Re: [PATCH v4 12/22] [media] em28xx: properly implement AC97 wait code

["Frank Schäfer" <fschaefer.oss@googlemail.com>]

Am 05.01.2014 16:44, schrieb Mauro Carvalho Chehab:
> Em Sun, 05 Jan 2014 11:20:02 -0200
> Mauro Carvalho Chehab <m.chehab@samsung.com> escreveu:
>
>> Em Sun, 05 Jan 2014 12:19:41 +0100
>> Frank Schäfer <fschaefer.oss@googlemail.com> escreveu:
>>
>>> Am 04.01.2014 11:55, schrieb Mauro Carvalho Chehab:
>>>> Instead of assuming that msleep() is precise, use a jiffies
>>>> based code to wait for AC97 to be available.
>>>>
>>>> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
>>>> ---
>>>>  drivers/media/usb/em28xx/em28xx-core.c | 7 +++++--
>>>>  drivers/media/usb/em28xx/em28xx.h      | 5 ++++-
>>>>  2 files changed, 9 insertions(+), 3 deletions(-)
>>>>
>>>> diff --git a/drivers/media/usb/em28xx/em28xx-core.c b/drivers/media/usb/em28xx/em28xx-core.c
>>>> index 818248d3fd28..36b2f1ab4474 100644
>>>> --- a/drivers/media/usb/em28xx/em28xx-core.c
>>>> +++ b/drivers/media/usb/em28xx/em28xx-core.c
>>>> @@ -23,6 +23,7 @@
>>>>   */
>>>>  
>>>>  #include <linux/init.h>
>>>> +#include <linux/jiffies.h>
>>>>  #include <linux/list.h>
>>>>  #include <linux/module.h>
>>>>  #include <linux/slab.h>
>>>> @@ -254,16 +255,18 @@ EXPORT_SYMBOL_GPL(em28xx_toggle_reg_bits);
>>>>   */
>>>>  static int em28xx_is_ac97_ready(struct em28xx *dev)
>>>>  {
>>>> -	int ret, i;
>>>> +	unsigned long timeout = jiffies + msecs_to_jiffies(EM2800_AC97_XFER_TIMEOUT);
>>>> +	int ret;
>>>>  
>>>>  	/* Wait up to 50 ms for AC97 command to complete */
>>>> -	for (i = 0; i < 10; i++, msleep(5)) {
>>>> +	while (time_is_after_jiffies(timeout)) {
>>> time_is_before_jiffies(timeout)
>>>
>>>>  		ret = em28xx_read_reg(dev, EM28XX_R43_AC97BUSY);
>>>>  		if (ret < 0)
>>>>  			return ret;
>>>>  
>>>>  		if (!(ret & 0x01))
>>>>  			return 0;
>>>> +		msleep (5);
>>>>  	}
>>>>  
>>>>  	em28xx_warn("AC97 command still being executed: not handled properly!\n");
>>>> diff --git a/drivers/media/usb/em28xx/em28xx.h b/drivers/media/usb/em28xx/em28xx.h
>>>> index 9d6f43e4681f..ac79501f5d9f 100644
>>>> --- a/drivers/media/usb/em28xx/em28xx.h
>>>> +++ b/drivers/media/usb/em28xx/em28xx.h
>>>> @@ -182,9 +182,12 @@
>>>>  
>>>>  #define EM28XX_INTERLACED_DEFAULT 1
>>>>  
>>>> -/* time in msecs to wait for i2c writes to finish */
>>>> +/* time in msecs to wait for i2c xfers to finish */
>>>>  #define EM2800_I2C_XFER_TIMEOUT		20
>>>>  
>>>> +/* time in msecs to wait for AC97 xfers to finish */
>>>> +#define EM2800_AC97_XFER_TIMEOUT	100
>>>> +
>>> I applies to all chips supporting AC97 audio, so call it
>>> EM28XX_AC97_XFER_TIMEOUT.
>> Ok.
>>
>>> Why did you increase the timeout from 50ms to 100ms ?
>>> 50ms already seems to be a lot !
> Err... actually, the code is currently waiting for up to 100ms.
>
> The thing is that msleep(5) will wait for a minimum time of 1/CONFIG_HZ.
>
> If CONFIG_HZ is equal to 100 (the minimum), that means that it will
> wait for 10 ms. So,
>
> for (i = 0; i < 10; i++, msleep(5)) {
> 	...
> }
>
> Will actually wait up to 10*10ms = 100ms at the worse case. 
>
> The patch is just keeping the maximum timeout equal for all setups, no
> matter what's set at CONFIG_HZ.
>
> Arbitrarily changing it to a lower value seems risky, as it may cause
> regressions on some boards.
>
> So, I'll just rename the define here, preserving 100ms as timeout.
>
> Regards,
> Mauro
Ok, then it makes sense.
I'm pretty sure 100ms is much more than needed and it seems to work for
those people using a CONFIG_HZ value higher 100.
But for now let's keep it as.

Reviewed-by: Frank Schäfer <fschaefer.oss@googlemail.com>