Messages in this thread | | | From | Jim Cromie <> | Date | Fri, 16 Dec 2011 22:10:48 -0700 | Subject | Re: [PATCH 18/25] dynamic_debug: Introduce global fake module param $module.dyndbg |
| |
On Thu, Dec 15, 2011 at 5:00 PM, Rusty Russell <rusty@rustcorp.com.au> wrote: > On Mon, 12 Dec 2011 16:12:44 -0700, jim.cromie@gmail.com wrote: >> From: Jim Cromie <jim.cromie@gmail.com> >> >> Rework Thomas Renninger's $module.ddebug boot-time debugging feature, >> from https://lkml.org/lkml/2010/9/15/397 > > OK, this looks cleaner! > > s/modname/doing/ here, since it's a generic callback. > > s/modname/doing/ (or /s/modname/unused/). > > Same as above. > > s/modname/doing/ here too, makes it clear. > > ... again....
All done, to be sent shortly. Thanks
> >> +/* find & process .dyndbg params (prefixed by modname) in cmdline */ >> +static __init void ddebug_boot_parse_args(void) >> +{ > > OK, I don't get why you are doing your own parsing for builtins? >
After some poking around I didnt want to thrash too much, and went for Thomas' original approach for builtins.
> Can't you call something from init/main.c:unknown_bootoption()? > > If that's too early, I have a patch we're working on which adds > parameter parsing to each initlevel. >
It looks too early (unless Ive misunderstood) dynamic_debug is enabled via arch_initcall
977 static int __init dynamic_debug_init(void) 978 { ... 1018 } 1019 ddebug_boot_parse_args(); ... 1027 }
1028 /* Allow early initialization for boot messages via boot param */ 1029 arch_initcall(dynamic_debug_init); 1030 /* Debugfs setup must be done later */ 1031 module_init(dynamic_debug_init_debugfs);
Do you have a pointer to that patch ? Is it targetted for v3.3 ? If so, it sounds like I should aim to use it.
BTW, is this/your desc of arg-processing still accurate/current/complete ? http://thread.gmane.org/gmane.linux.kernel/767172/focus=767361
> Thanks, > Rusty.
thank you Jim Cromie
| |