On 6/21/2018 13:16, Ranjani Sridharan wrote:
On Wed, 2018-06-20 at 11:53 +0100, Liam Girdwood wrote:
On Wed, 2018-06-20 at 15:42 +0800, Pan, Xiuli wrote:
index 0000000..59d9af8 --- /dev/null +++ b/topology/sof/pipe-dai-loopback.m4
This file is not needed if you reuse the standard DAI macros and add a new param to them.
Will try to use a flag implement, maybe a new m4 file is not needed. We may need to add more args in the old m4 files.
Adding a pram has no benefit then adding a new file. Since we already have to much prams in tplg-build.sh, adding one more parms means we need double all command for test case. But adding one more file can simplify all this work into the different file it will lower down the switch case. Or we may need a double size tplg-build.sh
This path leads to things like pipe-dai-loopback-tristate-option3- config4- blah5.m4 which becomes unmanageable.
Using the flag as a param means it will be 0 for most users except when we need a LBM kcontrol. This means that cnl.m4, apl.m4 etc can set this flag to enable LBM kcontrol automatically, other non Intel platform.m4 wont define it.
Even better is creating an ssp.m4 that defines all the flags for SSP and then is included by all the intel platforms like apl.m4. The SSP tristate kcontrol flag could go here too.
This make sense to me. Passing a parameter directly from tplg-build.sh would make the scripts more complex. Using some pre-define config macro would be better.
Not to mention the tplg-build.sh needs a bit of clean up too. Its on my radar once 1.2 is done.
Agree, it is very complex now. It is not easy to use any more.
Thanks Xiuli
Liam _______________________________________________ Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware