Audio support on OMAP 3430LDP or 4430SDP boards

Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

Can someone please hint on what audio is supported on these two OMAP
boards?  From what I can tell, the answer is either "nothing" or "hdmi",
both of which are useless to me if someone wants me to convert the OMAP
ASoC driver to DMA engine.

>From what I can see on the 4430SDP, there's a 3.5mm headset jack,
microphones and speakers on the board, so the board does have audio
hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
much standard practice with OMAP) there's a total lack of software
support in mainline.

This kind'a prevents me testing changes to the DMA support for OMAP
ASoC...

What can be done to remedy this situation?

Thanks.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Shilimkar, Santosh]

+ Peter and Liam to comment

On Fri, Apr 27, 2012 at 7:28 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> Can someone please hint on what audio is supported on these two OMAP
> boards?  From what I can tell, the answer is either "nothing" or "hdmi",
> both of which are useless to me if someone wants me to convert the OMAP
> ASoC driver to DMA engine.
>
> From what I can see on the 4430SDP, there's a 3.5mm headset jack,
> microphones and speakers on the board, so the board does have audio
> hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
> much standard practice with OMAP) there's a total lack of software
> support in mainline.
>
> This kind'a prevents me testing changes to the DMA support for OMAP
> ASoC...
>
> What can be done to remedy this situation?
>
> Thanks.
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Liam Girdwood]

On Fri, 2012-04-27 at 19:45 +0530, Shilimkar, Santosh wrote:
> + Peter and Liam to comment
> 
> On Fri, Apr 27, 2012 at 7:28 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> > Can someone please hint on what audio is supported on these two OMAP
> > boards?  From what I can tell, the answer is either "nothing" or "hdmi",
> > both of which are useless to me if someone wants me to convert the OMAP
> > ASoC driver to DMA engine.
> >
> > From what I can see on the 4430SDP, there's a 3.5mm headset jack,
> > microphones and speakers on the board, so the board does have audio
> > hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
> > much standard practice with OMAP) there's a total lack of software
> > support in mainline.
> >
> > This kind'a prevents me testing changes to the DMA support for OMAP
> > ASoC...
> >
> > What can be done to remedy this situation?
> >

Peter now has has the final patches for Legacy mode (i.e. without ABE)
now upstream in 3.4-rc. Legacy mode gives a direct SDMA -> McPDM ->
twl6040 audio path so would be a good starting point for DMA testing.

Regards

Liam

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 03:56:42PM +0100, Liam Girdwood wrote:
> On Fri, 2012-04-27 at 19:45 +0530, Shilimkar, Santosh wrote:
> > + Peter and Liam to comment
> > 
> > On Fri, Apr 27, 2012 at 7:28 PM, Russell King - ARM Linux
> > <linux@arm.linux.org.uk> wrote:
> > > Can someone please hint on what audio is supported on these two OMAP
> > > boards?  From what I can tell, the answer is either "nothing" or "hdmi",
> > > both of which are useless to me if someone wants me to convert the OMAP
> > > ASoC driver to DMA engine.
> > >
> > > From what I can see on the 4430SDP, there's a 3.5mm headset jack,
> > > microphones and speakers on the board, so the board does have audio
> > > hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
> > > much standard practice with OMAP) there's a total lack of software
> > > support in mainline.
> > >
> > > This kind'a prevents me testing changes to the DMA support for OMAP
> > > ASoC...
> > >
> > > What can be done to remedy this situation?
> > >
> 
> Peter now has has the final patches for Legacy mode (i.e. without ABE)
> now upstream in 3.4-rc. Legacy mode gives a direct SDMA -> McPDM ->
> twl6040 audio path so would be a good starting point for DMA testing.

Sorry, what's ABE?  Association of Business Executives?  Association of
Building Engineers?

I had noticed that there's sound/soc/omap/snd-soc-omap-abe-twl6040.ko
which needs a bunch of other modules, but even this doesn't cause any
cards to show up.

So, I'm guessing that the real answer to my problem is that there is no
functional audio support on any OMAP board I have, and, therefore, I am
unable to continue with the work which TI would like me to do.  That's
great.  That's really bloody well great.

I'm going to leave this in Santosh's hands to deal with, and arrange to
get more complete and functional board support for any board that TI send
me into mainline - for example, I _still_ don't have any working video
output on the 4430SDP board.  Utter crap.

In fact, the only things that do work on the 4430SDP are serial and
network.  At least the 3430LDP has video output on its LCD.

When will support for OMAP based platforms become a reasonable reality?
When will people stop pissing around with utterly shit incomplete board
support?

I'm shelving further OMAP work until board support for these targets
improves to a state where a reasonable amount of the platform is
supported.  Until that time I'm not wasting any more of my time trying
to work out what I can do to assist OMAP along.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

Hi Russell,

On Fri, Apr 27, 2012 at 4:58 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> Can someone please hint on what audio is supported on these two OMAP
> boards?  From what I can tell, the answer is either "nothing" or "hdmi",
> both of which are useless to me if someone wants me to convert the OMAP
> ASoC driver to DMA engine.
>
> From what I can see on the 4430SDP, there's a 3.5mm headset jack,
> microphones and speakers on the board, so the board does have audio
> hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
> much standard practice with OMAP) there's a total lack of software
> support in mainline.

3.4-rc have audio support for SDP4430/Blaze, PandaBoard4430,
PandaBoardES (4460).
OMAP3 support is there for a long time, not sure about 3430LDP, but
the omap3-SDP,
BeagleBoards, Nokia n900, Pandora have at least ASoC machine driver.

> This kind'a prevents me testing changes to the DMA support for OMAP
> ASoC...

The ASoC driver for DMA: sound/soc/omap/omap-pcm.c
We also have dma engine support in the ASoC core:
sound/soc/soc-dmaengine-pcm.c
It provides some level of abstraction for us to use DMA engine with audio.

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 06:53:39PM +0300, Ujfalusi, Peter wrote:
> Hi Russell,
> 
> On Fri, Apr 27, 2012 at 4:58 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> > Can someone please hint on what audio is supported on these two OMAP
> > boards?  From what I can tell, the answer is either "nothing" or "hdmi",
> > both of which are useless to me if someone wants me to convert the OMAP
> > ASoC driver to DMA engine.
> >
> > From what I can see on the 4430SDP, there's a 3.5mm headset jack,
> > microphones and speakers on the board, so the board does have audio
> > hardware.  However, it seems (as, I'm afraid to say, seems to be pretty
> > much standard practice with OMAP) there's a total lack of software
> > support in mainline.
> 
> 3.4-rc have audio support for SDP4430/Blaze, PandaBoard4430,
> PandaBoardES (4460).
> OMAP3 support is there for a long time, not sure about 3430LDP, but
> the omap3-SDP,
> BeagleBoards, Nokia n900, Pandora have at least ASoC machine driver.

Right, so the only platform I have out of that (as I said above) is
SDP4430.

And as I've already said, it does *not* appear to support any audio
of any kind:

root@omap-4430sdp:~# insmod snd-soc-twl6040.ko                                  
root@omap-4430sdp:~# insmod snd-soc-omap.ko                                     
root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko                               
root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko                         
root@omap-4430sdp:~# cat /proc/asound/cards                                     
--- no soundcards ---                                                           

and that's with HDMI audio via DSS enabled.

So, it does *NOT* appear to support any kind of audio.  (If it does,
software support for it is broken when built as modules.)

How do I get audio support working on this platform?
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

On Fri, Apr 27, 2012 at 7:04 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> Right, so the only platform I have out of that (as I said above) is
> SDP4430.
>
> And as I've already said, it does *not* appear to support any audio
> of any kind:
>
> root@omap-4430sdp:~# insmod snd-soc-twl6040.ko
> root@omap-4430sdp:~# insmod snd-soc-omap.ko
> root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko
> root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko
> root@omap-4430sdp:~# cat /proc/asound/cards
> --- no soundcards ---

You need to load the DMIC driver as well:
insmod snd-soc-omap-dmic.ko

> So, it does *NOT* appear to support any kind of audio.  (If it does,
> software support for it is broken when built as modules.)

Audio should work as built in and as modules.

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 07:32:42PM +0300, Ujfalusi, Peter wrote:
> On Fri, Apr 27, 2012 at 7:04 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> > Right, so the only platform I have out of that (as I said above) is
> > SDP4430.
> >
> > And as I've already said, it does *not* appear to support any audio
> > of any kind:
> >
> > root@omap-4430sdp:~# insmod snd-soc-twl6040.ko
> > root@omap-4430sdp:~# insmod snd-soc-omap.ko
> > root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko
> > root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko
> > root@omap-4430sdp:~# cat /proc/asound/cards
> > --- no soundcards ---
>
> You need to load the DMIC driver as well:
> insmod snd-soc-omap-dmic.ko

You must be joking...

root@omap-4430sdp:~# insmod snd-soc-twl6040.ko
root@omap-4430sdp:~# insmod snd-soc-omap.ko
root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko
root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko
root@omap-4430sdp:~# insmod snd-soc-omap-dmic.ko
root@omap-4430sdp:~# cat /proc/asound/cards
--- no soundcards ---

And...

root@omap-4430sdp:~# rmmod snd_soc_omap_abe_twl6040
root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko
omap-abe-twl6040 omap-abe-twl6040: ASoC: Failed to create card debugfs directory
root@omap-4430sdp:~# cat /sys/kernel/debug/asoc/SDP4430/dapm_pop_time
Unable to handle kernel paging request at virtual address bf011e7c
pgd = def1c000
[bf011e7c] *pgd=9f511811, *pte=00000000, *ppte=00000000
Internal error: Oops: 7 [#1] SMP ARM
Modules linked in: snd_soc_omap_abe_twl6040 snd_soc_omap_dmic snd_soc_omap_mcpdm
 snd_soc_twl6040 snd_soc_omap [last unloaded: snd_soc_omap_abe_twl6040]
CPU: 1    Not tainted  (3.4.0-rc3+ #380)
PC is at debugfs_u32_get+0xc/0x20
LR is at simple_attr_read+0x70/0xc4
pc : [<c015fda0>]    lr : [<c00d65e0>]    psr: 60000113
sp : df4f3ef8  ip : df4f3f08  fp : df4f3f04
r10: 00001000  r9 : df4f2000  r8 : be8b5be0
r7 : df53e440  r6 : df4f3f70  r5 : df53e400  r4 : 00000000
r3 : 00000000  r2 : 00000000  r1 : df4f3f10  r0 : bf011e7c
Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment user
Control: 10c5387d  Table: 9ef1c04a  DAC: 00000015
Process cat (pid: 1494, stack limit = 0xdf4f22f0)
Stack: (0xdf4f3ef8 to 0xdf4f4000)
3ee0:                                                       df4f3f3c df4f3f08
3f00: c00d65e0 c015fda0 00000001 df532f08 df4f3f44 df4f3f20 df532f00 00001000
3f20: be8b5be0 df4f3f70 00001000 00000000 df4f3f6c df4f3f40 c00b59a4 c00d657c
3f40: df4f2000 def5e000 df4f3f94 00000000 00000000 df532f00 be8b5be0 00001000
3f60: df4f3fa4 df4f3f70 c00b5af4 c00b58fc 00000000 00000000 be8b6ef1 00000000
3f80: c0014068 b6fe7880 00000003 be8b5be0 00000003 c0014068 00000000 df4f3fa8
3fa0: c0013ec0 c00b5abc b6fe7880 00000003 00000003 be8b5be0 00001000 00098728
3fc0: b6fe7880 00000003 be8b5be0 00000003 00000000 00000000 00000003 be8b5bd4
3fe0: 00001000 be8b5bc0 00013140 b6e8da2c 60000110 00000003 4ddc5d55 75f55179
Backtrace:
[<c015fd94>] (debugfs_u32_get+0x0/0x20) from [<c00d65e0>] (simple_attr_read+0x70/0xc4)
[<c00d6570>] (simple_attr_read+0x0/0xc4) from [<c00b59a4>] (vfs_read+0xb4/0x140)
[<c00b58f0>] (vfs_read+0x0/0x140) from [<c00b5af4>] (sys_read+0x44/0x70)
 r8:00001000 r7:be8b5be0 r6:df532f00 r5:00000000 r4:00000000
[<c00b5ab0>] (sys_read+0x0/0x70) from [<c0013ec0>] (ret_fast_syscall+0x0/0x30)
 r8:c0014068 r7:00000003 r6:be8b5be0 r5:00000003 r4:b6fe7880
Code: e89da800 e1a0c00d e92dd800 e24cb004 (e5903000)
---[ end trace ba2e3db072bb311d ]---
Segmentation fault
root@omap-4430sdp:~#

That seems to be because of an error in the ASoC design.  ASoC only
tears down the resources after a card has been fully instantiated.
If it's been partially instantiated, let's not bother and leave
everything registered with all the various subsystems...

Sorry, I'm shelving this until someone reports that OMAP has some
meaningful working audio support.  At the moment, I'm entirely
unconvinced that any of this stuff has been tested as working.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

On Fri, Apr 27, 2012 at 7:50 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
>> You need to load the DMIC driver as well:
>> insmod snd-soc-omap-dmic.ko
>
> You must be joking...

Well SDP4430/Blaze have the following setup:
twl6040 via McPDM
digital microphones via DMIC

So I would:
root@omap-4430sdp:~# insmod snd-soc-twl6040.ko
root@omap-4430sdp:~# insmod snd-soc-omap.ko
root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko
root@omap-4430sdp:~# insmod snd-soc-omap-dmic.ko
root@omap-4430sdp:~# insmod snd-soc-dmic.ko
root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko

Does this works for you?

Or
make modules_install
depmod -a
modprobe snd-soc-omap-abe-twl6040

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 08:04:52PM +0300, Ujfalusi, Peter wrote:
> On Fri, Apr 27, 2012 at 7:50 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> >> You need to load the DMIC driver as well:
> >> insmod snd-soc-omap-dmic.ko
> >
> > You must be joking...
>
> Well SDP4430/Blaze have the following setup:
> twl6040 via McPDM
> digital microphones via DMIC
>
> So I would:
> root@omap-4430sdp:~# insmod snd-soc-twl6040.ko
> root@omap-4430sdp:~# insmod snd-soc-omap.ko
> root@omap-4430sdp:~# insmod snd-soc-omap-mcpdm.ko
> root@omap-4430sdp:~# insmod snd-soc-omap-dmic.ko
> root@omap-4430sdp:~# insmod snd-soc-dmic.ko
> root@omap-4430sdp:~# insmod snd-soc-omap-abe-twl6040.ko
>
> Does this works for you?

root@omap-4430sdp:~# lsmod
Module                  Size  Used by
snd_soc_dmic            1492  0
snd_soc_omap4_hdmi      1724  0
snd_soc_omap_hdmi       1536  0
snd_soc_omap_abe_twl6040     4867  0
snd_soc_omap_dmic       4248  0
snd_soc_omap_mcpdm      3077  1 snd_soc_omap_abe_twl6040
snd_soc_omap            2987  0
snd_soc_twl6040        18601  1 snd_soc_omap_abe_twl6040
root@omap-4430sdp:~# cat /proc/asound/cards
--- no soundcards ---
root@omap-4430sdp:~# grep . /sys/kernel/debug/asoc/*
/sys/kernel/debug/asoc/codecs:dmic-codec
/sys/kernel/debug/asoc/codecs:snd-soc-dummy
/sys/kernel/debug/asoc/codecs:omapdss_hdmi
/sys/kernel/debug/asoc/dais:dmic-hifi
/sys/kernel/debug/asoc/dais:omap-dmic
/sys/kernel/debug/asoc/dais:omap-mcpdm
/sys/kernel/debug/asoc/dais:snd-soc-dummy-dai
/sys/kernel/debug/asoc/dais:hdmi-audio-codec
/sys/kernel/debug/asoc/platforms:omap-pcm-audio
/sys/kernel/debug/asoc/platforms:snd-soc-dummy

So the answer is... no.

> Or
> make modules_install
> depmod -a
> modprobe snd-soc-omap-abe-twl6040

That's not going to help one bit, as the only two modules that
snd-soc-omap-abe-twl6040 depends on are snd-soc-omap-mcpdm and
snd-soc-twl6040.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

On Fri, Apr 27, 2012 at 8:19 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> root@omap-4430sdp:~# lsmod
> Module                  Size  Used by
> snd_soc_dmic            1492  0
> snd_soc_omap4_hdmi      1724  0
> snd_soc_omap_hdmi       1536  0
> snd_soc_omap_abe_twl6040     4867  0
> snd_soc_omap_dmic       4248  0
> snd_soc_omap_mcpdm      3077  1 snd_soc_omap_abe_twl6040
> snd_soc_omap            2987  0
> snd_soc_twl6040        18601  1 snd_soc_omap_abe_twl6040
> root@omap-4430sdp:~# cat /proc/asound/cards
> --- no soundcards ---
> root@omap-4430sdp:~# grep . /sys/kernel/debug/asoc/*
> /sys/kernel/debug/asoc/codecs:dmic-codec
> /sys/kernel/debug/asoc/codecs:snd-soc-dummy
> /sys/kernel/debug/asoc/codecs:omapdss_hdmi

the twl6040 codec is missing.

> /sys/kernel/debug/asoc/dais:dmic-hifi
> /sys/kernel/debug/asoc/dais:omap-dmic
> /sys/kernel/debug/asoc/dais:omap-mcpdm
> /sys/kernel/debug/asoc/dais:snd-soc-dummy-dai
> /sys/kernel/debug/asoc/dais:hdmi-audio-codec

as well the dai for the twl6040.

> /sys/kernel/debug/asoc/platforms:omap-pcm-audio
> /sys/kernel/debug/asoc/platforms:snd-soc-dummy
>
> So the answer is... no.

Hrm.
I don't think you can build the twl6040 MFD core driver as a module.
Do you have CONFIG_TWL6040_CORE=y ?
We have MFD for twl6040 since it provides audio and vibra functionality as well.

Can you check if you have this commit:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=8eaeb9393397be8eb700ab38a69c450975463b77

This is the second patch to separate the twl6040 from twl-core.
If you are missing this patch twl6040 will not probe (twl core will
not register the
audio if we have twl6030).

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 09:14:23PM +0300, Ujfalusi, Peter wrote:
> On Fri, Apr 27, 2012 at 8:19 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> > root@omap-4430sdp:~# lsmod
> > Module                  Size  Used by
> > snd_soc_dmic            1492  0
> > snd_soc_omap4_hdmi      1724  0
> > snd_soc_omap_hdmi       1536  0
> > snd_soc_omap_abe_twl6040     4867  0
> > snd_soc_omap_dmic       4248  0
> > snd_soc_omap_mcpdm      3077  1 snd_soc_omap_abe_twl6040
> > snd_soc_omap            2987  0
> > snd_soc_twl6040        18601  1 snd_soc_omap_abe_twl6040
> > root@omap-4430sdp:~# cat /proc/asound/cards
> > --- no soundcards ---
> > root@omap-4430sdp:~# grep . /sys/kernel/debug/asoc/*
> > /sys/kernel/debug/asoc/codecs:dmic-codec
> > /sys/kernel/debug/asoc/codecs:snd-soc-dummy
> > /sys/kernel/debug/asoc/codecs:omapdss_hdmi
> 
> the twl6040 codec is missing.
> 
> > /sys/kernel/debug/asoc/dais:dmic-hifi
> > /sys/kernel/debug/asoc/dais:omap-dmic
> > /sys/kernel/debug/asoc/dais:omap-mcpdm
> > /sys/kernel/debug/asoc/dais:snd-soc-dummy-dai
> > /sys/kernel/debug/asoc/dais:hdmi-audio-codec
> 
> as well the dai for the twl6040.
> 
> > /sys/kernel/debug/asoc/platforms:omap-pcm-audio
> > /sys/kernel/debug/asoc/platforms:snd-soc-dummy
> >
> > So the answer is... no.
> 
> Hrm.
> I don't think you can build the twl6040 MFD core driver as a module.
> Do you have CONFIG_TWL6040_CORE=y ?
> We have MFD for twl6040 since it provides audio and vibra functionality as well.

CONFIG_TWL6040_CORE=y

> Can you check if you have this commit:
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=8eaeb9393397be8eb700ab38a69c450975463b77

I don't have that in the branch I'm building because I started the
DMA engine work against v3.4-rc3.

> This is the second patch to separate the twl6040 from twl-core.
> If you are missing this patch twl6040 will not probe (twl core will
> not register the audio if we have twl6030).

So it sounds like if I rebase stuff forward to -rc4, it should start
working?
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

Hi Russell,

On Fri, Apr 27, 2012 at 9:31 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
>> Can you check if you have this commit:
>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=8eaeb9393397be8eb700ab38a69c450975463b77
>
> I don't have that in the branch I'm building because I started the
> DMA engine work against v3.4-rc3.
>
>> This is the second patch to separate the twl6040 from twl-core.
>> If you are missing this patch twl6040 will not probe (twl core will
>> not register the audio if we have twl6030).
>
> So it sounds like if I rebase stuff forward to -rc4, it should start
> working?

Yes, it should.
Unfortunately the two patch went separately to avoid merge issues.
You will need to configure the Headset path to be able to hear the audio.
I can send my .aconf if you are interested.

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 07:31:16PM +0100, Russell King - ARM Linux wrote:
> I don't have that in the branch I'm building because I started the
> DMA engine work against v3.4-rc3.
> 
> > This is the second patch to separate the twl6040 from twl-core.
> > If you are missing this patch twl6040 will not probe (twl core will
> > not register the audio if we have twl6030).
> 
> So it sounds like if I rebase stuff forward to -rc4, it should start
> working?

Right, so that was the problem.  Why not say in the first place that
it required -rc4 instead of just saying that it was in v3.4-rc ?  It's
taken severeal frustrating hours to get to this point...

But.. if I repeatedly aplay /dev/zero, I get randomly spat out on the
console:

omap_hwmod: mcpdm: _wait_target_disable failed

If I play a raw PCM audio file with aplay (44.1kHz s16le 2ch), I get
several of these from time to time:

DMA synchronization event drop occurred with device 66                          

and the 'noise' from the small loudspeakers is just about recognizable -
but is not what I was expecting (it's slow, plus it sounds on those small
speakers that it's horribly distorted.)  I can't get the headset jack
working to be able to play it through a decent sound system, through which
I'll be able to actually _hear_ what the corruption is.  I've tried fiddling
with everything in alsamixer, including turning on the vibras in the hope
of some kind of output (at least that does spin the motors.)

Timing the playback shows that something is very definitely not right:

On the laptop:
rmk@rmk-PC:[linux-rmk]:<2154> time aplay -f s16_le -c 2 -r 44100 NYG.s16.2.le 
Playing raw data 'NYG.s16.2.le' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo
real	4m56.911s
user	0m0.287s
sys	0m0.571s
rmk@rmk-PC:[linux-rmk]:<2155> md5sum NYG.s16.2.le
291d33b87e8d60556fd461216291889f  NYG.s16.2.le

On the SDP4430:
root@omap-4430sdp:~# time aplay -f s16_le -c 2 -r 44100 NYG.s16.2.le 
Playing raw data 'NYG.s16.2.le' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo
real	5m 53.67s
user	0m 3.17s
sys	0m 0.29s
root@omap-4430sdp:~# md5sum NYG.s16.2.le
291d33b87e8d60556fd461216291889f  NYG.s16.2.le

I notice that aplay is having to upconvert the audio to 96kHz 32 bit, so
that could be an ALSA library problem.

Any ideas?

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Fri, Apr 27, 2012 at 08:31:29PM +0100, Russell King - ARM Linux wrote:
> and the 'noise' from the small loudspeakers is just about recognizable -
> but is not what I was expecting (it's slow, plus it sounds on those small
> speakers that it's horribly distorted.)  I can't get the headset jack
> working to be able to play it through a decent sound system, through which
> I'll be able to actually _hear_ what the corruption is.

I've just tried pluging a 3.5mm stereo jack plug into the headset jack,
and tried loading up the rip and ring to sleave with various different
loads.  I see no reaction on the detection IRQ line:

389:          0          0   twl6040  twl6040_irq_plug

which seems to suggest that the jack detection is non-functional.

With all the documentation for the 4430SDP board hidden away under NDAs
within TI, I am unable to do any investigation myself to work out what's
wrong.

Please can someone at TI take a current mainline kernel, build it, and
test out audio support on the 4430SDP - specifically output through the
headset jack into a line input, and report when it works.  Thanks.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Peter Ujfalusi]

[-- Attachment #1: Type: text/plain, Size: 1232 bytes --]

On 04/29/2012 11:35 PM, Russell King - ARM Linux wrote:
> Please can someone at TI take a current mainline kernel, build it, and
> test out audio support on the 4430SDP - specifically output through the
> headset jack into a line input, and report when it works.  Thanks.

Try to use my attached aconf for SDP4430 to configure the HS playback.
Note that the HS is not a line out, so on your line-in device it will
cause distortion. If you have a headset/headphone you will have better
quality.
alsactl -f blaze.aconf restore

To play samples with aplay:
aplay -Dplughw:0,0 <wav file>
The sample format fot Legacy McPDM is S32_LE, 96KHz, 24bit (32 bit
sample with 24 LSB).

The slow pace audio playback is because at this point we do not have the
full audio stack up. The AESS IP does not have driver. You need to
disable CONFIG_OMAP_RESET_CLOCKS. This will fic the audio playback speed.

As for the basic usage:
We support up to 5 channel playback on OMAP4 in Legacy mode:
channel 1/2 (stereo): Headset, earpiece
channel 3/4: Handsfree/Line out
channel 5: Vibra

So with stereo sample you can only play to the Headset/Earpiece.
You need to have 5 channel audio, and valid data on channel 5 to be able
to drive the vibra.

-- 
Péter

[-- Attachment #2: blaze.aconf --]
[-- Type: text/plain, Size: 3844 bytes --]

state.SDP4430 {
	control.1 {
		iface MIXER
		name 'Capture Preamplifier Volume'
		value.0 1
		value.1 1
		comment {
			access 'read write'
			type INTEGER
			count 2
			range '0 - 1'
			dbmin -600
			dbmax 0
			dbvalue.0 0
			dbvalue.1 0
		}
	}
	control.2 {
		iface MIXER
		name 'Capture Volume'
		value.0 1
		value.1 1
		comment {
			access 'read write'
			type INTEGER
			count 2
			range '0 - 4'
			dbmin 600
			dbmax 3000
			dbvalue.0 1200
			dbvalue.1 1200
		}
	}
	control.3 {
		iface MIXER
		name 'Aux FM Volume'
		value.0 3
		value.1 3
		comment {
			access 'read write'
			type INTEGER
			count 2
			range '0 - 7'
			dbmin -1800
			dbmax 2400
			dbvalue.0 0
			dbvalue.1 0
		}
	}
	control.4 {
		iface MIXER
		name 'Headset Playback Volume'
		value.0 13
		value.1 13
		comment {
			access 'read write'
			type INTEGER
			count 2
			range '0 - 15'
			dbmin -3000
			dbmax 0
			dbvalue.0 -400
			dbvalue.1 -400
		}
	}
	control.5 {
		iface MIXER
		name 'Handsfree Playback Volume'
		value.0 4
		value.1 4
		comment {
			access 'read write'
			type INTEGER
			count 2
			range '0 - 29'
			dbmin -5200
			dbmax 600
			dbvalue.0 -4400
			dbvalue.1 -4400
		}
	}
	control.6 {
		iface MIXER
		name 'Earphone Playback Volume'
		value 2
		comment {
			access 'read write'
			type INTEGER
			count 1
			range '0 - 15'
			dbmin -2400
			dbmax 600
			dbvalue.0 -2000
		}
	}
	control.7 {
		iface MIXER
		name 'Headset Power Mode'
		value High-Perfomance
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Low-Power
			item.1 High-Perfomance
		}
	}
	control.8 {
		iface MIXER
		name 'PLL Selection'
		value High-Perfomance
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Low-Power
			item.1 High-Perfomance
		}
	}
	control.9 {
		iface MIXER
		name 'AUXR Playback Switch'
		value false
		comment {
			access 'read write'
			type BOOLEAN
			count 1
		}
	}
	control.10 {
		iface MIXER
		name 'AUXL Playback Switch'
		value false
		comment {
			access 'read write'
			type BOOLEAN
			count 1
		}
	}
	control.11 {
		iface MIXER
		name 'Earphone Playback Switch'
		value false
		comment {
			access 'read write'
			type BOOLEAN
			count 1
		}
	}
	control.12 {
		iface MIXER
		name 'Vibra Right Playback'
		value 'Input FF'
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 'Input FF'
			item.1 'Audio PDM'
		}
	}
	control.13 {
		iface MIXER
		name 'Vibra Left Playback'
		value 'Input FF'
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 'Input FF'
			item.1 'Audio PDM'
		}
	}
	control.14 {
		iface MIXER
		name 'Headset Right Playback'
		value 'HS DAC'
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Off
			item.1 'HS DAC'
			item.2 'Line-In amp'
		}
	}
	control.15 {
		iface MIXER
		name 'Headset Left Playback'
		value 'HS DAC'
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Off
			item.1 'HS DAC'
			item.2 'Line-In amp'
		}
	}
	control.16 {
		iface MIXER
		name 'Handsfree Right Playback'
		value Off
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Off
			item.1 'HF DAC'
			item.2 'Line-In amp'
		}
	}
	control.17 {
		iface MIXER
		name 'Handsfree Left Playback'
		value Off
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 Off
			item.1 'HF DAC'
			item.2 'Line-In amp'
		}
	}
	control.18 {
		iface MIXER
		name 'Analog Right Capture Route'
		value Off
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 'Headset Mic'
			item.1 'Sub Mic'
			item.2 'Aux/FM Right'
			item.3 Off
		}
	}
	control.19 {
		iface MIXER
		name 'Analog Left Capture Route'
		value Off
		comment {
			access 'read write'
			type ENUMERATED
			count 1
			item.0 'Headset Mic'
			item.1 'Main Mic'
			item.2 'Aux/FM Left'
			item.3 Off
		}
	}
}

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Mon, Apr 30, 2012 at 10:26:45AM +0300, Peter Ujfalusi wrote:
> On 04/29/2012 11:35 PM, Russell King - ARM Linux wrote:
> > Please can someone at TI take a current mainline kernel, build it, and
> > test out audio support on the 4430SDP - specifically output through the
> > headset jack into a line input, and report when it works.  Thanks.
> 
> Try to use my attached aconf for SDP4430 to configure the HS playback.
> Note that the HS is not a line out, so on your line-in device it will
> cause distortion. If you have a headset/headphone you will have better
> quality.
> alsactl -f blaze.aconf restore

This had no effect on its own.

> To play samples with aplay:
> aplay -Dplughw:0,0 <wav file>
> The sample format fot Legacy McPDM is S32_LE, 96KHz, 24bit (32 bit
> sample with 24 LSB).
> 
> The slow pace audio playback is because at this point we do not have the
> full audio stack up. The AESS IP does not have driver. You need to
> disable CONFIG_OMAP_RESET_CLOCKS. This will fic the audio playback speed.

But with that configuration file and that option disabled, I now get
audio output via the HS jack.  Bellowhead sounds a lot better now.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Mon, Apr 30, 2012 at 10:26:45AM +0300, Peter Ujfalusi wrote:
> On 04/29/2012 11:35 PM, Russell King - ARM Linux wrote:
> > Please can someone at TI take a current mainline kernel, build it, and
> > test out audio support on the 4430SDP - specifically output through the
> > headset jack into a line input, and report when it works.  Thanks.
> 
> Try to use my attached aconf for SDP4430 to configure the HS playback.
> Note that the HS is not a line out, so on your line-in device it will
> cause distortion. If you have a headset/headphone you will have better
> quality.
> alsactl -f blaze.aconf restore
> 
> To play samples with aplay:
> aplay -Dplughw:0,0 <wav file>
> The sample format fot Legacy McPDM is S32_LE, 96KHz, 24bit (32 bit
> sample with 24 LSB).

I'm just having another go with OMAP4430 audio... and again I find that
I can't get anything out of the board.

I've followed what you've said above, and it doesn't work.  I've undone
my DMA engine changes.  Still doesn't work.  I've checked
/sys/kernel/debug/asoc for the codecs and DAIs, and they're there, and
the SDP4430 audio is shown in /proc/asound/cards too, and yes, you
guessed it, it still doesn't work.

It's just a really expensive /dev/null from what I can tell at the moment.

Frustrated.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Thu, Jun 21, 2012 at 01:59:03PM +0100, Russell King - ARM Linux wrote:
> On Mon, Apr 30, 2012 at 10:26:45AM +0300, Peter Ujfalusi wrote:
> > On 04/29/2012 11:35 PM, Russell King - ARM Linux wrote:
> > > Please can someone at TI take a current mainline kernel, build it, and
> > > test out audio support on the 4430SDP - specifically output through the
> > > headset jack into a line input, and report when it works.  Thanks.
> > 
> > Try to use my attached aconf for SDP4430 to configure the HS playback.
> > Note that the HS is not a line out, so on your line-in device it will
> > cause distortion. If you have a headset/headphone you will have better
> > quality.
> > alsactl -f blaze.aconf restore
> > 
> > To play samples with aplay:
> > aplay -Dplughw:0,0 <wav file>
> > The sample format fot Legacy McPDM is S32_LE, 96KHz, 24bit (32 bit
> > sample with 24 LSB).
> 
> I'm just having another go with OMAP4430 audio... and again I find that
> I can't get anything out of the board.
> 
> I've followed what you've said above, and it doesn't work.  I've undone
> my DMA engine changes.  Still doesn't work.  I've checked
> /sys/kernel/debug/asoc for the codecs and DAIs, and they're there, and
> the SDP4430 audio is shown in /proc/asound/cards too, and yes, you
> guessed it, it still doesn't work.
> 
> It's just a really expensive /dev/null from what I can tell at the moment.
> 
> Frustrated.

Okay, reverting back to v3.4-rc4 works.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Thu, Jun 21, 2012 at 02:16:42PM +0100, Russell King - ARM Linux wrote:
> On Thu, Jun 21, 2012 at 01:59:03PM +0100, Russell King - ARM Linux wrote:
> > On Mon, Apr 30, 2012 at 10:26:45AM +0300, Peter Ujfalusi wrote:
> > > On 04/29/2012 11:35 PM, Russell King - ARM Linux wrote:
> > > > Please can someone at TI take a current mainline kernel, build it, and
> > > > test out audio support on the 4430SDP - specifically output through the
> > > > headset jack into a line input, and report when it works.  Thanks.
> > > 
> > > Try to use my attached aconf for SDP4430 to configure the HS playback.
> > > Note that the HS is not a line out, so on your line-in device it will
> > > cause distortion. If you have a headset/headphone you will have better
> > > quality.
> > > alsactl -f blaze.aconf restore
> > > 
> > > To play samples with aplay:
> > > aplay -Dplughw:0,0 <wav file>
> > > The sample format fot Legacy McPDM is S32_LE, 96KHz, 24bit (32 bit
> > > sample with 24 LSB).
> > 
> > I'm just having another go with OMAP4430 audio... and again I find that
> > I can't get anything out of the board.
> > 
> > I've followed what you've said above, and it doesn't work.  I've undone
> > my DMA engine changes.  Still doesn't work.  I've checked
> > /sys/kernel/debug/asoc for the codecs and DAIs, and they're there, and
> > the SDP4430 audio is shown in /proc/asound/cards too, and yes, you
> > guessed it, it still doesn't work.
> > 
> > It's just a really expensive /dev/null from what I can tell at the moment.
> > 
> > Frustrated.
> 
> Okay, reverting back to v3.4-rc4 works.

As does reverting to v3.4, of course.  So, it was broken during the last
merge window.

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Ujfalusi, Peter]

Hi Russel,

On Thu, Jun 21, 2012 at 4:22 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
>> Okay, reverting back to v3.4-rc4 works.
>
> As does reverting to v3.4, of course.  So, it was broken during the last
> merge window.

Strange...
I had no problem with the SDP4430/Blaze audio with 3.5-rc2 last week.
Have you restored the mixers with
alsactl -f blaze.aconf restore
command?

Does aplay (seemingly) plays the file OK or is it times out with an
error after a while?

-- 
Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Audio support on OMAP 3430LDP or 4430SDP boards

[Russell King - ARM Linux]

On Thu, Jun 21, 2012 at 09:36:12PM +0300, Ujfalusi, Peter wrote:
> Hi Russel,
> 
> On Thu, Jun 21, 2012 at 4:22 PM, Russell King - ARM Linux
> <linux@arm.linux.org.uk> wrote:
> >> Okay, reverting back to v3.4-rc4 works.
> >
> > As does reverting to v3.4, of course.  So, it was broken during the last
> > merge window.
> 
> Strange...
> I had no problem with the SDP4430/Blaze audio with 3.5-rc2 last week.
> Have you restored the mixers with
> alsactl -f blaze.aconf restore
> command?
> 
> Does aplay (seemingly) plays the file OK or is it times out with an
> error after a while?

It is as I said:

v3.5-rc3:
# alsactl -f blaze.aconf restore
# aplay -f S16_LE -r 48000 -c 2 file.raw
... plays in about the right time but no audio ...

v3.5-rc3 with the diff from v3.4..v3.5-rc5 for sound/soc/omap totally
 reverted:
# alsactl -f blaze.aconf restore
# aplay -f S16_LE -r 48000 -c 2 file.raw
... plays perfectly ...

Exactly the same procedure.  Only difference is I've backed out _all_ the
updates to sound/soc/omap.  Clearly something in there is buggered.
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html