Hi, ok, understood this problem. So, why not set this in si470x_fops_open directly after the si470x_start? It seems more appropriate to enable the RDS interrupt after starting the radio. Bye the way, you pointed me to a bug. Instead of always setting de-emphasis in si470x_start: radio->registers[SYSCONFIG1] = SYSCONFIG1_DE; This should only be done, if requested by module parameter: radio->registers[SYSCONFIG1] = (de << 11) & SYSCONFIG1_DE; /* DE */ Bye, Toby Am Mittwoch 02 Dezember 2009 00:54:15 schrieb Joonyoung Shim: > Hi, Tobias. > > On 12/2/2009 8:39 AM, Tobias Lorenz wrote: > > Hi, > > > > what is the advantage in not setting SYSCONFIG1 into a known state? > > > > At patch 3/3, i am setting the SYSCONFIG1 register for RDS interrupt in > i2c probe function, so i need this patch. Do you have other idea? > > > Bye, > > Toby > > > > Am Mittwoch 18 November 2009 07:21:25 schrieb Joonyoung Shim: > >> We should use the or operation to set value to the SYSCONFIG1 register > >> on si470x_start(). > >> > >> Signed-off-by: Joonyoung Shim <jy0922.shim@xxxxxxxxxxx> > >> --- > >> drivers/media/radio/si470x/radio-si470x-common.c | 2 +- > >> 1 files changed, 1 insertions(+), 1 deletions(-) > >> > >> diff --git a/drivers/media/radio/si470x/radio-si470x-common.c b/drivers/media/radio/si470x/radio-si470x-common.c > >> index f33315f..09f631a 100644 > >> --- a/drivers/media/radio/si470x/radio-si470x-common.c > >> +++ b/drivers/media/radio/si470x/radio-si470x-common.c > >> @@ -357,7 +357,7 @@ int si470x_start(struct si470x_device *radio) > >> goto done; > >> > >> /* sysconfig 1 */ > >> - radio->registers[SYSCONFIG1] = SYSCONFIG1_DE; > >> + radio->registers[SYSCONFIG1] |= SYSCONFIG1_DE; > >> retval = si470x_set_register(radio, SYSCONFIG1); > >> if (retval < 0) > >> goto done; > >> > > > -- 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
