ASoC: soc-acpi-intel-ptl-match: add ptl_cs42l43_agg_l3_cs35l56_l12_ghost_rt722#5754
ASoC: soc-acpi-intel-ptl-match: add ptl_cs42l43_agg_l3_cs35l56_l12_ghost_rt722#5754bardliao wants to merge 1 commit intothesofproject:topic/sof-devfrom
Conversation
bardliao
requested review from
dbaluta,
kv2019i,
lgirdwood,
plbossart and
ranj063
as code owners
There was a problem hiding this comment.
Pull request overview
Adds a new PTL SoundWire ACPI machine match intended to handle BIOSes that advertise a “ghost” RT722/ALC722 codec _ADR even when the device is not physically present, by providing an _ADR entry with zero endpoints so the machine driver skips it.
Changes:
- Added a new cs42l43 aggregated _ADR table that includes an extra “ghost” RT722/ALC722 device with no endpoints.
- Added new cs35l56 2-amp _ADR tables using
spk_1..spk_4endpoints. - Added a new PTL SoundWire link table + machine entry wiring these pieces together.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| .adr = 0x000330025d072201ull, /* Ghost rt722 */ | ||
| .num_endpoints = 0, | ||
| .endpoints = NULL, | ||
| .name_prefix = "ghost" |
| { | ||
| .adr = 0x00033001FA424301ull, | ||
| .num_endpoints = ARRAY_SIZE(cs42l43_amp_spkagg_endpoints), | ||
| .endpoints = cs42l43_amp_spkagg_endpoints, |
There was a problem hiding this comment.
I will wait for the test result since I am not quite sure if cs42l43 amp is aggregated with cs35l56s. @charleskeepax FYI
There was a problem hiding this comment.
@charleskeepax Is there a way that we can support the same codec/amp combination but with different endpoints? To be more specific, cs42l43 with and without amp aggregated. The endpoint will be skipped by the machine driver, but the topology was selected before we check is the endpoint present.
There was a problem hiding this comment.
Yeah its not aggregated here, I think the host thinks it is because of the smart amps bits on the rt722. If we use match tables you can't tell the difference between two systems with identical parts one of which aggregates.
I kinda wonder if this match table approach is the best way to go, it makes me nervous because I don't like mixing match tables and function topologies. Feels like it introduces chances of things matching when they should actually have used function topologies. Two other approaches that I can think of:
- We could disable the realtek codec using a SSDT something like:
DefinitionBlock ("", "SSDT", 2, "Test", "rt722", 0x00001000)
{
External (_SB.PC00.HDAS.IDA.SNDW.SWD0, DeviceObj)
Scope (\_SB.PC00.HDAS.IDA.SNDW.SWD0)
{
Method (_STA, 0, NotSerialized)
{
Return (Zero)
}
}
}
That should make the kernel ignore the device. Leaves the kernel code the cleanest and since this is an ACPI problem perhaps its the right place to fix it? Does mean people have to install and SSDT to make the laptop work though.... so dunno.
- Could we do some sort of quirk thing, like the stuff in sdca_device.c? Get the core to ignore the rt722 just for this system.
There was a problem hiding this comment.
@charleskeepax If the first approach is to make the change in the BIOS, I am afraid the ASIS's answer is NO. We reported this issue to ASUS and asked them to fix it in the BIOS, but there is no issue on Windows and they don't want to take any risk to change the BIOS. But, yes, I agree fixing it in the BIOS is the right and the best way to fix the issue. Or are you talking about adding a new SSDT to overwrite the existing SSDT?
Regarding the second approach , there is a similar quirk in drivers/soundwire/dmi-quirks.c. However, I guess this project is not the only product that has the issue. I think it is common is the ODM use a new audio solution but didn't remove the previous codec information from the BIOS.
Back to the PR, on second thought, we can actually use function topology with acpi match table as long as we set .sof_tplg_filename = "sof-ptl-dummy.tplg". Let's give it a try.
…ost_rt722 Some BIOS include a ghost ALC722 codec ADR while the hardware is not physically present. That's why we need to create an acpi mach table with empty endpoint. The ghost ALC722 codec with empty endpoint will be skipped by the machine driver. Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Closes:thesofproject#5721
Some BIOS include a ghost ALC722 codec ADR while the hardware is not physically present. That's why we need to create an acpi mach table with empty endpoint. The ghost ALC722 codec with empty endpoint will be skipped by the machine driver.