encoder: Add V4L2 Stateful encoder. #75
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
bgrzesik
changed the title
bgrzesik
force-pushed
the
enc-v4l2
branch
10 times, most recently
from
962bc40 to
7cafa42
Compare
bgrzesik
force-pushed
the
enc-v4l2
branch
5 times, most recently
from
26fea55 to
bb95bd1
Compare
|
Hey, I'd like to start reviewing this, do you think you could rebase on the latest ToT? |
Hey, sure, I'll do this today 😄 |
|
The PR is now rebased 😄 |
Gnurou
reviewed
May 27, 2024
| @@ -424,12 +424,18 @@ where | |||
| let picture = picture.render().context("picture render")?; | |||
There was a problem hiding this comment.
Which parser fails without this? Please add this to the description so we have a way to test.
There was a problem hiding this comment.
Brought up offline.
Gnurou
reviewed
Jun 3, 2024
src/backend/v4l2/encoder.rs
Outdated
| Unsupported(UnsupportedError), | ||
|
|
||
| #[error(transparent)] | ||
| InitializationError(#[from] InitializationError), |
There was a problem hiding this comment.
I think we want to split this variant into its own error type - since create only returns InitializationError (with one exception that we can add to the new type), it removes to need for the double map_err while also making our errors more local.
There was a problem hiding this comment.
I think it should be better now. PTAL 😄
bgrzesik
commented
Jun 6, 2024
| @@ -424,12 +424,18 @@ where | |||
| let picture = picture.render().context("picture render")?; | |||
There was a problem hiding this comment.
Brought up offline.
src/backend/v4l2/encoder.rs
Outdated
| Unsupported(UnsupportedError), | ||
|
|
||
| #[error(transparent)] | ||
| InitializationError(#[from] InitializationError), |
There was a problem hiding this comment.
I think it should be better now. PTAL 😄
bgrzesik
force-pushed
the
enc-v4l2
branch
2 times, most recently
from
709c5d5 to
12a11e9
Compare
bgrzesik
commented
Jun 7, 2024
This change moves the code responsible for generation of test frames to encoder common module and modifies the logic to use lambda to deduplicate the generation code for both NV12 and P010. Futhermore the fill_test_frame was renamed to fill_test_frame_nv12 to better reflect its function.
This commit exports the code that generates t value using timestamp and max number of frame for the test frames.
This commit remove tests that only served a utility function rather then testing.
This commit renames StatelessVideoEncoder into VideoEncoder and moves it into the encoder module. This change is motivated that upcoming stateful encoder shares a lot of similarities and the same trait will be used.
This commit update drm dependency to fix compilation for other targets.
…size The field FrameMetadata::display_resolution was redundant, therefore removing it.
Some parsers use 0x00000001 byte sequence for NALu slicing instead of 0x000001, causing parsing failures. This change prepends 0x00 byte before slice NALu in order to prevent those failures.
This change enables direct use of DmabufFrame and UserPtrFrame directly with Vaapi encoders.
We need this library for `fstat` in order to acquire dmabuf size. It will be used for fixing encoding failure on Intel MTL devices.
This change fixes VA_STATUS_ERROR_ENCODING_ERROR status when DmabufFrame is used for encoding. It is fixed by providing fd/plane size information to libva.
This change add ability to allocate UserPtrFrame using provided FrameLayout for finer allocations for platform specific needs.
This commit moves codec structs and their EncoderConfig from src/encoder/stateless up to src/encoder. This is purely cosmetic change now, but since those structs are used for both stateless and future stateful implementation, thus should reside in common modules. In the future it will enable disable some implementations on the compile time.
This change just moves PredictionStructure enum definition to encoder module. This will enable conditional compilation of stateful and stateless modules in the future.
This commit adds v4l2 feature which will be used for V4L2 based codecs. The feature depends on v4l2r library, which was included in the Cargo.toml as well. The Cargo.lock was updated accordingly.
This commit adds V4L2 stateful encoder implementation using v4l2r. Currently only H.264 encoding is implemented.
This commit adds VP9 stateful encoder with additional necessary v4l2 controls.
This commit adds H.265 stateful encoder with additional necessary v4l2 controls.
This commit adds VP8 stateful encoder with additional necessary v4l2 controls.
|
Merge after a few final tweaks, thanks! |
|
Thank you! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR is based on top of #73.