12:33 -07:00

467 lines
19 KiB
GLSL
Raw Blame History

// SPDX-License-Identifier: Apache-2.0 OR MIT OR Unlicense
// The coarse rasterizer stage of the pipeline.
//
// As input we have the ordered partitions of paths from the binning phase and
// the annotated tile list of segments and backdrop per path.
//
// Each workgroup operating on one bin by stream compacting
// the elements corresponding to the bin.
//
// As output we have an ordered command stream per tile. Every tile from a path (backdrop + segment list) will be
// encoded.
#version 450
#extension GL_GOOGLE_include_directive : enable
#include "mem.h"
#include "setup.h"
layout(local_size_x = N_TILE, local_size_y = 1) in;
layout(binding = 1) readonly buffer ConfigBuf {
Config conf;
};
layout(binding = 2) readonly buffer SceneBuf {
uint[] scene;
};
#include "drawtag.h"
#include "bins.h"
#include "tile.h"
#include "ptcl.h"
#include "blend.h"
#define LG_N_PART_READ (7 + LG_WG_FACTOR)
#define N_PART_READ (1 << LG_N_PART_READ)
shared uint sh_elements[N_TILE];
// Number of elements in the partition; prefix sum.
shared uint sh_part_count[N_PART_READ];
shared Alloc sh_part_elements[N_PART_READ];
shared uint sh_bitmaps[N_SLICE][N_TILE];
shared uint sh_tile_count[N_TILE];
// The width of the tile rect for the element, intersected with this bin
shared uint sh_tile_width[N_TILE];
shared uint sh_tile_x0[N_TILE];
shared uint sh_tile_y0[N_TILE];
// These are set up so base + tile_y * stride + tile_x points to a Tile.
shared uint sh_tile_base[N_TILE];
shared uint sh_tile_stride[N_TILE];
#ifdef MEM_DEBUG
// Store allocs only when MEM_DEBUG to save shared memory traffic.
shared Alloc sh_tile_alloc[N_TILE];
void write_tile_alloc(uint el_ix, Alloc a) {
sh_tile_alloc[el_ix] = a;
}
Alloc read_tile_alloc(uint el_ix, bool mem_ok) {
return sh_tile_alloc[el_ix];
}
#else
void write_tile_alloc(uint el_ix, Alloc a) {
// No-op
}
Alloc read_tile_alloc(uint el_ix, bool mem_ok) {
// All memory.
return new_alloc(0, memory.length() * 4, mem_ok);
}
#endif
// The maximum number of commands per annotated element.
#define ANNO_COMMANDS 2
// Perhaps cmd_alloc should be a global? This is a style question.
bool alloc_cmd(inout Alloc cmd_alloc, inout CmdRef cmd_ref, inout uint cmd_limit) {
if (cmd_ref.offset < cmd_limit) {
return true;
}
MallocResult new_cmd = malloc(PTCL_INITIAL_ALLOC);
if (new_cmd.failed) {
return false;
}
CmdJump jump = CmdJump(new_cmd.alloc.offset);
Cmd_Jump_write(cmd_alloc, cmd_ref, jump);
cmd_alloc = new_cmd.alloc;
cmd_ref = CmdRef(cmd_alloc.offset);
// Reserve space for the maximum number of commands and a potential jump.
cmd_limit = cmd_alloc.offset + PTCL_INITIAL_ALLOC - (ANNO_COMMANDS + 1) * Cmd_size;
return true;
}
void write_fill(Alloc alloc, inout CmdRef cmd_ref, Tile tile, float linewidth) {
if (linewidth < 0.0) {
if (tile.tile.offset != 0) {
CmdFill cmd_fill = CmdFill(tile.tile.offset, tile.backdrop);
Cmd_Fill_write(alloc, cmd_ref, cmd_fill);
cmd_ref.offset += 4 + CmdFill_size;
} else {
Cmd_Solid_write(alloc, cmd_ref);
cmd_ref.offset += 4;
}
} else {
CmdStroke cmd_stroke = CmdStroke(tile.tile.offset, 0.5 * linewidth);
Cmd_Stroke_write(alloc, cmd_ref, cmd_stroke);
cmd_ref.offset += 4 + CmdStroke_size;
}
}
void main() {
// Could use either linear or 2d layouts for both dispatch and
// invocations within the workgroup. We'll use variables to abstract.
uint width_in_bins = (conf.width_in_tiles + N_TILE_X - 1) / N_TILE_X;
uint bin_ix = width_in_bins * gl_WorkGroupID.y + gl_WorkGroupID.x;
uint partition_ix = 0;
uint n_partitions = (conf.n_elements + N_TILE - 1) / N_TILE;
uint th_ix = gl_LocalInvocationID.x;
// Coordinates of top left of bin, in tiles.
uint bin_tile_x = N_TILE_X * gl_WorkGroupID.x;
uint bin_tile_y = N_TILE_Y * gl_WorkGroupID.y;
// Per-tile state
uint tile_x = gl_LocalInvocationID.x % N_TILE_X;
uint tile_y = gl_LocalInvocationID.x / N_TILE_X;
uint this_tile_ix = (bin_tile_y + tile_y) * conf.width_in_tiles + bin_tile_x + tile_x;
Alloc cmd_alloc = slice_mem(conf.ptcl_alloc, this_tile_ix * PTCL_INITIAL_ALLOC, PTCL_INITIAL_ALLOC);
CmdRef cmd_ref = CmdRef(cmd_alloc.offset);
// Reserve space for the maximum number of commands and a potential jump.
uint cmd_limit = cmd_ref.offset + PTCL_INITIAL_ALLOC - (ANNO_COMMANDS + 1) * Cmd_size;
// The nesting depth of the clip stack
uint clip_depth = 0;
// State for the "clip zero" optimization. If it's nonzero, then we are
// currently in a clip for which the entire tile has an alpha of zero, and
// the value is the depth after the "begin clip" of that element.
uint clip_zero_depth = 0;
// I'm sure we can figure out how to do this with at least one fewer register...
// Items up to rd_ix have been read from sh_elements
uint rd_ix = 0;
// Items up to wr_ix have been written into sh_elements
uint wr_ix = 0;
// Items between part_start_ix and ready_ix are ready to be transferred from sh_part_elements
uint part_start_ix = 0;
uint ready_ix = 0;
cmd_ref.offset += 4;
// Accounting for allocation of blend memory
uint render_blend_depth = 0;
uint max_blend_depth = 0;
uint drawmonoid_start = conf.drawmonoid_alloc.offset >> 2;
uint drawtag_start = conf.drawtag_offset >> 2;
uint drawdata_start = conf.drawdata_offset >> 2;
uint drawinfo_start = conf.drawinfo_alloc.offset >> 2;
bool mem_ok = mem_error == NO_ERROR;
while (true) {
for (uint i = 0; i < N_SLICE; i++) {
sh_bitmaps[i][th_ix] = 0;
}
// parallel read of input partitions
do {
if (ready_ix == wr_ix && partition_ix < n_partitions) {
part_start_ix = ready_ix;
uint count = 0;
if (th_ix < N_PART_READ && partition_ix + th_ix < n_partitions) {
uint in_ix = (conf.bin_alloc.offset >> 2) + ((partition_ix + th_ix) * N_TILE + bin_ix) * 2;
count = read_mem(conf.bin_alloc, in_ix);
uint offset = read_mem(conf.bin_alloc, in_ix + 1);
sh_part_elements[th_ix] = new_alloc(offset, count * BinInstance_size, mem_ok);
}
// prefix sum of counts
for (uint i = 0; i < LG_N_PART_READ; i++) {
if (th_ix < N_PART_READ) {
sh_part_count[th_ix] = count;
}
barrier();
if (th_ix < N_PART_READ) {
if (th_ix >= (1u << i)) {
count += sh_part_count[th_ix - (1u << i)];
}
}
barrier();
}
if (th_ix < N_PART_READ) {
sh_part_count[th_ix] = part_start_ix + count;
}
barrier();
ready_ix = sh_part_count[N_PART_READ - 1];
partition_ix += N_PART_READ;
}
// use binary search to find element to read
uint ix = rd_ix + th_ix;
if (ix >= wr_ix && ix < ready_ix && mem_ok) {
uint part_ix = 0;
for (uint i = 0; i < LG_N_PART_READ; i++) {
uint probe = part_ix + (uint(N_PART_READ / 2) >> i);
if (ix >= sh_part_count[probe - 1]) {
part_ix = probe;
}
}
ix -= part_ix > 0 ? sh_part_count[part_ix - 1] : part_start_ix;
Alloc bin_alloc = sh_part_elements[part_ix];
BinInstanceRef inst_ref = BinInstanceRef(bin_alloc.offset);
BinInstance inst = BinInstance_read(bin_alloc, BinInstance_index(inst_ref, ix));
sh_elements[th_ix] = inst.element_ix;
}
barrier();
wr_ix = min(rd_ix + N_TILE, ready_ix);
} while (wr_ix - rd_ix < N_TILE && (wr_ix < ready_ix || partition_ix < n_partitions));
// We've done the merge and filled the buffer.
// Read one element, compute coverage.
uint tag = Drawtag_Nop;
uint element_ix;
if (th_ix + rd_ix < wr_ix) {
element_ix = sh_elements[th_ix];
tag = scene[drawtag_start + element_ix];
}
// Bounding box of element in pixel coordinates.
uint tile_count;
switch (tag) {
case Drawtag_FillColor:
case Drawtag_FillImage:
case Drawtag_FillLinGradient:
case Drawtag_FillRadGradient:
case Drawtag_BeginClip:
case Drawtag_EndClip:
uint drawmonoid_base = drawmonoid_start + 4 * element_ix;
uint path_ix = memory[drawmonoid_base];
Path path = Path_read(conf.tile_alloc, PathRef(conf.tile_alloc.offset + path_ix * Path_size));
uint stride = path.bbox.z - path.bbox.x;
sh_tile_stride[th_ix] = stride;
int dx = int(path.bbox.x) - int(bin_tile_x);
int dy = int(path.bbox.y) - int(bin_tile_y);
int x0 = clamp(dx, 0, N_TILE_X);
int y0 = clamp(dy, 0, N_TILE_Y);
int x1 = clamp(int(path.bbox.z) - int(bin_tile_x), 0, N_TILE_X);
int y1 = clamp(int(path.bbox.w) - int(bin_tile_y), 0, N_TILE_Y);
sh_tile_width[th_ix] = uint(x1 - x0);
sh_tile_x0[th_ix] = x0;
sh_tile_y0[th_ix] = y0;
tile_count = uint(x1 - x0) * uint(y1 - y0);
// base relative to bin
uint base = path.tiles.offset - uint(dy * stride + dx) * Tile_size;
sh_tile_base[th_ix] = base;
Alloc path_alloc = new_alloc(path.tiles.offset,
(path.bbox.z - path.bbox.x) * (path.bbox.w - path.bbox.y) * Tile_size, mem_ok);
write_tile_alloc(th_ix, path_alloc);
break;
default:
tile_count = 0;
break;
}
// Prefix sum of sh_tile_count
sh_tile_count[th_ix] = tile_count;
for (uint i = 0; i < LG_N_TILE; i++) {
barrier();
if (th_ix >= (1u << i)) {
tile_count += sh_tile_count[th_ix - (1u << i)];
}
barrier();
sh_tile_count[th_ix] = tile_count;
}
barrier();
uint total_tile_count = sh_tile_count[N_TILE - 1];
for (uint ix = th_ix; ix < total_tile_count; ix += N_TILE) {
// Binary search to find element
uint el_ix = 0;
for (uint i = 0; i < LG_N_TILE; i++) {
uint probe = el_ix + (uint(N_TILE / 2) >> i);
if (ix >= sh_tile_count[probe - 1]) {
el_ix = probe;
}
}
uint element_ix = sh_elements[el_ix];
uint tag = scene[drawtag_start + element_ix];
uint seq_ix = ix - (el_ix > 0 ? sh_tile_count[el_ix - 1] : 0);
uint width = sh_tile_width[el_ix];
uint x = sh_tile_x0[el_ix] + seq_ix % width;
uint y = sh_tile_y0[el_ix] + seq_ix / width;
bool include_tile = false;
if (mem_ok) {
Tile tile = Tile_read(read_tile_alloc(el_ix, mem_ok),
TileRef(sh_tile_base[el_ix] + (sh_tile_stride[el_ix] * y + x) * Tile_size));
bool is_clip = (tag & 1) != 0;
// Always include the tile if it contains a path segment.
// For draws, include the tile if it is solid.
// For clips, include the tile if it is empty - this way, logic
// below will suppress the drawing of inner elements.
// For blends, include the tile if
// (blend_mode, composition_mode) != (Normal, SrcOver)
bool is_blend = false;
if (is_clip) {
uint drawmonoid_base = drawmonoid_start + 4 * element_ix;
uint scene_offset = memory[drawmonoid_base + 2];
uint dd = drawdata_start + (scene_offset >> 2);
uint blend = scene[dd];
is_blend = (blend != BlendComp_default);
}
include_tile = tile.tile.offset != 0 || (tile.backdrop == 0) == is_clip
|| is_blend;
}
if (include_tile) {
uint el_slice = el_ix / 32;
uint el_mask = 1u << (el_ix & 31);
atomicOr(sh_bitmaps[el_slice][y * N_TILE_X + x], el_mask);
}
}
barrier();
// Output draw objects for this tile. The thread does a sequential walk
// through the draw objects.
uint slice_ix = 0;
uint bitmap = sh_bitmaps[0][th_ix];
while (mem_ok) {
if (bitmap == 0) {
slice_ix++;
if (slice_ix == N_SLICE) {
break;
}
bitmap = sh_bitmaps[slice_ix][th_ix];
if (bitmap == 0) {
continue;
}
}
uint element_ref_ix = slice_ix * 32 + findLSB(bitmap);
uint element_ix = sh_elements[element_ref_ix];
// Clear LSB
bitmap &= bitmap - 1;
uint drawtag = scene[drawtag_start + element_ix];
if (clip_zero_depth == 0) {
Tile tile = Tile_read(read_tile_alloc(element_ref_ix, mem_ok),
TileRef(sh_tile_base[element_ref_ix] +
(sh_tile_stride[element_ref_ix] * tile_y + tile_x) * Tile_size));
uint drawmonoid_base = drawmonoid_start + 4 * element_ix;
uint scene_offset = memory[drawmonoid_base + 2];
uint info_offset = memory[drawmonoid_base + 3];
uint dd = drawdata_start + (scene_offset >> 2);
uint di = drawinfo_start + (info_offset >> 2);
switch (drawtag) {
case Drawtag_FillColor:
float linewidth = uintBitsToFloat(memory[di]);
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
write_fill(cmd_alloc, cmd_ref, tile, linewidth);
uint rgba = scene[dd];
Cmd_Color_write(cmd_alloc, cmd_ref, CmdColor(rgba));
cmd_ref.offset += 4 + CmdColor_size;
break;
case Drawtag_FillLinGradient:
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
linewidth = uintBitsToFloat(memory[di]);
write_fill(cmd_alloc, cmd_ref, tile, linewidth);
CmdLinGrad cmd_lin;
cmd_lin.index = scene[dd];
cmd_lin.line_x = uintBitsToFloat(memory[di + 1]);
cmd_lin.line_y = uintBitsToFloat(memory[di + 2]);
cmd_lin.line_c = uintBitsToFloat(memory[di + 3]);
Cmd_LinGrad_write(cmd_alloc, cmd_ref, cmd_lin);
cmd_ref.offset += 4 + CmdLinGrad_size;
break;
case Drawtag_FillRadGradient:
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
linewidth = uintBitsToFloat(memory[di]);
write_fill(cmd_alloc, cmd_ref, tile, linewidth);
CmdRadGrad cmd_rad;
cmd_rad.index = scene[dd];
// Given that this is basically a memcpy, we might consider
// letting the fine raster read the info itself.
cmd_rad.mat = uintBitsToFloat(uvec4(memory[di + 1], memory[di + 2],
memory[di + 3], memory[di + 4]));
cmd_rad.xlat = uintBitsToFloat(uvec2(memory[di + 5], memory[di + 6]));
cmd_rad.c1 = uintBitsToFloat(uvec2(memory[di + 7], memory[di + 8]));
cmd_rad.ra = uintBitsToFloat(memory[di + 9]);
cmd_rad.roff = uintBitsToFloat(memory[di + 10]);
Cmd_RadGrad_write(cmd_alloc, cmd_ref, cmd_rad);
cmd_ref.offset += 4 + CmdRadGrad_size;
break;
case Drawtag_FillImage:
linewidth = uintBitsToFloat(memory[di]);
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
write_fill(cmd_alloc, cmd_ref, tile, linewidth);
uint index = scene[dd];
uint raw1 = scene[dd + 1];
ivec2 offset = ivec2(int(raw1 << 16) >> 16, int(raw1) >> 16);
Cmd_Image_write(cmd_alloc, cmd_ref, CmdImage(index, offset));
cmd_ref.offset += 4 + CmdImage_size;
break;
case Drawtag_BeginClip:
if (tile.tile.offset == 0 && tile.backdrop == 0) {
clip_zero_depth = clip_depth + 1;
} else {
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
Cmd_BeginClip_write(cmd_alloc, cmd_ref);
cmd_ref.offset += 4;
render_blend_depth++;
max_blend_depth = max(max_blend_depth, render_blend_depth);
}
clip_depth++;
break;
case Drawtag_EndClip:
clip_depth--;
if (!alloc_cmd(cmd_alloc, cmd_ref, cmd_limit)) {
break;
}
write_fill(cmd_alloc, cmd_ref, tile, -1.0);
uint blend = scene[dd];
Cmd_EndClip_write(cmd_alloc, cmd_ref, CmdEndClip(blend));
cmd_ref.offset += 4 + CmdEndClip_size;
render_blend_depth--;
break;
}
} else {
// In "clip zero" state, suppress all drawing
switch (drawtag) {
case Drawtag_BeginClip:
clip_depth++;
break;
case Drawtag_EndClip:
if (clip_depth == clip_zero_depth) {
clip_zero_depth = 0;
}
clip_depth--;
break;
}
}
}
barrier();
rd_ix += N_TILE;
if (rd_ix >= ready_ix && partition_ix >= n_partitions)
break;
}
if (bin_tile_x + tile_x < conf.width_in_tiles && bin_tile_y + tile_y < conf.height_in_tiles) {
Cmd_End_write(cmd_alloc, cmd_ref);
if (max_blend_depth > BLEND_STACK_SPLIT) {
// TODO: allocate blend memory and write result
}
}
}
Reference in a new issue View git blame Copy permalink