vello/piet-gpu/shader/build.ninja

# Build file for shaders.

# You must have glslangValidator in your path, or patch here.

glslang_validator = glslangValidator

rule glsl
  command = $glslang_validator -V -o $out $in

build image.spv: glsl image.comp | scene.h


build elements.spv: glsl elements.comp | scene.h state.h annotated.h

build binning.spv: glsl binning.comp | annotated.h state.h bins.h setup.h

build tile_alloc.spv: glsl tile_alloc.comp | annotated.h tile.h setup.h

build path_coarse.spv: glsl path_coarse.comp | annotated.h tile.h setup.h

build backdrop.spv: glsl backdrop.comp | annotated.h tile.h setup.h

build coarse.spv: glsl coarse.comp | annotated.h bins.h ptcl.h setup.h

build kernel4.spv: glsl kernel4.comp | ptcl.h setup.h
Start image rendering Populates the piet-gpu subdir, with an extremely simple renderer. The main program saves the image to a PNG. Contains a few fixes (I was confused about the need for multiple bindings, as opposed to multiple descriptors within a binding). 2020-04-17 07:04:40 +10:00			`# Build file for shaders.`

			`# You must have glslangValidator in your path, or patch here.`

			`glslang_validator = glslangValidator`

			`rule glsl`
			`command = $glslang_validator -V -o $out $in`

Render circles WIP 2020-04-18 09:01:37 +10:00			`build image.spv: glsl image.comp \| scene.h`
First draft of kernel 1 Output of kernel 1 is validated by simple inspection, next step is to wire it up properly. 2020-04-21 10:15:36 +10:00
Experimenting with sort-middle Starting a prototype that explores the sort-middle approach. This commit has a prefix sum pass computing state per element. 2020-05-12 13:01:06 +10:00
Binning stage Adds a binning stage. This is a first draft, and a number of loose ends exist. 2020-05-13 06:38:26 +10:00			`build elements.spv: glsl elements.comp \| scene.h state.h annotated.h`

Rework right_edge computation in elements Trying to fit it into the fancy monad doesn't really work, so use a more straightforward approach to compute it from the aggregate. Also add yEdge logic (basically copying piet-metal). With a fix to ELEMENT_BINNING_RATIO (which I had simply gotten wrong), the example renders almost correctly, with small bounding box artifacts. 2020-05-21 09:36:09 +10:00			`build binning.spv: glsl binning.comp \| annotated.h state.h bins.h setup.h`
Starting coarse rasterizer Working down the pipeline. WIP 2020-05-14 08:35:19 +10:00
Experiment with new sorting scheme Path segments are unsorted, but other elements are using the same sort-middle approach as before. This is a checkpoint. At this point, there are unoptimized versions of tile init and coarse path raster, but it isn't wired up into a working pipeline. Also observing about a 3x performance regression in element processing, which needs to be investigated. 2020-06-03 10:10:20 +10:00			`build tile_alloc.spv: glsl tile_alloc.comp \| annotated.h tile.h setup.h`

			`build path_coarse.spv: glsl path_coarse.comp \| annotated.h tile.h setup.h`

Make fills work The backdrop propagation is slow but it does work. 2020-06-06 08:07:02 +10:00			`build backdrop.spv: glsl backdrop.comp \| annotated.h tile.h setup.h`

Starting coarse rasterizer Working down the pipeline. WIP 2020-05-14 08:35:19 +10:00			`build coarse.spv: glsl coarse.comp \| annotated.h bins.h ptcl.h setup.h`
Delete old-style kernels and buffers Pave the way for the coarse raster pass to write to the ptcl buffer. 2020-05-16 08:20:25 +10:00
			`build kernel4.spv: glsl kernel4.comp \| ptcl.h setup.h`