There are several issues on the current code: 1) msleep(1) is actually equivalent to msleep(10); 2) au8522_video_set() will set reg 0xa4 to the proper value for SIF, CVBS or S-Video. No need to force it to CVBS; 3) Let's not hardcode 0x9d for CBS on audio_set. Signed-off-by: Mauro Carvalho Chehab <m.chehab@xxxxxxxxxxx> --- drivers/media/dvb-frontends/au8522_decoder.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/media/dvb-frontends/au8522_decoder.c b/drivers/media/dvb-frontends/au8522_decoder.c index 569922232eb8..b971c20624bf 100644 --- a/drivers/media/dvb-frontends/au8522_decoder.c +++ b/drivers/media/dvb-frontends/au8522_decoder.c @@ -458,8 +458,9 @@ static void set_audio_input(struct au8522_state *state) au8522_writereg(state, AU8522_I2C_CONTROL_REG0_REG090H, 0x84); msleep(150); au8522_writereg(state, AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H, 0x00); - msleep(1); - au8522_writereg(state, AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H, 0x9d); + msleep(10); + au8522_writereg(state, AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H, + AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H_CVBS); msleep(50); au8522_writereg(state, AU8522_AUDIO_VOLUME_L_REG0F2H, 0x7F); au8522_writereg(state, AU8522_AUDIO_VOLUME_R_REG0F3H, 0x7F); @@ -585,12 +586,9 @@ static int au8522_s_stream(struct v4l2_subdev *sd, int enable) au8522_writereg(state, AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H, 0x01); - msleep(1); - au8522_writereg(state, AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H, - AU8522_SYSTEM_MODULE_CONTROL_0_REG0A4H_CVBS); + msleep(10); au8522_video_set(state); - set_audio_input(state); state->operational_mode = AU8522_ANALOG_MODE; -- 1.9.3 -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
