xtensa-build-all: Fix missed $j -> $platform renaming

[crazoes]

In renaming of $j -> $platform one instance of variable was missed which
has led to wrong configs being used.

Signed-off-by: Shreeya Patel shreeya.patel23498@gmail.com

[crazoes]

Fixes #2711 (comment) spotted by @marc-hb. Also, sorry about missing this one instance of renaming 😓

[marc-hb]

Considering some of the CI failures in https://sof-ci.01.org/sofpr/PR3167/build6570/devicetest/ are on BYT, I think we need to pay them more attention than usual.

@crazoes can you diff the .config files before vs after this fix and share the diff in a comment here if it's not too big? Thx

[marc-hb]

For small but potentially impactful fixes like this the commit message is important. Can you please add Fixes: 941dfcccb03 ("use switch case") in the body and change the subject to something more meaningful like: xtensa-build-all: fix gcc _defconfig for byt|cht|cnl|sue. Thx.

sorry about missing this

shellcheck is your friend.

[paulstelian97]

GH came back online :) Also, this change looks good for me.

[crazoes]

CC: @dbaluta

[crazoes]

I've pushed the amended commit however, it does not reflect in github. I pushed it when github was down, maybe they have some processing that failed and hence, not showing up here. 😕

[crazoes]

Will amend again so hash changes

[lgirdwood]

Alsabat CI clipping. @aiChaoSONG Btw - the sof-logger output on BSW/BYT tests does not align to the topology. i.e. I see component IDs 4.20 and 1.6 that are not in topology diagram.

[crazoes]

Configs prior and post fix and diff between the same - https://gist.github.com/crazoes/f8fd5557ba1d3e418f887a0704a434e5

cc @marc-hb

[aiChaoSONG]

Alsabat CI clipping. @aiChaoSONG Btw - the sof-logger output on BSW/BYT tests does not align to the topology. i.e. I see component IDs 4.20 and 1.6 that are not in topology diagram.

@lgirdwood There are graph section and widget section in the topology binary, right? The graph drawing is from the graph section. As I observed, not all widgets are linked in the graph. Take sof-byt-rt5640.tplg for example, besides the widgets you can see from the graph, there are extra widgets in the widget section: codec_out0, PIPELINE.1.SSP2.OUT, codec_out1, PIPELINE.2.SSP2.IN, codec_in0, PIPELINE.3.SSP2.OUT, codec_in1, ssp2 Rx, ssp2 Tx, ssp0 Tx, ssp0 Rx, modem_out, modem_in

Maybe some legacy m4 need a fix?

[lgirdwood]

@aiChaoSONG the sof-log showed these as buffers ?

[aiChaoSONG]

@lgirdwood No, they are dumped from binary topology(*.tplg), but seems I couldn't see them from the sof-logger