On Mon, Mar 16, 2009 at 01:36:09PM +0100, Frederic Weisbecker wrote:
> On Fri, Mar 13, 2009 at 10:39:36AM +0000, Lai Jiangshan wrote:
> > Commit-ID: e94142a67f8bad494c593f0a07c9fc2fbec98c0e
> > Gitweb: http://git.kernel.org/tip/e94142a67f8bad494c593f0a07c9fc2fbec98c0e
> > Author: Lai Jiangshan <laijs@xxxxxxxxxxxxxx>
> > AuthorDate: Fri, 13 Mar 2009 17:51:27 +0800
> > Commit: Ingo Molnar <mingo@xxxxxxx>
> > CommitDate: Fri, 13 Mar 2009 11:36:20 +0100
> >
> > ftrace: remove struct list_head from struct dyn_ftrace
> >
> > Impact: save memory
> >
> > The struct dyn_ftrace table is very large, this patch will save
> > about 50%.
> >
> > Signed-off-by: Lai Jiangshan <laijs@xxxxxxxxxxxxxx>
> > Cc: Steven Rostedt <srostedt@xxxxxxxxxx>
> > LKML-Reference: <49BA2C9F.8020009@xxxxxxxxxxxxxx>
> > Signed-off-by: Ingo Molnar <mingo@xxxxxxx>
> >
> >
> > ---
> > include/linux/ftrace.h | 1 -
> > kernel/trace/ftrace.c | 14 ++++++++------
> > 2 files changed, 8 insertions(+), 7 deletions(-)
> >
> > diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h
> > index c146c10..9d598bb 100644
> > --- a/include/linux/ftrace.h
> > +++ b/include/linux/ftrace.h
> > @@ -145,7 +145,6 @@ enum {
> > };
> >
> > struct dyn_ftrace {
> > - struct list_head list;
> > unsigned long ip; /* address of mcount call-site */
> > unsigned long flags;
> > struct dyn_arch_ftrace arch;
> > diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
> > index bf78a4c..90d5729 100644
> > --- a/kernel/trace/ftrace.c
> > +++ b/kernel/trace/ftrace.c
> > @@ -272,7 +272,7 @@ enum {
> >
> > static int ftrace_filtered;
> >
> > -static LIST_HEAD(ftrace_new_addrs);
> > +static struct dyn_ftrace *ftrace_new_addrs;
> >
> > static DEFINE_MUTEX(ftrace_regex_lock);
> >
> > @@ -409,8 +409,8 @@ ftrace_record_ip(unsigned long ip)
> > return NULL;
> >
> > rec->ip = ip;
> > -
> > - list_add(&rec->list, &ftrace_new_addrs);
> > + rec->flags = (unsigned long)ftrace_new_addrs;
> > + ftrace_new_addrs = rec;
>
>
> But... you are overwritting all the ftrace flags infrastructure by
> using it as a list pointer.
>
> This is breaking dynamic tracing.
May be I'm wrong. May be I misunderstood this patch.
But I don't understand how flags can be used at the same time as a pointer
and as a flag value.
> >
> > return rec;
> > }
> > @@ -716,19 +716,21 @@ unsigned long ftrace_update_tot_cnt;
> >
> > static int ftrace_update_code(struct module *mod)
> > {
> > - struct dyn_ftrace *p, *t;
> > + struct dyn_ftrace *p;
> > cycle_t start, stop;
> >
> > start = ftrace_now(raw_smp_processor_id());
> > ftrace_update_cnt = 0;
> >
> > - list_for_each_entry_safe(p, t, &ftrace_new_addrs, list) {
> > + while (ftrace_new_addrs) {
> >
> > /* If something went wrong, bail without enabling anything */
> > if (unlikely(ftrace_disabled))
> > return -1;
> >
> > - list_del_init(&p->list);
> > + p = ftrace_new_addrs;
> > + ftrace_new_addrs = (struct dyn_ftrace *)p->flags;
> > + p->flags = 0L;
> >
> > /* convert record (i.e, patch mcount-call with NOP) */
> > if (ftrace_code_disable(mod, p)) {
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-tip-commits" in
> > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-tip-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
|