diff options
Diffstat (limited to 'drivers/perf')
| -rw-r--r-- | drivers/perf/amlogic/meson_ddr_pmu_core.c | 2 | ||||
| -rw-r--r-- | drivers/perf/apple_m1_cpu_pmu.c | 105 | ||||
| -rw-r--r-- | drivers/perf/arm-ccn.c | 5 | ||||
| -rw-r--r-- | drivers/perf/arm-cmn.c | 16 | ||||
| -rw-r--r-- | drivers/perf/arm-ni.c | 42 | ||||
| -rw-r--r-- | drivers/perf/arm_cspmu/ampere_cspmu.c | 32 | ||||
| -rw-r--r-- | drivers/perf/arm_cspmu/arm_cspmu.c | 81 | ||||
| -rw-r--r-- | drivers/perf/arm_cspmu/arm_cspmu.h | 57 | ||||
| -rw-r--r-- | drivers/perf/arm_cspmu/nvidia_cspmu.c | 22 | ||||
| -rw-r--r-- | drivers/perf/arm_pmuv3.c | 11 | ||||
| -rw-r--r-- | drivers/perf/arm_spe_pmu.c | 4 | ||||
| -rw-r--r-- | drivers/perf/arm_v7_pmu.c | 50 | ||||
| -rw-r--r-- | drivers/perf/dwc_pcie_pmu.c | 25 | ||||
| -rw-r--r-- | drivers/perf/marvell_cn10k_ddr_pmu.c | 4 | ||||
| -rw-r--r-- | drivers/perf/thunderx2_pmu.c | 5 |
15 files changed, 230 insertions, 231 deletions
diff --git a/drivers/perf/amlogic/meson_ddr_pmu_core.c b/drivers/perf/amlogic/meson_ddr_pmu_core.c index 07446d784a1a..c1e755c356a3 100644 --- a/drivers/perf/amlogic/meson_ddr_pmu_core.c +++ b/drivers/perf/amlogic/meson_ddr_pmu_core.c @@ -511,7 +511,7 @@ int meson_ddr_pmu_create(struct platform_device *pdev) fmt_attr_fill(pmu->info.hw_info->fmt_attr); - pmu->cpu = smp_processor_id(); + pmu->cpu = raw_smp_processor_id(); name = devm_kasprintf(&pdev->dev, GFP_KERNEL, DDR_PERF_DEV_NAME); if (!name) diff --git a/drivers/perf/apple_m1_cpu_pmu.c b/drivers/perf/apple_m1_cpu_pmu.c index 06fd317529fc..df9a28ba69dc 100644 --- a/drivers/perf/apple_m1_cpu_pmu.c +++ b/drivers/perf/apple_m1_cpu_pmu.c @@ -12,6 +12,7 @@ #include <linux/of.h> #include <linux/perf/arm_pmu.h> +#include <linux/perf/arm_pmuv3.h> #include <linux/platform_device.h> #include <asm/apple_m1_pmu.h> @@ -120,6 +121,8 @@ enum m1_pmu_events { */ M1_PMU_CFG_COUNT_USER = BIT(8), M1_PMU_CFG_COUNT_KERNEL = BIT(9), + M1_PMU_CFG_COUNT_HOST = BIT(10), + M1_PMU_CFG_COUNT_GUEST = BIT(11), }; /* @@ -172,6 +175,17 @@ static const unsigned m1_pmu_perf_map[PERF_COUNT_HW_MAX] = { [PERF_COUNT_HW_BRANCH_MISSES] = M1_PMU_PERFCTR_BRANCH_MISPRED_NONSPEC, }; +#define M1_PMUV3_EVENT_MAP(pmuv3_event, m1_event) \ + [ARMV8_PMUV3_PERFCTR_##pmuv3_event] = M1_PMU_PERFCTR_##m1_event + +static const u16 m1_pmu_pmceid_map[ARMV8_PMUV3_MAX_COMMON_EVENTS] = { + [0 ... ARMV8_PMUV3_MAX_COMMON_EVENTS - 1] = HW_OP_UNSUPPORTED, + M1_PMUV3_EVENT_MAP(INST_RETIRED, INST_ALL), + M1_PMUV3_EVENT_MAP(CPU_CYCLES, CORE_ACTIVE_CYCLE), + M1_PMUV3_EVENT_MAP(BR_RETIRED, INST_BRANCH), + M1_PMUV3_EVENT_MAP(BR_MIS_PRED_RETIRED, BRANCH_MISPRED_NONSPEC), +}; + /* sysfs definitions */ static ssize_t m1_pmu_events_sysfs_show(struct device *dev, struct device_attribute *attr, @@ -327,11 +341,10 @@ static void m1_pmu_disable_counter_interrupt(unsigned int index) __m1_pmu_enable_counter_interrupt(index, false); } -static void m1_pmu_configure_counter(unsigned int index, u8 event, - bool user, bool kernel) +static void __m1_pmu_configure_event_filter(unsigned int index, bool user, + bool kernel, bool host) { - u64 val, user_bit, kernel_bit; - int shift; + u64 clear, set, user_bit, kernel_bit; switch (index) { case 0 ... 7: @@ -346,19 +359,27 @@ static void m1_pmu_configure_counter(unsigned int index, u8 event, BUG(); } - val = read_sysreg_s(SYS_IMP_APL_PMCR1_EL1); - + clear = set = 0; if (user) - val |= user_bit; + set |= user_bit; else - val &= ~user_bit; + clear |= user_bit; if (kernel) - val |= kernel_bit; + set |= kernel_bit; else - val &= ~kernel_bit; + clear |= kernel_bit; + + if (host) + sysreg_clear_set_s(SYS_IMP_APL_PMCR1_EL1, clear, set); + else if (is_kernel_in_hyp_mode()) + sysreg_clear_set_s(SYS_IMP_APL_PMCR1_EL12, clear, set); +} - write_sysreg_s(val, SYS_IMP_APL_PMCR1_EL1); +static void __m1_pmu_configure_eventsel(unsigned int index, u8 event) +{ + u64 clear = 0, set = 0; + int shift; /* * Counters 0 and 1 have fixed events. For anything else, @@ -371,21 +392,32 @@ static void m1_pmu_configure_counter(unsigned int index, u8 event, break; case 2 ... 5: shift = (index - 2) * 8; - val = read_sysreg_s(SYS_IMP_APL_PMESR0_EL1); - val &= ~((u64)0xff << shift); - val |= (u64)event << shift; - write_sysreg_s(val, SYS_IMP_APL_PMESR0_EL1); + clear |= (u64)0xff << shift; + set |= (u64)event << shift; + sysreg_clear_set_s(SYS_IMP_APL_PMESR0_EL1, clear, set); break; case 6 ... 9: shift = (index - 6) * 8; - val = read_sysreg_s(SYS_IMP_APL_PMESR1_EL1); - val &= ~((u64)0xff << shift); - val |= (u64)event << shift; - write_sysreg_s(val, SYS_IMP_APL_PMESR1_EL1); + clear |= (u64)0xff << shift; + set |= (u64)event << shift; + sysreg_clear_set_s(SYS_IMP_APL_PMESR1_EL1, clear, set); break; } } +static void m1_pmu_configure_counter(unsigned int index, unsigned long config_base) +{ + bool kernel = config_base & M1_PMU_CFG_COUNT_KERNEL; + bool guest = config_base & M1_PMU_CFG_COUNT_GUEST; + bool host = config_base & M1_PMU_CFG_COUNT_HOST; + bool user = config_base & M1_PMU_CFG_COUNT_USER; + u8 evt = config_base & M1_PMU_CFG_EVENT; + + __m1_pmu_configure_event_filter(index, user && host, kernel && host, true); + __m1_pmu_configure_event_filter(index, user && guest, kernel && guest, false); + __m1_pmu_configure_eventsel(index, evt); +} + /* arm_pmu backend */ static void m1_pmu_enable_event(struct perf_event *event) { @@ -396,11 +428,7 @@ static void m1_pmu_enable_event(struct perf_event *event) user = event->hw.config_base & M1_PMU_CFG_COUNT_USER; kernel = event->hw.config_base & M1_PMU_CFG_COUNT_KERNEL; - m1_pmu_disable_counter_interrupt(event->hw.idx); - m1_pmu_disable_counter(event->hw.idx); - isb(); - - m1_pmu_configure_counter(event->hw.idx, evt, user, kernel); + m1_pmu_configure_counter(event->hw.idx, event->hw.config_base); m1_pmu_enable_counter(event->hw.idx); m1_pmu_enable_counter_interrupt(event->hw.idx); isb(); @@ -538,6 +566,26 @@ static int m2_pmu_map_event(struct perf_event *event) return armpmu_map_event(event, &m1_pmu_perf_map, NULL, M1_PMU_CFG_EVENT); } +static int m1_pmu_map_pmuv3_event(unsigned int eventsel) +{ + u16 m1_event = HW_OP_UNSUPPORTED; + + if (eventsel < ARMV8_PMUV3_MAX_COMMON_EVENTS) + m1_event = m1_pmu_pmceid_map[eventsel]; + + return m1_event == HW_OP_UNSUPPORTED ? -EOPNOTSUPP : m1_event; +} + +static void m1_pmu_init_pmceid(struct arm_pmu *pmu) +{ + unsigned int event; + + for (event = 0; event < ARMV8_PMUV3_MAX_COMMON_EVENTS; event++) { + if (m1_pmu_map_pmuv3_event(event) >= 0) + set_bit(event, pmu->pmceid_bitmap); + } +} + static void m1_pmu_reset(void *info) { int i; @@ -558,7 +606,7 @@ static int m1_pmu_set_event_filter(struct hw_perf_event *event, { unsigned long config_base = 0; - if (!attr->exclude_guest) { + if (!attr->exclude_guest && !is_kernel_in_hyp_mode()) { pr_debug("ARM performance counters do not support mode exclusion\n"); return -EOPNOTSUPP; } @@ -566,6 +614,10 @@ static int m1_pmu_set_event_filter(struct hw_perf_event *event, config_base |= M1_PMU_CFG_COUNT_KERNEL; if (!attr->exclude_user) config_base |= M1_PMU_CFG_COUNT_USER; + if (!attr->exclude_host) + config_base |= M1_PMU_CFG_COUNT_HOST; + if (!attr->exclude_guest) + config_base |= M1_PMU_CFG_COUNT_GUEST; event->config_base = config_base; @@ -594,6 +646,9 @@ static int m1_pmu_init(struct arm_pmu *cpu_pmu, u32 flags) cpu_pmu->reset = m1_pmu_reset; cpu_pmu->set_event_filter = m1_pmu_set_event_filter; + cpu_pmu->map_pmuv3_event = m1_pmu_map_pmuv3_event; + m1_pmu_init_pmceid(cpu_pmu); + bitmap_set(cpu_pmu->cntr_mask, 0, M1_PMU_NR_COUNTERS); cpu_pmu->attr_groups[ARMPMU_ATTR_GROUP_EVENTS] = &m1_pmu_events_attr_group; cpu_pmu->attr_groups[ARMPMU_ATTR_GROUP_FORMATS] = &m1_pmu_format_attr_group; diff --git a/drivers/perf/arm-ccn.c b/drivers/perf/arm-ccn.c index d5fcea3d4328..1a0d0e1a2263 100644 --- a/drivers/perf/arm-ccn.c +++ b/drivers/perf/arm-ccn.c @@ -1273,9 +1273,8 @@ static int arm_ccn_pmu_init(struct arm_ccn *ccn) /* No overflow interrupt? Have to use a timer instead. */ if (!ccn->irq) { dev_info(ccn->dev, "No access to interrupts, using timer.\n"); - hrtimer_init(&ccn->dt.hrtimer, CLOCK_MONOTONIC, - HRTIMER_MODE_REL); - ccn->dt.hrtimer.function = arm_ccn_pmu_timer_handler; + hrtimer_setup(&ccn->dt.hrtimer, arm_ccn_pmu_timer_handler, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); } /* Pick one CPU which we will use to collect data from CCN... */ diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index ef959e66db7c..403850b1040d 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -727,8 +727,8 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, if ((chan == 5 && cmn->rsp_vc_num < 2) || (chan == 6 && cmn->dat_vc_num < 2) || - (chan == 7 && cmn->snp_vc_num < 2) || - (chan == 8 && cmn->req_vc_num < 2)) + (chan == 7 && cmn->req_vc_num < 2) || + (chan == 8 && cmn->snp_vc_num < 2)) return 0; } @@ -802,8 +802,6 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, CMN_EVENT_ATTR(_model, ccha_##_name, CMN_TYPE_CCHA, _event) #define CMN_EVENT_CCLA(_name, _event) \ CMN_EVENT_ATTR(CMN_ANY, ccla_##_name, CMN_TYPE_CCLA, _event) -#define CMN_EVENT_CCLA_RNI(_name, _event) \ - CMN_EVENT_ATTR(CMN_ANY, ccla_rni_##_name, CMN_TYPE_CCLA_RNI, _event) #define CMN_EVENT_HNS(_name, _event) \ CMN_EVENT_ATTR(CMN_ANY, hns_##_name, CMN_TYPE_HNS, _event) @@ -884,8 +882,8 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, _CMN_EVENT_XP(pub_##_name, (_event) | (4 << 5)), \ _CMN_EVENT_XP(rsp2_##_name, (_event) | (5 << 5)), \ _CMN_EVENT_XP(dat2_##_name, (_event) | (6 << 5)), \ - _CMN_EVENT_XP(snp2_##_name, (_event) | (7 << 5)), \ - _CMN_EVENT_XP(req2_##_name, (_event) | (8 << 5)) + _CMN_EVENT_XP(req2_##_name, (_event) | (7 << 5)), \ + _CMN_EVENT_XP(snp2_##_name, (_event) | (8 << 5)) #define CMN_EVENT_XP_DAT(_name, _event) \ _CMN_EVENT_XP_PORT(dat_##_name, (_event) | (3 << 5)), \ @@ -1798,6 +1796,9 @@ static int arm_cmn_event_init(struct perf_event *event) } else if (type == CMN_TYPE_XP && (cmn->part == PART_CMN700 || cmn->part == PART_CMN_S3)) { hw->wide_sel = true; + } else if (type == CMN_TYPE_RND) { + /* Secretly permit this as an alias for "rnid" events */ + type = CMN_TYPE_RNI; } /* This is sufficiently annoying to recalculate, so cache it */ @@ -2557,6 +2558,7 @@ static int arm_cmn_probe(struct platform_device *pdev) cmn->dev = &pdev->dev; cmn->part = (unsigned long)device_get_match_data(cmn->dev); + cmn->cpu = cpumask_local_spread(0, dev_to_node(cmn->dev)); platform_set_drvdata(pdev, cmn); if (cmn->part == PART_CMN600 && has_acpi_companion(cmn->dev)) { @@ -2584,7 +2586,6 @@ static int arm_cmn_probe(struct platform_device *pdev) if (err) return err; - cmn->cpu = cpumask_local_spread(0, dev_to_node(cmn->dev)); cmn->pmu = (struct pmu) { .module = THIS_MODULE, .parent = cmn->dev, @@ -2650,6 +2651,7 @@ static const struct acpi_device_id arm_cmn_acpi_match[] = { { "ARMHC600", PART_CMN600 }, { "ARMHC650" }, { "ARMHC700" }, + { "ARMHC003" }, {} }; MODULE_DEVICE_TABLE(acpi, arm_cmn_acpi_match); diff --git a/drivers/perf/arm-ni.c b/drivers/perf/arm-ni.c index fd7a5e60e963..9396d243415f 100644 --- a/drivers/perf/arm-ni.c +++ b/drivers/perf/arm-ni.c @@ -544,6 +544,8 @@ static int arm_ni_init_cd(struct arm_ni *ni, struct arm_ni_node *node, u64 res_s return err; cd->cpu = cpumask_local_spread(0, dev_to_node(ni->dev)); + irq_set_affinity(cd->irq, cpumask_of(cd->cpu)); + cd->pmu = (struct pmu) { .module = THIS_MODULE, .parent = ni->dev, @@ -575,6 +577,23 @@ static int arm_ni_init_cd(struct arm_ni *ni, struct arm_ni_node *node, u64 res_s return err; } +static void arm_ni_remove(struct platform_device *pdev) +{ + struct arm_ni *ni = platform_get_drvdata(pdev); + + for (int i = 0; i < ni->num_cds; i++) { + struct arm_ni_cd *cd = ni->cds + i; + + if (!cd->pmu_base) + continue; + + writel_relaxed(0, cd->pmu_base + NI_PMCR); + writel_relaxed(U32_MAX, cd->pmu_base + NI_PMINTENCLR); + perf_pmu_unregister(&cd->pmu); + cpuhp_state_remove_instance_nocalls(arm_ni_hp_state, &cd->cpuhp_node); + } +} + static void arm_ni_probe_domain(void __iomem *base, struct arm_ni_node *node) { u32 reg = readl_relaxed(base + NI_NODE_TYPE); @@ -643,6 +662,7 @@ static int arm_ni_probe(struct platform_device *pdev) ni->num_cds = num_cds; ni->part = part; ni->id = atomic_fetch_inc(&id); + platform_set_drvdata(pdev, ni); for (int v = 0; v < cfg.num_components; v++) { reg = readl_relaxed(cfg.base + NI_CHILD_PTR(v)); @@ -656,8 +676,11 @@ static int arm_ni_probe(struct platform_device *pdev) reg = readl_relaxed(pd.base + NI_CHILD_PTR(c)); arm_ni_probe_domain(base + reg, &cd); ret = arm_ni_init_cd(ni, &cd, res->start); - if (ret) + if (ret) { + ni->cds[cd.id].pmu_base = NULL; + arm_ni_remove(pdev); return ret; + } } } } @@ -665,23 +688,6 @@ static int arm_ni_probe(struct platform_device *pdev) return 0; } -static void arm_ni_remove(struct platform_device *pdev) -{ - struct arm_ni *ni = platform_get_drvdata(pdev); - - for (int i = 0; i < ni->num_cds; i++) { - struct arm_ni_cd *cd = ni->cds + i; - - if (!cd->pmu_base) - continue; - - writel_relaxed(0, cd->pmu_base + NI_PMCR); - writel_relaxed(U32_MAX, cd->pmu_base + NI_PMINTENCLR); - perf_pmu_unregister(&cd->pmu); - cpuhp_state_remove_instance_nocalls(arm_ni_hp_state, &cd->cpuhp_node); - } -} - #ifdef CONFIG_OF static const struct of_device_id arm_ni_of_match[] = { { .compatible = "arm,ni-700" }, diff --git a/drivers/perf/arm_cspmu/ampere_cspmu.c b/drivers/perf/arm_cspmu/ampere_cspmu.c index f72f5689923c..b8ca69fd9d1d 100644 --- a/drivers/perf/arm_cspmu/ampere_cspmu.c +++ b/drivers/perf/arm_cspmu/ampere_cspmu.c @@ -10,10 +10,10 @@ #include "arm_cspmu.h" -#define PMAUXR0 0xD80 -#define PMAUXR1 0xD84 -#define PMAUXR2 0xD88 -#define PMAUXR3 0xD8C +#define PMAUXR0 PMIMPDEF +#define PMAUXR1 (PMIMPDEF + 0x4) +#define PMAUXR2 (PMIMPDEF + 0x8) +#define PMAUXR3 (PMIMPDEF + 0xC) #define to_ampere_cspmu_ctx(cspmu) ((struct ampere_cspmu_ctx *)(cspmu->impl.ctx)) @@ -132,32 +132,20 @@ ampere_cspmu_get_name(const struct arm_cspmu *cspmu) return ctx->name; } -static u32 ampere_cspmu_event_filter(const struct perf_event *event) +static void ampere_cspmu_set_cc_filter(struct arm_cspmu *cspmu, + const struct perf_event *event) { /* - * PMEVFILTR or PMCCFILTR aren't used in Ampere SoC PMU but are marked - * as RES0. Make sure, PMCCFILTR is written zero. + * PMCCFILTR is RES0, so this is just a dummy callback to override + * the default implementation and avoid writing to it. */ - return 0; } static void ampere_cspmu_set_ev_filter(struct arm_cspmu *cspmu, - struct hw_perf_event *hwc, - u32 filter) + const struct perf_event *event) { - struct perf_event *event; - unsigned int idx; u32 threshold, rank, bank; - /* - * At this point, all the events have the same filter settings. - * Therefore, take the first event and use its configuration. - */ - idx = find_first_bit(cspmu->hw_events.used_ctrs, - cspmu->cycle_counter_logical_idx); - - event = cspmu->hw_events.events[idx]; - threshold = get_threshold(event); rank = get_rank(event); bank = get_bank(event); @@ -233,7 +221,7 @@ static int ampere_cspmu_init_ops(struct arm_cspmu *cspmu) cspmu->impl.ctx = ctx; - impl_ops->event_filter = ampere_cspmu_event_filter; + impl_ops->set_cc_filter = ampere_cspmu_set_cc_filter; impl_ops->set_ev_filter = ampere_cspmu_set_ev_filter; impl_ops->validate_event = ampere_cspmu_validate_event; impl_ops->get_name = ampere_cspmu_get_name; diff --git a/drivers/perf/arm_cspmu/arm_cspmu.c b/drivers/perf/arm_cspmu/arm_cspmu.c index 81e8b97e9353..efa9b229e701 100644 --- a/drivers/perf/arm_cspmu/arm_cspmu.c +++ b/drivers/perf/arm_cspmu/arm_cspmu.c @@ -40,51 +40,6 @@ ARM_CSPMU_EXT_ATTR(_name, arm_cspmu_cpumask_show, \ (unsigned long)_config) -/* - * CoreSight PMU Arch register offsets. - */ -#define PMEVCNTR_LO 0x0 -#define PMEVCNTR_HI 0x4 -#define PMEVTYPER 0x400 -#define PMCCFILTR 0x47C -#define PMEVFILTR 0xA00 -#define PMCNTENSET 0xC00 -#define PMCNTENCLR 0xC20 -#define PMINTENSET 0xC40 -#define PMINTENCLR 0xC60 -#define PMOVSCLR 0xC80 -#define PMOVSSET 0xCC0 -#define PMCFGR 0xE00 -#define PMCR 0xE04 -#define PMIIDR 0xE08 - -/* PMCFGR register field */ -#define PMCFGR_NCG GENMASK(31, 28) -#define PMCFGR_HDBG BIT(24) -#define PMCFGR_TRO BIT(23) -#define PMCFGR_SS BIT(22) -#define PMCFGR_FZO BIT(21) -#define PMCFGR_MSI BIT(20) -#define PMCFGR_UEN BIT(19) -#define PMCFGR_NA BIT(17) -#define PMCFGR_EX BIT(16) -#define PMCFGR_CCD BIT(15) -#define PMCFGR_CC BIT(14) -#define PMCFGR_SIZE GENMASK(13, 8) -#define PMCFGR_N GENMASK(7, 0) - -/* PMCR register field */ -#define PMCR_TRO BIT(11) -#define PMCR_HDBG BIT(10) -#define PMCR_FZO BIT(9) -#define PMCR_NA BIT(8) -#define PMCR_DP BIT(5) -#define PMCR_X BIT(4) -#define PMCR_D BIT(3) -#define PMCR_C BIT(2) -#define PMCR_P BIT(1) -#define PMCR_E BIT(0) - /* Each SET/CLR register supports up to 32 counters. */ #define ARM_CSPMU_SET_CLR_COUNTER_SHIFT 5 #define ARM_CSPMU_SET_CLR_COUNTER_NUM \ @@ -111,7 +66,9 @@ static unsigned long arm_cspmu_cpuhp_state; static DEFINE_MUTEX(arm_cspmu_lock); static void arm_cspmu_set_ev_filter(struct arm_cspmu *cspmu, - struct hw_perf_event *hwc, u32 filter); + const struct perf_event *event); +static void arm_cspmu_set_cc_filter(struct arm_cspmu *cspmu, + const struct perf_event *event); static struct acpi_apmt_node *arm_cspmu_apmt_node(struct device *dev) { @@ -226,6 +183,7 @@ arm_cspmu_event_attr_is_visible(struct kobject *kobj, static struct attribute *arm_cspmu_format_attrs[] = { ARM_CSPMU_FORMAT_EVENT_ATTR, ARM_CSPMU_FORMAT_FILTER_ATTR, + ARM_CSPMU_FORMAT_FILTER2_ATTR, NULL, }; @@ -250,11 +208,6 @@ static bool arm_cspmu_is_cycle_counter_event(const struct perf_event *event) return (event->attr.config == ARM_CSPMU_EVT_CYCLES_DEFAULT); } -static u32 arm_cspmu_event_filter(const struct perf_event *event) -{ - return event->attr.config1 & ARM_CSPMU_FILTER_MASK; -} - static ssize_t arm_cspmu_identifier_show(struct device *dev, struct device_attribute *attr, char *page) @@ -416,7 +369,7 @@ static int arm_cspmu_init_impl_ops(struct arm_cspmu *cspmu) DEFAULT_IMPL_OP(get_name), DEFAULT_IMPL_OP(is_cycle_counter_event), DEFAULT_IMPL_OP(event_type), - DEFAULT_IMPL_OP(event_filter), + DEFAULT_IMPL_OP(set_cc_filter), DEFAULT_IMPL_OP(set_ev_filter), DEFAULT_IMPL_OP(event_attr_is_visible), }; @@ -812,26 +765,28 @@ static inline void arm_cspmu_set_event(struct arm_cspmu *cspmu, } static void arm_cspmu_set_ev_filter(struct arm_cspmu *cspmu, - struct hw_perf_event *hwc, - u32 filter) + const struct perf_event *event) { - u32 offset = PMEVFILTR + (4 * hwc->idx); + u32 filter = event->attr.config1 & ARM_CSPMU_FILTER_MASK; + u32 filter2 = event->attr.config2 & ARM_CSPMU_FILTER_MASK; + u32 offset = 4 * event->hw.idx; - writel(filter, cspmu->base0 + offset); + writel(filter, cspmu->base0 + PMEVFILTR + offset); + writel(filter2, cspmu->base0 + PMEVFILT2R + offset); } -static inline void arm_cspmu_set_cc_filter(struct arm_cspmu *cspmu, u32 filter) +static void arm_cspmu_set_cc_filter(struct arm_cspmu *cspmu, + const struct perf_event *event) { - u32 offset = PMCCFILTR; + u32 filter = event->attr.config1 & ARM_CSPMU_FILTER_MASK; - writel(filter, cspmu->base0 + offset); + writel(filter, cspmu->base0 + PMCCFILTR); } static void arm_cspmu_start(struct perf_event *event, int pmu_flags) { struct arm_cspmu *cspmu = to_arm_cspmu(event->pmu); struct hw_perf_event *hwc = &event->hw; - u32 filter; /* We always reprogram the counter */ if (pmu_flags & PERF_EF_RELOAD) @@ -839,13 +794,11 @@ static void arm_cspmu_start(struct perf_event *event, int pmu_flags) arm_cspmu_set_event_period(event); - filter = cspmu->impl.ops.event_filter(event); - if (event->hw.extra_reg.idx == cspmu->cycle_counter_logical_idx) { - arm_cspmu_set_cc_filter(cspmu, filter); + cspmu->impl.ops.set_cc_filter(cspmu, event); } else { arm_cspmu_set_event(cspmu, hwc); - cspmu->impl.ops.set_ev_filter(cspmu, hwc, filter); + cspmu->impl.ops.set_ev_filter(cspmu, event); } hwc->state = 0; diff --git a/drivers/perf/arm_cspmu/arm_cspmu.h b/drivers/perf/arm_cspmu/arm_cspmu.h index 2621f3111148..19684b76bd96 100644 --- a/drivers/perf/arm_cspmu/arm_cspmu.h +++ b/drivers/perf/arm_cspmu/arm_cspmu.h @@ -47,6 +47,8 @@ /* Default filter format */ #define ARM_CSPMU_FORMAT_FILTER_ATTR \ ARM_CSPMU_FORMAT_ATTR(filter, "config1:0-31") +#define ARM_CSPMU_FORMAT_FILTER2_ATTR \ + ARM_CSPMU_FORMAT_ATTR(filter2, "config2:0-31") /* * This is the default event number for cycle count, if supported, since the @@ -65,6 +67,53 @@ /* The cycle counter, if implemented, is located at counter[31]. */ #define ARM_CSPMU_CYCLE_CNTR_IDX 31 +/* + * CoreSight PMU Arch register offsets. + */ +#define PMEVCNTR_LO 0x0 +#define PMEVCNTR_HI 0x4 +#define PMEVTYPER 0x400 +#define PMCCFILTR 0x47C +#define PMEVFILT2R 0x800 +#define PMEVFILTR 0xA00 +#define PMCNTENSET 0xC00 +#define PMCNTENCLR 0xC20 +#define PMINTENSET 0xC40 +#define PMINTENCLR 0xC60 +#define PMOVSCLR 0xC80 +#define PMOVSSET 0xCC0 +#define PMIMPDEF 0xD80 +#define PMCFGR 0xE00 +#define PMCR 0xE04 +#define PMIIDR 0xE08 + +/* PMCFGR register field */ +#define PMCFGR_NCG GENMASK(31, 28) +#define PMCFGR_HDBG BIT(24) +#define PMCFGR_TRO BIT(23) +#define PMCFGR_SS BIT(22) +#define PMCFGR_FZO BIT(21) +#define PMCFGR_MSI BIT(20) +#define PMCFGR_UEN BIT(19) +#define PMCFGR_NA BIT(17) +#define PMCFGR_EX BIT(16) +#define PMCFGR_CCD BIT(15) +#define PMCFGR_CC BIT(14) +#define PMCFGR_SIZE GENMASK(13, 8) +#define PMCFGR_N GENMASK(7, 0) + +/* PMCR register field */ +#define PMCR_TRO BIT(11) +#define PMCR_HDBG BIT(10) +#define PMCR_FZO BIT(9) +#define PMCR_NA BIT(8) +#define PMCR_DP BIT(5) +#define PMCR_X BIT(4) +#define PMCR_D BIT(3) +#define PMCR_C BIT(2) +#define PMCR_P BIT(1) +#define PMCR_E BIT(0) + /* PMIIDR register field */ #define ARM_CSPMU_PMIIDR_IMPLEMENTER GENMASK(11, 0) #define ARM_CSPMU_PMIIDR_PRODUCTID GENMASK(31, 20) @@ -103,11 +152,11 @@ struct arm_cspmu_impl_ops { bool (*is_cycle_counter_event)(const struct perf_event *event); /* Decode event type/id from configs */ u32 (*event_type)(const struct perf_event *event); - /* Decode filter value from configs */ - u32 (*event_filter)(const struct perf_event *event); - /* Set event filter */ + /* Set event filters */ + void (*set_cc_filter)(struct arm_cspmu *cspmu, + const struct perf_event *event); void (*set_ev_filter)(struct arm_cspmu *cspmu, - struct hw_perf_event *hwc, u32 filter); + const struct perf_event *event); /* Implementation specific event validation */ int (*validate_event)(struct arm_cspmu *cspmu, struct perf_event *event); diff --git a/drivers/perf/arm_cspmu/nvidia_cspmu.c b/drivers/perf/arm_cspmu/nvidia_cspmu.c index 8116c7846a46..dc6d4e3e2a1b 100644 --- a/drivers/perf/arm_cspmu/nvidia_cspmu.c +++ b/drivers/perf/arm_cspmu/nvidia_cspmu.c @@ -6,6 +6,7 @@ /* Support for NVIDIA specific attributes. */ +#include <linux/io.h> #include <linux/module.h> #include <linux/topology.h> @@ -183,6 +184,24 @@ static u32 nv_cspmu_event_filter(const struct perf_event *event) return filter_val; } +static void nv_cspmu_set_ev_filter(struct arm_cspmu *cspmu, + const struct perf_event *event) +{ + u32 filter = nv_cspmu_event_filter(event); + u32 offset = PMEVFILTR + (4 * event->hw.idx); + + writel(filter, cspmu->base0 + offset); +} + +static void nv_cspmu_set_cc_filter(struct arm_cspmu *cspmu, + const struct perf_event *event) +{ + u32 filter = nv_cspmu_event_filter(event); + + writel(filter, cspmu->base0 + PMCCFILTR); +} + + enum nv_cspmu_name_fmt { NAME_FMT_GENERIC, NAME_FMT_SOCKET @@ -322,7 +341,8 @@ static int nv_cspmu_init_ops(struct arm_cspmu *cspmu) cspmu->impl.ctx = ctx; /* NVIDIA specific callbacks. */ - impl_ops->event_filter = nv_cspmu_event_filter; + impl_ops->set_cc_filter = nv_cspmu_set_cc_filter; + impl_ops->set_ev_filter = nv_cspmu_set_ev_filter; impl_ops->get_event_attrs = nv_cspmu_get_event_attrs; impl_ops->get_format_attrs = nv_cspmu_get_format_attrs; impl_ops->get_name = nv_cspmu_get_name; diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index 0e360feb3432..e506d59654e7 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -795,11 +795,6 @@ static void armv8pmu_enable_user_access(struct arm_pmu *cpu_pmu) static void armv8pmu_enable_event(struct perf_event *event) { - /* - * Enable counter and interrupt, and set the counter to count - * the event that we're interested in. - */ - armv8pmu_disable_event_counter(event); armv8pmu_write_event_type(event); armv8pmu_enable_event_irq(event); armv8pmu_enable_event_counter(event); @@ -825,10 +820,10 @@ static void armv8pmu_start(struct arm_pmu *cpu_pmu) else armv8pmu_disable_user_access(); + kvm_vcpu_pmu_resync_el0(); + /* Enable all counters */ armv8pmu_pmcr_write(armv8pmu_pmcr_read() | ARMV8_PMU_PMCR_E); - - kvm_vcpu_pmu_resync_el0(); } static void armv8pmu_stop(struct arm_pmu *cpu_pmu) @@ -1369,6 +1364,7 @@ PMUV3_INIT_SIMPLE(armv8_neoverse_v1) PMUV3_INIT_SIMPLE(armv8_neoverse_v2) PMUV3_INIT_SIMPLE(armv8_neoverse_v3) PMUV3_INIT_SIMPLE(armv8_neoverse_v3ae) +PMUV3_INIT_SIMPLE(armv8_rainier) PMUV3_INIT_SIMPLE(armv8_nvidia_carmel) PMUV3_INIT_SIMPLE(armv8_nvidia_denver) @@ -1416,6 +1412,7 @@ static const struct of_device_id armv8_pmu_of_device_ids[] = { {.compatible = "arm,neoverse-v2-pmu", .data = armv8_neoverse_v2_pmu_init}, {.compatible = "arm,neoverse-v3-pmu", .data = armv8_neoverse_v3_pmu_init}, {.compatible = "arm,neoverse-v3ae-pmu", .data = armv8_neoverse_v3ae_pmu_init}, + {.compatible = "arm,rainier-pmu", .data = armv8_rainier_pmu_init}, {.compatible = "cavium,thunder-pmu", .data = armv8_cavium_thunder_pmu_init}, {.compatible = "brcm,vulcan-pmu", .data = armv8_brcm_vulcan_pmu_init}, {.compatible = "nvidia,carmel-pmu", .data = armv8_nvidia_carmel_pmu_init}, diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c index f5e6878db9d6..3efed8839a4e 100644 --- a/drivers/perf/arm_spe_pmu.c +++ b/drivers/perf/arm_spe_pmu.c @@ -50,7 +50,7 @@ static_assert((PERF_EVENT_FLAG_ARCH & SPE_PMU_HW_FLAGS_CX) == SPE_PMU_HW_FLAGS_C static void set_spe_event_has_cx(struct perf_event *event) { - if (IS_ENABLED(CONFIG_PID_IN_CONTEXTIDR) && !perf_allow_kernel(&event->attr)) + if (IS_ENABLED(CONFIG_PID_IN_CONTEXTIDR) && !perf_allow_kernel()) event->hw.flags |= SPE_PMU_HW_FLAGS_CX; } @@ -765,7 +765,7 @@ static int arm_spe_pmu_event_init(struct perf_event *event) set_spe_event_has_cx(event); reg = arm_spe_event_to_pmscr(event); if (reg & (PMSCR_EL1_PA | PMSCR_EL1_PCT)) - return perf_allow_kernel(&event->attr); + return perf_allow_kernel(); return 0; } diff --git a/drivers/perf/arm_v7_pmu.c b/drivers/perf/arm_v7_pmu.c index 420cadd108e7..17831e1920bd 100644 --- a/drivers/perf/arm_v7_pmu.c +++ b/drivers/perf/arm_v7_pmu.c @@ -858,16 +858,6 @@ static void armv7pmu_enable_event(struct perf_event *event) } /* - * Enable counter and interrupt, and set the counter to count - * the event that we're interested in. - */ - - /* - * Disable counter - */ - armv7_pmnc_disable_counter(idx); - - /* * Set event (if destined for PMNx counters) * We only need to set the event for the cycle counter if we * have the ability to perform event filtering. @@ -875,14 +865,7 @@ static void armv7pmu_enable_event(struct perf_event *event) if (cpu_pmu->set_event_filter || idx != ARMV7_IDX_CYCLE_COUNTER) armv7_pmnc_write_evtsel(idx, hwc->config_base); - /* - * Enable interrupt for this counter - */ armv7_pmnc_enable_intens(idx); - - /* - * Enable counter - */ armv7_pmnc_enable_counter(idx); } @@ -898,18 +881,7 @@ static void armv7pmu_disable_event(struct perf_event *event) return; } - /* - * Disable counter and interrupt - */ - - /* - * Disable counter - */ armv7_pmnc_disable_counter(idx); - - /* - * Disable interrupt for this counter - */ armv7_pmnc_disable_intens(idx); } @@ -1477,14 +1449,6 @@ static void krait_pmu_enable_event(struct perf_event *event) int idx = hwc->idx; /* - * Enable counter and interrupt, and set the counter to count - * the event that we're interested in. - */ - - /* Disable counter */ - armv7_pmnc_disable_counter(idx); - - /* * Set event (if destined for PMNx counters) * We set the event for the cycle counter because we * have the ability to perform event filtering. @@ -1494,10 +1458,7 @@ static void krait_pmu_enable_event(struct perf_event *event) else armv7_pmnc_write_evtsel(idx, hwc->config_base); - /* Enable interrupt for this counter */ armv7_pmnc_enable_intens(idx); - - /* Enable counter */ armv7_pmnc_enable_counter(idx); } @@ -1798,14 +1759,6 @@ static void scorpion_pmu_enable_event(struct perf_event *event) int idx = hwc->idx; /* - * Enable counter and interrupt, and set the counter to count - * the event that we're interested in. - */ - - /* Disable counter */ - armv7_pmnc_disable_counter(idx); - - /* * Set event (if destined for PMNx counters) * We don't set the event for the cycle counter because we * don't have the ability to perform event filtering. @@ -1815,10 +1768,7 @@ static void scorpion_pmu_enable_event(struct perf_event *event) else if (idx != ARMV7_IDX_CYCLE_COUNTER) armv7_pmnc_write_evtsel(idx, hwc->config_base); - /* Enable interrupt for this counter */ armv7_pmnc_enable_intens(idx); - - /* Enable counter */ armv7_pmnc_enable_counter(idx); } diff --git a/drivers/perf/dwc_pcie_pmu.c b/drivers/perf/dwc_pcie_pmu.c index f851e070760c..146ff57813fb 100644 --- a/drivers/perf/dwc_pcie_pmu.c +++ b/drivers/perf/dwc_pcie_pmu.c @@ -13,6 +13,7 @@ #include <linux/errno.h> #include <linux/kernel.h> #include <linux/list.h> +#include <linux/pcie-dwc.h> #include <linux/perf_event.h> #include <linux/pci.h> #include <linux/platform_device.h> @@ -99,26 +100,6 @@ struct dwc_pcie_dev_info { struct list_head dev_node; }; -struct dwc_pcie_pmu_vsec_id { - u16 vendor_id; - u16 vsec_id; - u8 vsec_rev; -}; - -/* - * VSEC IDs are allocated by the vendor, so a given ID may mean different - * things to different vendors. See PCIe r6.0, sec 7.9.5.2. - */ -static const struct dwc_pcie_pmu_vsec_id dwc_pcie_pmu_vsec_ids[] = { - { .vendor_id = PCI_VENDOR_ID_ALIBABA, - .vsec_id = 0x02, .vsec_rev = 0x4 }, - { .vendor_id = PCI_VENDOR_ID_AMPERE, - .vsec_id = 0x02, .vsec_rev = 0x4 }, - { .vendor_id = PCI_VENDOR_ID_QCOM, - .vsec_id = 0x02, .vsec_rev = 0x4 }, - {} /* terminator */ -}; - static ssize_t cpumask_show(struct device *dev, struct device_attribute *attr, char *buf) @@ -529,14 +510,14 @@ static void dwc_pcie_unregister_pmu(void *data) static u16 dwc_pcie_des_cap(struct pci_dev *pdev) { - const struct dwc_pcie_pmu_vsec_id *vid; + const struct dwc_pcie_vsec_id *vid; u16 vsec; u32 val; if (!pci_is_pcie(pdev) || !(pci_pcie_type(pdev) == PCI_EXP_TYPE_ROOT_PORT)) return 0; - for (vid = dwc_pcie_pmu_vsec_ids; vid->vendor_id; vid++) { + for (vid = dwc_pcie_rasdes_vsec_ids; vid->vendor_id; vid++) { vsec = pci_find_vsec_capability(pdev, vid->vendor_id, vid->vsec_id); if (vsec) { diff --git a/drivers/perf/marvell_cn10k_ddr_pmu.c b/drivers/perf/marvell_cn10k_ddr_pmu.c index 039feded9152..72ac17efd846 100644 --- a/drivers/perf/marvell_cn10k_ddr_pmu.c +++ b/drivers/perf/marvell_cn10k_ddr_pmu.c @@ -1064,8 +1064,8 @@ static int cn10k_ddr_perf_probe(struct platform_device *pdev) if (!name) return -ENOMEM; - hrtimer_init(&ddr_pmu->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - ddr_pmu->hrtimer.function = cn10k_ddr_pmu_timer_handler; + hrtimer_setup(&ddr_pmu->hrtimer, cn10k_ddr_pmu_timer_handler, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); cpuhp_state_add_instance_nocalls( CPUHP_AP_PERF_ARM_MARVELL_CN10K_DDR_ONLINE, diff --git a/drivers/perf/thunderx2_pmu.c b/drivers/perf/thunderx2_pmu.c index cadd60221b8f..6ed4707bd6bb 100644 --- a/drivers/perf/thunderx2_pmu.c +++ b/drivers/perf/thunderx2_pmu.c @@ -752,9 +752,8 @@ static int tx2_uncore_pmu_add_dev(struct tx2_uncore_pmu *tx2_pmu) tx2_pmu->cpu = cpu; if (tx2_pmu->hrtimer_callback) { - hrtimer_init(&tx2_pmu->hrtimer, - CLOCK_MONOTONIC, HRTIMER_MODE_REL); - tx2_pmu->hrtimer.function = tx2_pmu->hrtimer_callback; + hrtimer_setup(&tx2_pmu->hrtimer, tx2_pmu->hrtimer_callback, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); } ret = tx2_uncore_pmu_register(tx2_pmu); |