Re: [PATCH V2 05/11] brcmfmac: error messages should not be suppressed

["Arend van Spriel" <arend@broadcom.com>]

[-- Attachment #1: Type: text/plain, Size: 1599 bytes --]

On 12/06/2012 07:32 AM, Joe Perches wrote:
> On Wed, 2012-12-05 at 22:28 +0100, Arend van Spriel wrote:
>> The call to brcmf_dbg(ERROR, ...) only resulted in a log message
>> when compiled with -DDEBUG. Error messages are valuable for resolving
>> issues so this patch replaces it with brcmf_err(...) so they always
>> end up in the log.
> []
>> Reviewed-by: Joe Perches <joe@perches.com>
> 
> Don't add signatures unless the person actually gives one.
> I didn't.

Sorry about that.

> []
> 
>> diff --git a/drivers/net/wireless/brcm80211/brcmfmac/dhd_dbg.h b/drivers/net/wireless/brcm80211/brcmfmac/dhd_dbg.h
> []
>> @@ -35,20 +34,11 @@
>>  
>>  #if defined(DEBUG)
>>  
>> -#define brcmf_dbg(level, fmt, ...)					\
>> -do {									\
>> -	if (BRCMF_ERROR_VAL == BRCMF_##level##_VAL) {			\
>> -		if (brcmf_msg_level & BRCMF_##level##_VAL) {		\
>> -			if (net_ratelimit())				\
>> -				pr_debug("%s: " fmt,			\
>> -					 __func__, ##__VA_ARGS__);	\
>> -		}							\
>> -	} else {							\
>> -		if (brcmf_msg_level & BRCMF_##level##_VAL) {		\
>> -			pr_debug("%s: " fmt,				\
>> -				 __func__, ##__VA_ARGS__);		\
>> -		}							\
>> -	}								\
>> +#define brcmf_err(fmt, ...)	pr_err("%s: " fmt, __func__, ##__VA_ARGS__)
> 
> You still lost the net_ratelimit() for the DEBUG case.
> 
> This brcmf_err macro should probably be in dhd.h,
> not in a debug header and not guarded by DEBUG at all.

I thought I explained why in my email response to you, but now I don't
see that one on the mailing list archives (attached now). When we are
debugging we do not want error messages to be filtered.

Gr. AvS

[-- Attachment #2: Re: [PATCH 05_11] brcmfmac: error messages should not be suppressed.eml --]
[-- Type: application/x-extension-eml, Size: 1578 bytes --]