lkml.org 
[lkml]   [2009]   [Apr]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: + page-owner-tracking.patch added to -mm tree

* Pekka Enberg <penberg@cs.helsinki.fi> wrote:

> On Wed, 2009-04-01 at 15:49 +0200, Ingo Molnar wrote:
> > And this info could be added to that, and it would sure be nice to
> > hook it up to kmemtrace primarily, which does a lot of similar
> > looking work in the slab space. (but Eduard and Pekka will know how
> > feasible/interesting this is to them.)
>
> Yup, makes sense to me. Something like this is probably a good
> starting point for a proper patch.

looks like an excellent starting point.

> +kmemtrace_print_page_alloc_user(struct trace_iterator *iter,
> + struct kmemtrace_page_alloc_entry *entry)
> +{
> + struct kmemtrace_user_event_page_alloc *ev_alloc;
> + struct trace_seq *s = &iter->seq;
> + struct kmemtrace_user_event *ev;
> +
> + ev = trace_seq_reserve(s, sizeof(*ev));
> + if (!ev)
> + return TRACE_TYPE_PARTIAL_LINE;
> +
> + ev->event_id = KMEMTRACE_USER_PAGE_ALLOC;
> + ev->type_id = entry->type_id;
> + ev->event_size = sizeof(*ev) + sizeof(*ev_alloc);
> + ev->cpu = iter->cpu;
> + ev->timestamp = iter->ts;
> + ev->call_site = 0ULL; /* FIXME */
> + ev->ptr = 0ULL; /* FIXME */

Here we could call save_stack_trace(), in a way like this, to save
up to 8 entries of the allocation back-trace:

#define NR_ENTRIES 8

struct kmemtrace_user_event {
...
unsigned long entries[NR_ENTRIES];
...
};

struct stack_trace trace;

trace.nr_entries = 0;
trace.max_entries = NR_ENTRIES;
trace.entries = ev->entries;
trace.skip = 2;

save_stack_trace(&trace);

( the '2' for skip will skip the useless tracer-internal backtrace
bits. )

ftrace has built-in stacktrace capabilities as well - but they are
not hooked up to the binary-tracing pathway yet - right Steve?

Ingo


\
 
 \ /
  Last update: 2009-04-01 17:27    [W:0.066 / U:0.028 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site