[PATCH 2/2] ARM: dts: mx28: Fix sgtl5000 codec probe

[Fabio Estevam]

On Mon, Jun 24, 2013 at 11:07 AM, Shawn Guo <shawn.guo at linaro.org> wrote:
> On Mon, Jun 24, 2013 at 10:54:11AM -0300, Fabio Estevam wrote:
>> On Mon, Jun 24, 2013 at 4:29 AM, Shawn Guo <shawn.guo at linaro.org> wrote:
>>
>> > Yes, we should have sgtl5000 acquire its clock, but when it fails to do
>> > so, there is nothing wrong with deferring it to see if mxs-saif would
>> > provide the clock later.
>>
>> I didn't manage to get mx28evk to probe audio correctly by using -EPROBE_DEFER.
>
> When you was doing that, did you have the clock that sgtl5000 is
> requesting registered somewhere?

I registered it as dummy one (same as done in this patch).

Then regmap_read fails (from sgtl5000 driver), as it tries to read the
codec revision via I2C and the clock is not available yet. Then I
changed to return a -EDEFER_PROBE in case of failure:

[    1.754258] mxs-lradc 80050000.lradc: Touchscreen not enabled.
[    1.783755] ******* reading I2C failed
[    1.789748] i2c 0-000a: Driver sgtl5000 requests probe deferral
[    1.801720] mxs-sgtl5000 sound.12: ASoC: CODEC (null) not registered
[    1.809608] mxs-sgtl5000 sound.12: snd_soc_register_card failed (-517)
[    1.816675] platform sound.12: Driver mxs-sgtl5000 requests probe deferral
[    1.825852] TCP: cubic registered
[    1.829375] NET: Registered protocol family 17
[    1.835999] Key type dns_resolver registered
[    1.844549] mmc0: new SDHC card at address f955
[    1.852887] registered taskstats version 1
[    1.860535] mmcblk0: mmc0:f955 SD04G 3.69 GiB
[    1.878653] sgtl5000 0-000a: sgtl5000 revision 0x11
[    1.887430]  mmcblk0: p1 p2 p3
[    1.894824] mxs-sgtl5000 sound.12: ASoC: CODEC (null) not registered
[    1.901468] mxs-sgtl5000 sound.12: snd_soc_register_card failed (-517)

Then it succeeds in the following attemp, but snd_soc_register_card() fails.