- 04 Jan, 2019 2 commits
-
-
Hans de Goede authored
Export the DAC functionality to mix left + right together and then output the same (mixed) signal on both outputs. Various (x86) tablets with an ES8316 codec use a single speaker connected between the headhpone LOUT and ROUT pins, expecting the output to be in a mono differential mode. Presumably this is done to use the power of both the left and right outputs to allow the speaker to be louder. The ES8316 codec does not have a differential output mode, but we can emulate this by making both channels output the same through the mono mix switch, combined with setting the Playback Polarity control to "R Invert", which applias a 180 degrees phase inversion to the right channel. Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Hans de Goede authored
Adding jack-detect support may seem weird for a codec with only a single output, but it is necessary. The ES8316 appnote showing the intended usage uses a jack-receptacle which physically disconnects the speakers from the output when a jack is plugged in. But all 3 devices using the es8316 which I have (2 Cherry Trail devices and one Bay Trail CR device), use an analog mux to disconnect the speakers, driven by a GPIO. In order to enable/disable the speakers at the right time, we need jack-detect. The same goes for the microphone where we must correctly set the mux for the single ADC to either the internal or the headset microphone. All devices I have support the es8316's builtin jack-detect functionality. Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
- 03 Jan, 2019 38 commits
-
-
YueHaibing authored
Remove .owner field if calls are used which set it automatically Generated by: scripts/coccinelle/api/platform_no_drv_owner.cocci Signed-off-by: YueHaibing <yuehaibing@huawei.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Gustavo A. R. Silva authored
Currently, if platform_get_irq_byname() fails, the returned error turns into a huge value, once it is being store into a variable of type unsigned int, hence never actually reporting any error and causing unexpected behavior when using the values stored in aud_drv_data->s2mm_irq and aud_drv_data->mm2s_irq. Fix this by changing the type of variables s2mm_irq and mm2s_irq in structure xlnx_pcm_drv_data from unsigned int to int. Addresses-Coverity-ID: 1476096 ("Unsigned compared against 0") Fixes: 796175a94a7f ("ASoC: xlnx: add pcm formatter platform driver") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Katsuhiro Suzuki authored
This patch adds reset and precharge in shutdown of PCM device. ACODEC goes to silence if we change Fs to 44.1kHz from 48kHz. This workaround seems to work but I don't know this workaround is correct sequence or not for ACODEC. Signed-off-by: Katsuhiro Suzuki <katsuhiro@katsuster.net> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Katsuhiro Suzuki authored
This patch adds support for audio CODEC core of rk3328. Rockchip does not publish detail specification of this core but driver source code is opened on their GitHub repository. https://github.com/rockchip-linux/kernel So I ported this code to linux-next and added some trivial fixes. Signed-off-by: Katsuhiro Suzuki <katsuhiro@katsuster.net> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current simple-card is using asoc_simple_card_xxx() for function / data naming. Because of this long prefix, it is easy to be 80 character over. Let's reduce prefix from asoc_simple_card_xxx() to simple_xxx(). Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current simple-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. This patch cleanups the code by using asoc_simple_card_for_each_link() which judges normal link / DPCM link. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current simple-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. This patch adds/modifies counting and parsing function for "normal sound" and "DPCM sound", and call it from link loop. This is prepare for cleanup DAI link loop method. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current simple-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. To preparing cleanup code, this patch adds link_info which handles number of DAIs/Links/Codec Conf, and CPU/Codec turn. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
simple-card is now supporting normal sound and DPCM sound. For DPCM sound, original sound card (= simple-scu-card) had been supported 1 CPU : 1 Codec connection which uses hw_params_fixup() for convert-rate/channel. But, merged simple-card is completely forgeting about it. This patch re-support it. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
simple-card is now supporting normal sound and DPCM sound. For DPCM sound, original sound card (= simple-scu-card) had been supported 1 CPU : 1 Codec connection which uses hw_params_fixup() for convert-rate/channel. But, merged simple-card is completely forgeting about it. To re-support 1 CPU : 1 Codec DPCM for hw_params_fixup(), it need to judge whether it is DPCM by checking convert-rate/channel. For this purpose, this patch adds asoc_simple_card_get_conversion() as preparation Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current audio-graph-card is using asoc_graph_card_xxx() for function / data naming. Because of this long prefix, it is easy to be 80 character over. Let's reduce prefix from asoc_graph_card_xxx() to graph_xxx(). Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current audio-graph-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. This patch cleanups the code by using asoc_graph_card_for_each_link() which judges normal link / DPCM link. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current audio-graph-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. This patch adds/modifies counting and parsing function for "normal sound" and "DPCM sound", and call it from link loop. This is prepare for cleanup DAI link loop method. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
Current audio-graph-card is parsing DAI link for both "normal sound" and "DPCM sound". On this driver, it needs to count and parse DAIs/Links/Codec Conf from each links. Then, counting/parsing link loop are very similar, but using different implementation. Because of this background, the link loop code is very mysterious. Mystery code will be trouble in the future. To preparing cleanup code, this patch adds link_info which handles number of DAIs/Links/Codec Conf, and CPU/Codec turn. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
audio-graph-card is now supporting normal sound and DPCM sound. For DPCM sound, original sound card (= audio-graph-scu) had been supported 1 CPU : 1 Codec connection which uses hw_params_fixup() for convert-rate/channel. But, merged audio-graph-card is completely forgeting about it. This patch re-support it. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Maruthi Srinivas Bayyavarapu authored
Enable audio formatter driver build. Signed-off-by: Maruthi Srinivas Bayyavarapu <maruthi.srinivas.bayyavarapu@xilinx.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Maruthi Srinivas Bayyavarapu authored
The audio formatter PL IP supports DMA of two streams - mm2s and s2mm for playback and capture respectively. Apart from DMA, IP also does conversions like PCM to AES and viceversa. This patch adds DMA component driver for the IP. Signed-off-by: Maruthi Srinivas Bayyavarapu <maruthi.srinivas.bayyavarapu@xilinx.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Maruthi Srinivas Bayyavarapu authored
Added documentation for audio formatter IP core DT bindings. Signed-off-by: Maruthi Srinivas Bayyavarapu <maruthi.srinivas.bayyavarapu@xilinx.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
It is already merged into simple-card. simple-scu-card is no longer needed. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
It is already merged into simple-card. simple-scu-card is no longer needed. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
It is already merged into audio-graph-card. audio-graph-scu-card is no longer needed. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
It is already merged into audio-graph-card. audio-graph-scu-card is no longer needed. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
audio-graph-card is now supporting normal sound and DPCM sound. For DPCM sound, original sound card (= audio-graph-scu) had been supported 1 CPU : 1 Codec connection which uses hw_params_fixup() for convert-rate/channel. But, merged audio-graph-card is completely forgeting about it. To re-support 1 CPU : 1 Codec DPCM for hw_params_fixup(), it need to judge whether it is DPCM by checking convert-rate/channel. For this purpose, this patch adds asoc_graph_card_get_conversion() as preparation Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Kuninori Morimoto authored
We will get DAI ID from "reg" property if it has on DT, otherwise get it by counting port/endpoint. But in below case, we need to get DAI ID = 0 via port reg = <0>, but current implementation returns ID = 1, because it can't judge ID = 0 was from "non reg" or "reg = <0>". Thus, it will count port/endpoint number as "non reg" case. of_graph_parse_endpoint() implementation itself is not a problem, but because asoc_simple_card_get_dai_id() need to count port/endpoint number when "non reg" case, it need to know ID = 0 was from "non reg" or "reg = <0>". This patch fix this issue. port { reg = <0>; xxxx: endpoint@0 { }; => xxxx: endpoint@1 { }; }; Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Fix Sparse warnings with two machine drivers which weren't updated Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle skl-messages.c:419:5-32: WARNING: Comparison to bool skl-pcm.c:1426:6-33: WARNING: Comparison to bool Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Pierre-Louis Bossart authored
Detected with Coccinelle sound/soc/intel/skylake/skl-topology.c:3106:16-20: WARNING: casting value returned by memory allocation function to (char *) is useless. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Michał Mirosław authored
MCLK input is needed when accessing any register after enabling SYSCLK. This also fixes imbalance of clk_enable / clk_disable when transitioning between ON -> STANDBY -> ON bias levels. Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Michał Mirosław authored
Save 2x unsigned int of .rodata. Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Michał Mirosław authored
For platforms that use the audio-graph-card driver, the codec is not selected by SoC-platform driver. Make it available. Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Alexander Shiyan authored
This patch adds DT bindings documentation for Cirrus Logic CS4341 DAC. Signed-off-by: Alexander Shiyan <shc_work@mail.ru> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Alexander Shiyan authored
This patch adds Cirrus Logic CS4341. This is a very simple, playback only, stereo DAC. Signed-off-by: Alexander Shiyan <shc_work@mail.ru> Signed-off-by: Mark Brown <broonie@kernel.org>
-
Shuming Fan authored
The ADC mixer setting needs to restore to default value after calibration. Signed-off-by: Shuming Fan <shumingf@realtek.com> Signed-off-by: Mark Brown <broonie@kernel.org>
-