From 9cd8062b38e61f11054f13b3c98695c7b1d73b11 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Thu, 6 Jun 2024 19:50:05 +0300 Subject: ACPI/IORT: Switch to use kmemdup_array() Let the kememdup_array() take care about multiplication and possible overflows. Signed-off-by: Andy Shevchenko Acked-by: Hanjun Guo Link: https://lore.kernel.org/r/20240606165005.3031490-1-andriy.shevchenko@linux.intel.com Signed-off-by: Will Deacon --- drivers/acpi/arm64/iort.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c index 1b39e9ae7ac1..034d303c3df1 100644 --- a/drivers/acpi/arm64/iort.c +++ b/drivers/acpi/arm64/iort.c @@ -822,7 +822,7 @@ static struct iommu_iort_rmr_data *iort_rmr_alloc( return NULL; /* Create a copy of SIDs array to associate with this rmr_data */ - sids_copy = kmemdup(sids, num_sids * sizeof(*sids), GFP_KERNEL); + sids_copy = kmemdup_array(sids, num_sids, sizeof(*sids), GFP_KERNEL); if (!sids_copy) { kfree(rmr_data); return NULL; -- cgit v1.2.3 From 48b035121a564f2695462fbab0d2af51cb4f4b7a Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Wed, 31 Jul 2024 13:12:53 -0600 Subject: perf: arm_pmu: Use of_property_present() Use of_property_present() to test for property presence rather than of_find_property(). This is part of a larger effort to remove callers of of_find_property() and similar functions. of_find_property() leaks the DT struct property and data pointers which is a problem for dynamically allocated nodes which may be freed. Signed-off-by: Rob Herring (Arm) Acked-by: Mark Rutland Reviewed-by: Anshuman Khandual Link: https://lore.kernel.org/r/20240731191312.1710417-15-robh@kernel.org Signed-off-by: Will Deacon --- drivers/perf/arm_pmu_platform.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/arm_pmu_platform.c b/drivers/perf/arm_pmu_platform.c index 4b1a9a92ea11..118170a5cede 100644 --- a/drivers/perf/arm_pmu_platform.c +++ b/drivers/perf/arm_pmu_platform.c @@ -59,7 +59,7 @@ static int pmu_parse_percpu_irq(struct arm_pmu *pmu, int irq) static bool pmu_has_irq_affinity(struct device_node *node) { - return !!of_find_property(node, "interrupt-affinity", NULL); + return of_property_present(node, "interrupt-affinity"); } static int pmu_parse_irq_affinity(struct device *dev, int i) -- cgit v1.2.3 From bf5ffc8c80e0cf5205849cd0c9c3cb261d2beee6 Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Wed, 31 Jul 2024 10:51:18 -0600 Subject: perf: arm_pmu: Remove event index to counter remapping Xscale and Armv6 PMUs defined the cycle counter at 0 and event counters starting at 1 and had 1:1 event index to counter numbering. On Armv7 and later, this changed the cycle counter to 31 and event counters start at 0. The drivers for Armv7 and PMUv3 kept the old event index numbering and introduced an event index to counter conversion. The conversion uses masking to convert from event index to a counter number. This operation relies on having at most 32 counters so that the cycle counter index 0 can be transformed to counter number 31. Armv9.4 adds support for an additional fixed function counter (instructions) which increases possible counters to more than 32, and the conversion won't work anymore as a simple subtract and mask. The primary reason for the translation (other than history) seems to be to have a contiguous mask of counters 0-N. Keeping that would result in more complicated index to counter conversions. Instead, store a mask of available counters rather than just number of events. That provides more information in addition to the number of events. No (intended) functional changes. Acked-by: Mark Rutland Signed-off-by: Rob Herring (Arm) Tested-by: James Clark Link: https://lore.kernel.org/r/20240731-arm-pmu-3-9-icntr-v3-1-280a8d7ff465@kernel.org Signed-off-by: Will Deacon --- arch/arm64/kvm/pmu-emul.c | 6 ++-- drivers/perf/apple_m1_cpu_pmu.c | 4 +-- drivers/perf/arm_pmu.c | 11 +++--- drivers/perf/arm_pmuv3.c | 62 +++++++++++---------------------- drivers/perf/arm_v6_pmu.c | 6 ++-- drivers/perf/arm_v7_pmu.c | 77 ++++++++++++++++------------------------- drivers/perf/arm_xscale_pmu.c | 12 ++++--- include/linux/perf/arm_pmu.h | 2 +- include/linux/perf/arm_pmuv3.h | 1 + 9 files changed, 75 insertions(+), 106 deletions(-) (limited to 'drivers') diff --git a/arch/arm64/kvm/pmu-emul.c b/arch/arm64/kvm/pmu-emul.c index 82a2a003259c..0e598f6c42c0 100644 --- a/arch/arm64/kvm/pmu-emul.c +++ b/arch/arm64/kvm/pmu-emul.c @@ -910,10 +910,10 @@ u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm) struct arm_pmu *arm_pmu = kvm->arch.arm_pmu; /* - * The arm_pmu->num_events considers the cycle counter as well. - * Ignore that and return only the general-purpose counters. + * The arm_pmu->cntr_mask considers the fixed counter(s) as well. + * Ignore those and return only the general-purpose counters. */ - return arm_pmu->num_events - 1; + return bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS); } static void kvm_arm_set_pmu(struct kvm *kvm, struct arm_pmu *arm_pmu) diff --git a/drivers/perf/apple_m1_cpu_pmu.c b/drivers/perf/apple_m1_cpu_pmu.c index f322e5ca1114..c8f607912567 100644 --- a/drivers/perf/apple_m1_cpu_pmu.c +++ b/drivers/perf/apple_m1_cpu_pmu.c @@ -400,7 +400,7 @@ static irqreturn_t m1_pmu_handle_irq(struct arm_pmu *cpu_pmu) regs = get_irq_regs(); - for (idx = 0; idx < cpu_pmu->num_events; idx++) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, M1_PMU_NR_COUNTERS) { struct perf_event *event = cpuc->events[idx]; struct perf_sample_data data; @@ -560,7 +560,7 @@ 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->num_events = M1_PMU_NR_COUNTERS; + 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; return 0; diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c index 8458fe2cebb4..398cce3d76fc 100644 --- a/drivers/perf/arm_pmu.c +++ b/drivers/perf/arm_pmu.c @@ -522,7 +522,7 @@ static void armpmu_enable(struct pmu *pmu) { struct arm_pmu *armpmu = to_arm_pmu(pmu); struct pmu_hw_events *hw_events = this_cpu_ptr(armpmu->hw_events); - bool enabled = !bitmap_empty(hw_events->used_mask, armpmu->num_events); + bool enabled = !bitmap_empty(hw_events->used_mask, ARMPMU_MAX_HWEVENTS); /* For task-bound events we may be called on other CPUs */ if (!cpumask_test_cpu(smp_processor_id(), &armpmu->supported_cpus)) @@ -742,7 +742,7 @@ static void cpu_pm_pmu_setup(struct arm_pmu *armpmu, unsigned long cmd) struct perf_event *event; int idx; - for (idx = 0; idx < armpmu->num_events; idx++) { + for_each_set_bit(idx, armpmu->cntr_mask, ARMPMU_MAX_HWEVENTS) { event = hw_events->events[idx]; if (!event) continue; @@ -772,7 +772,7 @@ static int cpu_pm_pmu_notify(struct notifier_block *b, unsigned long cmd, { struct arm_pmu *armpmu = container_of(b, struct arm_pmu, cpu_pm_nb); struct pmu_hw_events *hw_events = this_cpu_ptr(armpmu->hw_events); - bool enabled = !bitmap_empty(hw_events->used_mask, armpmu->num_events); + bool enabled = !bitmap_empty(hw_events->used_mask, ARMPMU_MAX_HWEVENTS); if (!cpumask_test_cpu(smp_processor_id(), &armpmu->supported_cpus)) return NOTIFY_DONE; @@ -924,8 +924,9 @@ int armpmu_register(struct arm_pmu *pmu) if (ret) goto out_destroy; - pr_info("enabled with %s PMU driver, %d counters available%s\n", - pmu->name, pmu->num_events, + pr_info("enabled with %s PMU driver, %d (%*pb) counters available%s\n", + pmu->name, bitmap_weight(pmu->cntr_mask, ARMPMU_MAX_HWEVENTS), + ARMPMU_MAX_HWEVENTS, &pmu->cntr_mask, has_nmi ? ", using NMIs" : ""); kvm_host_pmu_init(pmu); diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index d246840797b6..b48773ec892e 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -454,9 +454,7 @@ static const struct attribute_group armv8_pmuv3_caps_attr_group = { /* * Perf Events' indices */ -#define ARMV8_IDX_CYCLE_COUNTER 0 -#define ARMV8_IDX_COUNTER0 1 -#define ARMV8_IDX_CYCLE_COUNTER_USER 32 +#define ARMV8_IDX_CYCLE_COUNTER 31 /* * We unconditionally enable ARMv8.5-PMU long event counter support @@ -489,19 +487,12 @@ static bool armv8pmu_event_is_chained(struct perf_event *event) return !armv8pmu_event_has_user_read(event) && armv8pmu_event_is_64bit(event) && !armv8pmu_has_long_event(cpu_pmu) && - (idx != ARMV8_IDX_CYCLE_COUNTER); + (idx < ARMV8_PMU_MAX_GENERAL_COUNTERS); } /* * ARMv8 low level PMU access */ - -/* - * Perf Event to low level counters mapping - */ -#define ARMV8_IDX_TO_COUNTER(x) \ - (((x) - ARMV8_IDX_COUNTER0) & ARMV8_PMU_COUNTER_MASK) - static u64 armv8pmu_pmcr_read(void) { return read_pmcr(); @@ -521,14 +512,12 @@ static int armv8pmu_has_overflowed(u32 pmovsr) static int armv8pmu_counter_has_overflowed(u32 pmnc, int idx) { - return pmnc & BIT(ARMV8_IDX_TO_COUNTER(idx)); + return pmnc & BIT(idx); } static u64 armv8pmu_read_evcntr(int idx) { - u32 counter = ARMV8_IDX_TO_COUNTER(idx); - - return read_pmevcntrn(counter); + return read_pmevcntrn(idx); } static u64 armv8pmu_read_hw_counter(struct perf_event *event) @@ -557,7 +546,7 @@ static bool armv8pmu_event_needs_bias(struct perf_event *event) return false; if (armv8pmu_has_long_event(cpu_pmu) || - idx == ARMV8_IDX_CYCLE_COUNTER) + idx >= ARMV8_PMU_MAX_GENERAL_COUNTERS) return true; return false; @@ -595,9 +584,7 @@ static u64 armv8pmu_read_counter(struct perf_event *event) static void armv8pmu_write_evcntr(int idx, u64 value) { - u32 counter = ARMV8_IDX_TO_COUNTER(idx); - - write_pmevcntrn(counter, value); + write_pmevcntrn(idx, value); } static void armv8pmu_write_hw_counter(struct perf_event *event, @@ -628,7 +615,6 @@ static void armv8pmu_write_counter(struct perf_event *event, u64 value) static void armv8pmu_write_evtype(int idx, unsigned long val) { - u32 counter = ARMV8_IDX_TO_COUNTER(idx); unsigned long mask = ARMV8_PMU_EVTYPE_EVENT | ARMV8_PMU_INCLUDE_EL2 | ARMV8_PMU_EXCLUDE_EL0 | @@ -638,7 +624,7 @@ static void armv8pmu_write_evtype(int idx, unsigned long val) mask |= ARMV8_PMU_EVTYPE_TC | ARMV8_PMU_EVTYPE_TH; val &= mask; - write_pmevtypern(counter, val); + write_pmevtypern(idx, val); } static void armv8pmu_write_event_type(struct perf_event *event) @@ -667,7 +653,7 @@ static void armv8pmu_write_event_type(struct perf_event *event) static u32 armv8pmu_event_cnten_mask(struct perf_event *event) { - int counter = ARMV8_IDX_TO_COUNTER(event->hw.idx); + int counter = event->hw.idx; u32 mask = BIT(counter); if (armv8pmu_event_is_chained(event)) @@ -726,8 +712,7 @@ static void armv8pmu_enable_intens(u32 mask) static void armv8pmu_enable_event_irq(struct perf_event *event) { - u32 counter = ARMV8_IDX_TO_COUNTER(event->hw.idx); - armv8pmu_enable_intens(BIT(counter)); + armv8pmu_enable_intens(BIT(event->hw.idx)); } static void armv8pmu_disable_intens(u32 mask) @@ -741,8 +726,7 @@ static void armv8pmu_disable_intens(u32 mask) static void armv8pmu_disable_event_irq(struct perf_event *event) { - u32 counter = ARMV8_IDX_TO_COUNTER(event->hw.idx); - armv8pmu_disable_intens(BIT(counter)); + armv8pmu_disable_intens(BIT(event->hw.idx)); } static u32 armv8pmu_getreset_flags(void) @@ -786,7 +770,8 @@ static void armv8pmu_enable_user_access(struct arm_pmu *cpu_pmu) struct pmu_hw_events *cpuc = this_cpu_ptr(cpu_pmu->hw_events); /* Clear any unused counters to avoid leaking their contents */ - for_each_clear_bit(i, cpuc->used_mask, cpu_pmu->num_events) { + for_each_andnot_bit(i, cpu_pmu->cntr_mask, cpuc->used_mask, + ARMPMU_MAX_HWEVENTS) { if (i == ARMV8_IDX_CYCLE_COUNTER) write_pmccntr(0); else @@ -869,7 +854,7 @@ static irqreturn_t armv8pmu_handle_irq(struct arm_pmu *cpu_pmu) * to prevent skews in group events. */ armv8pmu_stop(cpu_pmu); - for (idx = 0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMPMU_MAX_HWEVENTS) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -908,7 +893,7 @@ static int armv8pmu_get_single_idx(struct pmu_hw_events *cpuc, { int idx; - for (idx = ARMV8_IDX_COUNTER0; idx < cpu_pmu->num_events; idx++) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS) { if (!test_and_set_bit(idx, cpuc->used_mask)) return idx; } @@ -924,7 +909,9 @@ static int armv8pmu_get_chain_idx(struct pmu_hw_events *cpuc, * Chaining requires two consecutive event counters, where * the lower idx must be even. */ - for (idx = ARMV8_IDX_COUNTER0 + 1; idx < cpu_pmu->num_events; idx += 2) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS) { + if (!(idx & 0x1)) + continue; if (!test_and_set_bit(idx, cpuc->used_mask)) { /* Check if the preceding even counter is available */ if (!test_and_set_bit(idx - 1, cpuc->used_mask)) @@ -978,15 +965,7 @@ static int armv8pmu_user_event_idx(struct perf_event *event) if (!sysctl_perf_user_access || !armv8pmu_event_has_user_read(event)) return 0; - /* - * We remap the cycle counter index to 32 to - * match the offset applied to the rest of - * the counter indices. - */ - if (event->hw.idx == ARMV8_IDX_CYCLE_COUNTER) - return ARMV8_IDX_CYCLE_COUNTER_USER; - - return event->hw.idx; + return event->hw.idx + 1; } /* @@ -1211,10 +1190,11 @@ static void __armv8pmu_probe_pmu(void *info) probe->present = true; /* Read the nb of CNTx counters supported from PMNC */ - cpu_pmu->num_events = FIELD_GET(ARMV8_PMU_PMCR_N, armv8pmu_pmcr_read()); + bitmap_set(cpu_pmu->cntr_mask, + 0, FIELD_GET(ARMV8_PMU_PMCR_N, armv8pmu_pmcr_read())); /* Add the CPU cycles counter */ - cpu_pmu->num_events += 1; + set_bit(ARMV8_IDX_CYCLE_COUNTER, cpu_pmu->cntr_mask); pmceid[0] = pmceid_raw[0] = read_pmceid0(); pmceid[1] = pmceid_raw[1] = read_pmceid1(); diff --git a/drivers/perf/arm_v6_pmu.c b/drivers/perf/arm_v6_pmu.c index 0bb685b4bac5..b09615bb2bb2 100644 --- a/drivers/perf/arm_v6_pmu.c +++ b/drivers/perf/arm_v6_pmu.c @@ -64,6 +64,7 @@ enum armv6_counters { ARMV6_CYCLE_COUNTER = 0, ARMV6_COUNTER0, ARMV6_COUNTER1, + ARMV6_NUM_COUNTERS }; /* @@ -254,7 +255,7 @@ armv6pmu_handle_irq(struct arm_pmu *cpu_pmu) */ armv6_pmcr_write(pmcr); - for (idx = 0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV6_NUM_COUNTERS) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -391,7 +392,8 @@ static void armv6pmu_init(struct arm_pmu *cpu_pmu) cpu_pmu->start = armv6pmu_start; cpu_pmu->stop = armv6pmu_stop; cpu_pmu->map_event = armv6_map_event; - cpu_pmu->num_events = 3; + + bitmap_set(cpu_pmu->cntr_mask, 0, ARMV6_NUM_COUNTERS); } static int armv6_1136_pmu_init(struct arm_pmu *cpu_pmu) diff --git a/drivers/perf/arm_v7_pmu.c b/drivers/perf/arm_v7_pmu.c index 928ac3d626ed..420cadd108e7 100644 --- a/drivers/perf/arm_v7_pmu.c +++ b/drivers/perf/arm_v7_pmu.c @@ -649,24 +649,12 @@ static struct attribute_group armv7_pmuv2_events_attr_group = { /* * Perf Events' indices */ -#define ARMV7_IDX_CYCLE_COUNTER 0 -#define ARMV7_IDX_COUNTER0 1 -#define ARMV7_IDX_COUNTER_LAST(cpu_pmu) \ - (ARMV7_IDX_CYCLE_COUNTER + cpu_pmu->num_events - 1) - -#define ARMV7_MAX_COUNTERS 32 -#define ARMV7_COUNTER_MASK (ARMV7_MAX_COUNTERS - 1) - +#define ARMV7_IDX_CYCLE_COUNTER 31 +#define ARMV7_IDX_COUNTER_MAX 31 /* * ARMv7 low level PMNC access */ -/* - * Perf Event to low level counters mapping - */ -#define ARMV7_IDX_TO_COUNTER(x) \ - (((x) - ARMV7_IDX_COUNTER0) & ARMV7_COUNTER_MASK) - /* * Per-CPU PMNC: config reg */ @@ -725,19 +713,17 @@ static inline int armv7_pmnc_has_overflowed(u32 pmnc) static inline int armv7_pmnc_counter_valid(struct arm_pmu *cpu_pmu, int idx) { - return idx >= ARMV7_IDX_CYCLE_COUNTER && - idx <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); + return test_bit(idx, cpu_pmu->cntr_mask); } static inline int armv7_pmnc_counter_has_overflowed(u32 pmnc, int idx) { - return pmnc & BIT(ARMV7_IDX_TO_COUNTER(idx)); + return pmnc & BIT(idx); } static inline void armv7_pmnc_select_counter(int idx) { - u32 counter = ARMV7_IDX_TO_COUNTER(idx); - asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (counter)); + asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (idx)); isb(); } @@ -787,29 +773,25 @@ static inline void armv7_pmnc_write_evtsel(int idx, u32 val) static inline void armv7_pmnc_enable_counter(int idx) { - u32 counter = ARMV7_IDX_TO_COUNTER(idx); - asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (BIT(counter))); + asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (BIT(idx))); } static inline void armv7_pmnc_disable_counter(int idx) { - u32 counter = ARMV7_IDX_TO_COUNTER(idx); - asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (BIT(counter))); + asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (BIT(idx))); } static inline void armv7_pmnc_enable_intens(int idx) { - u32 counter = ARMV7_IDX_TO_COUNTER(idx); - asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (BIT(counter))); + asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (BIT(idx))); } static inline void armv7_pmnc_disable_intens(int idx) { - u32 counter = ARMV7_IDX_TO_COUNTER(idx); - asm volatile("mcr p15, 0, %0, c9, c14, 2" : : "r" (BIT(counter))); + asm volatile("mcr p15, 0, %0, c9, c14, 2" : : "r" (BIT(idx))); isb(); /* Clear the overflow flag in case an interrupt is pending. */ - asm volatile("mcr p15, 0, %0, c9, c12, 3" : : "r" (BIT(counter))); + asm volatile("mcr p15, 0, %0, c9, c12, 3" : : "r" (BIT(idx))); isb(); } @@ -853,15 +835,12 @@ static void armv7_pmnc_dump_regs(struct arm_pmu *cpu_pmu) asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (val)); pr_info("CCNT =0x%08x\n", val); - for (cnt = ARMV7_IDX_COUNTER0; - cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { + for_each_set_bit(cnt, cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX) { armv7_pmnc_select_counter(cnt); asm volatile("mrc p15, 0, %0, c9, c13, 2" : "=r" (val)); - pr_info("CNT[%d] count =0x%08x\n", - ARMV7_IDX_TO_COUNTER(cnt), val); + pr_info("CNT[%d] count =0x%08x\n", cnt, val); asm volatile("mrc p15, 0, %0, c9, c13, 1" : "=r" (val)); - pr_info("CNT[%d] evtsel=0x%08x\n", - ARMV7_IDX_TO_COUNTER(cnt), val); + pr_info("CNT[%d] evtsel=0x%08x\n", cnt, val); } } #endif @@ -958,7 +937,7 @@ static irqreturn_t armv7pmu_handle_irq(struct arm_pmu *cpu_pmu) */ regs = get_irq_regs(); - for (idx = 0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMPMU_MAX_HWEVENTS) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -1027,7 +1006,7 @@ static int armv7pmu_get_event_idx(struct pmu_hw_events *cpuc, * For anything other than a cycle counter, try and use * the events counters */ - for (idx = ARMV7_IDX_COUNTER0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX) { if (!test_and_set_bit(idx, cpuc->used_mask)) return idx; } @@ -1073,7 +1052,7 @@ static int armv7pmu_set_event_filter(struct hw_perf_event *event, static void armv7pmu_reset(void *info) { struct arm_pmu *cpu_pmu = (struct arm_pmu *)info; - u32 idx, nb_cnt = cpu_pmu->num_events, val; + u32 idx, val; if (cpu_pmu->secure_access) { asm volatile("mrc p15, 0, %0, c1, c1, 1" : "=r" (val)); @@ -1082,7 +1061,7 @@ static void armv7pmu_reset(void *info) } /* The counter and interrupt enable registers are unknown at reset. */ - for (idx = ARMV7_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMPMU_MAX_HWEVENTS) { armv7_pmnc_disable_counter(idx); armv7_pmnc_disable_intens(idx); } @@ -1161,20 +1140,22 @@ static void armv7pmu_init(struct arm_pmu *cpu_pmu) static void armv7_read_num_pmnc_events(void *info) { - int *nb_cnt = info; + int nb_cnt; + struct arm_pmu *cpu_pmu = info; /* Read the nb of CNTx counters supported from PMNC */ - *nb_cnt = (armv7_pmnc_read() >> ARMV7_PMNC_N_SHIFT) & ARMV7_PMNC_N_MASK; + nb_cnt = (armv7_pmnc_read() >> ARMV7_PMNC_N_SHIFT) & ARMV7_PMNC_N_MASK; + bitmap_set(cpu_pmu->cntr_mask, 0, nb_cnt); /* Add the CPU cycles counter */ - *nb_cnt += 1; + set_bit(ARMV7_IDX_CYCLE_COUNTER, cpu_pmu->cntr_mask); } static int armv7_probe_num_events(struct arm_pmu *arm_pmu) { return smp_call_function_any(&arm_pmu->supported_cpus, armv7_read_num_pmnc_events, - &arm_pmu->num_events, 1); + arm_pmu, 1); } static int armv7_a8_pmu_init(struct arm_pmu *cpu_pmu) @@ -1524,7 +1505,7 @@ static void krait_pmu_reset(void *info) { u32 vval, fval; struct arm_pmu *cpu_pmu = info; - u32 idx, nb_cnt = cpu_pmu->num_events; + u32 idx; armv7pmu_reset(info); @@ -1538,7 +1519,7 @@ static void krait_pmu_reset(void *info) venum_post_pmresr(vval, fval); /* Reset PMxEVNCTCR to sane default */ - for (idx = ARMV7_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX) { armv7_pmnc_select_counter(idx); asm volatile("mcr p15, 0, %0, c9, c15, 0" : : "r" (0)); } @@ -1562,7 +1543,7 @@ static int krait_event_to_bit(struct perf_event *event, unsigned int region, * Lower bits are reserved for use by the counters (see * armv7pmu_get_event_idx() for more info) */ - bit += ARMV7_IDX_COUNTER_LAST(cpu_pmu) + 1; + bit += bitmap_weight(cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX); return bit; } @@ -1845,7 +1826,7 @@ static void scorpion_pmu_reset(void *info) { u32 vval, fval; struct arm_pmu *cpu_pmu = info; - u32 idx, nb_cnt = cpu_pmu->num_events; + u32 idx; armv7pmu_reset(info); @@ -1860,7 +1841,7 @@ static void scorpion_pmu_reset(void *info) venum_post_pmresr(vval, fval); /* Reset PMxEVNCTCR to sane default */ - for (idx = ARMV7_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX) { armv7_pmnc_select_counter(idx); asm volatile("mcr p15, 0, %0, c9, c15, 0" : : "r" (0)); } @@ -1883,7 +1864,7 @@ static int scorpion_event_to_bit(struct perf_event *event, unsigned int region, * Lower bits are reserved for use by the counters (see * armv7pmu_get_event_idx() for more info) */ - bit += ARMV7_IDX_COUNTER_LAST(cpu_pmu) + 1; + bit += bitmap_weight(cpu_pmu->cntr_mask, ARMV7_IDX_COUNTER_MAX); return bit; } diff --git a/drivers/perf/arm_xscale_pmu.c b/drivers/perf/arm_xscale_pmu.c index 3d8b72d6b37f..638fea9b1263 100644 --- a/drivers/perf/arm_xscale_pmu.c +++ b/drivers/perf/arm_xscale_pmu.c @@ -53,6 +53,8 @@ enum xscale_counters { XSCALE_COUNTER2, XSCALE_COUNTER3, }; +#define XSCALE1_NUM_COUNTERS 3 +#define XSCALE2_NUM_COUNTERS 5 static const unsigned xscale_perf_map[PERF_COUNT_HW_MAX] = { PERF_MAP_ALL_UNSUPPORTED, @@ -168,7 +170,7 @@ xscale1pmu_handle_irq(struct arm_pmu *cpu_pmu) regs = get_irq_regs(); - for (idx = 0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, XSCALE1_NUM_COUNTERS) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -364,7 +366,8 @@ static int xscale1pmu_init(struct arm_pmu *cpu_pmu) cpu_pmu->start = xscale1pmu_start; cpu_pmu->stop = xscale1pmu_stop; cpu_pmu->map_event = xscale_map_event; - cpu_pmu->num_events = 3; + + bitmap_set(cpu_pmu->cntr_mask, 0, XSCALE1_NUM_COUNTERS); return 0; } @@ -500,7 +503,7 @@ xscale2pmu_handle_irq(struct arm_pmu *cpu_pmu) regs = get_irq_regs(); - for (idx = 0; idx < cpu_pmu->num_events; ++idx) { + for_each_set_bit(idx, cpu_pmu->cntr_mask, XSCALE2_NUM_COUNTERS) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -719,7 +722,8 @@ static int xscale2pmu_init(struct arm_pmu *cpu_pmu) cpu_pmu->start = xscale2pmu_start; cpu_pmu->stop = xscale2pmu_stop; cpu_pmu->map_event = xscale_map_event; - cpu_pmu->num_events = 5; + + bitmap_set(cpu_pmu->cntr_mask, 0, XSCALE2_NUM_COUNTERS); return 0; } diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h index b3b34f6670cf..e5d6d204beab 100644 --- a/include/linux/perf/arm_pmu.h +++ b/include/linux/perf/arm_pmu.h @@ -96,7 +96,7 @@ struct arm_pmu { void (*stop)(struct arm_pmu *); void (*reset)(void *); int (*map_event)(struct perf_event *event); - int num_events; + DECLARE_BITMAP(cntr_mask, ARMPMU_MAX_HWEVENTS); bool secure_access; /* 32-bit ARM only */ #define ARMV8_PMUV3_MAX_COMMON_EVENTS 0x40 DECLARE_BITMAP(pmceid_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS); diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h index 7867db04ec98..eccbdd8eb98f 100644 --- a/include/linux/perf/arm_pmuv3.h +++ b/include/linux/perf/arm_pmuv3.h @@ -6,6 +6,7 @@ #ifndef __PERF_ARM_PMUV3_H #define __PERF_ARM_PMUV3_H +#define ARMV8_PMU_MAX_GENERAL_COUNTERS 31 #define ARMV8_PMU_MAX_COUNTERS 32 #define ARMV8_PMU_COUNTER_MASK (ARMV8_PMU_MAX_COUNTERS - 1) -- cgit v1.2.3 From a4a6e2078d85a9d94bcc7eab77845cb8cd39f680 Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Wed, 31 Jul 2024 10:51:19 -0600 Subject: perf: arm_pmuv3: Prepare for more than 32 counters Various PMUv3 registers which are a mask of counters are 64-bit registers, but the accessor functions take a u32. This has been fine as the upper 32-bits have been RES0 as there has been a maximum of 32 counters prior to Armv9.4/8.9. With Armv9.4/8.9, a 33rd counter is added. Update the accessor functions to use a u64 instead. Acked-by: Mark Rutland Signed-off-by: Rob Herring (Arm) Tested-by: James Clark Link: https://lore.kernel.org/r/20240731-arm-pmu-3-9-icntr-v3-2-280a8d7ff465@kernel.org Signed-off-by: Will Deacon --- arch/arm64/include/asm/arm_pmuv3.h | 12 ++++++------ arch/arm64/include/asm/kvm_host.h | 8 ++++---- arch/arm64/kvm/pmu.c | 8 ++++---- drivers/perf/arm_pmuv3.c | 40 ++++++++++++++++++++------------------ include/kvm/arm_pmu.h | 4 ++-- 5 files changed, 37 insertions(+), 35 deletions(-) (limited to 'drivers') diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h index a4697a0b6835..19b3f9150058 100644 --- a/arch/arm64/include/asm/arm_pmuv3.h +++ b/arch/arm64/include/asm/arm_pmuv3.h @@ -71,22 +71,22 @@ static inline u64 read_pmccntr(void) return read_sysreg(pmccntr_el0); } -static inline void write_pmcntenset(u32 val) +static inline void write_pmcntenset(u64 val) { write_sysreg(val, pmcntenset_el0); } -static inline void write_pmcntenclr(u32 val) +static inline void write_pmcntenclr(u64 val) { write_sysreg(val, pmcntenclr_el0); } -static inline void write_pmintenset(u32 val) +static inline void write_pmintenset(u64 val) { write_sysreg(val, pmintenset_el1); } -static inline void write_pmintenclr(u32 val) +static inline void write_pmintenclr(u64 val) { write_sysreg(val, pmintenclr_el1); } @@ -96,12 +96,12 @@ static inline void write_pmccfiltr(u64 val) write_sysreg(val, pmccfiltr_el0); } -static inline void write_pmovsclr(u32 val) +static inline void write_pmovsclr(u64 val) { write_sysreg(val, pmovsclr_el0); } -static inline u32 read_pmovsclr(void) +static inline u64 read_pmovsclr(void) { return read_sysreg(pmovsclr_el0); } diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h index a33f5996ca9f..c0fc753aac87 100644 --- a/arch/arm64/include/asm/kvm_host.h +++ b/arch/arm64/include/asm/kvm_host.h @@ -1330,12 +1330,12 @@ void kvm_arch_vcpu_load_debug_state_flags(struct kvm_vcpu *vcpu); void kvm_arch_vcpu_put_debug_state_flags(struct kvm_vcpu *vcpu); #ifdef CONFIG_KVM -void kvm_set_pmu_events(u32 set, struct perf_event_attr *attr); -void kvm_clr_pmu_events(u32 clr); +void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr); +void kvm_clr_pmu_events(u64 clr); bool kvm_set_pmuserenr(u64 val); #else -static inline void kvm_set_pmu_events(u32 set, struct perf_event_attr *attr) {} -static inline void kvm_clr_pmu_events(u32 clr) {} +static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) {} +static inline void kvm_clr_pmu_events(u64 clr) {} static inline bool kvm_set_pmuserenr(u64 val) { return false; diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c index 329819806096..e633b4434c6a 100644 --- a/arch/arm64/kvm/pmu.c +++ b/arch/arm64/kvm/pmu.c @@ -35,7 +35,7 @@ struct kvm_pmu_events *kvm_get_pmu_events(void) * Add events to track that we may want to switch at guest entry/exit * time. */ -void kvm_set_pmu_events(u32 set, struct perf_event_attr *attr) +void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) { struct kvm_pmu_events *pmu = kvm_get_pmu_events(); @@ -51,7 +51,7 @@ void kvm_set_pmu_events(u32 set, struct perf_event_attr *attr) /* * Stop tracking events */ -void kvm_clr_pmu_events(u32 clr) +void kvm_clr_pmu_events(u64 clr) { struct kvm_pmu_events *pmu = kvm_get_pmu_events(); @@ -176,7 +176,7 @@ static void kvm_vcpu_pmu_disable_el0(unsigned long events) void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) { struct kvm_pmu_events *pmu; - u32 events_guest, events_host; + u64 events_guest, events_host; if (!kvm_arm_support_pmu_v3() || !has_vhe()) return; @@ -197,7 +197,7 @@ void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu) { struct kvm_pmu_events *pmu; - u32 events_guest, events_host; + u64 events_guest, events_host; if (!kvm_arm_support_pmu_v3() || !has_vhe()) return; diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index b48773ec892e..bd45fbcb9a5a 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -505,14 +505,14 @@ static void armv8pmu_pmcr_write(u64 val) write_pmcr(val); } -static int armv8pmu_has_overflowed(u32 pmovsr) +static int armv8pmu_has_overflowed(u64 pmovsr) { - return pmovsr & ARMV8_PMU_OVERFLOWED_MASK; + return !!(pmovsr & ARMV8_PMU_OVERFLOWED_MASK); } -static int armv8pmu_counter_has_overflowed(u32 pmnc, int idx) +static int armv8pmu_counter_has_overflowed(u64 pmnc, int idx) { - return pmnc & BIT(idx); + return !!(pmnc & BIT(idx)); } static u64 armv8pmu_read_evcntr(int idx) @@ -651,17 +651,17 @@ static void armv8pmu_write_event_type(struct perf_event *event) } } -static u32 armv8pmu_event_cnten_mask(struct perf_event *event) +static u64 armv8pmu_event_cnten_mask(struct perf_event *event) { int counter = event->hw.idx; - u32 mask = BIT(counter); + u64 mask = BIT(counter); if (armv8pmu_event_is_chained(event)) mask |= BIT(counter - 1); return mask; } -static void armv8pmu_enable_counter(u32 mask) +static void armv8pmu_enable_counter(u64 mask) { /* * Make sure event configuration register writes are visible before we @@ -674,7 +674,7 @@ static void armv8pmu_enable_counter(u32 mask) static void armv8pmu_enable_event_counter(struct perf_event *event) { struct perf_event_attr *attr = &event->attr; - u32 mask = armv8pmu_event_cnten_mask(event); + u64 mask = armv8pmu_event_cnten_mask(event); kvm_set_pmu_events(mask, attr); @@ -683,7 +683,7 @@ static void armv8pmu_enable_event_counter(struct perf_event *event) armv8pmu_enable_counter(mask); } -static void armv8pmu_disable_counter(u32 mask) +static void armv8pmu_disable_counter(u64 mask) { write_pmcntenclr(mask); /* @@ -696,7 +696,7 @@ static void armv8pmu_disable_counter(u32 mask) static void armv8pmu_disable_event_counter(struct perf_event *event) { struct perf_event_attr *attr = &event->attr; - u32 mask = armv8pmu_event_cnten_mask(event); + u64 mask = armv8pmu_event_cnten_mask(event); kvm_clr_pmu_events(mask); @@ -705,7 +705,7 @@ static void armv8pmu_disable_event_counter(struct perf_event *event) armv8pmu_disable_counter(mask); } -static void armv8pmu_enable_intens(u32 mask) +static void armv8pmu_enable_intens(u64 mask) { write_pmintenset(mask); } @@ -715,7 +715,7 @@ static void armv8pmu_enable_event_irq(struct perf_event *event) armv8pmu_enable_intens(BIT(event->hw.idx)); } -static void armv8pmu_disable_intens(u32 mask) +static void armv8pmu_disable_intens(u64 mask) { write_pmintenclr(mask); isb(); @@ -729,9 +729,9 @@ static void armv8pmu_disable_event_irq(struct perf_event *event) armv8pmu_disable_intens(BIT(event->hw.idx)); } -static u32 armv8pmu_getreset_flags(void) +static u64 armv8pmu_getreset_flags(void) { - u32 value; + u64 value; /* Read */ value = read_pmovsclr(); @@ -827,7 +827,7 @@ static void armv8pmu_stop(struct arm_pmu *cpu_pmu) static irqreturn_t armv8pmu_handle_irq(struct arm_pmu *cpu_pmu) { - u32 pmovsr; + u64 pmovsr; struct perf_sample_data data; struct pmu_hw_events *cpuc = this_cpu_ptr(cpu_pmu->hw_events); struct pt_regs *regs; @@ -1040,14 +1040,16 @@ static int armv8pmu_set_event_filter(struct hw_perf_event *event, static void armv8pmu_reset(void *info) { struct arm_pmu *cpu_pmu = (struct arm_pmu *)info; - u64 pmcr; + u64 pmcr, mask; + + bitmap_to_arr64(&mask, cpu_pmu->cntr_mask, ARMPMU_MAX_HWEVENTS); /* The counter and interrupt enable registers are unknown at reset. */ - armv8pmu_disable_counter(U32_MAX); - armv8pmu_disable_intens(U32_MAX); + armv8pmu_disable_counter(mask); + armv8pmu_disable_intens(mask); /* Clear the counters we flip at guest entry/exit */ - kvm_clr_pmu_events(U32_MAX); + kvm_clr_pmu_events(mask); /* * Initialize & Reset PMNC. Request overflow interrupt for diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h index 35d4ca4f6122..334d7c5503cf 100644 --- a/include/kvm/arm_pmu.h +++ b/include/kvm/arm_pmu.h @@ -19,8 +19,8 @@ struct kvm_pmc { }; struct kvm_pmu_events { - u32 events_host; - u32 events_guest; + u64 events_host; + u64 events_guest; }; struct kvm_pmu { -- cgit v1.2.3 From 126d7d7cce5e048fb82477a9842d088d10ff0df6 Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Wed, 31 Jul 2024 10:51:22 -0600 Subject: arm64: perf/kvm: Use a common PMU cycle counter define The PMUv3 and KVM code each have a define for the PMU cycle counter index. Move KVM's define to a shared location and use it for PMUv3 driver. Reviewed-by: Marc Zyngier Acked-by: Mark Rutland Signed-off-by: Rob Herring (Arm) Tested-by: James Clark Link: https://lore.kernel.org/r/20240731-arm-pmu-3-9-icntr-v3-5-280a8d7ff465@kernel.org Signed-off-by: Will Deacon --- arch/arm64/kvm/sys_regs.c | 1 + drivers/perf/arm_pmuv3.c | 19 +++++++------------ include/kvm/arm_pmu.h | 1 - include/linux/perf/arm_pmuv3.h | 3 +++ 4 files changed, 11 insertions(+), 13 deletions(-) (limited to 'drivers') diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c index 33497db257fb..7db24de37ed6 100644 --- a/arch/arm64/kvm/sys_regs.c +++ b/arch/arm64/kvm/sys_regs.c @@ -18,6 +18,7 @@ #include #include +#include #include #include #include diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index bd45fbcb9a5a..18046cf4b3a3 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -451,11 +451,6 @@ static const struct attribute_group armv8_pmuv3_caps_attr_group = { .attrs = armv8_pmuv3_caps_attrs, }; -/* - * Perf Events' indices - */ -#define ARMV8_IDX_CYCLE_COUNTER 31 - /* * We unconditionally enable ARMv8.5-PMU long event counter support * (64-bit events) where supported. Indicate if this arm_pmu has long @@ -574,7 +569,7 @@ static u64 armv8pmu_read_counter(struct perf_event *event) int idx = hwc->idx; u64 value; - if (idx == ARMV8_IDX_CYCLE_COUNTER) + if (idx == ARMV8_PMU_CYCLE_IDX) value = read_pmccntr(); else value = armv8pmu_read_hw_counter(event); @@ -607,7 +602,7 @@ static void armv8pmu_write_counter(struct perf_event *event, u64 value) value = armv8pmu_bias_long_counter(event, value); - if (idx == ARMV8_IDX_CYCLE_COUNTER) + if (idx == ARMV8_PMU_CYCLE_IDX) write_pmccntr(value); else armv8pmu_write_hw_counter(event, value); @@ -644,7 +639,7 @@ static void armv8pmu_write_event_type(struct perf_event *event) armv8pmu_write_evtype(idx - 1, hwc->config_base); armv8pmu_write_evtype(idx, chain_evt); } else { - if (idx == ARMV8_IDX_CYCLE_COUNTER) + if (idx == ARMV8_PMU_CYCLE_IDX) write_pmccfiltr(hwc->config_base); else armv8pmu_write_evtype(idx, hwc->config_base); @@ -772,7 +767,7 @@ static void armv8pmu_enable_user_access(struct arm_pmu *cpu_pmu) /* Clear any unused counters to avoid leaking their contents */ for_each_andnot_bit(i, cpu_pmu->cntr_mask, cpuc->used_mask, ARMPMU_MAX_HWEVENTS) { - if (i == ARMV8_IDX_CYCLE_COUNTER) + if (i == ARMV8_PMU_CYCLE_IDX) write_pmccntr(0); else armv8pmu_write_evcntr(i, 0); @@ -933,8 +928,8 @@ static int armv8pmu_get_event_idx(struct pmu_hw_events *cpuc, /* Always prefer to place a cycle counter into the cycle counter. */ if ((evtype == ARMV8_PMUV3_PERFCTR_CPU_CYCLES) && !armv8pmu_event_get_threshold(&event->attr)) { - if (!test_and_set_bit(ARMV8_IDX_CYCLE_COUNTER, cpuc->used_mask)) - return ARMV8_IDX_CYCLE_COUNTER; + if (!test_and_set_bit(ARMV8_PMU_CYCLE_IDX, cpuc->used_mask)) + return ARMV8_PMU_CYCLE_IDX; else if (armv8pmu_event_is_64bit(event) && armv8pmu_event_want_user_access(event) && !armv8pmu_has_long_event(cpu_pmu)) @@ -1196,7 +1191,7 @@ static void __armv8pmu_probe_pmu(void *info) 0, FIELD_GET(ARMV8_PMU_PMCR_N, armv8pmu_pmcr_read())); /* Add the CPU cycles counter */ - set_bit(ARMV8_IDX_CYCLE_COUNTER, cpu_pmu->cntr_mask); + set_bit(ARMV8_PMU_CYCLE_IDX, cpu_pmu->cntr_mask); pmceid[0] = pmceid_raw[0] = read_pmceid0(); pmceid[1] = pmceid_raw[1] = read_pmceid1(); diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h index 334d7c5503cf..871067fb2616 100644 --- a/include/kvm/arm_pmu.h +++ b/include/kvm/arm_pmu.h @@ -10,7 +10,6 @@ #include #include -#define ARMV8_PMU_CYCLE_IDX (ARMV8_PMU_MAX_COUNTERS - 1) #if IS_ENABLED(CONFIG_HW_PERF_EVENTS) && IS_ENABLED(CONFIG_KVM) struct kvm_pmc { diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h index 792b8e10b72a..f4ec76f725a3 100644 --- a/include/linux/perf/arm_pmuv3.h +++ b/include/linux/perf/arm_pmuv3.h @@ -9,6 +9,9 @@ #define ARMV8_PMU_MAX_GENERAL_COUNTERS 31 #define ARMV8_PMU_MAX_COUNTERS 32 +#define ARMV8_PMU_CYCLE_IDX 31 + + /* * Common architectural and microarchitectural event numbers. */ -- cgit v1.2.3 From d8226d8cfbaf5eb9771af8ad8b4e58697e2ffb74 Mon Sep 17 00:00:00 2001 From: "Rob Herring (Arm)" Date: Wed, 31 Jul 2024 10:51:24 -0600 Subject: perf: arm_pmuv3: Add support for Armv9.4 PMU instruction counter Armv9.4/8.9 PMU adds optional support for a fixed instruction counter similar to the fixed cycle counter. Support for the feature is indicated in the ID_AA64DFR1_EL1 register PMICNTR field. The counter is not accessible in AArch32. Existing userspace using direct counter access won't know how to handle the fixed instruction counter, so we have to avoid using the counter when user access is requested. Acked-by: Mark Rutland Signed-off-by: Rob Herring (Arm) Tested-by: James Clark Link: https://lore.kernel.org/r/20240731-arm-pmu-3-9-icntr-v3-7-280a8d7ff465@kernel.org Signed-off-by: Will Deacon --- arch/arm/include/asm/arm_pmuv3.h | 20 ++++++++++++++++++++ arch/arm64/include/asm/arm_pmuv3.h | 28 ++++++++++++++++++++++++++++ arch/arm64/kvm/pmu.c | 8 ++++++-- arch/arm64/tools/sysreg | 25 +++++++++++++++++++++++++ drivers/perf/arm_pmuv3.c | 25 +++++++++++++++++++++++++ include/linux/perf/arm_pmu.h | 8 ++++++-- include/linux/perf/arm_pmuv3.h | 6 ++++-- 7 files changed, 114 insertions(+), 6 deletions(-) (limited to 'drivers') diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h index a41b503b7dcd..f63ba8986b24 100644 --- a/arch/arm/include/asm/arm_pmuv3.h +++ b/arch/arm/include/asm/arm_pmuv3.h @@ -127,6 +127,12 @@ static inline u32 read_pmuver(void) return (dfr0 >> 24) & 0xf; } +static inline bool pmuv3_has_icntr(void) +{ + /* FEAT_PMUv3_ICNTR not accessible for 32-bit */ + return false; +} + static inline void write_pmcr(u32 val) { write_sysreg(val, PMCR); @@ -152,6 +158,13 @@ static inline u64 read_pmccntr(void) return read_sysreg(PMCCNTR); } +static inline void write_pmicntr(u64 val) {} + +static inline u64 read_pmicntr(void) +{ + return 0; +} + static inline void write_pmcntenset(u32 val) { write_sysreg(val, PMCNTENSET); @@ -177,6 +190,13 @@ static inline void write_pmccfiltr(u32 val) write_sysreg(val, PMCCFILTR); } +static inline void write_pmicfiltr(u64 val) {} + +static inline u64 read_pmicfiltr(void) +{ + return 0; +} + static inline void write_pmovsclr(u32 val) { write_sysreg(val, PMOVSR); diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h index 36c3e82b4eec..468a049bc63b 100644 --- a/arch/arm64/include/asm/arm_pmuv3.h +++ b/arch/arm64/include/asm/arm_pmuv3.h @@ -54,6 +54,14 @@ static inline u32 read_pmuver(void) ID_AA64DFR0_EL1_PMUVer_SHIFT); } +static inline bool pmuv3_has_icntr(void) +{ + u64 dfr1 = read_sysreg(id_aa64dfr1_el1); + + return !!cpuid_feature_extract_unsigned_field(dfr1, + ID_AA64DFR1_EL1_PMICNTR_SHIFT); +} + static inline void write_pmcr(u64 val) { write_sysreg(val, pmcr_el0); @@ -79,6 +87,16 @@ static inline u64 read_pmccntr(void) return read_sysreg(pmccntr_el0); } +static inline void write_pmicntr(u64 val) +{ + write_sysreg_s(val, SYS_PMICNTR_EL0); +} + +static inline u64 read_pmicntr(void) +{ + return read_sysreg_s(SYS_PMICNTR_EL0); +} + static inline void write_pmcntenset(u64 val) { write_sysreg(val, pmcntenset_el0); @@ -109,6 +127,16 @@ static inline u64 read_pmccfiltr(void) return read_sysreg(pmccfiltr_el0); } +static inline void write_pmicfiltr(u64 val) +{ + write_sysreg_s(val, SYS_PMICFILTR_EL0); +} + +static inline u64 read_pmicfiltr(void) +{ + return read_sysreg_s(SYS_PMICFILTR_EL0); +} + static inline void write_pmovsclr(u64 val) { write_sysreg(val, pmovsclr_el0); diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c index 215b74875815..0b3adf3e17b4 100644 --- a/arch/arm64/kvm/pmu.c +++ b/arch/arm64/kvm/pmu.c @@ -66,24 +66,28 @@ void kvm_clr_pmu_events(u64 clr) /* * Read a value direct from PMEVTYPER where idx is 0-30 - * or PMCCFILTR_EL0 where idx is ARMV8_PMU_CYCLE_IDX (31). + * or PMxCFILTR_EL0 where idx is 31-32. */ static u64 kvm_vcpu_pmu_read_evtype_direct(int idx) { if (idx == ARMV8_PMU_CYCLE_IDX) return read_pmccfiltr(); + else if (idx == ARMV8_PMU_INSTR_IDX) + return read_pmicfiltr(); return read_pmevtypern(idx); } /* * Write a value direct to PMEVTYPER where idx is 0-30 - * or PMCCFILTR_EL0 where idx is ARMV8_PMU_CYCLE_IDX (31). + * or PMxCFILTR_EL0 where idx is 31-32. */ static void kvm_vcpu_pmu_write_evtype_direct(int idx, u32 val) { if (idx == ARMV8_PMU_CYCLE_IDX) write_pmccfiltr(val); + else if (idx == ARMV8_PMU_INSTR_IDX) + write_pmicfiltr(val); else write_pmevtypern(idx, val); } diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg index 37aa7eaad07b..8d637ac4b7c6 100644 --- a/arch/arm64/tools/sysreg +++ b/arch/arm64/tools/sysreg @@ -2029,6 +2029,31 @@ Sysreg FAR_EL1 3 0 6 0 0 Field 63:0 ADDR EndSysreg +Sysreg PMICNTR_EL0 3 3 9 4 0 +Field 63:0 ICNT +EndSysreg + +Sysreg PMICFILTR_EL0 3 3 9 6 0 +Res0 63:59 +Field 58 SYNC +Field 57:56 VS +Res0 55:32 +Field 31 P +Field 30 U +Field 29 NSK +Field 28 NSU +Field 27 NSH +Field 26 M +Res0 25 +Field 24 SH +Field 23 T +Field 22 RLK +Field 21 RLU +Field 20 RLH +Res0 19:16 +Field 15:0 evtCount +EndSysreg + Sysreg PMSCR_EL1 3 0 9 9 0 Res0 63:8 Field 7:6 PCT diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index 18046cf4b3a3..4d000532a07f 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -571,6 +571,8 @@ static u64 armv8pmu_read_counter(struct perf_event *event) if (idx == ARMV8_PMU_CYCLE_IDX) value = read_pmccntr(); + else if (idx == ARMV8_PMU_INSTR_IDX) + value = read_pmicntr(); else value = armv8pmu_read_hw_counter(event); @@ -604,6 +606,8 @@ static void armv8pmu_write_counter(struct perf_event *event, u64 value) if (idx == ARMV8_PMU_CYCLE_IDX) write_pmccntr(value); + else if (idx == ARMV8_PMU_INSTR_IDX) + write_pmicntr(value); else armv8pmu_write_hw_counter(event, value); } @@ -641,6 +645,8 @@ static void armv8pmu_write_event_type(struct perf_event *event) } else { if (idx == ARMV8_PMU_CYCLE_IDX) write_pmccfiltr(hwc->config_base); + else if (idx == ARMV8_PMU_INSTR_IDX) + write_pmicfiltr(hwc->config_base); else armv8pmu_write_evtype(idx, hwc->config_base); } @@ -769,6 +775,8 @@ static void armv8pmu_enable_user_access(struct arm_pmu *cpu_pmu) ARMPMU_MAX_HWEVENTS) { if (i == ARMV8_PMU_CYCLE_IDX) write_pmccntr(0); + else if (i == ARMV8_PMU_INSTR_IDX) + write_pmicntr(0); else armv8pmu_write_evcntr(i, 0); } @@ -936,6 +944,19 @@ static int armv8pmu_get_event_idx(struct pmu_hw_events *cpuc, return -EAGAIN; } + /* + * Always prefer to place a instruction counter into the instruction counter, + * but don't expose the instruction counter to userspace access as userspace + * may not know how to handle it. + */ + if ((evtype == ARMV8_PMUV3_PERFCTR_INST_RETIRED) && + !armv8pmu_event_get_threshold(&event->attr) && + test_bit(ARMV8_PMU_INSTR_IDX, cpu_pmu->cntr_mask) && + !armv8pmu_event_want_user_access(event)) { + if (!test_and_set_bit(ARMV8_PMU_INSTR_IDX, cpuc->used_mask)) + return ARMV8_PMU_INSTR_IDX; + } + /* * Otherwise use events counters */ @@ -1193,6 +1214,10 @@ static void __armv8pmu_probe_pmu(void *info) /* Add the CPU cycles counter */ set_bit(ARMV8_PMU_CYCLE_IDX, cpu_pmu->cntr_mask); + /* Add the CPU instructions counter */ + if (pmuv3_has_icntr()) + set_bit(ARMV8_PMU_INSTR_IDX, cpu_pmu->cntr_mask); + pmceid[0] = pmceid_raw[0] = read_pmceid0(); pmceid[1] = pmceid_raw[1] = read_pmceid1(); diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h index e5d6d204beab..4b5b83677e3f 100644 --- a/include/linux/perf/arm_pmu.h +++ b/include/linux/perf/arm_pmu.h @@ -17,10 +17,14 @@ #ifdef CONFIG_ARM_PMU /* - * The ARMv7 CPU PMU supports up to 32 event counters. + * The Armv7 and Armv8.8 or less CPU PMU supports up to 32 event counters. + * The Armv8.9/9.4 CPU PMU supports up to 33 event counters. */ +#ifdef CONFIG_ARM #define ARMPMU_MAX_HWEVENTS 32 - +#else +#define ARMPMU_MAX_HWEVENTS 33 +#endif /* * ARM PMU hw_event flags */ diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h index 4f7a7f2222e5..3372c1b56486 100644 --- a/include/linux/perf/arm_pmuv3.h +++ b/include/linux/perf/arm_pmuv3.h @@ -8,7 +8,7 @@ #define ARMV8_PMU_MAX_GENERAL_COUNTERS 31 #define ARMV8_PMU_CYCLE_IDX 31 - +#define ARMV8_PMU_INSTR_IDX 32 /* Not accessible from AArch32 */ /* * Common architectural and microarchitectural event numbers. @@ -228,8 +228,10 @@ */ #define ARMV8_PMU_OVSR_P GENMASK(30, 0) #define ARMV8_PMU_OVSR_C BIT(31) +#define ARMV8_PMU_OVSR_F BIT_ULL(32) /* arm64 only */ /* Mask for writable bits is both P and C fields */ -#define ARMV8_PMU_OVERFLOWED_MASK (ARMV8_PMU_OVSR_P | ARMV8_PMU_OVSR_C) +#define ARMV8_PMU_OVERFLOWED_MASK (ARMV8_PMU_OVSR_P | ARMV8_PMU_OVSR_C | \ + ARMV8_PMU_OVSR_F) /* * PMXEVTYPER: Event selection reg -- cgit v1.2.3 From f3b78b470f28bb2a3a40e88bdf5c6de6a35a9b76 Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Wed, 31 Jul 2024 17:26:58 +0800 Subject: ACPI/IORT: Add PMCG platform information for HiSilicon HIP10/11 HiSilicon HIP10/11 platforms using the same SMMU PMCG with HIP09 and thus suffers the same erratum. List them in the PMCG platform information list without introducing a new SMMU PMCG Model. Update the silicon-errata.rst as well. Signed-off-by: Yicong Yang Link: https://lore.kernel.org/r/20240731092658.11012-1-yangyicong@huawei.com Signed-off-by: Will Deacon --- Documentation/arch/arm64/silicon-errata.rst | 4 ++-- drivers/acpi/arm64/iort.c | 7 +++++++ 2 files changed, 9 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/Documentation/arch/arm64/silicon-errata.rst b/Documentation/arch/arm64/silicon-errata.rst index 50327c05be8d..35892c28d831 100644 --- a/Documentation/arch/arm64/silicon-errata.rst +++ b/Documentation/arch/arm64/silicon-errata.rst @@ -249,8 +249,8 @@ stable kernels. +----------------+-----------------+-----------------+-----------------------------+ | Hisilicon | Hip08 SMMU PMCG | #162001800 | N/A | +----------------+-----------------+-----------------+-----------------------------+ -| Hisilicon | Hip08 SMMU PMCG | #162001900 | N/A | -| | Hip09 SMMU PMCG | | | +| Hisilicon | Hip{08,09,10,10C| #162001900 | N/A | +| | ,11} SMMU PMCG | | | +----------------+-----------------+-----------------+-----------------------------+ +----------------+-----------------+-----------------+-----------------------------+ | Qualcomm Tech. | Kryo/Falkor v1 | E1003 | QCOM_FALKOR_ERRATUM_1003 | diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c index 034d303c3df1..4c745a26226b 100644 --- a/drivers/acpi/arm64/iort.c +++ b/drivers/acpi/arm64/iort.c @@ -1703,6 +1703,13 @@ static struct acpi_platform_list pmcg_plat_info[] __initdata = { /* HiSilicon Hip09 Platform */ {"HISI ", "HIP09 ", 0, ACPI_SIG_IORT, greater_than_or_equal, "Erratum #162001900", IORT_SMMU_V3_PMCG_HISI_HIP09}, + /* HiSilicon Hip10/11 Platform uses the same SMMU IP with Hip09 */ + {"HISI ", "HIP10 ", 0, ACPI_SIG_IORT, greater_than_or_equal, + "Erratum #162001900", IORT_SMMU_V3_PMCG_HISI_HIP09}, + {"HISI ", "HIP10C ", 0, ACPI_SIG_IORT, greater_than_or_equal, + "Erratum #162001900", IORT_SMMU_V3_PMCG_HISI_HIP09}, + {"HISI ", "HIP11 ", 0, ACPI_SIG_IORT, greater_than_or_equal, + "Erratum #162001900", IORT_SMMU_V3_PMCG_HISI_HIP09}, { } }; -- cgit v1.2.3 From 3cce331ee2a4cc99328757a9bd297ad9a39609e8 Mon Sep 17 00:00:00 2001 From: Yangyu Chen Date: Wed, 7 Aug 2024 11:35:18 +0900 Subject: drivers/perf: apple_m1: add known PMU events This patch adds known PMU events that can be found on /usr/share/kpep in macOS. The m1_pmu_events and m1_pmu_event_affinity are generated from the script [1], which consumes the plist file from Apple. And then added these events to m1_pmu_perf_map and m1_pmu_event_attrs with Apple's documentation [2]. Link: https://github.com/cyyself/m1-pmu-gen [1] Link: https://developer.apple.com/download/apple-silicon-cpu-optimization-guide/ [2] Signed-off-by: Yangyu Chen Acked-by: Hector Martin Link: https://lore.kernel.org/r/tencent_C5DA658E64B8D13125210C8D707CD8823F08@qq.com Signed-off-by: Will Deacon --- drivers/perf/apple_m1_cpu_pmu.c | 178 ++++++++++++++++++++++++---------------- 1 file changed, 105 insertions(+), 73 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/apple_m1_cpu_pmu.c b/drivers/perf/apple_m1_cpu_pmu.c index c8f607912567..1d4d01e1275e 100644 --- a/drivers/perf/apple_m1_cpu_pmu.c +++ b/drivers/perf/apple_m1_cpu_pmu.c @@ -47,46 +47,79 @@ * implementations, we'll have to introduce per cpu-type tables. */ enum m1_pmu_events { - M1_PMU_PERFCTR_UNKNOWN_01 = 0x01, - M1_PMU_PERFCTR_CPU_CYCLES = 0x02, - M1_PMU_PERFCTR_INSTRUCTIONS = 0x8c, - M1_PMU_PERFCTR_UNKNOWN_8d = 0x8d, - M1_PMU_PERFCTR_UNKNOWN_8e = 0x8e, - M1_PMU_PERFCTR_UNKNOWN_8f = 0x8f, - M1_PMU_PERFCTR_UNKNOWN_90 = 0x90, - M1_PMU_PERFCTR_UNKNOWN_93 = 0x93, - M1_PMU_PERFCTR_UNKNOWN_94 = 0x94, - M1_PMU_PERFCTR_UNKNOWN_95 = 0x95, - M1_PMU_PERFCTR_UNKNOWN_96 = 0x96, - M1_PMU_PERFCTR_UNKNOWN_97 = 0x97, - M1_PMU_PERFCTR_UNKNOWN_98 = 0x98, - M1_PMU_PERFCTR_UNKNOWN_99 = 0x99, - M1_PMU_PERFCTR_UNKNOWN_9a = 0x9a, - M1_PMU_PERFCTR_UNKNOWN_9b = 0x9b, - M1_PMU_PERFCTR_UNKNOWN_9c = 0x9c, - M1_PMU_PERFCTR_UNKNOWN_9f = 0x9f, - M1_PMU_PERFCTR_UNKNOWN_bf = 0xbf, - M1_PMU_PERFCTR_UNKNOWN_c0 = 0xc0, - M1_PMU_PERFCTR_UNKNOWN_c1 = 0xc1, - M1_PMU_PERFCTR_UNKNOWN_c4 = 0xc4, - M1_PMU_PERFCTR_UNKNOWN_c5 = 0xc5, - M1_PMU_PERFCTR_UNKNOWN_c6 = 0xc6, - M1_PMU_PERFCTR_UNKNOWN_c8 = 0xc8, - M1_PMU_PERFCTR_UNKNOWN_ca = 0xca, - M1_PMU_PERFCTR_UNKNOWN_cb = 0xcb, - M1_PMU_PERFCTR_UNKNOWN_f5 = 0xf5, - M1_PMU_PERFCTR_UNKNOWN_f6 = 0xf6, - M1_PMU_PERFCTR_UNKNOWN_f7 = 0xf7, - M1_PMU_PERFCTR_UNKNOWN_f8 = 0xf8, - M1_PMU_PERFCTR_UNKNOWN_fd = 0xfd, - M1_PMU_PERFCTR_LAST = M1_PMU_CFG_EVENT, + M1_PMU_PERFCTR_RETIRE_UOP = 0x1, + M1_PMU_PERFCTR_CORE_ACTIVE_CYCLE = 0x2, + M1_PMU_PERFCTR_L1I_TLB_FILL = 0x4, + M1_PMU_PERFCTR_L1D_TLB_FILL = 0x5, + M1_PMU_PERFCTR_MMU_TABLE_WALK_INSTRUCTION = 0x7, + M1_PMU_PERFCTR_MMU_TABLE_WALK_DATA = 0x8, + M1_PMU_PERFCTR_L2_TLB_MISS_INSTRUCTION = 0xa, + M1_PMU_PERFCTR_L2_TLB_MISS_DATA = 0xb, + M1_PMU_PERFCTR_MMU_VIRTUAL_MEMORY_FAULT_NONSPEC = 0xd, + M1_PMU_PERFCTR_SCHEDULE_UOP = 0x52, + M1_PMU_PERFCTR_INTERRUPT_PENDING = 0x6c, + M1_PMU_PERFCTR_MAP_STALL_DISPATCH = 0x70, + M1_PMU_PERFCTR_MAP_REWIND = 0x75, + M1_PMU_PERFCTR_MAP_STALL = 0x76, + M1_PMU_PERFCTR_MAP_INT_UOP = 0x7c, + M1_PMU_PERFCTR_MAP_LDST_UOP = 0x7d, + M1_PMU_PERFCTR_MAP_SIMD_UOP = 0x7e, + M1_PMU_PERFCTR_FLUSH_RESTART_OTHER_NONSPEC = 0x84, + M1_PMU_PERFCTR_INST_ALL = 0x8c, + M1_PMU_PERFCTR_INST_BRANCH = 0x8d, + M1_PMU_PERFCTR_INST_BRANCH_CALL = 0x8e, + M1_PMU_PERFCTR_INST_BRANCH_RET = 0x8f, + M1_PMU_PERFCTR_INST_BRANCH_TAKEN = 0x90, + M1_PMU_PERFCTR_INST_BRANCH_INDIR = 0x93, + M1_PMU_PERFCTR_INST_BRANCH_COND = 0x94, + M1_PMU_PERFCTR_INST_INT_LD = 0x95, + M1_PMU_PERFCTR_INST_INT_ST = 0x96, + M1_PMU_PERFCTR_INST_INT_ALU = 0x97, + M1_PMU_PERFCTR_INST_SIMD_LD = 0x98, + M1_PMU_PERFCTR_INST_SIMD_ST = 0x99, + M1_PMU_PERFCTR_INST_SIMD_ALU = 0x9a, + M1_PMU_PERFCTR_INST_LDST = 0x9b, + M1_PMU_PERFCTR_INST_BARRIER = 0x9c, + M1_PMU_PERFCTR_UNKNOWN_9f = 0x9f, + M1_PMU_PERFCTR_L1D_TLB_ACCESS = 0xa0, + M1_PMU_PERFCTR_L1D_TLB_MISS = 0xa1, + M1_PMU_PERFCTR_L1D_CACHE_MISS_ST = 0xa2, + M1_PMU_PERFCTR_L1D_CACHE_MISS_LD = 0xa3, + M1_PMU_PERFCTR_LD_UNIT_UOP = 0xa6, + M1_PMU_PERFCTR_ST_UNIT_UOP = 0xa7, + M1_PMU_PERFCTR_L1D_CACHE_WRITEBACK = 0xa8, + M1_PMU_PERFCTR_LDST_X64_UOP = 0xb1, + M1_PMU_PERFCTR_LDST_XPG_UOP = 0xb2, + M1_PMU_PERFCTR_ATOMIC_OR_EXCLUSIVE_SUCC = 0xb3, + M1_PMU_PERFCTR_ATOMIC_OR_EXCLUSIVE_FAIL = 0xb4, + M1_PMU_PERFCTR_L1D_CACHE_MISS_LD_NONSPEC = 0xbf, + M1_PMU_PERFCTR_L1D_CACHE_MISS_ST_NONSPEC = 0xc0, + M1_PMU_PERFCTR_L1D_TLB_MISS_NONSPEC = 0xc1, + M1_PMU_PERFCTR_ST_MEMORY_ORDER_VIOLATION_NONSPEC = 0xc4, + M1_PMU_PERFCTR_BRANCH_COND_MISPRED_NONSPEC = 0xc5, + M1_PMU_PERFCTR_BRANCH_INDIR_MISPRED_NONSPEC = 0xc6, + M1_PMU_PERFCTR_BRANCH_RET_INDIR_MISPRED_NONSPEC = 0xc8, + M1_PMU_PERFCTR_BRANCH_CALL_INDIR_MISPRED_NONSPEC = 0xca, + M1_PMU_PERFCTR_BRANCH_MISPRED_NONSPEC = 0xcb, + M1_PMU_PERFCTR_L1I_TLB_MISS_DEMAND = 0xd4, + M1_PMU_PERFCTR_MAP_DISPATCH_BUBBLE = 0xd6, + M1_PMU_PERFCTR_L1I_CACHE_MISS_DEMAND = 0xdb, + M1_PMU_PERFCTR_FETCH_RESTART = 0xde, + M1_PMU_PERFCTR_ST_NT_UOP = 0xe5, + M1_PMU_PERFCTR_LD_NT_UOP = 0xe6, + M1_PMU_PERFCTR_UNKNOWN_f5 = 0xf5, + M1_PMU_PERFCTR_UNKNOWN_f6 = 0xf6, + M1_PMU_PERFCTR_UNKNOWN_f7 = 0xf7, + M1_PMU_PERFCTR_UNKNOWN_f8 = 0xf8, + M1_PMU_PERFCTR_UNKNOWN_fd = 0xfd, + M1_PMU_PERFCTR_LAST = M1_PMU_CFG_EVENT, /* * From this point onwards, these are not actual HW events, * but attributes that get stored in hw->config_base. */ - M1_PMU_CFG_COUNT_USER = BIT(8), - M1_PMU_CFG_COUNT_KERNEL = BIT(9), + M1_PMU_CFG_COUNT_USER = BIT(8), + M1_PMU_CFG_COUNT_KERNEL = BIT(9), }; /* @@ -96,46 +129,45 @@ enum m1_pmu_events { * counters had strange affinities. */ static const u16 m1_pmu_event_affinity[M1_PMU_PERFCTR_LAST + 1] = { - [0 ... M1_PMU_PERFCTR_LAST] = ANY_BUT_0_1, - [M1_PMU_PERFCTR_UNKNOWN_01] = BIT(7), - [M1_PMU_PERFCTR_CPU_CYCLES] = ANY_BUT_0_1 | BIT(0), - [M1_PMU_PERFCTR_INSTRUCTIONS] = BIT(7) | BIT(1), - [M1_PMU_PERFCTR_UNKNOWN_8d] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_8e] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_8f] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_90] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_93] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_94] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_95] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_96] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_97] = BIT(7), - [M1_PMU_PERFCTR_UNKNOWN_98] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_99] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_9a] = BIT(7), - [M1_PMU_PERFCTR_UNKNOWN_9b] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_9c] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_9f] = BIT(7), - [M1_PMU_PERFCTR_UNKNOWN_bf] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c0] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c1] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c4] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c5] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c6] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_c8] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_ca] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_cb] = ONLY_5_6_7, - [M1_PMU_PERFCTR_UNKNOWN_f5] = ONLY_2_4_6, - [M1_PMU_PERFCTR_UNKNOWN_f6] = ONLY_2_4_6, - [M1_PMU_PERFCTR_UNKNOWN_f7] = ONLY_2_4_6, - [M1_PMU_PERFCTR_UNKNOWN_f8] = ONLY_2_TO_7, - [M1_PMU_PERFCTR_UNKNOWN_fd] = ONLY_2_4_6, + [0 ... M1_PMU_PERFCTR_LAST] = ANY_BUT_0_1, + [M1_PMU_PERFCTR_RETIRE_UOP] = BIT(7), + [M1_PMU_PERFCTR_CORE_ACTIVE_CYCLE] = ANY_BUT_0_1 | BIT(0), + [M1_PMU_PERFCTR_INST_ALL] = BIT(7) | BIT(1), + [M1_PMU_PERFCTR_INST_BRANCH] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_BRANCH_CALL] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_BRANCH_RET] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_BRANCH_TAKEN] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_BRANCH_INDIR] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_BRANCH_COND] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_INT_LD] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_INT_ST] = BIT(7), + [M1_PMU_PERFCTR_INST_INT_ALU] = BIT(7), + [M1_PMU_PERFCTR_INST_SIMD_LD] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_SIMD_ST] = ONLY_5_6_7, + [M1_PMU_PERFCTR_INST_SIMD_ALU] = BIT(7), + [M1_PMU_PERFCTR_INST_LDST] = BIT(7), + [M1_PMU_PERFCTR_INST_BARRIER] = ONLY_5_6_7, + [M1_PMU_PERFCTR_UNKNOWN_9f] = BIT(7), + [M1_PMU_PERFCTR_L1D_CACHE_MISS_LD_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_L1D_CACHE_MISS_ST_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_L1D_TLB_MISS_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_ST_MEMORY_ORDER_VIOLATION_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_BRANCH_COND_MISPRED_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_BRANCH_INDIR_MISPRED_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_BRANCH_RET_INDIR_MISPRED_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_BRANCH_CALL_INDIR_MISPRED_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_BRANCH_MISPRED_NONSPEC] = ONLY_5_6_7, + [M1_PMU_PERFCTR_UNKNOWN_f5] = ONLY_2_4_6, + [M1_PMU_PERFCTR_UNKNOWN_f6] = ONLY_2_4_6, + [M1_PMU_PERFCTR_UNKNOWN_f7] = ONLY_2_4_6, + [M1_PMU_PERFCTR_UNKNOWN_f8] = ONLY_2_TO_7, + [M1_PMU_PERFCTR_UNKNOWN_fd] = ONLY_2_4_6, }; static const unsigned m1_pmu_perf_map[PERF_COUNT_HW_MAX] = { PERF_MAP_ALL_UNSUPPORTED, - [PERF_COUNT_HW_CPU_CYCLES] = M1_PMU_PERFCTR_CPU_CYCLES, - [PERF_COUNT_HW_INSTRUCTIONS] = M1_PMU_PERFCTR_INSTRUCTIONS, - /* No idea about the rest yet */ + [PERF_COUNT_HW_CPU_CYCLES] = M1_PMU_PERFCTR_CORE_ACTIVE_CYCLE, + [PERF_COUNT_HW_INSTRUCTIONS] = M1_PMU_PERFCTR_INST_ALL, }; /* sysfs definitions */ @@ -154,8 +186,8 @@ static ssize_t m1_pmu_events_sysfs_show(struct device *dev, PMU_EVENT_ATTR_ID(name, m1_pmu_events_sysfs_show, config) static struct attribute *m1_pmu_event_attrs[] = { - M1_PMU_EVENT_ATTR(cycles, M1_PMU_PERFCTR_CPU_CYCLES), - M1_PMU_EVENT_ATTR(instructions, M1_PMU_PERFCTR_INSTRUCTIONS), + M1_PMU_EVENT_ATTR(cycles, M1_PMU_PERFCTR_CORE_ACTIVE_CYCLE), + M1_PMU_EVENT_ATTR(instructions, M1_PMU_PERFCTR_INST_ALL), NULL, }; -- cgit v1.2.3 From a3dd920977dccc453c550260c4b7605b280b79c3 Mon Sep 17 00:00:00 2001 From: Jing Zhang Date: Thu, 22 Aug 2024 11:33:31 +0800 Subject: drivers/perf: Fix ali_drw_pmu driver interrupt status clearing The alibaba_uncore_pmu driver forgot to clear all interrupt status in the interrupt processing function. After the PMU counter overflow interrupt occurred, an interrupt storm occurred, causing the system to hang. Therefore, clear the correct interrupt status in the interrupt handling function to fix it. Fixes: cf7b61073e45 ("drivers/perf: add DDR Sub-System Driveway PMU driver for Yitian 710 SoC") Signed-off-by: Jing Zhang Reviewed-by: Shuai Xue Acked-by: Mark Rutland Link: https://lore.kernel.org/r/1724297611-20686-1-git-send-email-renyu.zj@linux.alibaba.com Signed-off-by: Will Deacon --- drivers/perf/alibaba_uncore_drw_pmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/alibaba_uncore_drw_pmu.c b/drivers/perf/alibaba_uncore_drw_pmu.c index 38a2947ae813..c6ff1bc7d336 100644 --- a/drivers/perf/alibaba_uncore_drw_pmu.c +++ b/drivers/perf/alibaba_uncore_drw_pmu.c @@ -400,7 +400,7 @@ static irqreturn_t ali_drw_pmu_isr(int irq_num, void *data) } /* clear common counter intr status */ - clr_status = FIELD_PREP(ALI_DRW_PMCOM_CNT_OV_INTR_MASK, 1); + clr_status = FIELD_PREP(ALI_DRW_PMCOM_CNT_OV_INTR_MASK, status); writel(clr_status, drw_pmu->cfg_base + ALI_DRW_PMU_OV_INTR_CLR); } -- cgit v1.2.3 From e669388537c472142804eb5a0449cc23d5409694 Mon Sep 17 00:00:00 2001 From: Krishna chaitanya chundru Date: Fri, 16 Aug 2024 20:47:20 +0530 Subject: perf/dwc_pcie: Fix registration issue in multi PCIe controller instances When there are multiple of instances of PCIe controllers, registration to perf driver fails with this error. sysfs: cannot create duplicate filename '/devices/platform/dwc_pcie_pmu.0' CPU: 0 PID: 166 Comm: modprobe Not tainted 6.10.0-rc2-next-20240607-dirty Hardware name: Qualcomm SA8775P Ride (DT) Call trace: dump_backtrace.part.8+0x98/0xf0 show_stack+0x14/0x1c dump_stack_lvl+0x74/0x88 dump_stack+0x14/0x1c sysfs_warn_dup+0x60/0x78 sysfs_create_dir_ns+0xe8/0x100 kobject_add_internal+0x94/0x224 kobject_add+0xa8/0x118 device_add+0x298/0x7b4 platform_device_add+0x1a0/0x228 platform_device_register_full+0x11c/0x148 dwc_pcie_register_dev+0x74/0xf0 [dwc_pcie_pmu] dwc_pcie_pmu_init+0x7c/0x1000 [dwc_pcie_pmu] do_one_initcall+0x58/0x1c0 do_init_module+0x58/0x208 load_module+0x1804/0x188c __do_sys_init_module+0x18c/0x1f0 __arm64_sys_init_module+0x14/0x1c invoke_syscall+0x40/0xf8 el0_svc_common.constprop.1+0x70/0xf4 do_el0_svc+0x18/0x20 el0_svc+0x28/0xb0 el0t_64_sync_handler+0x9c/0xc0 el0t_64_sync+0x160/0x164 kobject: kobject_add_internal failed for dwc_pcie_pmu.0 with -EEXIST, don't try to register things with the same name in the same directory. This is because of having same bdf value for devices under two different controllers. Update the logic to use sbdf which is a unique number in case of multi instance also. Fixes: af9597adc2f1 ("drivers/perf: add DesignWare PCIe PMU driver") Signed-off-by: Krishna chaitanya chundru Reviewed-by: Yicong Yang Link: https://lore.kernel.org/r/20240816-dwc_pmu_fix-v2-1-198b8ab1077c@quicinc.com Signed-off-by: Will Deacon --- drivers/perf/dwc_pcie_pmu.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/dwc_pcie_pmu.c b/drivers/perf/dwc_pcie_pmu.c index c5e328f23841..85a5155d6018 100644 --- a/drivers/perf/dwc_pcie_pmu.c +++ b/drivers/perf/dwc_pcie_pmu.c @@ -556,10 +556,10 @@ static int dwc_pcie_register_dev(struct pci_dev *pdev) { struct platform_device *plat_dev; struct dwc_pcie_dev_info *dev_info; - u32 bdf; + u32 sbdf; - bdf = PCI_DEVID(pdev->bus->number, pdev->devfn); - plat_dev = platform_device_register_data(NULL, "dwc_pcie_pmu", bdf, + sbdf = (pci_domain_nr(pdev->bus) << 16) | PCI_DEVID(pdev->bus->number, pdev->devfn); + plat_dev = platform_device_register_data(NULL, "dwc_pcie_pmu", sbdf, pdev, sizeof(*pdev)); if (IS_ERR(plat_dev)) @@ -611,15 +611,15 @@ static int dwc_pcie_pmu_probe(struct platform_device *plat_dev) struct pci_dev *pdev = plat_dev->dev.platform_data; struct dwc_pcie_pmu *pcie_pmu; char *name; - u32 bdf, val; + u32 sbdf, val; u16 vsec; int ret; vsec = pci_find_vsec_capability(pdev, pdev->vendor, DWC_PCIE_VSEC_RAS_DES_ID); pci_read_config_dword(pdev, vsec + PCI_VNDR_HEADER, &val); - bdf = PCI_DEVID(pdev->bus->number, pdev->devfn); - name = devm_kasprintf(&plat_dev->dev, GFP_KERNEL, "dwc_rootport_%x", bdf); + sbdf = plat_dev->id; + name = devm_kasprintf(&plat_dev->dev, GFP_KERNEL, "dwc_rootport_%x", sbdf); if (!name) return -ENOMEM; @@ -650,7 +650,7 @@ static int dwc_pcie_pmu_probe(struct platform_device *plat_dev) ret = cpuhp_state_add_instance(dwc_pcie_pmu_hp_state, &pcie_pmu->cpuhp_node); if (ret) { - pci_err(pdev, "Error %d registering hotplug @%x\n", ret, bdf); + pci_err(pdev, "Error %d registering hotplug @%x\n", ret, sbdf); return ret; } @@ -663,7 +663,7 @@ static int dwc_pcie_pmu_probe(struct platform_device *plat_dev) ret = perf_pmu_register(&pcie_pmu->pmu, name, -1); if (ret) { - pci_err(pdev, "Error %d registering PMU @%x\n", ret, bdf); + pci_err(pdev, "Error %d registering PMU @%x\n", ret, sbdf); return ret; } ret = devm_add_action_or_reset(&plat_dev->dev, dwc_pcie_unregister_pmu, -- cgit v1.2.3 From b94b05478fb6a09033bf70c6edd03f8930a0fe24 Mon Sep 17 00:00:00 2001 From: Krishna chaitanya chundru Date: Fri, 16 Aug 2024 20:47:22 +0530 Subject: perf/dwc_pcie: Always register for PCIe bus notifier When the PCIe devices are discovered late, the driver can't find the PCIe devices and returns in the init without registering with the bus notifier. Due to that the devices which are discovered late the driver can't register for this. Register for bus notifier & driver even if the device is not found as part of init. Fixes: af9597adc2f1 ("drivers/perf: add DesignWare PCIe PMU driver") Signed-off-by: Krishna chaitanya chundru Reviewed-by: Yicong Yang Link: https://lore.kernel.org/r/20240816-dwc_pmu_fix-v2-3-198b8ab1077c@quicinc.com Signed-off-by: Will Deacon --- drivers/perf/dwc_pcie_pmu.c | 5 ----- 1 file changed, 5 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/dwc_pcie_pmu.c b/drivers/perf/dwc_pcie_pmu.c index 85a5155d6018..f205ecad2e4c 100644 --- a/drivers/perf/dwc_pcie_pmu.c +++ b/drivers/perf/dwc_pcie_pmu.c @@ -726,7 +726,6 @@ static struct platform_driver dwc_pcie_pmu_driver = { static int __init dwc_pcie_pmu_init(void) { struct pci_dev *pdev = NULL; - bool found = false; int ret; for_each_pci_dev(pdev) { @@ -738,11 +737,7 @@ static int __init dwc_pcie_pmu_init(void) pci_dev_put(pdev); return ret; } - - found = true; } - if (!found) - return -ENODEV; ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, "perf/dwc_pcie_pmu:online", -- cgit v1.2.3 From db9e7a83d30821ba50a84e9726099946900abde8 Mon Sep 17 00:00:00 2001 From: Krishna chaitanya chundru Date: Fri, 16 Aug 2024 20:47:23 +0530 Subject: perf/dwc_pcie: Add support for QCOM vendor devices Update the vendor table with QCOM PCIe vendorid. Signed-off-by: Krishna chaitanya chundru Reviewed-by: Yicong Yang Link: https://lore.kernel.org/r/20240816-dwc_pmu_fix-v2-4-198b8ab1077c@quicinc.com Signed-off-by: Will Deacon --- drivers/perf/dwc_pcie_pmu.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers') diff --git a/drivers/perf/dwc_pcie_pmu.c b/drivers/perf/dwc_pcie_pmu.c index f205ecad2e4c..4ca50f9b6dfe 100644 --- a/drivers/perf/dwc_pcie_pmu.c +++ b/drivers/perf/dwc_pcie_pmu.c @@ -107,6 +107,7 @@ struct dwc_pcie_vendor_id { static const struct dwc_pcie_vendor_id dwc_pcie_vendor_ids[] = { {.vendor_id = PCI_VENDOR_ID_ALIBABA }, + {.vendor_id = PCI_VENDOR_ID_QCOM }, {} /* terminator */ }; -- cgit v1.2.3 From 5e9629d0ae977d6f6916d7e519724804e95f0b07 Mon Sep 17 00:00:00 2001 From: James Clark Date: Tue, 27 Aug 2024 15:51:12 +0100 Subject: drivers/perf: arm_spe: Use perf_allow_kernel() for permissions Use perf_allow_kernel() for 'pa_enable' (physical addresses), 'pct_enable' (physical timestamps) and context IDs. This means that perf_event_paranoid is now taken into account and LSM hooks can be used, which is more consistent with other perf_event_open calls. For example PERF_SAMPLE_PHYS_ADDR uses perf_allow_kernel() rather than just perfmon_capable(). This also indirectly fixes the following error message which is misleading because perf_event_paranoid is not taken into account by perfmon_capable(): $ perf record -e arm_spe/pa_enable/ Error: Access to performance monitoring and observability operations is limited. Consider adjusting /proc/sys/kernel/perf_event_paranoid setting ... Suggested-by: Al Grant Signed-off-by: James Clark Link: https://lore.kernel.org/r/20240827145113.1224604-1-james.clark@linaro.org Link: https://lore.kernel.org/all/20240807120039.GD37996@noisy.programming.kicks-ass.net/ Signed-off-by: Will Deacon --- drivers/perf/arm_spe_pmu.c | 9 ++++----- include/linux/perf_event.h | 8 +------- kernel/events/core.c | 9 +++++++++ 3 files changed, 14 insertions(+), 12 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c index 9100d82bfabc..3569050f9cf3 100644 --- a/drivers/perf/arm_spe_pmu.c +++ b/drivers/perf/arm_spe_pmu.c @@ -41,7 +41,7 @@ /* * Cache if the event is allowed to trace Context information. - * This allows us to perform the check, i.e, perfmon_capable(), + * This allows us to perform the check, i.e, perf_allow_kernel(), * in the context of the event owner, once, during the event_init(). */ #define SPE_PMU_HW_FLAGS_CX 0x00001 @@ -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) && perfmon_capable()) + if (IS_ENABLED(CONFIG_PID_IN_CONTEXTIDR) && !perf_allow_kernel(&event->attr)) event->hw.flags |= SPE_PMU_HW_FLAGS_CX; } @@ -745,9 +745,8 @@ 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 (!perfmon_capable() && - (reg & (PMSCR_EL1_PA | PMSCR_EL1_PCT))) - return -EACCES; + if (reg & (PMSCR_EL1_PA | PMSCR_EL1_PCT)) + return perf_allow_kernel(&event->attr); return 0; } diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 1a8942277dda..e336306b8c08 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1602,13 +1602,7 @@ static inline int perf_is_paranoid(void) return sysctl_perf_event_paranoid > -1; } -static inline int perf_allow_kernel(struct perf_event_attr *attr) -{ - if (sysctl_perf_event_paranoid > 1 && !perfmon_capable()) - return -EACCES; - - return security_perf_event_open(attr, PERF_SECURITY_KERNEL); -} +int perf_allow_kernel(struct perf_event_attr *attr); static inline int perf_allow_cpu(struct perf_event_attr *attr) { diff --git a/kernel/events/core.c b/kernel/events/core.c index aa3450bdc227..ae7d63c0c593 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -13351,6 +13351,15 @@ const struct perf_event_attr *perf_event_attrs(struct perf_event *event) return &event->attr; } +int perf_allow_kernel(struct perf_event_attr *attr) +{ + if (sysctl_perf_event_paranoid > 1 && !perfmon_capable()) + return -EACCES; + + return security_perf_event_open(attr, PERF_SECURITY_KERNEL); +} +EXPORT_SYMBOL_GPL(perf_allow_kernel); + /* * Inherit an event from parent task to child task. * -- cgit v1.2.3 From daecd3373a16a039ad241086e30a1ec46fc9d61f Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Thu, 29 Aug 2024 17:03:30 +0800 Subject: drivers/perf: hisi_pcie: Record hardware counts correctly Currently we set the period and record it as the initial value of the counter without checking it's set to the hardware successfully or not. However the counter maybe unwritable if the target event is unsupported by the device. In such case we will pass user a wrong count: [start counts when setting the period] hwc->prev_count = 0x8000000000000000 device.counter_value = 0 // the counter is not set as the period [when user reads the counter] event->count = device.counter_value - hwc->prev_count = 0x8000000000000000 // wrong. should be 0. Fix this by record the hardware counter counts correctly when setting the period. Fixes: 8404b0fbc7fb ("drivers/perf: hisi: Add driver for HiSilicon PCIe PMU") Signed-off-by: Yicong Yang Acked-by: Jonathan Cameron Link: https://lore.kernel.org/r/20240829090332.28756-2-yangyicong@huawei.com Signed-off-by: Will Deacon --- drivers/perf/hisilicon/hisi_pcie_pmu.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'drivers') diff --git a/drivers/perf/hisilicon/hisi_pcie_pmu.c b/drivers/perf/hisilicon/hisi_pcie_pmu.c index f06027574a24..fba569a8640c 100644 --- a/drivers/perf/hisilicon/hisi_pcie_pmu.c +++ b/drivers/perf/hisilicon/hisi_pcie_pmu.c @@ -452,10 +452,24 @@ static void hisi_pcie_pmu_set_period(struct perf_event *event) struct hisi_pcie_pmu *pcie_pmu = to_pcie_pmu(event->pmu); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; + u64 orig_cnt, cnt; + + orig_cnt = hisi_pcie_pmu_read_counter(event); local64_set(&hwc->prev_count, HISI_PCIE_INIT_VAL); hisi_pcie_pmu_writeq(pcie_pmu, HISI_PCIE_CNT, idx, HISI_PCIE_INIT_VAL); hisi_pcie_pmu_writeq(pcie_pmu, HISI_PCIE_EXT_CNT, idx, HISI_PCIE_INIT_VAL); + + /* + * The counter maybe unwritable if the target event is unsupported. + * Check this by comparing the counts after setting the period. If + * the counts stay unchanged after setting the period then update + * the hwc->prev_count correctly. Otherwise the final counts user + * get maybe totally wrong. + */ + cnt = hisi_pcie_pmu_read_counter(event); + if (orig_cnt == cnt) + local64_set(&hwc->prev_count, cnt); } static void hisi_pcie_pmu_enable_counter(struct hisi_pcie_pmu *pcie_pmu, struct hw_perf_event *hwc) -- cgit v1.2.3 From 17bf68aeb3642221e3e770399b5a52f370747ac1 Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Thu, 29 Aug 2024 17:03:31 +0800 Subject: drivers/perf: hisi_pcie: Fix TLP headers bandwidth counting We make the initial value of event ctrl register as HISI_PCIE_INIT_SET and modify according to the user options. This will make TLP headers bandwidth only counting never take effect since HISI_PCIE_INIT_SET configures to count the TLP payloads bandwidth. Fix this by making the initial value of event ctrl register as 0. Fixes: 17d573984d4d ("drivers/perf: hisi: Add TLP filter support") Signed-off-by: Yicong Yang Acked-by: Jonathan Cameron Link: https://lore.kernel.org/r/20240829090332.28756-3-yangyicong@huawei.com Signed-off-by: Will Deacon --- drivers/perf/hisilicon/hisi_pcie_pmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/hisilicon/hisi_pcie_pmu.c b/drivers/perf/hisilicon/hisi_pcie_pmu.c index fba569a8640c..f7d6c59d9930 100644 --- a/drivers/perf/hisilicon/hisi_pcie_pmu.c +++ b/drivers/perf/hisilicon/hisi_pcie_pmu.c @@ -208,7 +208,7 @@ static void hisi_pcie_pmu_writeq(struct hisi_pcie_pmu *pcie_pmu, u32 reg_offset, static u64 hisi_pcie_pmu_get_event_ctrl_val(struct perf_event *event) { u64 port, trig_len, thr_len, len_mode; - u64 reg = HISI_PCIE_INIT_SET; + u64 reg = 0; /* Config HISI_PCIE_EVENT_CTRL according to event. */ reg |= FIELD_PREP(HISI_PCIE_EVENT_M, hisi_pcie_get_real_event(event)); -- cgit v1.2.3 From d1c93d5c67ebc7cf5b70ecff7172a0c399975d55 Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Thu, 29 Aug 2024 17:03:32 +0800 Subject: drivers/perf: hisi_pcie: Export supported Root Ports [bdf_min, bdf_max] Currently users can get the Root Ports supported by the PCIe PMU by "bus" sysfs attributes which indicates the PCIe bus number where Root Ports are located. This maybe insufficient since Root Ports supported by different PCIe PMUs may be located on the same PCIe bus. So export the BDF range the Root Ports additionally. Signed-off-by: Yicong Yang Acked-by: Jonathan Cameron Link: https://lore.kernel.org/r/20240829090332.28756-4-yangyicong@huawei.com Signed-off-by: Will Deacon --- Documentation/admin-guide/perf/hisi-pcie-pmu.rst | 4 +++- drivers/perf/hisilicon/hisi_pcie_pmu.c | 18 ++++++++++++++++++ 2 files changed, 21 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/Documentation/admin-guide/perf/hisi-pcie-pmu.rst b/Documentation/admin-guide/perf/hisi-pcie-pmu.rst index 5541ff40e06a..083ca50de896 100644 --- a/Documentation/admin-guide/perf/hisi-pcie-pmu.rst +++ b/Documentation/admin-guide/perf/hisi-pcie-pmu.rst @@ -28,7 +28,9 @@ The "identifier" sysfs file allows users to identify the version of the PMU hardware device. The "bus" sysfs file allows users to get the bus number of Root Ports -monitored by PMU. +monitored by PMU. Furthermore users can get the Root Ports range in +[bdf_min, bdf_max] from "bdf_min" and "bdf_max" sysfs attributes +respectively. Example usage of perf:: diff --git a/drivers/perf/hisilicon/hisi_pcie_pmu.c b/drivers/perf/hisilicon/hisi_pcie_pmu.c index f7d6c59d9930..c5394d007b61 100644 --- a/drivers/perf/hisilicon/hisi_pcie_pmu.c +++ b/drivers/perf/hisilicon/hisi_pcie_pmu.c @@ -141,6 +141,22 @@ static ssize_t bus_show(struct device *dev, struct device_attribute *attr, char } static DEVICE_ATTR_RO(bus); +static ssize_t bdf_min_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hisi_pcie_pmu *pcie_pmu = to_pcie_pmu(dev_get_drvdata(dev)); + + return sysfs_emit(buf, "%#04x\n", pcie_pmu->bdf_min); +} +static DEVICE_ATTR_RO(bdf_min); + +static ssize_t bdf_max_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hisi_pcie_pmu *pcie_pmu = to_pcie_pmu(dev_get_drvdata(dev)); + + return sysfs_emit(buf, "%#04x\n", pcie_pmu->bdf_max); +} +static DEVICE_ATTR_RO(bdf_max); + static struct hisi_pcie_reg_pair hisi_pcie_parse_reg_value(struct hisi_pcie_pmu *pcie_pmu, u32 reg_off) { @@ -763,6 +779,8 @@ static const struct attribute_group hisi_pcie_pmu_format_group = { static struct attribute *hisi_pcie_pmu_bus_attrs[] = { &dev_attr_bus.attr, + &dev_attr_bdf_max.attr, + &dev_attr_bdf_min.attr, NULL }; -- cgit v1.2.3 From 0ba5b4ba61781f1eca843d9e5d499da329a8a275 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 30 Aug 2024 14:01:44 +0100 Subject: firmware/smccc: Call arch-specific hook on discovering KVM services arm64 will soon require its own callback to initialise services that are only available on this architecture. Introduce a hook that can be overloaded by the architecture. Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20240830130150.8568-2-will@kernel.org Signed-off-by: Will Deacon --- arch/arm/include/asm/hypervisor.h | 2 ++ arch/arm64/include/asm/hypervisor.h | 4 ++++ drivers/firmware/smccc/kvm_guest.c | 2 ++ 3 files changed, 8 insertions(+) (limited to 'drivers') diff --git a/arch/arm/include/asm/hypervisor.h b/arch/arm/include/asm/hypervisor.h index bd61502b9715..8a648e506540 100644 --- a/arch/arm/include/asm/hypervisor.h +++ b/arch/arm/include/asm/hypervisor.h @@ -7,4 +7,6 @@ void kvm_init_hyp_services(void); bool kvm_arm_hyp_service_available(u32 func_id); +static inline void kvm_arch_init_hyp_services(void) { }; + #endif diff --git a/arch/arm64/include/asm/hypervisor.h b/arch/arm64/include/asm/hypervisor.h index 0ae427f352c8..8cab2ab535b7 100644 --- a/arch/arm64/include/asm/hypervisor.h +++ b/arch/arm64/include/asm/hypervisor.h @@ -7,4 +7,8 @@ void kvm_init_hyp_services(void); bool kvm_arm_hyp_service_available(u32 func_id); +static inline void kvm_arch_init_hyp_services(void) +{ +}; + #endif diff --git a/drivers/firmware/smccc/kvm_guest.c b/drivers/firmware/smccc/kvm_guest.c index 89a68e7eeaa6..f3319be20b36 100644 --- a/drivers/firmware/smccc/kvm_guest.c +++ b/drivers/firmware/smccc/kvm_guest.c @@ -39,6 +39,8 @@ void __init kvm_init_hyp_services(void) pr_info("hypervisor services detected (0x%08lx 0x%08lx 0x%08lx 0x%08lx)\n", res.a3, res.a2, res.a1, res.a0); + + kvm_arch_init_hyp_services(); } bool kvm_arm_hyp_service_available(u32 func_id) -- cgit v1.2.3 From a06c3fad49a50d5d5eb078f93e70f4d3eca5d5a5 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 30 Aug 2024 14:01:45 +0100 Subject: drivers/virt: pkvm: Add initial support for running as a protected guest Implement a pKVM protected guest driver to probe the presence of pKVM and determine the memory protection granule using the HYP_MEMINFO hypercall. Acked-by: Marc Zyngier Link: https://lore.kernel.org/r/20240830130150.8568-3-will@kernel.org Signed-off-by: Will Deacon --- Documentation/virt/kvm/arm/hypercalls.rst | 22 ++++++++++++++++ arch/arm64/include/asm/hypervisor.h | 7 +++++ drivers/virt/coco/Kconfig | 2 ++ drivers/virt/coco/Makefile | 1 + drivers/virt/coco/pkvm-guest/Kconfig | 10 ++++++++ drivers/virt/coco/pkvm-guest/Makefile | 2 ++ drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c | 37 +++++++++++++++++++++++++++ include/linux/arm-smccc.h | 7 +++++ 8 files changed, 88 insertions(+) create mode 100644 drivers/virt/coco/pkvm-guest/Kconfig create mode 100644 drivers/virt/coco/pkvm-guest/Makefile create mode 100644 drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c (limited to 'drivers') diff --git a/Documentation/virt/kvm/arm/hypercalls.rst b/Documentation/virt/kvm/arm/hypercalls.rst index 17be111f493f..16515eb42149 100644 --- a/Documentation/virt/kvm/arm/hypercalls.rst +++ b/Documentation/virt/kvm/arm/hypercalls.rst @@ -44,3 +44,25 @@ Provides a discovery mechanism for other KVM/arm64 hypercalls. ---------------------------------------- See ptp_kvm.rst + +``ARM_SMCCC_KVM_FUNC_HYP_MEMINFO`` +---------------------------------- + +Query the memory protection parameters for a pKVM protected virtual machine. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; pKVM protected guests only. | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000002 | ++---------------------+----------+----+---------------------------------------------+ +| Arguments: | (uint64) | R1 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``INVALID_PARAMETER (-3)`` on error, else | +| | | | memory protection granule in bytes | ++---------------------+----------+----+---------------------------------------------+ diff --git a/arch/arm64/include/asm/hypervisor.h b/arch/arm64/include/asm/hypervisor.h index 8cab2ab535b7..409e239834d1 100644 --- a/arch/arm64/include/asm/hypervisor.h +++ b/arch/arm64/include/asm/hypervisor.h @@ -7,8 +7,15 @@ void kvm_init_hyp_services(void); bool kvm_arm_hyp_service_available(u32 func_id); +#ifdef CONFIG_ARM_PKVM_GUEST +void pkvm_init_hyp_services(void); +#else +static inline void pkvm_init_hyp_services(void) { }; +#endif + static inline void kvm_arch_init_hyp_services(void) { + pkvm_init_hyp_services(); }; #endif diff --git a/drivers/virt/coco/Kconfig b/drivers/virt/coco/Kconfig index 87d142c1f932..d9ff676bf48d 100644 --- a/drivers/virt/coco/Kconfig +++ b/drivers/virt/coco/Kconfig @@ -9,6 +9,8 @@ config TSM_REPORTS source "drivers/virt/coco/efi_secret/Kconfig" +source "drivers/virt/coco/pkvm-guest/Kconfig" + source "drivers/virt/coco/sev-guest/Kconfig" source "drivers/virt/coco/tdx-guest/Kconfig" diff --git a/drivers/virt/coco/Makefile b/drivers/virt/coco/Makefile index 18c1aba5edb7..b69c30c1c720 100644 --- a/drivers/virt/coco/Makefile +++ b/drivers/virt/coco/Makefile @@ -4,5 +4,6 @@ # obj-$(CONFIG_TSM_REPORTS) += tsm.o obj-$(CONFIG_EFI_SECRET) += efi_secret/ +obj-$(CONFIG_ARM_PKVM_GUEST) += pkvm-guest/ obj-$(CONFIG_SEV_GUEST) += sev-guest/ obj-$(CONFIG_INTEL_TDX_GUEST) += tdx-guest/ diff --git a/drivers/virt/coco/pkvm-guest/Kconfig b/drivers/virt/coco/pkvm-guest/Kconfig new file mode 100644 index 000000000000..d2f344f1f98f --- /dev/null +++ b/drivers/virt/coco/pkvm-guest/Kconfig @@ -0,0 +1,10 @@ +config ARM_PKVM_GUEST + bool "Arm pKVM protected guest driver" + depends on ARM64 + help + Protected guests running under the pKVM hypervisor on arm64 + are isolated from the host and must issue hypercalls to enable + interaction with virtual devices. This driver implements + support for probing and issuing these hypercalls. + + If unsure, say 'N'. diff --git a/drivers/virt/coco/pkvm-guest/Makefile b/drivers/virt/coco/pkvm-guest/Makefile new file mode 100644 index 000000000000..4bee24579423 --- /dev/null +++ b/drivers/virt/coco/pkvm-guest/Makefile @@ -0,0 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only +obj-$(CONFIG_ARM_PKVM_GUEST) += arm-pkvm-guest.o diff --git a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c new file mode 100644 index 000000000000..a5148701d2f1 --- /dev/null +++ b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c @@ -0,0 +1,37 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Support for the hypercall interface exposed to protected guests by + * pKVM. + * + * Author: Will Deacon + * Copyright (C) 2024 Google LLC + */ + +#include +#include +#include + +#include + +static size_t pkvm_granule; + +void pkvm_init_hyp_services(void) +{ + int i; + struct arm_smccc_res res; + const u32 funcs[] = { + ARM_SMCCC_KVM_FUNC_HYP_MEMINFO, + }; + + for (i = 0; i < ARRAY_SIZE(funcs); ++i) { + if (!kvm_arm_hyp_service_available(funcs[i])) + return; + } + + arm_smccc_1_1_invoke(ARM_SMCCC_VENDOR_HYP_KVM_HYP_MEMINFO_FUNC_ID, + 0, 0, 0, &res); + if (res.a0 > PAGE_SIZE) /* Includes error codes */ + return; + + pkvm_granule = res.a0; +} diff --git a/include/linux/arm-smccc.h b/include/linux/arm-smccc.h index 083f85653716..16b6dcc54e02 100644 --- a/include/linux/arm-smccc.h +++ b/include/linux/arm-smccc.h @@ -115,6 +115,7 @@ /* KVM "vendor specific" services */ #define ARM_SMCCC_KVM_FUNC_FEATURES 0 #define ARM_SMCCC_KVM_FUNC_PTP 1 +#define ARM_SMCCC_KVM_FUNC_HYP_MEMINFO 2 #define ARM_SMCCC_KVM_FUNC_FEATURES_2 127 #define ARM_SMCCC_KVM_NUM_FUNCS 128 @@ -137,6 +138,12 @@ ARM_SMCCC_OWNER_VENDOR_HYP, \ ARM_SMCCC_KVM_FUNC_PTP) +#define ARM_SMCCC_VENDOR_HYP_KVM_HYP_MEMINFO_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_HYP_MEMINFO) + /* ptp_kvm counter type ID */ #define KVM_PTP_VIRT_COUNTER 0 #define KVM_PTP_PHYS_COUNTER 1 -- cgit v1.2.3 From ebc59b120c588156feb7ce194a9636584ced18ba Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 30 Aug 2024 14:01:47 +0100 Subject: drivers/virt: pkvm: Hook up mem_encrypt API using pKVM hypercalls If we detect the presence of pKVM's SHARE and UNSHARE hypercalls, then register a backend implementation of the mem_encrypt API so that things like DMA buffers can be shared appropriately with the host. Acked-by: Marc Zyngier Link: https://lore.kernel.org/r/20240830130150.8568-5-will@kernel.org Signed-off-by: Will Deacon --- Documentation/virt/kvm/arm/hypercalls.rst | 50 ++++++++++++++++++++++++ drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c | 55 +++++++++++++++++++++++++++ include/linux/arm-smccc.h | 14 +++++++ 3 files changed, 119 insertions(+) (limited to 'drivers') diff --git a/Documentation/virt/kvm/arm/hypercalls.rst b/Documentation/virt/kvm/arm/hypercalls.rst index 16515eb42149..c42580e71bf8 100644 --- a/Documentation/virt/kvm/arm/hypercalls.rst +++ b/Documentation/virt/kvm/arm/hypercalls.rst @@ -66,3 +66,53 @@ Query the memory protection parameters for a pKVM protected virtual machine. | Return Values: | (int64) | R0 | ``INVALID_PARAMETER (-3)`` on error, else | | | | | memory protection granule in bytes | +---------------------+----------+----+---------------------------------------------+ + +``ARM_SMCCC_KVM_FUNC_MEM_SHARE`` +-------------------------------- + +Share a region of memory with the KVM host, granting it read, write and execute +permissions. The size of the region is equal to the memory protection granule +advertised by ``ARM_SMCCC_KVM_FUNC_HYP_MEMINFO``. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; pKVM protected guests only. | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000003 | ++---------------------+----------+----+---------------------------------------------+ +| Arguments: | (uint64) | R1 | Base IPA of memory region to share | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``SUCCESS (0)`` | +| | | +---------------------------------------------+ +| | | | ``INVALID_PARAMETER (-3)`` | ++---------------------+----------+----+---------------------------------------------+ + +``ARM_SMCCC_KVM_FUNC_MEM_UNSHARE`` +---------------------------------- + +Revoke access permission from the KVM host to a memory region previously shared +with ``ARM_SMCCC_KVM_FUNC_MEM_SHARE``. The size of the region is equal to the +memory protection granule advertised by ``ARM_SMCCC_KVM_FUNC_HYP_MEMINFO``. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; pKVM protected guests only. | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000004 | ++---------------------+----------+----+---------------------------------------------+ +| Arguments: | (uint64) | R1 | Base IPA of memory region to unshare | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``SUCCESS (0)`` | +| | | +---------------------------------------------+ +| | | | ``INVALID_PARAMETER (-3)`` | ++---------------------+----------+----+---------------------------------------------+ diff --git a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c index a5148701d2f1..8256cf68fd76 100644 --- a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c +++ b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c @@ -9,18 +9,72 @@ #include #include +#include #include #include static size_t pkvm_granule; +static int arm_smccc_do_one_page(u32 func_id, phys_addr_t phys) +{ + phys_addr_t end = phys + PAGE_SIZE; + + while (phys < end) { + struct arm_smccc_res res; + + arm_smccc_1_1_invoke(func_id, phys, 0, 0, &res); + if (res.a0 != SMCCC_RET_SUCCESS) + return -EPERM; + + phys += pkvm_granule; + } + + return 0; +} + +static int __set_memory_range(u32 func_id, unsigned long start, int numpages) +{ + void *addr = (void *)start, *end = addr + numpages * PAGE_SIZE; + + while (addr < end) { + int err; + + err = arm_smccc_do_one_page(func_id, virt_to_phys(addr)); + if (err) + return err; + + addr += PAGE_SIZE; + } + + return 0; +} + +static int pkvm_set_memory_encrypted(unsigned long addr, int numpages) +{ + return __set_memory_range(ARM_SMCCC_VENDOR_HYP_KVM_MEM_UNSHARE_FUNC_ID, + addr, numpages); +} + +static int pkvm_set_memory_decrypted(unsigned long addr, int numpages) +{ + return __set_memory_range(ARM_SMCCC_VENDOR_HYP_KVM_MEM_SHARE_FUNC_ID, + addr, numpages); +} + +static const struct arm64_mem_crypt_ops pkvm_crypt_ops = { + .encrypt = pkvm_set_memory_encrypted, + .decrypt = pkvm_set_memory_decrypted, +}; + void pkvm_init_hyp_services(void) { int i; struct arm_smccc_res res; const u32 funcs[] = { ARM_SMCCC_KVM_FUNC_HYP_MEMINFO, + ARM_SMCCC_KVM_FUNC_MEM_SHARE, + ARM_SMCCC_KVM_FUNC_MEM_UNSHARE, }; for (i = 0; i < ARRAY_SIZE(funcs); ++i) { @@ -34,4 +88,5 @@ void pkvm_init_hyp_services(void) return; pkvm_granule = res.a0; + arm64_mem_crypt_ops_register(&pkvm_crypt_ops); } diff --git a/include/linux/arm-smccc.h b/include/linux/arm-smccc.h index 16b6dcc54e02..9cb7c95920b0 100644 --- a/include/linux/arm-smccc.h +++ b/include/linux/arm-smccc.h @@ -116,6 +116,8 @@ #define ARM_SMCCC_KVM_FUNC_FEATURES 0 #define ARM_SMCCC_KVM_FUNC_PTP 1 #define ARM_SMCCC_KVM_FUNC_HYP_MEMINFO 2 +#define ARM_SMCCC_KVM_FUNC_MEM_SHARE 3 +#define ARM_SMCCC_KVM_FUNC_MEM_UNSHARE 4 #define ARM_SMCCC_KVM_FUNC_FEATURES_2 127 #define ARM_SMCCC_KVM_NUM_FUNCS 128 @@ -144,6 +146,18 @@ ARM_SMCCC_OWNER_VENDOR_HYP, \ ARM_SMCCC_KVM_FUNC_HYP_MEMINFO) +#define ARM_SMCCC_VENDOR_HYP_KVM_MEM_SHARE_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_MEM_SHARE) + +#define ARM_SMCCC_VENDOR_HYP_KVM_MEM_UNSHARE_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_MEM_UNSHARE) + /* ptp_kvm counter type ID */ #define KVM_PTP_VIRT_COUNTER 0 #define KVM_PTP_PHYS_COUNTER 1 -- cgit v1.2.3 From 0f12694958001c96bda811473fdb23f333c6d3ca Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 30 Aug 2024 14:01:49 +0100 Subject: drivers/virt: pkvm: Intercept ioremap using pKVM MMIO_GUARD hypercall Hook up pKVM's MMIO_GUARD hypercall so that ioremap() and friends will register the target physical address as MMIO with the hypervisor, allowing guest exits to that page to be emulated by the host with full syndrome information. Acked-by: Marc Zyngier Link: https://lore.kernel.org/r/20240830130150.8568-7-will@kernel.org Signed-off-by: Will Deacon --- Documentation/virt/kvm/arm/hypercalls.rst | 26 ++++++++++++++++++++ drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c | 35 +++++++++++++++++++++++++++ include/linux/arm-smccc.h | 7 ++++++ 3 files changed, 68 insertions(+) (limited to 'drivers') diff --git a/Documentation/virt/kvm/arm/hypercalls.rst b/Documentation/virt/kvm/arm/hypercalls.rst index c42580e71bf8..af7bc2c2e0cb 100644 --- a/Documentation/virt/kvm/arm/hypercalls.rst +++ b/Documentation/virt/kvm/arm/hypercalls.rst @@ -116,3 +116,29 @@ memory protection granule advertised by ``ARM_SMCCC_KVM_FUNC_HYP_MEMINFO``. | | | +---------------------------------------------+ | | | | ``INVALID_PARAMETER (-3)`` | +---------------------+----------+----+---------------------------------------------+ + +``ARM_SMCCC_KVM_FUNC_MMIO_GUARD`` +---------------------------------- + +Request that a given memory region is handled as MMIO by the hypervisor, +allowing accesses to this region to be emulated by the KVM host. The size of the +region is equal to the memory protection granule advertised by +``ARM_SMCCC_KVM_FUNC_HYP_MEMINFO``. + ++---------------------+-------------------------------------------------------------+ +| Presence: | Optional; pKVM protected guests only. | ++---------------------+-------------------------------------------------------------+ +| Calling convention: | HVC64 | ++---------------------+----------+--------------------------------------------------+ +| Function ID: | (uint32) | 0xC6000007 | ++---------------------+----------+----+---------------------------------------------+ +| Arguments: | (uint64) | R1 | Base IPA of MMIO memory region | +| +----------+----+---------------------------------------------+ +| | (uint64) | R2 | Reserved / Must be zero | +| +----------+----+---------------------------------------------+ +| | (uint64) | R3 | Reserved / Must be zero | ++---------------------+----------+----+---------------------------------------------+ +| Return Values: | (int64) | R0 | ``SUCCESS (0)`` | +| | | +---------------------------------------------+ +| | | | ``INVALID_PARAMETER (-3)`` | ++---------------------+----------+----+---------------------------------------------+ diff --git a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c index 8256cf68fd76..56a3859dda8a 100644 --- a/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c +++ b/drivers/virt/coco/pkvm-guest/arm-pkvm-guest.c @@ -9,8 +9,10 @@ #include #include +#include #include #include +#include #include @@ -67,6 +69,36 @@ static const struct arm64_mem_crypt_ops pkvm_crypt_ops = { .decrypt = pkvm_set_memory_decrypted, }; +static int mmio_guard_ioremap_hook(phys_addr_t phys, size_t size, + pgprot_t *prot) +{ + phys_addr_t end; + pteval_t protval = pgprot_val(*prot); + + /* + * We only expect MMIO emulation for regions mapped with device + * attributes. + */ + if (protval != PROT_DEVICE_nGnRE && protval != PROT_DEVICE_nGnRnE) + return 0; + + phys = PAGE_ALIGN_DOWN(phys); + end = phys + PAGE_ALIGN(size); + + while (phys < end) { + const int func_id = ARM_SMCCC_VENDOR_HYP_KVM_MMIO_GUARD_FUNC_ID; + int err; + + err = arm_smccc_do_one_page(func_id, phys); + if (err) + return err; + + phys += PAGE_SIZE; + } + + return 0; +} + void pkvm_init_hyp_services(void) { int i; @@ -89,4 +121,7 @@ void pkvm_init_hyp_services(void) pkvm_granule = res.a0; arm64_mem_crypt_ops_register(&pkvm_crypt_ops); + + if (kvm_arm_hyp_service_available(ARM_SMCCC_KVM_FUNC_MMIO_GUARD)) + arm64_ioremap_prot_hook_register(&mmio_guard_ioremap_hook); } diff --git a/include/linux/arm-smccc.h b/include/linux/arm-smccc.h index 9cb7c95920b0..e93c1f7cea70 100644 --- a/include/linux/arm-smccc.h +++ b/include/linux/arm-smccc.h @@ -118,6 +118,7 @@ #define ARM_SMCCC_KVM_FUNC_HYP_MEMINFO 2 #define ARM_SMCCC_KVM_FUNC_MEM_SHARE 3 #define ARM_SMCCC_KVM_FUNC_MEM_UNSHARE 4 +#define ARM_SMCCC_KVM_FUNC_MMIO_GUARD 7 #define ARM_SMCCC_KVM_FUNC_FEATURES_2 127 #define ARM_SMCCC_KVM_NUM_FUNCS 128 @@ -158,6 +159,12 @@ ARM_SMCCC_OWNER_VENDOR_HYP, \ ARM_SMCCC_KVM_FUNC_MEM_UNSHARE) +#define ARM_SMCCC_VENDOR_HYP_KVM_MMIO_GUARD_FUNC_ID \ + ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, \ + ARM_SMCCC_SMC_64, \ + ARM_SMCCC_OWNER_VENDOR_HYP, \ + ARM_SMCCC_KVM_FUNC_MMIO_GUARD) + /* ptp_kvm counter type ID */ #define KVM_PTP_VIRT_COUNTER 0 #define KVM_PTP_PHYS_COUNTER 1 -- cgit v1.2.3 From e79634b53e398966c49f803c49701bc74dc3ccf8 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:51:57 +0100 Subject: perf/arm-cmn: Refactor node ID handling. Again. The scope of the "extra device ports" configuration is not made clear by the CMN documentation - so far we've assumed it applies globally, based on the sole example which suggests as much. However it transpires that this is incorrect, and the format does in fact vary based on each individual XP's port configuration. As a consequence, we're currenly liable to decode the port/device indices from a node ID incorrectly, thus program the wrong event source in the DTM leading to bogus event counts, and also show device topology on the wrong ports in debugfs. To put this right, rework node IDs yet again to carry around the additional data necessary to decode them properly per-XP. At this point the notion of fully decomposing an ID becomes more impractical than it's worth, so unabstracting the XY mesh coordinates (where 2/3 users were just debug anyway) ends up leaving things a bit simpler overall. Fixes: 60d1504070c2 ("perf/arm-cmn: Support new IP features") Acked-by: Mark Rutland Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/5195f990152fc37adba5fbf5929a6b11063d9f09.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 94 +++++++++++++++++++++----------------------------- 1 file changed, 40 insertions(+), 54 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index c932d9d355cf..b59ae8513dce 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -24,14 +24,6 @@ #define CMN_NI_NODE_ID GENMASK_ULL(31, 16) #define CMN_NI_LOGICAL_ID GENMASK_ULL(47, 32) -#define CMN_NODEID_DEVID(reg) ((reg) & 3) -#define CMN_NODEID_EXT_DEVID(reg) ((reg) & 1) -#define CMN_NODEID_PID(reg) (((reg) >> 2) & 1) -#define CMN_NODEID_EXT_PID(reg) (((reg) >> 1) & 3) -#define CMN_NODEID_1x1_PID(reg) (((reg) >> 2) & 7) -#define CMN_NODEID_X(reg, bits) ((reg) >> (3 + (bits))) -#define CMN_NODEID_Y(reg, bits) (((reg) >> 3) & ((1U << (bits)) - 1)) - #define CMN_CHILD_INFO 0x0080 #define CMN_CI_CHILD_COUNT GENMASK_ULL(15, 0) #define CMN_CI_CHILD_PTR_OFFSET GENMASK_ULL(31, 16) @@ -280,8 +272,11 @@ struct arm_cmn_node { u16 id, logid; enum cmn_node_type type; + /* XP properties really, but replicated to children for convenience */ u8 dtm; s8 dtc; + u8 portid_bits:4; + u8 deviceid_bits:4; /* DN/HN-F/CXHA */ struct { u8 val : 4; @@ -357,49 +352,33 @@ struct arm_cmn { static int arm_cmn_hp_state; struct arm_cmn_nodeid { - u8 x; - u8 y; u8 port; u8 dev; }; static int arm_cmn_xyidbits(const struct arm_cmn *cmn) { - return fls((cmn->mesh_x - 1) | (cmn->mesh_y - 1) | 2); + return fls((cmn->mesh_x - 1) | (cmn->mesh_y - 1)); } -static struct arm_cmn_nodeid arm_cmn_nid(const struct arm_cmn *cmn, u16 id) +static struct arm_cmn_nodeid arm_cmn_nid(const struct arm_cmn_node *dn) { struct arm_cmn_nodeid nid; - if (cmn->num_xps == 1) { - nid.x = 0; - nid.y = 0; - nid.port = CMN_NODEID_1x1_PID(id); - nid.dev = CMN_NODEID_DEVID(id); - } else { - int bits = arm_cmn_xyidbits(cmn); - - nid.x = CMN_NODEID_X(id, bits); - nid.y = CMN_NODEID_Y(id, bits); - if (cmn->ports_used & 0xc) { - nid.port = CMN_NODEID_EXT_PID(id); - nid.dev = CMN_NODEID_EXT_DEVID(id); - } else { - nid.port = CMN_NODEID_PID(id); - nid.dev = CMN_NODEID_DEVID(id); - } - } + nid.dev = dn->id & ((1U << dn->deviceid_bits) - 1); + nid.port = (dn->id >> dn->deviceid_bits) & ((1U << dn->portid_bits) - 1); return nid; } static struct arm_cmn_node *arm_cmn_node_to_xp(const struct arm_cmn *cmn, const struct arm_cmn_node *dn) { - struct arm_cmn_nodeid nid = arm_cmn_nid(cmn, dn->id); - int xp_idx = cmn->mesh_x * nid.y + nid.x; + int id = dn->id >> (dn->portid_bits + dn->deviceid_bits); + int bits = arm_cmn_xyidbits(cmn); + int x = id >> bits; + int y = id & ((1U << bits) - 1); - return cmn->xps + xp_idx; + return cmn->xps + cmn->mesh_x * y + x; } static struct arm_cmn_node *arm_cmn_node(const struct arm_cmn *cmn, enum cmn_node_type type) @@ -485,13 +464,13 @@ static const char *arm_cmn_device_type(u8 type) } } -static void arm_cmn_show_logid(struct seq_file *s, int x, int y, int p, int d) +static void arm_cmn_show_logid(struct seq_file *s, const struct arm_cmn_node *xp, int p, int d) { struct arm_cmn *cmn = s->private; struct arm_cmn_node *dn; + u16 id = xp->id | d | (p << xp->deviceid_bits); for (dn = cmn->dns; dn->type; dn++) { - struct arm_cmn_nodeid nid = arm_cmn_nid(cmn, dn->id); int pad = dn->logid < 10; if (dn->type == CMN_TYPE_XP) @@ -500,7 +479,7 @@ static void arm_cmn_show_logid(struct seq_file *s, int x, int y, int p, int d) if (dn->type < CMN_TYPE_HNI) continue; - if (nid.x != x || nid.y != y || nid.port != p || nid.dev != d) + if (dn->id != id) continue; seq_printf(s, " %*c#%-*d |", pad + 1, ' ', 3 - pad, dn->logid); @@ -521,6 +500,7 @@ static int arm_cmn_map_show(struct seq_file *s, void *data) y = cmn->mesh_y; while (y--) { int xp_base = cmn->mesh_x * y; + struct arm_cmn_node *xp = cmn->xps + xp_base; u8 port[CMN_MAX_PORTS][CMN_MAX_DIMENSION]; for (x = 0; x < cmn->mesh_x; x++) @@ -528,16 +508,14 @@ static int arm_cmn_map_show(struct seq_file *s, void *data) seq_printf(s, "\n%-2d |", y); for (x = 0; x < cmn->mesh_x; x++) { - struct arm_cmn_node *xp = cmn->xps + xp_base + x; - for (p = 0; p < CMN_MAX_PORTS; p++) - port[p][x] = arm_cmn_device_connect_info(cmn, xp, p); + port[p][x] = arm_cmn_device_connect_info(cmn, xp + x, p); seq_printf(s, " XP #%-3d|", xp_base + x); } seq_puts(s, "\n |"); for (x = 0; x < cmn->mesh_x; x++) { - s8 dtc = cmn->xps[xp_base + x].dtc; + s8 dtc = xp[x].dtc; if (dtc < 0) seq_puts(s, " DTC ?? |"); @@ -554,10 +532,10 @@ static int arm_cmn_map_show(struct seq_file *s, void *data) seq_puts(s, arm_cmn_device_type(port[p][x])); seq_puts(s, "\n 0|"); for (x = 0; x < cmn->mesh_x; x++) - arm_cmn_show_logid(s, x, y, p, 0); + arm_cmn_show_logid(s, xp + x, p, 0); seq_puts(s, "\n 1|"); for (x = 0; x < cmn->mesh_x; x++) - arm_cmn_show_logid(s, x, y, p, 1); + arm_cmn_show_logid(s, xp + x, p, 1); } seq_puts(s, "\n-----+"); } @@ -1815,10 +1793,7 @@ static int arm_cmn_event_init(struct perf_event *event) } if (!hw->num_dns) { - struct arm_cmn_nodeid nid = arm_cmn_nid(cmn, nodeid); - - dev_dbg(cmn->dev, "invalid node 0x%x (%d,%d,%d,%d) type 0x%x\n", - nodeid, nid.x, nid.y, nid.port, nid.dev, type); + dev_dbg(cmn->dev, "invalid node 0x%x type 0x%x\n", nodeid, type); return -EINVAL; } @@ -1921,7 +1896,7 @@ static int arm_cmn_event_add(struct perf_event *event, int flags) arm_cmn_claim_wp_idx(dtm, event, d, wp_idx, i); writel_relaxed(cfg, dtm->base + CMN_DTM_WPn_CONFIG(wp_idx)); } else { - struct arm_cmn_nodeid nid = arm_cmn_nid(cmn, dn->id); + struct arm_cmn_nodeid nid = arm_cmn_nid(dn); if (cmn->multi_dtm) nid.port %= 2; @@ -2168,10 +2143,12 @@ static int arm_cmn_init_dtcs(struct arm_cmn *cmn) continue; xp = arm_cmn_node_to_xp(cmn, dn); + dn->portid_bits = xp->portid_bits; + dn->deviceid_bits = xp->deviceid_bits; dn->dtc = xp->dtc; dn->dtm = xp->dtm; if (cmn->multi_dtm) - dn->dtm += arm_cmn_nid(cmn, dn->id).port / 2; + dn->dtm += arm_cmn_nid(dn).port / 2; if (dn->type == CMN_TYPE_DTC) { int err = arm_cmn_init_dtc(cmn, dn, dtc_idx++); @@ -2341,18 +2318,27 @@ static int arm_cmn_discover(struct arm_cmn *cmn, unsigned int rgn_offset) arm_cmn_init_dtm(dtm++, xp, 0); /* * Keeping track of connected ports will let us filter out - * unnecessary XP events easily. We can also reliably infer the - * "extra device ports" configuration for the node ID format - * from this, since in that case we will see at least one XP - * with port 2 connected, for the HN-D. + * unnecessary XP events easily, and also infer the per-XP + * part of the node ID format. */ for (int p = 0; p < CMN_MAX_PORTS; p++) if (arm_cmn_device_connect_info(cmn, xp, p)) xp_ports |= BIT(p); - if (cmn->multi_dtm && (xp_ports & 0xc)) + if (cmn->num_xps == 1) { + xp->portid_bits = 3; + xp->deviceid_bits = 2; + } else if (xp_ports > 0x3) { + xp->portid_bits = 2; + xp->deviceid_bits = 1; + } else { + xp->portid_bits = 1; + xp->deviceid_bits = 2; + } + + if (cmn->multi_dtm && (xp_ports > 0x3)) arm_cmn_init_dtm(dtm++, xp, 1); - if (cmn->multi_dtm && (xp_ports & 0x30)) + if (cmn->multi_dtm && (xp_ports > 0xf)) arm_cmn_init_dtm(dtm++, xp, 2); cmn->ports_used |= xp_ports; -- cgit v1.2.3 From 88b63a82c84ed9bbcbdefb10cb6f75dd1dd04887 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:51:58 +0100 Subject: perf/arm-cmn: Fix CCLA register offset Apparently pmu_event_sel is offset by 8 for all CCLA nodes, not just the CCLA_RNI combination type. Fixes: 23760a014417 ("perf/arm-cmn: Add CMN-700 support") Acked-by: Mark Rutland Reviewed-by: Ilkka Koskinen Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/6e7bb06fef6046f83e7647aad0e5be544139763f.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index b59ae8513dce..4e2338afe669 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -70,7 +70,8 @@ /* Technically this is 4 bits wide on DNs, but we only use 2 there anyway */ #define CMN__PMU_OCCUP1_ID GENMASK_ULL(34, 32) -/* HN-Ps are weird... */ +/* Some types are designed to coexist with another device in the same node */ +#define CMN_CCLA_PMU_EVENT_SEL 0x008 #define CMN_HNP_PMU_EVENT_SEL 0x008 /* DTMs live in the PMU space of XP registers */ @@ -2393,10 +2394,13 @@ static int arm_cmn_discover(struct arm_cmn *cmn, unsigned int rgn_offset) case CMN_TYPE_CXHA: case CMN_TYPE_CCRA: case CMN_TYPE_CCHA: - case CMN_TYPE_CCLA: case CMN_TYPE_HNS: dn++; break; + case CMN_TYPE_CCLA: + dn->pmu_base += CMN_CCLA_PMU_EVENT_SEL; + dn++; + break; /* Nothing to see here */ case CMN_TYPE_MPAM_S: case CMN_TYPE_MPAM_NS: @@ -2414,7 +2418,7 @@ static int arm_cmn_discover(struct arm_cmn *cmn, unsigned int rgn_offset) case CMN_TYPE_HNP: case CMN_TYPE_CCLA_RNI: dn[1] = dn[0]; - dn[0].pmu_base += CMN_HNP_PMU_EVENT_SEL; + dn[0].pmu_base += CMN_CCLA_PMU_EVENT_SEL; dn[1].type = arm_cmn_subtype(dn->type); dn += 2; break; -- cgit v1.2.3 From 359414b33e00bae91e4eabf3e4ef8e76024c7673 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:51:59 +0100 Subject: perf/arm-cmn: Ensure dtm_idx is big enough While CMN_MAX_DIMENSION was bumped to 12 for CMN-650, that only supports up to a 10x10 mesh, so bumping dtm_idx to 256 bits at the time worked out OK in practice. However CMN-700 did finally support up to 144 XPs, and thus needs a worst-case 288 bits of dtm_idx for an aggregated XP event on a maxed-out config. Oops. Fixes: 23760a014417 ("perf/arm-cmn: Add CMN-700 support") Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/e771b358526a0d7fc06efee2c3a2fdc0c9f51d44.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 4e2338afe669..48863b31ccfb 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -35,6 +35,9 @@ #define CMN_MAX_XPS (CMN_MAX_DIMENSION * CMN_MAX_DIMENSION) #define CMN_MAX_DTMS (CMN_MAX_XPS + (CMN_MAX_DIMENSION - 1) * 4) +/* Currently XPs are the node type we can have most of; others top out at 128 */ +#define CMN_MAX_NODES_PER_EVENT CMN_MAX_XPS + /* The CFG node has various info besides the discovery tree */ #define CMN_CFGM_PERIPH_ID_01 0x0008 #define CMN_CFGM_PID0_PART_0 GENMASK_ULL(7, 0) @@ -564,7 +567,7 @@ static void arm_cmn_debugfs_init(struct arm_cmn *cmn, int id) {} struct arm_cmn_hw_event { struct arm_cmn_node *dn; - u64 dtm_idx[4]; + u64 dtm_idx[DIV_ROUND_UP(CMN_MAX_NODES_PER_EVENT * 2, 64)]; s8 dtc_idx[CMN_MAX_DTCS]; u8 num_dns; u8 dtm_offset; -- cgit v1.2.3 From ff436cee694ee8bc4173f2d42622ee7c17a085d3 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:52:00 +0100 Subject: perf/arm-cmn: Improve build-time assertion These days we can use static_assert() in the logical place rather than jamming a BUILD_BUG_ON() into the nearest function scope. Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/224ee8286f299100f1c768edb254edc898539f50.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 48863b31ccfb..5b0edeb69394 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -581,6 +581,7 @@ struct arm_cmn_hw_event { bool wide_sel; enum cmn_filter_select filter_sel; }; +static_assert(sizeof(struct arm_cmn_hw_event) <= offsetof(struct hw_perf_event, target)); #define for_each_hw_dn(hw, dn, i) \ for (i = 0, dn = hw->dn; i < hw->num_dns; i++, dn++) @@ -591,7 +592,6 @@ struct arm_cmn_hw_event { static struct arm_cmn_hw_event *to_cmn_hw(struct perf_event *event) { - BUILD_BUG_ON(sizeof(struct arm_cmn_hw_event) > offsetof(struct hw_perf_event, target)); return (struct arm_cmn_hw_event *)&event->hw; } -- cgit v1.2.3 From c5b15ddf11a8a82f5e9ccd9b44f7b765c59bffdd Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:52:01 +0100 Subject: perf/arm-cmn: Make cycle counts less surprising By default, CMN has automatic clock-gating with the implication that a DTC's cycle counter may not increment while the DTC is sufficiently idle. Given that we may have up to 4 DTCs to choose from when scheduling a cycles event, this may potentially lead to surprising results if trying to measure metrics based on activity in a different DTC domain from where cycles end up being counted. Furthermore, since the details of internal clock gating are not documented, we can't even reason about what "active" cycles for a DTC actually mean relative to the activity of other nodes within the same nominal DTC domain. Make the reasonable assumption that if the user wants to count cycles, they almost certainly want to count all of the cycles, and disable clock gating while a DTC's cycle counter is in use. Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/c47cfdc09e907b1d7753d142a7e659982cceb246.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 5b0edeb69394..2205c183ec1b 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -119,6 +119,7 @@ /* The DTC node is where the magic happens */ #define CMN_DT_DTC_CTL 0x0a00 #define CMN_DT_DTC_CTL_DT_EN BIT(0) +#define CMN_DT_DTC_CTL_CG_DISABLE BIT(10) /* DTC counters are paired in 64-bit registers on a 16-byte stride. Yuck */ #define _CMN_DT_CNT_REG(n) ((((n) / 2) * 4 + (n) % 2) * 4) @@ -1546,9 +1547,12 @@ static void arm_cmn_event_start(struct perf_event *event, int flags) int i; if (type == CMN_TYPE_DTC) { - i = hw->dtc_idx[0]; - writeq_relaxed(CMN_CC_INIT, cmn->dtc[i].base + CMN_DT_PMCCNTR); - cmn->dtc[i].cc_active = true; + struct arm_cmn_dtc *dtc = cmn->dtc + hw->dtc_idx[0]; + + writel_relaxed(CMN_DT_DTC_CTL_DT_EN | CMN_DT_DTC_CTL_CG_DISABLE, + dtc->base + CMN_DT_DTC_CTL); + writeq_relaxed(CMN_CC_INIT, dtc->base + CMN_DT_PMCCNTR); + dtc->cc_active = true; } else if (type == CMN_TYPE_WP) { u64 val = CMN_EVENT_WP_VAL(event); u64 mask = CMN_EVENT_WP_MASK(event); @@ -1577,8 +1581,10 @@ static void arm_cmn_event_stop(struct perf_event *event, int flags) int i; if (type == CMN_TYPE_DTC) { - i = hw->dtc_idx[0]; - cmn->dtc[i].cc_active = false; + struct arm_cmn_dtc *dtc = cmn->dtc + hw->dtc_idx[0]; + + dtc->cc_active = false; + writel_relaxed(CMN_DT_DTC_CTL_DT_EN, dtc->base + CMN_DT_DTC_CTL); } else if (type == CMN_TYPE_WP) { for_each_hw_dn(hw, dn, i) { void __iomem *base = dn->pmu_base + CMN_DTM_OFFSET(hw->dtm_offset); -- cgit v1.2.3 From 67acca3504eade75e03ea3ae87a329df6b10ae02 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:52:02 +0100 Subject: perf/arm-cmn: Refactor DTC PMU register access Annoyingly, we're soon going to have to cope with PMU registers moving about. This will mostly be straightforward, except for the hard-coding of CMN_PMU_OFFSET for the DTC PMU registers. As a first step, refactor those accessors to allow for encapsulating a variable offset without making a big mess all over. As a bonus, we can repack the arm_cmn_dtc structure to accommodate the new pointer without growing any larger, since irq_friend only encodes a range of +/-3. Acked-by: Mark Rutland Reviewed-by: Ilkka Koskinen Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/fc677576fae7b5b55780e5b245a4ef6ea1b30daf.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 64 ++++++++++++++++++++++++++++---------------------- 1 file changed, 36 insertions(+), 28 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 2205c183ec1b..ac7dd4c352e8 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -123,24 +123,24 @@ /* DTC counters are paired in 64-bit registers on a 16-byte stride. Yuck */ #define _CMN_DT_CNT_REG(n) ((((n) / 2) * 4 + (n) % 2) * 4) -#define CMN_DT_PMEVCNT(n) (CMN_PMU_OFFSET + _CMN_DT_CNT_REG(n)) -#define CMN_DT_PMCCNTR (CMN_PMU_OFFSET + 0x40) +#define CMN_DT_PMEVCNT(dtc, n) ((dtc)->pmu_base + _CMN_DT_CNT_REG(n)) +#define CMN_DT_PMCCNTR(dtc) ((dtc)->pmu_base + 0x40) -#define CMN_DT_PMEVCNTSR(n) (CMN_PMU_OFFSET + 0x50 + _CMN_DT_CNT_REG(n)) -#define CMN_DT_PMCCNTRSR (CMN_PMU_OFFSET + 0x90) +#define CMN_DT_PMEVCNTSR(dtc, n) ((dtc)->pmu_base + 0x50 + _CMN_DT_CNT_REG(n)) +#define CMN_DT_PMCCNTRSR(dtc) ((dtc)->pmu_base + 0x90) -#define CMN_DT_PMCR (CMN_PMU_OFFSET + 0x100) +#define CMN_DT_PMCR(dtc) ((dtc)->pmu_base + 0x100) #define CMN_DT_PMCR_PMU_EN BIT(0) #define CMN_DT_PMCR_CNTR_RST BIT(5) #define CMN_DT_PMCR_OVFL_INTR_EN BIT(6) -#define CMN_DT_PMOVSR (CMN_PMU_OFFSET + 0x118) -#define CMN_DT_PMOVSR_CLR (CMN_PMU_OFFSET + 0x120) +#define CMN_DT_PMOVSR(dtc) ((dtc)->pmu_base + 0x118) +#define CMN_DT_PMOVSR_CLR(dtc) ((dtc)->pmu_base + 0x120) -#define CMN_DT_PMSSR (CMN_PMU_OFFSET + 0x128) +#define CMN_DT_PMSSR(dtc) ((dtc)->pmu_base + 0x128) #define CMN_DT_PMSSR_SS_STATUS(n) BIT(n) -#define CMN_DT_PMSRR (CMN_PMU_OFFSET + 0x130) +#define CMN_DT_PMSRR(dtc) ((dtc)->pmu_base + 0x130) #define CMN_DT_PMSRR_SS_REQ BIT(0) #define CMN_DT_NUM_COUNTERS 8 @@ -307,8 +307,9 @@ struct arm_cmn_dtm { struct arm_cmn_dtc { void __iomem *base; + void __iomem *pmu_base; int irq; - int irq_friend; + s8 irq_friend; bool cc_active; struct perf_event *counters[CMN_DT_NUM_COUNTERS]; @@ -412,10 +413,15 @@ static enum cmn_model arm_cmn_model(const struct arm_cmn *cmn) }; } +static int arm_cmn_pmu_offset(const struct arm_cmn *cmn, const struct arm_cmn_node *dn) +{ + return CMN_PMU_OFFSET; +} + static u32 arm_cmn_device_connect_info(const struct arm_cmn *cmn, const struct arm_cmn_node *xp, int port) { - int offset = CMN_MXP__CONNECT_INFO(port); + int offset = CMN_MXP__CONNECT_INFO(port) - arm_cmn_pmu_offset(cmn, xp); if (port >= 2) { if (cmn->part == PART_CMN600 || cmn->part == PART_CMN650) @@ -428,7 +434,7 @@ static u32 arm_cmn_device_connect_info(const struct arm_cmn *cmn, offset += CI700_CONNECT_INFO_P2_5_OFFSET; } - return readl_relaxed(xp->pmu_base - CMN_PMU_OFFSET + offset); + return readl_relaxed(xp->pmu_base + offset); } static struct dentry *arm_cmn_debugfs; @@ -1398,7 +1404,7 @@ static u32 arm_cmn_wp_config(struct perf_event *event, int wp_idx) static void arm_cmn_set_state(struct arm_cmn *cmn, u32 state) { if (!cmn->state) - writel_relaxed(0, cmn->dtc[0].base + CMN_DT_PMCR); + writel_relaxed(0, CMN_DT_PMCR(&cmn->dtc[0])); cmn->state |= state; } @@ -1407,7 +1413,7 @@ static void arm_cmn_clear_state(struct arm_cmn *cmn, u32 state) cmn->state &= ~state; if (!cmn->state) writel_relaxed(CMN_DT_PMCR_PMU_EN | CMN_DT_PMCR_OVFL_INTR_EN, - cmn->dtc[0].base + CMN_DT_PMCR); + CMN_DT_PMCR(&cmn->dtc[0])); } static void arm_cmn_pmu_enable(struct pmu *pmu) @@ -1442,18 +1448,19 @@ static u64 arm_cmn_read_dtm(struct arm_cmn *cmn, struct arm_cmn_hw_event *hw, static u64 arm_cmn_read_cc(struct arm_cmn_dtc *dtc) { - u64 val = readq_relaxed(dtc->base + CMN_DT_PMCCNTR); + void __iomem *pmccntr = CMN_DT_PMCCNTR(dtc); + u64 val = readq_relaxed(pmccntr); - writeq_relaxed(CMN_CC_INIT, dtc->base + CMN_DT_PMCCNTR); + writeq_relaxed(CMN_CC_INIT, pmccntr); return (val - CMN_CC_INIT) & ((CMN_CC_INIT << 1) - 1); } static u32 arm_cmn_read_counter(struct arm_cmn_dtc *dtc, int idx) { - u32 val, pmevcnt = CMN_DT_PMEVCNT(idx); + void __iomem *pmevcnt = CMN_DT_PMEVCNT(dtc, idx); + u32 val = readl_relaxed(pmevcnt); - val = readl_relaxed(dtc->base + pmevcnt); - writel_relaxed(CMN_COUNTER_INIT, dtc->base + pmevcnt); + writel_relaxed(CMN_COUNTER_INIT, pmevcnt); return val - CMN_COUNTER_INIT; } @@ -1464,7 +1471,7 @@ static void arm_cmn_init_counter(struct perf_event *event) u64 count; for_each_hw_dtc_idx(hw, i, idx) { - writel_relaxed(CMN_COUNTER_INIT, cmn->dtc[i].base + CMN_DT_PMEVCNT(idx)); + writel_relaxed(CMN_COUNTER_INIT, CMN_DT_PMEVCNT(&cmn->dtc[i], idx)); cmn->dtc[i].counters[idx] = event; } @@ -1551,7 +1558,7 @@ static void arm_cmn_event_start(struct perf_event *event, int flags) writel_relaxed(CMN_DT_DTC_CTL_DT_EN | CMN_DT_DTC_CTL_CG_DISABLE, dtc->base + CMN_DT_DTC_CTL); - writeq_relaxed(CMN_CC_INIT, dtc->base + CMN_DT_PMCCNTR); + writeq_relaxed(CMN_CC_INIT, CMN_DT_PMCCNTR(dtc)); dtc->cc_active = true; } else if (type == CMN_TYPE_WP) { u64 val = CMN_EVENT_WP_VAL(event); @@ -2028,7 +2035,7 @@ static irqreturn_t arm_cmn_handle_irq(int irq, void *dev_id) irqreturn_t ret = IRQ_NONE; for (;;) { - u32 status = readl_relaxed(dtc->base + CMN_DT_PMOVSR); + u32 status = readl_relaxed(CMN_DT_PMOVSR(dtc)); u64 delta; int i; @@ -2050,7 +2057,7 @@ static irqreturn_t arm_cmn_handle_irq(int irq, void *dev_id) } } - writel_relaxed(status, dtc->base + CMN_DT_PMOVSR_CLR); + writel_relaxed(status, CMN_DT_PMOVSR_CLR(dtc)); if (!dtc->irq_friend) return ret; @@ -2104,15 +2111,16 @@ static int arm_cmn_init_dtc(struct arm_cmn *cmn, struct arm_cmn_node *dn, int id { struct arm_cmn_dtc *dtc = cmn->dtc + idx; - dtc->base = dn->pmu_base - CMN_PMU_OFFSET; + dtc->pmu_base = dn->pmu_base; + dtc->base = dtc->pmu_base - arm_cmn_pmu_offset(cmn, dn); dtc->irq = platform_get_irq(to_platform_device(cmn->dev), idx); if (dtc->irq < 0) return dtc->irq; writel_relaxed(CMN_DT_DTC_CTL_DT_EN, dtc->base + CMN_DT_DTC_CTL); - writel_relaxed(CMN_DT_PMCR_PMU_EN | CMN_DT_PMCR_OVFL_INTR_EN, dtc->base + CMN_DT_PMCR); - writeq_relaxed(0, dtc->base + CMN_DT_PMCCNTR); - writel_relaxed(0x1ff, dtc->base + CMN_DT_PMOVSR_CLR); + writel_relaxed(CMN_DT_PMCR_PMU_EN | CMN_DT_PMCR_OVFL_INTR_EN, CMN_DT_PMCR(dtc)); + writeq_relaxed(0, CMN_DT_PMCCNTR(dtc)); + writel_relaxed(0x1ff, CMN_DT_PMOVSR_CLR(dtc)); return 0; } @@ -2200,7 +2208,7 @@ static void arm_cmn_init_node_info(struct arm_cmn *cmn, u32 offset, struct arm_c node->id = FIELD_GET(CMN_NI_NODE_ID, reg); node->logid = FIELD_GET(CMN_NI_LOGICAL_ID, reg); - node->pmu_base = cmn->base + offset + CMN_PMU_OFFSET; + node->pmu_base = cmn->base + offset + arm_cmn_pmu_offset(cmn, node); if (node->type == CMN_TYPE_CFG) level = 0; -- cgit v1.2.3 From 0dc2f4963f7ef187b80d832d7d88f735a9dc99cb Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 2 Sep 2024 18:52:04 +0100 Subject: perf/arm-cmn: Support CMN S3 CMN S3 is the latest and greatest evolution for 2024, although most of the new features don't impact the PMU, so from our point of view it ends up looking a lot like CMN-700 r3 still. We have some new device types to ignore, a mildly irritating rearrangement of the register layouts, and a scary new configuration option that makes it potentially unsafe to even walk the full discovery tree, let alone attempt to use the PMU. Acked-by: Mark Rutland Reviewed-by: Ilkka Koskinen Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/2ec9eec5b6bf215a9886f3b69e3b00e4cd85095c.1725296395.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 119 +++++++++++++++++++++++++++++++------------------ 1 file changed, 76 insertions(+), 43 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index ac7dd4c352e8..0266ff041e0b 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -45,24 +45,28 @@ #define CMN_CFGM_PERIPH_ID_23 0x0010 #define CMN_CFGM_PID2_REVISION GENMASK_ULL(7, 4) -#define CMN_CFGM_INFO_GLOBAL 0x900 +#define CMN_CFGM_INFO_GLOBAL 0x0900 #define CMN_INFO_MULTIPLE_DTM_EN BIT_ULL(63) #define CMN_INFO_RSP_VC_NUM GENMASK_ULL(53, 52) #define CMN_INFO_DAT_VC_NUM GENMASK_ULL(51, 50) +#define CMN_INFO_DEVICE_ISO_ENABLE BIT_ULL(44) -#define CMN_CFGM_INFO_GLOBAL_1 0x908 +#define CMN_CFGM_INFO_GLOBAL_1 0x0908 #define CMN_INFO_SNP_VC_NUM GENMASK_ULL(3, 2) #define CMN_INFO_REQ_VC_NUM GENMASK_ULL(1, 0) /* XPs also have some local topology info which has uses too */ #define CMN_MXP__CONNECT_INFO(p) (0x0008 + 8 * (p)) -#define CMN__CONNECT_INFO_DEVICE_TYPE GENMASK_ULL(4, 0) +#define CMN__CONNECT_INFO_DEVICE_TYPE GENMASK_ULL(5, 0) #define CMN_MAX_PORTS 6 #define CI700_CONNECT_INFO_P2_5_OFFSET 0x10 /* PMU registers occupy the 3rd 4KB page of each node's region */ #define CMN_PMU_OFFSET 0x2000 +/* ...except when they don't :( */ +#define CMN_S3_DTM_OFFSET 0xa000 +#define CMN_S3_PMU_OFFSET 0xd900 /* For most nodes, this is all there is */ #define CMN_PMU_EVENT_SEL 0x000 @@ -195,10 +199,11 @@ enum cmn_model { CMN650 = 2, CMN700 = 4, CI700 = 8, + CMNS3 = 16, /* ...and then we can use bitmap tricks for commonality */ CMN_ANY = -1, NOT_CMN600 = -2, - CMN_650ON = CMN650 | CMN700, + CMN_650ON = CMN650 | CMN700 | CMNS3, }; /* Actual part numbers and revision IDs defined by the hardware */ @@ -207,6 +212,7 @@ enum cmn_part { PART_CMN650 = 0x436, PART_CMN700 = 0x43c, PART_CI700 = 0x43a, + PART_CMN_S3 = 0x43e, }; /* CMN-600 r0px shouldn't exist in silicon, thankfully */ @@ -258,6 +264,7 @@ enum cmn_node_type { CMN_TYPE_HNS = 0x200, CMN_TYPE_HNS_MPAM_S, CMN_TYPE_HNS_MPAM_NS, + CMN_TYPE_APB = 0x1000, /* Not a real node type */ CMN_TYPE_WP = 0x7770 }; @@ -408,6 +415,8 @@ static enum cmn_model arm_cmn_model(const struct arm_cmn *cmn) return CMN700; case PART_CI700: return CI700; + case PART_CMN_S3: + return CMNS3; default: return 0; }; @@ -415,6 +424,11 @@ static enum cmn_model arm_cmn_model(const struct arm_cmn *cmn) static int arm_cmn_pmu_offset(const struct arm_cmn *cmn, const struct arm_cmn_node *dn) { + if (cmn->part == PART_CMN_S3) { + if (dn->type == CMN_TYPE_XP) + return CMN_S3_DTM_OFFSET; + return CMN_S3_PMU_OFFSET; + } return CMN_PMU_OFFSET; } @@ -468,9 +482,14 @@ static const char *arm_cmn_device_type(u8 type) case 0x17: return "RN-F_C_E|"; case 0x18: return " RN-F_E |"; case 0x19: return "RN-F_E_E|"; + case 0x1a: return " HN-S |"; + case 0x1b: return " LCN |"; case 0x1c: return " MTSX |"; case 0x1d: return " HN-V |"; case 0x1e: return " CCG |"; + case 0x20: return " RN-F_F |"; + case 0x21: return "RN-F_F_E|"; + case 0x22: return " SN-F_F |"; default: return " ???? |"; } } @@ -779,8 +798,8 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, CMN_EVENT_ATTR(CMN_ANY, cxha_##_name, CMN_TYPE_CXHA, _event) #define CMN_EVENT_CCRA(_name, _event) \ CMN_EVENT_ATTR(CMN_ANY, ccra_##_name, CMN_TYPE_CCRA, _event) -#define CMN_EVENT_CCHA(_name, _event) \ - CMN_EVENT_ATTR(CMN_ANY, ccha_##_name, CMN_TYPE_CCHA, _event) +#define CMN_EVENT_CCHA(_model, _name, _event) \ + 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) \ @@ -1138,42 +1157,43 @@ static struct attribute *arm_cmn_event_attrs[] = { CMN_EVENT_CCRA(wdb_alloc, 0x59), CMN_EVENT_CCRA(ssb_alloc, 0x5a), - CMN_EVENT_CCHA(rddatbyp, 0x61), - CMN_EVENT_CCHA(chirsp_up_stall, 0x62), - CMN_EVENT_CCHA(chidat_up_stall, 0x63), - CMN_EVENT_CCHA(snppcrd_link0_stall, 0x64), - CMN_EVENT_CCHA(snppcrd_link1_stall, 0x65), - CMN_EVENT_CCHA(snppcrd_link2_stall, 0x66), - CMN_EVENT_CCHA(reqtrk_occ, 0x67), - CMN_EVENT_CCHA(rdb_occ, 0x68), - CMN_EVENT_CCHA(rdbyp_occ, 0x69), - CMN_EVENT_CCHA(wdb_occ, 0x6a), - CMN_EVENT_CCHA(snptrk_occ, 0x6b), - CMN_EVENT_CCHA(sdb_occ, 0x6c), - CMN_EVENT_CCHA(snphaz_occ, 0x6d), - CMN_EVENT_CCHA(reqtrk_alloc, 0x6e), - CMN_EVENT_CCHA(rdb_alloc, 0x6f), - CMN_EVENT_CCHA(rdbyp_alloc, 0x70), - CMN_EVENT_CCHA(wdb_alloc, 0x71), - CMN_EVENT_CCHA(snptrk_alloc, 0x72), - CMN_EVENT_CCHA(sdb_alloc, 0x73), - CMN_EVENT_CCHA(snphaz_alloc, 0x74), - CMN_EVENT_CCHA(pb_rhu_req_occ, 0x75), - CMN_EVENT_CCHA(pb_rhu_req_alloc, 0x76), - CMN_EVENT_CCHA(pb_rhu_pcie_req_occ, 0x77), - CMN_EVENT_CCHA(pb_rhu_pcie_req_alloc, 0x78), - CMN_EVENT_CCHA(pb_pcie_wr_req_occ, 0x79), - CMN_EVENT_CCHA(pb_pcie_wr_req_alloc, 0x7a), - CMN_EVENT_CCHA(pb_pcie_reg_req_occ, 0x7b), - CMN_EVENT_CCHA(pb_pcie_reg_req_alloc, 0x7c), - CMN_EVENT_CCHA(pb_pcie_rsvd_req_occ, 0x7d), - CMN_EVENT_CCHA(pb_pcie_rsvd_req_alloc, 0x7e), - CMN_EVENT_CCHA(pb_rhu_dat_occ, 0x7f), - CMN_EVENT_CCHA(pb_rhu_dat_alloc, 0x80), - CMN_EVENT_CCHA(pb_rhu_pcie_dat_occ, 0x81), - CMN_EVENT_CCHA(pb_rhu_pcie_dat_alloc, 0x82), - CMN_EVENT_CCHA(pb_pcie_wr_dat_occ, 0x83), - CMN_EVENT_CCHA(pb_pcie_wr_dat_alloc, 0x84), + CMN_EVENT_CCHA(CMN_ANY, rddatbyp, 0x61), + CMN_EVENT_CCHA(CMN_ANY, chirsp_up_stall, 0x62), + CMN_EVENT_CCHA(CMN_ANY, chidat_up_stall, 0x63), + CMN_EVENT_CCHA(CMN_ANY, snppcrd_link0_stall, 0x64), + CMN_EVENT_CCHA(CMN_ANY, snppcrd_link1_stall, 0x65), + CMN_EVENT_CCHA(CMN_ANY, snppcrd_link2_stall, 0x66), + CMN_EVENT_CCHA(CMN_ANY, reqtrk_occ, 0x67), + CMN_EVENT_CCHA(CMN_ANY, rdb_occ, 0x68), + CMN_EVENT_CCHA(CMN_ANY, rdbyp_occ, 0x69), + CMN_EVENT_CCHA(CMN_ANY, wdb_occ, 0x6a), + CMN_EVENT_CCHA(CMN_ANY, snptrk_occ, 0x6b), + CMN_EVENT_CCHA(CMN_ANY, sdb_occ, 0x6c), + CMN_EVENT_CCHA(CMN_ANY, snphaz_occ, 0x6d), + CMN_EVENT_CCHA(CMN_ANY, reqtrk_alloc, 0x6e), + CMN_EVENT_CCHA(CMN_ANY, rdb_alloc, 0x6f), + CMN_EVENT_CCHA(CMN_ANY, rdbyp_alloc, 0x70), + CMN_EVENT_CCHA(CMN_ANY, wdb_alloc, 0x71), + CMN_EVENT_CCHA(CMN_ANY, snptrk_alloc, 0x72), + CMN_EVENT_CCHA(CMN_ANY, db_alloc, 0x73), + CMN_EVENT_CCHA(CMN_ANY, snphaz_alloc, 0x74), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_req_occ, 0x75), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_req_alloc, 0x76), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_pcie_req_occ, 0x77), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_pcie_req_alloc, 0x78), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_wr_req_occ, 0x79), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_wr_req_alloc, 0x7a), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_reg_req_occ, 0x7b), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_reg_req_alloc, 0x7c), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_rsvd_req_occ, 0x7d), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_rsvd_req_alloc, 0x7e), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_dat_occ, 0x7f), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_dat_alloc, 0x80), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_pcie_dat_occ, 0x81), + CMN_EVENT_CCHA(CMN_ANY, pb_rhu_pcie_dat_alloc, 0x82), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_wr_dat_occ, 0x83), + CMN_EVENT_CCHA(CMN_ANY, pb_pcie_wr_dat_alloc, 0x84), + CMN_EVENT_CCHA(CMNS3, chirsp1_up_stall, 0x85), CMN_EVENT_CCLA(rx_cxs, 0x21), CMN_EVENT_CCLA(tx_cxs, 0x22), @@ -1779,7 +1799,8 @@ static int arm_cmn_event_init(struct perf_event *event) /* ...but the DTM may depend on which port we're watching */ if (cmn->multi_dtm) hw->dtm_offset = CMN_EVENT_WP_DEV_SEL(event) / 2; - } else if (type == CMN_TYPE_XP && cmn->part == PART_CMN700) { + } else if (type == CMN_TYPE_XP && + (cmn->part == PART_CMN700 || cmn->part == PART_CMN_S3)) { hw->wide_sel = true; } @@ -2266,7 +2287,17 @@ static int arm_cmn_discover(struct arm_cmn *cmn, unsigned int rgn_offset) reg = readl_relaxed(cfg_region + CMN_CFGM_PERIPH_ID_23); cmn->rev = FIELD_GET(CMN_CFGM_PID2_REVISION, reg); + /* + * With the device isolation feature, if firmware has neglected to enable + * an XP port then we risk locking up if we try to access anything behind + * it; however we also have no way to tell from Non-Secure whether any + * given port is disabled or not, so the only way to win is not to play... + */ reg = readq_relaxed(cfg_region + CMN_CFGM_INFO_GLOBAL); + if (reg & CMN_INFO_DEVICE_ISO_ENABLE) { + dev_err(cmn->dev, "Device isolation enabled, not continuing due to risk of lockup\n"); + return -ENODEV; + } cmn->multi_dtm = reg & CMN_INFO_MULTIPLE_DTM_EN; cmn->rsp_vc_num = FIELD_GET(CMN_INFO_RSP_VC_NUM, reg); cmn->dat_vc_num = FIELD_GET(CMN_INFO_DAT_VC_NUM, reg); @@ -2425,6 +2456,7 @@ static int arm_cmn_discover(struct arm_cmn *cmn, unsigned int rgn_offset) case CMN_TYPE_CXLA: case CMN_TYPE_HNS_MPAM_S: case CMN_TYPE_HNS_MPAM_NS: + case CMN_TYPE_APB: break; /* * Split "optimised" combination nodes into separate @@ -2610,6 +2642,7 @@ static const struct of_device_id arm_cmn_of_match[] = { { .compatible = "arm,cmn-600", .data = (void *)PART_CMN600 }, { .compatible = "arm,cmn-650" }, { .compatible = "arm,cmn-700" }, + { .compatible = "arm,cmn-s3" }, { .compatible = "arm,ci-700" }, {} }; -- cgit v1.2.3 From f04b611e66503336bbdac04eb5a76d62932ce2e3 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Wed, 4 Sep 2024 19:41:54 +0100 Subject: perf/arm-cmn: Clean up unnecessary NUMA_NO_NODE check Checking for NUMA_NO_NODE is a misleading and, on reflection, entirely unnecessary micro-optimisation. If it ever did happen that an incoming CPU has no NUMA affinity while the current CPU does, a questionably- useful PMU migration isn't the biggest thing wrong with that picture... Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/00634da33c21269a00844140afc7cc3a2ac1eb4d.1725474584.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 0266ff041e0b..62d4782da7e4 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -2023,7 +2023,7 @@ static int arm_cmn_pmu_online_cpu(unsigned int cpu, struct hlist_node *cpuhp_nod cmn = hlist_entry_safe(cpuhp_node, struct arm_cmn, cpuhp_node); node = dev_to_node(cmn->dev); - if (node != NUMA_NO_NODE && cpu_to_node(cmn->cpu) != node && cpu_to_node(cpu) == node) + if (cpu_to_node(cmn->cpu) != node && cpu_to_node(cpu) == node) arm_cmn_migrate(cmn, cpu); return 0; } -- cgit v1.2.3 From f32efa3e4bba5b3432d7932dc89bd2e36c5c0f49 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Wed, 4 Sep 2024 19:41:55 +0100 Subject: perf/arm-cmn: Improve format attr printing Take full advantage of our formats being stored in bitfield form, and make the printing even more robust and simple by letting printk do all the hard work of formatting bitlists. Signed-off-by: Robin Murphy Reviewed-by: Ilkka Koskinen Link: https://lore.kernel.org/r/50459f2d48fc62310a566863dbf8a7c14361d363.1725474584.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 62d4782da7e4..397a46410f7c 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -1280,15 +1280,11 @@ static ssize_t arm_cmn_format_show(struct device *dev, struct device_attribute *attr, char *buf) { struct arm_cmn_format_attr *fmt = container_of(attr, typeof(*fmt), attr); - int lo = __ffs(fmt->field), hi = __fls(fmt->field); - - if (lo == hi) - return sysfs_emit(buf, "config:%d\n", lo); if (!fmt->config) - return sysfs_emit(buf, "config:%d-%d\n", lo, hi); + return sysfs_emit(buf, "config:%*pbl\n", 64, &fmt->field); - return sysfs_emit(buf, "config%d:%d-%d\n", fmt->config, lo, hi); + return sysfs_emit(buf, "config%d:%*pbl\n", fmt->config, 64, &fmt->field); } #define _CMN_FORMAT_ATTR(_name, _cfg, _fld) \ -- cgit v1.2.3 From 4d5a7680f2b4d0c2955e1d9f9a594b050d637436 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Wed, 4 Sep 2024 18:34:03 +0100 Subject: perf: Add driver for Arm NI-700 interconnect PMU The Arm NI-700 Network-on-Chip Interconnect has a relatively straightforward design with a hierarchy of voltage, power, and clock domains, where each clock domain then contains a number of interface units and a PMU which can monitor events thereon. As such, it begets a relatively straightforward driver to interface those PMUs with perf. Even more so than with arm-cmn, users will require detailed knowledge of the wider system topology in order to meaningfully analyse anything, since the interconnect itself cannot know what lies beyond the boundary of each inscrutably-numbered interface. Given that, for now they are also expected to refer to the NI-700 documentation for the relevant event IDs to provide as well. An identifier is implemented so we can come back and add jevents if anyone really wants to. Signed-off-by: Robin Murphy Link: https://lore.kernel.org/r/9933058d0ab8138c78a61cd6852ea5d5ff48e393.1725470837.git.robin.murphy@arm.com Signed-off-by: Will Deacon --- Documentation/admin-guide/perf/arm-ni.rst | 17 + Documentation/admin-guide/perf/index.rst | 1 + drivers/perf/Kconfig | 7 + drivers/perf/Makefile | 1 + drivers/perf/arm-ni.c | 781 ++++++++++++++++++++++++++++++ 5 files changed, 807 insertions(+) create mode 100644 Documentation/admin-guide/perf/arm-ni.rst create mode 100644 drivers/perf/arm-ni.c (limited to 'drivers') diff --git a/Documentation/admin-guide/perf/arm-ni.rst b/Documentation/admin-guide/perf/arm-ni.rst new file mode 100644 index 000000000000..d26a8f697c36 --- /dev/null +++ b/Documentation/admin-guide/perf/arm-ni.rst @@ -0,0 +1,17 @@ +==================================== +Arm Network-on Chip Interconnect PMU +==================================== + +NI-700 and friends implement a distinct PMU for each clock domain within the +interconnect. Correspondingly, the driver exposes multiple PMU devices named +arm_ni__cd_, where is an (arbitrary) instance identifier and is +the clock domain ID within that particular instance. If multiple NI instances +exist within a system, the PMU devices can be correlated with the underlying +hardware instance via sysfs parentage. + +Each PMU exposes base event aliases for the interface types present in its clock +domain. These require qualifying with the "eventid" and "nodeid" parameters +to specify the event code to count and the interface at which to count it +(per the configured hardware ID as reflected in the xxNI_NODE_INFO register). +The exception is the "cycles" alias for the PMU cycle counter, which is encoded +with the PMU node type and needs no further qualification. diff --git a/Documentation/admin-guide/perf/index.rst b/Documentation/admin-guide/perf/index.rst index 7eb3dcd6f4da..8502bc174640 100644 --- a/Documentation/admin-guide/perf/index.rst +++ b/Documentation/admin-guide/perf/index.rst @@ -16,6 +16,7 @@ Performance monitor support starfive_starlink_pmu arm-ccn arm-cmn + arm-ni xgene-pmu arm_dsu_pmu thunderx2-pmu diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig index aa9530b4064f..bab8ba64162f 100644 --- a/drivers/perf/Kconfig +++ b/drivers/perf/Kconfig @@ -48,6 +48,13 @@ config ARM_CMN Support for PMU events monitoring on the Arm CMN-600 Coherent Mesh Network interconnect. +config ARM_NI + tristate "Arm NI-700 PMU support" + depends on ARM64 || COMPILE_TEST + help + Support for PMU events monitoring on the Arm NI-700 Network-on-Chip + interconnect and family. + config ARM_PMU depends on ARM || ARM64 bool "ARM PMU framework" diff --git a/drivers/perf/Makefile b/drivers/perf/Makefile index d43df81d52f7..8268f38e42c5 100644 --- a/drivers/perf/Makefile +++ b/drivers/perf/Makefile @@ -3,6 +3,7 @@ obj-$(CONFIG_ARM_CCI_PMU) += arm-cci.o obj-$(CONFIG_ARM_CCN) += arm-ccn.o obj-$(CONFIG_ARM_CMN) += arm-cmn.o obj-$(CONFIG_ARM_DSU_PMU) += arm_dsu_pmu.o +obj-$(CONFIG_ARM_NI) += arm-ni.o obj-$(CONFIG_ARM_PMU) += arm_pmu.o arm_pmu_platform.o obj-$(CONFIG_ARM_PMU_ACPI) += arm_pmu_acpi.o obj-$(CONFIG_ARM_PMUV3) += arm_pmuv3.o diff --git a/drivers/perf/arm-ni.c b/drivers/perf/arm-ni.c new file mode 100644 index 000000000000..b72df3aea93e --- /dev/null +++ b/drivers/perf/arm-ni.c @@ -0,0 +1,781 @@ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (C) 2022-2024 Arm Limited +// NI-700 Network-on-Chip PMU driver + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Common registers */ +#define NI_NODE_TYPE 0x000 +#define NI_NODE_TYPE_NODE_ID GENMASK(31, 16) +#define NI_NODE_TYPE_NODE_TYPE GENMASK(15, 0) + +#define NI_CHILD_NODE_INFO 0x004 +#define NI_CHILD_PTR(n) (0x008 + (n) * 4) + +#define NI700_PMUSELA 0x00c + +/* Config node */ +#define NI_PERIPHERAL_ID0 0xfe0 +#define NI_PIDR0_PART_7_0 GENMASK(7, 0) +#define NI_PERIPHERAL_ID1 0xfe4 +#define NI_PIDR1_PART_11_8 GENMASK(3, 0) +#define NI_PERIPHERAL_ID2 0xfe8 +#define NI_PIDR2_VERSION GENMASK(7, 4) + +/* PMU node */ +#define NI_PMEVCNTR(n) (0x008 + (n) * 8) +#define NI_PMCCNTR_L 0x0f8 +#define NI_PMCCNTR_U 0x0fc +#define NI_PMEVTYPER(n) (0x400 + (n) * 4) +#define NI_PMEVTYPER_NODE_TYPE GENMASK(12, 9) +#define NI_PMEVTYPER_NODE_ID GENMASK(8, 0) +#define NI_PMCNTENSET 0xc00 +#define NI_PMCNTENCLR 0xc20 +#define NI_PMINTENSET 0xc40 +#define NI_PMINTENCLR 0xc60 +#define NI_PMOVSCLR 0xc80 +#define NI_PMOVSSET 0xcc0 +#define NI_PMCFGR 0xe00 +#define NI_PMCR 0xe04 +#define NI_PMCR_RESET_CCNT BIT(2) +#define NI_PMCR_RESET_EVCNT BIT(1) +#define NI_PMCR_ENABLE BIT(0) + +#define NI_NUM_COUNTERS 8 +#define NI_CCNT_IDX 31 + +/* Event attributes */ +#define NI_CONFIG_TYPE GENMASK_ULL(15, 0) +#define NI_CONFIG_NODEID GENMASK_ULL(31, 16) +#define NI_CONFIG_EVENTID GENMASK_ULL(47, 32) + +#define NI_EVENT_TYPE(event) FIELD_GET(NI_CONFIG_TYPE, (event)->attr.config) +#define NI_EVENT_NODEID(event) FIELD_GET(NI_CONFIG_NODEID, (event)->attr.config) +#define NI_EVENT_EVENTID(event) FIELD_GET(NI_CONFIG_EVENTID, (event)->attr.config) + +enum ni_part { + PART_NI_700 = 0x43b, + PART_NI_710AE = 0x43d, +}; + +enum ni_node_type { + NI_GLOBAL, + NI_VOLTAGE, + NI_POWER, + NI_CLOCK, + NI_ASNI, + NI_AMNI, + NI_PMU, + NI_HSNI, + NI_HMNI, + NI_PMNI, +}; + +struct arm_ni_node { + void __iomem *base; + enum ni_node_type type; + u16 id; + u32 num_components; +}; + +struct arm_ni_unit { + void __iomem *pmusela; + enum ni_node_type type; + u16 id; + bool ns; + union { + __le64 pmusel; + u8 event[8]; + }; +}; + +struct arm_ni_cd { + void __iomem *pmu_base; + u16 id; + int num_units; + int irq; + int cpu; + struct hlist_node cpuhp_node; + struct pmu pmu; + struct arm_ni_unit *units; + struct perf_event *evcnt[NI_NUM_COUNTERS]; + struct perf_event *ccnt; +}; + +struct arm_ni { + struct device *dev; + void __iomem *base; + enum ni_part part; + int id; + int num_cds; + struct arm_ni_cd cds[] __counted_by(num_cds); +}; + +#define cd_to_ni(cd) container_of((cd), struct arm_ni, cds[(cd)->id]) +#define pmu_to_cd(p) container_of((p), struct arm_ni_cd, pmu) + +#define cd_for_each_unit(cd, u) \ + for (struct arm_ni_unit *u = cd->units; u < cd->units + cd->num_units; u++) + +static int arm_ni_hp_state; + +struct arm_ni_event_attr { + struct device_attribute attr; + enum ni_node_type type; +}; + +#define NI_EVENT_ATTR(_name, _type) \ + (&((struct arm_ni_event_attr[]) {{ \ + .attr = __ATTR(_name, 0444, arm_ni_event_show, NULL), \ + .type = _type, \ + }})[0].attr.attr) + +static ssize_t arm_ni_event_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct arm_ni_event_attr *eattr = container_of(attr, typeof(*eattr), attr); + + if (eattr->type == NI_PMU) + return sysfs_emit(buf, "type=0x%x\n", eattr->type); + + return sysfs_emit(buf, "type=0x%x,eventid=?,nodeid=?\n", eattr->type); +} + +static umode_t arm_ni_event_attr_is_visible(struct kobject *kobj, + struct attribute *attr, int unused) +{ + struct device *dev = kobj_to_dev(kobj); + struct arm_ni_cd *cd = pmu_to_cd(dev_get_drvdata(dev)); + struct arm_ni_event_attr *eattr; + + eattr = container_of(attr, typeof(*eattr), attr.attr); + + cd_for_each_unit(cd, unit) { + if (unit->type == eattr->type && unit->ns) + return attr->mode; + } + + return 0; +} + +static struct attribute *arm_ni_event_attrs[] = { + NI_EVENT_ATTR(asni, NI_ASNI), + NI_EVENT_ATTR(amni, NI_AMNI), + NI_EVENT_ATTR(cycles, NI_PMU), + NI_EVENT_ATTR(hsni, NI_HSNI), + NI_EVENT_ATTR(hmni, NI_HMNI), + NI_EVENT_ATTR(pmni, NI_PMNI), + NULL +}; + +static const struct attribute_group arm_ni_event_attrs_group = { + .name = "events", + .attrs = arm_ni_event_attrs, + .is_visible = arm_ni_event_attr_is_visible, +}; + +struct arm_ni_format_attr { + struct device_attribute attr; + u64 field; +}; + +#define NI_FORMAT_ATTR(_name, _fld) \ + (&((struct arm_ni_format_attr[]) {{ \ + .attr = __ATTR(_name, 0444, arm_ni_format_show, NULL), \ + .field = _fld, \ + }})[0].attr.attr) + +static ssize_t arm_ni_format_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct arm_ni_format_attr *fmt = container_of(attr, typeof(*fmt), attr); + + return sysfs_emit(buf, "config:%*pbl\n", 64, &fmt->field); +} + +static struct attribute *arm_ni_format_attrs[] = { + NI_FORMAT_ATTR(type, NI_CONFIG_TYPE), + NI_FORMAT_ATTR(nodeid, NI_CONFIG_NODEID), + NI_FORMAT_ATTR(eventid, NI_CONFIG_EVENTID), + NULL +}; + +static const struct attribute_group arm_ni_format_attrs_group = { + .name = "format", + .attrs = arm_ni_format_attrs, +}; + +static ssize_t arm_ni_cpumask_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct arm_ni_cd *cd = pmu_to_cd(dev_get_drvdata(dev)); + + return cpumap_print_to_pagebuf(true, buf, cpumask_of(cd->cpu)); +} + +static struct device_attribute arm_ni_cpumask_attr = + __ATTR(cpumask, 0444, arm_ni_cpumask_show, NULL); + +static ssize_t arm_ni_identifier_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct arm_ni *ni = cd_to_ni(pmu_to_cd(dev_get_drvdata(dev))); + u32 reg = readl_relaxed(ni->base + NI_PERIPHERAL_ID2); + int version = FIELD_GET(NI_PIDR2_VERSION, reg); + + return sysfs_emit(buf, "%03x%02x\n", ni->part, version); +} + +static struct device_attribute arm_ni_identifier_attr = + __ATTR(identifier, 0444, arm_ni_identifier_show, NULL); + +static struct attribute *arm_ni_other_attrs[] = { + &arm_ni_cpumask_attr.attr, + &arm_ni_identifier_attr.attr, + NULL +}; + +static const struct attribute_group arm_ni_other_attr_group = { + .attrs = arm_ni_other_attrs, + NULL +}; + +static const struct attribute_group *arm_ni_attr_groups[] = { + &arm_ni_event_attrs_group, + &arm_ni_format_attrs_group, + &arm_ni_other_attr_group, + NULL +}; + +static void arm_ni_pmu_enable(struct pmu *pmu) +{ + writel_relaxed(NI_PMCR_ENABLE, pmu_to_cd(pmu)->pmu_base + NI_PMCR); +} + +static void arm_ni_pmu_disable(struct pmu *pmu) +{ + writel_relaxed(0, pmu_to_cd(pmu)->pmu_base + NI_PMCR); +} + +struct arm_ni_val { + unsigned int evcnt; + unsigned int ccnt; +}; + +static bool arm_ni_val_count_event(struct perf_event *evt, struct arm_ni_val *val) +{ + if (is_software_event(evt)) + return true; + + if (NI_EVENT_TYPE(evt) == NI_PMU) { + val->ccnt++; + return val->ccnt <= 1; + } + + val->evcnt++; + return val->evcnt <= NI_NUM_COUNTERS; +} + +static int arm_ni_validate_group(struct perf_event *event) +{ + struct perf_event *sibling, *leader = event->group_leader; + struct arm_ni_val val = { 0 }; + + if (leader == event) + return 0; + + arm_ni_val_count_event(event, &val); + if (!arm_ni_val_count_event(leader, &val)) + return -EINVAL; + + for_each_sibling_event(sibling, leader) { + if (!arm_ni_val_count_event(sibling, &val)) + return -EINVAL; + } + return 0; +} + +static int arm_ni_event_init(struct perf_event *event) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + + if (event->attr.type != event->pmu->type) + return -ENOENT; + + if (is_sampling_event(event)) + return -EINVAL; + + event->cpu = cd->cpu; + if (NI_EVENT_TYPE(event) == NI_PMU) + return arm_ni_validate_group(event); + + cd_for_each_unit(cd, unit) { + if (unit->type == NI_EVENT_TYPE(event) && + unit->id == NI_EVENT_NODEID(event) && unit->ns) { + event->hw.config_base = (unsigned long)unit; + return arm_ni_validate_group(event); + } + } + return -EINVAL; +} + +static u64 arm_ni_read_ccnt(struct arm_ni_cd *cd) +{ + u64 l, u_old, u_new; + int retries = 3; /* 1st time unlucky, 2nd improbable, 3rd just broken */ + + u_new = readl_relaxed(cd->pmu_base + NI_PMCCNTR_U); + do { + u_old = u_new; + l = readl_relaxed(cd->pmu_base + NI_PMCCNTR_L); + u_new = readl_relaxed(cd->pmu_base + NI_PMCCNTR_U); + } while (u_new != u_old && --retries); + WARN_ON(!retries); + + return (u_new << 32) | l; +} + +static void arm_ni_event_read(struct perf_event *event) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + struct hw_perf_event *hw = &event->hw; + u64 count, prev; + bool ccnt = hw->idx == NI_CCNT_IDX; + + do { + prev = local64_read(&hw->prev_count); + if (ccnt) + count = arm_ni_read_ccnt(cd); + else + count = readl_relaxed(cd->pmu_base + NI_PMEVCNTR(hw->idx)); + } while (local64_cmpxchg(&hw->prev_count, prev, count) != prev); + + count -= prev; + if (!ccnt) + count = (u32)count; + local64_add(count, &event->count); +} + +static void arm_ni_event_start(struct perf_event *event, int flags) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + + writel_relaxed(1U << event->hw.idx, cd->pmu_base + NI_PMCNTENSET); +} + +static void arm_ni_event_stop(struct perf_event *event, int flags) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + + writel_relaxed(1U << event->hw.idx, cd->pmu_base + NI_PMCNTENCLR); + if (flags & PERF_EF_UPDATE) + arm_ni_event_read(event); +} + +static void arm_ni_init_ccnt(struct arm_ni_cd *cd) +{ + local64_set(&cd->ccnt->hw.prev_count, S64_MIN); + lo_hi_writeq_relaxed(S64_MIN, cd->pmu_base + NI_PMCCNTR_L); +} + +static void arm_ni_init_evcnt(struct arm_ni_cd *cd, int idx) +{ + local64_set(&cd->evcnt[idx]->hw.prev_count, S32_MIN); + writel_relaxed(S32_MIN, cd->pmu_base + NI_PMEVCNTR(idx)); +} + +static int arm_ni_event_add(struct perf_event *event, int flags) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + struct hw_perf_event *hw = &event->hw; + struct arm_ni_unit *unit; + enum ni_node_type type = NI_EVENT_TYPE(event); + u32 reg; + + if (type == NI_PMU) { + if (cd->ccnt) + return -ENOSPC; + hw->idx = NI_CCNT_IDX; + cd->ccnt = event; + arm_ni_init_ccnt(cd); + } else { + hw->idx = 0; + while (cd->evcnt[hw->idx]) { + if (++hw->idx == NI_NUM_COUNTERS) + return -ENOSPC; + } + cd->evcnt[hw->idx] = event; + unit = (void *)hw->config_base; + unit->event[hw->idx] = NI_EVENT_EVENTID(event); + arm_ni_init_evcnt(cd, hw->idx); + lo_hi_writeq_relaxed(le64_to_cpu(unit->pmusel), unit->pmusela); + + reg = FIELD_PREP(NI_PMEVTYPER_NODE_TYPE, type) | + FIELD_PREP(NI_PMEVTYPER_NODE_ID, NI_EVENT_NODEID(event)); + writel_relaxed(reg, cd->pmu_base + NI_PMEVTYPER(hw->idx)); + } + if (flags & PERF_EF_START) + arm_ni_event_start(event, 0); + return 0; +} + +static void arm_ni_event_del(struct perf_event *event, int flags) +{ + struct arm_ni_cd *cd = pmu_to_cd(event->pmu); + struct hw_perf_event *hw = &event->hw; + + arm_ni_event_stop(event, PERF_EF_UPDATE); + + if (hw->idx == NI_CCNT_IDX) + cd->ccnt = NULL; + else + cd->evcnt[hw->idx] = NULL; +} + +static irqreturn_t arm_ni_handle_irq(int irq, void *dev_id) +{ + struct arm_ni_cd *cd = dev_id; + irqreturn_t ret = IRQ_NONE; + u32 reg = readl_relaxed(cd->pmu_base + NI_PMOVSCLR); + + if (reg & (1U << NI_CCNT_IDX)) { + ret = IRQ_HANDLED; + if (!(WARN_ON(!cd->ccnt))) { + arm_ni_event_read(cd->ccnt); + arm_ni_init_ccnt(cd); + } + } + for (int i = 0; i < NI_NUM_COUNTERS; i++) { + if (!(reg & (1U << i))) + continue; + ret = IRQ_HANDLED; + if (!(WARN_ON(!cd->evcnt[i]))) { + arm_ni_event_read(cd->evcnt[i]); + arm_ni_init_evcnt(cd, i); + } + } + writel_relaxed(reg, cd->pmu_base + NI_PMOVSCLR); + return ret; +} + +static int arm_ni_init_cd(struct arm_ni *ni, struct arm_ni_node *node, u64 res_start) +{ + struct arm_ni_cd *cd = ni->cds + node->id; + const char *name; + int err; + + cd->id = node->id; + cd->num_units = node->num_components; + cd->units = devm_kcalloc(ni->dev, cd->num_units, sizeof(*(cd->units)), GFP_KERNEL); + if (!cd->units) + return -ENOMEM; + + for (int i = 0; i < cd->num_units; i++) { + u32 reg = readl_relaxed(node->base + NI_CHILD_PTR(i)); + void __iomem *unit_base = ni->base + reg; + struct arm_ni_unit *unit = cd->units + i; + + reg = readl_relaxed(unit_base + NI_NODE_TYPE); + unit->type = FIELD_GET(NI_NODE_TYPE_NODE_TYPE, reg); + unit->id = FIELD_GET(NI_NODE_TYPE_NODE_ID, reg); + + switch (unit->type) { + case NI_PMU: + reg = readl_relaxed(unit_base + NI_PMCFGR); + if (!reg) { + dev_info(ni->dev, "No access to PMU %d\n", cd->id); + devm_kfree(ni->dev, cd->units); + return 0; + } + unit->ns = true; + cd->pmu_base = unit_base; + break; + case NI_ASNI: + case NI_AMNI: + case NI_HSNI: + case NI_HMNI: + case NI_PMNI: + unit->pmusela = unit_base + NI700_PMUSELA; + writel_relaxed(1, unit->pmusela); + if (readl_relaxed(unit->pmusela) != 1) + dev_info(ni->dev, "No access to node 0x%04x%04x\n", unit->id, unit->type); + else + unit->ns = true; + break; + default: + /* + * e.g. FMU - thankfully bits 3:2 of FMU_ERR_FR0 are RES0 so + * can't alias any of the leaf node types we're looking for. + */ + dev_dbg(ni->dev, "Mystery node 0x%04x%04x\n", unit->id, unit->type); + break; + } + } + + res_start += cd->pmu_base - ni->base; + if (!devm_request_mem_region(ni->dev, res_start, SZ_4K, dev_name(ni->dev))) { + dev_err(ni->dev, "Failed to request PMU region 0x%llx\n", res_start); + return -EBUSY; + } + + writel_relaxed(NI_PMCR_RESET_CCNT | NI_PMCR_RESET_EVCNT, + cd->pmu_base + NI_PMCR); + writel_relaxed(U32_MAX, cd->pmu_base + NI_PMCNTENCLR); + writel_relaxed(U32_MAX, cd->pmu_base + NI_PMOVSCLR); + writel_relaxed(U32_MAX, cd->pmu_base + NI_PMINTENSET); + + cd->irq = platform_get_irq(to_platform_device(ni->dev), cd->id); + if (cd->irq < 0) + return cd->irq; + + err = devm_request_irq(ni->dev, cd->irq, arm_ni_handle_irq, + IRQF_NOBALANCING | IRQF_NO_THREAD, + dev_name(ni->dev), cd); + if (err) + return err; + + cd->cpu = cpumask_local_spread(0, dev_to_node(ni->dev)); + cd->pmu = (struct pmu) { + .module = THIS_MODULE, + .parent = ni->dev, + .attr_groups = arm_ni_attr_groups, + .capabilities = PERF_PMU_CAP_NO_EXCLUDE, + .task_ctx_nr = perf_invalid_context, + .pmu_enable = arm_ni_pmu_enable, + .pmu_disable = arm_ni_pmu_disable, + .event_init = arm_ni_event_init, + .add = arm_ni_event_add, + .del = arm_ni_event_del, + .start = arm_ni_event_start, + .stop = arm_ni_event_stop, + .read = arm_ni_event_read, + }; + + name = devm_kasprintf(ni->dev, GFP_KERNEL, "arm_ni_%d_cd_%d", ni->id, cd->id); + if (!name) + return -ENOMEM; + + err = cpuhp_state_add_instance_nocalls(arm_ni_hp_state, &cd->cpuhp_node); + if (err) + return err; + + err = perf_pmu_register(&cd->pmu, name, -1); + if (err) + cpuhp_state_remove_instance_nocalls(arm_ni_hp_state, &cd->cpuhp_node); + + return err; +} + +static void arm_ni_probe_domain(void __iomem *base, struct arm_ni_node *node) +{ + u32 reg = readl_relaxed(base + NI_NODE_TYPE); + + node->base = base; + node->type = FIELD_GET(NI_NODE_TYPE_NODE_TYPE, reg); + node->id = FIELD_GET(NI_NODE_TYPE_NODE_ID, reg); + node->num_components = readl_relaxed(base + NI_CHILD_NODE_INFO); +} + +static int arm_ni_probe(struct platform_device *pdev) +{ + struct arm_ni_node cfg, vd, pd, cd; + struct arm_ni *ni; + struct resource *res; + void __iomem *base; + static atomic_t id; + int num_cds; + u32 reg, part; + + /* + * We want to map the whole configuration space for ease of discovery, + * but the PMU pages are the only ones for which we can honestly claim + * exclusive ownership, so we'll request them explicitly once found. + */ + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + base = devm_ioremap(&pdev->dev, res->start, resource_size(res)); + if (IS_ERR(base)) + return PTR_ERR(base); + + arm_ni_probe_domain(base, &cfg); + if (cfg.type != NI_GLOBAL) + return -ENODEV; + + reg = readl_relaxed(cfg.base + NI_PERIPHERAL_ID0); + part = FIELD_GET(NI_PIDR0_PART_7_0, reg); + reg = readl_relaxed(cfg.base + NI_PERIPHERAL_ID1); + part |= FIELD_GET(NI_PIDR1_PART_11_8, reg) << 8; + + switch (part) { + case PART_NI_700: + case PART_NI_710AE: + break; + default: + dev_WARN(&pdev->dev, "Unknown part number: 0x%03x, this may go badly\n", part); + break; + } + + num_cds = 0; + for (int v = 0; v < cfg.num_components; v++) { + reg = readl_relaxed(cfg.base + NI_CHILD_PTR(v)); + arm_ni_probe_domain(base + reg, &vd); + for (int p = 0; p < vd.num_components; p++) { + reg = readl_relaxed(vd.base + NI_CHILD_PTR(p)); + arm_ni_probe_domain(base + reg, &pd); + num_cds += pd.num_components; + } + } + + ni = devm_kzalloc(&pdev->dev, struct_size(ni, cds, num_cds), GFP_KERNEL); + if (!ni) + return -ENOMEM; + + ni->dev = &pdev->dev; + ni->base = base; + ni->num_cds = num_cds; + ni->part = part; + ni->id = atomic_fetch_inc(&id); + + for (int v = 0; v < cfg.num_components; v++) { + reg = readl_relaxed(cfg.base + NI_CHILD_PTR(v)); + arm_ni_probe_domain(base + reg, &vd); + for (int p = 0; p < vd.num_components; p++) { + reg = readl_relaxed(vd.base + NI_CHILD_PTR(p)); + arm_ni_probe_domain(base + reg, &pd); + for (int c = 0; c < pd.num_components; c++) { + int ret; + + 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) + return ret; + } + } + } + + 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" }, + {} +}; +MODULE_DEVICE_TABLE(of, arm_ni_of_match); +#endif + +#ifdef CONFIG_ACPI +static const struct acpi_device_id arm_ni_acpi_match[] = { + { "ARMHCB70" }, + {} +}; +MODULE_DEVICE_TABLE(acpi, arm_ni_acpi_match); +#endif + +static struct platform_driver arm_ni_driver = { + .driver = { + .name = "arm-ni", + .of_match_table = of_match_ptr(arm_ni_of_match), + .acpi_match_table = ACPI_PTR(arm_ni_acpi_match), + }, + .probe = arm_ni_probe, + .remove = arm_ni_remove, +}; + +static void arm_ni_pmu_migrate(struct arm_ni_cd *cd, unsigned int cpu) +{ + perf_pmu_migrate_context(&cd->pmu, cd->cpu, cpu); + irq_set_affinity(cd->irq, cpumask_of(cpu)); + cd->cpu = cpu; +} + +static int arm_ni_pmu_online_cpu(unsigned int cpu, struct hlist_node *cpuhp_node) +{ + struct arm_ni_cd *cd; + int node; + + cd = hlist_entry_safe(cpuhp_node, struct arm_ni_cd, cpuhp_node); + node = dev_to_node(cd_to_ni(cd)->dev); + if (cpu_to_node(cd->cpu) != node && cpu_to_node(cpu) == node) + arm_ni_pmu_migrate(cd, cpu); + return 0; +} + +static int arm_ni_pmu_offline_cpu(unsigned int cpu, struct hlist_node *cpuhp_node) +{ + struct arm_ni_cd *cd; + unsigned int target; + int node; + + cd = hlist_entry_safe(cpuhp_node, struct arm_ni_cd, cpuhp_node); + if (cpu != cd->cpu) + return 0; + + node = dev_to_node(cd_to_ni(cd)->dev); + target = cpumask_any_and_but(cpumask_of_node(node), cpu_online_mask, cpu); + if (target >= nr_cpu_ids) + target = cpumask_any_but(cpu_online_mask, cpu); + + if (target < nr_cpu_ids) + arm_ni_pmu_migrate(cd, target); + return 0; +} + +static int __init arm_ni_init(void) +{ + int ret; + + ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, + "perf/arm/ni:online", + arm_ni_pmu_online_cpu, + arm_ni_pmu_offline_cpu); + if (ret < 0) + return ret; + + arm_ni_hp_state = ret; + + ret = platform_driver_register(&arm_ni_driver); + if (ret) + cpuhp_remove_multi_state(arm_ni_hp_state); + return ret; +} + +static void __exit arm_ni_exit(void) +{ + platform_driver_unregister(&arm_ni_driver); + cpuhp_remove_multi_state(arm_ni_hp_state); +} + +module_init(arm_ni_init); +module_exit(arm_ni_exit); + +MODULE_AUTHOR("Robin Murphy "); +MODULE_DESCRIPTION("Arm NI-700 PMU driver"); +MODULE_LICENSE("GPL v2"); -- cgit v1.2.3 From 5967a19f1c2ffb530f5d4589ddc4b4afbb6c7bd4 Mon Sep 17 00:00:00 2001 From: Ilkka Koskinen Date: Fri, 6 Sep 2024 12:15:39 -0700 Subject: perf: arm_pmuv3: Use BR_RETIRED for HW branch event if enabled The PMU driver attempts to use PC_WRITE_RETIRED for the HW branch event, if enabled. However, PC_WRITE_RETIRED counts only taken branches, whereas BR_RETIRED counts also non-taken ones. Furthermore, perf uses HW branch event to calculate branch misses ratio, implying BR_RETIRED is the correct event to count. We keep PC_WRITE_RETIRED still as an option in case BR_RETIRED isn't implemented. Signed-off-by: Ilkka Koskinen Link: https://lore.kernel.org/r/20240906191539.4847-1-ilkka@os.amperecomputing.com Signed-off-by: Will Deacon --- drivers/perf/arm_pmuv3.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c index 4d000532a07f..0afe02f879b4 100644 --- a/drivers/perf/arm_pmuv3.c +++ b/drivers/perf/arm_pmuv3.c @@ -1086,14 +1086,14 @@ static int __armv8_pmuv3_map_event_id(struct arm_pmu *armpmu, if (event->attr.type == PERF_TYPE_HARDWARE && event->attr.config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS) { - if (test_bit(ARMV8_PMUV3_PERFCTR_PC_WRITE_RETIRED, - armpmu->pmceid_bitmap)) - return ARMV8_PMUV3_PERFCTR_PC_WRITE_RETIRED; - if (test_bit(ARMV8_PMUV3_PERFCTR_BR_RETIRED, armpmu->pmceid_bitmap)) return ARMV8_PMUV3_PERFCTR_BR_RETIRED; + if (test_bit(ARMV8_PMUV3_PERFCTR_PC_WRITE_RETIRED, + armpmu->pmceid_bitmap)) + return ARMV8_PMUV3_PERFCTR_PC_WRITE_RETIRED; + return HW_OP_UNSUPPORTED; } -- cgit v1.2.3 From 2e091a805febb9a91cc7de2735d8d4ef0e640241 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 11 Sep 2024 10:39:59 +0300 Subject: perf: arm-ni: Fix an NULL vs IS_ERR() bug The devm_ioremap() function never returns error pointers, it returns a NULL pointer if there is an error. Fixes: 4d5a7680f2b4 ("perf: Add driver for Arm NI-700 interconnect PMU") Signed-off-by: Dan Carpenter Reviewed-by: Robin Murphy Link: https://lore.kernel.org/r/04d6ccc3-6d31-4f0f-ab0f-7a88342cc09a@stanley.mountain Signed-off-by: Will Deacon --- drivers/perf/arm-ni.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/perf/arm-ni.c b/drivers/perf/arm-ni.c index b72df3aea93e..90fcfe693439 100644 --- a/drivers/perf/arm-ni.c +++ b/drivers/perf/arm-ni.c @@ -603,8 +603,8 @@ static int arm_ni_probe(struct platform_device *pdev) */ res = platform_get_resource(pdev, IORESOURCE_MEM, 0); base = devm_ioremap(&pdev->dev, res->start, resource_size(res)); - if (IS_ERR(base)) - return PTR_ERR(base); + if (!base) + return -ENOMEM; arm_ni_probe_domain(base, &cfg); if (cfg.type != NI_GLOBAL) -- cgit v1.2.3