Merge tag 'perf-core-2021-10-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull perf updates from Thomas Gleixner:
 "Core:

   - Allow ftrace to instrument parts of the perf core code

   - Add a new mem_hops field to perf_mem_data_src which allows to
     represent intra-node/package or inter-node/off-package details to
     prepare for next generation systems which have more hieararchy
     within the node/pacakge level.

  Tools:

   - Update for the new mem_hops field in perf_mem_data_src

  Arch:

   - A set of constraints fixes for the Intel uncore PMU

   - The usual set of small fixes and improvements for x86 and PPC"

* tag 'perf-core-2021-10-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/intel: Fix ICL/SPR INST_RETIRED.PREC_DIST encodings
  powerpc/perf: Fix data source encodings for L2.1 and L3.1 accesses
  tools/perf: Add mem_hops field in perf_mem_data_src structure
  perf: Add mem_hops field in perf_mem_data_src structure
  perf: Add comment about current state of PERF_MEM_LVL_* namespace and remove an extra line
  perf/core: Allow ftrace for functions in kernel/event/core.c
  perf/x86: Add new event for AUX output counter index
  perf/x86: Add compiler barrier after updating BTS
  perf/x86/intel/uncore: Fix Intel SPR M3UPI event constraints
  perf/x86/intel/uncore: Fix Intel SPR M2PCIE event constraints
  perf/x86/intel/uncore: Fix Intel SPR IIO event constraints
  perf/x86/intel/uncore: Fix Intel SPR CHA event constraints
  perf/x86/intel/uncore: Fix Intel ICX IIO event constraints
  perf/x86/intel/uncore: Fix invalid unit check
  perf/x86/intel/uncore: Support extra IMC channel on Ice Lake server

7 files changed, 50 insertions, 7 deletions

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index 6dfa8ddaa60f..38b2c779146f 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -66,6 +66,8 @@ DEFINE_STATIC_CALL_NULL(x86_pmu_enable_all,  *x86_pmu.enable_all);
 DEFINE_STATIC_CALL_NULL(x86_pmu_enable,	     *x86_pmu.enable);
 DEFINE_STATIC_CALL_NULL(x86_pmu_disable,     *x86_pmu.disable);
 
+DEFINE_STATIC_CALL_NULL(x86_pmu_assign, *x86_pmu.assign);
+
 DEFINE_STATIC_CALL_NULL(x86_pmu_add,  *x86_pmu.add);
 DEFINE_STATIC_CALL_NULL(x86_pmu_del,  *x86_pmu.del);
 DEFINE_STATIC_CALL_NULL(x86_pmu_read, *x86_pmu.read);
@@ -1215,6 +1217,8 @@ static inline void x86_assign_hw_event(struct perf_event *event,
 	hwc->last_cpu = smp_processor_id();
 	hwc->last_tag = ++cpuc->tags[i];
 
+	static_call_cond(x86_pmu_assign)(event, idx);
+
 	switch (hwc->idx) {
 	case INTEL_PMC_IDX_FIXED_BTS:
 	case INTEL_PMC_IDX_FIXED_VLBR:
@@ -2005,6 +2009,8 @@ static void x86_pmu_static_call_update(void)
 	static_call_update(x86_pmu_enable, x86_pmu.enable);
 	static_call_update(x86_pmu_disable, x86_pmu.disable);
 
+	static_call_update(x86_pmu_assign, x86_pmu.assign);
+
 	static_call_update(x86_pmu_add, x86_pmu.add);
 	static_call_update(x86_pmu_del, x86_pmu.del);
 	static_call_update(x86_pmu_read, x86_pmu.read);
diff --git a/arch/x86/events/intel/bts.c b/arch/x86/events/intel/bts.c
index 6320d2cfd9d3..974e917e65b2 100644
--- a/arch/x86/events/intel/bts.c
+++ b/arch/x86/events/intel/bts.c
@@ -209,6 +209,12 @@ static void bts_update(struct bts_ctx *bts)
 	} else {
 		local_set(&buf->data_size, head);
 	}
+
+	/*
+	 * Since BTS is coherent, just add compiler barrier to ensure
+	 * BTS updating is ordered against bts::handle::event.
+	 */
+	barrier();
 }
 
 static int
diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index 9a044438072b..aa3f5b527dc0 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -243,7 +243,8 @@ static struct extra_reg intel_skl_extra_regs[] __read_mostly = {
 
 static struct event_constraint intel_icl_event_constraints[] = {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
-	FIXED_EVENT_CONSTRAINT(0x01c0, 0),	/* INST_RETIRED.PREC_DIST */
+	FIXED_EVENT_CONSTRAINT(0x01c0, 0),	/* old INST_RETIRED.PREC_DIST */
+	FIXED_EVENT_CONSTRAINT(0x0100, 0),	/* INST_RETIRED.PREC_DIST */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
 	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
 	FIXED_EVENT_CONSTRAINT(0x0400, 3),	/* SLOTS */
@@ -288,7 +289,7 @@ static struct extra_reg intel_spr_extra_regs[] __read_mostly = {
 
 static struct event_constraint intel_spr_event_constraints[] = {
 	FIXED_EVENT_CONSTRAINT(0x00c0, 0),	/* INST_RETIRED.ANY */
-	FIXED_EVENT_CONSTRAINT(0x01c0, 0),	/* INST_RETIRED.PREC_DIST */
+	FIXED_EVENT_CONSTRAINT(0x0100, 0),	/* INST_RETIRED.PREC_DIST */
 	FIXED_EVENT_CONSTRAINT(0x003c, 1),	/* CPU_CLK_UNHALTED.CORE */
 	FIXED_EVENT_CONSTRAINT(0x0300, 2),	/* CPU_CLK_UNHALTED.REF */
 	FIXED_EVENT_CONSTRAINT(0x0400, 3),	/* SLOTS */
@@ -2403,6 +2404,12 @@ static void intel_pmu_disable_event(struct perf_event *event)
 		intel_pmu_pebs_disable(event);
 }
 
+static void intel_pmu_assign_event(struct perf_event *event, int idx)
+{
+	if (is_pebs_pt(event))
+		perf_report_aux_output_id(event, idx);
+}
+
 static void intel_pmu_del_event(struct perf_event *event)
 {
 	if (needs_branch_stack(event))
@@ -4495,8 +4502,16 @@ static int intel_pmu_check_period(struct perf_event *event, u64 value)
 	return intel_pmu_has_bts_period(event, value) ? -EINVAL : 0;
 }
 
+static void intel_aux_output_init(void)
+{
+	/* Refer also intel_pmu_aux_output_match() */
+	if (x86_pmu.intel_cap.pebs_output_pt_available)
+		x86_pmu.assign = intel_pmu_assign_event;
+}
+
 static int intel_pmu_aux_output_match(struct perf_event *event)
 {
+	/* intel_pmu_assign_event() is needed, refer intel_aux_output_init() */
 	if (!x86_pmu.intel_cap.pebs_output_pt_available)
 		return 0;
 
@@ -6302,6 +6317,8 @@ __init int intel_pmu_init(void)
 	if (is_hybrid())
 		intel_pmu_check_hybrid_pmus((u64)fixed_mask);
 
+	intel_aux_output_init();
+
 	return 0;
 }
 
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 8647713276a7..4dbb55a43dad 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -923,7 +923,8 @@ struct event_constraint intel_skl_pebs_event_constraints[] = {
 };
 
 struct event_constraint intel_icl_pebs_event_constraints[] = {
-	INTEL_FLAGS_UEVENT_CONSTRAINT(0x1c0, 0x100000000ULL),	/* INST_RETIRED.PREC_DIST */
+	INTEL_FLAGS_UEVENT_CONSTRAINT(0x01c0, 0x100000000ULL),	/* old INST_RETIRED.PREC_DIST */
+	INTEL_FLAGS_UEVENT_CONSTRAINT(0x0100, 0x100000000ULL),	/* INST_RETIRED.PREC_DIST */
 	INTEL_FLAGS_UEVENT_CONSTRAINT(0x0400, 0x800000000ULL),	/* SLOTS */
 
 	INTEL_PLD_CONSTRAINT(0x1cd, 0xff),			/* MEM_TRANS_RETIRED.LOAD_LATENCY */
@@ -943,7 +944,7 @@ struct event_constraint intel_icl_pebs_event_constraints[] = {
 };
 
 struct event_constraint intel_spr_pebs_event_constraints[] = {
-	INTEL_FLAGS_UEVENT_CONSTRAINT(0x1c0, 0x100000000ULL),
+	INTEL_FLAGS_UEVENT_CONSTRAINT(0x100, 0x100000000ULL),	/* INST_RETIRED.PREC_DIST */
 	INTEL_FLAGS_UEVENT_CONSTRAINT(0x0400, 0x800000000ULL),
 
 	INTEL_FLAGS_EVENT_CONSTRAINT(0xc0, 0xfe),
diff --git a/arch/x86/events/intel/uncore_discovery.h b/arch/x86/events/intel/uncore_discovery.h
index 7280c8a3c831..6d735611c281 100644
--- a/arch/x86/events/intel/uncore_discovery.h
+++ b/arch/x86/events/intel/uncore_discovery.h
@@ -30,7 +30,7 @@
 
 
 #define uncore_discovery_invalid_unit(unit)			\
-	(!unit.table1 || !unit.ctl || !unit.table3 ||	\
+	(!unit.table1 || !unit.ctl || \
 	 unit.table1 == -1ULL || unit.ctl == -1ULL ||	\
 	 unit.table3 == -1ULL)
 
diff --git a/arch/x86/events/intel/uncore_snbep.c b/arch/x86/events/intel/uncore_snbep.c
index 5ddc0f30db6f..eb2c6cea9d0d 100644
--- a/arch/x86/events/intel/uncore_snbep.c
+++ b/arch/x86/events/intel/uncore_snbep.c
@@ -452,7 +452,7 @@
 #define ICX_M3UPI_PCI_PMON_BOX_CTL		0xa0
 
 /* ICX IMC */
-#define ICX_NUMBER_IMC_CHN			2
+#define ICX_NUMBER_IMC_CHN			3
 #define ICX_IMC_MEM_STRIDE			0x4
 
 /* SPR */
@@ -5076,8 +5076,10 @@ static struct event_constraint icx_uncore_iio_constraints[] = {
 	UNCORE_EVENT_CONSTRAINT(0x02, 0x3),
 	UNCORE_EVENT_CONSTRAINT(0x03, 0x3),
 	UNCORE_EVENT_CONSTRAINT(0x83, 0x3),
+	UNCORE_EVENT_CONSTRAINT(0x88, 0xc),
 	UNCORE_EVENT_CONSTRAINT(0xc0, 0xc),
 	UNCORE_EVENT_CONSTRAINT(0xc5, 0xc),
+	UNCORE_EVENT_CONSTRAINT(0xd5, 0xc),
 	EVENT_CONSTRAINT_END
 };
 
@@ -5463,7 +5465,7 @@ static struct intel_uncore_ops icx_uncore_mmio_ops = {
 static struct intel_uncore_type icx_uncore_imc = {
 	.name		= "imc",
 	.num_counters   = 4,
-	.num_boxes	= 8,
+	.num_boxes	= 12,
 	.perf_ctr_bits	= 48,
 	.fixed_ctr_bits	= 48,
 	.fixed_ctr	= SNR_IMC_MMIO_PMON_FIXED_CTR,
@@ -5647,6 +5649,7 @@ static struct intel_uncore_type spr_uncore_chabox = {
 	.event_mask		= SPR_CHA_PMON_EVENT_MASK,
 	.event_mask_ext		= SPR_RAW_EVENT_MASK_EXT,
 	.num_shared_regs	= 1,
+	.constraints		= skx_uncore_chabox_constraints,
 	.ops			= &spr_uncore_chabox_ops,
 	.format_group		= &spr_uncore_chabox_format_group,
 	.attr_update		= uncore_alias_groups,
@@ -5658,6 +5661,7 @@ static struct intel_uncore_type spr_uncore_iio = {
 	.event_mask_ext		= SNR_IIO_PMON_RAW_EVENT_MASK_EXT,
 	.format_group		= &snr_uncore_iio_format_group,
 	.attr_update		= uncore_alias_groups,
+	.constraints		= icx_uncore_iio_constraints,
 };
 
 static struct attribute *spr_uncore_raw_formats_attr[] = {
@@ -5686,9 +5690,16 @@ static struct intel_uncore_type spr_uncore_irp = {
 
 };
 
+static struct event_constraint spr_uncore_m2pcie_constraints[] = {
+	UNCORE_EVENT_CONSTRAINT(0x14, 0x3),
+	UNCORE_EVENT_CONSTRAINT(0x2d, 0x3),
+	EVENT_CONSTRAINT_END
+};
+
 static struct intel_uncore_type spr_uncore_m2pcie = {
 	SPR_UNCORE_COMMON_FORMAT(),
 	.name			= "m2pcie",
+	.constraints		= spr_uncore_m2pcie_constraints,
 };
 
 static struct intel_uncore_type spr_uncore_pcu = {
@@ -5765,6 +5776,7 @@ static struct intel_uncore_type spr_uncore_upi = {
 static struct intel_uncore_type spr_uncore_m3upi = {
 	SPR_UNCORE_PCI_COMMON_FORMAT(),
 	.name			= "m3upi",
+	.constraints		= icx_uncore_m3upi_constraints,
 };
 
 static struct intel_uncore_type spr_uncore_mdf = {
diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h
index e3ac05c97b5e..76436a55d9ba 100644
--- a/arch/x86/events/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -726,6 +726,7 @@ struct x86_pmu {
 	void		(*enable_all)(int added);
 	void		(*enable)(struct perf_event *);
 	void		(*disable)(struct perf_event *);
+	void		(*assign)(struct perf_event *event, int idx);
 	void		(*add)(struct perf_event *);
 	void		(*del)(struct perf_event *);
 	void		(*read)(struct perf_event *event);