On 8/2/24 8:16 AM, kan.liang@xxxxxxxxxxxxxxx wrote:
> From: Kan Liang <kan.liang@xxxxxxxxxxxxxxx>
>
> The idxd PMU is system-wide scope, which is supported by the generic
> perf_event subsystem now.
>
> Set the scope for the idxd PMU and remove all the cpumask and hotplug
> codes.
>
> Signed-off-by: Kan Liang <kan.liang@xxxxxxxxxxxxxxx>
> Cc: Fenghua Yu <fenghua.yu@xxxxxxxxx>
> Cc: Dave Jiang <dave.jiang@xxxxxxxxx>
> Cc: Vinod Koul <vkoul@xxxxxxxxxx>
> Cc: dmaengine@xxxxxxxxxxxxxxx
Cc: Tom Zanussi <tom.zanussi@xxxxxxxxx>
Reviewed-by: Dave Jiang <dave.jiang@xxxxxxxxx>
> ---
> drivers/dma/idxd/idxd.h | 7 ---
> drivers/dma/idxd/init.c | 3 --
> drivers/dma/idxd/perfmon.c | 98 +-------------------------------------
> 3 files changed, 1 insertion(+), 107 deletions(-)
>
> diff --git a/drivers/dma/idxd/idxd.h b/drivers/dma/idxd/idxd.h
> index 868b724a3b75..d84e21daa991 100644
> --- a/drivers/dma/idxd/idxd.h
> +++ b/drivers/dma/idxd/idxd.h
> @@ -124,7 +124,6 @@ struct idxd_pmu {
>
> struct pmu pmu;
> char name[IDXD_NAME_SIZE];
> - int cpu;
>
> int n_counters;
> int counter_width;
> @@ -135,8 +134,6 @@ struct idxd_pmu {
>
> unsigned long supported_filters;
> int n_filters;
> -
> - struct hlist_node cpuhp_node;
> };
>
> #define IDXD_MAX_PRIORITY 0xf
> @@ -803,14 +800,10 @@ void idxd_user_counter_increment(struct idxd_wq *wq, u32 pasid, int index);
> int perfmon_pmu_init(struct idxd_device *idxd);
> void perfmon_pmu_remove(struct idxd_device *idxd);
> void perfmon_counter_overflow(struct idxd_device *idxd);
> -void perfmon_init(void);
> -void perfmon_exit(void);
> #else
> static inline int perfmon_pmu_init(struct idxd_device *idxd) { return 0; }
> static inline void perfmon_pmu_remove(struct idxd_device *idxd) {}
> static inline void perfmon_counter_overflow(struct idxd_device *idxd) {}
> -static inline void perfmon_init(void) {}
> -static inline void perfmon_exit(void) {}
> #endif
>
> /* debugfs */
> diff --git a/drivers/dma/idxd/init.c b/drivers/dma/idxd/init.c
> index 21f6905b554d..5725ea82c409 100644
> --- a/drivers/dma/idxd/init.c
> +++ b/drivers/dma/idxd/init.c
> @@ -878,8 +878,6 @@ static int __init idxd_init_module(void)
> else
> support_enqcmd = true;
>
> - perfmon_init();
> -
> err = idxd_driver_register(&idxd_drv);
> if (err < 0)
> goto err_idxd_driver_register;
> @@ -928,7 +926,6 @@ static void __exit idxd_exit_module(void)
> idxd_driver_unregister(&idxd_drv);
> pci_unregister_driver(&idxd_pci_driver);
> idxd_cdev_remove();
> - perfmon_exit();
> idxd_remove_debugfs();
> }
> module_exit(idxd_exit_module);
> diff --git a/drivers/dma/idxd/perfmon.c b/drivers/dma/idxd/perfmon.c
> index 5e94247e1ea7..f511cf15845b 100644
> --- a/drivers/dma/idxd/perfmon.c
> +++ b/drivers/dma/idxd/perfmon.c
> @@ -6,29 +6,6 @@
> #include "idxd.h"
> #include "perfmon.h"
>
> -static ssize_t cpumask_show(struct device *dev, struct device_attribute *attr,
> - char *buf);
> -
> -static cpumask_t perfmon_dsa_cpu_mask;
> -static bool cpuhp_set_up;
> -static enum cpuhp_state cpuhp_slot;
> -
> -/*
> - * perf userspace reads this attribute to determine which cpus to open
> - * counters on. It's connected to perfmon_dsa_cpu_mask, which is
> - * maintained by the cpu hotplug handlers.
> - */
> -static DEVICE_ATTR_RO(cpumask);
> -
> -static struct attribute *perfmon_cpumask_attrs[] = {
> - &dev_attr_cpumask.attr,
> - NULL,
> -};
> -
> -static struct attribute_group cpumask_attr_group = {
> - .attrs = perfmon_cpumask_attrs,
> -};
> -
> /*
> * These attributes specify the bits in the config word that the perf
> * syscall uses to pass the event ids and categories to perfmon.
> @@ -67,16 +44,9 @@ static struct attribute_group perfmon_format_attr_group = {
>
> static const struct attribute_group *perfmon_attr_groups[] = {
> &perfmon_format_attr_group,
> - &cpumask_attr_group,
> NULL,
> };
>
> -static ssize_t cpumask_show(struct device *dev, struct device_attribute *attr,
> - char *buf)
> -{
> - return cpumap_print_to_pagebuf(true, buf, &perfmon_dsa_cpu_mask);
> -}
> -
> static bool is_idxd_event(struct idxd_pmu *idxd_pmu, struct perf_event *event)
> {
> return &idxd_pmu->pmu == event->pmu;
> @@ -217,7 +187,6 @@ static int perfmon_pmu_event_init(struct perf_event *event)
> return -EINVAL;
>
> event->hw.event_base = ioread64(PERFMON_TABLE_OFFSET(idxd));
> - event->cpu = idxd->idxd_pmu->cpu;
> event->hw.config = event->attr.config;
>
> if (event->group_leader != event)
> @@ -488,6 +457,7 @@ static void idxd_pmu_init(struct idxd_pmu *idxd_pmu)
> idxd_pmu->pmu.stop = perfmon_pmu_event_stop;
> idxd_pmu->pmu.read = perfmon_pmu_event_update;
> idxd_pmu->pmu.capabilities = PERF_PMU_CAP_NO_EXCLUDE;
> + idxd_pmu->pmu.scope = PERF_PMU_SCOPE_SYS_WIDE;
> idxd_pmu->pmu.module = THIS_MODULE;
> }
>
> @@ -496,59 +466,17 @@ void perfmon_pmu_remove(struct idxd_device *idxd)
> if (!idxd->idxd_pmu)
> return;
>
> - cpuhp_state_remove_instance(cpuhp_slot, &idxd->idxd_pmu->cpuhp_node);
> perf_pmu_unregister(&idxd->idxd_pmu->pmu);
> kfree(idxd->idxd_pmu);
> idxd->idxd_pmu = NULL;
> }
>
> -static int perf_event_cpu_online(unsigned int cpu, struct hlist_node *node)
> -{
> - struct idxd_pmu *idxd_pmu;
> -
> - idxd_pmu = hlist_entry_safe(node, typeof(*idxd_pmu), cpuhp_node);
> -
> - /* select the first online CPU as the designated reader */
> - if (cpumask_empty(&perfmon_dsa_cpu_mask)) {
> - cpumask_set_cpu(cpu, &perfmon_dsa_cpu_mask);
> - idxd_pmu->cpu = cpu;
> - }
> -
> - return 0;
> -}
> -
> -static int perf_event_cpu_offline(unsigned int cpu, struct hlist_node *node)
> -{
> - struct idxd_pmu *idxd_pmu;
> - unsigned int target;
> -
> - idxd_pmu = hlist_entry_safe(node, typeof(*idxd_pmu), cpuhp_node);
> -
> - if (!cpumask_test_and_clear_cpu(cpu, &perfmon_dsa_cpu_mask))
> - return 0;
> -
> - target = cpumask_any_but(cpu_online_mask, cpu);
> - /* migrate events if there is a valid target */
> - if (target < nr_cpu_ids) {
> - cpumask_set_cpu(target, &perfmon_dsa_cpu_mask);
> - perf_pmu_migrate_context(&idxd_pmu->pmu, cpu, target);
> - }
> -
> - return 0;
> -}
> -
> int perfmon_pmu_init(struct idxd_device *idxd)
> {
> union idxd_perfcap perfcap;
> struct idxd_pmu *idxd_pmu;
> int rc = -ENODEV;
>
> - /*
> - * perfmon module initialization failed, nothing to do
> - */
> - if (!cpuhp_set_up)
> - return -ENODEV;
> -
> /*
> * If perfmon_offset or num_counters is 0, it means perfmon is
> * not supported on this hardware.
> @@ -624,11 +552,6 @@ int perfmon_pmu_init(struct idxd_device *idxd)
> if (rc)
> goto free;
>
> - rc = cpuhp_state_add_instance(cpuhp_slot, &idxd_pmu->cpuhp_node);
> - if (rc) {
> - perf_pmu_unregister(&idxd->idxd_pmu->pmu);
> - goto free;
> - }
> out:
> return rc;
> free:
> @@ -637,22 +560,3 @@ int perfmon_pmu_init(struct idxd_device *idxd)
>
> goto out;
> }
> -
> -void __init perfmon_init(void)
> -{
> - int rc = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN,
> - "driver/dma/idxd/perf:online",
> - perf_event_cpu_online,
> - perf_event_cpu_offline);
> - if (WARN_ON(rc < 0))
> - return;
> -
> - cpuhp_slot = rc;
> - cpuhp_set_up = true;
> -}
> -
> -void __exit perfmon_exit(void)
> -{
> - if (cpuhp_set_up)
> - cpuhp_remove_multi_state(cpuhp_slot);
> -}
