Commit Graph
907 Commits
Author SHA1 Message Date
Dzmitry Malyshau 1cc548a7a2 Make descriptors generic over Label 2020-04-30 09:55:52 -04:00
Dzmitry Malyshau 4c1ea02553 trace: hook up RON output 2020-04-30 09:55:52 -04:00
Dzmitry Malyshau 5035ffdfe2 Split limits and private features, add tracing module 2020-04-30 09:55:52 -04:00
Dzmitry Malyshau 0e533547a1 Split serde feature into trace+replay 2020-04-30 09:55:52 -04:00
Gabriel Majeri d45323505c Add links to docs in README 2020-04-27 09:27:34 -04:00
bors[bot]andDzmitry Malyshau bc065e4abc Merge #615
615: Use General allocator at all times for now r=grovesNL a=kvark

In all user-managed resources, we don't have control of the lifetime. Since we don't know when it's released, we can't use any more specific allocator kind than `General`.

Previously, we assumed that buffers created for MAP_READ+COPY_SRC, for example, were one-time buffers created to copy data. However, there appear to be cases where they were used to fill data in once, and then persistently used as a copy source destination.

In the future, one WebGPU data transfer story is settled, we'll be able to use `Linear` kind again for all internally managed uploads. I.e. writeBuffer, writeTexture, and createBuffeMappedOnlyAtStart.

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-25 02:08:54 +00:00
Dzmitry Malyshau fc68b5ef54 Use General allocator at all times for now 2020-04-24 21:59:41 -04:00
Dzmitry Malyshau 09d3623885 Minor error reporting fixes (#607) 2020-04-24 21:58:30 -04:00
Gabriel Majeri b8f8dd78fd Run tests in CI 2020-04-24 22:43:14 -02:30
Igor Shaposhnik 1088683643 Fix windows nightly CI 2020-04-24 10:34:13 -02:30
Igor Shaposhnik 84cb39ee73 Install clippy on CI 2020-04-24 01:22:40 -02:30
Igor Shaposhnik 5c172dd475 Migrate CI to GitHub Actions (#605) 2020-04-22 09:03:26 -04:00
Dzmitry Malyshau baf435d8b8 Add is_power_of_two function (#603) 2020-04-21 09:29:10 -04:00
Dzmitry Malyshau 5e458b5d45 Separate public usage from internal use (#601) 2020-04-21 09:12:40 -04:00
Aron Granberg 21b441d087 Improve error messages for mismatched binding types (#593) 2020-04-20 14:41:40 -04:00
bors[bot]andDzmitry Malyshau 1234c96373 Merge #596
596: Remove wgpu-native and wgpu-remote r=grovesNL a=kvark

Closes #587 

I'm still not sure if it's a good idea. Feedback is welcome!
For instance, even after we remove these two things, the repo will still contain a bunch of logic that Gecko wouldn't necessarily be interested in, such as the surface/swapchains logic.

It's also not clear to me how to properly organize the workflow with the wgpu-native being separate (btw, it's https://github.com/gfx-rs/wgpu-native). Do we still have the workspace here? Or do we just introduce a separate repo that will include all the stuff as sub-modules and have a single workspace?

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-18 21:40:51 +00:00
bors[bot]andDzmitry Malyshau e612ac1fe2 Merge #600
600: Fix host mapping buffer barriers r=kvark a=kvark



Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-18 20:58:52 +00:00
Dzmitry Malyshau df2c97ff0c Fix host mapping buffer barriers 2020-04-18 16:56:05 -04:00
bors[bot]andDzmitry Malyshau f4550cc4e3 Merge #598
598: Enable READ access for texture storage r=kvark a=kvark

This is a short-term workaround until we properly implement #597 

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-18 17:30:21 +00:00
Dzmitry Malyshau 75d0157aa7 Enable READ access for texture storage 2020-04-18 13:29:17 -04:00
bors[bot]andAlmar Klein daf45362bc Merge #591
591: Warn when binding a buffer that is still mapped r=kvark a=almarklein

Fixes #510

I'd like to make more contributions wrt validation. Though I'm quite new to Rust, so let's start small :)  

From the discussion in #510 I concluded that in this case the buffer has actually been dropped, does the error message make sense like this?

Do the messages logged with `log::warn`, end up in the same logging system as the layer validation messages? So a solution to get those messages into Python (for wgpu-py) would work for both kinds of validation messages?


Co-authored-by: Almar Klein <almar.klein@gmail.com>
2020-04-17 21:15:19 +00:00
bors[bot]andDzmitry Malyshau 4bf3d57bb5 Merge #595
595: Derive swapchain layout off the load operation r=kvark a=kvark

Fixes https://github.com/gfx-rs/wgpu-rs/issues/258

It's fairly simple. We expect there to be exactly one `LoadOp::Clear` for the swapchain attachment, and for it to be the first pass in a frame using it. Any other passes need to do `LoadOp::Load`.

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-17 04:01:08 +00:00
Dzmitry Malyshau dbb261b7f1 Remove wgpu-native and wgpu-remote 2020-04-16 22:35:59 -04:00
Dzmitry Malyshau 8eefe269c1 Derive swapchain layout off the load operation 2020-04-16 21:49:24 -04:00
Almar Klein e7005548c0 Warn when binding a buffer that is still mapped 2020-04-16 11:35:03 +02:00
bors[bot]andAron Granberg 6b0751f1f8 Merge #586
586: Improve drawing error messages r=kvark a=HalfVoxel



Co-authored-by: Aron Granberg <aron.granberg@gmail.com>
2020-04-15 12:59:58 +00:00
bors[bot]andAron Granberg 2b120018f5 Merge #583
583: Improve error messages when passing buffers with the wrong usage flags r=kvark a=HalfVoxel

e.g.

```
thread 'main' panicked at 'An invalid setIndexBuffer call has been made. The buffer usage is COPY_DST | VERTEX which does not contain required usage INDEX
```

Co-authored-by: Aron Granberg <aron.granberg@gmail.com>
2020-04-15 12:30:48 +00:00
Aron Granberg 41a6425fb2 Improve drawing error messages 2020-04-15 02:29:37 +02:00
bors[bot]andDzmitry Malyshau 6de38a7792 Merge #584
584: Fix vertex format enum to match the native header r=kvark a=kvark

This was one of the problems to get wgpu-rs examples on the web.

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-14 21:15:30 +00:00
Dzmitry Malyshau 82a7791763 Fix vertex format enum to match the native header 2020-04-14 17:14:17 -04:00
Aron Granberg 5f6accc206 Improve error messages when passing buffers with the wrong usage flags 2020-04-14 20:18:00 +02:00
bors[bot]andDzmitry Malyshau 729ecb1404 Merge #582
582: Track pipeline layout lifetime r=grovesNL a=kvark

Fixes #580
This one is interesting: it's not instantly destroyed when dropped, like bind group layouts. And it's not tracked for GPU usage like the resources. It's somewhat in-between. We have the following classes of tracking now:
  1. no tracking, destroyed on drop. Applies to: bind group layouts, adapters
  2. only CPU-side tracking. They go to "suspected" list on drop of self or anything that can point to them. Applies to: pipeline layouts
  3. full GPU tracking. They go to "suspected" list on drop, then get associated with active submission before destruction. Applies to: buffers, textures, views, bind groups, pipelines

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-14 13:08:29 +00:00
bors[bot]andDzmitry Malyshau 96a4a34b83 Merge #581
581: Fix buffer unmap warning r=kvark a=kvark

It's more consistent if neither map_buffer or unmap_buffer care about the status.

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-14 11:24:19 +00:00
Dzmitry Malyshau 3cc4fa51bc Track pipeline layout lifetime 2020-04-14 00:58:30 -04:00
Dzmitry Malyshau a0701b41cb Fix buffer unmap warning 2020-04-14 00:08:04 -04:00
bors[bot]andDzmitry Malyshau 49dbe08f37 Merge #578
578: Comment fixes, header update r=kvark a=kvark

Matches review fixes in https://phabricator.services.mozilla.com/D70707

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-13 22:49:11 +00:00
Dzmitry Malyshau 379ea37982 Comment fixes, header update 2020-04-13 18:48:08 -04:00
bors[bot]andDzmitry Malyshau 8d1185f3e9 Merge #577
577: Remove array layer count from texture descriptor r=grovesNL a=kvark

Matches https://github.com/gpuweb/gpuweb/pull/613

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-13 18:16:06 +00:00
Dzmitry Malyshau d9609ec269 Remove array layer count from texture descriptor 2020-04-13 11:44:10 -04:00
bors[bot]andAron Granberg f50dd7a299 Merge #570
570: Improve error message when bind group and bind group layout have different number of entries r=kvark a=HalfVoxel



Co-authored-by: Aron Granberg <aron.granberg@gmail.com>
2020-04-13 00:17:43 +00:00
bors[bot]andDzmitry Malyshau 312fa3da62 Merge #573
573: Fix read-only flags r=kvark a=kvark



Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-12 00:51:26 +00:00
Dzmitry Malyshau 357ee9a447 Fix read-only flags 2020-04-11 20:50:20 -04:00
Aron Granberg 3db1e6c10d Improve error message when bind group and bind group layout have different number of entries 2020-04-11 11:22:40 +02:00
bors[bot]andDzmitry Malyshau 6b627451bc Merge #565
565: Usage tracking fixes for swapchain and command buffers r=kvark a=kvark

Fixes #561 
see description in commit messages

Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-10 04:40:15 +00:00
Dzmitry Malyshau 920c00931d Fix tracking of the initial state during replacement.
When multiple "replace" style transitions are happening,
we weren't properly retaining the "first" state, which
is required for proper stitching of command buffers.

This logic is fixed and fortified with a new set of
"change" and "merge" tests in the track module.
2020-04-10 00:39:23 -04:00
Dzmitry Malyshau 74e9e89fe7 Refactor tracking of swapchain images as attachments
We were improperly detecting if a swapchain image has already
been used by a command buffer. In this case, we need to assume
that it's already in the PRESENT state.
2020-04-10 00:38:53 -04:00
bors[bot]andDzmitry Malyshau 65a0f4afa0 Merge #560
560: Update ffi for wgpu-remote r=kvark a=kvark



Co-authored-by: Dzmitry Malyshau <kvarkus@gmail.com>
2020-04-09 16:18:55 +00:00
bors[bot]andAaron Loucks 0d5442d1c0 Merge #562
562: Increase default maximum vertex buffer inputs from 8 to 16 r=kvark a=aloucks

Additionally, the input states are now stored in a `SmallVec` to enable higher limits.

Fixes #558

Co-authored-by: Aaron Loucks <aloucks@cofront.net>
2020-04-08 13:49:35 +00:00
Dzmitry Malyshau ee2e56446c Update ffi for wgpu-remote 2020-04-07 22:16:46 -04:00
Aaron Loucks b6ec8440e9 Increase default maximum vertex buffer inputs from 8 to 16
Additionally, the input states are now stored in a `SmallVec` to
enable higher limits.

Fixes #558
2020-04-07 21:00:06 -04:00