On 04/02/16 19:23, Christoffer Dall wrote:

On Wed, Feb 03, 2016 at 06:00:13PM +0000, Marc Zyngier wrote:

quoted

When the kernel is running in HYP (with VHE), it is necessary to
include EL2 events if the user requests counting kernel or
hypervisor events.

Signed-off-by: Marc Zyngier <redacted>
---
 arch/arm64/kernel/perf_event.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c
index f7ab14c..6013a38 100644
--- a/arch/arm64/kernel/perf_event.c
+++ b/arch/arm64/kernel/perf_event.c

@@ -20,6 +20,7 @@
  */
 
 #include <asm/irq_regs.h>
+#include <asm/virt.h>
 
 #include <linux/of.h>
 #include <linux/perf/arm_pmu.h>

@@ -693,10 +694,15 @@ static int armv8pmu_set_event_filter(struct hw_perf_event *event,
 		return -EPERM;
 	if (attr->exclude_user)
 		config_base |= ARMV8_EXCLUDE_EL0;
-	if (attr->exclude_kernel)
-		config_base |= ARMV8_EXCLUDE_EL1;
-	if (!attr->exclude_hv)
-		config_base |= ARMV8_INCLUDE_EL2;
+	if (is_kernel_in_hyp_mode()) {
+		if (!attr->exclude_kernel || !attr->exclude_hv)
+			config_base |= ARMV8_INCLUDE_EL2;

is exclude_hv a perf generic construct?

Yes. It seems to be universally used on all architectures.

is it really correct to count everything the host kernel does (even
unrelated to running VMs) when counting hypervisor events?

With VHE, the roles are blurred, and the PMU can only count events
occurring when at a given exception level. It has no notion of context.

Other than my ignorance in this area, this patch looks fine to me.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...