* audio mcbsp1 broken for beagleboard-xm -OMAP
@ 2015-02-20 6:45 noman pouigt
2015-02-20 16:24 ` Peter Ujfalusi
2015-02-20 16:27 ` Peter Ujfalusi
0 siblings, 2 replies; 7+ messages in thread
From: noman pouigt @ 2015-02-20 6:45 UTC (permalink / raw)
To: peter.ujfalusi@ti.com, jarkko.nikula@bitmer.com,
broonie@kernel.org, jslaby@suse.cz
Cc: alsa-devel@alsa-project.org, zonque
Hello,
I am trying to integrate max98090 codec with $subject board. I am
running the mcBSP in slave mode as i have not changed the hw_parms for
omap-twl4030.c file. With below changes i see wait_for_avail is
getting stuck i.e. mcbsp is not getting any interrupt and nothing is
playing.
Kernel version:
VERSION = 3
PATCHLEVEL = 19
SUBLEVEL = 0
EXTRAVERSION =
NAME = Diseased Newt
Setup information:
Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
Connected only 4 pins.
Changed the omap-twl4030.c machine file as below to disable the twl4030:
static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
{
+ .name = "max98090",
+ .stream_name = "max98090",
.cpu_dai_name = "omap-mcbsp.1",
+ .codec_dai_name = "HiFi",
.platform_name = "omap-mcbsp.1",
+ .codec_name = "max98090.1-0010",
.ops = &omap_twl4030_ops,
},
DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
&i2c2 {
clock-frequency = <400000>;
+
+ max98090: max98090@10 {
+ compatible = "maxim,max98090";
+ reg = <0x10>;
+ };
};
omap/omap-twl4030.c
added in hw_params below function to set internal oscillator for 98090
for generating PLL for master clock.
snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
I am able to see that the cpu_dai is mapped with codec_dai and
configuration is ok. kernel logs at https://ideone.com/oezNSj
I think i don't need to configure the mcbsp1 pins right as it uses default.
Please guide me.I have tried IRC channel also but not getting the right folks.
https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
Thanks,
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-20 6:45 audio mcbsp1 broken for beagleboard-xm -OMAP noman pouigt
@ 2015-02-20 16:24 ` Peter Ujfalusi
2015-02-24 0:03 ` noman pouigt
2015-02-20 16:27 ` Peter Ujfalusi
1 sibling, 1 reply; 7+ messages in thread
From: Peter Ujfalusi @ 2015-02-20 16:24 UTC (permalink / raw)
To: noman pouigt, jarkko.nikula@bitmer.com, broonie@kernel.org,
jslaby@suse.cz
Cc: alsa-devel@alsa-project.org, zonque
Hi,
On 02/20/2015 08:45 AM, noman pouigt wrote:
> Hello,
>
> I am trying to integrate max98090 codec with $subject board. I am
> running the mcBSP in slave mode as i have not changed the hw_parms for
> omap-twl4030.c file. With below changes i see wait_for_avail is
> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
> playing.
I have tested the XM with twl4030 on linux-next and McBSP is working fine as
slave and master (I have a patch to add the McBSP master support for testing)
also.
>
> Kernel version:
> VERSION = 3
> PATCHLEVEL = 19
> SUBLEVEL = 0
> EXTRAVERSION =
> NAME = Diseased Newt
>
> Setup information:
> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
> Connected only 4 pins.
>
> Changed the omap-twl4030.c machine file as below to disable the twl4030:
> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
> {
> + .name = "max98090",
> + .stream_name = "max98090",
> .cpu_dai_name = "omap-mcbsp.1",
Hrm, mcbsp.1 is not in use by default on the board, 2 and 3 are the one which
is used by the omap-twl4030 driver.
> + .codec_dai_name = "HiFi",
> .platform_name = "omap-mcbsp.1",
> + .codec_name = "max98090.1-0010",
> .ops = &omap_twl4030_ops,
> },
>
> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>
> &i2c2 {
> clock-frequency = <400000>;
> +
> + max98090: max98090@10 {
> + compatible = "maxim,max98090";
> + reg = <0x10>;
> + };
> };
>
> omap/omap-twl4030.c
> added in hw_params below function to set internal oscillator for 98090
> for generating PLL for master clock.
> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>
> I am able to see that the cpu_dai is mapped with codec_dai and
> configuration is ok. kernel logs at https://ideone.com/oezNSj
>
> I think i don't need to configure the mcbsp1 pins right as it uses default.
I think you should check the pin mux for the McBSP1 pins. They might not set
to correct mode.
>
> Please guide me.I have tried IRC channel also but not getting the right folks.
>
> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>
> Thanks,
>
--
Péter
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-20 6:45 audio mcbsp1 broken for beagleboard-xm -OMAP noman pouigt
2015-02-20 16:24 ` Peter Ujfalusi
@ 2015-02-20 16:27 ` Peter Ujfalusi
2015-02-21 2:36 ` noman pouigt
1 sibling, 1 reply; 7+ messages in thread
From: Peter Ujfalusi @ 2015-02-20 16:27 UTC (permalink / raw)
To: noman pouigt, jarkko.nikula@bitmer.com, broonie@kernel.org,
jslaby@suse.cz
Cc: alsa-devel@alsa-project.org, zonque
On 02/20/2015 08:45 AM, noman pouigt wrote:
> Hello,
>
> I am trying to integrate max98090 codec with $subject board. I am
> running the mcBSP in slave mode as i have not changed the hw_parms for
> omap-twl4030.c file. With below changes i see wait_for_avail is
> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
> playing.
>
> Kernel version:
> VERSION = 3
> PATCHLEVEL = 19
> SUBLEVEL = 0
> EXTRAVERSION =
> NAME = Diseased Newt
>
> Setup information:
> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
> Connected only 4 pins.
>
> Changed the omap-twl4030.c machine file as below to disable the twl4030:
> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
> {
> + .name = "max98090",
> + .stream_name = "max98090",
> .cpu_dai_name = "omap-mcbsp.1",
> + .codec_dai_name = "HiFi",
> .platform_name = "omap-mcbsp.1",
> + .codec_name = "max98090.1-0010",
> .ops = &omap_twl4030_ops,
> },
>
> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>
> &i2c2 {
> clock-frequency = <400000>;
> +
> + max98090: max98090@10 {
> + compatible = "maxim,max98090";
> + reg = <0x10>;
> + };
> };
You also need to change the dts file's sound node, enable mcbsp1, set the
pinctrl entries, etc if you boot with DT.
>
> omap/omap-twl4030.c
> added in hw_params below function to set internal oscillator for 98090
> for generating PLL for master clock.
> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>
> I am able to see that the cpu_dai is mapped with codec_dai and
> configuration is ok. kernel logs at https://ideone.com/oezNSj
>
> I think i don't need to configure the mcbsp1 pins right as it uses default.
>
> Please guide me.I have tried IRC channel also but not getting the right folks.
>
> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>
> Thanks,
>
--
Péter
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-20 16:27 ` Peter Ujfalusi
@ 2015-02-21 2:36 ` noman pouigt
2015-02-25 9:40 ` Peter Ujfalusi
0 siblings, 1 reply; 7+ messages in thread
From: noman pouigt @ 2015-02-21 2:36 UTC (permalink / raw)
To: Peter Ujfalusi
Cc: alsa-devel@alsa-project.org, broonie@kernel.org, jslaby@suse.cz,
jarkko.nikula@bitmer.com, Daniel Mack
On Fri, Feb 20, 2015 at 8:27 AM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> On 02/20/2015 08:45 AM, noman pouigt wrote:
>> Hello,
>>
>> I am trying to integrate max98090 codec with $subject board. I am
>> running the mcBSP in slave mode as i have not changed the hw_parms for
>> omap-twl4030.c file. With below changes i see wait_for_avail is
>> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
>> playing.
>>
>> Kernel version:
>> VERSION = 3
>> PATCHLEVEL = 19
>> SUBLEVEL = 0
>> EXTRAVERSION =
>> NAME = Diseased Newt
>>
>> Setup information:
>> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
>> Connected only 4 pins.
>>
>> Changed the omap-twl4030.c machine file as below to disable the twl4030:
>> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
>> {
>> + .name = "max98090",
>> + .stream_name = "max98090",
>> .cpu_dai_name = "omap-mcbsp.1",
>> + .codec_dai_name = "HiFi",
>> .platform_name = "omap-mcbsp.1",
>> + .codec_name = "max98090.1-0010",
>> .ops = &omap_twl4030_ops,
>> },
>>
>> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>>
>> &i2c2 {
>> clock-frequency = <400000>;
>> +
>> + max98090: max98090@10 {
>> + compatible = "maxim,max98090";
>> + reg = <0x10>;
>> + };
>> };
>
> You also need to change the dts file's sound node, enable mcbsp1, set the
> pinctrl entries, etc if you boot with DT.
Yes and it is done as below:
&omap3_pmx_core {
mcbsp1_pins: pinmux_mcbsp1_pins {
pinctrl-single,pins = <
OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0)
OMAP3_CORE1_IOPAD(0x2194, PIN_INPUT | MUX_MODE0)
OMAP3_CORE1_IOPAD(0x2190, PIN_INPUT | MUX_MODE0)
OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0)
>;
};
}
sound {
- ti,mcbsp = <&mcbsp2>;
- ti,codec = <&twl_audio>;
+ ti,mcbsp = <&mcbsp1>;
+ ti,codec = <&max98090>;
}
&mcbsp1 {
pinctrl-names = "default";
pinctrl-0 = <&mcbsp1_pins>;
status = "okay";
};
&i2c2 {
clock-frequency = <400000>;
max98090: max98090@10 {
compatible = "maxim,max98090";
reg = <0x10>;
};
};
all other changes remain the same. Have i done muxing wrong? I am
still stuck with mcbsp not getting interrupt.
please find the dts file here http://ideone.com/TPFkBv
I connected digital output of 98090 to mcbsp1_dr and digital receive
to mcbsp1_dx. I can see the bit clock and lrclk is coming fine from
98090. As already mentioned i have connected lrclk, bit clock, DR
and DX only.
Kernel logs here: http://ideone.com/hDHq2l
>
>>
>> omap/omap-twl4030.c
>> added in hw_params below function to set internal oscillator for 98090
>> for generating PLL for master clock.
>> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>>
>> I am able to see that the cpu_dai is mapped with codec_dai and
>> configuration is ok. kernel logs at https://ideone.com/oezNSj
>>
>> I think i don't need to configure the mcbsp1 pins right as it uses default.
>>
>> Please guide me.I have tried IRC channel also but not getting the right folks.
>>
>> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>>
>> Thanks,
>>
>
>
> --
> Péter
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-20 16:24 ` Peter Ujfalusi
@ 2015-02-24 0:03 ` noman pouigt
0 siblings, 0 replies; 7+ messages in thread
From: noman pouigt @ 2015-02-24 0:03 UTC (permalink / raw)
To: Peter Ujfalusi
Cc: alsa-devel@alsa-project.org, broonie@kernel.org, jslaby@suse.cz,
jarkko.nikula@bitmer.com, Daniel Mack
On Fri, Feb 20, 2015 at 8:24 AM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> Hi,
>
> On 02/20/2015 08:45 AM, noman pouigt wrote:
>> Hello,
>>
>> I am trying to integrate max98090 codec with $subject board. I am
>> running the mcBSP in slave mode as i have not changed the hw_parms for
>> omap-twl4030.c file. With below changes i see wait_for_avail is
>> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
>> playing.
>
> I have tested the XM with twl4030 on linux-next and McBSP is working fine as
> slave and master (I have a patch to add the McBSP master support for testing)
> also.
>
>>
>> Kernel version:
>> VERSION = 3
>> PATCHLEVEL = 19
>> SUBLEVEL = 0
>> EXTRAVERSION =
>> NAME = Diseased Newt
>>
>> Setup information:
>> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
>> Connected only 4 pins.
>>
>> Changed the omap-twl4030.c machine file as below to disable the twl4030:
>> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
>> {
>> + .name = "max98090",
>> + .stream_name = "max98090",
>> .cpu_dai_name = "omap-mcbsp.1",
>
> Hrm, mcbsp.1 is not in use by default on the board, 2 and 3 are the one which
> is used by the omap-twl4030 driver.
I have used the mcbsp3 also as below where i am running codec in master and
mcbsp3 in slave mode. Configuration for that is below:
compatible = "ti,omap-twl4030";
ti,model = "omap3beagle";
- ti,mcbsp = <&mcbsp2>;
- ti,codec = <&twl_audio>;
+ ti,mcbsp = <&mcbsp3>;
+ ti,codec = <&max98090>;
&omap3_pmx_core {
+ mcbsp3_pins: pinmux_mcbsp3_pins {
+ pinctrl-single,pins = <
+ OMAP3_CORE1_IOPAD(0x216c, PIN_OUTPUT |
MUX_MODE1) /* mcbsp3_dx.uart2_cts */
+ OMAP3_CORE1_IOPAD(0x216e, PIN_INPUT |
MUX_MODE1) /* mcbsp3_dr.uart2_rts */
+ OMAP3_CORE1_IOPAD(0x2170, PIN_INPUT |
MUX_MODE1) /* mcbsp3_clk.uart2_tx */
+ OMAP3_CORE1_IOPAD(0x2172, PIN_INPUT |
MUX_MODE0) /* mcbsp3_fsx.uart2_rx */
+ >;
+ };
}
&mcbsp3 {
pinctrl-names = "default";
pinctrl-0 = <&mcbsp3_pins>;
status = "okay";
};
However still the interrupts doesn't get generated. Can you please
guide in right
direction? I have checked the mcbsp1_fsx and mcbsp1_clkx and
both the clocks generated are right i.e. one is 8k and other is
multiple of it.
On the data line i can see only zeroes. Probably dsp is filling it with
zeroes as generally done by Qualcomm dsp.
>
>> + .codec_dai_name = "HiFi",
>> .platform_name = "omap-mcbsp.1",
>> + .codec_name = "max98090.1-0010",
>> .ops = &omap_twl4030_ops,
>> },
>>
>> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>>
>> &i2c2 {
>> clock-frequency = <400000>;
>> +
>> + max98090: max98090@10 {
>> + compatible = "maxim,max98090";
>> + reg = <0x10>;
>> + };
>> };
>>
>> omap/omap-twl4030.c
>> added in hw_params below function to set internal oscillator for 98090
>> for generating PLL for master clock.
>> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>>
>> I am able to see that the cpu_dai is mapped with codec_dai and
>> configuration is ok. kernel logs at https://ideone.com/oezNSj
>>
>> I think i don't need to configure the mcbsp1 pins right as it uses default.
>
> I think you should check the pin mux for the McBSP1 pins. They might not set
> to correct mode.
>
>>
>> Please guide me.I have tried IRC channel also but not getting the right folks.
>>
>> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>>
>> Thanks,
>>
>
>
> --
> Péter
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-21 2:36 ` noman pouigt
@ 2015-02-25 9:40 ` Peter Ujfalusi
2015-02-25 20:09 ` noman pouigt
0 siblings, 1 reply; 7+ messages in thread
From: Peter Ujfalusi @ 2015-02-25 9:40 UTC (permalink / raw)
To: noman pouigt
Cc: alsa-devel@alsa-project.org, broonie@kernel.org, jslaby@suse.cz,
jarkko.nikula@bitmer.com, Daniel Mack
On 02/21/2015 04:36 AM, noman pouigt wrote:
> On Fri, Feb 20, 2015 at 8:27 AM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
>> On 02/20/2015 08:45 AM, noman pouigt wrote:
>>> Hello,
>>>
>>> I am trying to integrate max98090 codec with $subject board. I am
>>> running the mcBSP in slave mode as i have not changed the hw_parms for
>>> omap-twl4030.c file. With below changes i see wait_for_avail is
>>> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
>>> playing.
>>>
>>> Kernel version:
>>> VERSION = 3
>>> PATCHLEVEL = 19
>>> SUBLEVEL = 0
>>> EXTRAVERSION =
>>> NAME = Diseased Newt
>>>
>>> Setup information:
>>> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
>>> Connected only 4 pins.
>>>
>>> Changed the omap-twl4030.c machine file as below to disable the twl4030:
>>> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
>>> {
>>> + .name = "max98090",
>>> + .stream_name = "max98090",
>>> .cpu_dai_name = "omap-mcbsp.1",
>>> + .codec_dai_name = "HiFi",
>>> .platform_name = "omap-mcbsp.1",
>>> + .codec_name = "max98090.1-0010",
>>> .ops = &omap_twl4030_ops,
>>> },
>>>
>>> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>>>
>>> &i2c2 {
>>> clock-frequency = <400000>;
>>> +
>>> + max98090: max98090@10 {
>>> + compatible = "maxim,max98090";
>>> + reg = <0x10>;
>>> + };
>>> };
>>
>> You also need to change the dts file's sound node, enable mcbsp1, set the
>> pinctrl entries, etc if you boot with DT.
> Yes and it is done as below:
>
> &omap3_pmx_core {
> mcbsp1_pins: pinmux_mcbsp1_pins {
> pinctrl-single,pins = <
> OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0)
> OMAP3_CORE1_IOPAD(0x2194, PIN_INPUT | MUX_MODE0)
> OMAP3_CORE1_IOPAD(0x2190, PIN_INPUT | MUX_MODE0)
> OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0)
Should be:
OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0) /* CLKX */
OMAP3_CORE1_IOPAD(0x2196, PIN_INPUT | MUX_MODE0) /* FSX */
OMAP3_CORE1_IOPAD(0x2192, PIN_INPUT | MUX_MODE0) /* DR */
OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0) /* DX */
And make sure to connect lrclk/bclk to FSX/CLKX (not to FSR/CLKR) of McBSP1.
I have tested on on my xM that McBSP1 in master mode works, DMA is running (I
do not have anything connected to pins).
> >;
> };
> }
>
> sound {
> - ti,mcbsp = <&mcbsp2>;
> - ti,codec = <&twl_audio>;
> + ti,mcbsp = <&mcbsp1>;
> + ti,codec = <&max98090>;
> }
>
> &mcbsp1 {
> pinctrl-names = "default";
> pinctrl-0 = <&mcbsp1_pins>;
>
> status = "okay";
> };
>
> &i2c2 {
> clock-frequency = <400000>;
>
> max98090: max98090@10 {
> compatible = "maxim,max98090";
> reg = <0x10>;
> };
> };
>
> all other changes remain the same. Have i done muxing wrong? I am
> still stuck with mcbsp not getting interrupt.
> please find the dts file here http://ideone.com/TPFkBv
>
> I connected digital output of 98090 to mcbsp1_dr and digital receive
> to mcbsp1_dx. I can see the bit clock and lrclk is coming fine from
> 98090. As already mentioned i have connected lrclk, bit clock, DR
> and DX only.
>
> Kernel logs here: http://ideone.com/hDHq2l
>
>>
>>>
>>> omap/omap-twl4030.c
>>> added in hw_params below function to set internal oscillator for 98090
>>> for generating PLL for master clock.
>>> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>>>
>>> I am able to see that the cpu_dai is mapped with codec_dai and
>>> configuration is ok. kernel logs at https://ideone.com/oezNSj
>>>
>>> I think i don't need to configure the mcbsp1 pins right as it uses default.
>>>
>>> Please guide me.I have tried IRC channel also but not getting the right folks.
>>>
>>> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>>>
>>> Thanks,
>>>
>>
>>
>> --
>> Péter
--
Péter
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: audio mcbsp1 broken for beagleboard-xm -OMAP
2015-02-25 9:40 ` Peter Ujfalusi
@ 2015-02-25 20:09 ` noman pouigt
0 siblings, 0 replies; 7+ messages in thread
From: noman pouigt @ 2015-02-25 20:09 UTC (permalink / raw)
To: Peter Ujfalusi
Cc: alsa-devel@alsa-project.org, broonie@kernel.org, jslaby@suse.cz,
jarkko.nikula@bitmer.com, Daniel Mack
On Wed, Feb 25, 2015 at 1:40 AM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> On 02/21/2015 04:36 AM, noman pouigt wrote:
>> On Fri, Feb 20, 2015 at 8:27 AM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
>>> On 02/20/2015 08:45 AM, noman pouigt wrote:
>>>> Hello,
>>>>
>>>> I am trying to integrate max98090 codec with $subject board. I am
>>>> running the mcBSP in slave mode as i have not changed the hw_parms for
>>>> omap-twl4030.c file. With below changes i see wait_for_avail is
>>>> getting stuck i.e. mcbsp is not getting any interrupt and nothing is
>>>> playing.
>>>>
>>>> Kernel version:
>>>> VERSION = 3
>>>> PATCHLEVEL = 19
>>>> SUBLEVEL = 0
>>>> EXTRAVERSION =
>>>> NAME = Diseased Newt
>>>>
>>>> Setup information:
>>>> Using beagle board -xm and connected max98090 codec to mcbsp1 pins.
>>>> Connected only 4 pins.
>>>>
>>>> Changed the omap-twl4030.c machine file as below to disable the twl4030:
>>>> static struct snd_soc_dai_link omap_twl4030_dai_links[] = {
>>>> {
>>>> + .name = "max98090",
>>>> + .stream_name = "max98090",
>>>> .cpu_dai_name = "omap-mcbsp.1",
>>>> + .codec_dai_name = "HiFi",
>>>> .platform_name = "omap-mcbsp.1",
>>>> + .codec_name = "max98090.1-0010",
>>>> .ops = &omap_twl4030_ops,
>>>> },
>>>>
>>>> DTSI file change:arch/arm/boot/dts/omap3-beagle-xm.dts
>>>>
>>>> &i2c2 {
>>>> clock-frequency = <400000>;
>>>> +
>>>> + max98090: max98090@10 {
>>>> + compatible = "maxim,max98090";
>>>> + reg = <0x10>;
>>>> + };
>>>> };
>>>
>>> You also need to change the dts file's sound node, enable mcbsp1, set the
>>> pinctrl entries, etc if you boot with DT.
>> Yes and it is done as below:
>>
>> &omap3_pmx_core {
>> mcbsp1_pins: pinmux_mcbsp1_pins {
>> pinctrl-single,pins = <
>> OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0)
>> OMAP3_CORE1_IOPAD(0x2194, PIN_INPUT | MUX_MODE0)
>> OMAP3_CORE1_IOPAD(0x2190, PIN_INPUT | MUX_MODE0)
>> OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0)
>
> Should be:
> OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0) /* CLKX */
> OMAP3_CORE1_IOPAD(0x2196, PIN_INPUT | MUX_MODE0) /* FSX */
> OMAP3_CORE1_IOPAD(0x2192, PIN_INPUT | MUX_MODE0) /* DR */
> OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0) /* DX */
>
>
> And make sure to connect lrclk/bclk to FSX/CLKX (not to FSR/CLKR) of McBSP1.
>
> I have tested on on my xM that McBSP1 in master mode works, DMA is running (I
> do not have anything connected to pins).
Thanks for this. Thomas Niederprüm was kind enough to provide also the
same details.
So summarizing it as below:
Inorder to use mcbsp1 pins connected to the auxillary header in
beagleboard-xm we
have to do the configure the below pins as output/input depending on what is
who is running as master or slave.
As stated in the DM3730 TRM the register offsets for the mcbsp1 pins
are:
0x218C -> mcbsp1_clkr
0x218E -> mcbsp1_fsr
0x2190 -> mcbsp1_dx
0x2192 -> mcbsp1_dr
0x2194 -> mcbsp1_clks (not exposed)
0x2196 -> mcbsp1_fsx
0x2198 -> mcbsp1_clkx
Also note that the pins mcbsp1_clkr, mcbsp1_dx, mcbsp1_dr and
mcbsp1_fsx interfere with the mcspi4 pins. So if you use those pins it
might be worth to disable mcspi4:
&mcspi4 {
status = "disabled";
};
In my case i was running as mcbsp as slave and my codec as master.For
this configuration i did muxing as below in device tree file.
+ mcbsp1_pins: pinmux_mcbsp1_pins {
+ pinctrl-single,pins = <
+ OMAP3_CORE1_IOPAD(0x2196, PIN_INPUT | MUX_MODE0)
+ OMAP3_CORE1_IOPAD(0x2198, PIN_INPUT | MUX_MODE0)
+ OMAP3_CORE1_IOPAD(0x2190, PIN_OUTPUT | MUX_MODE0)
+ OMAP3_CORE1_IOPAD(0x2192, PIN_INPUT | MUX_MODE0)
+ >;
+ };
>
>> >;
>> };
>> }
>>
>> sound {
>> - ti,mcbsp = <&mcbsp2>;
>> - ti,codec = <&twl_audio>;
>> + ti,mcbsp = <&mcbsp1>;
>> + ti,codec = <&max98090>;
>> }
>>
>> &mcbsp1 {
>> pinctrl-names = "default";
>> pinctrl-0 = <&mcbsp1_pins>;
>>
>> status = "okay";
>> };
>>
>> &i2c2 {
>> clock-frequency = <400000>;
>>
>> max98090: max98090@10 {
>> compatible = "maxim,max98090";
>> reg = <0x10>;
>> };
>> };
>>
>> all other changes remain the same. Have i done muxing wrong? I am
>> still stuck with mcbsp not getting interrupt.
>> please find the dts file here http://ideone.com/TPFkBv
>>
>> I connected digital output of 98090 to mcbsp1_dr and digital receive
>> to mcbsp1_dx. I can see the bit clock and lrclk is coming fine from
>> 98090. As already mentioned i have connected lrclk, bit clock, DR
>> and DX only.
>>
>> Kernel logs here: http://ideone.com/hDHq2l
>>
>>>
>>>>
>>>> omap/omap-twl4030.c
>>>> added in hw_params below function to set internal oscillator for 98090
>>>> for generating PLL for master clock.
>>>> snd_soc_dai_set_sysclk(codec_dai, 26000000, 0, SND_SOC_CLOCK_IN);
>>>>
>>>> I am able to see that the cpu_dai is mapped with codec_dai and
>>>> configuration is ok. kernel logs at https://ideone.com/oezNSj
>>>>
>>>> I think i don't need to configure the mcbsp1 pins right as it uses default.
>>>>
>>>> Please guide me.I have tried IRC channel also but not getting the right folks.
>>>>
>>>> https://github.com/CircuitCo/BeagleBoard-xM-RevC/blob/master/BeagleBoard-xM_revC_SCH.pdf?raw=true
>>>>
>>>> Thanks,
>>>>
>>>
>>>
>>> --
>>> Péter
>
>
> --
> Péter
_______________________________________________
Alsa-devel mailing list
Alsa-devel@alsa-project.org
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2015-02-25 20:09 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-02-20 6:45 audio mcbsp1 broken for beagleboard-xm -OMAP noman pouigt
2015-02-20 16:24 ` Peter Ujfalusi
2015-02-24 0:03 ` noman pouigt
2015-02-20 16:27 ` Peter Ujfalusi
2015-02-21 2:36 ` noman pouigt
2015-02-25 9:40 ` Peter Ujfalusi
2015-02-25 20:09 ` noman pouigt
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.