Atmel: SAM: Convert XDMAC & SSC-I2S to devicetree

[galak]

No description provided.

[zephyrbot]

All checks passed.

checkpatch (informational only, not a failure)

-:183: WARNING:LONG_LINE: line over 80 characters
#183: FILE: drivers/i2s/i2s_sam_ssc.c:925:
+	dev_data->dev_dma = device_get_binding(DT_INST_DMAS_LABEL_BY_NAME(0, tx));

-:186: WARNING:LONG_LINE: line over 80 characters
#186: FILE: drivers/i2s/i2s_sam_ssc.c:927:
+		LOG_ERR("%s device not found", DT_INST_DMAS_LABEL_BY_NAME(0, tx));

- total: 0 errors, 2 warnings, 637 lines checked

NOTE: For some of the reported defects, checkpatch may be able to
      mechanically convert to the typical style using --fix or --fix-inplace.

Your patch has style problems, please review.

NOTE: Ignored message types: AVOID_EXTERNS BRACES CONFIG_EXPERIMENTAL CONST_STRUCT DATE_TIME FILE_PATH_CHANGES MINMAX NETWORKING_BLOCK_COMMENT_STYLE PRINTK_WITHOUT_KERN_LEVEL SPDX_LICENSE_TAG SPLIT_STRING VOLATILE

NOTE: If any of the errors are false positives, please report
      them to the maintainers.

Tip: The bot edits this comment instead of posting a new one, so you can check the comment's history to see earlier messages.

[mnkp]

It's going to complicate implementation, however it's a fairly common configuration where RK, RF pins are not connected.

It seems like at the moment there is no possibility to have RK, RF signals disconnected.

[mnkp]

I underestimated the new DT generation scheme. There is no need to change the implementation. It's enough that user defines their own pinctrl-0 property for the ssc node.

The one issue that remains is if we should provide 'default' pinctrl-0 property at the SoC level.

Atmel is a bit special, its pinmux is fairly rigid. In any given peripheral few pins, if any, can be assigned to an alternative location. But these choices exist and it feels improper to provide default configuration at the SoC level. Pin assignments are board specific and as soon as there exist a choice any default is wrong. Connecting incorrect signal to a pin can potentially damage the board.

At the same time if a peripheral doesn't allow to connect any of its pins to an alternate location forcing the user to define pinctrl-0 property at the board level also feels like an overkill.

[galak]

I tried to provide pinctrl-0 in the SoC dtsi files in the cases that either make sense or if there is only one option for a given peripheral. I'm find if because of how SSC is used if we want to move the pinctrl-0 setting to the board file instead.

[mnkp]

I've verified the PR with tests/drivers/i2s/* testcases on sam_e70_xplained board (after applying #25005). One of the tests was failing, however the error is unrelated / exists also on master.

There is one more issue: this PR uses the label "SSC_0" which is consistent with the node name. However, the tests rely on old "I2S_0" label. We could handle it also in a follow up PR.

Regarding location of pinctrl-0 property, let's discuss it at the dev-review meeting since more peripherals: e.g. uart, spi are affected.

[galak]

Should be updated with all the changes. Let me know if there's anything I missed.