cadence: unification of cadence module naming #9287

dnikodem · 2024-07-08T10:44:51Z

This change improves and standardizes the naming convention of the cadence module during its creation.

This change improves and standardizes the naming convention of the cadence module during its creation. Signed-off-by: Damian Nikodem <damian.nikodem@intel.com>

dnikodem · 2024-07-08T12:05:46Z

Sof CI tests fails - WiP

dbaluta · 2024-07-08T13:01:32Z

FYI: There is a compilation error problem, existing just before this patch!

                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/audio_stream.h:23,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/buffer.h:11,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/component.h:19,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/module_adapter/module/generic.h:16,
                 from /home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:14:
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:80:31: error: '_cadence_uuid' undeclared here (not in a function); did you mean 'cadence_api'?
   80 | #define _UUID(uuid_name)    (&_##uuid_name)
      |                               ^
/home/nxf25322/work/sof_dir5/sof/zephyr/include/sof/trace/../../../../src/include/sof/trace/trace.h:410:27: note: in definition of macro 'DECLARE_TR_CTX'
  410 |                 .uuid_p = uuid,                                 \
      |                           ^~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:148:29: note: in expansion of macro '_UUID'
  148 | #define SOF_UUID(uuid_name) _UUID(uuid_name)
      |                             ^~~~~
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:23:28: note: in expansion of macro 'SOF_UUID'
   23 | DECLARE_TR_CTX(cadence_tr, SOF_UUID(cadence_uuid), LOG_LEVEL_INFO);
      |                            ^~~~~~~~
In file included from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/buffer.h:18:
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:921:43: error: 'cadence_uuid' undeclared here (not in a function); did you mean 'cadence_api'?
  921 | DECLARE_MODULE_ADAPTER(cadence_interface, cadence_uuid, cadence_tr);
      |                                           ^~~~~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:81:31: note: in definition of macro '_RT_UUID'
   81 | #define _RT_UUID(uuid_name) (&uuid_name)
      |                               ^~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/module_adapter/module/generic.h:52:16: note: in expansion of macro 'SOF_RT_UUID'
   52 |         .uid = SOF_RT_UUID(uuid), \
      |                ^~~~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:921:1: note: in expansion of macro 'DECLARE_MODULE_ADAPTER'
  921 | DECLARE_MODULE_ADAPTER(cadence_interface, cadence_uuid, cadence_tr);
      | ^~~~~~~~~~~~~~~~~~~~~~
[24/68] Building C object modules/hal_nxp/hal_nxp/CMakeFiles/..__modules__hal__nxp.dir/mcux/mcux-sdk/drivers/iuart/fsl_uart.c.obj

I think it was introduce by @andyross fdb039e / #9261

commit fdb039ef5f77e37e2566e343c8a831d1436e9d91
Author: Andy Ross <andyross@google.com>
Date:   Sat Jun 22 13:31:39 2024 -0700

    everywhere: Normalize UUID naming

dbaluta · 2024-07-08T13:04:33Z

@dnikodem Your PR not only unifies the cadence module naming! but also fixes a compilation problem!

dnikodem · 2024-07-08T14:08:55Z

FYI: There is a compilation error problem, existing just before this patch!

                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/audio_stream.h:23,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/buffer.h:11,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/component.h:19,
                 from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/module_adapter/module/generic.h:16,
                 from /home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:14:
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:80:31: error: '_cadence_uuid' undeclared here (not in a function); did you mean 'cadence_api'?
   80 | #define _UUID(uuid_name)    (&_##uuid_name)
      |                               ^
/home/nxf25322/work/sof_dir5/sof/zephyr/include/sof/trace/../../../../src/include/sof/trace/trace.h:410:27: note: in definition of macro 'DECLARE_TR_CTX'
  410 |                 .uuid_p = uuid,                                 \
      |                           ^~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:148:29: note: in expansion of macro '_UUID'
  148 | #define SOF_UUID(uuid_name) _UUID(uuid_name)
      |                             ^~~~~
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:23:28: note: in expansion of macro 'SOF_UUID'
   23 | DECLARE_TR_CTX(cadence_tr, SOF_UUID(cadence_uuid), LOG_LEVEL_INFO);
      |                            ^~~~~~~~
In file included from /home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/buffer.h:18:
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:921:43: error: 'cadence_uuid' undeclared here (not in a function); did you mean 'cadence_api'?
  921 | DECLARE_MODULE_ADAPTER(cadence_interface, cadence_uuid, cadence_tr);
      |                                           ^~~~~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/lib/uuid.h:81:31: note: in definition of macro '_RT_UUID'
   81 | #define _RT_UUID(uuid_name) (&uuid_name)
      |                               ^~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/include/sof/audio/module_adapter/module/generic.h:52:16: note: in expansion of macro 'SOF_RT_UUID'
   52 |         .uid = SOF_RT_UUID(uuid), \
      |                ^~~~~~~~~~~
/home/nxf25322/work/sof_dir5/sof/src/audio/module_adapter/module/cadence.c:921:1: note: in expansion of macro 'DECLARE_MODULE_ADAPTER'
  921 | DECLARE_MODULE_ADAPTER(cadence_interface, cadence_uuid, cadence_tr);
      | ^~~~~~~~~~~~~~~~~~~~~~
[24/68] Building C object modules/hal_nxp/hal_nxp/CMakeFiles/..__modules__hal__nxp.dir/mcux/mcux-sdk/drivers/iuart/fsl_uart.c.obj

I think it was introduce by @andyross

commit fdb039ef5f77e37e2566e343c8a831d1436e9d91
Author: Andy Ross <andyross@google.com>
Date:   Sat Jun 22 13:31:39 2024 -0700

    everywhere: Normalize UUID naming

correct, this is why I prepared this PR. The problem is that I see a lot of fails in CI (probably not related to this change) - need to verify it.

andyross

That's the correct way to adjust the naming. Is this not buildable in CI or otherwise from e.g. xtensa_build_{all,zephyr} or rebuild_testbench.sh? What's the correct way to validate this component? If the validation isn't in-tree, can we add an issue to make sure it gets a test?

(Honestly I sorta worry that there are others like this hiding. SOF has a lot of seemingly dead/zombied code and this was a pretty invasive PR that plays poorly in that kind of bathtub.)

dnikodem · 2024-07-09T07:53:28Z

Looks like the issues are not related to my PR:

sof-logger failed on all PRs
LNL SDW's failures are also seen on other PRs

I think that this is ready to merge.

kv2019i · 2024-07-09T08:05:34Z

@dnikodem wrote:

Looks like the issues are not related to my PR:

sof-logger failed on all PRs

LNL SDW's failures are also seen on other PRs

I think that this is ready to merge.

Ack, (1) is fixed in main via sof-test PRs and (2) is tracked as thesofproject/linux#5098

Proceeding with merge.

dbaluta · 2024-07-09T11:40:48Z

That's the correct way to adjust the naming. Is this not buildable in CI or otherwise from e.g. xtensa_build_{all,zephyr} or rebuild_testbench.sh? What's the correct way to validate this component? If the validation isn't in-tree, can we add an issue to make sure it gets a test?

In order to validate this component you will need some proprietary libraries from Cadence. Intel / NXP are using different libraries.

We validate this internally, using our CI - and it usually takes 1 or 2 days until we get signalled about issues.

I think Intel does the same.

andyross · 2024-07-09T15:50:38Z

In order to validate this component you will need some proprietary libraries from Cadence. Intel / NXP are using different libraries.

The normal pattern for that sort of thing is to provide a "stub" implementation selected via kconfig that emulates the proprietary bits as some kind of noop. Then you build that in a default configuration (the fuzzer seems to be evolving in that direction as the official "enable everything for testing" config) that runs in CI, so you're guaranteed you don't get surprised by this kind of API change.

cadence: unification of cadence module naming

d1ad10d

This change improves and standardizes the naming convention of the cadence module during its creation. Signed-off-by: Damian Nikodem <damian.nikodem@intel.com>

dnikodem requested a review from andyross July 8, 2024 10:45

dnikodem marked this pull request as ready for review July 8, 2024 12:01

dnikodem requested review from lgirdwood, plbossart, mmaka1, lbetlej, dbaluta and kv2019i as code owners July 8, 2024 12:01

dnikodem changed the title ~~cadence: unification of cadence module naming~~ [DNM] cadence: unification of cadence module naming Jul 8, 2024

wszypelt approved these changes Jul 8, 2024

View reviewed changes

dbaluta approved these changes Jul 8, 2024

View reviewed changes

lyakh mentioned this pull request Jul 8, 2024

LLEXT: fix failures and make DRC an LLEXT module by default on MTL #9116

Merged

marc-hb mentioned this pull request Jul 8, 2024

case-lib: remove check for log ldc file thesofproject/sof-test#1215

Merged

andyross approved these changes Jul 9, 2024

View reviewed changes

dnikodem changed the title ~~[DNM] cadence: unification of cadence module naming~~ cadence: unification of cadence module naming Jul 9, 2024

kv2019i merged commit d3ae43c into thesofproject:main Jul 9, 2024
44 of 47 checks passed

andyross mentioned this pull request Jul 9, 2024

[BUG] Cadence codec is dead code in-tree, needs a stub #9292

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

cadence: unification of cadence module naming #9287

cadence: unification of cadence module naming #9287

dnikodem commented Jul 8, 2024

dnikodem commented Jul 8, 2024

dbaluta commented Jul 8, 2024 •

edited by marc-hb

Loading

dbaluta commented Jul 8, 2024

dnikodem commented Jul 8, 2024

andyross left a comment

dnikodem commented Jul 9, 2024

kv2019i commented Jul 9, 2024

dbaluta commented Jul 9, 2024

andyross commented Jul 9, 2024

cadence: unification of cadence module naming #9287

cadence: unification of cadence module naming #9287

Conversation

dnikodem commented Jul 8, 2024

dnikodem commented Jul 8, 2024

dbaluta commented Jul 8, 2024 • edited by marc-hb Loading

dbaluta commented Jul 8, 2024

dnikodem commented Jul 8, 2024

andyross left a comment

Choose a reason for hiding this comment

dnikodem commented Jul 9, 2024

kv2019i commented Jul 9, 2024

dbaluta commented Jul 9, 2024

andyross commented Jul 9, 2024

dbaluta commented Jul 8, 2024 •

edited by marc-hb

Loading