From: xujianqun <[email protected]>
Since hardware may not MUST to use IRQ pin of max98090 as jack detect, the
driver can work well without it, can report jack trigger to CPU by a GPIO.
But here driver will register fail caused by failed to request irq.
Signed-off-by: xujianqun <[email protected]>
---
sound/soc/codecs/max98090.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c
index 566919c..bc124ff 100644
--- a/sound/soc/codecs/max98090.c
+++ b/sound/soc/codecs/max98090.c
@@ -2515,6 +2515,8 @@ static int max98090_probe(struct snd_soc_codec *codec)
max98090_add_widgets(codec);
+ return 0;
+
err_access:
return ret;
}
--
1.7.9.5
On 07/30/2014 04:50 AM, jianqun wrote:
> From: xujianqun <[email protected]>
>
> Since hardware may not MUST to use IRQ pin of max98090 as jack detect, the
> driver can work well without it, can report jack trigger to CPU by a GPIO.
>
> But here driver will register fail caused by failed to request irq.
>
> Signed-off-by: xujianqun <[email protected]>
> ---
> sound/soc/codecs/max98090.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c
> index 566919c..bc124ff 100644
> --- a/sound/soc/codecs/max98090.c
> +++ b/sound/soc/codecs/max98090.c
> @@ -2515,6 +2515,8 @@ static int max98090_probe(struct snd_soc_codec *codec)
>
> max98090_add_widgets(codec);
>
> + return 0;
> +
> err_access:
> return ret;
> }
I would say it's better to call request_threaded_irq() conditionally
when max98090->irq is valid in the same max98090_probe(). Also code
should return instantly in case of request_threaded_irq() fails for
valid irq.
Now code is still printing needless error message in case max98090->irq
is invalid or not specified and continue probing in case
request_threaded_irq() fails for valid irq.
--
Jarkko
From: Jianqun <[email protected]>
Since IRQ pin from max98090 may NC, the irq number will be zero, that is
invalid for request_threaded_irq, so just add irq valid check there.
Since hardware may not MUST to use IRQ pin of max98090 as jack detect, the
driver can work well without it, can report jack trigger to CPU by a GPIO.
But here driver will register fail caused by failed to request irq.
Signed-off-by: Jianqun <[email protected]>
---
sound/soc/codecs/max98090.c | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c
index 566919c..9dc0e8c 100644
--- a/sound/soc/codecs/max98090.c
+++ b/sound/soc/codecs/max98090.c
@@ -2478,12 +2478,14 @@ static int max98090_probe(struct snd_soc_codec *codec)
/* Register for interrupts */
dev_dbg(codec->dev, "irq = %d\n", max98090->irq);
- ret = request_threaded_irq(max98090->irq, NULL,
- max98090_interrupt, IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
- "max98090_interrupt", codec);
- if (ret < 0) {
- dev_err(codec->dev, "request_irq failed: %d\n",
- ret);
+ if (max98090->irq) {
+ ret = request_threaded_irq(max98090->irq, NULL,
+ max98090_interrupt, IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
+ "max98090_interrupt", codec);
+ if (ret < 0) {
+ dev_err(codec->dev, "request_irq failed: %d\n",
+ ret);
+ }
}
/*
--
1.7.9.5
On 2014年07月30日 14:29, Jarkko Nikula wrote:
> On 07/30/2014 04:50 AM, jianqun wrote:
>> From: xujianqun <[email protected]>
>>
>> Since hardware may not MUST to use IRQ pin of max98090 as jack detect, the
>> driver can work well without it, can report jack trigger to CPU by a GPIO.
>>
>> But here driver will register fail caused by failed to request irq.
>>
>> Signed-off-by: xujianqun <[email protected]>
>> ---
>> sound/soc/codecs/max98090.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c
>> index 566919c..bc124ff 100644
>> --- a/sound/soc/codecs/max98090.c
>> +++ b/sound/soc/codecs/max98090.c
>> @@ -2515,6 +2515,8 @@ static int max98090_probe(struct snd_soc_codec *codec)
>> max98090_add_widgets(codec);
>> + return 0;
>> +
>> err_access:
>> return ret;
>> }
> I would say it's better to call request_threaded_irq() conditionally when max98090->irq is valid in the same max98090_probe(). Also code should return instantly in case of request_threaded_irq() fails for valid irq.
>
> Now code is still printing needless error message in case max98090->irq is invalid or not specified and continue probing in case request_threaded_irq() fails for valid irq.
>
Add a valid check could be better, thanks Jarkko~
I'm working a board with max98090, this is a really problem for me now,
the IRQ shouldn't cause the driver fail to register.
--
-------------
许 剑 群 Jay
Rockchip Electronics Co.Ltd
****************************************************************************
*IMPORTANT NOTICE:*This email is from Fuzhou Rockchip Electronics Co.,
Ltd .The contents of this email and any attachments may contain
information that is privileged, confidential and/or exempt from
disclosure under applicable law and relevant NDA. If you are not the
intended recipient, you are hereby notified that any disclosure,
copying, distribution, or use of the information is STRICTLY PROHIBITED.
Please immediately contact the sender as soon as possible and destroy
the material in its entirety in any format. Thank you.
****************************************************************************
Hi
On 07/30/2014 11:54 AM, jianqun wrote:
> From: Jianqun <[email protected]>
>
> Since IRQ pin from max98090 may NC, the irq number will be zero, that is
> invalid for request_threaded_irq, so just add irq valid check there.
>
> Since hardware may not MUST to use IRQ pin of max98090 as jack detect, the
> driver can work well without it, can report jack trigger to CPU by a GPIO.
>
> But here driver will register fail caused by failed to request irq.
>
> Signed-off-by: Jianqun <[email protected]>
> ---
> sound/soc/codecs/max98090.c | 14 ++++++++------
> 1 file changed, 8 insertions(+), 6 deletions(-)
>
> diff --git a/sound/soc/codecs/max98090.c b/sound/soc/codecs/max98090.c
> index 566919c..9dc0e8c 100644
> --- a/sound/soc/codecs/max98090.c
> +++ b/sound/soc/codecs/max98090.c
> @@ -2478,12 +2478,14 @@ static int max98090_probe(struct snd_soc_codec *codec)
> /* Register for interrupts */
> dev_dbg(codec->dev, "irq = %d\n", max98090->irq);
>
> - ret = request_threaded_irq(max98090->irq, NULL,
> - max98090_interrupt, IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
> - "max98090_interrupt", codec);
> - if (ret < 0) {
> - dev_err(codec->dev, "request_irq failed: %d\n",
> - ret);
> + if (max98090->irq) {
> + ret = request_threaded_irq(max98090->irq, NULL,
> + max98090_interrupt, IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
> + "max98090_interrupt", codec);
> + if (ret < 0) {
> + dev_err(codec->dev, "request_irq failed: %d\n",
> + ret);
> + }
> }
>
Remember to add also "return ret;" at the end of if () {} in case
request_threaded_irq() fails since code should not continue probing now
after adding a test for valid irq.
Should the if (max98090->irq) test be >0? I'm not sure is there
possibility that can i2c->irq actually pass both -1 and 0 depending how
struct i2c_board_info etc are initialized?
I forgot that I have a fix changing request_threaded_irq() to
devm_request_threaded_irq() so you should redo your patch on top of
for-next branch of
git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git.
--
Jarkko