-----Original Message----- From: Carlo Caione [mailto:carlo@endlessm.com] Sent: Thursday, October 19, 2017 11:55 PM To: Pierre-Louis Bossart Cc: Carlo Caione; Linux Upstreaming Team; Bard Liao; Oder Chiou; Mark Brown; alsa-devel@alsa-project.org; linux-kernel@vger.kernel.org; tiwai@suse.com; Albert Chen; Edgar Shen Subject: Re: [alsa-devel] [PATCH 1/2] ASoC: rt5651: Enable jack detection on JD1_1
On Thu, Oct 19, 2017 at 4:47 PM, Pierre-Louis Bossart pierre-louis.bossart@linux.intel.com wrote:
On 10/19/17 6:03 AM, Carlo Caione wrote:
From: Carlo Caione carlo@endlessm.com
Enable jack detection or the RT5651 codec on the JD1_1 pin.
Nice, but the codec supports a second jack detection on JD1 and has a
second
JD2 pin. I will bet that some devices will have a different routing and I wonder if we could just add support for all options.
I can write support for that but I have no hardware to actually test it, that's why I left those cases out.
The codec has no means to detect the type of the jack connected so we assume that the jack is always an headset jack.
that's odd, was this confirmed by Realtek?
The Realtek people are in CC :) Probably there is way but in the datasheet there is nothing about that (or did I miss it?)
Yes, rt5651 has the capability of jack type detection. Please see the following code for reference. +static int rt5651_jack_detect(struct snd_soc_codec *codec, int jack_insert) +{ + struct snd_soc_dapm_context *dapm = snd_soc_codec_get_dapm(codec); + int jack_type; + + if (jack_insert) { + snd_soc_dapm_force_enable_pin(dapm, "LDO"); + snd_soc_dapm_sync(dapm); + + snd_soc_update_bits(codec, RT5651_MICBIAS, + RT5651_MIC1_OVCD_MASK | RT5651_MIC1_OVTH_MASK | + RT5651_PWR_CLK12M_MASK | RT5651_PWR_MB_MASK, + RT5651_MIC1_OVCD_EN | RT5651_MIC1_OVTH_600UA | + RT5651_PWR_MB_PU | RT5651_PWR_CLK12M_PU); + msleep(100); + if (snd_soc_read(codec, RT5651_IRQ_CTRL2) & RT5651_MB1_OC_CLR) + jack_type = SND_JACK_HEADPHONE; + else + jack_type = SND_JACK_HEADSET; + snd_soc_update_bits(codec, RT5651_IRQ_CTRL2, + RT5651_MB1_OC_CLR, 0); + } else { /* jack out */ + jack_type = 0; + + snd_soc_update_bits(codec, RT5651_MICBIAS, + RT5651_MIC1_OVCD_MASK, RT5651_MIC1_OVCD_DIS); + } + + return jack_type; +}
Signed-off-by: Carlo Caione carlo@endlessm.com
include/sound/rt5651.h | 7 ++++ sound/soc/codecs/rt5651.c | 91 +++++++++++++++++++++++++++++++++++++++++++++-- sound/soc/codecs/rt5651.h | 3 ++ 3 files changed, 99 insertions(+), 2 deletions(-)
diff --git a/include/sound/rt5651.h b/include/sound/rt5651.h index d35de758dfb5..c563383149c4 100644 --- a/include/sound/rt5651.h +++ b/include/sound/rt5651.h @@ -11,11 +11,18 @@ #ifndef __LINUX_SND_RT5651_H #define __LINUX_SND_RT5651_H +enum rt5651_jd_src {
RT5651_JD_NULL,RT5651_JD1_1,+};
- struct rt5651_platform_data { /* IN2 can optionally be differential */ bool in2_diff; bool dmic_en;
enum rt5651_jd_src jd_src;I don't see code that sets this platform data, is there a quirk or of_property missing in this patchset?
Yes, it is supposed to be enabled by a quirk. In general (personal taste) I prefer to post the quirk enabling code after the base code has been ACKed. If you feel like it I can post also the quirk code together with the next respin of this patchset.
Thank you,
-- Carlo Caione | +39.340.80.30.096 | Endless
------Please consider the environment before printing this e-mail.