[PATCH v2] x86/platform/uv: Dont use smp_processor_id while preemptible

[PATCH v2] x86/platform/uv: Dont use smp_processor_id while preemptible

[Mike Travis]

To avoid a "BUG: using smp_processor_id() in preemptible" debug warning
message, disable preemption around use of the processor id.  This code
sequence merely decides which portal that this CPU uses to read the RTC.
It does this to avoid thrashing the cache but even if preempted it still
reads the same time from the single RTC clock.

Signed-off-by: Mike Travis <mike.travis@hpe.com>
Reviewed-by: Steve Wahl <steve.wahl@hpe.com>
Reviewed-by: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Cc: stable@vger.kernel.org
---
v2: modify patch description, add Cc:stable tag
---
 arch/x86/platform/uv/uv_time.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/arch/x86/platform/uv/uv_time.c b/arch/x86/platform/uv/uv_time.c
index 54663f3e00cb..094190814a28 100644
--- a/arch/x86/platform/uv/uv_time.c
+++ b/arch/x86/platform/uv/uv_time.c
@@ -275,14 +275,17 @@ static int uv_rtc_unset_timer(int cpu, int force)
  */
 static u64 uv_read_rtc(struct clocksource *cs)
 {
-	unsigned long offset;
+	unsigned long offset, time;
+	unsigned int cpu = get_cpu();
 
 	if (uv_get_min_hub_revision_id() == 1)
 		offset = 0;
 	else
-		offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE;
+		offset = (uv_cpu_blade_processor_id(cpu) * L1_CACHE_BYTES) % PAGE_SIZE;
 
-	return (u64)uv_read_local_mmr(UVH_RTC | offset);
+	time = (u64)uv_read_local_mmr(UVH_RTC | offset);
+	put_cpu();
+	return time;
 }
 
 /*
-- 
2.26.2

Re: [PATCH v2] x86/platform/uv: Dont use smp_processor_id while preemptible

[Andy Shevchenko]

On Thu, Sep 22, 2022 at 11:01 PM Mike Travis <mike.travis@hpe.com> wrote:
>
> To avoid a "BUG: using smp_processor_id() in preemptible" debug warning
> message, disable preemption around use of the processor id.  This code
> sequence merely decides which portal that this CPU uses to read the RTC.
> It does this to avoid thrashing the cache but even if preempted it still
> reads the same time from the single RTC clock.

...

> Signed-off-by: Mike Travis <mike.travis@hpe.com>
> Reviewed-by: Steve Wahl <steve.wahl@hpe.com>
> Reviewed-by: Dimitri Sivanich <dimitri.sivanich@hpe.com>
> Cc: stable@vger.kernel.org

No kernel version? No Fixes tag?

...

> -               offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE;
> +               offset = (uv_cpu_blade_processor_id(cpu) * L1_CACHE_BYTES) % PAGE_SIZE;

Perhaps it can be transformed to use offset_in_page() at the same time.

-- 
With Best Regards,
Andy Shevchenko

[PATCH v2] x86/platform/uv: Dont use smp_processor_id while preemptible

[Mike Travis]

To avoid a "BUG: using smp_processor_id() in preemptible" debug
warning message, disable preemption around use of the processor id.

Signed-off-by: Mike Travis <mike.travis@hpe.com>
Reviewed-by: Steve Wahl <steve.wahl@hpe.com>
Reviewed-by: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Cc: stable@vger.kernel.org
---
v2: Add Cc: stable tag
---
 arch/x86/platform/uv/uv_time.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/arch/x86/platform/uv/uv_time.c b/arch/x86/platform/uv/uv_time.c
index 54663f3e00cb..094190814a28 100644
--- a/arch/x86/platform/uv/uv_time.c
+++ b/arch/x86/platform/uv/uv_time.c
@@ -275,14 +275,17 @@ static int uv_rtc_unset_timer(int cpu, int force)
  */
 static u64 uv_read_rtc(struct clocksource *cs)
 {
-	unsigned long offset;
+	unsigned long offset, time;
+	unsigned int cpu = get_cpu();
 
 	if (uv_get_min_hub_revision_id() == 1)
 		offset = 0;
 	else
-		offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE;
+		offset = (uv_cpu_blade_processor_id(cpu) * L1_CACHE_BYTES) % PAGE_SIZE;
 
-	return (u64)uv_read_local_mmr(UVH_RTC | offset);
+	time = (u64)uv_read_local_mmr(UVH_RTC | offset);
+	put_cpu();
+	return time;
 }
 
 /*
-- 
2.26.2