Messages in this thread | | | Date | Wed, 7 May 2014 14:19:56 -0400 | From | Steven Rostedt <> | Subject | Re: [PATCH] plist: replace pr_debug with printk in plist_test() |
| |
On Wed, 07 May 2014 11:10:38 -0700 Joe Perches <joe@perches.com> wrote:
> On Wed, 2014-05-07 at 10:35 -0400, Steven Rostedt wrote: > > On Wed, 7 May 2014 10:21:28 -0400 Dan Streetman <ddstreet@ieee.org> wrote: > > > > It would be even better if the note could clarify that sometimes it is > > > ok to use printk(KERN_DEBUG > > > > Exactly. I think it's rather stupid to have to do a #define DEBUG to > > have pr_debug() print in general. > > > > I see no reason to have pr_debug() be anything different than the other > > pr_*() functions. > > pr_debug is meant to be disabled and have _no_ runtime > effect unless DEBUG is #defined.
I understand why it does it, but having pr_debug() named just like pr_info(), pr_notice(), pr_warning(), pr_err(), pr_crit(), pr_alert(), pr_emerg(), where all those are just printk(<LOGLEVEL>...) *except* for pr_debug(). That's inconsistent and wrong.
pr_debug() should have been just printk(KERN_DEBUG ...) as that follows convention. Not something that gets disabled by default. For that, we should have given it a different name. That's the point I was trying to make.
Yes, it's somewhat too late as pr_debug() is all over the place, but maybe when things slow down (Ha! like that will ever happen ... "are we done yet?"), then we could do a massive clean up and rename pr_debug() to something not so confusing in its usage compared to the other pr_*() prints.
-- Steve
| |