Peter Ujfalusi wrote:
On Tuesday 06 October 2009 10:29:39 ext Lopez Cruz, Misael wrote:
In order to have TWL6030 CODEC driver as a platform driver, codec data should be passed through twl_platform_data structure.
For twl6030 audio codec, the following data may be passed:
- audpwron_gpio: gpio line used to power-up/down the codec. A low-to-high transition powers codec up. Setting audpwron_gpio to a negative value means that codec will use manual power sequence
instead of automatic sequence - naudint_irq: irq line for audio interrupt. twl6030 drives NAUDINT line to low when an interrupt (codec ready, plug insertion/removal, etc) is detected
However, codec driver can operate if any or none of them are passed.
How does the twl4030 series would fit with this modification? I have also added the twl4030 codec as a child for the twl MFD, but I have not finished the cleanup of the codec fully.
twl6030 codec data won't conflict with twl4030's as it's dependent of TWL6030_CORE. We can add parallely the required data for twl4030 codec.
Signed-off-by: Misael Lopez Cruz x0052729@ti.com
drivers/mfd/twl-core.c | 15 +++++++++++++++ include/linux/i2c/twl.h | 7 +++++++ 2 files changed, 22 insertions(+), 0 deletions(-)
diff --git a/drivers/mfd/twl-core.c b/drivers/mfd/twl-core.c index af4cf47..6432af1 100644 --- a/drivers/mfd/twl-core.c +++ b/drivers/mfd/twl-core.c @@ -108,6 +108,12 @@ #define twl_has_mmc() false #endif
+#if defined(CONFIG_SND_SOC_TWL6030) || defined(CONFIG_SND_SOC_TWL6030_MODULE) +#define twl_has_codec() true +#else +#define twl_has_codec() false +#endif
#if defined(CONFIG_TWL4030_USB) ||
defined(CONFIG_TWL4030_USB_MODULE)
#define twl_has_usb() true #else @@ -190,6 +196,7 @@ #define BCI_SUB_CHIP_ID SUB_CHIP_ID1 #define GPIO_SUB_CHIP_ID 0 /* NOT SUPPORTED IN TWL6030 */ #define KEYPAD_SUB_CHIP_ID 0 /* ADDED FOR COMPILATION ONLY */ +#define CODEC_SUB_CHIP_ID SUB_CHIP_ID3
TWL4030 codec is under ID2 address group.
CODEC_SUB_CHIP_ID for twl6030 is under #ifdef CONFIG_TWL6030_CORE. There are a different set of _CHIP_IDs for twl4030 children, which doesn't have a definition for CODEC yet.
/* subchip/slave 0 0x48 - POWER */ #define TWL6030_BASEADD_RTC 0x0000 @@ -632,6 +639,14 @@ add_children(struct twl_platform_data *pdata, unsigned long features) if (IS_ERR(child)) return PTR_ERR(child); }
- if (twl_has_codec()) {
child = add_child(CODEC_SUB_CHIP_ID, "twl6030_codec",pdata->codec, sizeof(*pdata->codec), false,0, 0);if (IS_ERR(child))return PTR_ERR(child);- }
We are going to have the twl4030 as child as well.
#endif
if (twl_has_usb() && pdata->usb) { diff --git a/include/linux/i2c/twl.h b/include/linux/i2c/twl.h index b687a8b..e76ca9b 100644 --- a/include/linux/i2c/twl.h +++ b/include/linux/i2c/twl.h @@ -472,6 +472,12 @@ struct twl_usb_data { enum twl_usb_mode usb_mode; };
+struct twl_codec_data {
- /* twl6030 */
- int audpwron_gpio; /* audio power-on gpio */
- int naudint_irq; /* audio interrupt */
+};
These are not applicable for the twl4030 codec, should we have different twl_codec_data for twl6030 and twl4030 series?
Currently, the support for TWL4030 and TWL6030 makes them mutually exclusive, TWL6030_CORE depends on !TWL4030_CORE. So we can have separate twl_codec_data definitions to avoid unused space in the structure.
Should I merge here the things which will be used for the twl4030 codec?
I think merging them would suit only if both twl versions can be enabled at the same time, but not sure it that makes sense.
At the moment, I only have audio_mclk in my codec_data, but most probably the HS ramp related things will be also merged here in the future.
struct twl_platform_data { unsigned irq_base, irq_end; struct twl_bci_platform_data *bci; @@ -479,6 +485,7 @@ struct twl_platform_data { struct twl_madc_platform_data *madc; struct twl_keypad_data *keypad; struct twl_usb_data *usb;
struct twl_codec_data *codec;
/* LDO regulators common to TWL4030/TWL6030 */ struct regulator_init_data *vdac;