Merge tag 'perf-core-for-mingo-2.1' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements from Arnaldo Carvalho de Melo:

User visible changes:

  - Add record.build-id config option to 'perf record', to allow configuring
    in the ~/.perfconfig file if and how build-ids should be processed, allowing
    a permanent setting for options such as -B and -N: (Namhyung Kim)

    $ perf record -h -B -N

     Usage: perf record [<options>] [<command>]
        or: perf record [<options>] -- <command> [<options>]

        -B, --no-buildid       do not collect buildids in perf.data
        -N, --no-buildid-cache do not update the buildid cache

    $

Infrastructure changes:

  - Move code for options parsing and subcommand handling from tools/perf/
    to tools/lib/subcmd/, so that it can be used by other tools/ living
    utilities (Josh Poimboeuf)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
diff --git a/tools/build/Makefile.feature b/tools/build/Makefile.feature
index 37ff4c9..6c0519d 100644
--- a/tools/build/Makefile.feature
+++ b/tools/build/Makefile.feature
@@ -7,7 +7,7 @@
 
 feature_check = $(eval $(feature_check_code))
 define feature_check_code
-  feature-$(1) := $(shell $(MAKE) OUTPUT=$(OUTPUT_FEATURES) CFLAGS="$(EXTRA_CFLAGS) $(FEATURE_CHECK_CFLAGS-$(1))" LDFLAGS="$(LDFLAGS) $(FEATURE_CHECK_LDFLAGS-$(1))" -C $(feature_dir) test-$1.bin >/dev/null 2>/dev/null && echo 1 || echo 0)
+  feature-$(1) := $(shell $(MAKE) OUTPUT=$(OUTPUT_FEATURES) CFLAGS="$(EXTRA_CFLAGS) $(FEATURE_CHECK_CFLAGS-$(1))" LDFLAGS="$(LDFLAGS) $(FEATURE_CHECK_LDFLAGS-$(1))" -C $(feature_dir) $(OUTPUT_FEATURES)test-$1.bin >/dev/null 2>/dev/null && echo 1 || echo 0)
 endef
 
 feature_set = $(eval $(feature_set_code))
@@ -101,7 +101,6 @@
   #
   $(foreach feat,$(FEATURE_TESTS),$(call feature_set,$(feat)))
 else
-  $(shell $(MAKE) OUTPUT=$(OUTPUT_FEATURES) CFLAGS="$(EXTRA_CFLAGS)" LDFLAGS=$(LDFLAGS) -i -j -C $(feature_dir) $(addsuffix .bin,$(FEATURE_TESTS)) >/dev/null 2>&1)
   $(foreach feat,$(FEATURE_TESTS),$(call feature_check,$(feat)))
 endif
 
diff --git a/tools/build/feature/Makefile b/tools/build/feature/Makefile
index cea04ce9..bf8f035 100644
--- a/tools/build/feature/Makefile
+++ b/tools/build/feature/Makefile
@@ -1,4 +1,3 @@
-
 FILES=					\
 	test-all.bin			\
 	test-backtrace.bin		\
@@ -38,38 +37,40 @@
 	test-bpf.bin			\
 	test-get_cpuid.bin
 
+FILES := $(addprefix $(OUTPUT),$(FILES))
+
 CC := $(CROSS_COMPILE)gcc -MD
 PKG_CONFIG := $(CROSS_COMPILE)pkg-config
 
 all: $(FILES)
 
-__BUILD = $(CC) $(CFLAGS) -Wall -Werror -o $(OUTPUT)$@ $(patsubst %.bin,%.c,$@) $(LDFLAGS)
-  BUILD = $(__BUILD) > $(OUTPUT)$(@:.bin=.make.output) 2>&1
+__BUILD = $(CC) $(CFLAGS) -Wall -Werror -o $@ $(patsubst %.bin,%.c,$(@F)) $(LDFLAGS)
+  BUILD = $(__BUILD) > $(@:.bin=.make.output) 2>&1
 
 ###############################
 
-test-all.bin:
+$(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
 
-test-hello.bin:
+$(OUTPUT)test-hello.bin:
 	$(BUILD)
 
-test-pthread-attr-setaffinity-np.bin:
+$(OUTPUT)test-pthread-attr-setaffinity-np.bin:
 	$(BUILD) -D_GNU_SOURCE -lpthread
 
-test-stackprotector-all.bin:
+$(OUTPUT)test-stackprotector-all.bin:
 	$(BUILD) -fstack-protector-all
 
-test-fortify-source.bin:
+$(OUTPUT)test-fortify-source.bin:
 	$(BUILD) -O2 -D_FORTIFY_SOURCE=2
 
-test-bionic.bin:
+$(OUTPUT)test-bionic.bin:
 	$(BUILD)
 
-test-libelf.bin:
+$(OUTPUT)test-libelf.bin:
 	$(BUILD) -lelf
 
-test-glibc.bin:
+$(OUTPUT)test-glibc.bin:
 	$(BUILD)
 
 DWARFLIBS := -ldw
@@ -77,37 +78,37 @@
 DWARFLIBS += -lelf -lebl -lz -llzma -lbz2
 endif
 
-test-dwarf.bin:
+$(OUTPUT)test-dwarf.bin:
 	$(BUILD) $(DWARFLIBS)
 
-test-libelf-mmap.bin:
+$(OUTPUT)test-libelf-mmap.bin:
 	$(BUILD) -lelf
 
-test-libelf-getphdrnum.bin:
+$(OUTPUT)test-libelf-getphdrnum.bin:
 	$(BUILD) -lelf
 
-test-libnuma.bin:
+$(OUTPUT)test-libnuma.bin:
 	$(BUILD) -lnuma
 
-test-numa_num_possible_cpus.bin:
+$(OUTPUT)test-numa_num_possible_cpus.bin:
 	$(BUILD) -lnuma
 
-test-libunwind.bin:
+$(OUTPUT)test-libunwind.bin:
 	$(BUILD) -lelf
 
-test-libunwind-debug-frame.bin:
+$(OUTPUT)test-libunwind-debug-frame.bin:
 	$(BUILD) -lelf
 
-test-libaudit.bin:
+$(OUTPUT)test-libaudit.bin:
 	$(BUILD) -laudit
 
-test-libslang.bin:
+$(OUTPUT)test-libslang.bin:
 	$(BUILD) -I/usr/include/slang -lslang
 
-test-gtk2.bin:
+$(OUTPUT)test-gtk2.bin:
 	$(BUILD) $(shell $(PKG_CONFIG) --libs --cflags gtk+-2.0 2>/dev/null)
 
-test-gtk2-infobar.bin:
+$(OUTPUT)test-gtk2-infobar.bin:
 	$(BUILD) $(shell $(PKG_CONFIG) --libs --cflags gtk+-2.0 2>/dev/null)
 
 grep-libs  = $(filter -l%,$(1))
@@ -119,63 +120,63 @@
 PERL_EMBED_CCOPTS = `perl -MExtUtils::Embed -e ccopts 2>/dev/null`
 FLAGS_PERL_EMBED=$(PERL_EMBED_CCOPTS) $(PERL_EMBED_LDOPTS)
 
-test-libperl.bin:
+$(OUTPUT)test-libperl.bin:
 	$(BUILD) $(FLAGS_PERL_EMBED)
 
-test-libpython.bin:
+$(OUTPUT)test-libpython.bin:
 	$(BUILD)
 
-test-libpython-version.bin:
+$(OUTPUT)test-libpython-version.bin:
 	$(BUILD)
 
-test-libbfd.bin:
+$(OUTPUT)test-libbfd.bin:
 	$(BUILD) -DPACKAGE='"perf"' -lbfd -lz -liberty -ldl
 
-test-liberty.bin:
-	$(CC) $(CFLAGS) -Wall -Werror -o $(OUTPUT)$@ test-libbfd.c -DPACKAGE='"perf"' $(LDFLAGS) -lbfd -ldl -liberty
+$(OUTPUT)test-liberty.bin:
+	$(CC) $(CFLAGS) -Wall -Werror -o $@ test-libbfd.c -DPACKAGE='"perf"' $(LDFLAGS) -lbfd -ldl -liberty
 
-test-liberty-z.bin:
-	$(CC) $(CFLAGS) -Wall -Werror -o $(OUTPUT)$@ test-libbfd.c -DPACKAGE='"perf"' $(LDFLAGS) -lbfd -ldl -liberty -lz
+$(OUTPUT)test-liberty-z.bin:
+	$(CC) $(CFLAGS) -Wall -Werror -o $@ test-libbfd.c -DPACKAGE='"perf"' $(LDFLAGS) -lbfd -ldl -liberty -lz
 
-test-cplus-demangle.bin:
+$(OUTPUT)test-cplus-demangle.bin:
 	$(BUILD) -liberty
 
-test-backtrace.bin:
+$(OUTPUT)test-backtrace.bin:
 	$(BUILD)
 
-test-timerfd.bin:
+$(OUTPUT)test-timerfd.bin:
 	$(BUILD)
 
-test-libdw-dwarf-unwind.bin:
+$(OUTPUT)test-libdw-dwarf-unwind.bin:
 	$(BUILD) # -ldw provided by $(FEATURE_CHECK_LDFLAGS-libdw-dwarf-unwind)
 
-test-libbabeltrace.bin:
+$(OUTPUT)test-libbabeltrace.bin:
 	$(BUILD) # -lbabeltrace provided by $(FEATURE_CHECK_LDFLAGS-libbabeltrace)
 
-test-sync-compare-and-swap.bin:
+$(OUTPUT)test-sync-compare-and-swap.bin:
 	$(BUILD)
 
-test-compile-32.bin:
-	$(CC) -m32 -o $(OUTPUT)$@ test-compile.c
+$(OUTPUT)test-compile-32.bin:
+	$(CC) -m32 -o $@ test-compile.c
 
-test-compile-x32.bin:
-	$(CC) -mx32 -o $(OUTPUT)$@ test-compile.c
+$(OUTPUT)test-compile-x32.bin:
+	$(CC) -mx32 -o $@ test-compile.c
 
-test-zlib.bin:
+$(OUTPUT)test-zlib.bin:
 	$(BUILD) -lz
 
-test-lzma.bin:
+$(OUTPUT)test-lzma.bin:
 	$(BUILD) -llzma
 
-test-get_cpuid.bin:
+$(OUTPUT)test-get_cpuid.bin:
 	$(BUILD)
 
-test-bpf.bin:
+$(OUTPUT)test-bpf.bin:
 	$(BUILD)
 
--include *.d
+-include $(OUTPUT)*.d
 
 ###############################
 
 clean:
-	rm -f $(FILES) *.d $(FILES:.bin=.make.output)
+	rm -f $(FILES) $(OUTPUT)*.d $(FILES:.bin=.make.output)
diff --git a/tools/include/linux/string.h b/tools/include/linux/string.h
index 2e2f736..e26223f 100644
--- a/tools/include/linux/string.h
+++ b/tools/include/linux/string.h
@@ -8,4 +8,8 @@
 
 int strtobool(const char *s, bool *res);
 
+#ifndef __UCLIBC__
+extern size_t strlcpy(char *dest, const char *src, size_t size);
+#endif
+
 #endif /* _LINUX_STRING_H_ */
diff --git a/tools/lib/string.c b/tools/lib/string.c
index 065e54f..bd239bc 100644
--- a/tools/lib/string.c
+++ b/tools/lib/string.c
@@ -16,6 +16,7 @@
 #include <string.h>
 #include <errno.h>
 #include <linux/string.h>
+#include <linux/compiler.h>
 
 /**
  * memdup - duplicate region of memory
@@ -60,3 +61,29 @@
 	}
 	return 0;
 }
+
+/**
+ * strlcpy - Copy a C-string into a sized buffer
+ * @dest: Where to copy the string to
+ * @src: Where to copy the string from
+ * @size: size of destination buffer
+ *
+ * Compatible with *BSD: the result is always a valid
+ * NUL-terminated string that fits in the buffer (unless,
+ * of course, the buffer size is zero). It does not pad
+ * out the result like strncpy() does.
+ *
+ * If libc has strlcpy() then that version will override this
+ * implementation:
+ */
+size_t __weak strlcpy(char *dest, const char *src, size_t size)
+{
+	size_t ret = strlen(src);
+
+	if (size) {
+		size_t len = (ret >= size) ? size - 1 : ret;
+		memcpy(dest, src, len);
+		dest[len] = '\0';
+	}
+	return ret;
+}
diff --git a/tools/lib/subcmd/Build b/tools/lib/subcmd/Build
new file mode 100644
index 0000000..ee31288
--- /dev/null
+++ b/tools/lib/subcmd/Build
@@ -0,0 +1,7 @@
+libsubcmd-y += exec-cmd.o
+libsubcmd-y += help.o
+libsubcmd-y += pager.o
+libsubcmd-y += parse-options.o
+libsubcmd-y += run-command.o
+libsubcmd-y += sigchain.o
+libsubcmd-y += subcmd-config.o
diff --git a/tools/lib/subcmd/Makefile b/tools/lib/subcmd/Makefile
new file mode 100644
index 0000000..629cf8c
--- /dev/null
+++ b/tools/lib/subcmd/Makefile
@@ -0,0 +1,48 @@
+include ../../scripts/Makefile.include
+include ../../perf/config/utilities.mak		# QUIET_CLEAN
+
+ifeq ($(srctree),)
+srctree := $(patsubst %/,%,$(dir $(shell pwd)))
+srctree := $(patsubst %/,%,$(dir $(srctree)))
+srctree := $(patsubst %/,%,$(dir $(srctree)))
+#$(info Determined 'srctree' to be $(srctree))
+endif
+
+CC = $(CROSS_COMPILE)gcc
+AR = $(CROSS_COMPILE)ar
+RM = rm -f
+
+MAKEFLAGS += --no-print-directory
+
+LIBFILE = $(OUTPUT)libsubcmd.a
+
+CFLAGS := $(EXTRA_WARNINGS) $(EXTRA_CFLAGS)
+CFLAGS += -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=2 -fPIC
+CFLAGS += -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE
+
+CFLAGS += -I$(srctree)/tools/include/
+CFLAGS += -I$(srctree)/include/uapi
+CFLAGS += -I$(srctree)/include
+
+SUBCMD_IN := $(OUTPUT)libsubcmd-in.o
+
+all:
+
+export srctree OUTPUT CC LD CFLAGS V
+include $(srctree)/tools/build/Makefile.include
+
+all: fixdep $(LIBFILE)
+
+$(SUBCMD_IN): FORCE
+	@$(MAKE) $(build)=libsubcmd
+
+$(LIBFILE): $(SUBCMD_IN)
+	$(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(SUBCMD_IN)
+
+clean:
+	$(call QUIET_CLEAN, libsubcmd) $(RM) $(LIBFILE); \
+	find $(if $(OUTPUT),$(OUTPUT),.) -name \*.o -or -name \*.o.cmd -or -name \*.o.d | xargs $(RM)
+
+FORCE:
+
+.PHONY: clean FORCE
diff --git a/tools/lib/subcmd/exec-cmd.c b/tools/lib/subcmd/exec-cmd.c
new file mode 100644
index 0000000..1ae833a
--- /dev/null
+++ b/tools/lib/subcmd/exec-cmd.c
@@ -0,0 +1,209 @@
+#include <linux/compiler.h>
+#include <linux/string.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <string.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include "subcmd-util.h"
+#include "exec-cmd.h"
+#include "subcmd-config.h"
+
+#define MAX_ARGS	32
+#define PATH_MAX	4096
+
+static const char *argv_exec_path;
+static const char *argv0_path;
+
+void exec_cmd_init(const char *exec_name, const char *prefix,
+		   const char *exec_path, const char *exec_path_env)
+{
+	subcmd_config.exec_name		= exec_name;
+	subcmd_config.prefix		= prefix;
+	subcmd_config.exec_path		= exec_path;
+	subcmd_config.exec_path_env	= exec_path_env;
+}
+
+#define is_dir_sep(c) ((c) == '/')
+
+static int is_absolute_path(const char *path)
+{
+	return path[0] == '/';
+}
+
+static const char *get_pwd_cwd(void)
+{
+	static char cwd[PATH_MAX + 1];
+	char *pwd;
+	struct stat cwd_stat, pwd_stat;
+	if (getcwd(cwd, PATH_MAX) == NULL)
+		return NULL;
+	pwd = getenv("PWD");
+	if (pwd && strcmp(pwd, cwd)) {
+		stat(cwd, &cwd_stat);
+		if (!stat(pwd, &pwd_stat) &&
+		    pwd_stat.st_dev == cwd_stat.st_dev &&
+		    pwd_stat.st_ino == cwd_stat.st_ino) {
+			strlcpy(cwd, pwd, PATH_MAX);
+		}
+	}
+	return cwd;
+}
+
+static const char *make_nonrelative_path(const char *path)
+{
+	static char buf[PATH_MAX + 1];
+
+	if (is_absolute_path(path)) {
+		if (strlcpy(buf, path, PATH_MAX) >= PATH_MAX)
+			die("Too long path: %.*s", 60, path);
+	} else {
+		const char *cwd = get_pwd_cwd();
+		if (!cwd)
+			die("Cannot determine the current working directory");
+		if (snprintf(buf, PATH_MAX, "%s/%s", cwd, path) >= PATH_MAX)
+			die("Too long path: %.*s", 60, path);
+	}
+	return buf;
+}
+
+char *system_path(const char *path)
+{
+	char *buf = NULL;
+
+	if (is_absolute_path(path))
+		return strdup(path);
+
+	astrcatf(&buf, "%s/%s", subcmd_config.prefix, path);
+
+	return buf;
+}
+
+const char *extract_argv0_path(const char *argv0)
+{
+	const char *slash;
+
+	if (!argv0 || !*argv0)
+		return NULL;
+	slash = argv0 + strlen(argv0);
+
+	while (argv0 <= slash && !is_dir_sep(*slash))
+		slash--;
+
+	if (slash >= argv0) {
+		argv0_path = strndup(argv0, slash - argv0);
+		return argv0_path ? slash + 1 : NULL;
+	}
+
+	return argv0;
+}
+
+void set_argv_exec_path(const char *exec_path)
+{
+	argv_exec_path = exec_path;
+	/*
+	 * Propagate this setting to external programs.
+	 */
+	setenv(subcmd_config.exec_path_env, exec_path, 1);
+}
+
+
+/* Returns the highest-priority location to look for subprograms. */
+char *get_argv_exec_path(void)
+{
+	char *env;
+
+	if (argv_exec_path)
+		return strdup(argv_exec_path);
+
+	env = getenv(subcmd_config.exec_path_env);
+	if (env && *env)
+		return strdup(env);
+
+	return system_path(subcmd_config.exec_path);
+}
+
+static void add_path(char **out, const char *path)
+{
+	if (path && *path) {
+		if (is_absolute_path(path))
+			astrcat(out, path);
+		else
+			astrcat(out, make_nonrelative_path(path));
+
+		astrcat(out, ":");
+	}
+}
+
+void setup_path(void)
+{
+	const char *old_path = getenv("PATH");
+	char *new_path = NULL;
+	char *tmp = get_argv_exec_path();
+
+	add_path(&new_path, tmp);
+	add_path(&new_path, argv0_path);
+	free(tmp);
+
+	if (old_path)
+		astrcat(&new_path, old_path);
+	else
+		astrcat(&new_path, "/usr/local/bin:/usr/bin:/bin");
+
+	setenv("PATH", new_path, 1);
+
+	free(new_path);
+}
+
+static const char **prepare_exec_cmd(const char **argv)
+{
+	int argc;
+	const char **nargv;
+
+	for (argc = 0; argv[argc]; argc++)
+		; /* just counting */
+	nargv = malloc(sizeof(*nargv) * (argc + 2));
+
+	nargv[0] = subcmd_config.exec_name;
+	for (argc = 0; argv[argc]; argc++)
+		nargv[argc + 1] = argv[argc];
+	nargv[argc + 1] = NULL;
+	return nargv;
+}
+
+int execv_cmd(const char **argv) {
+	const char **nargv = prepare_exec_cmd(argv);
+
+	/* execvp() can only ever return if it fails */
+	execvp(subcmd_config.exec_name, (char **)nargv);
+
+	free(nargv);
+	return -1;
+}
+
+
+int execl_cmd(const char *cmd,...)
+{
+	int argc;
+	const char *argv[MAX_ARGS + 1];
+	const char *arg;
+	va_list param;
+
+	va_start(param, cmd);
+	argv[0] = cmd;
+	argc = 1;
+	while (argc < MAX_ARGS) {
+		arg = argv[argc++] = va_arg(param, char *);
+		if (!arg)
+			break;
+	}
+	va_end(param);
+	if (MAX_ARGS <= argc) {
+		fprintf(stderr, " Error: too many args to run %s\n", cmd);
+		return -1;
+	}
+
+	argv[argc] = NULL;
+	return execv_cmd(argv);
+}
diff --git a/tools/lib/subcmd/exec-cmd.h b/tools/lib/subcmd/exec-cmd.h
new file mode 100644
index 0000000..5d08bda
--- /dev/null
+++ b/tools/lib/subcmd/exec-cmd.h
@@ -0,0 +1,16 @@
+#ifndef __SUBCMD_EXEC_CMD_H
+#define __SUBCMD_EXEC_CMD_H
+
+extern void exec_cmd_init(const char *exec_name, const char *prefix,
+			  const char *exec_path, const char *exec_path_env);
+
+extern void set_argv_exec_path(const char *exec_path);
+extern const char *extract_argv0_path(const char *path);
+extern void setup_path(void);
+extern int execv_cmd(const char **argv); /* NULL terminated */
+extern int execl_cmd(const char *cmd, ...);
+/* get_argv_exec_path and system_path return malloc'd string, caller must free it */
+extern char *get_argv_exec_path(void);
+extern char *system_path(const char *path);
+
+#endif /* __SUBCMD_EXEC_CMD_H */
diff --git a/tools/lib/subcmd/help.c b/tools/lib/subcmd/help.c
new file mode 100644
index 0000000..e228c3c
--- /dev/null
+++ b/tools/lib/subcmd/help.c
@@ -0,0 +1,268 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <termios.h>
+#include <sys/ioctl.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <dirent.h>
+#include "subcmd-util.h"
+#include "help.h"
+#include "exec-cmd.h"
+
+void add_cmdname(struct cmdnames *cmds, const char *name, size_t len)
+{
+	struct cmdname *ent = malloc(sizeof(*ent) + len + 1);
+
+	ent->len = len;
+	memcpy(ent->name, name, len);
+	ent->name[len] = 0;
+
+	ALLOC_GROW(cmds->names, cmds->cnt + 1, cmds->alloc);
+	cmds->names[cmds->cnt++] = ent;
+}
+
+void clean_cmdnames(struct cmdnames *cmds)
+{
+	unsigned int i;
+
+	for (i = 0; i < cmds->cnt; ++i)
+		zfree(&cmds->names[i]);
+	zfree(&cmds->names);
+	cmds->cnt = 0;
+	cmds->alloc = 0;
+}
+
+int cmdname_compare(const void *a_, const void *b_)
+{
+	struct cmdname *a = *(struct cmdname **)a_;
+	struct cmdname *b = *(struct cmdname **)b_;
+	return strcmp(a->name, b->name);
+}
+
+void uniq(struct cmdnames *cmds)
+{
+	unsigned int i, j;
+
+	if (!cmds->cnt)
+		return;
+
+	for (i = j = 1; i < cmds->cnt; i++)
+		if (strcmp(cmds->names[i]->name, cmds->names[i-1]->name))
+			cmds->names[j++] = cmds->names[i];
+
+	cmds->cnt = j;
+}
+
+void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes)
+{
+	size_t ci, cj, ei;
+	int cmp;
+
+	ci = cj = ei = 0;
+	while (ci < cmds->cnt && ei < excludes->cnt) {
+		cmp = strcmp(cmds->names[ci]->name, excludes->names[ei]->name);
+		if (cmp < 0)
+			cmds->names[cj++] = cmds->names[ci++];
+		else if (cmp == 0)
+			ci++, ei++;
+		else if (cmp > 0)
+			ei++;
+	}
+
+	while (ci < cmds->cnt)
+		cmds->names[cj++] = cmds->names[ci++];
+
+	cmds->cnt = cj;
+}
+
+static void get_term_dimensions(struct winsize *ws)
+{
+	char *s = getenv("LINES");
+
+	if (s != NULL) {
+		ws->ws_row = atoi(s);
+		s = getenv("COLUMNS");
+		if (s != NULL) {
+			ws->ws_col = atoi(s);
+			if (ws->ws_row && ws->ws_col)
+				return;
+		}
+	}
+#ifdef TIOCGWINSZ
+	if (ioctl(1, TIOCGWINSZ, ws) == 0 &&
+	    ws->ws_row && ws->ws_col)
+		return;
+#endif
+	ws->ws_row = 25;
+	ws->ws_col = 80;
+}
+
+static void pretty_print_string_list(struct cmdnames *cmds, int longest)
+{
+	int cols = 1, rows;
+	int space = longest + 1; /* min 1 SP between words */
+	struct winsize win;
+	int max_cols;
+	int i, j;
+
+	get_term_dimensions(&win);
+	max_cols = win.ws_col - 1; /* don't print *on* the edge */
+
+	if (space < max_cols)
+		cols = max_cols / space;
+	rows = (cmds->cnt + cols - 1) / cols;
+
+	for (i = 0; i < rows; i++) {
+		printf("  ");
+
+		for (j = 0; j < cols; j++) {
+			unsigned int n = j * rows + i;
+			unsigned int size = space;
+
+			if (n >= cmds->cnt)
+				break;
+			if (j == cols-1 || n + rows >= cmds->cnt)
+				size = 1;
+			printf("%-*s", size, cmds->names[n]->name);
+		}
+		putchar('\n');
+	}
+}
+
+static int is_executable(const char *name)
+{
+	struct stat st;
+
+	if (stat(name, &st) || /* stat, not lstat */
+	    !S_ISREG(st.st_mode))
+		return 0;
+
+	return st.st_mode & S_IXUSR;
+}
+
+static int has_extension(const char *filename, const char *ext)
+{
+	size_t len = strlen(filename);
+	size_t extlen = strlen(ext);
+
+	return len > extlen && !memcmp(filename + len - extlen, ext, extlen);
+}
+
+static void list_commands_in_dir(struct cmdnames *cmds,
+					 const char *path,
+					 const char *prefix)
+{
+	int prefix_len;
+	DIR *dir = opendir(path);
+	struct dirent *de;
+	char *buf = NULL;
+
+	if (!dir)
+		return;
+	if (!prefix)
+		prefix = "perf-";
+	prefix_len = strlen(prefix);
+
+	astrcatf(&buf, "%s/", path);
+
+	while ((de = readdir(dir)) != NULL) {
+		int entlen;
+
+		if (prefixcmp(de->d_name, prefix))
+			continue;
+
+		astrcat(&buf, de->d_name);
+		if (!is_executable(buf))
+			continue;
+
+		entlen = strlen(de->d_name) - prefix_len;
+		if (has_extension(de->d_name, ".exe"))
+			entlen -= 4;
+
+		add_cmdname(cmds, de->d_name + prefix_len, entlen);
+	}
+	closedir(dir);
+	free(buf);
+}
+
+void load_command_list(const char *prefix,
+		struct cmdnames *main_cmds,
+		struct cmdnames *other_cmds)
+{
+	const char *env_path = getenv("PATH");
+	char *exec_path = get_argv_exec_path();
+
+	if (exec_path) {
+		list_commands_in_dir(main_cmds, exec_path, prefix);
+		qsort(main_cmds->names, main_cmds->cnt,
+		      sizeof(*main_cmds->names), cmdname_compare);
+		uniq(main_cmds);
+	}
+
+	if (env_path) {
+		char *paths, *path, *colon;
+		path = paths = strdup(env_path);
+		while (1) {
+			if ((colon = strchr(path, ':')))
+				*colon = 0;
+			if (!exec_path || strcmp(path, exec_path))
+				list_commands_in_dir(other_cmds, path, prefix);
+
+			if (!colon)
+				break;
+			path = colon + 1;
+		}
+		free(paths);
+
+		qsort(other_cmds->names, other_cmds->cnt,
+		      sizeof(*other_cmds->names), cmdname_compare);
+		uniq(other_cmds);
+	}
+	free(exec_path);
+	exclude_cmds(other_cmds, main_cmds);
+}
+
+void list_commands(const char *title, struct cmdnames *main_cmds,
+		   struct cmdnames *other_cmds)
+{
+	unsigned int i, longest = 0;
+
+	for (i = 0; i < main_cmds->cnt; i++)
+		if (longest < main_cmds->names[i]->len)
+			longest = main_cmds->names[i]->len;
+	for (i = 0; i < other_cmds->cnt; i++)
+		if (longest < other_cmds->names[i]->len)
+			longest = other_cmds->names[i]->len;
+
+	if (main_cmds->cnt) {
+		char *exec_path = get_argv_exec_path();
+		printf("available %s in '%s'\n", title, exec_path);
+		printf("----------------");
+		mput_char('-', strlen(title) + strlen(exec_path));
+		putchar('\n');
+		pretty_print_string_list(main_cmds, longest);
+		putchar('\n');
+		free(exec_path);
+	}
+
+	if (other_cmds->cnt) {
+		printf("%s available from elsewhere on your $PATH\n", title);
+		printf("---------------------------------------");
+		mput_char('-', strlen(title));
+		putchar('\n');
+		pretty_print_string_list(other_cmds, longest);
+		putchar('\n');
+	}
+}
+
+int is_in_cmdlist(struct cmdnames *c, const char *s)
+{
+	unsigned int i;
+
+	for (i = 0; i < c->cnt; i++)
+		if (!strcmp(s, c->names[i]->name))
+			return 1;
+	return 0;
+}
diff --git a/tools/perf/util/help.h b/tools/lib/subcmd/help.h
similarity index 73%
rename from tools/perf/util/help.h
rename to tools/lib/subcmd/help.h
index 7f5c6de..e145a02 100644
--- a/tools/perf/util/help.h
+++ b/tools/lib/subcmd/help.h
@@ -1,12 +1,14 @@
-#ifndef __PERF_HELP_H
-#define __PERF_HELP_H
+#ifndef __SUBCMD_HELP_H
+#define __SUBCMD_HELP_H
+
+#include <sys/types.h>
 
 struct cmdnames {
 	size_t alloc;
 	size_t cnt;
 	struct cmdname {
 		size_t len; /* also used for similarity index in help.c */
-		char name[FLEX_ARRAY];
+		char name[];
 	} **names;
 };
 
@@ -20,10 +22,13 @@
 		struct cmdnames *main_cmds,
 		struct cmdnames *other_cmds);
 void add_cmdname(struct cmdnames *cmds, const char *name, size_t len);
+void clean_cmdnames(struct cmdnames *cmds);
+int cmdname_compare(const void *a, const void *b);
+void uniq(struct cmdnames *cmds);
 /* Here we require that excludes is a sorted list. */
 void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes);
 int is_in_cmdlist(struct cmdnames *c, const char *s);
 void list_commands(const char *title, struct cmdnames *main_cmds,
 		   struct cmdnames *other_cmds);
 
-#endif /* __PERF_HELP_H */
+#endif /* __SUBCMD_HELP_H */
diff --git a/tools/perf/util/pager.c b/tools/lib/subcmd/pager.c
similarity index 84%
rename from tools/perf/util/pager.c
rename to tools/lib/subcmd/pager.c
index 53ef006..d50f3b58 100644
--- a/tools/perf/util/pager.c
+++ b/tools/lib/subcmd/pager.c
@@ -1,6 +1,12 @@
-#include "cache.h"
+#include <sys/select.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <signal.h>
+#include "pager.h"
 #include "run-command.h"
 #include "sigchain.h"
+#include "subcmd-config.h"
 
 /*
  * This is split up from the rest of git so that we can do
@@ -9,6 +15,11 @@
 
 static int spawned_pager;
 
+void pager_init(const char *pager_env)
+{
+	subcmd_config.pager_env = pager_env;
+}
+
 static void pager_preexec(void)
 {
 	/*
@@ -46,7 +57,7 @@
 
 void setup_pager(void)
 {
-	const char *pager = getenv("PERF_PAGER");
+	const char *pager = getenv(subcmd_config.pager_env);
 
 	if (!isatty(1))
 		return;
@@ -85,11 +96,5 @@
 
 int pager_in_use(void)
 {
-	const char *env;
-
-	if (spawned_pager)
-		return 1;
-
-	env = getenv("PERF_PAGER_IN_USE");
-	return env ? perf_config_bool("PERF_PAGER_IN_USE", env) : 0;
+	return spawned_pager;
 }
diff --git a/tools/lib/subcmd/pager.h b/tools/lib/subcmd/pager.h
new file mode 100644
index 0000000..8b83714
--- /dev/null
+++ b/tools/lib/subcmd/pager.h
@@ -0,0 +1,9 @@
+#ifndef __SUBCMD_PAGER_H
+#define __SUBCMD_PAGER_H
+
+extern void pager_init(const char *pager_env);
+
+extern void setup_pager(void);
+extern int pager_in_use(void);
+
+#endif /* __SUBCMD_PAGER_H */
diff --git a/tools/perf/util/parse-options.c b/tools/lib/subcmd/parse-options.c
similarity index 78%
rename from tools/perf/util/parse-options.c
rename to tools/lib/subcmd/parse-options.c
index de3290b..981bb44 100644
--- a/tools/perf/util/parse-options.c
+++ b/tools/lib/subcmd/parse-options.c
@@ -1,37 +1,66 @@
-#include "util.h"
+#include <linux/compiler.h>
+#include <linux/types.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
+#include <ctype.h>
+#include "subcmd-util.h"
 #include "parse-options.h"
-#include "cache.h"
-#include "header.h"
-#include <linux/string.h>
+#include "subcmd-config.h"
+#include "pager.h"
 
 #define OPT_SHORT 1
 #define OPT_UNSET 2
 
-static struct strbuf error_buf = STRBUF_INIT;
+char *error_buf;
 
 static int opterror(const struct option *opt, const char *reason, int flags)
 {
 	if (flags & OPT_SHORT)
-		return error("switch `%c' %s", opt->short_name, reason);
-	if (flags & OPT_UNSET)
-		return error("option `no-%s' %s", opt->long_name, reason);
-	return error("option `%s' %s", opt->long_name, reason);
+		fprintf(stderr, " Error: switch `%c' %s", opt->short_name, reason);
+	else if (flags & OPT_UNSET)
+		fprintf(stderr, " Error: option `no-%s' %s", opt->long_name, reason);
+	else
+		fprintf(stderr, " Error: option `%s' %s", opt->long_name, reason);
+
+	return -1;
+}
+
+static const char *skip_prefix(const char *str, const char *prefix)
+{
+	size_t len = strlen(prefix);
+	return strncmp(str, prefix, len) ? NULL : str + len;
+}
+
+static void optwarning(const struct option *opt, const char *reason, int flags)
+{
+	if (flags & OPT_SHORT)
+		fprintf(stderr, " Warning: switch `%c' %s", opt->short_name, reason);
+	else if (flags & OPT_UNSET)
+		fprintf(stderr, " Warning: option `no-%s' %s", opt->long_name, reason);
+	else
+		fprintf(stderr, " Warning: option `%s' %s", opt->long_name, reason);
 }
 
 static int get_arg(struct parse_opt_ctx_t *p, const struct option *opt,
 		   int flags, const char **arg)
 {
+	const char *res;
+
 	if (p->opt) {
-		*arg = p->opt;
+		res = p->opt;
 		p->opt = NULL;
 	} else if ((opt->flags & PARSE_OPT_LASTARG_DEFAULT) && (p->argc == 1 ||
 		    **(p->argv + 1) == '-')) {
-		*arg = (const char *)opt->defval;
+		res = (const char *)opt->defval;
 	} else if (p->argc > 1) {
 		p->argc--;
-		*arg = *++p->argv;
+		res = *++p->argv;
 	} else
 		return opterror(opt, "requires a value", flags);
+	if (arg)
+		*arg = res;
 	return 0;
 }
 
@@ -55,11 +84,11 @@
 
 			if (((flags & OPT_SHORT) && p->excl_opt->short_name) ||
 			    p->excl_opt->long_name == NULL) {
-				scnprintf(msg, sizeof(msg), "cannot be used with switch `%c'",
-					  p->excl_opt->short_name);
+				snprintf(msg, sizeof(msg), "cannot be used with switch `%c'",
+					 p->excl_opt->short_name);
 			} else {
-				scnprintf(msg, sizeof(msg), "cannot be used with %s",
-					  p->excl_opt->long_name);
+				snprintf(msg, sizeof(msg), "cannot be used with %s",
+					 p->excl_opt->long_name);
 			}
 			opterror(opt, msg, flags);
 			return -3;
@@ -91,6 +120,64 @@
 		}
 	}
 
+	if (opt->flags & PARSE_OPT_NOBUILD) {
+		char reason[128];
+		bool noarg = false;
+
+		err = snprintf(reason, sizeof(reason),
+				opt->flags & PARSE_OPT_CANSKIP ?
+					"is being ignored because %s " :
+					"is not available because %s",
+				opt->build_opt);
+		reason[sizeof(reason) - 1] = '\0';
+
+		if (err < 0)
+			strncpy(reason, opt->flags & PARSE_OPT_CANSKIP ?
+					"is being ignored" :
+					"is not available",
+					sizeof(reason));
+
+		if (!(opt->flags & PARSE_OPT_CANSKIP))
+			return opterror(opt, reason, flags);
+
+		err = 0;
+		if (unset)
+			noarg = true;
+		if (opt->flags & PARSE_OPT_NOARG)
+			noarg = true;
+		if (opt->flags & PARSE_OPT_OPTARG && !p->opt)
+			noarg = true;
+
+		switch (opt->type) {
+		case OPTION_BOOLEAN:
+		case OPTION_INCR:
+		case OPTION_BIT:
+		case OPTION_SET_UINT:
+		case OPTION_SET_PTR:
+		case OPTION_END:
+		case OPTION_ARGUMENT:
+		case OPTION_GROUP:
+			noarg = true;
+			break;
+		case OPTION_CALLBACK:
+		case OPTION_STRING:
+		case OPTION_INTEGER:
+		case OPTION_UINTEGER:
+		case OPTION_LONG:
+		case OPTION_U64:
+		default:
+			break;
+		}
+
+		if (!noarg)
+			err = get_arg(p, opt, flags, NULL);
+		if (err)
+			return err;
+
+		optwarning(opt, reason, flags);
+		return 0;
+	}
+
 	switch (opt->type) {
 	case OPTION_BIT:
 		if (unset)
@@ -327,14 +414,16 @@
 		return get_value(p, options, flags);
 	}
 
-	if (ambiguous_option)
-		return error("Ambiguous option: %s "
-			"(could be --%s%s or --%s%s)",
-			arg,
-			(ambiguous_flags & OPT_UNSET) ?  "no-" : "",
-			ambiguous_option->long_name,
-			(abbrev_flags & OPT_UNSET) ?  "no-" : "",
-			abbrev_option->long_name);
+	if (ambiguous_option) {
+		 fprintf(stderr,
+			 " Error: Ambiguous option: %s (could be --%s%s or --%s%s)",
+			 arg,
+			 (ambiguous_flags & OPT_UNSET) ?  "no-" : "",
+			 ambiguous_option->long_name,
+			 (abbrev_flags & OPT_UNSET) ?  "no-" : "",
+			 abbrev_option->long_name);
+		 return -1;
+	}
 	if (abbrev_option)
 		return get_value(p, abbrev_option, abbrev_flags);
 	return -2;
@@ -346,7 +435,7 @@
 		return;
 
 	if (!prefixcmp(arg, "no-")) {
-		error ("did you mean `--%s` (with two dashes ?)", arg);
+		fprintf(stderr, " Error: did you mean `--%s` (with two dashes ?)", arg);
 		exit(129);
 	}
 
@@ -354,14 +443,14 @@
 		if (!options->long_name)
 			continue;
 		if (!prefixcmp(options->long_name, arg)) {
-			error ("did you mean `--%s` (with two dashes ?)", arg);
+			fprintf(stderr, " Error: did you mean `--%s` (with two dashes ?)", arg);
 			exit(129);
 		}
 	}
 }
 
-void parse_options_start(struct parse_opt_ctx_t *ctx,
-			 int argc, const char **argv, int flags)
+static void parse_options_start(struct parse_opt_ctx_t *ctx,
+				int argc, const char **argv, int flags)
 {
 	memset(ctx, 0, sizeof(*ctx));
 	ctx->argc = argc - 1;
@@ -378,9 +467,9 @@
 				       const struct option *, int,
 				       struct parse_opt_ctx_t *);
 
-int parse_options_step(struct parse_opt_ctx_t *ctx,
-		       const struct option *options,
-		       const char * const usagestr[])
+static int parse_options_step(struct parse_opt_ctx_t *ctx,
+			      const struct option *options,
+			      const char * const usagestr[])
 {
 	int internal_help = !(ctx->flags & PARSE_OPT_NO_INTERNAL_HELP);
 	int excl_short_opt = 1;
@@ -489,7 +578,7 @@
 	return PARSE_OPT_HELP;
 }
 
-int parse_options_end(struct parse_opt_ctx_t *ctx)
+static int parse_options_end(struct parse_opt_ctx_t *ctx)
 {
 	memmove(ctx->out + ctx->cpidx, ctx->argv, ctx->argc * sizeof(*ctx->out));
 	ctx->out[ctx->cpidx + ctx->argc] = NULL;
@@ -503,18 +592,18 @@
 
 	/* build usage string if it's not provided */
 	if (subcommands && !usagestr[0]) {
-		struct strbuf buf = STRBUF_INIT;
+		char *buf = NULL;
 
-		strbuf_addf(&buf, "perf %s [<options>] {", argv[0]);
+		astrcatf(&buf, "%s %s [<options>] {", subcmd_config.exec_name, argv[0]);
+
 		for (int i = 0; subcommands[i]; i++) {
 			if (i)
-				strbuf_addstr(&buf, "|");
-			strbuf_addstr(&buf, subcommands[i]);
+				astrcat(&buf, "|");
+			astrcat(&buf, subcommands[i]);
 		}
-		strbuf_addstr(&buf, "}");
+		astrcat(&buf, "}");
 
-		usagestr[0] = strdup(buf.buf);
-		strbuf_release(&buf);
+		usagestr[0] = buf;
 	}
 
 	parse_options_start(&ctx, argc, argv, flags);
@@ -539,13 +628,11 @@
 		putchar('\n');
 		exit(130);
 	default: /* PARSE_OPT_UNKNOWN */
-		if (ctx.argv[0][1] == '-') {
-			strbuf_addf(&error_buf, "unknown option `%s'",
-				    ctx.argv[0] + 2);
-		} else {
-			strbuf_addf(&error_buf, "unknown switch `%c'",
-				    *ctx.opt);
-		}
+		if (ctx.argv[0][1] == '-')
+			astrcatf(&error_buf, "unknown option `%s'",
+				 ctx.argv[0] + 2);
+		else
+			astrcatf(&error_buf, "unknown switch `%c'", *ctx.opt);
 		usage_with_options(usagestr, options);
 	}
 
@@ -645,6 +732,10 @@
 		pad = USAGE_OPTS_WIDTH;
 	}
 	fprintf(stderr, "%*s%s\n", pad + USAGE_GAP, "", opts->help);
+	if (opts->flags & PARSE_OPT_NOBUILD)
+		fprintf(stderr, "%*s(not built-in because %s)\n",
+			USAGE_OPTS_WIDTH + USAGE_GAP, "",
+			opts->build_opt);
 }
 
 static int option__cmp(const void *va, const void *vb)
@@ -670,16 +761,18 @@
 
 static struct option *options__order(const struct option *opts)
 {
-	int nr_opts = 0;
+	int nr_opts = 0, len;
 	const struct option *o = opts;
 	struct option *ordered;
 
 	for (o = opts; o->type != OPTION_END; o++)
 		++nr_opts;
 
-	ordered = memdup(opts, sizeof(*o) * (nr_opts + 1));
-	if (ordered == NULL)
+	len = sizeof(*o) * (nr_opts + 1);
+	ordered = malloc(len);
+	if (!ordered)
 		goto out;
+	memcpy(ordered, opts, len);
 
 	qsort(ordered, nr_opts, sizeof(*o), option__cmp);
 out:
@@ -717,9 +810,9 @@
 	return false;
 }
 
-int usage_with_options_internal(const char * const *usagestr,
-				const struct option *opts, int full,
-				struct parse_opt_ctx_t *ctx)
+static int usage_with_options_internal(const char * const *usagestr,
+				       const struct option *opts, int full,
+				       struct parse_opt_ctx_t *ctx)
 {
 	struct option *ordered;
 
@@ -728,9 +821,9 @@
 
 	setup_pager();
 
-	if (strbuf_avail(&error_buf)) {
-		fprintf(stderr, "  Error: %s\n", error_buf.buf);
-		strbuf_release(&error_buf);
+	if (error_buf) {
+		fprintf(stderr, "  Error: %s\n", error_buf);
+		zfree(&error_buf);
 	}
 
 	fprintf(stderr, "\n Usage: %s\n", *usagestr++);
@@ -774,11 +867,15 @@
 			    const struct option *opts, const char *fmt, ...)
 {
 	va_list ap;
+	char *tmp = error_buf;
 
 	va_start(ap, fmt);
-	strbuf_addv(&error_buf, fmt, ap);
+	if (vasprintf(&error_buf, fmt, ap) == -1)
+		die("vasprintf failed");
 	va_end(ap);
 
+	free(tmp);
+
 	usage_with_options_internal(usagestr, opts, 0, NULL);
 	exit(129);
 }
@@ -848,15 +945,39 @@
 	return 0;
 }
 
-void set_option_flag(struct option *opts, int shortopt, const char *longopt,
-		     int flag)
+static struct option *
+find_option(struct option *opts, int shortopt, const char *longopt)
 {
 	for (; opts->type != OPTION_END; opts++) {
 		if ((shortopt && opts->short_name == shortopt) ||
 		    (opts->long_name && longopt &&
-		     !strcmp(opts->long_name, longopt))) {
-			opts->flags |= flag;
-			break;
-		}
+		     !strcmp(opts->long_name, longopt)))
+			return opts;
 	}
+	return NULL;
+}
+
+void set_option_flag(struct option *opts, int shortopt, const char *longopt,
+		     int flag)
+{
+	struct option *opt = find_option(opts, shortopt, longopt);
+
+	if (opt)
+		opt->flags |= flag;
+	return;
+}
+
+void set_option_nobuild(struct option *opts, int shortopt,
+			const char *longopt,
+			const char *build_opt,
+			bool can_skip)
+{
+	struct option *opt = find_option(opts, shortopt, longopt);
+
+	if (!opt)
+		return;
+
+	opt->flags |= PARSE_OPT_NOBUILD;
+	opt->flags |= can_skip ? PARSE_OPT_CANSKIP : 0;
+	opt->build_opt = build_opt;
 }
diff --git a/tools/perf/util/parse-options.h b/tools/lib/subcmd/parse-options.h
similarity index 94%
rename from tools/perf/util/parse-options.h
rename to tools/lib/subcmd/parse-options.h
index a8e407b..13a2cc1 100644
--- a/tools/perf/util/parse-options.h
+++ b/tools/lib/subcmd/parse-options.h
@@ -1,8 +1,8 @@
-#ifndef __PERF_PARSE_OPTIONS_H
-#define __PERF_PARSE_OPTIONS_H
+#ifndef __SUBCMD_PARSE_OPTIONS_H
+#define __SUBCMD_PARSE_OPTIONS_H
 
-#include <linux/kernel.h>
 #include <stdbool.h>
+#include <stdint.h>
 
 enum parse_opt_type {
 	/* special types */
@@ -41,6 +41,8 @@
 	PARSE_OPT_DISABLED = 32,
 	PARSE_OPT_EXCLUSIVE = 64,
 	PARSE_OPT_NOEMPTY  = 128,
+	PARSE_OPT_NOBUILD  = 256,
+	PARSE_OPT_CANSKIP  = 512,
 };
 
 struct option;
@@ -96,6 +98,7 @@
 	void *value;
 	const char *argh;
 	const char *help;
+	const char *build_opt;
 
 	int flags;
 	parse_opt_cb *callback;
@@ -149,6 +152,9 @@
 /* parse_options() will filter out the processed options and leave the
  * non-option argments in argv[].
  * Returns the number of arguments left in argv[].
+ *
+ * NOTE: parse_options() and parse_options_subcommand() may call exit() in the
+ * case of an error (or for 'special' options like --list-cmds or --list-opts).
  */
 extern int parse_options(int argc, const char **argv,
                          const struct option *options,
@@ -195,15 +201,6 @@
 			       const char *optstr,
 			       bool short_opt);
 
-extern void parse_options_start(struct parse_opt_ctx_t *ctx,
-				int argc, const char **argv, int flags);
-
-extern int parse_options_step(struct parse_opt_ctx_t *ctx,
-			      const struct option *options,
-			      const char * const usagestr[]);
-
-extern int parse_options_end(struct parse_opt_ctx_t *ctx);
-
 
 /*----- some often used options -----*/
 extern int parse_opt_abbrev_cb(const struct option *, const char *, int);
@@ -226,4 +223,7 @@
 extern const char *parse_options_fix_filename(const char *prefix, const char *file);
 
 void set_option_flag(struct option *opts, int sopt, const char *lopt, int flag);
-#endif /* __PERF_PARSE_OPTIONS_H */
+void set_option_nobuild(struct option *opts, int shortopt, const char *longopt,
+			const char *build_opt, bool can_skip);
+
+#endif /* __SUBCMD_PARSE_OPTIONS_H */
diff --git a/tools/perf/util/run-command.c b/tools/lib/subcmd/run-command.c
similarity index 90%
rename from tools/perf/util/run-command.c
rename to tools/lib/subcmd/run-command.c
index 34622b5..f4f6c9e 100644
--- a/tools/perf/util/run-command.c
+++ b/tools/lib/subcmd/run-command.c
@@ -1,7 +1,15 @@
-#include "cache.h"
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <string.h>
+#include <errno.h>
+#include <sys/wait.h>
+#include "subcmd-util.h"
 #include "run-command.h"
-#include "exec_cmd.h"
-#include "debug.h"
+#include "exec-cmd.h"
+
+#define STRERR_BUFSIZE 128
 
 static inline void close_pair(int fd[2])
 {
@@ -112,8 +120,8 @@
 		}
 		if (cmd->preexec_cb)
 			cmd->preexec_cb();
-		if (cmd->perf_cmd) {
-			execv_perf_cmd(cmd->argv);
+		if (cmd->exec_cmd) {
+			execv_cmd(cmd->argv);
 		} else {
 			execvp(cmd->argv[0], (char *const*) cmd->argv);
 		}
@@ -164,8 +172,8 @@
 		if (waiting < 0) {
 			if (errno == EINTR)
 				continue;
-			error("waitpid failed (%s)",
-			      strerror_r(errno, sbuf, sizeof(sbuf)));
+			fprintf(stderr, " Error: waitpid failed (%s)",
+				strerror_r(errno, sbuf, sizeof(sbuf)));
 			return -ERR_RUN_COMMAND_WAITPID;
 		}
 		if (waiting != pid)
@@ -207,7 +215,7 @@
 	memset(cmd, 0, sizeof(*cmd));
 	cmd->argv = argv;
 	cmd->no_stdin = opt & RUN_COMMAND_NO_STDIN ? 1 : 0;
-	cmd->perf_cmd = opt & RUN_PERF_CMD ? 1 : 0;
+	cmd->exec_cmd = opt & RUN_EXEC_CMD ? 1 : 0;
 	cmd->stdout_to_stderr = opt & RUN_COMMAND_STDOUT_TO_STDERR ? 1 : 0;
 }
 
diff --git a/tools/perf/util/run-command.h b/tools/lib/subcmd/run-command.h
similarity index 86%
rename from tools/perf/util/run-command.h
rename to tools/lib/subcmd/run-command.h
index 1ef264d..fe2befe 100644
--- a/tools/perf/util/run-command.h
+++ b/tools/lib/subcmd/run-command.h
@@ -1,5 +1,7 @@
-#ifndef __PERF_RUN_COMMAND_H
-#define __PERF_RUN_COMMAND_H
+#ifndef __SUBCMD_RUN_COMMAND_H
+#define __SUBCMD_RUN_COMMAND_H
+
+#include <unistd.h>
 
 enum {
 	ERR_RUN_COMMAND_FORK = 10000,
@@ -41,7 +43,7 @@
 	unsigned no_stdin:1;
 	unsigned no_stdout:1;
 	unsigned no_stderr:1;
-	unsigned perf_cmd:1; /* if this is to be perf sub-command */
+	unsigned exec_cmd:1; /* if this is to be external sub-command */
 	unsigned stdout_to_stderr:1;
 	void (*preexec_cb)(void);
 };
@@ -51,8 +53,8 @@
 int run_command(struct child_process *);
 
 #define RUN_COMMAND_NO_STDIN 1
-#define RUN_PERF_CMD	     2	/*If this is to be perf sub-command */
+#define RUN_EXEC_CMD	     2	/*If this is to be external sub-command */
 #define RUN_COMMAND_STDOUT_TO_STDERR 4
 int run_command_v_opt(const char **argv, int opt);
 
-#endif /* __PERF_RUN_COMMAND_H */
+#endif /* __SUBCMD_RUN_COMMAND_H */
diff --git a/tools/perf/util/sigchain.c b/tools/lib/subcmd/sigchain.c
similarity index 95%
rename from tools/perf/util/sigchain.c
rename to tools/lib/subcmd/sigchain.c
index ba785e9..3537c34 100644
--- a/tools/perf/util/sigchain.c
+++ b/tools/lib/subcmd/sigchain.c
@@ -1,5 +1,6 @@
+#include <signal.h>
+#include "subcmd-util.h"
 #include "sigchain.h"
-#include "cache.h"
 
 #define SIGCHAIN_MAX_SIGNALS 32
 
diff --git a/tools/lib/subcmd/sigchain.h b/tools/lib/subcmd/sigchain.h
new file mode 100644
index 0000000..0c919f2
--- /dev/null
+++ b/tools/lib/subcmd/sigchain.h
@@ -0,0 +1,10 @@
+#ifndef __SUBCMD_SIGCHAIN_H
+#define __SUBCMD_SIGCHAIN_H
+
+typedef void (*sigchain_fun)(int);
+
+int sigchain_pop(int sig);
+
+void sigchain_push_common(sigchain_fun f);
+
+#endif /* __SUBCMD_SIGCHAIN_H */
diff --git a/tools/lib/subcmd/subcmd-config.c b/tools/lib/subcmd/subcmd-config.c
new file mode 100644
index 0000000..d017c72
--- /dev/null
+++ b/tools/lib/subcmd/subcmd-config.c
@@ -0,0 +1,11 @@
+#include "subcmd-config.h"
+
+#define UNDEFINED "SUBCMD_HAS_NOT_BEEN_INITIALIZED"
+
+struct subcmd_config subcmd_config = {
+	.exec_name	= UNDEFINED,
+	.prefix		= UNDEFINED,
+	.exec_path	= UNDEFINED,
+	.exec_path_env	= UNDEFINED,
+	.pager_env	= UNDEFINED,
+};
diff --git a/tools/lib/subcmd/subcmd-config.h b/tools/lib/subcmd/subcmd-config.h
new file mode 100644
index 0000000..cc85140
--- /dev/null
+++ b/tools/lib/subcmd/subcmd-config.h
@@ -0,0 +1,14 @@
+#ifndef __PERF_SUBCMD_CONFIG_H
+#define __PERF_SUBCMD_CONFIG_H
+
+struct subcmd_config {
+	const char *exec_name;
+	const char *prefix;
+	const char *exec_path;
+	const char *exec_path_env;
+	const char *pager_env;
+};
+
+extern struct subcmd_config subcmd_config;
+
+#endif /* __PERF_SUBCMD_CONFIG_H */
diff --git a/tools/lib/subcmd/subcmd-util.h b/tools/lib/subcmd/subcmd-util.h
new file mode 100644
index 0000000..fc2e45d
--- /dev/null
+++ b/tools/lib/subcmd/subcmd-util.h
@@ -0,0 +1,91 @@
+#ifndef __SUBCMD_UTIL_H
+#define __SUBCMD_UTIL_H
+
+#include <stdarg.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NORETURN __attribute__((__noreturn__))
+
+static inline void report(const char *prefix, const char *err, va_list params)
+{
+	char msg[1024];
+	vsnprintf(msg, sizeof(msg), err, params);
+	fprintf(stderr, " %s%s\n", prefix, msg);
+}
+
+static NORETURN inline void die(const char *err, ...)
+{
+	va_list params;
+
+	va_start(params, err);
+	report(" Fatal: ", err, params);
+	exit(128);
+	va_end(params);
+}
+
+#define zfree(ptr) ({ free(*ptr); *ptr = NULL; })
+
+#define alloc_nr(x) (((x)+16)*3/2)
+
+/*
+ * Realloc the buffer pointed at by variable 'x' so that it can hold
+ * at least 'nr' entries; the number of entries currently allocated
+ * is 'alloc', using the standard growing factor alloc_nr() macro.
+ *
+ * DO NOT USE any expression with side-effect for 'x' or 'alloc'.
+ */
+#define ALLOC_GROW(x, nr, alloc) \
+	do { \
+		if ((nr) > alloc) { \
+			if (alloc_nr(alloc) < (nr)) \
+				alloc = (nr); \
+			else \
+				alloc = alloc_nr(alloc); \
+			x = xrealloc((x), alloc * sizeof(*(x))); \
+		} \
+	} while(0)
+
+static inline void *xrealloc(void *ptr, size_t size)
+{
+	void *ret = realloc(ptr, size);
+	if (!ret && !size)
+		ret = realloc(ptr, 1);
+	if (!ret) {
+		ret = realloc(ptr, size);
+		if (!ret && !size)
+			ret = realloc(ptr, 1);
+		if (!ret)
+			die("Out of memory, realloc failed");
+	}
+	return ret;
+}
+
+#define astrcatf(out, fmt, ...)						\
+({									\
+	char *tmp = *(out);						\
+	if (asprintf((out), "%s" fmt, tmp ?: "", ## __VA_ARGS__) == -1)	\
+		die("asprintf failed");					\
+	free(tmp);							\
+})
+
+static inline void astrcat(char **out, const char *add)
+{
+	char *tmp = *out;
+
+	if (asprintf(out, "%s%s", tmp ?: "", add) == -1)
+		die("asprintf failed");
+
+	free(tmp);
+}
+
+static inline int prefixcmp(const char *str, const char *prefix)
+{
+	for (; ; str++, prefix++)
+		if (!*prefix)
+			return 0;
+		else if (*str != *prefix)
+			return (unsigned char)*prefix - (unsigned char)*str;
+}
+
+#endif /* __SUBCMD_UTIL_H */
diff --git a/tools/perf/Build b/tools/perf/Build
index 2a41217..00c4b8c 100644
--- a/tools/perf/Build
+++ b/tools/perf/Build
@@ -36,7 +36,10 @@
 
 CFLAGS_builtin-help.o      += $(paths)
 CFLAGS_builtin-timechart.o += $(paths)
-CFLAGS_perf.o              += -DPERF_HTML_PATH="BUILD_STR($(htmldir_SQ))" -include $(OUTPUT)PERF-VERSION-FILE
+CFLAGS_perf.o              += -DPERF_HTML_PATH="BUILD_STR($(htmldir_SQ))"	\
+			      -DPERF_EXEC_PATH="BUILD_STR($(perfexecdir_SQ))"	\
+			      -DPREFIX="BUILD_STR($(prefix_SQ))"		\
+			      -include $(OUTPUT)PERF-VERSION-FILE
 CFLAGS_builtin-trace.o	   += -DSTRACE_GROUPS_DIR="BUILD_STR($(STRACE_GROUPS_DIR_SQ))"
 
 libperf-y += util/
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index e630a7d..3a1a32f 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -207,11 +207,23 @@
 In per-thread mode with inheritance mode on (default), samples are captured only when
 the thread executes on the designated CPUs. Default is to monitor all CPUs.
 
+-B::
+--no-buildid::
+Do not save the build ids of binaries in the perf.data files. This skips
+post processing after recording, which sometimes makes the final step in
+the recording process to take a long time, as it needs to process all
+events looking for mmap records. The downside is that it can misresolve
+symbols if the workload binaries used when recording get locally rebuilt
+or upgraded, because the only key available in this case is the
+pathname. You can also set the "record.build-id" config variable to
+'skip to have this behaviour permanently.
+
 -N::
 --no-buildid-cache::
 Do not update the buildid cache. This saves some overhead in situations
 where the information in the perf.data file (which includes buildids)
-is sufficient.
+is sufficient.  You can also set the "record.build-id" config variable to
+'no-cache' to have the same effect.
 
 -G name,...::
 --cgroup name,...::
@@ -314,11 +326,17 @@
 Record context switch events i.e. events of type PERF_RECORD_SWITCH or
 PERF_RECORD_SWITCH_CPU_WIDE.
 
---clang-path::
+--clang-path=PATH::
 Path to clang binary to use for compiling BPF scriptlets.
+(enabled when BPF support is on)
 
---clang-opt::
+--clang-opt=OPTIONS::
 Options passed to clang when compiling BPF scriptlets.
+(enabled when BPF support is on)
+
+--vmlinux=PATH::
+Specify vmlinux path which has debuginfo.
+(enabled when BPF prologue is on)
 
 SEE ALSO
 --------
diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST
index 2562eac..ce3932e 100644
--- a/tools/perf/MANIFEST
+++ b/tools/perf/MANIFEST
@@ -20,6 +20,7 @@
 tools/lib/bpf
 tools/lib/api
 tools/lib/bpf
+tools/lib/subcmd
 tools/lib/hweight.c
 tools/lib/rbtree.c
 tools/lib/string.c
diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf
index 929a32b..569fcf0 100644
--- a/tools/perf/Makefile.perf
+++ b/tools/perf/Makefile.perf
@@ -145,9 +145,10 @@
 STRIP   = strip
 AWK     = awk
 
-LIB_DIR          = $(srctree)/tools/lib/api/
+LIB_DIR		= $(srctree)/tools/lib/api/
 TRACE_EVENT_DIR = $(srctree)/tools/lib/traceevent/
-BPF_DIR = $(srctree)/tools/lib/bpf/
+BPF_DIR		= $(srctree)/tools/lib/bpf/
+SUBCMD_DIR	= $(srctree)/tools/lib/subcmd/
 
 # include config/Makefile by default and rule out
 # non-config cases
@@ -184,15 +185,17 @@
 ifneq ($(OUTPUT),)
   TE_PATH=$(OUTPUT)
   BPF_PATH=$(OUTPUT)
+  SUBCMD_PATH=$(OUTPUT)
 ifneq ($(subdir),)
-  LIB_PATH=$(OUTPUT)/../lib/api/
+  API_PATH=$(OUTPUT)/../lib/api/
 else
-  LIB_PATH=$(OUTPUT)
+  API_PATH=$(OUTPUT)
 endif
 else
   TE_PATH=$(TRACE_EVENT_DIR)
-  LIB_PATH=$(LIB_DIR)
+  API_PATH=$(LIB_DIR)
   BPF_PATH=$(BPF_DIR)
+  SUBCMD_PATH=$(SUBCMD_DIR)
 endif
 
 LIBTRACEEVENT = $(TE_PATH)libtraceevent.a
@@ -201,11 +204,13 @@
 LIBTRACEEVENT_DYNAMIC_LIST = $(TE_PATH)libtraceevent-dynamic-list
 LIBTRACEEVENT_DYNAMIC_LIST_LDFLAGS = -Xlinker --dynamic-list=$(LIBTRACEEVENT_DYNAMIC_LIST)
 
-LIBAPI = $(LIB_PATH)libapi.a
+LIBAPI = $(API_PATH)libapi.a
 export LIBAPI
 
 LIBBPF = $(BPF_PATH)libbpf.a
 
+LIBSUBCMD = $(SUBCMD_PATH)libsubcmd.a
+
 # python extension build directories
 PYTHON_EXTBUILD     := $(OUTPUT)python_ext_build/
 PYTHON_EXTBUILD_LIB := $(PYTHON_EXTBUILD)lib/
@@ -257,7 +262,7 @@
 
 LIB_FILE=$(OUTPUT)libperf.a
 
-PERFLIBS = $(LIB_FILE) $(LIBAPI) $(LIBTRACEEVENT)
+PERFLIBS = $(LIB_FILE) $(LIBAPI) $(LIBTRACEEVENT) $(LIBSUBCMD)
 ifndef NO_LIBBPF
   PERFLIBS += $(LIBBPF)
 endif
@@ -437,6 +442,13 @@
 	$(call QUIET_CLEAN, libbpf)
 	$(Q)$(MAKE) -C $(BPF_DIR) O=$(OUTPUT) clean >/dev/null
 
+$(LIBSUBCMD): fixdep FORCE
+	$(Q)$(MAKE) -C $(SUBCMD_DIR) O=$(OUTPUT) $(OUTPUT)libsubcmd.a
+
+$(LIBSUBCMD)-clean:
+	$(call QUIET_CLEAN, libsubcmd)
+	$(Q)$(MAKE) -C $(SUBCMD_DIR) O=$(OUTPUT) clean
+
 help:
 	@echo 'Perf make targets:'
 	@echo '  doc		- make *all* documentation (see below)'
@@ -582,15 +594,16 @@
 #
 config-clean:
 	$(call QUIET_CLEAN, config)
-	$(Q)$(MAKE) -C $(srctree)/tools/build/feature/ clean >/dev/null
+	$(Q)$(MAKE) -C $(srctree)/tools/build/feature/ $(if $(OUTPUT),OUTPUT=$(OUTPUT)feature/,) clean >/dev/null
 
-clean: $(LIBTRACEEVENT)-clean $(LIBAPI)-clean $(LIBBPF)-clean config-clean
+clean: $(LIBTRACEEVENT)-clean $(LIBAPI)-clean $(LIBBPF)-clean $(LIBSUBCMD)-clean config-clean
 	$(call QUIET_CLEAN, core-objs)  $(RM) $(LIB_FILE) $(OUTPUT)perf-archive $(OUTPUT)perf-with-kcore $(LANG_BINDINGS)
-	$(Q)find . -name '*.o' -delete -o -name '\.*.cmd' -delete -o -name '\.*.d' -delete
+	$(Q)find $(if $(OUTPUT),$(OUTPUT),.) -name '*.o' -delete -o -name '\.*.cmd' -delete -o -name '\.*.d' -delete
 	$(Q)$(RM) $(OUTPUT).config-detected
 	$(call QUIET_CLEAN, core-progs) $(RM) $(ALL_PROGRAMS) perf perf-read-vdso32 perf-read-vdsox32
 	$(call QUIET_CLEAN, core-gen)   $(RM)  *.spec *.pyc *.pyo */*.pyc */*.pyo $(OUTPUT)common-cmds.h TAGS tags cscope* $(OUTPUT)PERF-VERSION-FILE $(OUTPUT)FEATURE-DUMP $(OUTPUT)util/*-bison* $(OUTPUT)util/*-flex* \
-		$(OUTPUT)util/intel-pt-decoder/inat-tables.c
+		$(OUTPUT)util/intel-pt-decoder/inat-tables.c $(OUTPUT)fixdep \
+		$(OUTPUT)tests/llvm-src-{base,kbuild,prologue}.c
 	$(QUIET_SUBDIR0)Documentation $(QUIET_SUBDIR1) clean
 	$(python-clean)
 
diff --git a/tools/perf/arch/x86/util/intel-pt.c b/tools/perf/arch/x86/util/intel-pt.c
index b02af06..b64d462 100644
--- a/tools/perf/arch/x86/util/intel-pt.c
+++ b/tools/perf/arch/x86/util/intel-pt.c
@@ -26,7 +26,7 @@
 #include "../../util/evlist.h"
 #include "../../util/evsel.h"
 #include "../../util/cpumap.h"
-#include "../../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../../util/parse-events.h"
 #include "../../util/pmu.h"
 #include "../../util/debug.h"
diff --git a/tools/perf/bench/futex-hash.c b/tools/perf/bench/futex-hash.c
index fc9bebd..0999ac5 100644
--- a/tools/perf/bench/futex-hash.c
+++ b/tools/perf/bench/futex-hash.c
@@ -11,7 +11,7 @@
 #include "../perf.h"
 #include "../util/util.h"
 #include "../util/stat.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "bench.h"
 #include "futex.h"
diff --git a/tools/perf/bench/futex-lock-pi.c b/tools/perf/bench/futex-lock-pi.c
index bc6a16a..6a18ce2 100644
--- a/tools/perf/bench/futex-lock-pi.c
+++ b/tools/perf/bench/futex-lock-pi.c
@@ -5,7 +5,7 @@
 #include "../perf.h"
 #include "../util/util.h"
 #include "../util/stat.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "bench.h"
 #include "futex.h"
diff --git a/tools/perf/bench/futex-requeue.c b/tools/perf/bench/futex-requeue.c
index ad0d9b5..7182386 100644
--- a/tools/perf/bench/futex-requeue.c
+++ b/tools/perf/bench/futex-requeue.c
@@ -11,7 +11,7 @@
 #include "../perf.h"
 #include "../util/util.h"
 #include "../util/stat.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "bench.h"
 #include "futex.h"
diff --git a/tools/perf/bench/futex-wake-parallel.c b/tools/perf/bench/futex-wake-parallel.c
index 6d8c9fa..91aaf2a 100644
--- a/tools/perf/bench/futex-wake-parallel.c
+++ b/tools/perf/bench/futex-wake-parallel.c
@@ -10,7 +10,7 @@
 #include "../perf.h"
 #include "../util/util.h"
 #include "../util/stat.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "bench.h"
 #include "futex.h"
diff --git a/tools/perf/bench/futex-wake.c b/tools/perf/bench/futex-wake.c
index e5e41d3..f416bd7 100644
--- a/tools/perf/bench/futex-wake.c
+++ b/tools/perf/bench/futex-wake.c
@@ -11,7 +11,7 @@
 #include "../perf.h"
 #include "../util/util.h"
 #include "../util/stat.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "bench.h"
 #include "futex.h"
diff --git a/tools/perf/bench/mem-functions.c b/tools/perf/bench/mem-functions.c
index 9419b94..a91aa85 100644
--- a/tools/perf/bench/mem-functions.c
+++ b/tools/perf/bench/mem-functions.c
@@ -8,7 +8,7 @@
 
 #include "../perf.h"
 #include "../util/util.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/header.h"
 #include "../util/cloexec.h"
 #include "bench.h"
diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c
index 492df27..5049d63 100644
--- a/tools/perf/bench/numa.c
+++ b/tools/perf/bench/numa.c
@@ -7,7 +7,7 @@
 #include "../perf.h"
 #include "../builtin.h"
 #include "../util/util.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../util/cloexec.h"
 
 #include "bench.h"
diff --git a/tools/perf/bench/sched-messaging.c b/tools/perf/bench/sched-messaging.c
index d4ff1b5..bfaf950 100644
--- a/tools/perf/bench/sched-messaging.c
+++ b/tools/perf/bench/sched-messaging.c
@@ -11,7 +11,7 @@
 
 #include "../perf.h"
 #include "../util/util.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../builtin.h"
 #include "bench.h"
 
diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c
index 005cc28..1dc2d13 100644
--- a/tools/perf/bench/sched-pipe.c
+++ b/tools/perf/bench/sched-pipe.c
@@ -10,7 +10,7 @@
  */
 #include "../perf.h"
 #include "../util/util.h"
-#include "../util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "../builtin.h"
 #include "bench.h"
 
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 1f00dc7..e18f1b9 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -21,7 +21,7 @@
 #include "util/evsel.h"
 #include "util/annotate.h"
 #include "util/event.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 #include "util/thread.h"
 #include "util/sort.h"
diff --git a/tools/perf/builtin-bench.c b/tools/perf/builtin-bench.c
index b17aed3..a1cddc6 100644
--- a/tools/perf/builtin-bench.c
+++ b/tools/perf/builtin-bench.c
@@ -16,7 +16,7 @@
  */
 #include "perf.h"
 #include "util/util.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "builtin.h"
 #include "bench/bench.h"
 
diff --git a/tools/perf/builtin-buildid-cache.c b/tools/perf/builtin-buildid-cache.c
index 7b8450c..d93bff7 100644
--- a/tools/perf/builtin-buildid-cache.c
+++ b/tools/perf/builtin-buildid-cache.c
@@ -16,7 +16,7 @@
 #include "util/cache.h"
 #include "util/debug.h"
 #include "util/header.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/strlist.h"
 #include "util/build-id.h"
 #include "util/session.h"
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
index 6419f57..5e914ee 100644
--- a/tools/perf/builtin-buildid-list.c
+++ b/tools/perf/builtin-buildid-list.c
@@ -12,7 +12,7 @@
 #include "util/build-id.h"
 #include "util/cache.h"
 #include "util/debug.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/session.h"
 #include "util/symbol.h"
 #include "util/data.h"
diff --git a/tools/perf/builtin-config.c b/tools/perf/builtin-config.c
index 427ea7a..f04e804 100644
--- a/tools/perf/builtin-config.c
+++ b/tools/perf/builtin-config.c
@@ -9,7 +9,7 @@
 #include "perf.h"
 
 #include "util/cache.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/util.h"
 #include "util/debug.h"
 
diff --git a/tools/perf/builtin-data.c b/tools/perf/builtin-data.c
index d6525bc..b97bc15 100644
--- a/tools/perf/builtin-data.c
+++ b/tools/perf/builtin-data.c
@@ -2,7 +2,7 @@
 #include "builtin.h"
 #include "perf.h"
 #include "debug.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 #include "data-convert-bt.h"
 
 typedef int (*data_cmd_fn_t)(int argc, const char **argv, const char *prefix);
diff --git a/tools/perf/builtin-evlist.c b/tools/perf/builtin-evlist.c
index f4d6251..08a7d36 100644
--- a/tools/perf/builtin-evlist.c
+++ b/tools/perf/builtin-evlist.c
@@ -12,7 +12,7 @@
 #include "util/evlist.h"
 #include "util/evsel.h"
 #include "util/parse-events.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/session.h"
 #include "util/data.h"
 #include "util/debug.h"
diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
index a7d588b..96c1a4c 100644
--- a/tools/perf/builtin-help.c
+++ b/tools/perf/builtin-help.c
@@ -6,11 +6,11 @@
 #include "perf.h"
 #include "util/cache.h"
 #include "builtin.h"
-#include "util/exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "common-cmds.h"
-#include "util/parse-options.h"
-#include "util/run-command.h"
-#include "util/help.h"
+#include <subcmd/parse-options.h>
+#include <subcmd/run-command.h>
+#include <subcmd/help.h>
 #include "util/debug.h"
 
 static struct man_viewer_list {
@@ -407,7 +407,7 @@
 #ifndef open_html
 static void open_html(const char *path)
 {
-	execl_perf_cmd("web--browse", "-c", "help.browser", path, NULL);
+	execl_cmd("web--browse", "-c", "help.browser", path, NULL);
 }
 #endif
 
diff --git a/tools/perf/builtin-inject.c b/tools/perf/builtin-inject.c
index 99d127f..0022e02 100644
--- a/tools/perf/builtin-inject.c
+++ b/tools/perf/builtin-inject.c
@@ -18,7 +18,7 @@
 #include "util/data.h"
 #include "util/auxtrace.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 
 #include <linux/list.h>
 
diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c
index 93ce665..1180105 100644
--- a/tools/perf/builtin-kmem.c
+++ b/tools/perf/builtin-kmem.c
@@ -12,7 +12,7 @@
 #include "util/tool.h"
 #include "util/callchain.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/trace-event.h"
 #include "util/data.h"
 #include "util/cpumap.h"
diff --git a/tools/perf/builtin-kvm.c b/tools/perf/builtin-kvm.c
index 031f9f5..4418d92 100644
--- a/tools/perf/builtin-kvm.c
+++ b/tools/perf/builtin-kvm.c
@@ -10,7 +10,7 @@
 #include "util/header.h"
 #include "util/session.h"
 #include "util/intlist.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/trace-event.h"
 #include "util/debug.h"
 #include "util/tool.h"
diff --git a/tools/perf/builtin-list.c b/tools/perf/builtin-list.c
index bf679e2..5e22db4 100644
--- a/tools/perf/builtin-list.c
+++ b/tools/perf/builtin-list.c
@@ -14,7 +14,7 @@
 #include "util/parse-events.h"
 #include "util/cache.h"
 #include "util/pmu.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 
 int cmd_list(int argc, const char **argv, const char *prefix __maybe_unused)
 {
diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c
index de16aae..ce3bfb4 100644
--- a/tools/perf/builtin-lock.c
+++ b/tools/perf/builtin-lock.c
@@ -9,7 +9,7 @@
 #include "util/thread.h"
 #include "util/header.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/trace-event.h"
 
 #include "util/debug.h"
diff --git a/tools/perf/builtin-mem.c b/tools/perf/builtin-mem.c
index 80170aa..3901700 100644
--- a/tools/perf/builtin-mem.c
+++ b/tools/perf/builtin-mem.c
@@ -1,7 +1,7 @@
 #include "builtin.h"
 #include "perf.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/trace-event.h"
 #include "util/tool.h"
 #include "util/session.h"
diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c
index 132afc9..9af859b 100644
--- a/tools/perf/builtin-probe.c
+++ b/tools/perf/builtin-probe.c
@@ -37,7 +37,7 @@
 #include "util/strfilter.h"
 #include "util/symbol.h"
 #include "util/debug.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/probe-finder.h"
 #include "util/probe-event.h"
 #include "util/probe-file.h"
@@ -249,6 +249,9 @@
 
 	return ret;
 }
+#else
+# define opt_show_lines NULL
+# define opt_show_vars NULL
 #endif
 static int opt_add_probe_event(const struct option *opt,
 			      const char *str, int unset __maybe_unused)
@@ -473,7 +476,6 @@
 		opt_add_probe_event),
 	OPT_BOOLEAN('f', "force", &probe_conf.force_add, "forcibly add events"
 		    " with existing name"),
-#ifdef HAVE_DWARF_SUPPORT
 	OPT_CALLBACK('L', "line", NULL,
 		     "FUNC[:RLN[+NUM|-RLN2]]|SRC:ALN[+NUM|-ALN2]",
 		     "Show source code lines.", opt_show_lines),
@@ -490,7 +492,6 @@
 		   "directory", "path to kernel source"),
 	OPT_BOOLEAN('\0', "no-inlines", &probe_conf.no_inlines,
 		"Don't search inlined functions"),
-#endif
 	OPT__DRY_RUN(&probe_event_dry_run),
 	OPT_INTEGER('\0', "max-probes", &probe_conf.max_probes,
 		 "Set how many probe points can be found for a probe."),
@@ -521,6 +522,16 @@
 #ifdef HAVE_DWARF_SUPPORT
 	set_option_flag(options, 'L', "line", PARSE_OPT_EXCLUSIVE);
 	set_option_flag(options, 'V', "vars", PARSE_OPT_EXCLUSIVE);
+#else
+# define set_nobuild(s, l, c) set_option_nobuild(options, s, l, "NO_DWARF=1", c)
+	set_nobuild('L', "line", false);
+	set_nobuild('V', "vars", false);
+	set_nobuild('\0', "externs", false);
+	set_nobuild('\0', "range", false);
+	set_nobuild('k', "vmlinux", true);
+	set_nobuild('s', "source", true);
+	set_nobuild('\0', "no-inlines", true);
+# undef set_nobuild
 #endif
 	set_option_flag(options, 'F', "funcs", PARSE_OPT_EXCLUSIVE);
 
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 199fc31..1435ef6 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -11,7 +11,7 @@
 
 #include "util/build-id.h"
 #include "util/util.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 
 #include "util/callchain.h"
@@ -837,6 +837,19 @@
 
 static int perf_record_config(const char *var, const char *value, void *cb)
 {
+	struct record *rec = cb;
+
+	if (!strcmp(var, "record.build-id")) {
+		if (!strcmp(value, "cache"))
+			rec->no_buildid_cache = false;
+		else if (!strcmp(value, "no-cache"))
+			rec->no_buildid_cache = true;
+		else if (!strcmp(value, "skip"))
+			rec->no_buildid = true;
+		else
+			return -1;
+		return 0;
+	}
 	if (!strcmp(var, "record.call-graph"))
 		var = "call-graph.record-mode"; /* fall-through */
 
@@ -1113,12 +1126,12 @@
 			"per thread proc mmap processing timeout in ms"),
 	OPT_BOOLEAN(0, "switch-events", &record.opts.record_switch_events,
 		    "Record context switch events"),
-#ifdef HAVE_LIBBPF_SUPPORT
 	OPT_STRING(0, "clang-path", &llvm_param.clang_path, "clang path",
 		   "clang binary to use for compiling BPF scriptlets"),
 	OPT_STRING(0, "clang-opt", &llvm_param.clang_opt, "clang options",
 		   "options passed to clang when compiling BPF scriptlets"),
-#endif
+	OPT_STRING(0, "vmlinux", &symbol_conf.vmlinux_name,
+		   "file", "vmlinux pathname"),
 	OPT_END()
 };
 
@@ -1130,6 +1143,27 @@
 	struct record *rec = &record;
 	char errbuf[BUFSIZ];
 
+#ifndef HAVE_LIBBPF_SUPPORT
+# define set_nobuild(s, l, c) set_option_nobuild(record_options, s, l, "NO_LIBBPF=1", c)
+	set_nobuild('\0', "clang-path", true);
+	set_nobuild('\0', "clang-opt", true);
+# undef set_nobuild
+#endif
+
+#ifndef HAVE_BPF_PROLOGUE
+# if !defined (HAVE_DWARF_SUPPORT)
+#  define REASON  "NO_DWARF=1"
+# elif !defined (HAVE_LIBBPF_SUPPORT)
+#  define REASON  "NO_LIBBPF=1"
+# else
+#  define REASON  "this architecture doesn't support BPF prologue"
+# endif
+# define set_nobuild(s, l, c) set_option_nobuild(record_options, s, l, REASON, c)
+	set_nobuild('\0', "vmlinux", true);
+# undef set_nobuild
+# undef REASON
+#endif
+
 	rec->evlist = perf_evlist__new();
 	if (rec->evlist == NULL)
 		return -ENOMEM;
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 5a45466..2a7330b 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -27,7 +27,7 @@
 #include "util/session.h"
 #include "util/tool.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 
 #include "util/thread.h"
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index e3d3e32..871b55ae 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -12,7 +12,7 @@
 #include "util/tool.h"
 #include "util/cloexec.h"
 
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/trace-event.h"
 
 #include "util/debug.h"
diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index d259e9a..bcc3542 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -3,9 +3,9 @@
 #include "perf.h"
 #include "util/cache.h"
 #include "util/debug.h"
-#include "util/exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "util/header.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/perf_regs.h"
 #include "util/session.h"
 #include "util/tool.h"
@@ -1408,7 +1408,7 @@
 	char first_half[BUFSIZ];
 	char *script_root;
 
-	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
+	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", get_argv_exec_path());
 
 	scripts_dir = opendir(scripts_path);
 	if (!scripts_dir)
@@ -1529,7 +1529,7 @@
 	if (!session)
 		return -1;
 
-	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
+	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", get_argv_exec_path());
 
 	scripts_dir = opendir(scripts_path);
 	if (!scripts_dir) {
@@ -1587,7 +1587,7 @@
 	char lang_path[MAXPATHLEN];
 	char *__script_root;
 
-	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
+	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", get_argv_exec_path());
 
 	scripts_dir = opendir(scripts_path);
 	if (!scripts_dir)
@@ -1823,7 +1823,7 @@
 		scripting_max_stack = itrace_synth_opts.callchain_sz;
 
 	/* make sure PERF_EXEC_PATH is set for scripts */
-	perf_set_argv_exec_path(perf_exec_path());
+	set_argv_exec_path(get_argv_exec_path());
 
 	if (argc && !script_name && !rec_script_path && !rep_script_path) {
 		int live_pipe[2];
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 25a95f4..bbf42ee 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -45,7 +45,7 @@
 #include "builtin.h"
 #include "util/cgroup.h"
 #include "util/util.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 #include "util/pmu.h"
 #include "util/event.h"
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 30e5962..bd7a775 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -30,7 +30,7 @@
 
 #include "perf.h"
 #include "util/header.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 #include "util/event.h"
 #include "util/session.h"
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index 92fe963..9ebd67a 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -34,7 +34,7 @@
 #include "util/top.h"
 #include "util/util.h"
 #include <linux/rbtree.h>
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-events.h"
 #include "util/cpumap.h"
 #include "util/xyarray.h"
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index c783d8f..20916dd 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -22,11 +22,11 @@
 #include "util/color.h"
 #include "util/debug.h"
 #include "util/evlist.h"
-#include "util/exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "util/machine.h"
 #include "util/session.h"
 #include "util/thread.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/strlist.h"
 #include "util/intlist.h"
 #include "util/thread_map.h"
diff --git a/tools/perf/perf.c b/tools/perf/perf.c
index 59ea48c..cb1d249 100644
--- a/tools/perf/perf.c
+++ b/tools/perf/perf.c
@@ -9,12 +9,12 @@
 #include "builtin.h"
 
 #include "util/env.h"
-#include "util/exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "util/cache.h"
 #include "util/quote.h"
-#include "util/run-command.h"
+#include <subcmd/run-command.h>
 #include "util/parse-events.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/bpf-loader.h"
 #include "util/debug.h"
 #include <api/fs/tracing_path.h>
@@ -119,7 +119,7 @@
 {
 	switch (use_pager) {
 	case 0:
-		setenv("PERF_PAGER", "cat", 1);
+		setenv(PERF_PAGER_ENVIRONMENT, "cat", 1);
 		break;
 	case 1:
 		/* setup_pager(); */
@@ -183,9 +183,9 @@
 		if (!prefixcmp(cmd, CMD_EXEC_PATH)) {
 			cmd += strlen(CMD_EXEC_PATH);
 			if (*cmd == '=')
-				perf_set_argv_exec_path(cmd + 1);
+				set_argv_exec_path(cmd + 1);
 			else {
-				puts(perf_exec_path());
+				puts(get_argv_exec_path());
 				exit(0);
 			}
 		} else if (!strcmp(cmd, "--html-path")) {
@@ -530,11 +530,15 @@
 	const char *cmd;
 	char sbuf[STRERR_BUFSIZE];
 
+	/* libsubcmd init */
+	exec_cmd_init("perf", PREFIX, PERF_EXEC_PATH, EXEC_PATH_ENVIRONMENT);
+	pager_init(PERF_PAGER_ENVIRONMENT);
+
 	/* The page_size is placed in util object. */
 	page_size = sysconf(_SC_PAGE_SIZE);
 	cacheline_size = sysconf(_SC_LEVEL1_DCACHE_LINESIZE);
 
-	cmd = perf_extract_argv0_path(argv[0]);
+	cmd = extract_argv0_path(argv[0]);
 	if (!cmd)
 		cmd = "perf-help";
 
diff --git a/tools/perf/tests/Build b/tools/perf/tests/Build
index 0ff8a97..f23fb7e 100644
--- a/tools/perf/tests/Build
+++ b/tools/perf/tests/Build
@@ -35,21 +35,21 @@
 perf-y += bpf.o
 perf-y += topology.o
 
-$(OUTPUT)tests/llvm-src-base.c: tests/bpf-script-example.c
+$(OUTPUT)tests/llvm-src-base.c: tests/bpf-script-example.c tests/Build
 	$(call rule_mkdir)
 	$(Q)echo '#include <tests/llvm.h>' > $@
 	$(Q)echo 'const char test_llvm__bpf_base_prog[] =' >> $@
 	$(Q)sed -e 's/"/\\"/g' -e 's/\(.*\)/"\1\\n"/g' $< >> $@
 	$(Q)echo ';' >> $@
 
-$(OUTPUT)tests/llvm-src-kbuild.c: tests/bpf-script-test-kbuild.c
+$(OUTPUT)tests/llvm-src-kbuild.c: tests/bpf-script-test-kbuild.c tests/Build
 	$(call rule_mkdir)
 	$(Q)echo '#include <tests/llvm.h>' > $@
 	$(Q)echo 'const char test_llvm__bpf_test_kbuild_prog[] =' >> $@
 	$(Q)sed -e 's/"/\\"/g' -e 's/\(.*\)/"\1\\n"/g' $< >> $@
 	$(Q)echo ';' >> $@
 
-$(OUTPUT)tests/llvm-src-prologue.c: tests/bpf-script-test-prologue.c
+$(OUTPUT)tests/llvm-src-prologue.c: tests/bpf-script-test-prologue.c tests/Build
 	$(call rule_mkdir)
 	$(Q)echo '#include <tests/llvm.h>' > $@
 	$(Q)echo 'const char test_llvm__bpf_test_prologue_prog[] =' >> $@
diff --git a/tools/perf/tests/attr.c b/tools/perf/tests/attr.c
index b66730eb9..28d1605 100644
--- a/tools/perf/tests/attr.c
+++ b/tools/perf/tests/attr.c
@@ -24,7 +24,7 @@
 #include <linux/kernel.h>
 #include "../perf.h"
 #include "util.h"
-#include "exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "tests.h"
 
 #define ENV "PERF_TEST_ATTR"
@@ -164,7 +164,7 @@
 		return run_dir("./tests", "./perf");
 
 	/* Then installed path. */
-	snprintf(path_dir,  PATH_MAX, "%s/tests", perf_exec_path());
+	snprintf(path_dir,  PATH_MAX, "%s/tests", get_argv_exec_path());
 	snprintf(path_perf, PATH_MAX, "%s/perf", BINDIR);
 
 	if (!lstat(path_dir, &st) &&
diff --git a/tools/perf/tests/builtin-test.c b/tools/perf/tests/builtin-test.c
index 2b1ade1..0372d59 100644
--- a/tools/perf/tests/builtin-test.c
+++ b/tools/perf/tests/builtin-test.c
@@ -11,7 +11,7 @@
 #include "tests.h"
 #include "debug.h"
 #include "color.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 #include "symbol.h"
 
 struct test __weak arch_tests[] = {
@@ -236,6 +236,9 @@
 				dup2(STDOUT_FILENO, STDERR_FILENO);
 				close(nullfd);
 			}
+		} else {
+			signal(SIGSEGV, sighandler_dump_stack);
+			signal(SIGFPE, sighandler_dump_stack);
 		}
 
 		err = test->func(subtest);
diff --git a/tools/perf/tests/hists_common.c b/tools/perf/tests/hists_common.c
index 46f453b..bcfd081 100644
--- a/tools/perf/tests/hists_common.c
+++ b/tools/perf/tests/hists_common.c
@@ -88,8 +88,8 @@
 	}
 
 	if (machine__create_kernel_maps(machine)) {
-		pr_debug("Not enough memory for machine setup\n");
-		goto out;
+		pr_debug("Cannot create kernel maps\n");
+		return NULL;
 	}
 
 	for (i = 0; i < ARRAY_SIZE(fake_threads); i++) {
@@ -155,7 +155,6 @@
 out:
 	pr_debug("Not enough memory for machine setup\n");
 	machine__delete_threads(machine);
-	machine__delete(machine);
 	return NULL;
 }
 
diff --git a/tools/perf/tests/make b/tools/perf/tests/make
index 8ea3dff..c1fbb8e 100644
--- a/tools/perf/tests/make
+++ b/tools/perf/tests/make
@@ -259,7 +259,8 @@
 tarpkg:
 	@cmd="$(PERF)/tests/perf-targz-src-pkg $(PERF)"; \
 	echo "- $@: $$cmd" && echo $$cmd > $@ && \
-	( eval $$cmd ) >> $@ 2>&1
+	( eval $$cmd ) >> $@ 2>&1 && \
+	rm -f $@
 
 make_kernelsrc:
 	@echo "- make -C <kernelsrc> tools/perf"
diff --git a/tools/perf/util/Build b/tools/perf/util/Build
index 65fef59..94b1099 100644
--- a/tools/perf/util/Build
+++ b/tools/perf/util/Build
@@ -9,13 +9,10 @@
 libperf-y += event.o
 libperf-y += evlist.o
 libperf-y += evsel.o
-libperf-y += exec_cmd.o
 libperf-y += find_next_bit.o
-libperf-y += help.o
 libperf-y += kallsyms.o
 libperf-y += levenshtein.o
 libperf-y += llvm-utils.o
-libperf-y += parse-options.o
 libperf-y += parse-events.o
 libperf-y += perf_regs.o
 libperf-y += path.o
@@ -23,7 +20,6 @@
 libperf-y += libstring.o
 libperf-y += bitmap.o
 libperf-y += hweight.o
-libperf-y += run-command.o
 libperf-y += quote.o
 libperf-y += strbuf.o
 libperf-y += string.o
@@ -32,11 +28,9 @@
 libperf-y += top.o
 libperf-y += usage.o
 libperf-y += wrapper.o
-libperf-y += sigchain.o
 libperf-y += dso.o
 libperf-y += symbol.o
 libperf-y += color.o
-libperf-y += pager.o
 libperf-y += header.o
 libperf-y += callchain.o
 libperf-y += values.o
@@ -87,6 +81,7 @@
 libperf-y += parse-branch-options.o
 libperf-y += parse-regs-options.o
 libperf-y += term.o
+libperf-y += help-unknown-cmd.o
 
 libperf-$(CONFIG_LIBBPF) += bpf-loader.o
 libperf-$(CONFIG_BPF_PROLOGUE) += bpf-prologue.o
@@ -112,7 +107,6 @@
 libperf-$(CONFIG_LZMA) += lzma.o
 
 CFLAGS_config.o   += -DETC_PERFCONFIG="BUILD_STR($(ETC_PERFCONFIG_SQ))"
-CFLAGS_exec_cmd.o += -DPERF_EXEC_PATH="BUILD_STR($(perfexecdir_SQ))" -DPREFIX="BUILD_STR($(prefix_SQ))"
 
 $(OUTPUT)util/parse-events-flex.c: util/parse-events.l $(OUTPUT)util/parse-events-bison.c
 	$(call rule_mkdir)
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index 7f10430a..360fda0 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -45,7 +45,7 @@
 #include "event.h"
 #include "session.h"
 #include "debug.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 
 #include "intel-pt.h"
 #include "intel-bts.h"
diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h
index 4c2b764..07b5d63 100644
--- a/tools/perf/util/cache.h
+++ b/tools/perf/util/cache.h
@@ -4,9 +4,12 @@
 #include <stdbool.h>
 #include "util.h"
 #include "strbuf.h"
+#include <subcmd/pager.h>
 #include "../perf.h"
 #include "../ui/ui.h"
 
+#include <linux/string.h>
+
 #define CMD_EXEC_PATH "--exec-path"
 #define CMD_PERF_DIR "--perf-dir="
 #define CMD_WORK_TREE "--work-tree="
@@ -18,6 +21,7 @@
 #define DEFAULT_PERF_DIR_ENVIRONMENT ".perf"
 #define PERF_DEBUGFS_ENVIRONMENT "PERF_DEBUGFS_DIR"
 #define PERF_TRACEFS_ENVIRONMENT "PERF_TRACEFS_DIR"
+#define PERF_PAGER_ENVIRONMENT "PERF_PAGER"
 
 typedef int (*config_fn_t)(const char *, const char *, void *);
 extern int perf_default_config(const char *, const char *, void *);
@@ -28,10 +32,6 @@
 extern int config_error_nonbool(const char *);
 extern const char *perf_config_dirname(const char *, const char *);
 
-/* pager.c */
-extern void setup_pager(void);
-extern int pager_in_use(void);
-
 char *alias_lookup(const char *alias);
 int split_cmdline(char *cmdline, const char ***argv);
 
@@ -70,9 +70,4 @@
 extern char *perf_pathdup(const char *fmt, ...)
 	__attribute__((format (printf, 1, 2)));
 
-#ifndef __UCLIBC__
-/* Matches the libc/libbsd function attribute so we declare this unconditionally: */
-extern size_t strlcpy(char *dest, const char *src, size_t size);
-#endif
-
 #endif /* __PERF_CACHE_H */
diff --git a/tools/perf/util/cgroup.c b/tools/perf/util/cgroup.c
index 32e12ecf..90aa1b4 100644
--- a/tools/perf/util/cgroup.c
+++ b/tools/perf/util/cgroup.c
@@ -1,6 +1,6 @@
 #include "util.h"
 #include "../perf.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 #include "evsel.h"
 #include "cgroup.h"
 #include "evlist.h"
diff --git a/tools/perf/util/config.c b/tools/perf/util/config.c
index 2e452ac..d3e12e3 100644
--- a/tools/perf/util/config.c
+++ b/tools/perf/util/config.c
@@ -10,7 +10,7 @@
  */
 #include "util.h"
 #include "cache.h"
-#include "exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "util/hist.h"  /* perf_hist_config */
 #include "util/llvm-utils.h"   /* perf_llvm_config */
 
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index d1b6c20..8c44aad 100644
--- a/tools/perf/util/evlist.c
+++ b/tools/perf/util/evlist.c
@@ -18,7 +18,7 @@
 #include <unistd.h>
 
 #include "parse-events.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 
 #include <sys/mman.h>
 
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index 47f0330..544e440 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -36,6 +36,7 @@
 	bool cloexec;
 	bool clockid;
 	bool clockid_wrong;
+	bool lbr_flags;
 } perf_missing_features;
 
 static clockid_t clockid;
@@ -574,7 +575,9 @@
 			} else {
 				perf_evsel__set_sample_bit(evsel, BRANCH_STACK);
 				attr->branch_sample_type = PERF_SAMPLE_BRANCH_USER |
-							PERF_SAMPLE_BRANCH_CALL_STACK;
+							PERF_SAMPLE_BRANCH_CALL_STACK |
+							PERF_SAMPLE_BRANCH_NO_CYCLES |
+							PERF_SAMPLE_BRANCH_NO_FLAGS;
 			}
 		} else
 			 pr_warning("Cannot use LBR callstack with branch stack. "
@@ -1337,6 +1340,9 @@
 		evsel->attr.mmap2 = 0;
 	if (perf_missing_features.exclude_guest)
 		evsel->attr.exclude_guest = evsel->attr.exclude_host = 0;
+	if (perf_missing_features.lbr_flags)
+		evsel->attr.branch_sample_type &= ~(PERF_SAMPLE_BRANCH_NO_FLAGS |
+				     PERF_SAMPLE_BRANCH_NO_CYCLES);
 retry_sample_id:
 	if (perf_missing_features.sample_id_all)
 		evsel->attr.sample_id_all = 0;
@@ -1455,6 +1461,12 @@
 	} else if (!perf_missing_features.sample_id_all) {
 		perf_missing_features.sample_id_all = true;
 		goto retry_sample_id;
+	} else if (!perf_missing_features.lbr_flags &&
+			(evsel->attr.branch_sample_type &
+			 (PERF_SAMPLE_BRANCH_NO_CYCLES |
+			  PERF_SAMPLE_BRANCH_NO_FLAGS))) {
+		perf_missing_features.lbr_flags = true;
+		goto fallback_missing_features;
 	}
 
 out_close:
diff --git a/tools/perf/util/exec_cmd.c b/tools/perf/util/exec_cmd.c
deleted file mode 100644
index 1099e92..0000000
--- a/tools/perf/util/exec_cmd.c
+++ /dev/null
@@ -1,149 +0,0 @@
-#include "cache.h"
-#include "exec_cmd.h"
-#include "quote.h"
-
-#include <string.h>
-
-#define MAX_ARGS	32
-
-static const char *argv_exec_path;
-static const char *argv0_path;
-
-char *system_path(const char *path)
-{
-	static const char *prefix = PREFIX;
-	struct strbuf d = STRBUF_INIT;
-
-	if (is_absolute_path(path))
-		return strdup(path);
-
-	strbuf_addf(&d, "%s/%s", prefix, path);
-	path = strbuf_detach(&d, NULL);
-	return (char *)path;
-}
-
-const char *perf_extract_argv0_path(const char *argv0)
-{
-	const char *slash;
-
-	if (!argv0 || !*argv0)
-		return NULL;
-	slash = argv0 + strlen(argv0);
-
-	while (argv0 <= slash && !is_dir_sep(*slash))
-		slash--;
-
-	if (slash >= argv0) {
-		argv0_path = strndup(argv0, slash - argv0);
-		return argv0_path ? slash + 1 : NULL;
-	}
-
-	return argv0;
-}
-
-void perf_set_argv_exec_path(const char *exec_path)
-{
-	argv_exec_path = exec_path;
-	/*
-	 * Propagate this setting to external programs.
-	 */
-	setenv(EXEC_PATH_ENVIRONMENT, exec_path, 1);
-}
-
-
-/* Returns the highest-priority, location to look for perf programs. */
-char *perf_exec_path(void)
-{
-	char *env;
-
-	if (argv_exec_path)
-		return strdup(argv_exec_path);
-
-	env = getenv(EXEC_PATH_ENVIRONMENT);
-	if (env && *env)
-		return strdup(env);
-
-	return system_path(PERF_EXEC_PATH);
-}
-
-static void add_path(struct strbuf *out, const char *path)
-{
-	if (path && *path) {
-		if (is_absolute_path(path))
-			strbuf_addstr(out, path);
-		else
-			strbuf_addstr(out, make_nonrelative_path(path));
-
-		strbuf_addch(out, PATH_SEP);
-	}
-}
-
-void setup_path(void)
-{
-	const char *old_path = getenv("PATH");
-	struct strbuf new_path = STRBUF_INIT;
-	char *tmp = perf_exec_path();
-
-	add_path(&new_path, tmp);
-	add_path(&new_path, argv0_path);
-	free(tmp);
-
-	if (old_path)
-		strbuf_addstr(&new_path, old_path);
-	else
-		strbuf_addstr(&new_path, "/usr/local/bin:/usr/bin:/bin");
-
-	setenv("PATH", new_path.buf, 1);
-
-	strbuf_release(&new_path);
-}
-
-static const char **prepare_perf_cmd(const char **argv)
-{
-	int argc;
-	const char **nargv;
-
-	for (argc = 0; argv[argc]; argc++)
-		; /* just counting */
-	nargv = malloc(sizeof(*nargv) * (argc + 2));
-
-	nargv[0] = "perf";
-	for (argc = 0; argv[argc]; argc++)
-		nargv[argc + 1] = argv[argc];
-	nargv[argc + 1] = NULL;
-	return nargv;
-}
-
-int execv_perf_cmd(const char **argv) {
-	const char **nargv = prepare_perf_cmd(argv);
-
-	/* execvp() can only ever return if it fails */
-	execvp("perf", (char **)nargv);
-
-	free(nargv);
-	return -1;
-}
-
-
-int execl_perf_cmd(const char *cmd,...)
-{
-	int argc;
-	const char *argv[MAX_ARGS + 1];
-	const char *arg;
-	va_list param;
-
-	va_start(param, cmd);
-	argv[0] = cmd;
-	argc = 1;
-	while (argc < MAX_ARGS) {
-		arg = argv[argc++] = va_arg(param, char *);
-		if (!arg)
-			break;
-	}
-	va_end(param);
-	if (MAX_ARGS <= argc)
-		return error("too many args to run %s", cmd);
-
-	argv[argc] = NULL;
-	return execv_perf_cmd(argv);
-}
diff --git a/tools/perf/util/exec_cmd.h b/tools/perf/util/exec_cmd.h
deleted file mode 100644
index 48b4175..0000000
--- a/tools/perf/util/exec_cmd.h
+++ /dev/null
@@ -1,13 +0,0 @@
-#ifndef __PERF_EXEC_CMD_H
-#define __PERF_EXEC_CMD_H
-
-extern void perf_set_argv_exec_path(const char *exec_path);
-extern const char *perf_extract_argv0_path(const char *path);
-extern void setup_path(void);
-extern int execv_perf_cmd(const char **argv); /* NULL terminated */
-extern int execl_perf_cmd(const char *cmd, ...);
-/* perf_exec_path and system_path return malloc'd string, caller must free it */
-extern char *perf_exec_path(void);
-extern char *system_path(const char *path);
-
-#endif /* __PERF_EXEC_CMD_H */
diff --git a/tools/perf/util/help-unknown-cmd.c b/tools/perf/util/help-unknown-cmd.c
new file mode 100644
index 0000000..dc1e41c
--- /dev/null
+++ b/tools/perf/util/help-unknown-cmd.c
@@ -0,0 +1,103 @@
+#include "cache.h"
+#include <subcmd/help.h>
+#include "../builtin.h"
+#include "levenshtein.h"
+
+static int autocorrect;
+static struct cmdnames aliases;
+
+static int perf_unknown_cmd_config(const char *var, const char *value, void *cb)
+{
+	if (!strcmp(var, "help.autocorrect"))
+		autocorrect = perf_config_int(var,value);
+	/* Also use aliases for command lookup */
+	if (!prefixcmp(var, "alias."))
+		add_cmdname(&aliases, var + 6, strlen(var + 6));
+
+	return perf_default_config(var, value, cb);
+}
+
+static int levenshtein_compare(const void *p1, const void *p2)
+{
+	const struct cmdname *const *c1 = p1, *const *c2 = p2;
+	const char *s1 = (*c1)->name, *s2 = (*c2)->name;
+	int l1 = (*c1)->len;
+	int l2 = (*c2)->len;
+	return l1 != l2 ? l1 - l2 : strcmp(s1, s2);
+}
+
+static void add_cmd_list(struct cmdnames *cmds, struct cmdnames *old)
+{
+	unsigned int i;
+
+	ALLOC_GROW(cmds->names, cmds->cnt + old->cnt, cmds->alloc);
+
+	for (i = 0; i < old->cnt; i++)
+		cmds->names[cmds->cnt++] = old->names[i];
+	zfree(&old->names);
+	old->cnt = 0;
+}
+
+const char *help_unknown_cmd(const char *cmd)
+{
+	unsigned int i, n = 0, best_similarity = 0;
+	struct cmdnames main_cmds, other_cmds;
+
+	memset(&main_cmds, 0, sizeof(main_cmds));
+	memset(&other_cmds, 0, sizeof(main_cmds));
+	memset(&aliases, 0, sizeof(aliases));
+
+	perf_config(perf_unknown_cmd_config, NULL);
+
+	load_command_list("perf-", &main_cmds, &other_cmds);
+
+	add_cmd_list(&main_cmds, &aliases);
+	add_cmd_list(&main_cmds, &other_cmds);
+	qsort(main_cmds.names, main_cmds.cnt,
+	      sizeof(main_cmds.names), cmdname_compare);
+	uniq(&main_cmds);
+
+	if (main_cmds.cnt) {
+		/* This reuses cmdname->len for similarity index */
+		for (i = 0; i < main_cmds.cnt; ++i)
+			main_cmds.names[i]->len =
+				levenshtein(cmd, main_cmds.names[i]->name, 0, 2, 1, 4);
+
+		qsort(main_cmds.names, main_cmds.cnt,
+		      sizeof(*main_cmds.names), levenshtein_compare);
+
+		best_similarity = main_cmds.names[0]->len;
+		n = 1;
+		while (n < main_cmds.cnt && best_similarity == main_cmds.names[n]->len)
+			++n;
+	}
+
+	if (autocorrect && n == 1) {
+		const char *assumed = main_cmds.names[0]->name;
+
+		main_cmds.names[0] = NULL;
+		clean_cmdnames(&main_cmds);
+		fprintf(stderr, "WARNING: You called a perf program named '%s', "
+			"which does not exist.\n"
+			"Continuing under the assumption that you meant '%s'\n",
+			cmd, assumed);
+		if (autocorrect > 0) {
+			fprintf(stderr, "in %0.1f seconds automatically...\n",
+				(float)autocorrect/10.0);
+			poll(NULL, 0, autocorrect * 100);
+		}
+		return assumed;
+	}
+
+	fprintf(stderr, "perf: '%s' is not a perf-command. See 'perf --help'.\n", cmd);
+
+	if (main_cmds.cnt && best_similarity < 6) {
+		fprintf(stderr, "\nDid you mean %s?\n",
+			n < 2 ? "this": "one of these");
+
+		for (i = 0; i < n; i++)
+			fprintf(stderr, "\t%s\n", main_cmds.names[i]->name);
+	}
+
+	exit(1);
+}
diff --git a/tools/perf/util/help-unknown-cmd.h b/tools/perf/util/help-unknown-cmd.h
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/tools/perf/util/help-unknown-cmd.h
diff --git a/tools/perf/util/help.c b/tools/perf/util/help.c
deleted file mode 100644
index 929c93f..0000000
--- a/tools/perf/util/help.c
+++ /dev/null
@@ -1,334 +0,0 @@
-#include "cache.h"
-#include "../builtin.h"
-#include "exec_cmd.h"
-#include "levenshtein.h"
-#include "help.h"
-#include <termios.h>
-
-void add_cmdname(struct cmdnames *cmds, const char *name, size_t len)
-{
-	struct cmdname *ent = malloc(sizeof(*ent) + len + 1);
-
-	ent->len = len;
-	memcpy(ent->name, name, len);
-	ent->name[len] = 0;
-
-	ALLOC_GROW(cmds->names, cmds->cnt + 1, cmds->alloc);
-	cmds->names[cmds->cnt++] = ent;
-}
-
-static void clean_cmdnames(struct cmdnames *cmds)
-{
-	unsigned int i;
-
-	for (i = 0; i < cmds->cnt; ++i)
-		zfree(&cmds->names[i]);
-	zfree(&cmds->names);
-	cmds->cnt = 0;
-	cmds->alloc = 0;
-}
-
-static int cmdname_compare(const void *a_, const void *b_)
-{
-	struct cmdname *a = *(struct cmdname **)a_;
-	struct cmdname *b = *(struct cmdname **)b_;
-	return strcmp(a->name, b->name);
-}
-
-static void uniq(struct cmdnames *cmds)
-{
-	unsigned int i, j;
-
-	if (!cmds->cnt)
-		return;
-
-	for (i = j = 1; i < cmds->cnt; i++)
-		if (strcmp(cmds->names[i]->name, cmds->names[i-1]->name))
-			cmds->names[j++] = cmds->names[i];
-
-	cmds->cnt = j;
-}
-
-void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes)
-{
-	size_t ci, cj, ei;
-	int cmp;
-
-	ci = cj = ei = 0;
-	while (ci < cmds->cnt && ei < excludes->cnt) {
-		cmp = strcmp(cmds->names[ci]->name, excludes->names[ei]->name);
-		if (cmp < 0)
-			cmds->names[cj++] = cmds->names[ci++];
-		else if (cmp == 0)
-			ci++, ei++;
-		else if (cmp > 0)
-			ei++;
-	}
-
-	while (ci < cmds->cnt)
-		cmds->names[cj++] = cmds->names[ci++];
-
-	cmds->cnt = cj;
-}
-
-static void pretty_print_string_list(struct cmdnames *cmds, int longest)
-{
-	int cols = 1, rows;
-	int space = longest + 1; /* min 1 SP between words */
-	struct winsize win;
-	int max_cols;
-	int i, j;
-
-	get_term_dimensions(&win);
-	max_cols = win.ws_col - 1; /* don't print *on* the edge */
-
-	if (space < max_cols)
-		cols = max_cols / space;
-	rows = (cmds->cnt + cols - 1) / cols;
-
-	for (i = 0; i < rows; i++) {
-		printf("  ");
-
-		for (j = 0; j < cols; j++) {
-			unsigned int n = j * rows + i;
-			unsigned int size = space;
-
-			if (n >= cmds->cnt)
-				break;
-			if (j == cols-1 || n + rows >= cmds->cnt)
-				size = 1;
-			printf("%-*s", size, cmds->names[n]->name);
-		}
-		putchar('\n');
-	}
-}
-
-static int is_executable(const char *name)
-{
-	struct stat st;
-
-	if (stat(name, &st) || /* stat, not lstat */
-	    !S_ISREG(st.st_mode))
-		return 0;
-
-	return st.st_mode & S_IXUSR;
-}
-
-static void list_commands_in_dir(struct cmdnames *cmds,
-					 const char *path,
-					 const char *prefix)
-{
-	int prefix_len;
-	DIR *dir = opendir(path);
-	struct dirent *de;
-	struct strbuf buf = STRBUF_INIT;
-	int len;
-
-	if (!dir)
-		return;
-	if (!prefix)
-		prefix = "perf-";
-	prefix_len = strlen(prefix);
-
-	strbuf_addf(&buf, "%s/", path);
-	len = buf.len;
-
-	while ((de = readdir(dir)) != NULL) {
-		int entlen;
-
-		if (prefixcmp(de->d_name, prefix))
-			continue;
-
-		strbuf_setlen(&buf, len);
-		strbuf_addstr(&buf, de->d_name);
-		if (!is_executable(buf.buf))
-			continue;
-
-		entlen = strlen(de->d_name) - prefix_len;
-		if (has_extension(de->d_name, ".exe"))
-			entlen -= 4;
-
-		add_cmdname(cmds, de->d_name + prefix_len, entlen);
-	}
-	closedir(dir);
-	strbuf_release(&buf);
-}
-
-void load_command_list(const char *prefix,
-		struct cmdnames *main_cmds,
-		struct cmdnames *other_cmds)
-{
-	const char *env_path = getenv("PATH");
-	char *exec_path = perf_exec_path();
-
-	if (exec_path) {
-		list_commands_in_dir(main_cmds, exec_path, prefix);
-		qsort(main_cmds->names, main_cmds->cnt,
-		      sizeof(*main_cmds->names), cmdname_compare);
-		uniq(main_cmds);
-	}
-
-	if (env_path) {
-		char *paths, *path, *colon;
-		path = paths = strdup(env_path);
-		while (1) {
-			if ((colon = strchr(path, PATH_SEP)))
-				*colon = 0;
-			if (!exec_path || strcmp(path, exec_path))
-				list_commands_in_dir(other_cmds, path, prefix);
-
-			if (!colon)
-				break;
-			path = colon + 1;
-		}
-		free(paths);
-
-		qsort(other_cmds->names, other_cmds->cnt,
-		      sizeof(*other_cmds->names), cmdname_compare);
-		uniq(other_cmds);
-	}
-	free(exec_path);
-	exclude_cmds(other_cmds, main_cmds);
-}
-
-void list_commands(const char *title, struct cmdnames *main_cmds,
-		   struct cmdnames *other_cmds)
-{
-	unsigned int i, longest = 0;
-
-	for (i = 0; i < main_cmds->cnt; i++)
-		if (longest < main_cmds->names[i]->len)
-			longest = main_cmds->names[i]->len;
-	for (i = 0; i < other_cmds->cnt; i++)
-		if (longest < other_cmds->names[i]->len)
-			longest = other_cmds->names[i]->len;
-
-	if (main_cmds->cnt) {
-		char *exec_path = perf_exec_path();
-		printf("available %s in '%s'\n", title, exec_path);
-		printf("----------------");
-		mput_char('-', strlen(title) + strlen(exec_path));
-		putchar('\n');
-		pretty_print_string_list(main_cmds, longest);
-		putchar('\n');
-		free(exec_path);
-	}
-
-	if (other_cmds->cnt) {
-		printf("%s available from elsewhere on your $PATH\n", title);
-		printf("---------------------------------------");
-		mput_char('-', strlen(title));
-		putchar('\n');
-		pretty_print_string_list(other_cmds, longest);
-		putchar('\n');
-	}
-}
-
-int is_in_cmdlist(struct cmdnames *c, const char *s)
-{
-	unsigned int i;
-
-	for (i = 0; i < c->cnt; i++)
-		if (!strcmp(s, c->names[i]->name))
-			return 1;
-	return 0;
-}
-
-static int autocorrect;
-static struct cmdnames aliases;
-
-static int perf_unknown_cmd_config(const char *var, const char *value, void *cb)
-{
-	if (!strcmp(var, "help.autocorrect"))
-		autocorrect = perf_config_int(var,value);
-	/* Also use aliases for command lookup */
-	if (!prefixcmp(var, "alias."))
-		add_cmdname(&aliases, var + 6, strlen(var + 6));
-
-	return perf_default_config(var, value, cb);
-}
-
-static int levenshtein_compare(const void *p1, const void *p2)
-{
-	const struct cmdname *const *c1 = p1, *const *c2 = p2;
-	const char *s1 = (*c1)->name, *s2 = (*c2)->name;
-	int l1 = (*c1)->len;
-	int l2 = (*c2)->len;
-	return l1 != l2 ? l1 - l2 : strcmp(s1, s2);
-}
-
-static void add_cmd_list(struct cmdnames *cmds, struct cmdnames *old)
-{
-	unsigned int i;
-
-	ALLOC_GROW(cmds->names, cmds->cnt + old->cnt, cmds->alloc);
-
-	for (i = 0; i < old->cnt; i++)
-		cmds->names[cmds->cnt++] = old->names[i];
-	zfree(&old->names);
-	old->cnt = 0;
-}
-
-const char *help_unknown_cmd(const char *cmd)
-{
-	unsigned int i, n = 0, best_similarity = 0;
-	struct cmdnames main_cmds, other_cmds;
-
-	memset(&main_cmds, 0, sizeof(main_cmds));
-	memset(&other_cmds, 0, sizeof(main_cmds));
-	memset(&aliases, 0, sizeof(aliases));
-
-	perf_config(perf_unknown_cmd_config, NULL);
-
-	load_command_list("perf-", &main_cmds, &other_cmds);
-
-	add_cmd_list(&main_cmds, &aliases);
-	add_cmd_list(&main_cmds, &other_cmds);
-	qsort(main_cmds.names, main_cmds.cnt,
-	      sizeof(main_cmds.names), cmdname_compare);
-	uniq(&main_cmds);
-
-	if (main_cmds.cnt) {
-		/* This reuses cmdname->len for similarity index */
-		for (i = 0; i < main_cmds.cnt; ++i)
-			main_cmds.names[i]->len =
-				levenshtein(cmd, main_cmds.names[i]->name, 0, 2, 1, 4);
-
-		qsort(main_cmds.names, main_cmds.cnt,
-		      sizeof(*main_cmds.names), levenshtein_compare);
-
-		best_similarity = main_cmds.names[0]->len;
-		n = 1;
-		while (n < main_cmds.cnt && best_similarity == main_cmds.names[n]->len)
-			++n;
-	}
-
-	if (autocorrect && n == 1) {
-		const char *assumed = main_cmds.names[0]->name;
-
-		main_cmds.names[0] = NULL;
-		clean_cmdnames(&main_cmds);
-		fprintf(stderr, "WARNING: You called a perf program named '%s', "
-			"which does not exist.\n"
-			"Continuing under the assumption that you meant '%s'\n",
-			cmd, assumed);
-		if (autocorrect > 0) {
-			fprintf(stderr, "in %0.1f seconds automatically...\n",
-				(float)autocorrect/10.0);
-			poll(NULL, 0, autocorrect * 100);
-		}
-		return assumed;
-	}
-
-	fprintf(stderr, "perf: '%s' is not a perf-command. See 'perf --help'.\n", cmd);
-
-	if (main_cmds.cnt && best_similarity < 6) {
-		fprintf(stderr, "\nDid you mean %s?\n",
-			n < 2 ? "this": "one of these");
-
-		for (i = 0; i < n; i++)
-			fprintf(stderr, "\t%s\n", main_cmds.names[i]->name);
-	}
-
-	exit(1);
-}
diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 97f963a..81a2eb7 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1744,7 +1744,7 @@
 	auxtrace_heap__free(&pt->heap);
 	intel_pt_free_events(session);
 	session->auxtrace = NULL;
-	thread__delete(pt->unknown_thread);
+	thread__put(pt->unknown_thread);
 	free(pt);
 }
 
@@ -2153,7 +2153,7 @@
 	return 0;
 
 err_delete_thread:
-	thread__delete(pt->unknown_thread);
+	thread__zput(pt->unknown_thread);
 err_free_queues:
 	intel_pt_log_disable();
 	auxtrace_queues__free(&pt->queues);
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 1407d51..ad79297 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -352,13 +352,18 @@
 	}
 
 	th->mg = map_groups__get(leader->mg);
-
+out_put:
+	thread__put(leader);
 	return;
-
 out_err:
 	pr_err("Failed to join map groups for %d:%d\n", th->pid_, th->tid);
+	goto out_put;
 }
 
+/*
+ * Caller must eventually drop thread->refcnt returned with a successfull
+ * lookup/new thread inserted.
+ */
 static struct thread *____machine__findnew_thread(struct machine *machine,
 						  pid_t pid, pid_t tid,
 						  bool create)
@@ -376,7 +381,7 @@
 	if (th != NULL) {
 		if (th->tid == tid) {
 			machine__update_thread_pid(machine, th, pid);
-			return th;
+			return thread__get(th);
 		}
 
 		machine->last_match = NULL;
@@ -389,7 +394,7 @@
 		if (th->tid == tid) {
 			machine->last_match = th;
 			machine__update_thread_pid(machine, th, pid);
-			return th;
+			return thread__get(th);
 		}
 
 		if (tid < th->tid)
@@ -417,7 +422,7 @@
 		if (thread__init_map_groups(th, machine)) {
 			rb_erase_init(&th->rb_node, &machine->threads);
 			RB_CLEAR_NODE(&th->rb_node);
-			thread__delete(th);
+			thread__put(th);
 			return NULL;
 		}
 		/*
@@ -441,7 +446,7 @@
 	struct thread *th;
 
 	pthread_rwlock_wrlock(&machine->threads_lock);
-	th = thread__get(__machine__findnew_thread(machine, pid, tid));
+	th = __machine__findnew_thread(machine, pid, tid);
 	pthread_rwlock_unlock(&machine->threads_lock);
 	return th;
 }
@@ -451,7 +456,7 @@
 {
 	struct thread *th;
 	pthread_rwlock_rdlock(&machine->threads_lock);
-	th =  thread__get(____machine__findnew_thread(machine, pid, tid, false));
+	th =  ____machine__findnew_thread(machine, pid, tid, false);
 	pthread_rwlock_unlock(&machine->threads_lock);
 	return th;
 }
diff --git a/tools/perf/util/parse-branch-options.c b/tools/perf/util/parse-branch-options.c
index 355eecf..afc088d 100644
--- a/tools/perf/util/parse-branch-options.c
+++ b/tools/perf/util/parse-branch-options.c
@@ -1,7 +1,7 @@
 #include "perf.h"
 #include "util/util.h"
 #include "util/debug.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-branch-options.h"
 
 #define BRANCH_OPT(n, m) \
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 6fc8cd7..4f7b0ef 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -4,9 +4,9 @@
 #include "../perf.h"
 #include "evlist.h"
 #include "evsel.h"
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 #include "parse-events.h"
-#include "exec_cmd.h"
+#include <subcmd/exec-cmd.h>
 #include "string.h"
 #include "symbol.h"
 #include "cache.h"
diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index 4f2c1c2..646ecf7 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -1,7 +1,7 @@
 #include "perf.h"
 #include "util/util.h"
 #include "util/debug.h"
-#include "util/parse-options.h"
+#include <subcmd/parse-options.h>
 #include "util/parse-regs-options.h"
 
 int
diff --git a/tools/perf/util/path.c b/tools/perf/util/path.c
index 5d13cb4..3654d96 100644
--- a/tools/perf/util/path.c
+++ b/tools/perf/util/path.c
@@ -22,24 +22,6 @@
 	return ".";
 }
 
-/*
- * If libc has strlcpy() then that version will override this
- * implementation:
- */
-size_t __weak strlcpy(char *dest, const char *src, size_t size)
-{
-	size_t ret = strlen(src);
-
-	if (size) {
-		size_t len = (ret >= size) ? size - 1 : ret;
-
-		memcpy(dest, src, len);
-		dest[len] = '\0';
-	}
-
-	return ret;
-}
-
 static char *get_pathname(void)
 {
 	static char pathname_array[4][PATH_MAX];
diff --git a/tools/perf/util/sigchain.h b/tools/perf/util/sigchain.h
deleted file mode 100644
index 959d64e..0000000
--- a/tools/perf/util/sigchain.h
+++ /dev/null
@@ -1,10 +0,0 @@
-#ifndef __PERF_SIGCHAIN_H
-#define __PERF_SIGCHAIN_H
-
-typedef void (*sigchain_fun)(int);
-
-int sigchain_pop(int sig);
-
-void sigchain_push_common(sigchain_fun f);
-
-#endif /* __PERF_SIGCHAIN_H */
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index 3122885..86f05e7 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -18,7 +18,7 @@
 #include "debug.h"
 #include "header.h"
 
-#include "parse-options.h"
+#include <subcmd/parse-options.h>
 #include "parse-events.h"
 #include "hist.h"
 #include "thread.h"
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index 0a9ae80..dfd00c6 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -19,8 +19,10 @@
 		thread->mg = map_groups__new(machine);
 	} else {
 		leader = __machine__findnew_thread(machine, pid, pid);
-		if (leader)
+		if (leader) {
 			thread->mg = map_groups__get(leader->mg);
+			thread__put(leader);
+		}
 	}
 
 	return thread->mg ? 0 : -1;
@@ -53,7 +55,7 @@
 			goto err_thread;
 
 		list_add(&comm->list, &thread->comm_list);
-		atomic_set(&thread->refcnt, 0);
+		atomic_set(&thread->refcnt, 1);
 		RB_CLEAR_NODE(&thread->rb_node);
 	}
 
@@ -95,6 +97,10 @@
 void thread__put(struct thread *thread)
 {
 	if (thread && atomic_dec_and_test(&thread->refcnt)) {
+		/*
+		 * Remove it from the dead_threads list, as last reference
+		 * is gone.
+		 */
 		list_del_init(&thread->node);
 		thread__delete(thread);
 	}
diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 07da970..aff0cfd 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -352,7 +352,8 @@
 {
 	psignal(sig, "perf");
 	dump_stack();
-	exit(sig);
+	signal(sig, SIG_DFL);
+	raise(sig);
 }
 
 int parse_nsec_time(const char *str, u64 *ptime)
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 150858f..4b519c5 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -151,12 +151,6 @@
 extern int prefixcmp(const char *str, const char *prefix);
 extern void set_buildid_dir(const char *dir);
 
-static inline const char *skip_prefix(const char *str, const char *prefix)
-{
-	size_t len = strlen(prefix);
-	return strncmp(str, prefix, len) ? NULL : str + len;
-}
-
 #ifdef __GLIBC_PREREQ
 #if __GLIBC_PREREQ(2, 1)
 #define HAVE_STRCHRNUL
@@ -187,14 +181,6 @@
 
 #define zfree(ptr) ({ free(*ptr); *ptr = NULL; })
 
-static inline int has_extension(const char *filename, const char *ext)
-{
-	size_t len = strlen(filename);
-	size_t extlen = strlen(ext);
-
-	return len > extlen && !memcmp(filename + len - extlen, ext, extlen);
-}
-
 /* Sane ctype - no locale, and works with signed chars */
 #undef isascii
 #undef isspace