Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[PATCH] x86, perfctr: remove unused func avail_to_resrv_perfctr_nmi()

2 views
Skip to first unread message

Naga Chumbalkar

unread,
Dec 23, 2009, 9:00:02 PM12/23/09
to
avail_to_resrv_perfctr_nmi() is neither EXPORT'd, nor used in the file.
So remove it.

Signed-off-by: Naga Chumbalkar <nagananda....@hp.com>
---
arch/x86/include/asm/nmi.h | 1 -
arch/x86/kernel/cpu/perfctr-watchdog.c | 11 -----------
2 files changed, 0 insertions(+), 12 deletions(-)

diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h
index 139d4c1..93da9c3 100644
--- a/arch/x86/include/asm/nmi.h
+++ b/arch/x86/include/asm/nmi.h
@@ -19,7 +19,6 @@ extern void die_nmi(char *str, struct pt_regs *regs, int do_panic);
extern int check_nmi_watchdog(void);
extern int nmi_watchdog_enabled;
extern int avail_to_resrv_perfctr_nmi_bit(unsigned int);
-extern int avail_to_resrv_perfctr_nmi(unsigned int);
extern int reserve_perfctr_nmi(unsigned int);
extern void release_perfctr_nmi(unsigned int);
extern int reserve_evntsel_nmi(unsigned int);
diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c
index 898df97..74f4e85 100644
--- a/arch/x86/kernel/cpu/perfctr-watchdog.c
+++ b/arch/x86/kernel/cpu/perfctr-watchdog.c
@@ -115,17 +115,6 @@ int avail_to_resrv_perfctr_nmi_bit(unsigned int counter)

return !test_bit(counter, perfctr_nmi_owner);
}
-
-/* checks the an msr for availability */
-int avail_to_resrv_perfctr_nmi(unsigned int msr)
-{
- unsigned int counter;
-
- counter = nmi_perfctr_msr_to_bit(msr);
- BUG_ON(counter > NMI_MAX_COUNTER_BITS);
-
- return !test_bit(counter, perfctr_nmi_owner);
-}
EXPORT_SYMBOL(avail_to_resrv_perfctr_nmi_bit);

int reserve_perfctr_nmi(unsigned int msr)
--
1.6.2.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Cyrill Gorcunov

unread,
Dec 24, 2009, 10:10:02 AM12/24/09
to
On Thu, Dec 24, 2009 at 01:54:47AM +0000, Naga Chumbalkar wrote:
> avail_to_resrv_perfctr_nmi() is neither EXPORT'd, nor used in the file.
> So remove it.
>
> Signed-off-by: Naga Chumbalkar <nagananda....@hp.com>
> ---
> arch/x86/include/asm/nmi.h | 1 -
> arch/x86/kernel/cpu/perfctr-watchdog.c | 11 -----------
> 2 files changed, 0 insertions(+), 12 deletions(-)
>

Hi Naga, this looks good to me! Thanks.

Is there any plans to use this routine in future
so we shouldn't touch this snippet?

-- Cyrill

Ingo Molnar

unread,
Dec 28, 2009, 3:40:01 AM12/28/09
to

* Cyrill Gorcunov <gorc...@gmail.com> wrote:

> On Thu, Dec 24, 2009 at 01:54:47AM +0000, Naga Chumbalkar wrote:
> > avail_to_resrv_perfctr_nmi() is neither EXPORT'd, nor used in the file.
> > So remove it.
> >
> > Signed-off-by: Naga Chumbalkar <nagananda....@hp.com>
> > ---
> > arch/x86/include/asm/nmi.h | 1 -
> > arch/x86/kernel/cpu/perfctr-watchdog.c | 11 -----------
> > 2 files changed, 0 insertions(+), 12 deletions(-)
> >
>
> Hi Naga, this looks good to me! Thanks.
>
> Is there any plans to use this routine in future
> so we shouldn't touch this snippet?

Not that i know of.

In fact we should transform/migrate the NMI watchdog driver by making it based
on a kernel-internal created perf event. (which is what the NMI watchdog
really is: a periodic NMI event occuring once per second and running a
callback function.)

That would give us NMI watchdog support on all future x86 CPUs, automatically
- without having to extend the arch/x86/kernel/cpu/perfctr-watchdog.c code for
it.

Ingo

Cyrill Gorcunov

unread,
Dec 28, 2009, 5:00:01 AM12/28/09
to
On Mon, Dec 28, 2009 at 09:38:45AM +0100, Ingo Molnar wrote:
>
> * Cyrill Gorcunov <gorc...@gmail.com> wrote:
>
> > On Thu, Dec 24, 2009 at 01:54:47AM +0000, Naga Chumbalkar wrote:
> > > avail_to_resrv_perfctr_nmi() is neither EXPORT'd, nor used in the file.
> > > So remove it.
> > >
> > > Signed-off-by: Naga Chumbalkar <nagananda....@hp.com>
> > > ---
> > > arch/x86/include/asm/nmi.h | 1 -
> > > arch/x86/kernel/cpu/perfctr-watchdog.c | 11 -----------
> > > 2 files changed, 0 insertions(+), 12 deletions(-)
> > >
> >
> > Hi Naga, this looks good to me! Thanks.
> >
> > Is there any plans to use this routine in future
> > so we shouldn't touch this snippet?
>
> Not that i know of.

ok, then we could apply this patch I think, at least for
a while.

>
> In fact we should transform/migrate the NMI watchdog driver by making it based
> on a kernel-internal created perf event. (which is what the NMI watchdog
> really is: a periodic NMI event occuring once per second and running a
> callback function.)

Yes, this would be great. I'll try to find out some time for this
task, though no promises :) If someone get it done earlier I would
really appreciate.

>
> That would give us NMI watchdog support on all future x86 CPUs, automatically
> - without having to extend the arch/x86/kernel/cpu/perfctr-watchdog.c code for
> it.
>
> Ingo
>

-- Cyrill

tip-bot for Naga Chumbalkar

unread,
Dec 28, 2009, 5:10:02 AM12/28/09
to
Commit-ID: fd2a50a0240f5f5b59070474eabd83a85720a406
Gitweb: http://git.kernel.org/tip/fd2a50a0240f5f5b59070474eabd83a85720a406
Author: Naga Chumbalkar <nagananda....@hp.com>
AuthorDate: Thu, 24 Dec 2009 01:54:47 +0000
Committer: Ingo Molnar <mi...@elte.hu>
CommitDate: Mon, 28 Dec 2009 09:36:46 +0100

x86, perfctr: Remove unused func avail_to_resrv_perfctr_nmi()

avail_to_resrv_perfctr_nmi() is neither EXPORT'd, nor used in
the file. So remove it.

Signed-off-by: Naga Chumbalkar <nagananda....@hp.com>
Acked-by: Cyrill Gorcunov <gorc...@gmail.com>
Cc: oprofi...@lists.sf.net
LKML-Reference: <20091224015441.600...@localhost.localdomain>
Signed-off-by: Ingo Molnar <mi...@elte.hu>

Ingo Molnar

unread,
Dec 28, 2009, 5:10:01 AM12/28/09
to

* Cyrill Gorcunov <gorc...@gmail.com> wrote:

> > Not that i know of.
>
> ok, then we could apply this patch I think, at least for a while.

yeah, i have applied it.

> > In fact we should transform/migrate the NMI watchdog driver by making it
> > based on a kernel-internal created perf event. (which is what the NMI
> > watchdog really is: a periodic NMI event occuring once per second and
> > running a callback function.)
>
> Yes, this would be great. I'll try to find out some time for this task,
> though no promises :) If someone get it done earlier I would really
> appreciate.

Feel free to do it. We could do it gradual. Maybe we could also then gradually
remove perfctr code for CPU variants that already have perf events support.

Plus later on we could carefully identify bits of APIC support code in
perfctr.c and turn those into minimal perf events PMU drivers.

Ingo

0 new messages