Re: [PATCH 0/3] dyndbg: dev_dbg bugfix + 2 trivials

[Joe Perches <joe@perches.com>]

On Tue, 2012-07-24 at 15:40 -0600, Jim Cromie wrote:
> On Fri, Jul 20, 2012 at 2:38 PM, Jason Baron <jbaron@redhat.com> wrote:
> > On Thu, Jul 19, 2012 at 01:46:19PM -0600, Jim Cromie wrote:
> >> 3 patches here, 1st is bugfix, others are trivial.
> >>
> >> 1- fix __dev_printk, which broke dev_dbg() prefix under CONFIG_DYNAMIC_DEBUG.

I'd rather what I posted be tested/implemented instead.
It works with the printk changes already in -next via mm.

Try your patch against -next and see it doesn't work.
It requires additional changes for SOH handling.

> >> 2- change dyndbg prefix interfield separator from ':' to '.'

I think this part of the patch completely unnecessary.
What value does changing the prefix from one char to another really add?
Trying to parse  logfiles via cut will always be imperfect.

dmesg -r|grep "^<7>" seems to work well enough for debug messages

> Going forward, the proper priorities for dyndbg are IMO (CMIIW)
> 
> - get jumplabels working
> 
>   Ive coded this, and it looks ok, but Im stuck on the circular
> include problem seen
>   when I add #include <linux/jumplabel.h> needed for the ddebug.key field.
>   Ive tried a few permutations, but I dont think thats gonna
> illuminate a working fix.
>   I think I'll post a minimal RF-Help patch to demonstrate the problem ..

Seems reasonable.

> - propose a pr_debug_flags("flagchars", "fmt", ...)
> 
>   "flagchars" would be uppercase chars defined for each module
>   DBG_DEFINE_FLAG( mod_debug_str, "A", "alpha flag controls .. blah blah..");

what's the fmt,... for?
Why not just a per-module set of flags?