[PATCH] staging: brcmfmac: Consolidate debugging macros
zajec5 at gmail.com
Sat Aug 27 00:29:23 PDT 2011
W dniu 26 sierpnia 2011 11:24 użytkownik Arend van Spriel
<arend at broadcom.com> napisał:
> On 08/26/2011 11:19 AM, Rafał Miłecki wrote:
>> W dniu 26 sierpnia 2011 11:02 użytkownik Joe Perches<joe at perches.com>
>>> On Fri, 2011-08-26 at 10:55 +0200, Rafał Miłecki wrote:
>>>> 2011/8/25 Joe Perches<joe at perches.com>:
>>>>> Convert multiple BRCMF_<debug_level> macros to a single
>>>>> brcmf_dbg(level, fmt, ...) macro.
>>>> I'm not sure if that "_dbg" suffix is a really good choice. You use
>>>> "_dbg" for all the kinds of messages, while "DEBUG" is already one
>>>> kind of messages.
>>> No, not really. These _are_ debugging uses.
>>> The #include is called dhd_dbg.h and
>>> the whole block is guarded by
>>> #if defined(BCMDBG)
>>>> We have also other types like EMERG, ALERT, CRIT,
>>>> ERR, WARNING, NOTICE, INFO and DEBUG.
>>> Except for the last, those aren't debugging uses,
>>> these are.
>> It makes sense, maybe some messages should be just always printed. It
>> seems some of they may be important for common-user having problem
>> with his card. Just few random ones:
>> brcmf_dbg(ERROR, "device attach failed\n");
>> brcmf_dbg(ERROR, "Failed to enable F1 Err: 0x%08x\n", err_ret);
>> brcmf_dbg(ERROR, "MACEVENT: %s [unsupported version --> brcmf
>> version:%d dongle version:%d]\n",
>> brcmf_dbg(ERROR, "mismatched OUI, bailing\n");
>> brcmf_dbg(ERROR, "brcmf_net_attach failed, err %d\n",
>> brcmf_dbg(ERROR, "failed to bring up cfg80211\n");
>> brcmf_dbg(ERROR, "HT Avail request error: %d\n", err);
> Hi Rafał,
> The feedback from Greg was that no user would be interested in it so we
> cleaned it up and put it under Kconfig option.
Having messages like "Entered xxx function" are really useless for
common user, but I believe errors should be printed, to let use known
sth is wrong.
Did Greg really mean all that messages? Including error ones? Greg?
More information about the devel