alex/vello
1
0
Fork 0
mirror of https://github.com/italicsjenga/vello.git synced 2025-01-08 20:01:30 +11:00
Code Issues Packages Projects Releases Wiki Activity
912 commits 1 branch 0 tags 6.8 MiB
Commit graph

912 commits

This branch
This branch
All branches
Author SHA1 Message Date
Raph Levien 4470cdeccc Add unlicense 
Also take the opportunity to add more links to the README and adopt
the Rust code of conduct.
 2020-07-02 17:32:14 -07:00
Raph Levien 2c0f1d3cdd
Merge pull request #25 from ztlpn/fix-mapping-size 
Fix read mapping size in the Vulkan adapter
 2020-07-02 11:45:50 -07:00
Alexey Zatelepin 9667c34e7e Fix read mapping size (size is number of bytes, not number of items) 2020-07-02 21:32:17 +03:00
Raph Levien 368721e317
Merge pull request #24 from msiglreith/docs 
Shader documentation and a slight cleanup
 2020-07-01 17:30:16 -07:00
Raph Levien 46d0e7cea2
Merge pull request #22 from kant/patch-1 
Minor fix on paragraph 35
 2020-06-29 07:21:59 -07:00
msiglreith 1cc5c7ac0d Shader documentation and a slight cleanup 2020-06-28 15:37:27 +02:00
Darío Hereñú 9b87de055f
Minor fix on paragraph 35 2020-06-15 11:15:32 -03:00
Raph Levien 258bf6890a
Merge pull request #21 from msiglreith/update-winit 
Update winit example
 2020-06-14 14:42:48 -07:00
msiglreith eed71721eb Update winit example 2020-06-14 23:32:59 +02:00
Raph Levien dc5facd198
Merge pull request #20 from linebender/sorta 
A sorta-middle architecture
 2020-06-13 13:40:48 -07:00
Raph Levien 65f802894c Merge branch 'master' into sorta 2020-06-13 07:30:40 -07:00
Raph Levien daa7c9dd64 Merge branch 'sort_middle' 2020-06-11 16:17:40 -07:00
Raph Levien 73df5534a1
Merge pull request #19 from linebender/sort_middle 
Bring sort_middle branch to master
 2020-06-11 16:16:10 -07:00
Raph Levien b23113461b Minor cleanups 
Get rid of warnings. Do cargo update to bump deps.
 2020-06-10 14:10:28 -07:00
Raph Levien 79cc9da811 Fancy flattening 
Implement same flattening algorithm as kurbo.
 2020-06-09 20:45:19 -07:00
Raph Levien eaa1d261c3 Sederberg error metric 
Use proper math to compute number of subdivisions. This works but is not
very satisfying, as it over-subdivides.
 2020-06-09 18:43:49 -07:00
Raph Levien b571e0d10c Continue wiring up gpu-side flattening 
All segments given to path coarse raster are cubics. Flatten to
quadratics.

This works but the quality is not (yet) good.
 2020-06-09 17:56:11 -07:00
Raph Levien 0f44bc8b78 Start GPU-side flattening 
This starts the work on GPU-side flattening by plumbing curves through.
 2020-06-09 16:01:47 -07:00
Raph Levien 3a8227d025 Non-load balanced coarse path raster 
This is a bit of a revert of the load-balanced ("more parallel") coarse
path rasterizer, but includes fills and also uses atomicExchange.

I'm doing it this way because it should be considerably easier to do
flattening in this structure, even though there will be some performance
regression.
 2020-06-09 15:09:53 -07:00
Raph Levien 7118c8efc1 Fix backdrop of segments to left of viewport 
Make sure we account for backdrop in segments clipped out of viewport.
 2020-06-09 10:25:22 -07:00
Raph Levien 6db4e20bbb More parallel backdrop propagation 
This is a nice improvement but still not great on tiger.
 2020-06-06 08:23:40 -07:00
Raph Levien af0a1af8e1 Make fills work 
The backdrop propagation is slow but it does work.
 2020-06-05 22:40:44 -07:00
Raph Levien feeaa31fd1
Merge pull request #17 from linebender/cli_args 
Command line arguments
 2020-06-05 09:04:55 -07:00
Raph Levien f9f5961428 Use atomicExchange over atomicCompSwap 
Significant perf win (approx 2x in the path coarse rasterizer)
 2020-06-05 08:24:26 -07:00
Raph Levien e5dd9ae01e More parallel path coarse raster 
Use fancier load balancing algorithm for coarse rendering of paths.

Seems to work and an improvement in some cases.
 2020-06-04 17:42:33 -07:00
Raph Levien 877da4a98e Faster coarse raster 
Store a lot more tile context in shared memory and do the work from
that.
 2020-06-04 10:39:08 -07:00
Raph Levien e1aa9b2f5d Remove bbox guard 
It's probably not necessary.

This development still work in progress.
 2020-06-03 20:59:19 -07:00
Raph Levien 7f4a6523a8 Filter sparse tiles 
Have a more-parallel read of the tile structures based on bbox coverage,
and only set the bit when the tile isn't empty.

This is a speedup, but there is some duplicated work and it is possible
to improve it further.
 2020-06-03 17:55:42 -07:00
Raph Levien 63ba45c774 Fix performance issues 
Use larger workgroup for tile initialization (utilization was poor).
Provide correct element count to coarse rasterizer.
 2020-06-03 15:32:58 -07:00
Raph Levien ff8cee059c Optimize tile allocation 
Use parallel scheme to zero out tiles.
 2020-06-03 14:46:41 -07:00
Raph Levien 70a9c17e23 Continue building out pipeline 
Plumbs the new tiling scheme to k4. This works (stroke only) but still
has some performance issues.
 2020-06-03 12:21:09 -07:00
Raph Levien 294f6fd1db 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 09:29:25 -07:00
Raph Levien f3cb904f86 Add command line args for loading svg 2020-05-31 09:57:25 -07:00
Raph Levien f6ef1c16ab
Merge pull request #16 from linebender/new_merge 
More parallel merging in coarse raster
 2020-05-31 09:53:20 -07:00
Raph Levien c603cafc6c Merge branch 'more_svg' into new_merge 2020-05-31 09:19:34 -07:00
Raph Levien 2c185c3718 Simplify ringbuf 
We don't really need a ring buffer, as we only read what we're actually
going to process.
 2020-05-30 21:20:48 -07:00
Raph Levien 192ddc5eab Parallel merge 
The fancy stuff :)
 2020-05-30 21:11:13 -07:00
Raph Levien 121f29fef6 Merge one segment at a time 
No parallelism yet, but seems to improve performance.
 2020-05-30 08:51:52 -07:00
Raph Levien 894ef156e1 Change to new merge strategy in binning 
WIP

We get "device lost" on NV :/
 2020-05-29 20:06:16 -07:00
Raph Levien 9c0bdd664d
Merge pull request #14 from linebender/faster_k4 
Output multiple pixels per thread in k4
 2020-05-29 07:54:58 -07:00
Raph Levien 3e83972606 Improve SVG parsing 
WIP
 2020-05-28 11:48:36 -07:00
Raph Levien 319aa703c4 Output multiple pixels per thread in k4 
In kernel 4, compute a chunk of pixels rather than just one per thread.
This is a dramatic speedup.

(This commit cherry-picked from another working branch)
 2020-05-28 07:54:24 -07:00
Raph Levien 37a6f6aa51
Merge pull request #13 from linebender/par_coarse 
Parallelized segment output in coarse raster
 2020-05-28 07:46:42 -07:00
Raph Levien e16f68d89d Fix buffer overrun 
Was a little too eager zeroing out sh_is_segment[]
 2020-05-26 22:47:28 -07:00
Raph Levien dbcffb10db Reinstate fills 
Add fills back in.
 2020-05-25 15:27:03 -07:00
Raph Levien 3d422d9243 Allocate segment chunks in slabs 
Another speedup might be to special-case when the number of chunks in a
stroke or fill command is 1, then the segment header doesn't need
allocation and memory traffic is reduced. But right now we'll avoid the
complexity.
 2020-05-25 12:22:29 -07:00
Raph Levien 8eaf49a04d Checkpoint parallel output 
Parallel segment output seems to be working for strokes.
 2020-05-25 12:14:18 -07:00
Raph Levien 24b3def0a1 Start work on parallel segment output 
Output of segments is in parallel. Getting closer, some problems with
chaining but mostly correct.
 2020-05-24 21:02:19 -07:00
Raph Levien 55df3e6cc8 Fix linewidth math 
Coarse rasterization wasn't entirely taking line width into account.

Also fix swizzle in matrix (not yet used). And fix missing End command
in ptcl output (hasn't been a problem because buffer was cleared).
 2020-05-24 09:43:41 -07:00
Raph Levien b5e96b5b87
Merge pull request #12 from linebender/fills 
Make fills work in sort-middle pipeline
 2020-05-23 10:26:29 -07:00