This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [Patch 0/2] dynamic debug


On Thu, Feb 05, 2009 at 11:45:20AM -0500, Jason Baron wrote:
> hi,
> 
> This patchset combines Greg Bank's dprintk() patchset with the dynamic printk
> patchset. We are renaming the combined solution 'dynamic debug'. As a reminder,
> the patchset allows debug printk style statements to be dynamically 
> enabled/disabled at runtime.
> 
> The key new feature of this patchset is a richer /debugfs control file
> interface, (an example output from my system is at the bottom), which allows
> fined grained control over the the debug output. The output can be controlled by
> function, file and line number, module, and format string. 

That's nice, but do we really need to have that kind of control?  It
seems like overkill.

> For example, to enable all debug statement in module 'nf_conntrack':
> 
> #echo -n 'module nf_conntrack +p' > /mnt/debugfs/dynamic_debug/control
> 
> A further explanation can be found in the documentation patch.
> 
> While adding these new features we have dropped a couple that we intend to add
> back in subsequent patches. The ability to turn on/off all debugging at once,
> either via the /debugfs control file, or via the kernel command line.
> 
> thanks,
> 
> -Jason and Greg
> 
> 
> # filename:lineno [module]function flags format
> arch/x86/kernel/setup_percpu.c:260 [setup_percpu]numa_set_node - "Setting\040node\040for\040non-present\040cpu\040%d\012"

Why the \040 syntax?  Can't we handle a space in a string?  Makes it
very hard to grep for things...

I'll give these a spin in my tree to see how it works for now.

thanks,

greg k-h


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]