Merge tag 'perf-tools-for-v6.11-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools

Pull perf tools updates from Namhyung Kim:
 "Build:

   - Build each directory as a library so that depedency check for the
     python extension module can be automatic

   - Use pkg-config to check libtraceevent and libtracefs

  perf sched:

   - Add --task-name and --fuzzy-name options for `perf sched map`

     It focuses on selected tasks only by removing unrelated tasks in
     the output. It matches the task comm with the given string and the
     --fuzzy-name option allows the partial matching:

       $ sudo perf sched record -a sleep 1

       $ sudo perf sched map --task-name kworker --fuzzy-name
          .   .   .   .   -  *A0  .   .    481065.315131 secs A0 => kworker/5:2-i91:438521
          .   .   .   .   -  *-   .   .    481065.315160 secs
         *B0  .   .   .   -   .   .   .    481065.316435 secs B0 => kworker/0:0-i91:437860
         *-   .   .   .   .   .   .   .    481065.316441 secs
          .   .   .   .   .  *A0  .   .    481065.318703 secs
          .   .   .   .   .  *-   .   .    481065.318717 secs
          .   .  *C0  .   .   .   .   .    481065.320544 secs C0 => kworker/u16:30-:430186
          .   .  *-   .   .   .   .   .    481065.320555 secs
          .   .  *D0  .   .   .   .   .    481065.328524 secs D0 => kworker/2:0-kdm:429654
         *B0  .   D0  .   -   .   .   .    481065.328527 secs
         *-   .   D0  .   -   .   .   .    481065.328535 secs
          .   .  *-   .   .   .   .   .    481065.328535 secs

   - Fix -r/--repeat option of perf sched replay

     The documentation said -1 will work as infinity but it didn't
     accept the value. Update the code and document to use 0 instead

   - Fix perf sched timehist to account the delay time for preempted
     tasks

  Perf event filtering:

   - perf top gained filtering support on regular events using BPF like
     perf record. Previously it was able to use it for tracepoints only

   - The BPF filter now supports filtering by UID/GID. This should be
     preferred than -u <UID> option as it's racy to scan /proc to check
     tasks for the user and fails to open an event for the task if it's
     already gone

       $ sudo perf top -e cycles --filter "uid == $(id -u)"

  perf report:

   - Skip dummy events in the group output by default. The --skip-empty
     option controls display of empty events without samples. But perf
     report can force display all events in a group

     In this case, auto-added a dummy event (for a system-wide record)
     ends up in the output. Now it can skip those empty events even in
     the group display mode

     To preserve the old behavior, run this:

       $ perf report --group --no-skip-empty

  perf stat:

   - Choose the most disaggregate option when multiple aggregation
     options are given. It used to pick the last option in the command
     line but it can be confusing and not consistent. Now it'll choose
     the smallest unit

     For example, it'd aggregate the result per-core when the user gave
     both --per-socket and --per-core options at the same time

  Internals:

   - Fix `perf bench` when some CPUs are offline

   - Fix handling of JIT symbol mappings to accept "/tmp/perf-${PID}.map
     patterns only so that it can not be confused by other /tmp/perf-*
     files

   - Many improvements and fixes for `perf test`

  Others:

   - Support some new instructions for Intel-PT

   - Fix syscall ID mapping in perf trace

   - Document AMD IBS PMU usages

   - Change `perf lock info` to show map and thread info by default

  Vendor JSON events:

   - Update Intel events and metrics

   - Add i.MX9[35] DDR metrics"

* tag 'perf-tools-for-v6.11-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools: (125 commits)
  perf trace: Fix iteration of syscall ids in syscalltbl->entries
  perf dso: Fix address sanitizer build
  perf mem: Warn if memory events are not supported on all CPUs
  perf arm-spe: Support multiple Arm SPE PMUs
  perf build x86: Fix SC2034 error in syscalltbl.sh
  perf record: Fix memset out-of-range error
  perf sched map: Add --fuzzy-name option for fuzzy matching in task names
  perf sched map: Add support for multiple task names using CSV
  perf sched map: Add task-name option to filter the output map
  perf build: Conditionally add feature check flags for libtrace{event,fs}
  perf install: Don't propagate subdir to Documentation submake
  perf vendor events arm64:: Add i.MX95 DDR Performance Monitor metrics
  perf vendor events arm64:: Add i.MX93 DDR Performance Monitor metrics
  perf dsos: When adding a dso into sorted dsos maintain the sort order
  perf comm str: Avoid sort during insert
  perf report: Calling available function for stats printing
  perf intel-pt: Fix exclude_guest setting
  perf intel-pt: Fix aux_watermark calculation for 64-bit size
  perf sched replay: Fix -r/--repeat command line option for infinity
  perf: pmus: Remove unneeded semicolon
  ...

Author: torvalds

tools/lib/api/io.h

@@ -43,48 +43,55 @@ static inline void io__init(struct io *io, int fd,
 	io->eof = false;
 }
 
-/* Reads one character from the "io" file with similar semantics to fgetc. */
-static inline int io__get_char(struct io *io)
+/* Read from fd filling the buffer. Called when io->data == io->end. */
+static inline int io__fill_buffer(struct io *io)
 {
-	char *ptr = io->data;
+	ssize_t n;
 
 	if (io->eof)
 		return -1;
 
-	if (ptr == io->end) {
-		ssize_t n;
-
-		if (io->timeout_ms != 0) {
-			struct pollfd pfds[] = {
-				{
-					.fd = io->fd,
-					.events = POLLIN,
-				},
-			};
-
-			n = poll(pfds, 1, io->timeout_ms);
-			if (n == 0)
-				errno = ETIMEDOUT;
-			if (n > 0 && !(pfds[0].revents & POLLIN)) {
-				errno = EIO;
-				n = -1;
-			}
-			if (n <= 0) {
-				io->eof = true;
-				return -1;
-			}
+	if (io->timeout_ms != 0) {
+		struct pollfd pfds[] = {
+			{
+				.fd = io->fd,
+				.events = POLLIN,
+			},
+		};
+
+		n = poll(pfds, 1, io->timeout_ms);
+		if (n == 0)
+			errno = ETIMEDOUT;
+		if (n > 0 && !(pfds[0].revents & POLLIN)) {
+			errno = EIO;
+			n = -1;
 		}
-		n = read(io->fd, io->buf, io->buf_len);
-
 		if (n <= 0) {
 			io->eof = true;
 			return -1;
 		}
-		ptr = &io->buf[0];
-		io->end = &io->buf[n];
 	}
-	io->data = ptr + 1;
-	return *ptr;
+	n = read(io->fd, io->buf, io->buf_len);
+
+	if (n <= 0) {
+		io->eof = true;
+		return -1;
+	}
+	io->data = &io->buf[0];
+	io->end = &io->buf[n];
+	return 0;
+}
+
+/* Reads one character from the "io" file with similar semantics to fgetc. */
+static inline int io__get_char(struct io *io)
+{
+	if (io->data == io->end) {
+		int ret = io__fill_buffer(io);
+
+		if (ret)
+			return ret;
+	}
+	return *io->data++;
 }
 
 /* Read a hexadecimal value with no 0x prefix into the out argument hex. If the

tools/lib/perf/include/perf/event.h

@@ -77,6 +77,12 @@ struct perf_record_lost_samples {
 	__u64			 lost;
 };
 
+#define MAX_ID_HDR_ENTRIES  6
+struct perf_record_lost_samples_and_ids {
+	struct perf_record_lost_samples lost;
+	__u64 sample_ids[MAX_ID_HDR_ENTRIES];
+};
+
 /*
  * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID | PERF_FORMAT_LOST
  */

tools/perf/Build

@@ -1,4 +1,4 @@
-perf-y += builtin-bench.o
+perf-bench-y += builtin-bench.o
 perf-y += builtin-annotate.o
 perf-y += builtin-config.o
 perf-y += builtin-diff.o
@@ -35,8 +35,8 @@ endif
 
 perf-$(CONFIG_LIBELF) += builtin-probe.o
 
-perf-y += bench/
-perf-y += tests/
+perf-bench-y += bench/
+perf-test-y += tests/
 
 perf-y += perf.o
 
@@ -53,10 +53,12 @@ CFLAGS_builtin-trace.o	   += -DSTRACE_GROUPS_DIR="BUILD_STR($(STRACE_GROUPS_DIR_
 CFLAGS_builtin-report.o	   += -DTIPDIR="BUILD_STR($(tipdir_SQ))"
 CFLAGS_builtin-report.o	   += -DDOCDIR="BUILD_STR($(srcdir_SQ)/Documentation)"
 
-perf-y += util/
+perf-util-y += util/
+perf-util-y += arch/
 perf-y += arch/
-perf-y += ui/
-perf-y += scripts/
+perf-test-y += arch/
+perf-ui-y += ui/
+perf-util-y += scripts/
 
 gtk-y += ui/gtk/
 

tools/perf/Documentation/perf-amd-ibs.txt

@@ -0,0 +1,189 @@
+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
+
+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]

tools/perf/Documentation/perf-kwork.txt

@@ -1,4 +1,4 @@
-perf-kowrk(1)
+perf-kwork(1)
 =============
 
 NAME
@@ -35,7 +35,7 @@ There are several variants of 'perf kwork':
         perf kwork top
         perf kwork top -b
 
-   By default it shows the individual work events such as irq, workqeueu,
+   By default it shows the individual work events such as irq, workqueue,
    including the run time and delay (time between raise and actually entry):
 
       Runtime start      Runtime end        Cpu     Kwork name                 Runtime     Delaytime

tools/perf/Documentation/perf-lock.txt

@@ -111,11 +111,11 @@ INFO OPTIONS
 
 -t::
 --threads::
-	dump thread list in perf.data
+	dump only the thread list in perf.data
 
 -m::
 --map::
-	dump map of lock instances (address:name table)
+	dump only the map of lock instances (address:name table)
 
 
 CONTENTION OPTIONS

tools/perf/Documentation/perf-mem.txt

@@ -21,7 +21,7 @@ and stores are sampled. Use the -t option to limit to loads or stores.
 
 Note that on Intel systems the memory latency reported is the use-latency,
 not the pure load (or store latency). Use latency includes any pipeline
-queueing delays in addition to the memory subsystem latency.
+queuing delays in addition to the memory subsystem latency.
 
 On Arm64 this uses SPE to sample load and store operations, therefore hardware
 and kernel support is required. See linkperf:perf-arm-spe[1] for a setup guide.

tools/perf/Documentation/perf-record.txt

@@ -200,7 +200,7 @@ OPTIONS
 	  ip, id, tid, pid, cpu, time, addr, period, txn, weight, phys_addr,
 	  code_pgsz, data_pgsz, weight1, weight2, weight3, ins_lat, retire_lat,
 	  p_stage_cyc, mem_op, mem_lvl, mem_snoop, mem_remote, mem_lock,
-	  mem_dtlb, mem_blk, mem_hops
+	  mem_dtlb, mem_blk, mem_hops, uid, gid
 
 	The <operator> can be one of:
 	  ==, !=, >, >=, <, <=, &
@@ -311,7 +311,7 @@ OPTIONS
 	User can change the size by passing the size after comma like
 	"--call-graph dwarf,4096".
 
-	When "fp" recording is used, perf tries to save stack enties
+	When "fp" recording is used, perf tries to save stack entries
 	up to the number specified in sysctl.kernel.perf_event_max_stack
 	by default.  User can change the number by passing it after comma
 	like "--call-graph fp,32".