[alsa-devel] [PATCH 10/10] ASoC: Intel: Add ACPI match table entry for ES8316 codec on BYTCR platform

[Pierre-Louis Bossart]

>>>       },
>>> +    {
>>> +        .id = "ESSX8316",
>>> +        .drv_name = "bytcht_es8316",
>>> +        .fw_filename = "intel/fw_sst_0f28.bin",
>>> +        .board = "bytcht_es8316",
>>
>> Coming from the holiday break I saw this .board field and couldn't 
>> recall why we needed it. None of the more recent boards set this 
>> field. Is this legacy stuff that isn't used by anyone?
>
> I would expect you to know that better then I do :)  I just copied 
> this from the
> other entries.

I don't, it's at least 5 years old....

The initial tables in atom/sst/sst_acpi.c had this (April 2015).

+static struct sst_machines sst_acpi_bytcr[] = {
+    {"10EC5640", "T100", "bytt100_rt5640", NULL, "intel/fw_sst_0f28.bin",
+                        &byt_rvp_platform_data },
+    {},
+};

then it became

+static struct sst_acpi_mach sst_acpi_bytcr[] = {
+    {"10EC5640", "bytt100_rt5640", "intel/fw_sst_0f28.bin", "T100", NULL,

then

  static struct sst_acpi_mach sst_acpi_bytcr[] = {
-    {"10EC5640", "bytt100_rt5640", "intel/fw_sst_0f28.bin", "T100", NULL,
+    {"10EC5640", "bytcr_rt5640", "intel/fw_sst_0f28.bin", 
"bytcr_rt5640", NULL,
                          &byt_rvp_platform_data },

and then

+    {
+        .id = "10EC5640",
+        .drv_name = "bytcr_rt5640",
+        .fw_filename = "intel/fw_sst_0f28.bin",
+        .board = "bytcr_rt5640",
+        .machine_quirk = byt_quirk,
+        .pdata = &byt_rvp_platform_data,
+    },

we only use the drv_name to select the machine driver.

Keyon, Liam, do you have any memories of why this board field was needed 
in the first place?

>
> If you want to submit a patch removing the .board entries I'm all for it.

I was planning to remove the new_mach_data already so this is the second 
strike on this soc_acpi_machine structure...

-Pierre