Merge tag 'perf-tools-2020-06-02' of git://git.kernel.org/pub/scm/lin…

…ux/kernel/git/acme/linux

Pull perf tooling updates from Arnaldo Carvalho de Melo:
 "These are additional changes to the perf tools, on top of what Ingo
  already submitted.

   - Further Intel PT call-trace fixes

   - Improve SELinux docs and tool warnings

   - Fix race at exit in 'perf record' using eventfd.

   - Add missing build tests to the default set of 'make -C tools/perf
     build-test'

   - Sync msr-index.h getting new AMD MSRs to decode and filter in 'perf
     trace'.

   - Fix fallback to libaudit in 'perf trace' for arches not using
     per-arch *.tbl files.

   - Fixes for 'perf ftrace'.

   - Fixes and improvements for the 'perf stat' metrics.

   - Use dummy event to get PERF_RECORD_{FORK,MMAP,etc} while
     synthesizing those metadata events for pre-existing threads.

   - Fix leaks detected using clang tooling.

   - Improvements to PMU event metric testing.

   - Report summary for 'perf stat' interval mode at the end, summing up
     all the intervals.

   - Improve pipe mode, i.e. this now works as expected, continuously
     dumping samples:

        # perf record -g -e raw_syscalls:sys_enter | perf --no-pager script

   - Fixes for event grouping, detecting incompatible groups such as:

        # perf stat -e '{cycles,power/energy-cores/}' -v
        WARNING: group events cpu maps do not match, disabling group:
          anon group { power/energy-cores/, cycles }
            power/energy-cores/: 0
            cycles: 0-7

   - Fixes for 'perf probe': blacklist address checking, number of
     kretprobe instances, etc.

   - JIT processing improvements and fixes plus the addition of a 'perf
     test' entry for the java demangler.

   - Add support for synthesizing first/last level cache, TLB and remove
     access events from HW tracing in the auxtrace code, first to use is
     ARM SPE.

   - Vendor events updates and fixes, including for POWER9 and Intel.

   - Allow using ~/.perfconfig for removing the ',' separators in 'perf
     stat' output.

   - Opt-in support for libpfm4"

* tag 'perf-tools-2020-06-02' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (120 commits)
  perf tools: Remove some duplicated includes
  perf symbols: Fix kernel maps for kcore and eBPF
  tools arch x86: Sync the msr-index.h copy with the kernel sources
  perf stat: Ensure group is defined on top of the same cpu mask
  perf libdw: Fix off-by 1 relative directory includes
  perf arm-spe: Support synthetic events
  perf auxtrace: Add four itrace options
  perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
  perf test: Initialize memory in dwarf-unwind
  perf tests: Don't tail call optimize in unwind test
  tools compiler.h: Add attribute to disable tail calls
  perf build: Add a LIBPFM4=1 build test entry
  perf tools: Add optional support for libpfm4
  perf tools: Correct license on jsmn JSON parser
  perf jit: Fix inaccurate DWARF line table
  perf jvmti: Remove redundant jitdump line table entries
  perf build: Add NO_SDT=1 to the default set of build tests
  perf build: Add NO_LIBCRYPTO=1 to the default set of build tests
  perf build: Add NO_SYSCALL_TABLE=1 to the build tests
  perf build: Remove libaudit from the default feature checks
  ...

tools/arch/x86/include/asm/msr-index.h

@@ -301,6 +301,9 @@
 #define MSR_PP1_ENERGY_STATUS		0x00000641
 #define MSR_PP1_POLICY			0x00000642
 
+#define MSR_AMD_PKG_ENERGY_STATUS	0xc001029b
+#define MSR_AMD_RAPL_POWER_UNIT		0xc0010299
+
 /* Config TDP MSRs */
 #define MSR_CONFIG_TDP_NOMINAL		0x00000648
 #define MSR_CONFIG_TDP_LEVEL_1		0x00000649

tools/build/Makefile.feature

@@ -40,7 +40,6 @@ FEATURE_TESTS_BASIC :=                  \
         glibc                           \
         gtk2                            \
         gtk2-infobar                    \
-        libaudit                        \
         libbfd                          \
         libcap                          \
         libelf                          \
@@ -112,7 +111,6 @@ FEATURE_DISPLAY ?=              \
          dwarf_getlocations     \
          glibc                  \
          gtk2                   \
-         libaudit               \
          libbfd                 \
          libcap                 \
          libelf                 \

tools/build/feature/Makefile

@@ -91,7 +91,7 @@ __BUILDXX = $(CXX) $(CXXFLAGS) -MD -Wall -Werror -o $@ $(patsubst %.bin,%.cpp,$(
 ###############################
 
 $(OUTPUT)test-all.bin:
-	$(BUILD) -fstack-protector-all -O2 -D_FORTIFY_SOURCE=2 -ldw -lelf -lnuma -lelf -laudit -I/usr/include/slang -lslang $(shell $(PKG_CONFIG) --libs --cflags gtk+-2.0 2>/dev/null) $(FLAGS_PERL_EMBED) $(FLAGS_PYTHON_EMBED) -DPACKAGE='"perf"' -lbfd -ldl -lz -llzma
+	$(BUILD) -fstack-protector-all -O2 -D_FORTIFY_SOURCE=2 -ldw -lelf -lnuma -lelf -I/usr/include/slang -lslang $(shell $(PKG_CONFIG) --libs --cflags gtk+-2.0 2>/dev/null) $(FLAGS_PERL_EMBED) $(FLAGS_PYTHON_EMBED) -DPACKAGE='"perf"' -lbfd -ldl -lz -llzma
 
 $(OUTPUT)test-hello.bin:
 	$(BUILD)

tools/build/feature/test-all.c

@@ -74,10 +74,6 @@
 # include "test-libunwind.c"
 #undef main
 
-#define main main_test_libaudit
-# include "test-libaudit.c"
-#undef main
-
 #define main main_test_libslang
 # include "test-libslang.c"
 #undef main
@@ -208,7 +204,6 @@ int main(int argc, char *argv[])
 	main_test_libelf_gelf_getnote();
 	main_test_libelf_getshdrstrndx();
 	main_test_libunwind();
-	main_test_libaudit();
 	main_test_libslang();
 	main_test_gtk2(argc, argv);
 	main_test_gtk2_infobar(argc, argv);

tools/include/linux/compiler-gcc.h

@@ -27,6 +27,18 @@
 #define  __pure		__attribute__((pure))
 #endif
 #define  noinline	__attribute__((noinline))
+#ifdef __has_attribute
+#if __has_attribute(disable_tail_calls)
+#define __no_tail_call	__attribute__((disable_tail_calls))
+#endif
+#endif
+#ifndef __no_tail_call
+#if GCC_VERSION > 40201
+#define __no_tail_call	__attribute__((optimize("no-optimize-sibling-calls")))
+#else
+#define __no_tail_call
+#endif
+#endif
 #ifndef __packed
 #define __packed	__attribute__((packed))
 #endif

tools/include/linux/compiler.h

@@ -47,6 +47,9 @@
 #ifndef noinline
 #define noinline
 #endif
+#ifndef __no_tail_call
+#define __no_tail_call
+#endif
 
 /* Are two types/vars the same type (ignoring qualifiers)? */
 #ifndef __same_type

tools/perf/Documentation/itrace.txt

@@ -1,5 +1,5 @@
 		i	synthesize instructions events
-		b	synthesize branches events
+		b	synthesize branches events (branch misses for Arm SPE)
 		c	synthesize branches events (calls only)
 		r	synthesize branches events (returns only)
 		x	synthesize transactions events
@@ -9,6 +9,10 @@
 			of aux-output (refer to perf record)
 		e	synthesize error events
 		d	create a debug log
+		f	synthesize first level cache events
+		m	synthesize last level cache events
+		t	synthesize TLB events
+		a	synthesize remote access events
 		g	synthesize a call chain (use with i or x)
 		G	synthesize a call chain on existing event records
 		l	synthesize last branch entries (use with i or x)

tools/perf/Documentation/perf-c2c.txt

@@ -40,7 +40,7 @@ RECORD OPTIONS
 --------------
 -e::
 --event=::
-	Select the PMU event. Use 'perf mem record -e list'
+	Select the PMU event. Use 'perf c2c record -e list'
 	to list available events.
 
 -v::

tools/perf/Documentation/perf-config.txt

@@ -667,6 +667,11 @@ convert.*::
 		Limit the size of ordered_events queue, so we could control
 		allocation size of perf data files without proper finished
 		round events.
+stat.*::
+
+	stat.big-num::
+		(boolean) Change the default for "--big-num". To make
+		"--no-big-num" the default, set "stat.big-num=false".
 
 intel-pt.*::
 

tools/perf/Documentation/perf-intel-pt.txt

@@ -687,7 +687,7 @@ The v4.2 kernel introduced support for a context switch metadata event,
 PERF_RECORD_SWITCH, which allows unprivileged users to see when their processes
 are scheduled out and in, just not by whom, which is left for the
 PERF_RECORD_SWITCH_CPU_WIDE, that is only accessible in system wide context,
-which in turn requires CAP_SYS_ADMIN.
+which in turn requires CAP_PERFMON or CAP_SYS_ADMIN.
 
 Please see the 45ac1403f564 ("perf: Add PERF_RECORD_SWITCH to indicate context
 switches") commit, that introduces these metadata events for further info.

tools/perf/Documentation/perf-record.txt

@@ -458,7 +458,9 @@ This option sets the time out limit. The default value is 500 ms.
 
 --switch-events::
 Record context switch events i.e. events of type PERF_RECORD_SWITCH or
-PERF_RECORD_SWITCH_CPU_WIDE.
+PERF_RECORD_SWITCH_CPU_WIDE. In some cases (e.g. Intel PT or CoreSight)
+switch events will be enabled automatically, which can be suppressed by
+by the option --no-switch-events.
 
 --clang-path=PATH::
 Path to clang binary to use for compiling BPF scriptlets.
@@ -613,6 +615,17 @@ appended unit character - B/K/M/G
 	The number of threads to run when synthesizing events for existing processes.
 	By default, the number of threads equals 1.
 
+ifdef::HAVE_LIBPFM[]
+--pfm-events events::
+Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
+including support for event filters. For example '--pfm-events
+inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
+option using the comma separator. Hardware events and generic hardware
+events cannot be mixed together. The latter must be used with the -e
+option. The -e option and this one can be mixed and matched.  Events
+can be grouped using the {} notation.
+endif::HAVE_LIBPFM[]
+
 SEE ALSO
 --------
 linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-intel-pt[1]

tools/perf/Documentation/perf-stat.txt

@@ -71,6 +71,16 @@ report::
 --tid=<tid>::
         stat events on existing thread id (comma separated list)
 
+ifdef::HAVE_LIBPFM[]
+--pfm-events events::
+Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
+including support for event filters. For example '--pfm-events
+inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
+option using the comma separator. Hardware events and generic hardware
+events cannot be mixed together. The latter must be used with the -e
+option. The -e option and this one can be mixed and matched.  Events
+can be grouped using the {} notation.
+endif::HAVE_LIBPFM[]
 
 -a::
 --all-cpus::
@@ -93,7 +103,9 @@ report::
 
 -B::
 --big-num::
-        print large numbers with thousands' separators according to locale
+        print large numbers with thousands' separators according to locale.
+	Enabled by default. Use "--no-big-num" to disable.
+	Default setting can be changed with "perf config stat.big-num=false".
 
 -C::
 --cpu=::
@@ -234,6 +246,25 @@ filter out the startup phase of the program, which is often very different.
 
 Print statistics of transactional execution if supported.
 
+--metric-no-group::
+By default, events to compute a metric are placed in weak groups. The
+group tries to enforce scheduling all or none of the events. The
+--metric-no-group option places events outside of groups and may
+increase the chance of the event being scheduled - leading to more
+accuracy. However, as events may not be scheduled together accuracy
+for metrics like instructions per cycle can be lower - as both metrics
+may no longer be being measured at the same time.
+
+--metric-no-merge::
+By default metric events in different weak groups can be shared if one
+group contains all the events needed by another. In such cases one
+group will be eliminated reducing event multiplexing and making it so
+that certain groups of metrics sum to 100%. A downside to sharing a
+group is that the group may require multiplexing and so accuracy for a
+small group that need not have multiplexing is lowered. This option
+forbids the event merging logic from sharing events between groups and
+may be used to increase accuracy in this case.
+
 STAT RECORD
 -----------
 Stores stat data into perf data file.

tools/perf/Documentation/perf-top.txt

@@ -329,6 +329,17 @@ Default is to monitor all CPUS.
 	The known limitations include exception handing such as
 	setjmp/longjmp will have calls/returns not match.
 
+ifdef::HAVE_LIBPFM[]
+--pfm-events events::
+Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
+including support for event filters. For example '--pfm-events
+inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
+option using the comma separator. Hardware events and generic hardware
+events cannot be mixed together. The latter must be used with the -e
+option. The -e option and this one can be mixed and matched.  Events
+can be grouped using the {} notation.
+endif::HAVE_LIBPFM[]
+
 INTERACTIVE PROMPTING KEYS
 --------------------------