diff --git a/arch/x86/events/amd/ibs.c b/arch/x86/events/amd/ibs.c
index 855503e9e1fd110bec741e95704d3ec378ed9aef..405f30cf62da4332e6b620116741ded41caacba2 100644
--- a/arch/x86/events/amd/ibs.c
+++ b/arch/x86/events/amd/ibs.c
@@ -266,6 +266,14 @@ static int validate_group(struct perf_event *event)
 	return 0;
 }
 
+static bool perf_ibs_ldlat_event(struct perf_ibs *perf_ibs,
+				 struct perf_event *event)
+{
+	return perf_ibs == &perf_ibs_op &&
+	       (ibs_caps & IBS_CAPS_OPLDLAT) &&
+	       (event->attr.config1 & 0xFFF);
+}
+
 static int perf_ibs_init(struct perf_event *event)
 {
 	struct hw_perf_event *hwc = &event->hw;
@@ -324,6 +332,17 @@ static int perf_ibs_init(struct perf_event *event)
 			return -EINVAL;
 	}
 
+	if (perf_ibs_ldlat_event(perf_ibs, event)) {
+		u64 ldlat = event->attr.config1 & 0xFFF;
+
+		if (ldlat < 128 || ldlat > 2048)
+			return -EINVAL;
+		ldlat >>= 7;
+
+		config |= (ldlat - 1) << 59;
+		config |= IBS_OP_L3MISSONLY | IBS_OP_LDLAT_EN;
+	}
+
 	/*
 	 * If we modify hwc->sample_period, we also need to update
 	 * hwc->last_period and hwc->period_left.
@@ -602,7 +621,10 @@ PMU_FORMAT_ATTR(rand_en,	"config:57");
 PMU_FORMAT_ATTR(cnt_ctl,	"config:19");
 PMU_EVENT_ATTR_STRING(l3missonly, fetch_l3missonly, "config:59");
 PMU_EVENT_ATTR_STRING(l3missonly, op_l3missonly, "config:16");
+PMU_EVENT_ATTR_STRING(ldlat, ibs_op_ldlat_format, "config1:0-11");
 PMU_EVENT_ATTR_STRING(zen4_ibs_extensions, zen4_ibs_extensions, "1");
+PMU_EVENT_ATTR_STRING(ldlat, ibs_op_ldlat_cap, "1");
+PMU_EVENT_ATTR_STRING(dtlb_pgsize, ibs_op_dtlb_pgsize_cap, "1");
 
 static umode_t
 zen4_ibs_extensions_is_visible(struct kobject *kobj, struct attribute *attr, int i)
@@ -610,6 +632,18 @@ zen4_ibs_extensions_is_visible(struct kobject *kobj, struct attribute *attr, int
 	return ibs_caps & IBS_CAPS_ZEN4 ? attr->mode : 0;
 }
 
+static umode_t
+ibs_op_ldlat_is_visible(struct kobject *kobj, struct attribute *attr, int i)
+{
+	return ibs_caps & IBS_CAPS_OPLDLAT ? attr->mode : 0;
+}
+
+static umode_t
+ibs_op_dtlb_pgsize_is_visible(struct kobject *kobj, struct attribute *attr, int i)
+{
+	return ibs_caps & IBS_CAPS_OPDTLBPGSIZE ? attr->mode : 0;
+}
+
 static struct attribute *rand_en_attrs[] = {
 	&format_attr_rand_en.attr,
 	NULL,
@@ -625,6 +659,16 @@ static struct attribute *zen4_ibs_extensions_attrs[] = {
 	NULL,
 };
 
+static struct attribute *ibs_op_ldlat_cap_attrs[] = {
+	&ibs_op_ldlat_cap.attr.attr,
+	NULL,
+};
+
+static struct attribute *ibs_op_dtlb_pgsize_cap_attrs[] = {
+	&ibs_op_dtlb_pgsize_cap.attr.attr,
+	NULL,
+};
+
 static struct attribute_group group_rand_en = {
 	.name = "format",
 	.attrs = rand_en_attrs,
@@ -642,6 +686,18 @@ static struct attribute_group group_zen4_ibs_extensions = {
 	.is_visible = zen4_ibs_extensions_is_visible,
 };
 
+static struct attribute_group group_ibs_op_ldlat_cap = {
+	.name = "caps",
+	.attrs = ibs_op_ldlat_cap_attrs,
+	.is_visible = ibs_op_ldlat_is_visible,
+};
+
+static struct attribute_group group_ibs_op_dtlb_pgsize_cap = {
+	.name = "caps",
+	.attrs = ibs_op_dtlb_pgsize_cap_attrs,
+	.is_visible = ibs_op_dtlb_pgsize_is_visible,
+};
+
 static const struct attribute_group *fetch_attr_groups[] = {
 	&group_rand_en,
 	&empty_caps_group,
@@ -670,6 +726,11 @@ static struct attribute *op_l3missonly_attrs[] = {
 	NULL,
 };
 
+static struct attribute *ibs_op_ldlat_format_attrs[] = {
+	&ibs_op_ldlat_format.attr.attr,
+	NULL,
+};
+
 static struct attribute_group group_cnt_ctl = {
 	.name = "format",
 	.attrs = cnt_ctl_attrs,
@@ -682,10 +743,19 @@ static struct attribute_group group_op_l3missonly = {
 	.is_visible = zen4_ibs_extensions_is_visible,
 };
 
+static struct attribute_group group_ibs_op_ldlat_format = {
+	.name = "format",
+	.attrs = ibs_op_ldlat_format_attrs,
+	.is_visible = ibs_op_ldlat_is_visible,
+};
+
 static const struct attribute_group *op_attr_update[] = {
 	&group_cnt_ctl,
 	&group_op_l3missonly,
 	&group_zen4_ibs_extensions,
+	&group_ibs_op_ldlat_cap,
+	&group_ibs_op_ldlat_format,
+	&group_ibs_op_dtlb_pgsize_cap,
 	NULL,
 };
 
@@ -936,6 +1006,10 @@ static void perf_ibs_get_tlb_lvl(union ibs_op_data3 *op_data3,
 	if (!op_data3->dc_lin_addr_valid)
 		return;
 
+	if ((ibs_caps & IBS_CAPS_OPDTLBPGSIZE) &&
+	    !op_data3->dc_phy_addr_valid)
+		return;
+
 	if (!op_data3->dc_l1tlb_miss) {
 		data_src->mem_dtlb = PERF_MEM_TLB_L1 | PERF_MEM_TLB_HIT;
 		return;
@@ -1040,15 +1114,25 @@ static void perf_ibs_parse_ld_st_data(__u64 sample_type,
 	}
 }
 
-static int perf_ibs_get_offset_max(struct perf_ibs *perf_ibs, u64 sample_type,
+static bool perf_ibs_is_mem_sample_type(struct perf_ibs *perf_ibs,
+					struct perf_event *event)
+{
+	u64 sample_type = event->attr.sample_type;
+
+	return perf_ibs == &perf_ibs_op &&
+	       sample_type & (PERF_SAMPLE_DATA_SRC |
+			      PERF_SAMPLE_WEIGHT_TYPE |
+			      PERF_SAMPLE_ADDR |
+			      PERF_SAMPLE_PHYS_ADDR);
+}
+
+static int perf_ibs_get_offset_max(struct perf_ibs *perf_ibs,
+				   struct perf_event *event,
 				   int check_rip)
 {
-	if (sample_type & PERF_SAMPLE_RAW ||
-	    (perf_ibs == &perf_ibs_op &&
-	     (sample_type & PERF_SAMPLE_DATA_SRC ||
-	      sample_type & PERF_SAMPLE_WEIGHT_TYPE ||
-	      sample_type & PERF_SAMPLE_ADDR ||
-	      sample_type & PERF_SAMPLE_PHYS_ADDR)))
+	if (event->attr.sample_type & PERF_SAMPLE_RAW ||
+	    perf_ibs_is_mem_sample_type(perf_ibs, event) ||
+	    perf_ibs_ldlat_event(perf_ibs, event))
 		return perf_ibs->offset_max;
 	else if (check_rip)
 		return 3;
@@ -1103,7 +1187,7 @@ static int perf_ibs_handle_irq(struct perf_ibs *perf_ibs, struct pt_regs *iregs)
 	offset = 1;
 	check_rip = (perf_ibs == &perf_ibs_op && (ibs_caps & IBS_CAPS_RIPINVALIDCHK));
 
-	offset_max = perf_ibs_get_offset_max(perf_ibs, event->attr.sample_type, check_rip);
+	offset_max = perf_ibs_get_offset_max(perf_ibs, event, check_rip);
 
 	do {
 		rdmsrl(msr + offset, *buf++);
@@ -1112,6 +1196,22 @@ static int perf_ibs_handle_irq(struct perf_ibs *perf_ibs, struct pt_regs *iregs)
 				       perf_ibs->offset_max,
 				       offset + 1);
 	} while (offset < offset_max);
+
+	if (perf_ibs_ldlat_event(perf_ibs, event)) {
+		union ibs_op_data3 op_data3;
+
+		op_data3.val = ibs_data.regs[ibs_op_msr_idx(MSR_AMD64_IBSOPDATA3)];
+		/*
+		 * Opening event is errored out if load latency threshold is
+		 * outside of [128, 2048] range. Since the event has reached
+		 * interrupt handler, we can safely assume the threshold is
+		 * within [128, 2048] range.
+		 */
+		if (!op_data3.ld_op || !op_data3.dc_miss ||
+		    op_data3.dc_miss_lat <= (event->attr.config1 & 0xFFF))
+			goto out;
+	}
+
 	/*
 	 * Read IbsBrTarget, IbsOpData4, and IbsExtdCtl separately
 	 * depending on their availability.
diff --git a/arch/x86/include/asm/amd-ibs.h b/arch/x86/include/asm/amd-ibs.h
index cb2a5e113daa75d50ae8dd1ed338bd1c77db9273..77f3a589a99ac58fdd0e41291adcf1e48a44e2c2 100644
--- a/arch/x86/include/asm/amd-ibs.h
+++ b/arch/x86/include/asm/amd-ibs.h
@@ -64,7 +64,8 @@ union ibs_op_ctl {
 			opmaxcnt_ext:7,	/* 20-26: upper 7 bits of periodic op maximum count */
 			reserved0:5,	/* 27-31: reserved */
 			opcurcnt:27,	/* 32-58: periodic op counter current count */
-			reserved1:5;	/* 59-63: reserved */
+			ldlat_thrsh:4,	/* 59-62: Load Latency threshold */
+			ldlat_en:1;	/* 63: Load Latency enabled */
 	};
 };
 
diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h
index 338b8a7042d26691726eaae3bfd75d497850c4a8..e671cce1b4179491a49e7bd163803966af87709a 100644
--- a/arch/x86/include/asm/perf_event.h
+++ b/arch/x86/include/asm/perf_event.h
@@ -540,6 +540,8 @@ struct pebs_cntr_header {
 #define IBS_CAPS_FETCHCTLEXTD		(1U<<9)
 #define IBS_CAPS_OPDATA4		(1U<<10)
 #define IBS_CAPS_ZEN4			(1U<<11)
+#define IBS_CAPS_OPLDLAT		(1U<<12)
+#define IBS_CAPS_OPDTLBPGSIZE		(1U<<19)
 
 #define IBS_CAPS_DEFAULT		(IBS_CAPS_AVAIL		\
 					 | IBS_CAPS_FETCHSAM	\
@@ -565,6 +567,8 @@ struct pebs_cntr_header {
  * The lower 7 bits of the current count are random bits
  * preloaded by hardware and ignored in software
  */
+#define IBS_OP_LDLAT_EN		(1ULL<<63)
+#define IBS_OP_LDLAT_THRSH	(0xFULL<<59)
 #define IBS_OP_CUR_CNT		(0xFFF80ULL<<32)
 #define IBS_OP_CUR_CNT_RAND	(0x0007FULL<<32)
 #define IBS_OP_CUR_CNT_EXT_MASK	(0x7FULL<<52)
diff --git a/tools/arch/x86/include/asm/amd-ibs.h b/tools/arch/x86/include/asm/amd-ibs.h
index 93807b437e4deb85b2d96e61df2bfd0b0262e711..cb1740bc3da2df521a2282e900e7bc127dbe550d 100644
--- a/tools/arch/x86/include/asm/amd-ibs.h
+++ b/tools/arch/x86/include/asm/amd-ibs.h
@@ -64,7 +64,8 @@ union ibs_op_ctl {
 			opmaxcnt_ext:7,	/* 20-26: upper 7 bits of periodic op maximum count */
 			reserved0:5,	/* 27-31: reserved */
 			opcurcnt:27,	/* 32-58: periodic op counter current count */
-			reserved1:5;	/* 59-63: reserved */
+			ldlat_thrsh:4,	/* 59-62: Load Latency threshold */
+			ldlat_en:1;	/* 63: Load Latency enabled */
 	};
 };
 
diff --git a/tools/perf/Documentation/perf-amd-ibs.txt b/tools/perf/Documentation/perf-amd-ibs.txt
new file mode 100644
index 0000000000000000000000000000000000000000..55f80beae0375a72e58fd4bdb571b29b018ac45a
--- /dev/null
+++ b/tools/perf/Documentation/perf-amd-ibs.txt
@@ -0,0 +1,198 @@
+perf-amd-ibs(1)
+===============
+
+NAME
+----
+perf-amd-ibs - Support for AMD Instruction-Based Sampling (IBS) with perf tool
+
+SYNOPSIS
+--------
+[verse]
+'perf record' -e ibs_op//
+'perf record' -e ibs_fetch//
+
+DESCRIPTION
+-----------
+
+Instruction-Based Sampling (IBS) provides precise Instruction Pointer (IP)
+profiling support on AMD platforms. IBS has two independent components: IBS
+Op and IBS Fetch. IBS Op sampling provides information about instruction
+execution (micro-op execution to be precise) with details like d-cache
+hit/miss, d-TLB hit/miss, cache miss latency, load/store data source, branch
+behavior etc. IBS Fetch sampling provides information about instruction fetch
+with details like i-cache hit/miss, i-TLB hit/miss, fetch latency etc. IBS is
+per-smt-thread i.e. each SMT hardware thread contains standalone IBS units.
+
+Both, IBS Op and IBS Fetch, are exposed as PMUs by Linux and can be exploited
+using the Linux perf utility. The following files will be created at boot time
+if IBS is supported by the hardware and kernel.
+
+  /sys/bus/event_source/devices/ibs_op/
+  /sys/bus/event_source/devices/ibs_fetch/
+
+IBS Op PMU supports two events: cycles and micro ops. IBS Fetch PMU supports
+one event: fetch ops.
+
+IBS PMUs do not have user/kernel filtering capability and thus it requires
+CAP_SYS_ADMIN or CAP_PERFMON privilege.
+
+IBS VS. REGULAR CORE PMU
+------------------------
+
+IBS gives samples with precise IP, i.e. the IP recorded with IBS sample has
+no skid. Whereas the IP recorded by regular core PMU will have some skid
+(sample was generated at IP X but perf would record it at IP X+n). Hence,
+regular core PMU might not help for profiling with instruction level
+precision. Further, IBS provides additional information about the sample in
+question. On the other hand, regular core PMU has it's own advantages like
+plethora of events, counting mode (less interference), up to 6 parallel
+counters, event grouping support, filtering capabilities etc.
+
+Three regular core PMU events are internally forwarded to IBS Op PMU when
+precise_ip attribute is set:
+
+	-e cpu-cycles:p becomes -e ibs_op//
+	-e r076:p becomes -e ibs_op//
+	-e r0C1:p becomes -e ibs_op/cnt_ctl=1/
+
+EXAMPLES
+--------
+
+IBS Op PMU
+~~~~~~~~~~
+
+System-wide profile, cycles event, sampling period: 100000
+
+	# perf record -e ibs_op// -c 100000 -a
+
+Per-cpu profile (cpu10), cycles event, sampling period: 100000
+
+	# perf record -e ibs_op// -c 100000 -C 10
+
+Per-cpu profile (cpu10), cycles event, sampling freq: 1000
+
+	# perf record -e ibs_op// -F 1000 -C 10
+
+System-wide profile, uOps event, sampling period: 100000
+
+	# perf record -e ibs_op/cnt_ctl=1/ -c 100000 -a
+
+Same command, but also capture IBS register raw dump along with perf sample:
+
+	# perf record -e ibs_op/cnt_ctl=1/ -c 100000 -a --raw-samples
+
+System-wide profile, uOps event, sampling period: 100000, L3MissOnly (Zen4 onward)
+
+	# perf record -e ibs_op/cnt_ctl=1,l3missonly=1/ -c 100000 -a
+
+System-wide profile, cycles event, sampling period: 100000, LdLat filtering (Zen5
+onward)
+
+	# perf record -e ibs_op/ldlat=128/ -c 100000 -a
+
+	Supported load latency threshold values are 128 to 2048 (both inclusive).
+	Latency value which is a multiple of 128 incurs a little less profiling
+	overhead compared to other values.
+
+Per process(upstream v6.2 onward), uOps event, sampling period: 100000
+
+	# perf record -e ibs_op/cnt_ctl=1/ -c 100000 -p 1234
+
+Per process(upstream v6.2 onward), uOps event, sampling period: 100000
+
+	# perf record -e ibs_op/cnt_ctl=1/ -c 100000 -- ls
+
+To analyse recorded profile in aggregate mode
+
+	# perf report
+	/* Select a line and press 'a' to drill down at instruction level. */
+
+To go over each sample
+
+	# perf script
+
+Raw dump of IBS registers when profiled with --raw-samples
+
+	# perf report -D
+	/* Look for PERF_RECORD_SAMPLE */
+
+	Example register raw dump:
+
+	ibs_op_ctl:     000002c30006186a MaxCnt    100000 L3MissOnly 0 En 1
+		Val 1 CntCtl 0=cycles CurCnt       707
+	IbsOpRip:       ffffffff8204aea7
+	ibs_op_data:    0000010002550001 CompToRetCtr     1 TagToRetCtr   597
+		BrnRet 0  RipInvalid 0 BrnFuse 0 Microcode 1
+	ibs_op_data2:   0000000000000013 RmtNode 1 DataSrc 3=DRAM
+	ibs_op_data3:   0000000031960092 LdOp 0 StOp 1 DcL1TlbMiss 0
+		DcL2TlbMiss 0 DcL1TlbHit2M 1 DcL1TlbHit1G 0 DcL2TlbHit2M 0
+		DcMiss 1 DcMisAcc 0 DcWcMemAcc 0 DcUcMemAcc 0 DcLockedOp 0
+		DcMissNoMabAlloc 0 DcLinAddrValid 1 DcPhyAddrValid 1
+		DcL2TlbHit1G 0 L2Miss 1 SwPf 0 OpMemWidth 32 bytes
+		OpDcMissOpenMemReqs 12 DcMissLat     0 TlbRefillLat     0
+	IbsDCLinAd:     ff110008a5398920
+	IbsDCPhysAd:    00000008a5398920
+
+IBS applied in a real world usecase
+
+	~90% regression was observed in tbench with specific scheduler hint
+	which was counter intuitive. IBS profile of good and bad run captured
+	using perf helped in identifying exact cause of the problem:
+
+	https://lore.kernel.org/r/20220921063638.2489-1-kprateek.nayak@amd.com
+
+IBS Fetch PMU
+~~~~~~~~~~~~~
+
+Similar commands can be used with Fetch PMU as well.
+
+System-wide profile, fetch ops event, sampling period: 100000
+
+	# perf record -e ibs_fetch// -c 100000 -a
+
+System-wide profile, fetch ops event, sampling period: 100000, Random enable
+
+	# perf record -e ibs_fetch/rand_en=1/ -c 100000 -a
+
+	Random enable adds small degree of variability to sample period. This
+	helps in cases like long running loops where PMU is tagging the same
+	instruction over and over because of fixed sample period.
+
+etc.
+
+PERF MEM AND PERF C2C
+---------------------
+
+perf mem is a memory access profiler tool and perf c2c is a shared data
+cacheline analyser tool. Both of them internally uses IBS Op PMU on AMD.
+Below is a simple example of the perf mem tool.
+
+	# perf mem record -c 100000 -- make
+	# perf mem report
+
+A normal perf mem report output will provide detailed memory access profile.
+However, it can also be aggregated based on output fields. For example:
+
+	# perf mem report -F mem,sample,snoop
+	Samples: 3M of event 'ibs_op//', Event count (approx.): 23524876
+	Memory access                                 Samples  Snoop
+	N/A                                           1903343  N/A
+	L1 hit                                        1056754  N/A
+	L2 hit                                          75231  N/A
+	L3 hit                                           9496  HitM
+	L3 hit                                           2270  N/A
+	RAM hit                                          8710  N/A
+	Remote node, same socket RAM hit                 3241  N/A
+	Remote core, same node Any cache hit             1572  HitM
+	Remote core, same node Any cache hit              514  N/A
+	Remote node, same socket Any cache hit           1216  HitM
+	Remote node, same socket Any cache hit            350  N/A
+	Uncached hit                                       18  N/A
+
+Please refer to their man page for more detail.
+
+SEE ALSO
+--------
+
+linkperf:perf-record[1], linkperf:perf-script[1], linkperf:perf-report[1],
+linkperf:perf-mem[1], linkperf:perf-c2c[1]
diff --git a/tools/perf/Documentation/perf-c2c.txt b/tools/perf/Documentation/perf-c2c.txt
index 192ab0415ee9e855ff61c74f88d15632ee3364b2..c9c3dc18bb9d34c64fcc4ec48ce3f2ac6f1dbb85 100644
--- a/tools/perf/Documentation/perf-c2c.txt
+++ b/tools/perf/Documentation/perf-c2c.txt
@@ -54,8 +54,15 @@ RECORD OPTIONS
 
 -l::
 --ldlat::
-	Configure mem-loads latency. Supported on Intel and Arm64 processors
-	only. Ignored on other archs.
+	Configure mem-loads latency. Supported on Intel, Arm64 and some AMD
+	processors. Ignored on other archs.
+
+	On supported AMD processors:
+	- /sys/bus/event_source/devices/ibs_op/caps/ldlat file contains '1'.
+	- Supported latency values are 128 to 2048 (both inclusive).
+	- Latency value which is a multiple of 128 incurs a little less profiling
+	  overhead compared to other values.
+	- Load latency filtering is disabled by default.
 
 -k::
 --all-kernel::
diff --git a/tools/perf/Documentation/perf-mem.txt b/tools/perf/Documentation/perf-mem.txt
index 19862572e3f2b64dcfc40f0fbf175baadfc284b4..117b90b91e1f5c7809b8d13c59062fd5d2e24544 100644
--- a/tools/perf/Documentation/perf-mem.txt
+++ b/tools/perf/Documentation/perf-mem.txt
@@ -28,6 +28,8 @@ and kernel support is required. See linkperf:perf-arm-spe[1] for a setup guide.
 Due to the statistical nature of SPE sampling, not every memory operation will
 be sampled.
 
+On AMD this use IBS Op PMU to sample load-store operations.
+
 OPTIONS
 -------
 <command>...::
@@ -90,8 +92,15 @@ RECORD OPTIONS
 	Be more verbose (show counter open errors, etc)
 
 --ldlat <n>::
-	Specify desired latency for loads event. Supported on Intel and Arm64
-	processors only. Ignored on other archs.
+	Specify desired latency for loads event. Supported on Intel, Arm64 and
+	some AMD processors. Ignored on other archs.
+
+	On supported AMD processors:
+	- /sys/bus/event_source/devices/ibs_op/caps/ldlat file contains '1'.
+	- Supported latency values are 128 to 2048 (both inclusive).
+	- Latency value which is a multiple of 128 incurs a little less profiling
+	  overhead compared to other values.
+	- Load latency filtering is disabled by default.
 
 In addition, for report all perf report options are valid, and for record
 all perf record options.
diff --git a/tools/perf/Documentation/perf.txt b/tools/perf/Documentation/perf.txt
index ba3df49c169d329223f6d03fb6cab122e737b757..fa20f54e347bbe4d06bd79dd91d817586ba14b7c 100644
--- a/tools/perf/Documentation/perf.txt
+++ b/tools/perf/Documentation/perf.txt
@@ -77,7 +77,8 @@ linkperf:perf-stat[1], linkperf:perf-top[1],
 linkperf:perf-record[1], linkperf:perf-report[1],
 linkperf:perf-list[1]
 
-linkperf:perf-annotate[1],linkperf:perf-archive[1],linkperf:perf-arm-spe[1],
+linkperf:perf-amd-ibs[1], linkperf:perf-annotate[1],
+linkperf:perf-archive[1], linkperf:perf-arm-spe[1],
 linkperf:perf-bench[1], linkperf:perf-buildid-cache[1],
 linkperf:perf-buildid-list[1], linkperf:perf-c2c[1],
 linkperf:perf-config[1], linkperf:perf-data[1], linkperf:perf-diff[1],
diff --git a/tools/perf/arch/arm/util/pmu.c b/tools/perf/arch/arm/util/pmu.c
index 7f3af3b97f3bac2a35faddc2a0a0ad180198a618..8b7cb68ba1a8a24ec86593570756304aa3fdfc43 100644
--- a/tools/perf/arch/arm/util/pmu.c
+++ b/tools/perf/arch/arm/util/pmu.c
@@ -13,6 +13,7 @@
 #include "hisi-ptt.h"
 #include "../../../util/pmu.h"
 #include "../../../util/cs-etm.h"
+#include "../../arm64/util/mem-events.h"
 
 void perf_pmu__arch_init(struct perf_pmu *pmu __maybe_unused)
 {
@@ -26,6 +27,8 @@ void perf_pmu__arch_init(struct perf_pmu *pmu __maybe_unused)
 		pmu->selectable = true;
 		pmu->is_uncore = false;
 		pmu->perf_event_attr_init_default = arm_spe_pmu_default_config;
+		if (!strcmp(pmu->name, "arm_spe_0"))
+			pmu->mem_events = perf_mem_events_arm;
 	} else if (strstarts(pmu->name, HISI_PTT_PMU_NAME)) {
 		pmu->selectable = true;
 #endif
diff --git a/tools/perf/arch/arm64/util/mem-events.c b/tools/perf/arch/arm64/util/mem-events.c
index 3bcc5c7035c21429eb1f55ae8891ac642249f6e2..9f8da7937255cc9b14c3c58a1119b40bd0c76f6b 100644
--- a/tools/perf/arch/arm64/util/mem-events.c
+++ b/tools/perf/arch/arm64/util/mem-events.c
@@ -1,37 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "map_symbol.h"
+#include "util/map_symbol.h"
+#include "util/mem-events.h"
 #include "mem-events.h"
 
-#define E(t, n, s) { .tag = t, .name = n, .sysfs_name = s }
+#define E(t, n, s, l, a) { .tag = t, .name = n, .event_name = s, .ldlat = l, .aux_event = a }
 
-static struct perf_mem_event perf_mem_events[PERF_MEM_EVENTS__MAX] = {
-	E("spe-load",	"arm_spe_0/ts_enable=1,pa_enable=1,load_filter=1,store_filter=0,min_latency=%u/",	"arm_spe_0"),
-	E("spe-store",	"arm_spe_0/ts_enable=1,pa_enable=1,load_filter=0,store_filter=1/",			"arm_spe_0"),
-	E("spe-ldst",	"arm_spe_0/ts_enable=1,pa_enable=1,load_filter=1,store_filter=1,min_latency=%u/",	"arm_spe_0"),
+struct perf_mem_event perf_mem_events_arm[PERF_MEM_EVENTS__MAX] = {
+	E("spe-load",	"%s/ts_enable=1,pa_enable=1,load_filter=1,store_filter=0,min_latency=%u/",	NULL,	true,	0),
+	E("spe-store",	"%s/ts_enable=1,pa_enable=1,load_filter=0,store_filter=1/",			NULL,	false,	0),
+	E("spe-ldst",	"%s/ts_enable=1,pa_enable=1,load_filter=1,store_filter=1,min_latency=%u/",	NULL,	true,	0),
 };
-
-static char mem_ev_name[100];
-
-struct perf_mem_event *perf_mem_events__ptr(int i)
-{
-	if (i >= PERF_MEM_EVENTS__MAX)
-		return NULL;
-
-	return &perf_mem_events[i];
-}
-
-const char *perf_mem_events__name(int i, const char *pmu_name __maybe_unused)
-{
-	struct perf_mem_event *e = perf_mem_events__ptr(i);
-
-	if (i >= PERF_MEM_EVENTS__MAX)
-		return NULL;
-
-	if (i == PERF_MEM_EVENTS__LOAD || i == PERF_MEM_EVENTS__LOAD_STORE)
-		scnprintf(mem_ev_name, sizeof(mem_ev_name),
-			  e->name, perf_mem_events__loads_ldlat);
-	else /* PERF_MEM_EVENTS__STORE */
-		scnprintf(mem_ev_name, sizeof(mem_ev_name), e->name);
-
-	return mem_ev_name;
-}
diff --git a/tools/perf/arch/arm64/util/mem-events.h b/tools/perf/arch/arm64/util/mem-events.h
new file mode 100644
index 0000000000000000000000000000000000000000..5fc50be4be3802fd647c50524f107d0f59a750c9
--- /dev/null
+++ b/tools/perf/arch/arm64/util/mem-events.h
@@ -0,0 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _ARM64_MEM_EVENTS_H
+#define _ARM64_MEM_EVENTS_H
+
+extern struct perf_mem_event perf_mem_events_arm[PERF_MEM_EVENTS__MAX];
+
+#endif /* _ARM64_MEM_EVENTS_H */
diff --git a/tools/perf/arch/powerpc/util/Build b/tools/perf/arch/powerpc/util/Build
index 9889245c555c4cfb9730997d2ae5b1cdfbf1b3f9..1d323f3a3322bc394a22380c0cef41cd57905f6e 100644
--- a/tools/perf/arch/powerpc/util/Build
+++ b/tools/perf/arch/powerpc/util/Build
@@ -2,6 +2,7 @@ perf-y += header.o
 perf-$(CONFIG_LIBTRACEEVENT) += kvm-stat.o
 perf-y += perf_regs.o
 perf-y += mem-events.o
+perf-y += pmu.o
 perf-y += sym-handling.o
 perf-y += evsel.o
 perf-y += event.o
diff --git a/tools/perf/arch/powerpc/util/mem-events.c b/tools/perf/arch/powerpc/util/mem-events.c
index 78b986e5268d95ec7f80a9b2a91a0596041a4608..765d4a054b0a42362f34af60ede150e0eacdca28 100644
--- a/tools/perf/arch/powerpc/util/mem-events.c
+++ b/tools/perf/arch/powerpc/util/mem-events.c
@@ -1,12 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0
-#include "map_symbol.h"
+#include "util/map_symbol.h"
+#include "util/mem-events.h"
 #include "mem-events.h"
 
-/* PowerPC does not support 'ldlat' parameter. */
-const char *perf_mem_events__name(int i, const char *pmu_name __maybe_unused)
-{
-	if (i == PERF_MEM_EVENTS__LOAD)
-		return "cpu/mem-loads/";
+#define E(t, n, s, l, a) { .tag = t, .name = n, .event_name = s, .ldlat = l, .aux_event = a }
 
-	return "cpu/mem-stores/";
-}
+struct perf_mem_event perf_mem_events_power[PERF_MEM_EVENTS__MAX] = {
+	E("ldlat-loads",	"%s/mem-loads/",	"mem-loads",	false,	0),
+	E("ldlat-stores",	"%s/mem-stores/",	"mem-stores",	false,	0),
+	E(NULL,			NULL,			NULL,		false,	0),
+};
diff --git a/tools/perf/arch/powerpc/util/mem-events.h b/tools/perf/arch/powerpc/util/mem-events.h
new file mode 100644
index 0000000000000000000000000000000000000000..6acc3d1b687318669d9834aa9d08788301d28e93
--- /dev/null
+++ b/tools/perf/arch/powerpc/util/mem-events.h
@@ -0,0 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _POWER_MEM_EVENTS_H
+#define _POWER_MEM_EVENTS_H
+
+extern struct perf_mem_event perf_mem_events_power[PERF_MEM_EVENTS__MAX];
+
+#endif /* _POWER_MEM_EVENTS_H */
diff --git a/tools/perf/arch/powerpc/util/pmu.c b/tools/perf/arch/powerpc/util/pmu.c
new file mode 100644
index 0000000000000000000000000000000000000000..554675deef7ba2890da2d413369d56f3c695462a
--- /dev/null
+++ b/tools/perf/arch/powerpc/util/pmu.c
@@ -0,0 +1,12 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <string.h>
+
+#include "../../../util/pmu.h"
+#include "mem-events.h"
+
+void perf_pmu__arch_init(struct perf_pmu *pmu)
+{
+	if (pmu->is_core)
+		pmu->mem_events = perf_mem_events_power;
+}
diff --git a/tools/perf/arch/x86/util/mem-events.c b/tools/perf/arch/x86/util/mem-events.c
index f8d9aecbf2f28763b5361d48b03dba54022affa3..aee029421eaf534df147a211aad219ae91ab074a 100644
--- a/tools/perf/arch/x86/util/mem-events.c
+++ b/tools/perf/arch/x86/util/mem-events.c
@@ -7,38 +7,34 @@
 #include "linux/string.h"
 #include "env.h"
 
-static char mem_loads_name[100];
-static bool mem_loads_name__init;
-static char mem_stores_name[100];
-
 #define MEM_LOADS_AUX		0x8203
-#define MEM_LOADS_AUX_NAME     "{%s/mem-loads-aux/,%s/mem-loads,ldlat=%u/}:P"
 
-#define E(t, n, s) { .tag = t, .name = n, .sysfs_name = s }
+#define E(t, n, s, l, a) { .tag = t, .name = n, .event_name = s, .ldlat = l, .aux_event = a }
 
-static struct perf_mem_event perf_mem_events_intel[PERF_MEM_EVENTS__MAX] = {
-	E("ldlat-loads",	"%s/mem-loads,ldlat=%u/P",	"%s/events/mem-loads"),
-	E("ldlat-stores",	"%s/mem-stores/P",		"%s/events/mem-stores"),
-	E(NULL,			NULL,				NULL),
+struct perf_mem_event perf_mem_events_intel[PERF_MEM_EVENTS__MAX] = {
+	E("ldlat-loads",	"%s/mem-loads,ldlat=%u/P",	"mem-loads",	true,	0),
+	E("ldlat-stores",	"%s/mem-stores/P",		"mem-stores",	false,	0),
+	E(NULL,			NULL,				NULL,		false,	0),
 };
 
-static struct perf_mem_event perf_mem_events_amd[PERF_MEM_EVENTS__MAX] = {
-	E(NULL,		NULL,		NULL),
-	E(NULL,		NULL,		NULL),
-	E("mem-ldst",	"ibs_op//",	"ibs_op"),
+struct perf_mem_event perf_mem_events_intel_aux[PERF_MEM_EVENTS__MAX] = {
+	E("ldlat-loads",	"{%s/mem-loads-aux/,%s/mem-loads,ldlat=%u/}:P",	"mem-loads",	true,	MEM_LOADS_AUX),
+	E("ldlat-stores",	"%s/mem-stores/P",		"mem-stores",	false,	0),
+	E(NULL,			NULL,				NULL,		false,	0),
 };
 
-struct perf_mem_event *perf_mem_events__ptr(int i)
-{
-	if (i >= PERF_MEM_EVENTS__MAX)
-		return NULL;
-
-	if (x86__is_amd_cpu() || x86__is_hygon_cpu())
-		return &perf_mem_events_amd[i];
+struct perf_mem_event perf_mem_events_amd[PERF_MEM_EVENTS__MAX] = {
+	E(NULL,		NULL,		NULL,	false,	0),
+	E(NULL,		NULL,		NULL,	false,	0),
+	E("mem-ldst",	"%s//",		NULL,	false,	0),
+};
 
-	return &perf_mem_events_intel[i];
-}
 
+struct perf_mem_event perf_mem_events_amd_ldlat[PERF_MEM_EVENTS__MAX] = {
+	E(NULL,		NULL,		NULL,	false,	0),
+	E(NULL,		NULL,		NULL,	false,	0),
+	E("mem-ldst",	"%s/ldlat=%u/",	NULL,	true,	0),
+};
 bool is_mem_loads_aux_event(struct evsel *leader)
 {
 	struct perf_pmu *pmu = perf_pmus__find("cpu");
@@ -51,43 +47,3 @@ bool is_mem_loads_aux_event(struct evsel *leader)
 
 	return leader->core.attr.config == MEM_LOADS_AUX;
 }
-
-const char *perf_mem_events__name(int i, const char *pmu_name)
-{
-	struct perf_mem_event *e = perf_mem_events__ptr(i);
-
-	if (!e)
-		return NULL;
-
-	if (i == PERF_MEM_EVENTS__LOAD) {
-		if (mem_loads_name__init && !pmu_name)
-			return mem_loads_name;
-
-		if (!pmu_name) {
-			mem_loads_name__init = true;
-			pmu_name = "cpu";
-		}
-
-		if (perf_pmus__have_event(pmu_name, "mem-loads-aux")) {
-			scnprintf(mem_loads_name, sizeof(mem_loads_name),
-				  MEM_LOADS_AUX_NAME, pmu_name, pmu_name,
-				  perf_mem_events__loads_ldlat);
-		} else {
-			scnprintf(mem_loads_name, sizeof(mem_loads_name),
-				  e->name, pmu_name,
-				  perf_mem_events__loads_ldlat);
-		}
-		return mem_loads_name;
-	}
-
-	if (i == PERF_MEM_EVENTS__STORE) {
-		if (!pmu_name)
-			pmu_name = "cpu";
-
-		scnprintf(mem_stores_name, sizeof(mem_stores_name),
-			  e->name, pmu_name);
-		return mem_stores_name;
-	}
-
-	return e->name;
-}
diff --git a/tools/perf/arch/x86/util/mem-events.h b/tools/perf/arch/x86/util/mem-events.h
new file mode 100644
index 0000000000000000000000000000000000000000..11e09a256f5bb084b8c272f8830f512037616bc4
--- /dev/null
+++ b/tools/perf/arch/x86/util/mem-events.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _X86_MEM_EVENTS_H
+#define _X86_MEM_EVENTS_H
+
+extern struct perf_mem_event perf_mem_events_intel[PERF_MEM_EVENTS__MAX];
+extern struct perf_mem_event perf_mem_events_intel_aux[PERF_MEM_EVENTS__MAX];
+
+extern struct perf_mem_event perf_mem_events_amd[PERF_MEM_EVENTS__MAX];
+extern struct perf_mem_event perf_mem_events_amd_ldlat[PERF_MEM_EVENTS__MAX];
+
+#endif /* _X86_MEM_EVENTS_H */
diff --git a/tools/perf/arch/x86/util/pmu.c b/tools/perf/arch/x86/util/pmu.c
index bd36c1c885b67aa043b396e2fbf784b765bfc735..c5c7a93b635a1a93b4e021ac95b2ea58f3f66f8a 100644
--- a/tools/perf/arch/x86/util/pmu.c
+++ b/tools/perf/arch/x86/util/pmu.c
@@ -15,10 +15,13 @@
 #include "../../../util/pmu.h"
 #include "../../../util/fncache.h"
 #include "../../../util/pmus.h"
+#include "mem-events.h"
 #include "env.h"
 
-void perf_pmu__arch_init(struct perf_pmu *pmu __maybe_unused)
+void perf_pmu__arch_init(struct perf_pmu *pmu)
 {
+	struct perf_pmu_caps *ldlat_cap;
+
 #ifdef HAVE_AUXTRACE_SUPPORT
 	if (!strcmp(pmu->name, INTEL_PT_PMU_NAME)) {
 		pmu->auxtrace = true;
@@ -30,6 +33,28 @@ void perf_pmu__arch_init(struct perf_pmu *pmu __maybe_unused)
 		pmu->selectable = true;
 	}
 #endif
+
+	if (x86__is_amd_cpu()) {
+		if (strcmp(pmu->name, "ibs_op"))
+			return;
+
+		pmu->mem_events = perf_mem_events_amd;
+
+		if (!perf_pmu__caps_parse(pmu))
+			return;
+
+		ldlat_cap = perf_pmu__get_cap(pmu, "ldlat");
+		if (!ldlat_cap || strcmp(ldlat_cap->value, "1"))
+			return;
+
+		perf_mem_events__loads_ldlat = 0;
+		pmu->mem_events = perf_mem_events_amd_ldlat;
+	} else if (pmu->is_core) {
+		if (perf_pmu__have_event(pmu, "mem-loads-aux"))
+			pmu->mem_events = perf_mem_events_intel_aux;
+		else
+			pmu->mem_events = perf_mem_events_intel;
+	}
 }
 
 int perf_pmus__num_mem_pmus(void)
diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c
index ef7ed53a4b4e2dc5da83ff6a458266d2acb9089f..5e136b4786843fae1e531397a30dcbbbb2ac2ad5 100644
--- a/tools/perf/builtin-c2c.c
+++ b/tools/perf/builtin-c2c.c
@@ -3211,12 +3211,19 @@ static int parse_record_events(const struct option *opt,
 			       const char *str, int unset __maybe_unused)
 {
 	bool *event_set = (bool *) opt->value;
+	struct perf_pmu *pmu;
+
+	pmu = perf_mem_events_find_pmu();
+	if (!pmu) {
+		pr_err("failed: there is no PMU that supports perf c2c\n");
+		exit(-1);
+	}
 
 	if (!strcmp(str, "list")) {
-		perf_mem_events__list();
+		perf_pmu__mem_events_list(pmu);
 		exit(0);
 	}
-	if (perf_mem_events__parse(str))
+	if (perf_pmu__mem_events_parse(pmu, str))
 		exit(-1);
 
 	*event_set = true;
@@ -3241,6 +3248,7 @@ static int perf_c2c__record(int argc, const char **argv)
 	bool all_user = false, all_kernel = false;
 	bool event_set = false;
 	struct perf_mem_event *e;
+	struct perf_pmu *pmu;
 	struct option options[] = {
 	OPT_CALLBACK('e', "event", &event_set, "event",
 		     "event selector. Use 'perf c2c record -e list' to list available events",
@@ -3252,7 +3260,13 @@ static int perf_c2c__record(int argc, const char **argv)
 	OPT_END()
 	};
 
-	if (perf_mem_events__init()) {
+	pmu = perf_mem_events_find_pmu();
+	if (!pmu) {
+		pr_err("failed: no PMU supports the memory events\n");
+		return -1;
+	}
+
+	if (perf_pmu__mem_events_init(pmu)) {
 		pr_err("failed: memory events not supported\n");
 		return -1;
 	}
@@ -3276,7 +3290,7 @@ static int perf_c2c__record(int argc, const char **argv)
 	rec_argv[i++] = "record";
 
 	if (!event_set) {
-		e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD_STORE);
+		e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD_STORE);
 		/*
 		 * The load and store operations are required, use the event
 		 * PERF_MEM_EVENTS__LOAD_STORE if it is supported.
@@ -3285,15 +3299,15 @@ static int perf_c2c__record(int argc, const char **argv)
 			e->record = true;
 			rec_argv[i++] = "-W";
 		} else {
-			e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD);
+			e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD);
 			e->record = true;
 
-			e = perf_mem_events__ptr(PERF_MEM_EVENTS__STORE);
+			e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__STORE);
 			e->record = true;
 		}
 	}
 
-	e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD);
+	e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD);
 	if (e->record)
 		rec_argv[i++] = "-W";
 
diff --git a/tools/perf/builtin-mem.c b/tools/perf/builtin-mem.c
index 286105be91cec9516fce1657a95dc6c9cd044032..a65d3981ff57dc5ddfa9de3551a8270204fa9e60 100644
--- a/tools/perf/builtin-mem.c
+++ b/tools/perf/builtin-mem.c
@@ -43,12 +43,19 @@ static int parse_record_events(const struct option *opt,
 			       const char *str, int unset __maybe_unused)
 {
 	struct perf_mem *mem = *(struct perf_mem **)opt->value;
+	struct perf_pmu *pmu;
+
+	pmu = perf_mem_events_find_pmu();
+	if (!pmu) {
+		pr_err("failed: there is no PMU that supports perf mem\n");
+		exit(-1);
+	}
 
 	if (!strcmp(str, "list")) {
-		perf_mem_events__list();
+		perf_pmu__mem_events_list(pmu);
 		exit(0);
 	}
-	if (perf_mem_events__parse(str))
+	if (perf_pmu__mem_events_parse(pmu, str))
 		exit(-1);
 
 	mem->operation = 0;
@@ -72,6 +79,7 @@ static int __cmd_record(int argc, const char **argv, struct perf_mem *mem)
 	int ret;
 	bool all_user = false, all_kernel = false;
 	struct perf_mem_event *e;
+	struct perf_pmu *pmu;
 	struct option options[] = {
 	OPT_CALLBACK('e', "event", &mem, "event",
 		     "event selector. use 'perf mem record -e list' to list available events",
@@ -84,7 +92,13 @@ static int __cmd_record(int argc, const char **argv, struct perf_mem *mem)
 	OPT_END()
 	};
 
-	if (perf_mem_events__init()) {
+	pmu = perf_mem_events_find_pmu();
+	if (!pmu) {
+		pr_err("failed: no PMU supports the memory events\n");
+		return -1;
+	}
+
+	if (perf_pmu__mem_events_init(pmu)) {
 		pr_err("failed: memory events not supported\n");
 		return -1;
 	}
@@ -113,7 +127,7 @@ static int __cmd_record(int argc, const char **argv, struct perf_mem *mem)
 
 	rec_argv[i++] = "record";
 
-	e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD_STORE);
+	e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD_STORE);
 
 	/*
 	 * The load and store operations are required, use the event
@@ -126,17 +140,17 @@ static int __cmd_record(int argc, const char **argv, struct perf_mem *mem)
 		rec_argv[i++] = "-W";
 	} else {
 		if (mem->operation & MEM_OPERATION_LOAD) {
-			e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD);
+			e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD);
 			e->record = true;
 		}
 
 		if (mem->operation & MEM_OPERATION_STORE) {
-			e = perf_mem_events__ptr(PERF_MEM_EVENTS__STORE);
+			e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__STORE);
 			e->record = true;
 		}
 	}
 
-	e = perf_mem_events__ptr(PERF_MEM_EVENTS__LOAD);
+	e = perf_pmu__mem_events_ptr(pmu, PERF_MEM_EVENTS__LOAD);
 	if (e->record)
 		rec_argv[i++] = "-W";
 
diff --git a/tools/perf/tests/shell/test_data_symbol.sh b/tools/perf/tests/shell/test_data_symbol.sh
index 3dfa91832aa87f89b8f0ef0c1ba51df6d130d2d2..b7d5ddc6b41ba80ecf2760bfd90414113f71a74d 100755
--- a/tools/perf/tests/shell/test_data_symbol.sh
+++ b/tools/perf/tests/shell/test_data_symbol.sh
@@ -55,11 +55,34 @@ trap cleanup_files exit term int
 
 echo "Recording workload..."
 
-# perf mem/c2c internally uses IBS PMU on AMD CPU which doesn't support
-# user/kernel filtering and per-process monitoring, spin program on
-# specific CPU and test in per-CPU mode.
 is_amd=$(grep -E -c 'vendor_id.*AuthenticAMD' /proc/cpuinfo)
 if (($is_amd >= 1)); then
+	mem_events="$(perf mem record -v -e list 2>&1)"
+	if ! [[ "$mem_events" =~ ^mem\-ldst.*ibs_op/(.*)/.*available ]]; then
+		echo "ERROR: mem-ldst event is not matching"
+		exit 1
+	fi
+
+	# --ldlat on AMD:
+	# o Zen4 and earlier uarch does not support ldlat
+	# o Even on supported platforms, it's disabled (--ldlat=0) by default.
+	ldlat=${BASH_REMATCH[1]}
+	if [[ -n $ldlat ]]; then
+		if ! [[ "$ldlat" =~ ldlat=0 ]]; then
+			echo "ERROR: ldlat not initialized to 0?"
+			exit 1
+		fi
+
+		mem_events="$(perf mem record -v --ldlat=150 -e list 2>&1)"
+		if ! [[ "$mem_events" =~ ^mem-ldst.*ibs_op/ldlat=150/.*available ]]; then
+			echo "ERROR: --ldlat not honored?"
+			exit 1
+		fi
+	fi
+
+	# perf mem/c2c internally uses IBS PMU on AMD CPU which doesn't
+	# support user/kernel filtering and per-process monitoring on older
+	# kernels, spin program on specific CPU and test in per-CPU mode.
 	perf mem record -vvv -o ${PERF_DATA} -C 0 -- taskset -c 0 $TEST_PROGRAM 2>"${ERR_FILE}" &
 else
 	perf mem record -vvv --all-user -o ${PERF_DATA} -- $TEST_PROGRAM 2>"${ERR_FILE}" &
diff --git a/tools/perf/util/amd-sample-raw.c b/tools/perf/util/amd-sample-raw.c
index 9d0ce88e90e450c72804f0c5c3aacec5c5caa47d..022c9eb3950909226ea273a564137aa41460ed37 100644
--- a/tools/perf/util/amd-sample-raw.c
+++ b/tools/perf/util/amd-sample-raw.c
@@ -19,6 +19,8 @@
 
 static u32 cpu_family, cpu_model, ibs_fetch_type, ibs_op_type;
 static bool zen4_ibs_extensions;
+static bool ldlat_cap;
+static bool dtlb_pgsize_cap;
 
 static void pr_ibs_fetch_ctl(union ibs_fetch_ctl reg)
 {
@@ -78,14 +80,20 @@ static void pr_ic_ibs_extd_ctl(union ic_ibs_extd_ctl reg)
 static void pr_ibs_op_ctl(union ibs_op_ctl reg)
 {
 	char l3_miss_only[sizeof(" L3MissOnly _")] = "";
+	char ldlat[sizeof(" LdLatThrsh __ LdLatEn _")] = "";
 
 	if (zen4_ibs_extensions)
 		snprintf(l3_miss_only, sizeof(l3_miss_only), " L3MissOnly %d", reg.l3_miss_only);
 
-	printf("ibs_op_ctl:\t%016llx MaxCnt %9d%s En %d Val %d CntCtl %d=%s CurCnt %9d\n",
+	if (ldlat_cap) {
+		snprintf(ldlat, sizeof(ldlat), " LdLatThrsh %2d LdLatEn %d",
+			 reg.ldlat_thrsh, reg.ldlat_en);
+	}
+
+	printf("ibs_op_ctl:\t%016llx MaxCnt %9d%s En %d Val %d CntCtl %d=%s CurCnt %9d%s\n",
 		reg.val, ((reg.opmaxcnt_ext << 16) | reg.opmaxcnt) << 4, l3_miss_only,
 		reg.op_en, reg.op_val, reg.cnt_ctl,
-		reg.cnt_ctl ? "uOps" : "cycles", reg.opcurcnt);
+		reg.cnt_ctl ? "uOps" : "cycles", reg.opcurcnt, ldlat);
 }
 
 static void pr_ibs_op_data(union ibs_op_data reg)
@@ -154,9 +162,20 @@ static void pr_ibs_op_data2(union ibs_op_data2 reg)
 
 static void pr_ibs_op_data3(union ibs_op_data3 reg)
 {
-	char l2_miss_str[sizeof(" L2Miss _")] = "";
-	char op_mem_width_str[sizeof(" OpMemWidth _____ bytes")] = "";
+	static const char * const dc_page_sizes[] = {
+		"  4K",
+		"  2M",
+		"  1G",
+		"  ??",
+	};
 	char op_dc_miss_open_mem_reqs_str[sizeof(" OpDcMissOpenMemReqs __")] = "";
+	char dc_l1_l2tlb_miss_str[sizeof(" DcL1TlbMiss _ DcL2TlbMiss _")] = "";
+	char dc_l1tlb_hit_str[sizeof(" DcL1TlbHit2M _ DcL1TlbHit1G _")] = "";
+	char op_mem_width_str[sizeof(" OpMemWidth _____ bytes")] = "";
+	char dc_l2tlb_hit_2m_str[sizeof(" DcL2TlbHit2M _")] = "";
+	char dc_l2tlb_hit_1g_str[sizeof(" DcL2TlbHit1G _")] = "";
+	char dc_page_size_str[sizeof(" DcPageSize ____")] = "";
+	char l2_miss_str[sizeof(" L2Miss _")] = "";
 
 	/*
 	 * Erratum #1293
@@ -172,16 +191,40 @@ static void pr_ibs_op_data3(union ibs_op_data3 reg)
 		snprintf(op_mem_width_str, sizeof(op_mem_width_str),
 			 " OpMemWidth %2d bytes", 1 << (reg.op_mem_width - 1));
 
-	printf("ibs_op_data3:\t%016llx LdOp %d StOp %d DcL1TlbMiss %d DcL2TlbMiss %d "
-		"DcL1TlbHit2M %d DcL1TlbHit1G %d DcL2TlbHit2M %d DcMiss %d DcMisAcc %d "
-		"DcWcMemAcc %d DcUcMemAcc %d DcLockedOp %d DcMissNoMabAlloc %d DcLinAddrValid %d "
-		"DcPhyAddrValid %d DcL2TlbHit1G %d%s SwPf %d%s%s DcMissLat %5d TlbRefillLat %5d\n",
-		reg.val, reg.ld_op, reg.st_op, reg.dc_l1tlb_miss, reg.dc_l2tlb_miss,
-		reg.dc_l1tlb_hit_2m, reg.dc_l1tlb_hit_1g, reg.dc_l2tlb_hit_2m, reg.dc_miss,
-		reg.dc_mis_acc, reg.dc_wc_mem_acc, reg.dc_uc_mem_acc, reg.dc_locked_op,
-		reg.dc_miss_no_mab_alloc, reg.dc_lin_addr_valid, reg.dc_phy_addr_valid,
-		reg.dc_l2_tlb_hit_1g, l2_miss_str, reg.sw_pf, op_mem_width_str,
-		op_dc_miss_open_mem_reqs_str, reg.dc_miss_lat, reg.tlb_refill_lat);
+	if (dtlb_pgsize_cap) {
+		if (reg.dc_phy_addr_valid) {
+			int idx = (reg.dc_l1tlb_hit_1g << 1) | reg.dc_l1tlb_hit_2m;
+
+			snprintf(dc_l1_l2tlb_miss_str, sizeof(dc_l1_l2tlb_miss_str),
+				 " DcL1TlbMiss %d DcL2TlbMiss %d",
+				 reg.dc_l1tlb_miss, reg.dc_l2tlb_miss);
+			snprintf(dc_page_size_str, sizeof(dc_page_size_str),
+				 " DcPageSize %4s", dc_page_sizes[idx]);
+		}
+	} else {
+		snprintf(dc_l1_l2tlb_miss_str, sizeof(dc_l1_l2tlb_miss_str),
+			 " DcL1TlbMiss %d DcL2TlbMiss %d",
+			 reg.dc_l1tlb_miss, reg.dc_l2tlb_miss);
+		snprintf(dc_l1tlb_hit_str, sizeof(dc_l1tlb_hit_str),
+			 " DcL1TlbHit2M %d DcL1TlbHit1G %d",
+			 reg.dc_l1tlb_hit_2m, reg.dc_l1tlb_hit_1g);
+		snprintf(dc_l2tlb_hit_2m_str, sizeof(dc_l2tlb_hit_2m_str),
+			 " DcL2TlbHit2M %d", reg.dc_l2tlb_hit_2m);
+		snprintf(dc_l2tlb_hit_1g_str, sizeof(dc_l2tlb_hit_1g_str),
+			 " DcL2TlbHit1G %d", reg.dc_l2_tlb_hit_1g);
+	}
+
+	printf("ibs_op_data3:\t%016llx LdOp %d StOp %d%s%s%s DcMiss %d DcMisAcc %d "
+		"DcWcMemAcc %d DcUcMemAcc %d DcLockedOp %d DcMissNoMabAlloc %d "
+		"DcLinAddrValid %d DcPhyAddrValid %d%s%s SwPf %d%s%s "
+		"DcMissLat %5d TlbRefillLat %5d\n",
+		reg.val, reg.ld_op, reg.st_op, dc_l1_l2tlb_miss_str,
+		dtlb_pgsize_cap ? dc_page_size_str : dc_l1tlb_hit_str,
+		dc_l2tlb_hit_2m_str, reg.dc_miss, reg.dc_mis_acc, reg.dc_wc_mem_acc,
+		reg.dc_uc_mem_acc, reg.dc_locked_op, reg.dc_miss_no_mab_alloc,
+		reg.dc_lin_addr_valid, reg.dc_phy_addr_valid, dc_l2tlb_hit_1g_str,
+		l2_miss_str, reg.sw_pf, op_mem_width_str, op_dc_miss_open_mem_reqs_str,
+		reg.dc_miss_lat, reg.tlb_refill_lat);
 }
 
 /*
@@ -331,6 +374,12 @@ bool evlist__has_amd_ibs(struct evlist *evlist)
 	if (perf_env__find_pmu_cap(env, "ibs_op", "zen4_ibs_extensions"))
 		zen4_ibs_extensions = 1;
 
+	if (perf_env__find_pmu_cap(env, "ibs_op", "ldlat"))
+		ldlat_cap = 1;
+
+	if (perf_env__find_pmu_cap(env, "ibs_op", "dtlb_pgsize"))
+		dtlb_pgsize_cap = 1;
+
 	if (ibs_fetch_type || ibs_op_type) {
 		if (!cpu_family)
 			parse_cpuid(env);
diff --git a/tools/perf/util/mem-events.c b/tools/perf/util/mem-events.c
index 3a2e3687878c1862c64d0f723496a76ceb2f8229..32890848bb3dfc4f682329695b3fdd19b34cb1a6 100644
--- a/tools/perf/util/mem-events.c
+++ b/tools/perf/util/mem-events.c
@@ -17,43 +17,94 @@
 
 unsigned int perf_mem_events__loads_ldlat = 30;
 
-#define E(t, n, s) { .tag = t, .name = n, .sysfs_name = s }
+#define E(t, n, s, l, a) { .tag = t, .name = n, .event_name = s, .ldlat = l, .aux_event = a }
 
-static struct perf_mem_event perf_mem_events[PERF_MEM_EVENTS__MAX] = {
-	E("ldlat-loads",	"cpu/mem-loads,ldlat=%u/P",	"cpu/events/mem-loads"),
-	E("ldlat-stores",	"cpu/mem-stores/P",		"cpu/events/mem-stores"),
-	E(NULL,			NULL,				NULL),
+struct perf_mem_event perf_mem_events[PERF_MEM_EVENTS__MAX] = {
+	E("ldlat-loads",	"%s/mem-loads,ldlat=%u/P",	"mem-loads",	true,	0),
+	E("ldlat-stores",	"%s/mem-stores/P",		"mem-stores",	false,	0),
+	E(NULL,			NULL,				NULL,		false,	0),
 };
 #undef E
 
 static char mem_loads_name[100];
-static bool mem_loads_name__init;
+static char mem_stores_name[100];
 
-struct perf_mem_event * __weak perf_mem_events__ptr(int i)
+struct perf_mem_event *perf_pmu__mem_events_ptr(struct perf_pmu *pmu, int i)
 {
-	if (i >= PERF_MEM_EVENTS__MAX)
+	if (i >= PERF_MEM_EVENTS__MAX || !pmu)
 		return NULL;
 
-	return &perf_mem_events[i];
+	return &pmu->mem_events[i];
 }
 
-const char * __weak perf_mem_events__name(int i, const char *pmu_name  __maybe_unused)
+static struct perf_pmu *perf_pmus__scan_mem(struct perf_pmu *pmu)
 {
-	struct perf_mem_event *e = perf_mem_events__ptr(i);
+	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
+		if (pmu->mem_events)
+			return pmu;
+	}
+	return NULL;
+}
+
+struct perf_pmu *perf_mem_events_find_pmu(void)
+{
+	/*
+	 * The current perf mem doesn't support per-PMU configuration.
+	 * The exact same configuration is applied to all the
+	 * mem_events supported PMUs.
+	 * Return the first mem_events supported PMU.
+	 *
+	 * Notes: The only case which may support multiple mem_events
+	 * supported PMUs is Intel hybrid. The exact same mem_events
+	 * is shared among the PMUs. Only configure the first PMU
+	 * is good enough as well.
+	 */
+	return perf_pmus__scan_mem(NULL);
+}
+
+static const char *perf_pmu__mem_events_name(int i, struct perf_pmu *pmu)
+{
+	struct perf_mem_event *e;
+
+	if (i >= PERF_MEM_EVENTS__MAX || !pmu)
+		return NULL;
 
+	e = &pmu->mem_events[i];
 	if (!e)
 		return NULL;
 
-	if (i == PERF_MEM_EVENTS__LOAD) {
-		if (!mem_loads_name__init) {
-			mem_loads_name__init = true;
-			scnprintf(mem_loads_name, sizeof(mem_loads_name),
-				  e->name, perf_mem_events__loads_ldlat);
+	if (i == PERF_MEM_EVENTS__LOAD || i == PERF_MEM_EVENTS__LOAD_STORE) {
+		if (e->ldlat) {
+			if (!e->aux_event) {
+				/* ARM and Most of Intel */
+				scnprintf(mem_loads_name, sizeof(mem_loads_name),
+					  e->name, pmu->name,
+					  perf_mem_events__loads_ldlat);
+			} else {
+				/* Intel with mem-loads-aux event */
+				scnprintf(mem_loads_name, sizeof(mem_loads_name),
+					  e->name, pmu->name, pmu->name,
+					  perf_mem_events__loads_ldlat);
+			}
+		} else {
+			if (!e->aux_event) {
+				/* AMD and POWER */
+				scnprintf(mem_loads_name, sizeof(mem_loads_name),
+					  e->name, pmu->name);
+			} else
+				return NULL;
 		}
+
 		return mem_loads_name;
 	}
 
-	return e->name;
+	if (i == PERF_MEM_EVENTS__STORE) {
+		scnprintf(mem_stores_name, sizeof(mem_stores_name),
+			  e->name, pmu->name);
+		return mem_stores_name;
+	}
+
+	return NULL;
 }
 
 __weak bool is_mem_loads_aux_event(struct evsel *leader __maybe_unused)
@@ -61,7 +112,7 @@ __weak bool is_mem_loads_aux_event(struct evsel *leader __maybe_unused)
 	return false;
 }
 
-int perf_mem_events__parse(const char *str)
+int perf_pmu__mem_events_parse(struct perf_pmu *pmu, const char *str)
 {
 	char *tok, *saveptr = NULL;
 	bool found = false;
@@ -79,7 +130,7 @@ int perf_mem_events__parse(const char *str)
 
 	while (tok) {
 		for (j = 0; j < PERF_MEM_EVENTS__MAX; j++) {
-			struct perf_mem_event *e = perf_mem_events__ptr(j);
+			struct perf_mem_event *e = perf_pmu__mem_events_ptr(pmu, j);
 
 			if (!e->tag)
 				continue;
@@ -100,19 +151,21 @@ int perf_mem_events__parse(const char *str)
 	return -1;
 }
 
-static bool perf_mem_event__supported(const char *mnt, struct perf_pmu *pmu,
+static bool perf_pmu__mem_events_supported(const char *mnt, struct perf_pmu *pmu,
 				      struct perf_mem_event *e)
 {
-	char sysfs_name[100];
 	char path[PATH_MAX];
 	struct stat st;
 
-	scnprintf(sysfs_name, sizeof(sysfs_name), e->sysfs_name, pmu->name);
-	scnprintf(path, PATH_MAX, "%s/devices/%s", mnt, sysfs_name);
+	if (!e->event_name)
+		return true;
+
+	scnprintf(path, PATH_MAX, "%s/devices/%s/events/%s", mnt, pmu->name, e->event_name);
+
 	return !stat(path, &st);
 }
 
-int perf_mem_events__init(void)
+int perf_pmu__mem_events_init(struct perf_pmu *pmu)
 {
 	const char *mnt = sysfs__mount();
 	bool found = false;
@@ -122,8 +175,7 @@ int perf_mem_events__init(void)
 		return -ENOENT;
 
 	for (j = 0; j < PERF_MEM_EVENTS__MAX; j++) {
-		struct perf_mem_event *e = perf_mem_events__ptr(j);
-		struct perf_pmu *pmu = NULL;
+		struct perf_mem_event *e = perf_pmu__mem_events_ptr(pmu, j);
 
 		/*
 		 * If the event entry isn't valid, skip initialization
@@ -132,86 +184,59 @@ int perf_mem_events__init(void)
 		if (!e->tag)
 			continue;
 
-		/*
-		 * Scan all PMUs not just core ones, since perf mem/c2c on
-		 * platforms like AMD uses IBS OP PMU which is independent
-		 * of core PMU.
-		 */
-		while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-			e->supported |= perf_mem_event__supported(mnt, pmu, e);
-			if (e->supported) {
-				found = true;
-				break;
-			}
-		}
+		e->supported |= perf_pmu__mem_events_supported(mnt, pmu, e);
+		if (e->supported)
+			found = true;
 	}
 
 	return found ? 0 : -ENOENT;
 }
 
-void perf_mem_events__list(void)
+void perf_pmu__mem_events_list(struct perf_pmu *pmu)
 {
 	int j;
 
 	for (j = 0; j < PERF_MEM_EVENTS__MAX; j++) {
-		struct perf_mem_event *e = perf_mem_events__ptr(j);
+		struct perf_mem_event *e = perf_pmu__mem_events_ptr(pmu, j);
 
 		fprintf(stderr, "%-*s%-*s%s",
 			e->tag ? 13 : 0,
 			e->tag ? : "",
 			e->tag && verbose > 0 ? 25 : 0,
-			e->tag && verbose > 0 ? perf_mem_events__name(j, NULL) : "",
+			e->tag && verbose > 0 ? perf_pmu__mem_events_name(j, pmu) : "",
 			e->supported ? ": available\n" : "");
 	}
 }
 
-static void perf_mem_events__print_unsupport_hybrid(struct perf_mem_event *e,
-						    int idx)
-{
-	const char *mnt = sysfs__mount();
-	struct perf_pmu *pmu = NULL;
-
-	while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-		if (!perf_mem_event__supported(mnt, pmu, e)) {
-			pr_err("failed: event '%s' not supported\n",
-			       perf_mem_events__name(idx, pmu->name));
-		}
-	}
-}
-
 int perf_mem_events__record_args(const char **rec_argv, int *argv_nr,
 				 char **rec_tmp, int *tmp_nr)
 {
 	const char *mnt = sysfs__mount();
+	struct perf_pmu *pmu = NULL;
 	int i = *argv_nr, k = 0;
 	struct perf_mem_event *e;
 
-	for (int j = 0; j < PERF_MEM_EVENTS__MAX; j++) {
-		e = perf_mem_events__ptr(j);
-		if (!e->record)
-			continue;
 
-		if (perf_pmus__num_mem_pmus() == 1) {
-			if (!e->supported) {
-				pr_err("failed: event '%s' not supported\n",
-				       perf_mem_events__name(j, NULL));
-				return -1;
-			}
+	while ((pmu = perf_pmus__scan_mem(pmu)) != NULL) {
+		for (int j = 0; j < PERF_MEM_EVENTS__MAX; j++) {
+			e = perf_pmu__mem_events_ptr(pmu, j);
 
-			rec_argv[i++] = "-e";
-			rec_argv[i++] = perf_mem_events__name(j, NULL);
-		} else {
-			struct perf_pmu *pmu = NULL;
+			if (!e->record)
+				continue;
 
 			if (!e->supported) {
-				perf_mem_events__print_unsupport_hybrid(e, j);
+				pr_err("failed: event '%s' not supported\n",
+					perf_pmu__mem_events_name(j, pmu));
 				return -1;
 			}
 
-			while ((pmu = perf_pmus__scan(pmu)) != NULL) {
-				const char *s = perf_mem_events__name(j, pmu->name);
+			if (perf_pmus__num_mem_pmus() == 1) {
+				rec_argv[i++] = "-e";
+				rec_argv[i++] = perf_pmu__mem_events_name(j, pmu);
+			} else {
+				const char *s = perf_pmu__mem_events_name(j, pmu);
 
-				if (!perf_mem_event__supported(mnt, pmu, e))
+				if (!perf_pmu__mem_events_supported(mnt, pmu, e))
 					continue;
 
 				rec_argv[i++] = "-e";
diff --git a/tools/perf/util/mem-events.h b/tools/perf/util/mem-events.h
index b40ad6ea93fcdb70d0111b20abb8cbab475ec120..f817a507b106ae8653c7109f8a94b89eae960d3b 100644
--- a/tools/perf/util/mem-events.h
+++ b/tools/perf/util/mem-events.h
@@ -14,9 +14,11 @@
 struct perf_mem_event {
 	bool		record;
 	bool		supported;
+	bool		ldlat;
+	u32		aux_event;
 	const char	*tag;
 	const char	*name;
-	const char	*sysfs_name;
+	const char	*event_name;
 };
 
 struct mem_info {
@@ -34,15 +36,16 @@ enum {
 };
 
 extern unsigned int perf_mem_events__loads_ldlat;
+extern struct perf_mem_event perf_mem_events[PERF_MEM_EVENTS__MAX];
 
-int perf_mem_events__parse(const char *str);
-int perf_mem_events__init(void);
+int perf_pmu__mem_events_parse(struct perf_pmu *pmu, const char *str);
+int perf_pmu__mem_events_init(struct perf_pmu *pmu);
 
-const char *perf_mem_events__name(int i, const char *pmu_name);
-struct perf_mem_event *perf_mem_events__ptr(int i);
+struct perf_mem_event *perf_pmu__mem_events_ptr(struct perf_pmu *pmu, int i);
+struct perf_pmu *perf_mem_events_find_pmu(void);
 bool is_mem_loads_aux_event(struct evsel *leader);
 
-void perf_mem_events__list(void);
+void perf_pmu__mem_events_list(struct perf_pmu *pmu);
 int perf_mem_events__record_args(const char **rec_argv, int *argv_nr,
 				 char **rec_tmp, int *tmp_nr);
 
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 5edfeac95c444c5ae6e9667d4bf15516c85df202..15eac86f15d13bdb85490e1e32348bd54929a745 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -1032,8 +1032,10 @@ static int pmu_max_precise(int dirfd, struct perf_pmu *pmu)
 }
 
 void __weak
-perf_pmu__arch_init(struct perf_pmu *pmu __maybe_unused)
+perf_pmu__arch_init(struct perf_pmu *pmu)
 {
+	if (pmu->is_core)
+		pmu->mem_events = perf_mem_events;
 }
 
 struct perf_pmu *perf_pmu__lookup(struct list_head *pmus, int dirfd, const char *name)
@@ -1939,6 +1941,17 @@ static void perf_pmu__del_caps(struct perf_pmu *pmu)
 	}
 }
 
+struct perf_pmu_caps *perf_pmu__get_cap(struct perf_pmu *pmu, const char *name)
+{
+	struct perf_pmu_caps *caps;
+
+	list_for_each_entry(caps, &pmu->caps, list) {
+		if (!strcmp(caps->name, name))
+			return caps;
+	}
+	return NULL;
+}
+
 /*
  * Reading/parsing the given pmu capabilities, which should be located at:
  * /sys/bus/event_source/devices/<dev>/caps as sysfs group attributes.
diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h
index 1765c35080d4db4d08d83bb19a74eb7f04dc802e..9e85d58ad224988e02f968e4db9335f8413e821c 100644
--- a/tools/perf/util/pmu.h
+++ b/tools/perf/util/pmu.h
@@ -10,6 +10,8 @@
 #include <stdio.h>
 #include "parse-events.h"
 #include "pmu-events/pmu-events.h"
+#include "map_symbol.h"
+#include "mem-events.h"
 
 struct evsel_config_term;
 struct perf_cpu_map;
@@ -169,6 +171,11 @@ struct perf_pmu {
 		 */
 		bool exclude_guest;
 	} missing_features;
+
+	/**
+	 * @mem_events: List of the supported mem events
+	 */
+	struct perf_mem_event *mem_events;
 };
 
 /** @perf_pmu__fake: A special global PMU used for testing. */
@@ -254,6 +261,8 @@ bool pmu_uncore_identifier_match(const char *compat, const char *id);
 
 int perf_pmu__convert_scale(const char *scale, char **end, double *sval);
 
+struct perf_pmu_caps *perf_pmu__get_cap(struct perf_pmu *pmu, const char *name);
+
 int perf_pmu__caps_parse(struct perf_pmu *pmu);
 
 void perf_pmu__warn_invalid_config(struct perf_pmu *pmu, __u64 config,