Chad Brokaw
e328bea0b8
Merge pull request #184 from dfrg/multi-surface
...
Separate Instance and Surface creation in HAL
2022-07-15 13:16:47 -04:00
Chad Brokaw
d6ffe970f9
Make graphics support non-optional for Vulkan
...
This also removes the new present flag, deletes some commented code and explains the lack of surface validation in device selection.
2022-07-15 13:14:12 -04:00
Chad Brokaw
fb952de1f3
Update piet-gpu/bin/android.rs
...
Co-authored-by: Raph Levien <raph.levien@gmail.com>
2022-07-15 12:49:40 -04:00
Chad Brokaw
007ec6cdc0
cargo fmt
2022-07-14 15:29:43 -04:00
Chad Brokaw
f9bbb25a4b
fix formatting
2022-07-14 15:08:18 -04:00
Chad Brokaw
ffaf8b06a4
restore .gitignore
2022-07-14 15:07:29 -04:00
Chad Brokaw
58721dc5c2
fix new API use for android example
2022-07-14 15:03:27 -04:00
Chad Brokaw
8de34f8728
remove shader gen directories
2022-07-14 14:57:17 -04:00
Chad Brokaw
af55dbfa4e
make it work on mac/metal
2022-07-14 14:52:44 -04:00
Chad Brokaw
9626eaa19b
separate instance and surface creation
...
This separates creation of Instance and Surfaces, allowing for rendering to multiple windows.
2022-07-14 14:46:46 -04:00
Commit by GitHub Action
bbdd4432f5
commit compiled shaders
2022-07-14 14:28:45 +00:00
Commit by GitHub Action
d529d3b0e8
merge from dev branch - dev
2022-07-14 14:28:25 +00:00
Raph Levien
bfa4abf642
Merge pull request #181 from linebender/mem2
...
Implement robust dynamic memory
2022-07-14 07:27:38 -07:00
Raph Levien
78bdd08f01
Add ref_name to commit message
2022-07-14 07:25:18 -07:00
Commit by GitHub Action
cac14a0aa7
merge from dev branch
2022-07-14 04:33:57 +00:00
Raph Levien
d5b04a69dd
Merge pull request #183 from linebender/winit_0_26
...
Update winit to 0.26.1
2022-07-13 21:33:11 -07:00
Raph Levien
1a20fe78f6
Portability fixes
...
Makes the changes work on Windows and Android.
2022-07-13 17:43:28 -07:00
Raph Levien
61598d2da0
Gather timing statistics from RenderDriver
...
Also change command line binaries to use new abstraction.
2022-07-13 17:01:48 -07:00
Raph Levien
169f053003
Update winit to 0.26.1
...
This is mostly to silence security warnings about a nix vulnerability.
2022-07-13 12:45:17 -07:00
Raph Levien
9930c9d1de
rustfmt
2022-07-13 12:35:21 -07:00
Raph Levien
240f44a228
Implement robust dynamic memory
...
This is the core logic for robust dynamic memory. There are changes to both shaders and the driver logic.
On the shader side, failure information is more useful and fine grained. In particular, it now reports which stage failed and how much memory would have been required to make that stage succeed.
On the driver side, there is a new RenderDriver abstraction which owns command buffers (and associated query pools) and runs the logic to retry and reallocate buffers when necessary. There's also a fairly significant rework of the logic to produce the config block, as that overlaps the robust memory.
The RenderDriver abstraction may not stay. It was done this way to minimize code disruption, but arguably it should just be combined with Renderer.
Another change: the GLSL length() method on a buffer requires additional infrastructure (at least on Metal, where it needs a binding of its own), so we now pass that in as a field in the config.
This also moves blend memory to its own buffer. This worked out well because coarse rasterization can simply report the size of the blend buffer and it can be reallocated without needing to rerun the pipeline. In the previous state, blend allocations and ptcl writes were interleaved in coarse rasterization, so a failure of the former would require rerunning coarse. This should fix #83 (finally!)
There are a few loose ends. The binaries haven't (yet) been updated (I've been testing using a hand-written test program). Gradients weren't touched so still have a fixed size allocation. And the logic to calculate the new buffer size on allocation failure could be smarter.
Closes #175
2022-07-13 12:34:51 -07:00
Commit by GitHub Action
f6ea9308ba
commit compiled shaders
2022-07-13 19:27:07 +00:00
Raph Levien
64e6268059
Remove generated shaders from dev branch
2022-07-13 12:22:11 -07:00
Raph Levien
b77df99159
Merge pull request #178 from linebender/ci2
...
Shader compilation in GitHub Actions
2022-07-13 12:21:37 -07:00
Raph Levien
57b94a3bed
Merge pull request #180 from DJMcNab/pgpu_render_not_apple
...
Make `pgpu-render` compile on non apple platforms
2022-07-13 12:10:47 -07:00
Raph Levien
db49bed50c
Doc improvements
...
Explain the shader compilation approach. Also add links while I'm at it.
2022-07-13 12:07:55 -07:00
Raph Levien
a2f9e106cc
Cover all generated shader directories
...
Also remove 'shader/gen' gitignore line in main (it's present in dev).
2022-07-13 11:22:04 -07:00
Raph Levien
52f1ab228b
Merge branch 'main' into ci2
2022-07-13 10:40:46 -07:00
Raph Levien
b8b4aad9c4
Merge pull request #179 from linebender/desc_update
...
Add HAL methods to update descriptor set
2022-07-11 08:25:38 -07:00
Daniel McNab
35fe9b4d1f
Disable the android example when not compiling for android
2022-06-25 14:16:20 +01:00
Daniel McNab
28cbc8f199
Make pgpu-render compile on non apple platforms
...
Also add `resolver="2"` to the entire workspace, in case we gain a wgpu
dependency
2022-06-25 14:11:13 +01:00
Raph Levien
b5b75cc42d
Add Vulkan and DX12 backends to descriptor update
2022-06-23 15:44:42 -07:00
Raph Levien
95081971fe
Add HAL method to update buffer descriptor
...
This is WIP because only the Metal implementation is added.
Part of the work for #175
2022-06-23 12:05:09 -07:00
Raph Levien
682e4b769f
Shader compilation in GitHub Actions
...
This patch sets up very basic CI (right now just cargo fmt) but more importantly compiles shaders in a GitHub Action.
Any PR to branches other than main will run shader compilation. Any push to the dev branch will run shader compilation and then merge to main.
Closes #177
2022-06-16 21:19:39 -07:00
Raph Levien
60d197bb4e
Merge pull request #173 from linebender/blend_mem
...
Blend mem
2022-05-20 14:27:58 -07:00
Raph Levien
9df486ef35
Fix blend memory issues
2022-05-20 14:02:24 -07:00
Raph Levien
368954a643
Merge branch 'master' into blend_mem
...
This does the merge and also rebuilds the generated shaders.
2022-05-19 15:42:45 -07:00
Raph Levien
a4299c1f02
Merge pull request #172 from linebender/query_pool_size
...
Fix query pool size
2022-05-19 15:39:31 -07:00
Raph Levien
d62a40ea34
Update piet-gpu/bin/cli.rs
...
Co-authored-by: Chad Brokaw <cbrokaw@gmail.com>
2022-05-19 15:35:15 -07:00
Raph Levien
693ecd69e1
Make query pool size an associated constant
2022-05-19 15:28:46 -07:00
Chad Brokaw
663607dd12
Merge pull request #171 from dfrg/kurbo_conv
...
Add kurbo conversions to scene API
2022-05-19 18:18:47 -04:00
Chad Brokaw
5252a33c5d
Oops. Delete one remaining lossy conversion
2022-05-19 18:16:30 -04:00
Chad Brokaw
d30750e8a7
Remove poorly named functions
...
Moves the more descriptive comments to the free functions.
2022-05-19 18:07:07 -04:00
Chad Brokaw
39b773c611
Replace lossy conversions with methods
2022-05-19 18:04:00 -04:00
Raph Levien
ccac8f1ddd
Merge pull request #169 from linebender/fix_blends
...
Fix blending math
2022-05-19 14:35:00 -07:00
Raph Levien
ee6694729b
Fix query pool size
...
This was causing harmless warnings on mac but causing rendering to fail
on Pixel 6.
2022-05-19 14:31:00 -07:00
Chad Brokaw
938d6fc052
Add clip mode for blends
...
This matches the changes in the fix_blends PR.
2022-05-19 16:59:14 -04:00
Chad Brokaw
7f37355ba0
Merge branch 'linebender:master' into kurbo_conv
2022-05-18 19:05:04 -04:00
Raph Levien
708cb3d0b8
Merge pull request #170 from armansito/fix-links
...
Fix some broken links in vision document
2022-05-18 15:42:01 -07:00
Raph Levien
aac6513409
Compile shaders on Windows
...
Updates DXIL on generated shaders.
2022-05-18 15:41:00 -07:00
