Greg,
On 07/01/2020 20:55, Greg Kroah-Hartman wrote:
From: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
[ Upstream commit 730766bae3280a25d40ea76a53dc6342e84e6513 ]
During a perf session we try to allocate buffers on the "node" associated
with the CPU the event is bound to. If it is not bound to a CPU, we
use the current CPU node, using smp_processor_id(). However this is unsafe
in a pre-emptible context and could generate the splats as below :
BUG: using smp_processor_id() in preemptible [00000000] code: perf/2544
Use NUMA_NO_NODE hint instead of using the current node for events
not bound to CPUs.
Fixes: 2997aa4063d97fdb39 ("coresight: etb10: implementing AUX API")
Cc: Mathieu Poirier <mathieu.poirier@xxxxxxxxxx>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
Cc: stable <stable@xxxxxxxxxxxxxxx> # 4.6+
Signed-off-by: Mathieu Poirier <mathieu.poirier@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20190620221237.3536-5-mathieu.poirier@xxxxxxxxxx
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
drivers/hwtracing/coresight/coresight-etb10.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c
index 0dad8626bcfb..0a59bf3af40b 100644
--- a/drivers/hwtracing/coresight/coresight-etb10.c
+++ b/drivers/hwtracing/coresight/coresight-etb10.c
@@ -275,9 +275,7 @@ static void *etb_alloc_buffer(struct coresight_device *csdev, int cpu,
int node;
struct cs_buffers *buf;
- if (cpu == -1)
- cpu = smp_processor_id();
- node = cpu_to_node(cpu);
+ node = (event->cpu == -1) ? NUMA_NO_NODE : cpu_to_node(event->cpu);
Please drop this patch too, from the list as it will break the build
with undefined "event" variable. I will post a backport soon.
Suzuki
