Linus,
please pull the latest timers-urgent-for-linus git tree from:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus
The following trilogy of patches brings you:
* Fix for a long standing math overflow issue with HZ < 60
* An onliner fix for a corner case in the dreaded tick broadcast
mechanism affecting a certain range of AMD machines which are
infested with the infamous automagic C1E power control misfeature
* A fix for one of the ARM platforms which allows the kernel to
proceed and boot instead of stupidly panicing for no good
reason. The patch is slightly larger than necessary, but it's less
ugly than the alternative 5 liner.
Thanks,
tglx
------------------>
Mikulas Patocka (1):
time: Fix overflow when HZ is smaller than 60
Thomas Gleixner (1):
tick: Clear broadcast pending bit when switching to oneshot
Tim Kryger (1):
clocksource: Kona: Print warning rather than panic
drivers/clocksource/bcm_kona_timer.c | 54 ++++++++++++++++++----------------
kernel/time/jiffies.c | 6 ++++
kernel/time/tick-broadcast.c | 1 +
3 files changed, 35 insertions(+), 26 deletions(-)
diff --git a/drivers/clocksource/bcm_kona_timer.c b/drivers/clocksource/bcm_kona_timer.c
index 974b2db..0595dc6 100644
--- a/drivers/clocksource/bcm_kona_timer.c
+++ b/drivers/clocksource/bcm_kona_timer.c
@@ -99,31 +99,6 @@ kona_timer_get_counter(void *timer_base, uint32_t *msw, uint32_t *lsw)
return;
}
-static void __init kona_timers_init(struct device_node *node)
-{
- u32 freq;
- struct clk *external_clk;
-
- external_clk = of_clk_get_by_name(node, NULL);
-
- if (!IS_ERR(external_clk)) {
- arch_timer_rate = clk_get_rate(external_clk);
- clk_prepare_enable(external_clk);
- } else if (!of_property_read_u32(node, "clock-frequency", &freq)) {
- arch_timer_rate = freq;
- } else {
- panic("unable to determine clock-frequency");
- }
-
- /* Setup IRQ numbers */
- timers.tmr_irq = irq_of_parse_and_map(node, 0);
-
- /* Setup IO addresses */
- timers.tmr_regs = of_iomap(node, 0);
-
- kona_timer_disable_and_clear(timers.tmr_regs);
-}
-
static int kona_timer_set_next_event(unsigned long clc,
struct clock_event_device *unused)
{
@@ -198,7 +173,34 @@ static struct irqaction kona_timer_irq = {
static void __init kona_timer_init(struct device_node *node)
{
- kona_timers_init(node);
+ u32 freq;
+ struct clk *external_clk;
+
+ if (!of_device_is_available(node)) {
+ pr_info("Kona Timer v1 marked as disabled in device tree\n");
+ return;
+ }
+
+ external_clk = of_clk_get_by_name(node, NULL);
+
+ if (!IS_ERR(external_clk)) {
+ arch_timer_rate = clk_get_rate(external_clk);
+ clk_prepare_enable(external_clk);
+ } else if (!of_property_read_u32(node, "clock-frequency", &freq)) {
+ arch_timer_rate = freq;
+ } else {
+ pr_err("Kona Timer v1 unable to determine clock-frequency");
+ return;
+ }
+
+ /* Setup IRQ numbers */
+ timers.tmr_irq = irq_of_parse_and_map(node, 0);
+
+ /* Setup IO addresses */
+ timers.tmr_regs = of_iomap(node, 0);
+
+ kona_timer_disable_and_clear(timers.tmr_regs);
+
kona_timer_clockevents_init();
setup_irq(timers.tmr_irq, &kona_timer_irq);
kona_timer_set_next_event((arch_timer_rate / HZ), NULL);
diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c
index 7a925ba..a6a5bf5 100644
--- a/kernel/time/jiffies.c
+++ b/kernel/time/jiffies.c
@@ -51,7 +51,13 @@
* HZ shrinks, so values greater than 8 overflow 32bits when
* HZ=100.
*/
+#if HZ < 34
+#define JIFFIES_SHIFT 6
+#elif HZ < 67
+#define JIFFIES_SHIFT 7
+#else
#define JIFFIES_SHIFT 8
+#endif
static cycle_t jiffies_read(struct clocksource *cs)
{
diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c
index 43780ab..98977a5 100644
--- a/kernel/time/tick-broadcast.c
+++ b/kernel/time/tick-broadcast.c
@@ -756,6 +756,7 @@ out:
static void tick_broadcast_clear_oneshot(int cpu)
{
cpumask_clear_cpu(cpu, tick_broadcast_oneshot_mask);
+ cpumask_clear_cpu(cpu, tick_broadcast_pending_mask);
}
static void tick_broadcast_init_next_event(struct cpumask *mask,
Hi Thomas,
With regard to the patch: "tick: Clear broadcast pending bit when
switching to oneshot"
isn't BROADCAST_EXIT called atleast after initializing the high
resolution timers while
handling the broadcast IPI and before re-entering idle? This should
clear the cpu bit
in the pending mask right?
idle()
|__BROADCAST_ENTER
|__halt()
|__broadcast_ipi
|__init_highres()->clear oneshot mask
|__return from interrupt->BROADCAST_EXIT (clear pending mask)
|__re-enter idle
When the cpu re-enters idle and does a BROADCAST_ENTER again, the
pending mask is cleared
and hence should not trigger a WARN_ON().
Thanks
Regards
Preeti U Murthy
On Sun, Feb 16, 2014 at 12:51 AM, Thomas Gleixner <[email protected]> wrote:
> Linus,
>
> please pull the latest timers-urgent-for-linus git tree from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus
>
> The following trilogy of patches brings you:
>
> * Fix for a long standing math overflow issue with HZ < 60
>
> * An onliner fix for a corner case in the dreaded tick broadcast
> mechanism affecting a certain range of AMD machines which are
> infested with the infamous automagic C1E power control misfeature
>
> * A fix for one of the ARM platforms which allows the kernel to
> proceed and boot instead of stupidly panicing for no good
> reason. The patch is slightly larger than necessary, but it's less
> ugly than the alternative 5 liner.
>
>
> Thanks,
>
> tglx
>
> ------------------>
> Mikulas Patocka (1):
> time: Fix overflow when HZ is smaller than 60
>
> Thomas Gleixner (1):
> tick: Clear broadcast pending bit when switching to oneshot
>
> Tim Kryger (1):
> clocksource: Kona: Print warning rather than panic
>
>
> drivers/clocksource/bcm_kona_timer.c | 54 ++++++++++++++++++----------------
> kernel/time/jiffies.c | 6 ++++
> kernel/time/tick-broadcast.c | 1 +
> 3 files changed, 35 insertions(+), 26 deletions(-)
>
> diff --git a/drivers/clocksource/bcm_kona_timer.c b/drivers/clocksource/bcm_kona_timer.c
> index 974b2db..0595dc6 100644
> --- a/drivers/clocksource/bcm_kona_timer.c
> +++ b/drivers/clocksource/bcm_kona_timer.c
> @@ -99,31 +99,6 @@ kona_timer_get_counter(void *timer_base, uint32_t *msw, uint32_t *lsw)
> return;
> }
>
> -static void __init kona_timers_init(struct device_node *node)
> -{
> - u32 freq;
> - struct clk *external_clk;
> -
> - external_clk = of_clk_get_by_name(node, NULL);
> -
> - if (!IS_ERR(external_clk)) {
> - arch_timer_rate = clk_get_rate(external_clk);
> - clk_prepare_enable(external_clk);
> - } else if (!of_property_read_u32(node, "clock-frequency", &freq)) {
> - arch_timer_rate = freq;
> - } else {
> - panic("unable to determine clock-frequency");
> - }
> -
> - /* Setup IRQ numbers */
> - timers.tmr_irq = irq_of_parse_and_map(node, 0);
> -
> - /* Setup IO addresses */
> - timers.tmr_regs = of_iomap(node, 0);
> -
> - kona_timer_disable_and_clear(timers.tmr_regs);
> -}
> -
> static int kona_timer_set_next_event(unsigned long clc,
> struct clock_event_device *unused)
> {
> @@ -198,7 +173,34 @@ static struct irqaction kona_timer_irq = {
>
> static void __init kona_timer_init(struct device_node *node)
> {
> - kona_timers_init(node);
> + u32 freq;
> + struct clk *external_clk;
> +
> + if (!of_device_is_available(node)) {
> + pr_info("Kona Timer v1 marked as disabled in device tree\n");
> + return;
> + }
> +
> + external_clk = of_clk_get_by_name(node, NULL);
> +
> + if (!IS_ERR(external_clk)) {
> + arch_timer_rate = clk_get_rate(external_clk);
> + clk_prepare_enable(external_clk);
> + } else if (!of_property_read_u32(node, "clock-frequency", &freq)) {
> + arch_timer_rate = freq;
> + } else {
> + pr_err("Kona Timer v1 unable to determine clock-frequency");
> + return;
> + }
> +
> + /* Setup IRQ numbers */
> + timers.tmr_irq = irq_of_parse_and_map(node, 0);
> +
> + /* Setup IO addresses */
> + timers.tmr_regs = of_iomap(node, 0);
> +
> + kona_timer_disable_and_clear(timers.tmr_regs);
> +
> kona_timer_clockevents_init();
> setup_irq(timers.tmr_irq, &kona_timer_irq);
> kona_timer_set_next_event((arch_timer_rate / HZ), NULL);
> diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c
> index 7a925ba..a6a5bf5 100644
> --- a/kernel/time/jiffies.c
> +++ b/kernel/time/jiffies.c
> @@ -51,7 +51,13 @@
> * HZ shrinks, so values greater than 8 overflow 32bits when
> * HZ=100.
> */
> +#if HZ < 34
> +#define JIFFIES_SHIFT 6
> +#elif HZ < 67
> +#define JIFFIES_SHIFT 7
> +#else
> #define JIFFIES_SHIFT 8
> +#endif
>
> static cycle_t jiffies_read(struct clocksource *cs)
> {
> diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c
> index 43780ab..98977a5 100644
> --- a/kernel/time/tick-broadcast.c
> +++ b/kernel/time/tick-broadcast.c
> @@ -756,6 +756,7 @@ out:
> static void tick_broadcast_clear_oneshot(int cpu)
> {
> cpumask_clear_cpu(cpu, tick_broadcast_oneshot_mask);
> + cpumask_clear_cpu(cpu, tick_broadcast_pending_mask);
> }
>
> static void tick_broadcast_init_next_event(struct cpumask *mask,
> --
> 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/
On Tue, 18 Feb 2014, Preeti Murthy wrote:
> Hi Thomas,
>
> With regard to the patch: "tick: Clear broadcast pending bit when
> switching to oneshot"
> isn't BROADCAST_EXIT called atleast after initializing the high
> resolution timers while
> handling the broadcast IPI and before re-entering idle? This should
> clear the cpu bit
> in the pending mask right?
>
> idle()
> |__BROADCAST_ENTER
> |__halt()
> |__broadcast_ipi
> |__init_highres()->clear oneshot mask
> |__return from interrupt->BROADCAST_EXIT (clear pending mask)
Unfortunately not:
if (cpumask_test_and_clear_cpu(cpu, tick_broadcast_oneshot_mask)) {
clear pending mask;
}
> |__re-enter idle
Go figure :)
Thanks,
tglx
On 02/19/2014 12:10 AM, Thomas Gleixner wrote:
> On Tue, 18 Feb 2014, Preeti Murthy wrote:
>
>> Hi Thomas,
>>
>> With regard to the patch: "tick: Clear broadcast pending bit when
>> switching to oneshot"
>> isn't BROADCAST_EXIT called atleast after initializing the high
>> resolution timers while
>> handling the broadcast IPI and before re-entering idle? This should
>> clear the cpu bit
>> in the pending mask right?
>>
>> idle()
>> |__BROADCAST_ENTER
>> |__halt()
>> |__broadcast_ipi
>> |__init_highres()->clear oneshot mask
>> |__return from interrupt->BROADCAST_EXIT (clear pending mask)
>
> Unfortunately not:
>
> if (cpumask_test_and_clear_cpu(cpu, tick_broadcast_oneshot_mask)) {
> clear pending mask;
> }
>
>> |__re-enter idle
>
> Go figure :)
Ah right! Thanks :)
Regards
Preeti U Murthy
>
> Thanks,
>
> tglx
>