diff mbox

[v5,4/7] ASoC: sunxi: Add sun8i I2S driver

Message ID bee315efb909c1ee8b5fe7d3aeb9a33d713cb562.1477142934.git.moinejf@free.fr (mailing list archive)
State New, archived
Headers show

Commit Message

Jean-Francois Moine Oct. 21, 2016, 8:36 a.m. UTC
This patch adds I2S support to sun8i SoCs as the A83T and H3.

Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
---
Note: This driver is closed to the sun4i-i2s except that:
- it handles the H3
- it creates the sound card (with sun4i-i2s, the sound card is created
  by the CODECs)
---
 .../devicetree/bindings/sound/sun4i-i2s.txt        |  38 +-
 sound/soc/sunxi/Kconfig                            |   8 +
 sound/soc/sunxi/Makefile                           |   3 +
 sound/soc/sunxi/sun8i-i2s.c                        | 700 +++++++++++++++++++++
 4 files changed, 744 insertions(+), 5 deletions(-)
 create mode 100644 sound/soc/sunxi/sun8i-i2s.c

Comments

Chen-Yu Tsai Oct. 23, 2016, 1:33 a.m. UTC | #1
Hi,

On Fri, Oct 21, 2016 at 4:36 PM, Jean-Francois Moine <moinejf@free.fr> wrote:
> This patch adds I2S support to sun8i SoCs as the A83T and H3.
>
> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> ---
> Note: This driver is closed to the sun4i-i2s except that:
> - it handles the H3

If it's close to sun4i-i2s, you should probably rework that one to support
the newer SoCs.

> - it creates the sound card (with sun4i-i2s, the sound card is created
>   by the CODECs)

I think this is wrong. I2S is only the DAI. You typically have a separate
platform driver for the whole card, or just use simple-card.

> ---
>  .../devicetree/bindings/sound/sun4i-i2s.txt        |  38 +-
>  sound/soc/sunxi/Kconfig                            |   8 +
>  sound/soc/sunxi/Makefile                           |   3 +
>  sound/soc/sunxi/sun8i-i2s.c                        | 700 +++++++++++++++++++++
>  4 files changed, 744 insertions(+), 5 deletions(-)
>  create mode 100644 sound/soc/sunxi/sun8i-i2s.c
>
> diff --git a/Documentation/devicetree/bindings/sound/sun4i-i2s.txt b/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
> index 7b526ec..2fb0a7a 100644
> --- a/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
> +++ b/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
> @@ -1,4 +1,4 @@
> -* Allwinner A10 I2S controller
> +* Allwinner A10/A38T/H3 I2S controller
>
>  The I2S bus (Inter-IC sound bus) is a serial link for digital
>  audio data transfer between devices in the system.
> @@ -6,20 +6,30 @@ audio data transfer between devices in the system.
>  Required properties:
>
>  - compatible: should be one of the followings
> -   - "allwinner,sun4i-a10-i2s"
> +  - "allwinner,sun4i-a10-i2s"
> +       "allwinner,sun8i-a83t-i2s"
> +       "allwinner,sun8i-h3-i2s"
>  - reg: physical base address of the controller and length of memory mapped
>    region.
> -- interrupts: should contain the I2S interrupt.
>  - dmas: DMA specifiers for tx and rx dma. See the DMA client binding,
>         Documentation/devicetree/bindings/dma/dma.txt
> -- dma-names: should include "tx" and "rx".
> +- dma-names: must include "tx" and/or "rx".
>  - clocks: a list of phandle + clock-specifer pairs, one for each entry in clock-names.
>  - clock-names: should contain followings:
>     - "apb" : clock for the I2S bus interface
>     - "mod" : module clock for the I2S controller
>  - #sound-dai-cells : Must be equal to 0
>
> -Example:
> +Optional properties:
> +
> +- interrupts: I2S interrupt
> +- resets: phandle to the reset of the device
> +
> +Required nodes:
> +
> + - port: link to the associated CODEC (DAC, HDMI...)

Note here you are changing an existing binding, adding a required node.
If it were truely different, you probably should've started a new binding.

Regards
ChenYu

> +
> +Example 1:
>
>  i2s0: i2s@01c22400 {
>         #sound-dai-cells = <0>;
> @@ -32,3 +42,21 @@ i2s0: i2s@01c22400 {
>                <&dma SUN4I_DMA_NORMAL 3>;
>         dma-names = "rx", "tx";
>  };
> +
> +Example 2:
> +
> +i2s2: i2s@1c22800 {
> +       compatible = "allwinner,sun8i-a83t-i2s";
> +       reg = <0x01c22800 0x60>;
> +       clocks = <&ccu CLK_BUS_I2S2>, <&ccu CLK_I2S2>;
> +       clock-names = "apb", "mod";
> +       resets = <&ccu RST_I2S2>;
> +       dmas = <&dma 27>;
> +       dma-names = "tx";
> +       status = "disabled";
> +       port {
> +               i2s2_hdmi: endpoint {
> +                       remote-endpoint = <&hdmi_i2s2>;
> +               };
> +       };
> +};
Jean-Francois Moine Oct. 23, 2016, 7:45 a.m. UTC | #2
On Sun, 23 Oct 2016 09:33:16 +0800
Chen-Yu Tsai <wens@csie.org> wrote:

> > Note: This driver is closed to the sun4i-i2s except that:
> > - it handles the H3
> 
> If it's close to sun4i-i2s, you should probably rework that one to support
> the newer SoCs.
> 
> > - it creates the sound card (with sun4i-i2s, the sound card is created
> >   by the CODECs)
> 
> I think this is wrong. I2S is only the DAI. You typically have a separate
> platform driver for the whole card, or just use simple-card.

An other device is not needed. The layout is simple:
	I2S_controller (CPU DAI) <-> HDMI_CODEC (CODEC DAI)
The HDMI CODEC is supported by the HDMI video driver (only one device),
so, it cannot be the card device.
ASoC does not use the CPU DAI device (I2S_controller), so, it is
natural to use it to handle the card.
Otherwise, the simple-card asks for a node definition in the DT and
this node is a pure Linux software entity. On the other side, the
simple-graph-card from Kuninori is not useful for this simple case.
Maxime Ripard Oct. 24, 2016, 12:34 p.m. UTC | #3
Hi,

On Sun, Oct 23, 2016 at 09:45:03AM +0200, Jean-Francois Moine wrote:
> On Sun, 23 Oct 2016 09:33:16 +0800
> Chen-Yu Tsai <wens@csie.org> wrote:
> 
> > > Note: This driver is closed to the sun4i-i2s except that:
> > > - it handles the H3
> > 
> > If it's close to sun4i-i2s, you should probably rework that one to support
> > the newer SoCs.
> > 
> > > - it creates the sound card (with sun4i-i2s, the sound card is created
> > >   by the CODECs)
> > 
> > I think this is wrong. I2S is only the DAI. You typically have a separate
> > platform driver for the whole card, or just use simple-card.
> 
> An other device is not needed. The layout is simple:
> 	I2S_controller (CPU DAI) <-> HDMI_CODEC (CODEC DAI)
> The HDMI CODEC is supported by the HDMI video driver (only one device),
> so, it cannot be the card device.
> ASoC does not use the CPU DAI device (I2S_controller), so, it is
> natural to use it to handle the card.

Still, duplicating the driver is not the solution. I agree with
Chen-Yu that we want to leverage the driver that is already there.

Maxime
Jean-Francois Moine Oct. 27, 2016, 3:13 p.m. UTC | #4
On Mon, 24 Oct 2016 14:34:49 +0200
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> Hi,
> 
> On Sun, Oct 23, 2016 at 09:45:03AM +0200, Jean-Francois Moine wrote:
> > On Sun, 23 Oct 2016 09:33:16 +0800
> > Chen-Yu Tsai <wens@csie.org> wrote:
> > 
> > > > Note: This driver is closed to the sun4i-i2s except that:
> > > > - it handles the H3
> > > 
> > > If it's close to sun4i-i2s, you should probably rework that one to support
> > > the newer SoCs.
> > > 
> > > > - it creates the sound card (with sun4i-i2s, the sound card is created
> > > >   by the CODECs)
> > > 
> > > I think this is wrong. I2S is only the DAI. You typically have a separate
> > > platform driver for the whole card, or just use simple-card.
> > 
> > An other device is not needed. The layout is simple:
> > 	I2S_controller (CPU DAI) <-> HDMI_CODEC (CODEC DAI)
> > The HDMI CODEC is supported by the HDMI video driver (only one device),
> > so, it cannot be the card device.
> > ASoC does not use the CPU DAI device (I2S_controller), so, it is
> > natural to use it to handle the card.
> 
> Still, duplicating the driver is not the solution. I agree with
> Chen-Yu that we want to leverage the driver that is already there.

Hi Maxime and Chen-Yu,

After looking at the sun4i-i2s, I found 2 solutions for re-using its
code in the DE2 HDMI context:

1) either to split the sun4i-i2s driver into common I/O functions and
   slave CPU DAI,

2) or to move the sun4i-i2s into a master CPU DAI.

(
 some explanation about 'master' and 'slave': the master is the
 component the device of which is also the sound card.
 As the sound card uses the 'drvdata' of the device, this drvdata pointer
 cannot be used by the master.
 In the actual implementations:
  - sun4i-i2s
	master:	card dev = codec dev, drvdata -> card
	slave:	i2s dev (CPU DAI), drvdata -> i2s data
  - sun8i-i2s
	master:	card dev = i2s dev (CPU DAI), drvdata -> card
	slave:	codec dev (hdmi), drvdata -> codec data (audio/video)
)

In the case 1, there is no functional change, just a source split.
The sun8i-i2s will then use the common I/O functions.

In the case 2, the CODECs using the sun4i-i2s would have to move to
slave CODEC DAI, i.e. the card is created by the sun4i-i2s code.
In the 4.9, there is only one codec (sun4i-codec), so, the change
is just to move the card creation and the use of drvdata in both
codes.

In either cases, I could not check if this changes raise some
regression on the sun4i SoCs side. Then, I'd be glad to know:
- which solution suits you?
- are you ready to do and test the needed changes at the sun4i side?
Maxime Ripard Oct. 27, 2016, 8:01 p.m. UTC | #5
On Thu, Oct 27, 2016 at 05:13:25PM +0200, Jean-Francois Moine wrote:
> On Mon, 24 Oct 2016 14:34:49 +0200
> Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
> 
> > Hi,
> > 
> > On Sun, Oct 23, 2016 at 09:45:03AM +0200, Jean-Francois Moine wrote:
> > > On Sun, 23 Oct 2016 09:33:16 +0800
> > > Chen-Yu Tsai <wens@csie.org> wrote:
> > > 
> > > > > Note: This driver is closed to the sun4i-i2s except that:
> > > > > - it handles the H3
> > > > 
> > > > If it's close to sun4i-i2s, you should probably rework that one to support
> > > > the newer SoCs.
> > > > 
> > > > > - it creates the sound card (with sun4i-i2s, the sound card is created
> > > > >   by the CODECs)
> > > > 
> > > > I think this is wrong. I2S is only the DAI. You typically have a separate
> > > > platform driver for the whole card, or just use simple-card.
> > > 
> > > An other device is not needed. The layout is simple:
> > > 	I2S_controller (CPU DAI) <-> HDMI_CODEC (CODEC DAI)
> > > The HDMI CODEC is supported by the HDMI video driver (only one device),
> > > so, it cannot be the card device.
> > > ASoC does not use the CPU DAI device (I2S_controller), so, it is
> > > natural to use it to handle the card.
> > 
> > Still, duplicating the driver is not the solution. I agree with
> > Chen-Yu that we want to leverage the driver that is already there.
> 
> Hi Maxime and Chen-Yu,
> 
> After looking at the sun4i-i2s, I found 2 solutions for re-using its
> code in the DE2 HDMI context:
> 
> 1) either to split the sun4i-i2s driver into common I/O functions and
>    slave CPU DAI,
> 
> 2) or to move the sun4i-i2s into a master CPU DAI.
> 
> (
>  some explanation about 'master' and 'slave': the master is the
>  component the device of which is also the sound card.
>  As the sound card uses the 'drvdata' of the device, this drvdata pointer
>  cannot be used by the master.
>  In the actual implementations:
>   - sun4i-i2s
> 	master:	card dev = codec dev, drvdata -> card
> 	slave:	i2s dev (CPU DAI), drvdata -> i2s data
>   - sun8i-i2s
> 	master:	card dev = i2s dev (CPU DAI), drvdata -> card
> 	slave:	codec dev (hdmi), drvdata -> codec data (audio/video)
> )
> 
> In the case 1, there is no functional change, just a source split.
> The sun8i-i2s will then use the common I/O functions.
> 
> In the case 2, the CODECs using the sun4i-i2s would have to move to
> slave CODEC DAI, i.e. the card is created by the sun4i-i2s code.
> In the 4.9, there is only one codec (sun4i-codec), so, the change
> is just to move the card creation and the use of drvdata in both
> codes.

I think you're mistaken. sun4i-codec has nothing to do with the I2S
driver. It's a driver for the (poorly named) Allwinner's Audio Codec
which features it's own DAI and Codec directly into the SoC.

The DAI being different from the I2S one.

However, we want to use any codec driver with the i2s driver,
including those in sound/soc/codecs, and we already have drivers for
them.

So I'm not sure either solution is a good one. Why not just make the
HDMI part a codec itself, and use the i2s driver as the CPU DAI, like
any other codec?

Maxime
Jean-Francois Moine Nov. 6, 2016, 6:02 p.m. UTC | #6
On Sun, 23 Oct 2016 09:33:16 +0800
Chen-Yu Tsai <wens@csie.org> wrote:

> On Fri, Oct 21, 2016 at 4:36 PM, Jean-Francois Moine <moinejf@free.fr> wrote:
> > This patch adds I2S support to sun8i SoCs as the A83T and H3.
> >
> > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > ---
> > Note: This driver is closed to the sun4i-i2s except that:
> > - it handles the H3
> 
> If it's close to sun4i-i2s, you should probably rework that one to support
> the newer SoCs.

I started to add the H3 into the sun4i-i2s, but I am blocked with
regmap.
Many H3 registers are common with the A10, but some of them have more
or less fields, the fields may be at different offsets. And, finally,
some registers are completely different.
This would not raise any problem, except with regmap which is really
painful.

As I may understood, regmap is used to simplify suspend/resume, but, is
it useful to save the I2S register on suspend?
Practically, I am streaming some tune on my device. I suspend it for
any reason. The next morning, I resume it. Are you sure I want to
continue to hear the end of the tune?

I better think that streaming should be simply stopped on suspend.
Then, there is no need to save the playing registers, and, here I am,
there is no need to use regmap.

May I go this way?
Maxime Ripard Nov. 7, 2016, 8:05 p.m. UTC | #7
Hi,

On Sun, Nov 06, 2016 at 07:02:48PM +0100, Jean-Francois Moine wrote:
> On Sun, 23 Oct 2016 09:33:16 +0800
> Chen-Yu Tsai <wens@csie.org> wrote:
> 
> > On Fri, Oct 21, 2016 at 4:36 PM, Jean-Francois Moine <moinejf@free.fr> wrote:
> > > This patch adds I2S support to sun8i SoCs as the A83T and H3.
> > >
> > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > ---
> > > Note: This driver is closed to the sun4i-i2s except that:
> > > - it handles the H3
> > 
> > If it's close to sun4i-i2s, you should probably rework that one to support
> > the newer SoCs.
> 
> I started to add the H3 into the sun4i-i2s, but I am blocked with
> regmap.
> Many H3 registers are common with the A10, but some of them have more
> or less fields, the fields may be at different offsets. And, finally,
> some registers are completely different.
> This would not raise any problem, except with regmap which is really
> painful.

That's weird, because regmap's regmap_field should make that much
easier.

> As I may understood, regmap is used to simplify suspend/resume, but, is
> it useful to save the I2S register on suspend?
> Practically, I am streaming some tune on my device. I suspend it for
> any reason. The next morning, I resume it. Are you sure I want to
> continue to hear the end of the tune?
> 
> I better think that streaming should be simply stopped on suspend.

You're mistaken. The code in there is for *runtime* suspend, ie when
the device is no longer used, so that case shouldn't even happen at
all.

(And real suspend isn't supported anyway)

> Then, there is no need to save the playing registers, and, here I am,
> there is no need to use regmap.
> 
> May I go this way?

No, please don't. regmap is also providing very useful features, such
as access to all the registers through debugfs, or tracing. What
exactly feels painful to you?

Maxime
Jean-Francois Moine Nov. 8, 2016, 10:51 a.m. UTC | #8
On Mon, 7 Nov 2016 21:05:05 +0100
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> Hi,
> 
> On Sun, Nov 06, 2016 at 07:02:48PM +0100, Jean-Francois Moine wrote:
> > On Sun, 23 Oct 2016 09:33:16 +0800
> > Chen-Yu Tsai <wens@csie.org> wrote:
> > 
> > > On Fri, Oct 21, 2016 at 4:36 PM, Jean-Francois Moine <moinejf@free.fr> wrote:
> > > > This patch adds I2S support to sun8i SoCs as the A83T and H3.
> > > >
> > > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > > ---
> > > > Note: This driver is closed to the sun4i-i2s except that:
> > > > - it handles the H3
> > > 
> > > If it's close to sun4i-i2s, you should probably rework that one to support
> > > the newer SoCs.
> > 
> > I started to add the H3 into the sun4i-i2s, but I am blocked with
> > regmap.
> > Many H3 registers are common with the A10, but some of them have more
> > or less fields, the fields may be at different offsets. And, finally,
> > some registers are completely different.
> > This would not raise any problem, except with regmap which is really
> > painful.
> 
> That's weird, because regmap's regmap_field should make that much
> easier.

#define field_relaxed(addr, mask, val) \
	writel_relaxed((readl_relaxed(addr) & mask) | val, addr)

> > As I may understood, regmap is used to simplify suspend/resume, but, is
> > it useful to save the I2S register on suspend?
> > Practically, I am streaming some tune on my device. I suspend it for
> > any reason. The next morning, I resume it. Are you sure I want to
> > continue to hear the end of the tune?
> > 
> > I better think that streaming should be simply stopped on suspend.
> 
> You're mistaken. The code in there is for *runtime* suspend, ie when
> the device is no longer used, so that case shouldn't even happen at
> all.
> 
> (And real suspend isn't supported anyway)

Is it time to remove this useless code?

> > Then, there is no need to save the playing registers, and, here I am,
> > there is no need to use regmap.
> > 
> > May I go this way?
> 
> No, please don't. regmap is also providing very useful features, such
> as access to all the registers through debugfs, or tracing. What
> exactly feels painful to you?

When the I/O registers are in memory (that's the case), you may access
them (read and write) thru /dev/mem.
Also, is a register access trace really needed in this driver?

The pain is to define the regmap_config (which registers can be
read/write/volatile and which can be the values the u-boot let us in
the registers at startup time), and the lot of code which is run instead
of simple load/store machine instructions.
Maxime Ripard Nov. 14, 2016, 9:22 p.m. UTC | #9
On Tue, Nov 08, 2016 at 11:51:29AM +0100, Jean-Francois Moine wrote:
> On Mon, 7 Nov 2016 21:05:05 +0100
> Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
> 
> > Hi,
> > 
> > On Sun, Nov 06, 2016 at 07:02:48PM +0100, Jean-Francois Moine wrote:
> > > On Sun, 23 Oct 2016 09:33:16 +0800
> > > Chen-Yu Tsai <wens@csie.org> wrote:
> > > 
> > > > On Fri, Oct 21, 2016 at 4:36 PM, Jean-Francois Moine <moinejf@free.fr> wrote:
> > > > > This patch adds I2S support to sun8i SoCs as the A83T and H3.
> > > > >
> > > > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > > > ---
> > > > > Note: This driver is closed to the sun4i-i2s except that:
> > > > > - it handles the H3
> > > > 
> > > > If it's close to sun4i-i2s, you should probably rework that one to support
> > > > the newer SoCs.
> > > 
> > > I started to add the H3 into the sun4i-i2s, but I am blocked with
> > > regmap.
> > > Many H3 registers are common with the A10, but some of them have more
> > > or less fields, the fields may be at different offsets. And, finally,
> > > some registers are completely different.
> > > This would not raise any problem, except with regmap which is really
> > > painful.
> > 
> > That's weird, because regmap's regmap_field should make that much
> > easier.
> 
> #define field_relaxed(addr, mask, val) \
> 	writel_relaxed((readl_relaxed(addr) & mask) | val, addr)

I'm not sure what you mean here.

> > > As I may understood, regmap is used to simplify suspend/resume, but, is
> > > it useful to save the I2S register on suspend?
> > > Practically, I am streaming some tune on my device. I suspend it for
> > > any reason. The next morning, I resume it. Are you sure I want to
> > > continue to hear the end of the tune?
> > > 
> > > I better think that streaming should be simply stopped on suspend.
> > 
> > You're mistaken. The code in there is for *runtime* suspend, ie when
> > the device is no longer used, so that case shouldn't even happen at
> > all.
> > 
> > (And real suspend isn't supported anyway)
> 
> Is it time to remove this useless code?

Which useless code?

> > > Then, there is no need to save the playing registers, and, here I am,
> > > there is no need to use regmap.
> > > 
> > > May I go this way?
> > 
> > No, please don't. regmap is also providing very useful features, such
> > as access to all the registers through debugfs, or tracing. What
> > exactly feels painful to you?
> 
> When the I/O registers are in memory (that's the case), you may access
> them (read and write) thru /dev/mem.

For all the registers if you want to dump all of them. It needs
scripting, it needs root access, and it needs some tool (either devmem
or a custom one) to dump the values. And this is if you have the right
kernel configuration options (devmem enabled, with the protection
against mapped devices disabled).

It just works with debugfs.

> Also, is a register access trace really needed in this driver?

Yes.

> The pain is to define the regmap_config (which registers can be
> read/write/volatile and which can be the values the u-boot let us in
> the registers at startup time), and the lot of code which is run instead
> of simple load/store machine instructions.

This is only needed if you want to use caching, and caching is
optional.

Maxime
diff mbox

Patch

diff --git a/Documentation/devicetree/bindings/sound/sun4i-i2s.txt b/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
index 7b526ec..2fb0a7a 100644
--- a/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
+++ b/Documentation/devicetree/bindings/sound/sun4i-i2s.txt
@@ -1,4 +1,4 @@ 
-* Allwinner A10 I2S controller
+* Allwinner A10/A38T/H3 I2S controller
 
 The I2S bus (Inter-IC sound bus) is a serial link for digital
 audio data transfer between devices in the system.
@@ -6,20 +6,30 @@  audio data transfer between devices in the system.
 Required properties:
 
 - compatible: should be one of the followings
-   - "allwinner,sun4i-a10-i2s"
+  - "allwinner,sun4i-a10-i2s"
+	"allwinner,sun8i-a83t-i2s"
+	"allwinner,sun8i-h3-i2s"
 - reg: physical base address of the controller and length of memory mapped
   region.
-- interrupts: should contain the I2S interrupt.
 - dmas: DMA specifiers for tx and rx dma. See the DMA client binding,
 	Documentation/devicetree/bindings/dma/dma.txt
-- dma-names: should include "tx" and "rx".
+- dma-names: must include "tx" and/or "rx".
 - clocks: a list of phandle + clock-specifer pairs, one for each entry in clock-names.
 - clock-names: should contain followings:
    - "apb" : clock for the I2S bus interface
    - "mod" : module clock for the I2S controller
 - #sound-dai-cells : Must be equal to 0
 
-Example:
+Optional properties:
+
+- interrupts: I2S interrupt
+- resets: phandle to the reset of the device
+
+Required nodes:
+
+ - port: link to the associated CODEC (DAC, HDMI...)
+
+Example 1:
 
 i2s0: i2s@01c22400 {
 	#sound-dai-cells = <0>;
@@ -32,3 +42,21 @@  i2s0: i2s@01c22400 {
 	       <&dma SUN4I_DMA_NORMAL 3>;
 	dma-names = "rx", "tx";
 };
+
+Example 2:
+
+i2s2: i2s@1c22800 {
+	compatible = "allwinner,sun8i-a83t-i2s";
+	reg = <0x01c22800 0x60>;
+	clocks = <&ccu CLK_BUS_I2S2>, <&ccu CLK_I2S2>;
+	clock-names = "apb", "mod";
+	resets = <&ccu RST_I2S2>;
+	dmas = <&dma 27>;
+	dma-names = "tx";
+	status = "disabled";
+	port {
+		i2s2_hdmi: endpoint {
+			remote-endpoint = <&hdmi_i2s2>;
+		};
+	};
+};
diff --git a/sound/soc/sunxi/Kconfig b/sound/soc/sunxi/Kconfig
index dd23682..d89b2da 100644
--- a/sound/soc/sunxi/Kconfig
+++ b/sound/soc/sunxi/Kconfig
@@ -26,4 +26,12 @@  config SND_SUN4I_SPDIF
 	help
 	  Say Y or M to add support for the S/PDIF audio block in the Allwinner
 	  A10 and affiliated SoCs.
+
+config SND_SUN8I_I2S
+	tristate "Allwinner sun8i I2S Support"
+	depends on OF
+	select SND_SOC_GENERIC_DMAENGINE_PCM
+	help
+	  Say Y or M if you want to add support for SoC audio on
+	  Allwinner sun8i boards.
 endmenu
diff --git a/sound/soc/sunxi/Makefile b/sound/soc/sunxi/Makefile
index 604c7b84..bcb871b 100644
--- a/sound/soc/sunxi/Makefile
+++ b/sound/soc/sunxi/Makefile
@@ -1,3 +1,6 @@ 
 obj-$(CONFIG_SND_SUN4I_CODEC) += sun4i-codec.o
 obj-$(CONFIG_SND_SUN4I_I2S) += sun4i-i2s.o
 obj-$(CONFIG_SND_SUN4I_SPDIF) += sun4i-spdif.o
+
+snd-soc-sun8i-i2s-objs := sun8i-i2s.o
+obj-$(CONFIG_SND_SUN8I_I2S) += snd-soc-sun8i-i2s.o
diff --git a/sound/soc/sunxi/sun8i-i2s.c b/sound/soc/sunxi/sun8i-i2s.c
new file mode 100644
index 0000000..ba15d62
--- /dev/null
+++ b/sound/soc/sunxi/sun8i-i2s.c
@@ -0,0 +1,700 @@ 
+/*
+ * Allwinner sun8i I2S sound card
+ *
+ * Copyright (C) 2016 Jean-Francois Moine <moinejf@free.fr>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/clk.h>
+#include <linux/reset.h>
+#include <sound/pcm_params.h>
+#include <sound/dmaengine_pcm.h>
+#include <linux/of_device.h>
+#include <linux/of_graph.h>
+
+/* --- hardware --- */
+
+#define I2S_CTL 	  	0x00
+	/* common */
+	#define I2S_CTL_SDO3EN		BIT(11)
+	#define I2S_CTL_SDO2EN		BIT(10)
+	#define I2S_CTL_SDO1EN		BIT(9)
+	#define I2S_CTL_SDO0EN		BIT(8)
+	#define I2S_CTL_TXEN		BIT(2)
+	#define I2S_CTL_RXEN		BIT(1)
+	#define I2S_CTL_GEN		BIT(0)
+	/* a83t */
+	#define I2S_CTL_A83T_MS		BIT(5)
+	#define I2S_CTL_A83T_PCM	BIT(4)
+	/* h3 */
+	#define I2S_CTL_H3_BCLKOUT	BIT(18)
+	#define I2S_CTL_H3_LRCKOUT	BIT(17)
+	#define I2S_CTL_H3_MODE_MSK	(3 << 4)
+		#define I2S_CTL_H3_MODE_I2S	(1 << 4)
+
+#define I2S_FAT0 		0x04
+	/* common */
+	/* a83t */
+	#define I2S_FAT0_A83T_LRCP		BIT(7)
+	#define I2S_FAT0_A83T_BCP		BIT(6)
+	#define I2S_FAT0_A83T_SR_16BIT		(0 << 4)
+	#define I2S_FAT0_A83T_SR_24BIT		(2 << 4)
+	#define I2S_FAT0_A83T_SR_MSK		(3 << 4)
+	#define I2S_FAT0_A83T_WSS_32BCLK	(3 << 2)
+	#define I2S_FAT0_A83T_FMT_I2S1		(0 << 0)
+	#define I2S_FAT0_A83T_FMT_LFT		(1 << 0)
+	#define I2S_FAT0_A83T_FMT_RGT		(2 << 0)
+	#define I2S_FAT0_A83T_FMT_MSK		(3 << 0)
+	/* h3 */
+	#define I2S_FAT0_H3_LRCKR_PERIOD(v) ((v) << 20)
+	#define I2S_FAT0_H3_LRCKR_PERIOD_MSK (0x3ff << 20)
+	#define I2S_FAT0_H3_LRCK_POLARITY	BIT(19)
+	#define I2S_FAT0_H3_LRCK_PERIOD(v)	((v) << 8)
+	#define I2S_FAT0_H3_LRCK_PERIOD_MSK (0x3ff << 8)
+	#define I2S_FAT0_H3_BCLK_POLARITY	BIT(7)
+	#define I2S_FAT0_H3_SR_16		(3 << 4)
+	#define I2S_FAT0_H3_SR_24		(5 << 4)
+	#define I2S_FAT0_H3_SR_MSK		(7 << 4)
+	#define I2S_FAT0_H3_SW_16		(3 << 0)
+	#define I2S_FAT0_H3_SW_32		(7 << 0)
+	#define I2S_FAT0_H3_SW_MSK		(7 << 0)
+
+#define I2S_FAT1		0x08
+
+#define I2S_FCTL		0x14
+	#define I2S_FCTL_FTX		BIT(25)
+	#define I2S_FCTL_FRX		BIT(24)
+	#define I2S_FCTL_TXTL(v)	((v) << 12)
+	#define I2S_FCTL_TXIM		BIT(2)
+
+#define I2S_INT    		0x1c
+	#define I2S_INT_TXDRQEN		BIT(7)
+
+#define I2S_TXFIFO		0x20
+	
+#define I2S_CLKD   		0x24
+	/* common */
+	#define I2S_CLKD_BCLKDIV(v)	((v) << 4)
+	#define I2S_CLKD_MCLKDIV(v)	((v) << 0)
+	/* a83t */
+	#define I2S_CLKD_A83T_MCLKOEN	BIT(7)
+	/* h3 */
+	#define I2S_CLKD_H3_MCLKOEN	BIT(8)
+
+#define I2S_TXCNT  		0x28
+
+#define I2S_RXCNT  		0x2c
+
+/* --- A83T --- */
+#define I2S_TXCHSEL_A83T	0x30
+	#define I2S_TXCHSEL_A83T_CHNUM(v)	(((v) - 1) << 0)
+	#define I2S_TXCHSEL_A83T_CHNUM_MSK	(7 << 0)
+
+#define I2S_TXCHMAP_A83T	0x34
+
+/* --- H3 --- */
+#define I2S_TXCHCFG_H3		0x30
+	#define I2S_TXCHCFG_H3_TX_SLOT_NUM_MSK (7 << 0)
+	#define I2S_TXCHCFG_H3_TX_SLOT_NUM(v) ((v) << 0)
+
+#define I2S_TX0CHSEL_H3		0x34		/* 0..3 */
+	#define I2S_TXn_H3_OFFSET_MSK	(3 << 12)
+	#define I2S_TXn_H3_OFFSET(v)	((v) << 12)
+	#define I2S_TXn_H3_CHEN_MSK	(0xff << 4)
+	#define I2S_TXn_H3_CHEN(v)	((v) << 4)
+	#define I2S_TXn_H3_CHSEL_MSK	(7 << 0)
+	#define I2S_TXn_H3_CHSEL(v)	((v) << 0)
+
+#define I2S_TX0CHMAP_H3		0x44		/* 0..3 */
+
+/* --- driver --- */
+
+#define DRV_NAME "sun8i-audio"
+
+#define I2S_FORMATS \
+	(SNDRV_PCM_FMTBIT_S16_LE | \
+	 SNDRV_PCM_FMTBIT_S20_3LE | \
+	 SNDRV_PCM_FMTBIT_S24_LE | \
+	 SNDRV_PCM_FMTBIT_S32_LE)
+
+#define PCM_LRCK_PERIOD 32
+#define PCM_LRCKR_PERIOD 1
+
+struct priv {
+	void __iomem *mmio;
+	struct clk *clk;
+	struct clk *gate;
+	struct reset_control *rstc;
+	int type;
+#define SOC_A83T 0
+#define SOC_H3 1
+	struct snd_dmaengine_dai_dma_data dma_data;
+};
+
+static const struct of_device_id sun8i_i2s_of_match[] = {
+	{ .compatible = "allwinner,sun8i-a83t-i2s",
+				.data = (void *) SOC_A83T },
+	{ .compatible = "allwinner,sun8i-h3-i2s",
+				.data = (void *) SOC_H3 },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, sun8i_i2s_of_match);
+
+/* --- CPU DAI --- */
+
+static void sun8i_i2s_init(struct priv *priv)
+{
+	u32 reg;
+
+	/* disable global */
+	reg = readl(priv->mmio + I2S_CTL);
+	reg &= ~(I2S_CTL_GEN |
+		 I2S_CTL_RXEN |
+		 I2S_CTL_TXEN);
+	writel(reg, priv->mmio + I2S_CTL);
+
+	/* A83T */
+	if (priv->type == SOC_A83T) {
+		reg &= ~(I2S_CTL_A83T_MS |	/* codec clk & FRM slave */
+			 I2S_CTL_A83T_PCM);	/* I2S mode */
+		writel(reg, priv->mmio + I2S_CTL);
+
+		reg = readl(priv->mmio + I2S_FAT0);
+		reg &= ~I2S_FAT0_A83T_FMT_MSK;
+		reg |= I2S_FAT0_A83T_FMT_I2S1;
+
+		reg &= ~(I2S_FAT0_A83T_LRCP | I2S_FAT0_A83T_BCP);
+		writel(reg, priv->mmio + I2S_FAT0);
+
+		reg = I2S_FCTL_TXIM |			/* fifo */
+			 I2S_FCTL_TXTL(0x40);
+		writel(reg, priv->mmio + I2S_FCTL);
+
+		reg = readl(priv->mmio + I2S_FAT0);
+		reg &= ~(I2S_FAT0_A83T_LRCP |	/* normal bit clock + frame */
+			 I2S_FAT0_A83T_BCP);
+		writel(reg, priv->mmio + I2S_FAT0);
+
+	/* H3 */
+	} else {
+		reg = readl(priv->mmio + I2S_FCTL);
+		reg &= ~(I2S_FCTL_FRX | I2S_FCTL_FTX);	/* clear the FIFOs */
+		writel(reg, priv->mmio + I2S_FCTL);
+
+		writel(0, priv->mmio + I2S_TXCNT);	/* FIFO counters */
+		writel(0, priv->mmio + I2S_RXCNT);
+
+		reg = readl(priv->mmio + I2S_CTL);
+		reg |= I2S_CTL_H3_LRCKOUT | I2S_CTL_H3_BCLKOUT;
+						/* codec clk & FRM slave */
+
+		reg &= ~I2S_CTL_H3_MODE_MSK;
+		reg |= I2S_CTL_H3_MODE_I2S;		/* I2S mode */
+		writel(reg, priv->mmio + I2S_CTL);
+
+		reg = readl(priv->mmio + I2S_TX0CHSEL_H3) &
+					~I2S_TXn_H3_OFFSET_MSK;
+		reg |= I2S_TXn_H3_OFFSET(1);
+		writel(reg, priv->mmio + I2S_TX0CHSEL_H3);
+
+		reg = readl(priv->mmio + I2S_FAT0);
+		reg &= ~(I2S_FAT0_H3_BCLK_POLARITY | /* normal bclk & frame */
+			 I2S_FAT0_H3_LRCK_POLARITY);
+		writel(reg, priv->mmio + I2S_FAT0);
+	}
+}
+
+static int sun8i_i2s_set_clock(struct priv *priv,
+				unsigned long rate)
+{
+	unsigned long freq;
+	int ret, i, div;
+	u32 reg;
+	static const u8 div_tb[] = {
+		1, 2, 4, 6, 8, 12, 16, 24,
+	};
+
+	/* compute the sys clock rate and divide values */
+	if (rate % 1000 == 0)
+		freq = 24576000;
+	else
+		freq = 22579200;
+	div = freq / 2 / PCM_LRCK_PERIOD / rate;
+	if (priv->type == SOC_A83T)
+		div /= 2;			/* bclk_div==0 => mclk/2 */
+	for (i = 0; i < ARRAY_SIZE(div_tb) - 1; i++)
+		if (div_tb[i] >= div)
+			break;
+
+	ret = clk_set_rate(priv->clk, freq);
+	if (ret) {
+		pr_info("Setting sysclk rate failed %d\n", ret);
+		return ret;
+	}
+
+	/* set the mclk and bclk dividor register */
+	if (priv->type == SOC_A83T) {
+		reg = I2S_CLKD_A83T_MCLKOEN | I2S_CLKD_MCLKDIV(i);
+	} else {
+		reg = I2S_CLKD_H3_MCLKOEN | I2S_CLKD_MCLKDIV(1) |
+				I2S_CLKD_BCLKDIV(i + 1);
+	}
+	writel(reg, priv->mmio + I2S_CLKD);
+
+	/* format */
+	reg = readl(priv->mmio + I2S_FAT0);
+	if (priv->type == SOC_A83T) {
+		reg |= I2S_FAT0_A83T_WSS_32BCLK;
+		reg &= ~I2S_FAT0_A83T_SR_MSK;
+		reg |= I2S_FAT0_A83T_SR_16BIT;
+	} else {
+		reg &= ~(I2S_FAT0_H3_LRCKR_PERIOD_MSK |
+			 I2S_FAT0_H3_LRCK_PERIOD_MSK);
+		reg |= I2S_FAT0_H3_LRCK_PERIOD(PCM_LRCK_PERIOD - 1) |
+			I2S_FAT0_H3_LRCKR_PERIOD(PCM_LRCKR_PERIOD - 1);
+
+		reg &= ~I2S_FAT0_H3_SW_MSK;
+		reg |= I2S_FAT0_H3_SW_16;
+
+		reg &= ~I2S_FAT0_H3_SR_MSK;
+		reg |= I2S_FAT0_H3_SR_16;
+	}
+	writel(reg, priv->mmio + I2S_FAT0);
+
+	writel(0, priv->mmio + I2S_FAT1);
+
+	return 0;
+}
+
+static int sun8i_i2s_prepare(struct snd_pcm_substream *substream,
+			     struct snd_soc_dai *dai)
+{
+	struct snd_soc_card *card = snd_soc_dai_get_drvdata(dai);
+	struct priv *priv = snd_soc_card_get_drvdata(card);
+	int nchan = substream->runtime->channels;
+	u32 reg;
+
+	if (priv->type == SOC_A83T) {
+		reg = readl(priv->mmio + I2S_TXCHSEL_A83T);
+		reg &= ~I2S_TXCHSEL_A83T_CHNUM_MSK;
+		reg |= I2S_TXCHSEL_A83T_CHNUM(substream->runtime->channels);
+		writel(reg, priv->mmio + I2S_TXCHSEL_A83T);
+
+		switch (substream->runtime->channels) {
+		case 1:
+			reg = 0x76543200;
+			break;
+		case 8:
+			reg = 0x54762310;
+			break;
+		default:
+/* left/right inversion of channels 0 and 1 */
+			reg = 0x76543201;
+			break;
+		}
+		writel(reg, priv->mmio + I2S_TXCHMAP_A83T);
+	} else {
+		reg = readl(priv->mmio + I2S_TXCHCFG_H3) &
+					~I2S_TXCHCFG_H3_TX_SLOT_NUM_MSK;
+		if (nchan != 1)
+			reg |= I2S_TXCHCFG_H3_TX_SLOT_NUM(1);
+		writel(reg, priv->mmio + I2S_TXCHCFG_H3);
+
+		reg = readl(priv->mmio + I2S_TX0CHSEL_H3);
+		reg &= ~(I2S_TXn_H3_CHEN_MSK |
+			 I2S_TXn_H3_CHSEL_MSK);
+		reg |= I2S_TXn_H3_CHEN(3) |
+			I2S_TXn_H3_CHSEL(1);
+		writel(reg, priv->mmio + I2S_TX0CHSEL_H3);
+
+		reg = nchan == 1 ? 0 : 0x10;
+		writel(reg, priv->mmio + I2S_TX0CHMAP_H3);
+	}
+
+	reg = readl(priv->mmio + I2S_CTL);
+	reg &= ~(I2S_CTL_SDO3EN |
+		 I2S_CTL_SDO2EN |
+		 I2S_CTL_SDO1EN);
+	if (nchan >= 7)
+		reg |= I2S_CTL_SDO3EN;
+	if (nchan >= 5)
+		reg |= I2S_CTL_SDO2EN;
+	if (nchan >= 3)
+		reg |= I2S_CTL_SDO1EN;
+	reg |= I2S_CTL_SDO0EN;
+	writel(reg, priv->mmio + I2S_CTL);
+
+	writel(0, priv->mmio + I2S_TXCNT);
+
+	return 0;
+}
+
+static void sun8i_i2s_shutdown(struct snd_pcm_substream *substream,
+			      struct snd_soc_dai *dai)
+{
+	struct snd_soc_card *card = snd_soc_dai_get_drvdata(dai);
+	struct priv *priv = snd_soc_card_get_drvdata(card);
+	u32 reg;
+
+	reg = readl(priv->mmio + I2S_CTL);
+	reg &= ~I2S_CTL_GEN;
+	writel(reg, priv->mmio + I2S_CTL);
+}
+
+static int sun8i_i2s_hw_params(struct snd_pcm_substream *substream,
+				struct snd_pcm_hw_params *params,
+				struct snd_soc_dai *dai)
+{
+	struct snd_soc_card *card = snd_soc_dai_get_drvdata(dai);
+	struct priv *priv = snd_soc_card_get_drvdata(card);
+	u32 reg, reg2;
+	int sample_resolution;
+	int ret;
+
+	ret = sun8i_i2s_set_clock(priv, params_rate(params));
+	if (ret)
+		return ret;
+
+	switch (params_format(params)) {
+	case SNDRV_PCM_FORMAT_S16_LE:
+		priv->dma_data.addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		sample_resolution = 16;
+		break;
+	case SNDRV_PCM_FORMAT_S20_3LE:
+	case SNDRV_PCM_FORMAT_S24_LE:
+	case SNDRV_PCM_FORMAT_S32_LE:
+		priv->dma_data.addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		sample_resolution = 24;
+		break;
+	default:
+		return -EINVAL;
+	}
+	reg = readl(priv->mmio + I2S_FAT0);
+	reg2 = readl(priv->mmio + I2S_FCTL);
+	if (priv->type == SOC_A83T) {
+		reg &= ~I2S_FAT0_A83T_SR_MSK;
+		if (sample_resolution == 16) {
+			reg |= I2S_FAT0_A83T_SR_16BIT;
+			reg2 |= I2S_FCTL_TXIM;
+		} else {
+			reg |= I2S_FAT0_A83T_SR_24BIT;
+			reg2 &= ~I2S_FCTL_TXIM;
+		}
+	} else {
+		reg &= ~(I2S_FAT0_H3_SR_MSK | I2S_FAT0_H3_SW_MSK);
+		if (sample_resolution == 16) {
+			reg |= I2S_FAT0_H3_SR_16 |
+					I2S_FAT0_H3_SW_16;
+			reg2 |= I2S_FCTL_TXIM;
+		} else {
+			reg |= I2S_FAT0_H3_SR_24 |
+					I2S_FAT0_H3_SW_32;
+			reg2 &= ~I2S_FCTL_TXIM;
+		}
+	}
+	writel(reg, priv->mmio + I2S_FAT0);
+	writel(reg2, priv->mmio + I2S_FCTL);
+
+	/* enable audio interface */
+	reg = readl(priv->mmio + I2S_CTL);
+	reg |= I2S_CTL_GEN;
+	writel(reg, priv->mmio + I2S_CTL);
+	msleep(10);
+
+	/* flush TX FIFO */
+	reg = readl(priv->mmio + I2S_FCTL);
+	reg |= I2S_FCTL_FTX;
+	writel(reg, priv->mmio + I2S_FCTL);
+
+	return 0;
+}
+
+static int sun8i_i2s_trigger(struct snd_pcm_substream *substream,
+				int cmd, struct snd_soc_dai *dai)
+{
+	struct snd_soc_card *card = snd_soc_dai_get_drvdata(dai);
+	struct priv *priv = snd_soc_card_get_drvdata(card);
+	u32 reg;
+
+	switch (cmd) {
+	case SNDRV_PCM_TRIGGER_START:
+	case SNDRV_PCM_TRIGGER_RESUME:
+	case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
+		reg = readl(priv->mmio + I2S_CTL);
+		reg |= I2S_CTL_TXEN;
+		writel(reg, priv->mmio + I2S_CTL);
+
+		/* enable DMA DRQ mode */
+		reg = readl(priv->mmio + I2S_INT);
+		reg |= I2S_INT_TXDRQEN;
+		writel(reg, priv->mmio + I2S_INT);
+		break;
+	case SNDRV_PCM_TRIGGER_STOP:
+	case SNDRV_PCM_TRIGGER_SUSPEND:
+	case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
+		reg = readl(priv->mmio + I2S_INT);
+		reg &= ~I2S_INT_TXDRQEN;
+		writel(reg, priv->mmio + I2S_INT);
+
+		reg = readl(priv->mmio + I2S_CTL);
+		reg &= ~I2S_CTL_TXEN;
+		writel(reg, priv->mmio + I2S_CTL);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const struct snd_soc_dai_ops sun8i_i2s_dai_ops = {
+	.hw_params = sun8i_i2s_hw_params,
+	.prepare = sun8i_i2s_prepare,
+	.trigger = sun8i_i2s_trigger,
+	.shutdown = sun8i_i2s_shutdown,
+};
+
+static int sun8i_i2s_dai_probe(struct snd_soc_dai *dai)
+{
+	struct snd_soc_card *card = snd_soc_dai_get_drvdata(dai);
+	struct priv *priv = snd_soc_card_get_drvdata(card);
+
+	snd_soc_dai_init_dma_data(dai, &priv->dma_data, NULL);
+
+	return 0;
+}
+
+static struct snd_soc_dai_driver sun8i_i2s_dai = {
+	.probe = sun8i_i2s_dai_probe,
+	.playback = {
+		.stream_name = "Playback",
+		.channels_min = 1,
+		.channels_max = 8,
+		.rates = SNDRV_PCM_RATE_CONTINUOUS,
+		.rate_min = 32000,
+		.rate_max = 192000,
+		.formats = I2S_FORMATS,
+	},
+	.ops = &sun8i_i2s_dai_ops,
+};
+
+static const struct snd_soc_component_driver i2s_component = {
+	.name = DRV_NAME,
+};
+
+/* --- dma --- */
+
+static const struct snd_pcm_hardware sun8i_i2s_pcm_hardware = {
+	.info = SNDRV_PCM_INFO_INTERLEAVED | SNDRV_PCM_INFO_BLOCK_TRANSFER |
+		SNDRV_PCM_INFO_MMAP | SNDRV_PCM_INFO_MMAP_VALID |
+		SNDRV_PCM_INFO_PAUSE | SNDRV_PCM_INFO_RESUME,
+	.formats = I2S_FORMATS,
+	.rates = SNDRV_PCM_RATE_CONTINUOUS,
+	.rate_min = 32000,
+	.rate_max = 192000,
+	.channels_min = 1,
+	.channels_max = 8,
+	.buffer_bytes_max = 1024 * 1024,
+	.period_bytes_min = 156,
+	.period_bytes_max = 1024 * 1024,
+	.periods_min = 1,
+	.periods_max = 8,
+	.fifo_size = 128,
+};
+
+static const struct snd_dmaengine_pcm_config sun8i_i2s_config = {
+	.prepare_slave_config = snd_dmaengine_pcm_prepare_slave_config,
+	.pcm_hardware = &sun8i_i2s_pcm_hardware,
+	.prealloc_buffer_size = 1024 * 1024,
+};
+
+/* --- audio card --- */
+
+static struct device_node *sun8i_get_codec(struct device *dev)
+{
+	struct device_node *ep, *remote;
+
+	ep = of_graph_get_next_endpoint(dev->of_node, NULL);
+	if (!ep)
+		return NULL;
+	remote = of_graph_get_remote_port_parent(ep);
+	of_node_put(ep);
+
+	return remote;
+}
+
+static int sun8i_card_create(struct device *dev, struct priv *priv)
+{
+	struct snd_soc_card *card;
+	struct snd_soc_dai_link *dai_link;
+	struct snd_soc_dai_link_component *codec;
+
+	card = devm_kzalloc(dev, sizeof(*card), GFP_KERNEL);
+	if (!card)
+		return -ENOMEM;
+	dai_link = devm_kzalloc(dev, sizeof(*dai_link), GFP_KERNEL);
+	if (!dai_link)
+		return -ENOMEM;
+	codec = devm_kzalloc(dev, sizeof(*codec), GFP_KERNEL);
+	if (!codec)
+		return -ENOMEM;
+
+	codec->of_node = sun8i_get_codec(dev);
+	if (!codec->of_node) {
+		dev_err(dev, "no port node\n");
+		return -ENXIO;
+	}
+
+	card->name = codec->of_node->name;
+	card->dai_link = dai_link;
+	card->num_links = 1;
+	dai_link->name = codec->of_node->name;
+	dai_link->stream_name = codec->of_node->name;
+	dai_link->platform_name = dev_name(dev);
+	dai_link->cpu_name = dev_name(dev);
+
+	dai_link->codecs = codec;
+	dai_link->num_codecs = 1;
+
+	/* the DAI name must be the name of codec node */
+	codec->dai_name = codec->of_node->name;
+
+	card->dev = dev;
+	dev_set_drvdata(dev, card);
+	snd_soc_card_set_drvdata(card, priv);
+
+	return devm_snd_soc_register_card(dev, card);
+}
+
+/* --- module init --- */
+
+static int sun8i_i2s_dev_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct priv *priv;
+	struct resource *mem;
+	int ret;
+
+	if (!dev->of_node) {
+		dev_err(dev, "no DT!\n");
+		return -EINVAL;
+	}
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	/* get the resources */
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	priv->mmio = devm_ioremap_resource(dev, mem);
+	if (IS_ERR(priv->mmio))
+		return PTR_ERR(priv->mmio);
+
+	/* get SoC type */
+	priv->type = (int) of_match_device(sun8i_i2s_of_match,
+						&pdev->dev)->data;
+
+	/* get and enable the clocks */
+	priv->gate = devm_clk_get(dev, "apb");	/* optional */
+	priv->clk = devm_clk_get(dev, "mod");
+	if (IS_ERR(priv->clk)) {
+		dev_err(dev, "no clock\n");
+		return PTR_ERR(priv->clk);
+	}
+	ret = clk_set_rate(priv->clk, 24576000);
+	if (ret) {
+		dev_err(dev, "cannot set rate of i2s clock %d\n", ret);
+		return ret;
+	}
+
+	priv->rstc = devm_reset_control_get_optional(dev, NULL);
+	if (!IS_ERR(priv->rstc)) {
+		ret = reset_control_deassert(priv->rstc);
+		if (ret < 0)
+			return ret;
+	}
+
+	if (!IS_ERR(priv->gate)) {
+		ret = clk_prepare_enable(priv->gate);
+		if (ret < 0)
+			goto err_gate;
+	}
+
+	ret = clk_prepare_enable(priv->clk);
+	if (ret < 0)
+		goto err_enable;
+
+	/* activate the audio subsystem */
+	sun8i_i2s_init(priv);
+
+	ret = devm_snd_soc_register_component(dev, &i2s_component,
+						&sun8i_i2s_dai, 1);
+	if (ret) {
+		dev_err(dev, "snd_soc_register_component failed %d\n", ret);
+		goto err_register;
+	}
+
+	ret = devm_snd_dmaengine_pcm_register(dev, &sun8i_i2s_config, 0);
+	if (ret) {
+		dev_err(dev, "pcm_register failed %d\n", ret);
+		goto err_register;
+	}
+
+	priv->dma_data.maxburst = priv->type == SOC_A83T ? 8 : 4;
+	priv->dma_data.addr = mem->start + I2S_TXFIFO;
+	priv->dma_data.addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+
+	ret = sun8i_card_create(dev, priv);
+	if (ret) {
+		if (ret != -EPROBE_DEFER)
+			dev_err(dev, "register card failed %d\n", ret);
+		goto err_register;
+	}
+
+	return 0;
+
+err_register:
+	clk_disable_unprepare(priv->clk);
+err_enable:
+	clk_disable_unprepare(priv->gate);
+err_gate:
+	if (!IS_ERR(priv->rstc))
+		reset_control_assert(priv->rstc);
+
+	return ret;
+}
+
+static int sun8i_i2s_dev_remove(struct platform_device *pdev)
+{
+	struct snd_soc_card *card = dev_get_drvdata(&pdev->dev);
+	struct priv *priv = snd_soc_card_get_drvdata(card); 
+
+	clk_disable_unprepare(priv->clk);
+	clk_disable_unprepare(priv->gate);
+	if (!IS_ERR_OR_NULL(priv->rstc))
+		reset_control_assert(priv->rstc);
+
+	return 0;
+}
+
+static struct platform_driver sun8i_i2s_driver = {
+	.probe  = sun8i_i2s_dev_probe,
+	.remove = sun8i_i2s_dev_remove,
+	.driver = {
+		.name = DRV_NAME,
+		.of_match_table = of_match_ptr(sun8i_i2s_of_match),
+	},
+};
+
+module_platform_driver(sun8i_i2s_driver);
+
+MODULE_AUTHOR("Jean-Francois Moine <moinejf@free.fr>");
+MODULE_DESCRIPTION("Allwinner sun8i I2S ASoC Interface");
+MODULE_LICENSE("GPL v2");