Re: [PATCH v2] drivers/perf: arm_spe: Fix consistency of SYS_PMSCR_EL1.CX

[Suzuki K Poulose]

On 13/07/2022 09:59, Anshuman Khandual wrote:
> The arm_spe_pmu driver will enable SYS_PMSCR_EL1.CX in order to add CONTEXT
> packets into the traces, if the owner of the perf event runs with required
> capabilities i.e CAP_PERFMON or CAP_SYS_ADMIN via perfmon_capable() helper.
>
> The value of this bit is computed in the arm_spe_event_to_pmscr() function
> but the check for capabilities happens in the pmu event init callback i.e
> arm_spe_pmu_event_init(). This suggests that the value of the CX bit should
> remain consistent for the duration of the perf session.
>
> However, the function arm_spe_event_to_pmscr() may be called later during
> the event start callback i.e arm_spe_pmu_start() when the "current" process
> is not the owner of the perf session, hence the CX bit setting is currently
> not consistent.
>
> One way to fix this, is by caching the required value of the CX bit during
> the initialization of the PMU event, so that it remains consistent for the
> duration of the session. It uses currently unused 'event->hw.flags' element
> to cache perfmon_capable() value, which can be referred during event start
> callback to compute SYS_PMSCR_EL1.CX. This ensures consistent availability
> of context packets in the trace as per event owner capabilities.
>
> Cc: Will Deacon <will@xxxxxxxxxx>
> Cc: Mark Rutland <mark.rutland@xxxxxxx>
> Cc: Alexey Budankov <alexey.budankov@xxxxxxxxxxxxxxx>
> Cc: linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> Cc: linux-kernel@xxxxxxxxxxxxxxx
> Fixes: cea7d0d4a59b ("drivers/perf: Open access for CAP_PERFMON privileged process")
> Reported-by: German Gomez <german.gomez@xxxxxxx>
> Signed-off-by: Anshuman Khandual <anshuman.khandual@xxxxxxx>
> ---
> Changes in V2:
>
> - Moved CONFIG_PID_IN_CONTEXTIDR config check inside the helper per Suzuki
> - Changed the comment per Suzuki
> - Renamed the helpers Per Suzuki
> - Added "Fixes: " tag per German
>
> Changes in V1:
>
> https://lore.kernel.org/all/20220712051404.2546851-1-anshuman.khandual@xxxxxxx/
>
>   drivers/perf/arm_spe_pmu.c | 23 +++++++++++++++++++++--
>   1 file changed, 21 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c
> index db670b265897..c4290b0492fd 100644
> --- a/drivers/perf/arm_spe_pmu.c
> +++ b/drivers/perf/arm_spe_pmu.c
> @@ -39,6 +39,24 @@
>   #include <asm/mmu.h>
>   #include <asm/sysreg.h>
>   
> +/*
> + * Cache if the event is allowed to trace Context information.
> + * This allows us to perform the check, i.e, perfmon_capable(),
> + * in the context of the event owner, once, during the event_init().
> + */
> +#define SPE_PMU_HW_FLAGS_CX			BIT(0)
> +
> +static void set_spe_event_has_cx(struct perf_event *event)
> +{
> +	if (IS_ENABLED(CONFIG_PID_IN_CONTEXTIDR) && perfmon_capable())
> +		event->hw.flags |= SPE_PMU_HW_FLAGS_CX;
> +}
> +
> +static bool get_spe_event_has_cx(struct perf_event *event)
> +{
> +	return !!(event->hw.flags & SPE_PMU_HW_FLAGS_CX);
> +}
> +
>   #define ARM_SPE_BUF_PAD_BYTE			0
>   
>   struct arm_spe_pmu_buf {
> @@ -272,7 +290,7 @@ static u64 arm_spe_event_to_pmscr(struct perf_event *event)
>   	if (!attr->exclude_kernel)
>   		reg |= BIT(SYS_PMSCR_EL1_E1SPE_SHIFT);
>   
> -	if (IS_ENABLED(CONFIG_PID_IN_CONTEXTIDR) && perfmon_capable())
> +	if (get_spe_event_has_cx(event))
>   		reg |= BIT(SYS_PMSCR_EL1_CX_SHIFT);
>   
>   	return reg;
> @@ -710,7 +728,8 @@ static int arm_spe_pmu_event_init(struct perf_event *event)
>   		return -EOPNOTSUPP;
>   



>   	reg = arm_spe_event_to_pmscr(event);
> -	if (!perfmon_capable() &&
> +	set_spe_event_has_cx(event);

This seems to be wrong. We need to set the event_has_cx() *before*
we call arm_spe_event_to_pmscr(), as the latter uses
get_spe_event_has_cx().

> +	if (!get_spe_event_has_cx(event) &&
>   	    (reg & (BIT(SYS_PMSCR_EL1_PA_SHIFT) |

And we must retain the perfmon_capable() check here to ensure that any 
of the following options are usable without CX. e.g,
if CONFIG_PID_IN_CONTEXTIDR is not enabled, !get_spe_event_has_cx() 
doesn't imply !perfmon_capable().


>   		    BIT(SYS_PMSCR_EL1_CX_SHIFT) |
>   		    BIT(SYS_PMSCR_EL1_PCT_SHIFT))))

Suzuki