P20 dcblock#11000
Open
piotrhoppeintel wants to merge 3 commits into
Open
Conversation
Signed-off-by: Piotr Hoppe <piotr.hoppe@intel.com>
Signed-off-by: Piotr Hoppe <piotr.hoppe@intel.com>
Convert the DC blocking filter from the deprecated audio_stream / process_audio_stream interface to the modern sof_source / sof_sink processing API. The module .process callback now acquires the source and sink circular buffers once via source_get_data() / sink_get_buffer() and passes them to the processing functions as struct cir_buf_source / struct cir_buf_sink views, mirroring the volume component. The generic path uses the shared cir_buf_samples_without_wrap_*() and cir_buf_wrap() helpers, while the HiFi3/HiFi4 variants derive the circular buffer bounds from the same views. Channel count is cached in comp_data during prepare(). Signed-off-by: Piotr Hoppe <piotr.hoppe@intel.com>
piotrhoppeintel
requested review from
a team,
dbaluta,
kv2019i,
lbetlej,
lgirdwood,
mmaka1 and
plbossart
as code owners
July 15, 2026 07:28
Contributor
There was a problem hiding this comment.
Pull request overview
Note
Copilot couldn't run its full agentic review because no GitHub Actions runner was available. Make sure your repository has a runner available to run Copilot's review, or add a copilot-setup-steps.yml file specifying one with the runs-on attribute. See the docs for more details.
Adds DC-block unit tests and updates the DC-block component to process audio via the SOF source/sink APIs (using circular buffer views) with updated DSP backends.
Changes:
- Add CMocka math-level and end-to-end process tests for dcblock (S16/S24/S32).
- Refactor dcblock processing functions (generic + HiFi3/HiFi4) to operate on circular buffer views and return status codes.
- Update module adapter integration to use
.processwith source/sink APIs; enable dcblock tests in unit-test config/CMake.
Reviewed changes
Copilot reviewed 10 out of 10 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
| test/cmocka/src/audio/dcblock/dcblock_process.c | New end-to-end processing test driving the module through source/sink APIs and verifying against a float reference. |
| test/cmocka/src/audio/dcblock/dcblock_math_test.c | New math-level bitexact/DC-removal/saturation tests for dcblock processing kernels. |
| test/cmocka/src/audio/dcblock/CMakeLists.txt | Adds dcblock tests and a minimal audio_for_dcblock library for linking. |
| test/cmocka/src/audio/CMakeLists.txt | Conditionally adds the dcblock test subdirectory when CONFIG_COMP_DCBLOCK is enabled. |
| src/audio/dcblock/dcblock_hifi4.c | Updates HiFi4 kernel to use circular buffer views and return int. |
| src/audio/dcblock/dcblock_hifi3.c | Updates HiFi3 kernel to use circular buffer views; manual wrap logic added. |
| src/audio/dcblock/dcblock_generic.c | Refactors generic kernels to use circular buffer views and return int. |
| src/audio/dcblock/dcblock.h | Updates dcblock_func signature and stores channel count in comp_data. |
| src/audio/dcblock/dcblock.c | Switches module processing to source/sink API; acquires/releases circular buffer views; validates channels. |
| src/arch/xtensa/configs/unit_test_defconfig | Enables CONFIG_COMP_DCBLOCK for unit test builds. |
Comment on lines
+229
to
+240
| cd->source_format = source_get_frm_fmt(sources[0]); | ||
|
|
||
| /* get sink data format */ | ||
| cd->sink_format = sink_get_frm_fmt(sinks[0]); | ||
|
|
||
| /* The processing uses a single channel count as the frame stride for | ||
| * both source and sink, so they must match to avoid corrupt output. | ||
| */ | ||
| if (source_get_channels(sources[0]) != sink_get_channels(sinks[0])) { | ||
| comp_err(dev, "mismatch source/sink stream channels"); | ||
| return -EINVAL; | ||
| } |
Comment on lines
+298
to
+300
| func = dcblock_find_func(SOF_IPC_FRAME_S24_4LE); | ||
| assert_non_null(func); | ||
| func(&cd, &csrc, &csnk, TEST_FRAMES); |
Comment on lines
+190
to
+193
| for (i = 0; i < TEST_FRAMES * TEST_CHANNELS; i++) { | ||
| assert_true(dst[i] <= INT32_MAX); | ||
| assert_true(dst[i] >= INT32_MIN); | ||
| } |
Comment on lines
+90
to
+96
| remaining -= n; | ||
| dst += n; | ||
| if (dst >= y_end) | ||
| dst -= y_size; | ||
| src += n; | ||
| if (src >= x_end) | ||
| src -= x_size; |
Comment on lines
+141
to
+143
| ret = ops->set_configuration(mod, 0, MODULE_CFG_FRAGMENT_SINGLE, | ||
| coeff_size, (const uint8_t *)cdata, | ||
| coeff_size, NULL, 0); |
Comment on lines
+218
to
+221
| dev = comp_new((struct sof_ipc_comp *)ipc); | ||
| free(ipc); | ||
| if (!dev) | ||
| return -EINVAL; |
Comment on lines
+227
to
+229
| ret = dcblock_send_config(mod); | ||
| if (ret) | ||
| return ret; |
Comment on lines
+244
to
+246
| ret = module_prepare(mod, sources, 1, sinks, 1); | ||
| if (ret) | ||
| return ret; |
lyakh
reviewed
Jul 16, 2026
| endif() | ||
| if(CONFIG_COMP_DCBLOCK) | ||
| add_subdirectory(dcblock) | ||
| endif() |
Collaborator
There was a problem hiding this comment.
same question re: cmocka vs. ztest
kv2019i
approved these changes
Jul 16, 2026
kv2019i
left a comment
kv2019i
left a comment
Collaborator
There was a problem hiding this comment.
I'll let @tmleman comment on the cmocka/ztest transition as he has been converting tests away from cmocka (we do have some left). A test is great to have, no matter which the framework. One question inline, but otherwise code looks good.
| comp_dbg(dev, "entry"); | ||
|
|
||
| frames = MIN(frames, sink_frames); | ||
| frames = MIN(frames, dev->frames); |
Collaborator
There was a problem hiding this comment.
Btw why this min to limit to "dev->frames". This is not done in non-p20 implementation. Would be easier to review if the patch would only map from API to another and not do other changes at the same time.
This file contains hidden or 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
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.
Convert the DC blocking filter from the deprecated audio_stream /
process_audio_stream interface to the modern sof_source / sof_sink
processing API.
The module .process callback now acquires the source and sink circular
buffers once via source_get_data() / sink_get_buffer() and passes them
to the processing functions as struct cir_buf_source / struct cir_buf_sink
views, mirroring the volume component. The generic path uses the shared
cir_buf_samples_without_wrap_*() and cir_buf_wrap() helpers, while the
HiFi3/HiFi4 variants derive the circular buffer bounds from the same
views. Channel count is cached in comp_data during prepare().