To use the MCT as a sched_clock, the timer value has to be accessed vi
an MCT register which is extremely slow. To improve performance on Arm64
SoCs, use the Arm architected timer as the default clocksource. Note, we
can't completely disable the MCT on Arm64 since it needs to be used as
the wakeup source for the arch_timer to exit the "c2" idle state.

Since ARM SoCs don't have an architectured timer, the MCT will continue
to be the default clocksource. Detailed discussion on this topic can be
found at [1].

[1] https://lore.kernel.org/linux-samsung-soc/1400188079-21832-1-git-send-email-chirantan@chromium.org/

Signed-off-by: Donghoon Yu <hoony.yu@samsung.com>
Signed-off-by: Youngmin Nam <youngmin.nam@samsung.com>
[Original commit from https://android.googlesource.com/kernel/gs/+/630817f7080e92c5e0216095ff52f6eb8dd00727
Signed-off-by: Will McVicker <willmcvicker@google.com>
Reviewed-by: Youngmin Nam <youngmin.nam@samsung.com>
---
 drivers/clocksource/exynos_mct.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c
index da09f467a6bb..96361d5dc57d 100644
--- a/drivers/clocksource/exynos_mct.c
+++ b/drivers/clocksource/exynos_mct.c
@@ -219,12 +219,18 @@ static struct clocksource mct_frc = {
 	.resume		= exynos4_frc_resume,
 };
 
+/*
+ * Since ARM devices do not have an architected timer, they need to continue
+ * using the MCT as the main clocksource for timekeeping, sched_clock, and the
+ * delay timer. For AARCH64 SoCs, the architected timer is the preferred
+ * clocksource due to it's superior performance.
+ */
+#if defined(CONFIG_ARM)
 static u64 notrace exynos4_read_sched_clock(void)
 {
 	return exynos4_read_count_32();
 }
 
-#if defined(CONFIG_ARM)
 static struct delay_timer exynos4_delay_timer;
 
 static cycles_t exynos4_read_current_timer(void)
@@ -250,12 +256,13 @@ static int __init exynos4_clocksource_init(bool frc_shared)
 	exynos4_delay_timer.read_current_timer = &exynos4_read_current_timer;
 	exynos4_delay_timer.freq = clk_rate;
 	register_current_timer_delay(&exynos4_delay_timer);
+
+	sched_clock_register(exynos4_read_sched_clock, 32, clk_rate);
 #endif
 
 	if (clocksource_register_hz(&mct_frc, clk_rate))
 		panic("%s: can't register clocksource\n", mct_frc.name);
 
-	sched_clock_register(exynos4_read_sched_clock, 32, clk_rate);
 
 	return 0;
 }
-- 
2.49.0.472.ge94155a9ec-goog

On Wed, Apr 2, 2025 at 4:34 PM 'Will McVicker' via kernel-team
<kernel-team@android.com> wrote:
>
> To use the MCT as a sched_clock, the timer value has to be accessed vi
> an MCT register which is extremely slow. To improve performance on Arm64
> SoCs, use the Arm architected timer as the default clocksource. Note, we

Nit:  sched_clock is sort of separate from the "default clocksource",
and after this patch we still register MCT as a clocksource, so this
doesn't sound quite right.

I'd probably reword this slightly to:
"The MCT register is unfortunately very slow to access, but importantly
does not halt in the c2 idle state.  So for ARM64, we can improve
performance by not registering the MCT for sched_clock, allowing the
system to use the faster ARM architected timer for sched_clock instead.

The MCT is still registered as a clocksource, and a clockevent in order
to be a wakeup source for the arch_timer to exit the "c2" idle state.

Since ARM32 SoCs don't have an architected timer, the MCT must continue
to be used for sched_clock. Detailed discussion on this topic can be
found at [1]. "

> can't completely disable the MCT on Arm64 since it needs to be used as
> the wakeup source for the arch_timer to exit the "c2" idle state.
>
> Since ARM SoCs don't have an architectured timer, the MCT will continue
> to be the default clocksource. Detailed discussion on this topic can be
> found at [1].
>
> [1] https://lore.kernel.org/linux-samsung-soc/1400188079-21832-1-git-send-email-chirantan@chromium.org/
>
> Signed-off-by: Donghoon Yu <hoony.yu@samsung.com>
> Signed-off-by: Youngmin Nam <youngmin.nam@samsung.com>
> [Original commit from https://android.googlesource.com/kernel/gs/+/630817f7080e92c5e0216095ff52f6eb8dd00727
> Signed-off-by: Will McVicker <willmcvicker@google.com>
> Reviewed-by: Youngmin Nam <youngmin.nam@samsung.com>

Otherwise, looks good.
Acked-by: John Stultz <jstultz@google.com>

thanks
-john

On 04/02/2025, John Stultz wrote:
> On Wed, Apr 2, 2025 at 4:34 PM 'Will McVicker' via kernel-team
> <kernel-team@android.com> wrote:
> >
> > To use the MCT as a sched_clock, the timer value has to be accessed vi
> > an MCT register which is extremely slow. To improve performance on Arm64
> > SoCs, use the Arm architected timer as the default clocksource. Note, we
> 
> Nit:  sched_clock is sort of separate from the "default clocksource",
> and after this patch we still register MCT as a clocksource, so this
> doesn't sound quite right.
> 
> I'd probably reword this slightly to:
> "The MCT register is unfortunately very slow to access, but importantly
> does not halt in the c2 idle state.  So for ARM64, we can improve
> performance by not registering the MCT for sched_clock, allowing the
> system to use the faster ARM architected timer for sched_clock instead.
> 
> The MCT is still registered as a clocksource, and a clockevent in order
> to be a wakeup source for the arch_timer to exit the "c2" idle state.
> 
> Since ARM32 SoCs don't have an architected timer, the MCT must continue
> to be used for sched_clock. Detailed discussion on this topic can be
> found at [1]. "

Thanks John for the suggestion! I'll give more time for review feedback before
sending the v3 with this update.

Regards,
Will

> 
> > can't completely disable the MCT on Arm64 since it needs to be used as
> > the wakeup source for the arch_timer to exit the "c2" idle state.
> >
> > Since ARM SoCs don't have an architectured timer, the MCT will continue
> > to be the default clocksource. Detailed discussion on this topic can be
> > found at [1].
> >
> > [1] https://lore.kernel.org/linux-samsung-soc/1400188079-21832-1-git-send-email-chirantan@chromium.org/
> >
> > Signed-off-by: Donghoon Yu <hoony.yu@samsung.com>
> > Signed-off-by: Youngmin Nam <youngmin.nam@samsung.com>
> > [Original commit from https://android.googlesource.com/kernel/gs/+/630817f7080e92c5e0216095ff52f6eb8dd00727
> > Signed-off-by: Will McVicker <willmcvicker@google.com>
> > Reviewed-by: Youngmin Nam <youngmin.nam@samsung.com>
> 
> Otherwise, looks good.
> Acked-by: John Stultz <jstultz@google.com>
> 
> thanks
> -john