Repository navigation
Abi alignment for WoV - #1304
Conversation
There was a problem hiding this comment.
Looks like we don't use this enum in FW at all. Therefore I would remove it.
There was a problem hiding this comment.
thanks for reviewing @mrajwa, good point, let me remove this enum type from the header file from both FW and driver.
For selector and kpb, we have switched to use byte kcontrols, so no ipc structures needed to align b/w FW and driver, here remove them to align with driver side. The sof_ipc_process_type enum is not needed, remove it also. Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>
|
updated, please refer to thesofproject/linux#864 on driver side, they don't need to be merged at the same time as no ABI bump needed. |
No abi bump is needed, as driver already run with this header for weeks, here it is only to fix the gap b/w FW and driver. Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>
| } __attribute__((packed)); | ||
|
|
||
| /* create new component kpb - SOF_IPC_TPLG_KPB_NEW */ | ||
| struct sof_ipc_comp_kpb { |
There was a problem hiding this comment.
In general it's OK for me. I only think sof_ipc_process_type shall be removed as we don't use it at all.
After a talk with validation team I think we should keep components structures here in src/include/uapi/ipc/topology.h and remove them from components header files. Validation team is referenced as code-owner of this file and any change in common structures is automaticly visible by them so they can quickly provide a feedback and/or apply changes in the slim driver.
So in case of KPB for example we should move struct sof_kpb_config here. Same thing for detector, selector etc.
There was a problem hiding this comment.
@mrajwa as what we aligned, let's move sof_kpb_config to src/include/uapi/user/kpb.h, that should belong to another PR.
|
@keyonjie can you confirm this aligns FW to existing driver code ? |
yep, @lgirdwood please see it here: https://github.com/thesofproject/linux/blob/topic/sof-dev/include/sound/sof/topology.h#L38 |
|
@keyonjie And maybe WoV to Key Word Detection? |
it's somewhat messed up for the ABI version at the moment, Kernel already have WoV support with 3.5, FW don't have that yet but with newer 3.6, I think ABI was bump up too much. we have some kernel patch to make them aligned, will send it soon after those contents are actually aligned: |
|
@keyonjie |
good point, let me add that to kernel. |
|
@keyonjie this should be the last of the ABI patches that were created before the new process was announced. It also appears we had an ABI minor race between kernel and FW prior to this too. Do we align with the latest kernel version sent upstream with this PR or does the kernel also need a PR to align ? And what ABI minor number should we align at for KBP ? |
|
@lgirdwood @plbossart @keyonjie And we also need to check about the upstream patches for the ABI version and the real header changes. |
We need PR on driver side also, one open is that I only checked topology.h and abi.h, for other headers, we still need to go through all of them, maybe better to arrange a short meeting to let everybody align on it.
|
lgirdwood
left a comment
There was a problem hiding this comment.
Ok, this should align the structure, but we will still have a feature offset until MINOR 7
This is only aimed to align headers for WoV component structures, they are already updated on driver side for weeks, No abi bump is needed, just to fix gap.