[alsa-devel] [patch] ASoC: adau1701: signedness bug in adau1701_write()
"ret" is supposed to be signed here. The current code will only return -EIO on error, instead of a more appropriate error code such as -EAGAIN etc.
Signed-off-by: Dan Carpenter error27@gmail.com
diff --git a/sound/soc/codecs/adau1701.c b/sound/soc/codecs/adau1701.c index 6c01bb6..2758d5f 100644 --- a/sound/soc/codecs/adau1701.c +++ b/sound/soc/codecs/adau1701.c @@ -140,9 +140,10 @@ static unsigned int adau1701_register_size(struct snd_soc_codec *codec, static int adau1701_write(struct snd_soc_codec *codec, unsigned int reg, unsigned int value) { - unsigned int i, ret; + unsigned int i; unsigned int size; uint8_t buf[4]; + int ret;
size = adau1701_register_size(codec, reg); if (size == 0)
On 06/20/2011 09:11 AM, Dan Carpenter wrote:
"ret" is supposed to be signed here. The current code will only return -EIO on error, instead of a more appropriate error code such as -EAGAIN etc.
Signed-off-by: Dan Carpenter error27@gmail.com
Acked-by: Lars-Peter Clausen lars@metafoo.de
Thanks
diff --git a/sound/soc/codecs/adau1701.c b/sound/soc/codecs/adau1701.c index 6c01bb6..2758d5f 100644 --- a/sound/soc/codecs/adau1701.c +++ b/sound/soc/codecs/adau1701.c @@ -140,9 +140,10 @@ static unsigned int adau1701_register_size(struct snd_soc_codec *codec, static int adau1701_write(struct snd_soc_codec *codec, unsigned int reg, unsigned int value) {
- unsigned int i, ret;
unsigned int i; unsigned int size; uint8_t buf[4];
int ret;
size = adau1701_register_size(codec, reg); if (size == 0)
On 20/06/11 11:03, Lars-Peter Clausen wrote:
On 06/20/2011 09:11 AM, Dan Carpenter wrote:
"ret" is supposed to be signed here. The current code will only return -EIO on error, instead of a more appropriate error code such as -EAGAIN etc.
Signed-off-by: Dan Carpenter error27@gmail.com
Acked-by: Lars-Peter Clausen lars@metafoo.de
Acked-by: Liam Girdwood lrg@ti.com
On Mon, Jun 20, 2011 at 10:11:25AM +0300, Dan Carpenter wrote:
"ret" is supposed to be signed here. The current code will only return -EIO on error, instead of a more appropriate error code such as -EAGAIN etc.
That changelog is not accurate, if nothing else the code will return -EINVAL if it can't figure out a length for a register. However, the change itself is good so applied, thanks.
participants (4)
-
Dan Carpenter -
Lars-Peter Clausen -
Liam Girdwood -
Mark Brown