Re: [PATCH 1/5] Revert "kmemtrace: fix printk format warnings"

[Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>]

* Pekka Enberg (penberg@cs.helsinki.fi) wrote:
> Eduard - Gabriel Munteanu wrote:
>> On Tue, Aug 19, 2008 at 02:16:53PM -0400, Mathieu Desnoyers wrote:
>>> Question :
>>>
>>> Is this kmemtrace marker meant to be exposed to userspace ?
>>>
>>> I suspect not. In all case, not directly. I expect in-kernel probes to
>>> be connected on these markers which will get the arguments they need,
>>> and maybe access the inner data structures. Anyhow, tracepoints should
>>> be used for that, not markers. You can later put markers in the probes
>>> which are themselves connected to tracepoints.
>>>
>>> Tracepoints = in-kernel tracing API.
>>>
>>> Markers = Data-formatting tracing API, meant to export the data either
>>> to user-space in text or binary format.
>>>
>>> See
>>>
>>> http://git.kernel.org/?p=linux/kernel/git/compudj/linux-2.6-lttng.git;a=shortlog
>>>
>>> tracepoint-related patches.
>> I think we're ready to try tracepoints. Pekka, could you merge Mathieu's
>> tracepoints or otherwise provide a branch where I could submit a
>> tracepoint conversion patch for kmemtrace?
>
> Sorry, that's too much of a hassle for me. I'll happily take your 
> conversion patch once tracepoints hit mainline. Mathieu, are you aiming for 
> 2.6.28?

Probably. it's in -tip right now, and the new ftrace depends on it. So
at least 2.6.28 yes.

Mathieu

-- 
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68