Hi,
On 24-12-18 10:53, Stephan Gerhold wrote:
On Mon, Dec 24, 2018 at 10:01:53AM +0100, Hans de Goede wrote:
H,
On 22-12-18 15:59, Stephan Gerhold wrote:
On Sat, Dec 22, 2018 at 03:47:12PM +0100, Stephan Gerhold wrote:
Add quirks to select the correct input map, jack-detect options and channel map to make sound work on the ASUS MeMO Pad 7 (ME176C).
Everything seems to work fine with the quirks below, although I have to say that the internal microphone quality is really bad. There are always high-pitched sounds in the background. Not sure if the hardware is just really bad or if there is any way to improve it. (I never use it so I have not investigated much further..)
One possible way of improving it which is worthwhile testing is adding the BYT_RT5640_DIFF_MIC quirk.
Note almost all quirk table entries not using a DMIC mapping have this set (it is also set in INPUT_DEFAULTS) because it does not really hurt to enable it, while likely only a few models really have a diff mic.
I thought that the DIFF_MIC quirk might help, so I tested it shortly before sending this patch: I was not able to hear any difference in the recorded sample, which is why I didn't add it to the quirks.
Would you prefer to have it enabled even if it (seemingly) makes no difference?
No there is no need for that, I just wanted to make sure that it was not necessary.
Regards,
Hans
Note: Although sound works out of the box, jack detection currently requires overriding the ACPI DSDT table. This is necessary because the rt5640 ACPI device (10EC5640) has the wrong GPIO listed as interrupt (one of the Bluetooth GPIOs). The correct GPIO is GPO2 0x0004 (listed as the first GPIO in the Intel(R) Audio Machine Driver - AMCR0F28 device).
At some point it might be possible to add a workaround for this using that AMCR0F28 device which has the correct GPIO. However, at the moment I still need the DSDT override for Bluetooth to work (with no simple workaround), so it's probably easiest if we just document it here for now. Eventually I will investigate this later..
+1 for just documenting, as long as this device needs a DSDT override anyways I don't think it is a good idea to add code to the bytcr_rt5640.c machine driver just to avoid that. If the machine driver is the only one needing the DSDT override and some code can avoid it then the balance changes, but for now I believe that just documenting it is best.
Agreed. :)
Regards,
Hans
Signed-off-by: Stephan Gerhold stephan@gerhold.net
sound/soc/intel/boards/bytcr_rt5640.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
diff --git a/sound/soc/intel/boards/bytcr_rt5640.c b/sound/soc/intel/boards/bytcr_rt5640.c index a22366ce33c4..ca8b4d5ff70f 100644 --- a/sound/soc/intel/boards/bytcr_rt5640.c +++ b/sound/soc/intel/boards/bytcr_rt5640.c @@ -428,6 +428,18 @@ static const struct dmi_system_id byt_rt5640_quirk_table[] = { BYT_RT5640_SSP0_AIF1 | BYT_RT5640_MCLK_EN), },
- {
.matches = {DMI_EXACT_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "ME176C"),},.driver_data = (void *)(BYT_RT5640_IN1_MAP |BYT_RT5640_JD_SRC_JD2_IN4N |BYT_RT5640_OVCD_TH_2000UA |BYT_RT5640_OVCD_SF_0P75 |BYT_RT5640_SSP0_AIF1 |BYT_RT5640_MCLK_EN),- }, { .matches = { DMI_EXACT_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-- 2.20.1