On 04/02/2012 03:32 PM, Borislav Petkov wrote:
On Mon, Apr 02, 2012 at 12:45:09PM -0500, Larry Finger wrote:
The current situation is fine with me. Any user with an embedded
system can save a lot of memory by setting CONFIG_RTLWIFI_DEBUG to
n. If a normal distro chooses to do that, then when some user of
that distro reports a problem and is unable to run with the module
parameter "debug=X", I get to say "Sorry, I cannot help you." :)
Of course, if you want to go through the source and convert every
instance of "RT_TRACE(rtlpriv, COMP_ERR, DBG_EMERG, ...)" to the
equivalent "pr_err(...), then I would probably sign off on it.
I see. Well, I suppose when the driver matures, you're probably going
to remove most of the debugging code yourself because you won't need it
anymore and turn important printk statements into pr_err or pr_info or
whatever.
However, the way it is default y now, people who do 'yes "" | make
oldconfig' will be surprised to see a whole lot of debugging output
spewed from the wireless driver upon next reboot, and I don't think you
want that.
I dunno, maybe CONFIG_DYNAMIC_DEBUG could be of help. There's also a
howto:<Documentation/dynamic-debug-howto.txt>
Users will see nothing more than before!
Before CONFIG_RTLWIFI_DEBUG was allowed to be set, the driver always compiled in
the debug code, but only those messages with the DBG_EMERG code were actually
output. With the parameter set to Y, nothing changed. If the module is loaded
with the "debug=X" option, where X is greater than 0, then other messages will
be seen. That dynamic setting of the level is enough.
The only change when the configuration parameter is set to N is that the entire
debug setup compiles away. The driver is a lot smaller, but there is no debugging.
Larry
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
