Hi Randy,
* Randy Dunlap ([EMAIL PROTECTED]) wrote:
> On Thu, 6 Sep 2007 16:05:37 -0400 Mathieu Desnoyers wrote:
>
> > Scheduler Profiling - Use Immediate Values
> >
> > Use immediate values with lower d-cache hit in optimized version as a
> > condition for scheduler profiling call.
> >
> > Depends on the Immediate values patches, applies to 2.6.23-rc4-mm1.
> >
> > Signed-off-by: Mathieu Desnoyers <[EMAIL PROTECTED]>
> > ---
> > drivers/kvm/svm.c | 2 +-
> > drivers/kvm/vmx.c | 2 +-
> > include/linux/profile.h | 10 ++++------
> > kernel/profile.c | 38 ++++++++++++++++++++++++++------------
> > kernel/sched.c | 3 ++-
> > 5 files changed, 34 insertions(+), 21 deletions(-)
> >
> > Index: linux-2.6-lttng/kernel/profile.c
> > ===================================================================
> > --- linux-2.6-lttng.orig/kernel/profile.c 2007-09-04 11:53:23.000000000
> > -0400
> > +++ linux-2.6-lttng/kernel/profile.c 2007-09-04 12:12:16.000000000
> > -0400
> > @@ -42,9 +42,6 @@ static int (*timer_hook)(struct pt_regs
> > static atomic_t *prof_buffer;
> > static unsigned long prof_len, prof_shift;
> >
> > -int prof_on __read_mostly;
> > -EXPORT_SYMBOL_GPL(prof_on);
> > -
> > static cpumask_t prof_cpu_mask = CPU_MASK_ALL;
> > #ifdef CONFIG_SMP
> > static DEFINE_PER_CPU(struct profile_hit *[2], cpu_profile_hits);
> > @@ -52,6 +49,14 @@ static DEFINE_PER_CPU(int, cpu_profile_f
> > static DEFINE_MUTEX(profile_flip_mutex);
> > #endif /* CONFIG_SMP */
> >
> > +/* Immediate values */
> > +immediate_char_t sleep_profiling __read_mostly,
> > + sched_profiling __read_mostly,
> > + kvm_profiling __read_mostly,
> > + cpu_profiling __read_mostly;
> > +EXPORT_SYMBOL_GPL(kvm_profiling);
> > +EXPORT_SYMBOL_GPL(cpu_profiling);
> > +
> > static int __init profile_setup(char * str)
> > {
> > static char __initdata schedstr[] = "schedule";
> > @@ -60,7 +65,7 @@ static int __init profile_setup(char * s
> > int par;
> >
> > if (!strncmp(str, sleepstr, strlen(sleepstr))) {
> > - prof_on = SLEEP_PROFILING;
> > + immediate_set_early(&sleep_profiling, 1);
>
> These are booleans, eh? so one could use , true);
>
> etc.
>
> > if (str[strlen(sleepstr)] == ',')
> > str += strlen(sleepstr) + 1;
> > if (get_option(&str, &par))
> > @@ -99,7 +104,10 @@ __setup("profile=", profile_setup);
> >
> > void __init profile_init(void)
> > {
> > - if (!prof_on)
> > + if (!_immediate_read(&sleep_profiling) &&
> > + !_immediate_read(&sched_profiling) &&
> > + !_immediate_read(&kvm_profiling) &&
> > + !_immediate_read(&cpu_profiling))
>
> Looks like this (in multiple places) needs to use prof_on, where
> prof_on is set earlier as the logical 'or' of all (4 now, could be more)
> profiling flags.
I first did this patch when the immediate values only supported
booleans. I just redid it encoding the original prof_on values in a
signle immediate_char_t variable. It makes the change must less
intrusive.
Since there seem to be only one type of profiling possible at a given
time, there is no need for a bitfield : just doing the same as prof_on
currently does seems ok.
> Or below, sometimes you just check prof_buffer. Is that sufficient
> here also?
>
Yes, before we already do the check early in profile_hit (inline
function checking the same condition). In this patch, I moved the check
out of profile_hit, but since I now re-use prof_on with values, I can
put it back in profile_hit().
> > return;
> >
> > /* only text is profiled */
> > @@ -288,7 +296,7 @@ void profile_hits(int type, void *__pc,
> > int i, j, cpu;
> > struct profile_hit *hits;
> >
> > - if (prof_on != type || !prof_buffer)
> > + if (!prof_buffer)
> > return;
> > pc = min((pc - (unsigned long)_stext) >> prof_shift, prof_len - 1);
> > i = primary = (pc & (NR_PROFILE_GRP - 1)) << PROFILE_GRPSHIFT;
> > @@ -398,7 +406,7 @@ void profile_hits(int type, void *__pc,
> > {
> > unsigned long pc;
> >
> > - if (prof_on != type || !prof_buffer)
> > + if (!prof_buffer)
> > return;
> > pc = ((unsigned long)__pc - (unsigned long)_stext) >> prof_shift;
> > atomic_add(nr_hits, &prof_buffer[min(pc, prof_len - 1)]);
> > @@ -555,7 +563,10 @@ static int __init create_hash_tables(voi
> > }
> > return 0;
> > out_cleanup:
> > - prof_on = 0;
> > + immediate_set_early(&sleep_profiling, 0);
> > + immediate_set_early(&sched_profiling, 0);
> > + immediate_set_early(&kvm_profiling, 0);
> > + immediate_set_early(&cpu_profiling, 0);
> > smp_mb();
> > on_each_cpu(profile_nop, NULL, 0, 1);
> > for_each_online_cpu(cpu) {
> > @@ -582,7 +593,10 @@ static int __init create_proc_profile(vo
> > {
> > struct proc_dir_entry *entry;
> >
> > - if (!prof_on)
> > + if (!_immediate_read(&sleep_profiling) &&
> > + !_immediate_read(&sched_profiling) &&
> > + !_immediate_read(&kvm_profiling) &&
> > + !_immediate_read(&cpu_profiling))
> > return 0;
> > if (create_hash_tables())
> > return -1;
>
>
> ---
> ~Randy
> *** Remember to use Documentation/SubmitChecklist when testing your code ***
--
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
