#include <asm/cpufeature.h>
#include <asm/hardirq.h>
+#include <asm/intel-family.h>
#include <asm/apic.h>
#include "../perf_event.h"
INTEL_UEVENT_CONSTRAINT(0x04a3, 0xf), /* CYCLE_ACTIVITY.CYCLES_NO_DISPATCH */
INTEL_UEVENT_CONSTRAINT(0x02a3, 0x4), /* CYCLE_ACTIVITY.CYCLES_L1D_PENDING */
+ /*
+ * When HT is off these events can only run on the bottom 4 counters
+ * When HT is on, they are impacted by the HT bug and require EXCL access
+ */
INTEL_EXCLEVT_CONSTRAINT(0xd0, 0xf), /* MEM_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */
INTEL_UEVENT_CONSTRAINT(0x0ca3, 0x4), /* CYCLE_ACTIVITY.STALLS_L1D_PENDING */
INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */
+ /*
+ * When HT is off these events can only run on the bottom 4 counters
+ * When HT is on, they are impacted by the HT bug and require EXCL access
+ */
INTEL_EXCLEVT_CONSTRAINT(0xd0, 0xf), /* MEM_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */
EVENT_CONSTRAINT_END
};
-struct event_constraint intel_skl_event_constraints[] = {
+static struct event_constraint intel_skl_event_constraints[] = {
FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
INTEL_UEVENT_CONSTRAINT(0x1c0, 0x2), /* INST_RETIRED.PREC_DIST */
+
+ /*
+ * when HT is off, these can only run on the bottom 4 counters
+ */
+ INTEL_EVENT_CONSTRAINT(0xd0, 0xf), /* MEM_INST_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_L3_HIT_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xcd, 0xf), /* MEM_TRANS_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xc6, 0xf), /* FRONTEND_RETIRED.* */
+
EVENT_CONSTRAINT_END
};
static struct extra_reg intel_knl_extra_regs[] __read_mostly = {
- INTEL_UEVENT_EXTRA_REG(0x01b7,
- MSR_OFFCORE_RSP_0, 0x7f9ffbffffull, RSP_0),
- INTEL_UEVENT_EXTRA_REG(0x02b7,
- MSR_OFFCORE_RSP_1, 0x3f9ffbffffull, RSP_1),
+ INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0x799ffbb6e7ull, RSP_0),
+ INTEL_UEVENT_EXTRA_REG(0x02b7, MSR_OFFCORE_RSP_1, 0x399ffbffe7ull, RSP_1),
EVENT_EXTRA_END
};
EVENT_ATTR_STR(mem-loads, mem_ld_snb, "event=0xcd,umask=0x1,ldlat=3");
EVENT_ATTR_STR(mem-stores, mem_st_snb, "event=0xcd,umask=0x2");
-struct attribute *nhm_events_attrs[] = {
+static struct attribute *nhm_events_attrs[] = {
EVENT_PTR(mem_ld_nhm),
NULL,
};
-struct attribute *snb_events_attrs[] = {
+/*
+ * topdown events for Intel Core CPUs.
+ *
+ * The events are all in slots, which is a free slot in a 4 wide
+ * pipeline. Some events are already reported in slots, for cycle
+ * events we multiply by the pipeline width (4).
+ *
+ * With Hyper Threading on, topdown metrics are either summed or averaged
+ * between the threads of a core: (count_t0 + count_t1).
+ *
+ * For the average case the metric is always scaled to pipeline width,
+ * so we use factor 2 ((count_t0 + count_t1) / 2 * 4)
+ */
+
+EVENT_ATTR_STR_HT(topdown-total-slots, td_total_slots,
+ "event=0x3c,umask=0x0", /* cpu_clk_unhalted.thread */
+ "event=0x3c,umask=0x0,any=1"); /* cpu_clk_unhalted.thread_any */
+EVENT_ATTR_STR_HT(topdown-total-slots.scale, td_total_slots_scale, "4", "2");
+EVENT_ATTR_STR(topdown-slots-issued, td_slots_issued,
+ "event=0xe,umask=0x1"); /* uops_issued.any */
+EVENT_ATTR_STR(topdown-slots-retired, td_slots_retired,
+ "event=0xc2,umask=0x2"); /* uops_retired.retire_slots */
+EVENT_ATTR_STR(topdown-fetch-bubbles, td_fetch_bubbles,
+ "event=0x9c,umask=0x1"); /* idq_uops_not_delivered_core */
+EVENT_ATTR_STR_HT(topdown-recovery-bubbles, td_recovery_bubbles,
+ "event=0xd,umask=0x3,cmask=1", /* int_misc.recovery_cycles */
+ "event=0xd,umask=0x3,cmask=1,any=1"); /* int_misc.recovery_cycles_any */
+EVENT_ATTR_STR_HT(topdown-recovery-bubbles.scale, td_recovery_bubbles_scale,
+ "4", "2");
+
+static struct attribute *snb_events_attrs[] = {
EVENT_PTR(mem_ld_snb),
EVENT_PTR(mem_st_snb),
+ EVENT_PTR(td_slots_issued),
+ EVENT_PTR(td_slots_retired),
+ EVENT_PTR(td_fetch_bubbles),
+ EVENT_PTR(td_total_slots),
+ EVENT_PTR(td_total_slots_scale),
+ EVENT_PTR(td_recovery_bubbles),
+ EVENT_PTR(td_recovery_bubbles_scale),
NULL,
};
/* CYCLE_ACTIVITY.CYCLES_NO_EXECUTE */
INTEL_UEVENT_CONSTRAINT(0x04a3, 0xf),
+ /*
+ * When HT is off these events can only run on the bottom 4 counters
+ * When HT is on, they are impacted by the HT bug and require EXCL access
+ */
INTEL_EXCLEVT_CONSTRAINT(0xd0, 0xf), /* MEM_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */
INTEL_EXCLEVT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */
EVENT_CONSTRAINT_END
};
-struct event_constraint intel_bdw_event_constraints[] = {
+static struct event_constraint intel_bdw_event_constraints[] = {
FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */
FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */
FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */
INTEL_UEVENT_CONSTRAINT(0x148, 0x4), /* L1D_PEND_MISS.PENDING */
INTEL_UBIT_EVENT_CONSTRAINT(0x8a3, 0x4), /* CYCLE_ACTIVITY.CYCLES_L1D_MISS */
+ /*
+ * when HT is off, these can only run on the bottom 4 counters
+ */
+ INTEL_EVENT_CONSTRAINT(0xd0, 0xf), /* MEM_INST_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_L3_HIT_RETIRED.* */
+ INTEL_EVENT_CONSTRAINT(0xcd, 0xf), /* MEM_TRANS_RETIRED.* */
EVENT_CONSTRAINT_END
};
},
};
+EVENT_ATTR_STR(topdown-total-slots, td_total_slots_slm, "event=0x3c");
+EVENT_ATTR_STR(topdown-total-slots.scale, td_total_slots_scale_slm, "2");
+/* no_alloc_cycles.not_delivered */
+EVENT_ATTR_STR(topdown-fetch-bubbles, td_fetch_bubbles_slm,
+ "event=0xca,umask=0x50");
+EVENT_ATTR_STR(topdown-fetch-bubbles.scale, td_fetch_bubbles_scale_slm, "2");
+/* uops_retired.all */
+EVENT_ATTR_STR(topdown-slots-issued, td_slots_issued_slm,
+ "event=0xc2,umask=0x10");
+/* uops_retired.all */
+EVENT_ATTR_STR(topdown-slots-retired, td_slots_retired_slm,
+ "event=0xc2,umask=0x10");
+
+static struct attribute *slm_events_attrs[] = {
+ EVENT_PTR(td_total_slots_slm),
+ EVENT_PTR(td_total_slots_scale_slm),
+ EVENT_PTR(td_fetch_bubbles_slm),
+ EVENT_PTR(td_fetch_bubbles_scale_slm),
+ EVENT_PTR(td_slots_issued_slm),
+ EVENT_PTR(td_slots_retired_slm),
+ NULL
+};
+
static struct extra_reg intel_slm_extra_regs[] __read_mostly =
{
/* must define OFFCORE_RSP_X first, see intel_fixup_er() */
u32 rev = UINT_MAX; /* default to broken for unknown models */
switch (cpu_data(cpu).x86_model) {
- case 42: /* SNB */
+ case INTEL_FAM6_SANDYBRIDGE:
rev = 0x28;
break;
- case 45: /* SNB-EP */
+ case INTEL_FAM6_SANDYBRIDGE_X:
switch (cpu_data(cpu).x86_mask) {
case 6: rev = 0x618; break;
case 7: rev = 0x70c; break;
}
}
+static bool is_lbr_from(unsigned long msr)
+{
+ unsigned long lbr_from_nr = x86_pmu.lbr_from + x86_pmu.lbr_nr;
+
+ return x86_pmu.lbr_from <= msr && msr < lbr_from_nr;
+}
+
/*
* Under certain circumstances, access certain MSR may cause #GP.
* The function tests if the input MSR can be safely accessed.
* Only change the bits which can be updated by wrmsrl.
*/
val_tmp = val_old ^ mask;
+
+ if (is_lbr_from(msr))
+ val_tmp = lbr_from_signext_quirk_wr(val_tmp);
+
if (wrmsrl_safe(msr, val_tmp) ||
rdmsrl_safe(msr, &val_new))
return false;
+ /*
+ * Quirk only affects validation in wrmsr(), so wrmsrl()'s value
+ * should equal rdmsrl()'s even with the quirk.
+ */
if (val_new != val_tmp)
return false;
+ if (is_lbr_from(msr))
+ val_old = lbr_from_signext_quirk_wr(val_old);
+
/* Here it's sure that the MSR can be safely accessed.
* Restore the old value and return.
*/
EVENT_PTR(cycles_ct),
EVENT_PTR(mem_ld_hsw),
EVENT_PTR(mem_st_hsw),
+ EVENT_PTR(td_slots_issued),
+ EVENT_PTR(td_slots_retired),
+ EVENT_PTR(td_fetch_bubbles),
+ EVENT_PTR(td_total_slots),
+ EVENT_PTR(td_total_slots_scale),
+ EVENT_PTR(td_recovery_bubbles),
+ EVENT_PTR(td_recovery_bubbles_scale),
NULL
};
* Install the hw-cache-events table:
*/
switch (boot_cpu_data.x86_model) {
- case 14: /* 65nm Core "Yonah" */
+ case INTEL_FAM6_CORE_YONAH:
pr_cont("Core events, ");
break;
- case 15: /* 65nm Core2 "Merom" */
+ case INTEL_FAM6_CORE2_MEROM:
x86_add_quirk(intel_clovertown_quirk);
- case 22: /* 65nm Core2 "Merom-L" */
- case 23: /* 45nm Core2 "Penryn" */
- case 29: /* 45nm Core2 "Dunnington (MP) */
+ case INTEL_FAM6_CORE2_MEROM_L:
+ case INTEL_FAM6_CORE2_PENRYN:
+ case INTEL_FAM6_CORE2_DUNNINGTON:
memcpy(hw_cache_event_ids, core2_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
pr_cont("Core2 events, ");
break;
- case 30: /* 45nm Nehalem */
- case 26: /* 45nm Nehalem-EP */
- case 46: /* 45nm Nehalem-EX */
+ case INTEL_FAM6_NEHALEM:
+ case INTEL_FAM6_NEHALEM_EP:
+ case INTEL_FAM6_NEHALEM_EX:
memcpy(hw_cache_event_ids, nehalem_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, nehalem_hw_cache_extra_regs,
pr_cont("Nehalem events, ");
break;
- case 28: /* 45nm Atom "Pineview" */
- case 38: /* 45nm Atom "Lincroft" */
- case 39: /* 32nm Atom "Penwell" */
- case 53: /* 32nm Atom "Cloverview" */
- case 54: /* 32nm Atom "Cedarview" */
+ case INTEL_FAM6_ATOM_PINEVIEW:
+ case INTEL_FAM6_ATOM_LINCROFT:
+ case INTEL_FAM6_ATOM_PENWELL:
+ case INTEL_FAM6_ATOM_CLOVERVIEW:
+ case INTEL_FAM6_ATOM_CEDARVIEW:
memcpy(hw_cache_event_ids, atom_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
pr_cont("Atom events, ");
break;
- case 55: /* 22nm Atom "Silvermont" */
- case 76: /* 14nm Atom "Airmont" */
- case 77: /* 22nm Atom "Silvermont Avoton/Rangely" */
+ case INTEL_FAM6_ATOM_SILVERMONT1:
+ case INTEL_FAM6_ATOM_SILVERMONT2:
+ case INTEL_FAM6_ATOM_AIRMONT:
memcpy(hw_cache_event_ids, slm_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, slm_hw_cache_extra_regs,
x86_pmu.pebs_constraints = intel_slm_pebs_event_constraints;
x86_pmu.extra_regs = intel_slm_extra_regs;
x86_pmu.flags |= PMU_FL_HAS_RSP_1;
+ x86_pmu.cpu_events = slm_events_attrs;
pr_cont("Silvermont events, ");
break;
- case 92: /* 14nm Atom "Goldmont" */
- case 95: /* 14nm Atom "Goldmont Denverton" */
+ case INTEL_FAM6_ATOM_GOLDMONT:
+ case INTEL_FAM6_ATOM_DENVERTON:
memcpy(hw_cache_event_ids, glm_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, glm_hw_cache_extra_regs,
pr_cont("Goldmont events, ");
break;
- case 37: /* 32nm Westmere */
- case 44: /* 32nm Westmere-EP */
- case 47: /* 32nm Westmere-EX */
+ case INTEL_FAM6_WESTMERE:
+ case INTEL_FAM6_WESTMERE_EP:
+ case INTEL_FAM6_WESTMERE_EX:
memcpy(hw_cache_event_ids, westmere_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, nehalem_hw_cache_extra_regs,
pr_cont("Westmere events, ");
break;
- case 42: /* 32nm SandyBridge */
- case 45: /* 32nm SandyBridge-E/EN/EP */
+ case INTEL_FAM6_SANDYBRIDGE:
+ case INTEL_FAM6_SANDYBRIDGE_X:
x86_add_quirk(intel_sandybridge_quirk);
x86_add_quirk(intel_ht_bug);
memcpy(hw_cache_event_ids, snb_hw_cache_event_ids,
x86_pmu.event_constraints = intel_snb_event_constraints;
x86_pmu.pebs_constraints = intel_snb_pebs_event_constraints;
x86_pmu.pebs_aliases = intel_pebs_aliases_snb;
- if (boot_cpu_data.x86_model == 45)
+ if (boot_cpu_data.x86_model == INTEL_FAM6_SANDYBRIDGE_X)
x86_pmu.extra_regs = intel_snbep_extra_regs;
else
x86_pmu.extra_regs = intel_snb_extra_regs;
pr_cont("SandyBridge events, ");
break;
- case 58: /* 22nm IvyBridge */
- case 62: /* 22nm IvyBridge-EP/EX */
+ case INTEL_FAM6_IVYBRIDGE:
+ case INTEL_FAM6_IVYBRIDGE_X:
x86_add_quirk(intel_ht_bug);
memcpy(hw_cache_event_ids, snb_hw_cache_event_ids,
sizeof(hw_cache_event_ids));
x86_pmu.pebs_constraints = intel_ivb_pebs_event_constraints;
x86_pmu.pebs_aliases = intel_pebs_aliases_ivb;
x86_pmu.pebs_prec_dist = true;
- if (boot_cpu_data.x86_model == 62)
+ if (boot_cpu_data.x86_model == INTEL_FAM6_IVYBRIDGE_X)
x86_pmu.extra_regs = intel_snbep_extra_regs;
else
x86_pmu.extra_regs = intel_snb_extra_regs;
break;
- case 60: /* 22nm Haswell Core */
- case 63: /* 22nm Haswell Server */
- case 69: /* 22nm Haswell ULT */
- case 70: /* 22nm Haswell + GT3e (Intel Iris Pro graphics) */
+ case INTEL_FAM6_HASWELL_CORE:
+ case INTEL_FAM6_HASWELL_X:
+ case INTEL_FAM6_HASWELL_ULT:
+ case INTEL_FAM6_HASWELL_GT3E:
x86_add_quirk(intel_ht_bug);
x86_pmu.late_ack = true;
memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
pr_cont("Haswell events, ");
break;
- case 61: /* 14nm Broadwell Core-M */
- case 86: /* 14nm Broadwell Xeon D */
- case 71: /* 14nm Broadwell + GT3e (Intel Iris Pro graphics) */
- case 79: /* 14nm Broadwell Server */
+ case INTEL_FAM6_BROADWELL_CORE:
+ case INTEL_FAM6_BROADWELL_XEON_D:
+ case INTEL_FAM6_BROADWELL_GT3E:
+ case INTEL_FAM6_BROADWELL_X:
x86_pmu.late_ack = true;
memcpy(hw_cache_event_ids, hsw_hw_cache_event_ids, sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, hsw_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
pr_cont("Broadwell events, ");
break;
- case 87: /* Knights Landing Xeon Phi */
+ case INTEL_FAM6_XEON_PHI_KNL:
memcpy(hw_cache_event_ids,
slm_hw_cache_event_ids, sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs,
pr_cont("Knights Landing events, ");
break;
- case 142: /* 14nm Kabylake Mobile */
- case 158: /* 14nm Kabylake Desktop */
- case 78: /* 14nm Skylake Mobile */
- case 94: /* 14nm Skylake Desktop */
- case 85: /* 14nm Skylake Server */
+ case INTEL_FAM6_SKYLAKE_MOBILE:
+ case INTEL_FAM6_SKYLAKE_DESKTOP:
+ case INTEL_FAM6_SKYLAKE_X:
+ case INTEL_FAM6_KABYLAKE_MOBILE:
+ case INTEL_FAM6_KABYLAKE_DESKTOP:
x86_pmu.late_ack = true;
memcpy(hw_cache_event_ids, skl_hw_cache_event_ids, sizeof(hw_cache_event_ids));
memcpy(hw_cache_extra_regs, skl_hw_cache_extra_regs, sizeof(hw_cache_extra_regs));
intel_pmu_lbr_init_skl();
+ /* INT_MISC.RECOVERY_CYCLES has umask 1 in Skylake */
+ event_attr_td_recovery_bubbles.event_str_noht =
+ "event=0xd,umask=0x1,cmask=1";
+ event_attr_td_recovery_bubbles.event_str_ht =
+ "event=0xd,umask=0x1,cmask=1,any=1";
+
x86_pmu.event_constraints = intel_skl_event_constraints;
x86_pmu.pebs_constraints = intel_skl_pebs_event_constraints;
x86_pmu.extra_regs = intel_skl_extra_regs;
x86_pmu.lbr_nr = 0;
}
+ if (x86_pmu.lbr_nr)
+ pr_cont("%d-deep LBR, ", x86_pmu.lbr_nr);
/*
* Access extra MSR may cause #GP under certain circumstances.
* E.g. KVM doesn't support offcore event
*/
static __init int fixup_ht_bug(void)
{
- int cpu = smp_processor_id();
- int w, c;
+ int c;
/*
* problem not present on this CPU model, nothing to do
*/
if (!(x86_pmu.flags & PMU_FL_EXCL_ENABLED))
return 0;
- w = cpumask_weight(topology_sibling_cpumask(cpu));
- if (w > 1) {
+ if (topology_max_smt_threads() > 1) {
pr_info("PMU erratum BJ122, BV98, HSD29 worked around, HT is on\n");
return 0;
}
projects / cascardo / linux.git / blobdiff