| Message ID | 1403197706-29966-1-git-send-email-dianders@chromium.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 06/19/2014 07:08 PM, Doug Anderson wrote: > In (93bfb76 clocksource: exynos_mct: register sched_clock callback) we > supported using the MCT as a scheduler clock. Hi Thomas, is it possible you update the tip/timers/core so I can get visibility on the above patch to apply this one ? Thanks -- Daniel > We properly marked > exynos4_read_sched_clock() as notrace. However, we then went and > called another function that _wasn't_ notrace. That means if you do: > > cd /sys/kernel/debug/tracing/ > echo function_graph > current_tracer > > You'll get a crash. > > Fix this (but still let other readers of the MCT be trace-enabled) by > adding an extra function. It's important to keep other users of MCT > traceable because the MCT is actually quite slow to access and we want > exynos4_frc_read() to show up in ftrace profiles if it's the > bottleneck. > > Signed-off-by: Doug Anderson <dianders@chromium.org> > --- > Changes in v2: > - Split out from other patches so this can go into 3.16. > - Better comment about why exynos4_frc_read() should be traceable. > - No more useless inline. > > drivers/clocksource/exynos_mct.c | 9 +++++++-- > 1 file changed, 7 insertions(+), 2 deletions(-) > > diff --git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c > index f71d55f..5ce99c0 100644 > --- a/drivers/clocksource/exynos_mct.c > +++ b/drivers/clocksource/exynos_mct.c > @@ -162,7 +162,7 @@ static void exynos4_mct_frc_start(void) > exynos4_mct_write(reg, EXYNOS4_MCT_G_TCON); > } > > -static cycle_t exynos4_frc_read(struct clocksource *cs) > +static cycle_t notrace _exynos4_frc_read(void) > { > unsigned int lo, hi; > u32 hi2 = __raw_readl(reg_base + EXYNOS4_MCT_G_CNT_U); > @@ -176,6 +176,11 @@ static cycle_t exynos4_frc_read(struct clocksource *cs) > return ((cycle_t)hi << 32) | lo; > } > > +static cycle_t exynos4_frc_read(struct clocksource *cs) > +{ > + return _exynos4_frc_read(); > +} > + > static void exynos4_frc_resume(struct clocksource *cs) > { > exynos4_mct_frc_start(); > @@ -192,7 +197,7 @@ struct clocksource mct_frc = { > > static u64 notrace exynos4_read_sched_clock(void) > { > - return exynos4_frc_read(&mct_frc); > + return _exynos4_frc_read(); > } > > static void __init exynos4_clocksource_init(void) >
diff --git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c index f71d55f..5ce99c0 100644 --- a/drivers/clocksource/exynos_mct.c +++ b/drivers/clocksource/exynos_mct.c @@ -162,7 +162,7 @@ static void exynos4_mct_frc_start(void) exynos4_mct_write(reg, EXYNOS4_MCT_G_TCON); } -static cycle_t exynos4_frc_read(struct clocksource *cs) +static cycle_t notrace _exynos4_frc_read(void) { unsigned int lo, hi; u32 hi2 = __raw_readl(reg_base + EXYNOS4_MCT_G_CNT_U); @@ -176,6 +176,11 @@ static cycle_t exynos4_frc_read(struct clocksource *cs) return ((cycle_t)hi << 32) | lo; } +static cycle_t exynos4_frc_read(struct clocksource *cs) +{ + return _exynos4_frc_read(); +} + static void exynos4_frc_resume(struct clocksource *cs) { exynos4_mct_frc_start(); @@ -192,7 +197,7 @@ struct clocksource mct_frc = { static u64 notrace exynos4_read_sched_clock(void) { - return exynos4_frc_read(&mct_frc); + return _exynos4_frc_read(); } static void __init exynos4_clocksource_init(void)
In (93bfb76 clocksource: exynos_mct: register sched_clock callback) we supported using the MCT as a scheduler clock. We properly marked exynos4_read_sched_clock() as notrace. However, we then went and called another function that _wasn't_ notrace. That means if you do: cd /sys/kernel/debug/tracing/ echo function_graph > current_tracer You'll get a crash. Fix this (but still let other readers of the MCT be trace-enabled) by adding an extra function. It's important to keep other users of MCT traceable because the MCT is actually quite slow to access and we want exynos4_frc_read() to show up in ftrace profiles if it's the bottleneck. Signed-off-by: Doug Anderson <dianders@chromium.org> --- Changes in v2: - Split out from other patches so this can go into 3.16. - Better comment about why exynos4_frc_read() should be traceable. - No more useless inline. drivers/clocksource/exynos_mct.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)