From mboxrd@z Thu Jan 1 00:00:00 1970 From: eric@eukrea.com (=?ISO-8859-15?Q?Eric_B=E9nard?=) Date: Fri, 28 May 2010 11:17:22 +0200 Subject: [PATCH v2 3/4] eukrea_mbimx27: add audio codec In-Reply-To: <20100528065840.GF23664@pengutronix.de> References: <1274950737-26498-1-git-send-email-eric@eukrea.com> <1274950737-26498-2-git-send-email-eric@eukrea.com> <1274950737-26498-3-git-send-email-eric@eukrea.com> <20100528065840.GF23664@pengutronix.de> Message-ID: <4BFF8A22.40704@eukrea.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Hi Sascha, Le 28/05/2010 08:58, Sascha Hauer a ?crit : > Hi Eric, > > The patch does not apply on current master. Can you please rebase it? > It's better we delay the defconfig patch until the code is merged, so > give me a ping when it's done. > in fact it's based on top of my patchset which updates our board's support : http://lists.infradead.org/pipermail/linux-arm-kernel/2010-May/thread.html#15821 I can rebase it but maybe you can also get most of the patches of this thread ? > > On Thu, May 27, 2010 at 10:58:56AM +0200, Eric B?nard wrote: >> Signed-off-by: Eric B?nard >> Cc: Sascha Hauer >> Cc: Mark Brown >> Cc: Liam Girdwood >> --- >> arch/arm/mach-mx2/eukrea_mbimx27-baseboard.c | 44 +++++++++++++++++++++++++- >> 1 files changed, 43 insertions(+), 1 deletions(-) >> >> diff --git a/arch/arm/mach-mx2/eukrea_mbimx27-baseboard.c b/arch/arm/mach-mx2/eukrea_mbimx27-baseboard.c >> index 9038e1f..e7fd65a 100644 >> --- a/arch/arm/mach-mx2/eukrea_mbimx27-baseboard.c >> +++ b/arch/arm/mach-mx2/eukrea_mbimx27-baseboard.c >> @@ -1,5 +1,5 @@ >> /* >> - * Copyright (C) 2009 Eric Benard - eric at eukrea.com >> + * Copyright (C) 2009-2010 Eric Benard - eric at eukrea.com >> * >> * Based on pcm970-baseboard.c which is : >> * Copyright (C) 2008 Juergen Beisert (kernel at pengutronix.de) >> @@ -37,6 +37,8 @@ >> #include >> #include >> #include >> +#include >> +#include >> >> #include "devices.h" >> >> @@ -93,6 +95,13 @@ static int eukrea_mbimx27_pins[] = { >> PD29_PF_CSPI1_SCLK, >> PD30_PF_CSPI1_MISO, >> PD31_PF_CSPI1_MOSI, >> + /* SSI4 */ >> +#if defined(CONFIG_SND_SOC_EUKREA_TLV320) >> + PC16_PF_SSI4_FS, >> + PC17_PF_SSI4_RXD | GPIO_PUEN, >> + PC18_PF_SSI4_TXD | GPIO_PUEN, >> + PC19_PF_SSI4_CLK, >> +#endif > > You should remove the ifdefs. The pin muxer should not depend on compile > time options. > well on this kind of board yes as the IO can be used as plain GPIO if the codec is not mounted. >> }; >> >> static const uint32_t eukrea_mbimx27_keymap[] = { >> @@ -292,6 +301,12 @@ static struct spi_board_info eukrea_mbimx27_spi_board_info[] __initdata = { >> }, >> }; >> >> +static struct i2c_board_info eukrea_mbimx27_i2c_devices[] = { >> + { >> + I2C_BOARD_INFO("tlv320aic23", 0x1a), >> + }, >> +}; >> + >> static int eukrea_mbimx27_spi_cs[] = {GPIO_PORTD | 28}; >> >> static struct spi_imx_master eukrea_mbimx27_spi_0_data = { >> @@ -308,6 +323,10 @@ static struct imxmmc_platform_data sdhc_pdata = { >> .dat3_card_detect = 1, >> }; >> >> +struct imx_ssi_platform_data eukrea_mbimx27_ssi_pdata = { >> + .flags = IMX_SSI_DMA | IMX_SSI_USE_I2S_SLAVE, >> +}; >> + >> /* >> * system init for baseboard usage. Will be called by cpuimx27 init. >> * >> @@ -319,6 +338,24 @@ void __init eukrea_mbimx27_baseboard_init(void) >> mxc_gpio_setup_multiple_pins(eukrea_mbimx27_pins, >> ARRAY_SIZE(eukrea_mbimx27_pins), "MBIMX27"); >> >> +#if defined(CONFIG_SND_SOC_EUKREA_TLV320) >> + /* SSI unit master I2S codec connected to SSI_PINS_4*/ >> + mxc_audmux_v1_configure_port(MX27_AUDMUX_HPCR1_SSI0, >> + MXC_AUDMUX_V1_PCR_SYN | >> + MXC_AUDMUX_V1_PCR_TFSDIR | >> + MXC_AUDMUX_V1_PCR_TCLKDIR | >> + MXC_AUDMUX_V1_PCR_RFSDIR | >> + MXC_AUDMUX_V1_PCR_RCLKDIR | >> + MXC_AUDMUX_V1_PCR_TFCSEL(MX27_AUDMUX_HPCR3_SSI_PINS_4) | >> + MXC_AUDMUX_V1_PCR_RFCSEL(MX27_AUDMUX_HPCR3_SSI_PINS_4) | >> + MXC_AUDMUX_V1_PCR_RXDSEL(MX27_AUDMUX_HPCR3_SSI_PINS_4) >> + ); >> + mxc_audmux_v1_configure_port(MX27_AUDMUX_HPCR3_SSI_PINS_4, >> + MXC_AUDMUX_V1_PCR_SYN | >> + MXC_AUDMUX_V1_PCR_RXDSEL(MX27_AUDMUX_HPCR1_SSI0) >> + ); >> +#endif > > Does it hurt to go without the ifdefs? > same thing as for the IOMUX. Thanks, Eric