Merge v2.6.37-rc8 into powerpc/next

89 files changed, 4366 insertions, 2054 deletions

diff --git a/tools/perf/Documentation/perf-annotate.txt b/tools/perf/Documentation/perf-annotate.txt
index 5164a655c39f..b2c63309a651 100644
--- a/tools/perf/Documentation/perf-annotate.txt
+++ b/tools/perf/Documentation/perf-annotate.txt
@@ -8,7 +8,7 @@ perf-annotate - Read perf.data (created by perf record) and display annotated co
 SYNOPSIS
 --------
 [verse]
-'perf annotate' [-i <file> | --input=file] symbol_name
+'perf annotate' [-i <file> | --input=file] [symbol_name]
 
 DESCRIPTION
 -----------
@@ -24,6 +24,13 @@ OPTIONS
 --input=::
         Input file name. (default: perf.data)
 
+--stdio:: Use the stdio interface.
+
+--tui:: Use the TUI interface Use of --tui requires a tty, if one is not
+        present, as when piping to other commands, the stdio interface is
+        used. This interfaces starts by centering on the line with more
+        samples, TAB/UNTAB cycles thru the lines with more samples.
+
 SEE ALSO
 --------
-linkperf:perf-record[1]
+linkperf:perf-record[1], linkperf:perf-report[1]
diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 43e3dd284b90..399751befeed 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -15,6 +15,23 @@ DESCRIPTION
 This command displays the symbolic event types which can be selected in the
 various perf commands with the -e option.
 
+EVENT MODIFIERS
+---------------
+
+Events can optionally have a modifer by appending a colon and one or
+more modifiers.  Modifiers allow the user to restrict when events are
+counted with 'u' for user-space, 'k' for kernel, 'h' for hypervisor.
+
+The 'p' modifier can be used for specifying how precise the instruction
+address should be. The 'p' modifier is currently only implemented for
+Intel PEBS and can be specified multiple times:
+  0 - SAMPLE_IP can have arbitrary skid
+  1 - SAMPLE_IP must have constant skid
+  2 - SAMPLE_IP requested to have 0 skid
+  3 - SAMPLE_IP must have 0 skid
+
+The PEBS implementation now supports up to 2.
+
 RAW HARDWARE EVENT DESCRIPTOR
 -----------------------------
 Even when an event is not available in a symbolic form within perf right now,
diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt
index 27d52dae5a43..62de1b7f4e76 100644
--- a/tools/perf/Documentation/perf-probe.txt
+++ b/tools/perf/Documentation/perf-probe.txt
@@ -16,7 +16,9 @@ or
 or
 'perf probe' --list
 or
-'perf probe' --line='FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]'
+'perf probe' [options] --line='FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]'
+or
+'perf probe' [options] --vars='PROBEPOINT'
 
 DESCRIPTION
 -----------
@@ -31,6 +33,11 @@ OPTIONS
 --vmlinux=PATH::
         Specify vmlinux path which has debuginfo (Dwarf binary).
 
+-m::
+--module=MODNAME::
+        Specify module name in which perf-probe searches probe points
+        or lines.
+
 -s::
 --source=PATH::
         Specify path to kernel source.
@@ -57,6 +64,15 @@ OPTIONS
         Show source code lines which can be probed. This needs an argument
         which specifies a range of the source code. (see LINE SYNTAX for detail)
 
+-V::
+--vars=::
+        Show available local variables at given probe point. The argument
+        syntax is same as PROBE SYNTAX, but NO ARGs.
+
+--externs::
+        (Only for --vars) Show external defined variables in addition to local
+        variables.
+
 -f::
 --force::
         Forcibly add events with existing name.
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index 3ee27dccfde9..a91f9f9e6e5c 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -83,6 +83,10 @@ OPTIONS
 --call-graph::
         Do call-graph (stack chain/backtrace) recording.
 
+-q::
+--quiet::
+        Don't print any message, useful for scripting.
+
 -v::
 --verbose::
         Be more verbose (show counter open errors, etc).
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index abfabe9147a4..12052c9ed0ba 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -65,6 +65,13 @@ OPTIONS
                  the tree is considered as a new profiled object. +
         Default: fractal,0.5.
 
+--stdio:: Use the stdio interface.
+
+--tui:: Use the TUI interface, that is integrated with annotate and allows
+        zooming into DSOs or threads, among other features. Use of --tui
+        requires a tty, if one is not present, as when piping to other
+        commands, the stdio interface is used.
+
 SEE ALSO
 --------
 linkperf:perf-stat[1]
diff --git a/tools/perf/Documentation/perf-trace.txt b/tools/perf/Documentation/perf-trace.txt
index 122ec9dc4853..26aff6bf9e50 100644
--- a/tools/perf/Documentation/perf-trace.txt
+++ b/tools/perf/Documentation/perf-trace.txt
@@ -8,7 +8,11 @@ perf-trace - Read perf.data (created by perf record) and display trace output
 SYNOPSIS
 --------
 [verse]
-'perf trace' {record <script> | report <script> [args] }
+'perf trace' [<options>]
+'perf trace' [<options>] record <script> [<record-options>] <command>
+'perf trace' [<options>] report <script> [script-args]
+'perf trace' [<options>] <script> <required-script-args> [<record-options>] <command>
+'perf trace' [<options>] <top-script> [script-args]
 
 DESCRIPTION
 -----------
@@ -24,23 +28,53 @@ There are several variants of perf trace:
   available via 'perf trace -l').  The following variants allow you to
   record and run those scripts:
 
-  'perf trace record <script>' to record the events required for 'perf
-  trace report'.  <script> is the name displayed in the output of
-  'perf trace --list' i.e. the actual script name minus any language
-  extension.
+  'perf trace record <script> <command>' to record the events required
+  for 'perf trace report'.  <script> is the name displayed in the
+  output of 'perf trace --list' i.e. the actual script name minus any
+  language extension.  If <command> is not specified, the events are
+  recorded using the -a (system-wide) 'perf record' option.
 
-  'perf trace report <script>' to run and display the results of
-  <script>.  <script> is the name displayed in the output of 'perf
+  'perf trace report <script> [args]' to run and display the results
+  of <script>.  <script> is the name displayed in the output of 'perf
   trace --list' i.e. the actual script name minus any language
   extension.  The perf.data output from a previous run of 'perf trace
   record <script>' is used and should be present for this command to
-  succeed.
+  succeed.  [args] refers to the (mainly optional) args expected by
+  the script.
+
+  'perf trace <script> <required-script-args> <command>' to both
+  record the events required for <script> and to run the <script>
+  using 'live-mode' i.e. without writing anything to disk.  <script>
+  is the name displayed in the output of 'perf trace --list' i.e. the
+  actual script name minus any language extension.  If <command> is
+  not specified, the events are recorded using the -a (system-wide)
+  'perf record' option.  If <script> has any required args, they
+  should be specified before <command>.  This mode doesn't allow for
+  optional script args to be specified; if optional script args are
+  desired, they can be specified using separate 'perf trace record'
+  and 'perf trace report' commands, with the stdout of the record step
+  piped to the stdin of the report script, using the '-o -' and '-i -'
+  options of the corresponding commands.
+
+  'perf trace <top-script>' to both record the events required for
+  <top-script> and to run the <top-script> using 'live-mode'
+  i.e. without writing anything to disk.  <top-script> is the name
+  displayed in the output of 'perf trace --list' i.e. the actual
+  script name minus any language extension; a <top-script> is defined
+  as any script name ending with the string 'top'.
+
+  [<record-options>] can be passed to the record steps of 'perf trace
+  record' and 'live-mode' variants; this isn't possible however for
+  <top-script> 'live-mode' or 'perf trace report' variants.
 
   See the 'SEE ALSO' section for links to language-specific
   information on how to write and run your own trace scripts.
 
 OPTIONS
 -------
+<command>...::
+        Any command you can specify in a shell.
+
 -D::
 --dump-raw-trace=::
         Display verbose dump of the trace data.
@@ -64,6 +98,13 @@ OPTIONS
         Generate perf-trace.[ext] starter script for given language,
         using current perf.data.
 
+-a::
+        Force system-wide collection.  Scripts run without a <command>
+        normally use -a by default, while scripts run with a <command>
+        normally don't - this option allows the latter to be run in
+        system-wide mode.
+
+
 SEE ALSO
 --------
 linkperf:perf-record[1], linkperf:perf-trace-perl[1],
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 26f626d45a9e..d1db0f676a4b 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -5,6 +5,12 @@ endif
 # The default target of this Makefile is...
 all::
 
+ifneq ($(OUTPUT),)
+# check that the output directory actually exists
+OUTDIR := $(shell cd $(OUTPUT) && /bin/pwd)
+$(if $(OUTDIR),, $(error output directory "$(OUTPUT)" does not exist))
+endif
+
 # Define V=1 to have a more verbose compile.
 # Define V=2 to have an even more verbose compile.
 #
@@ -157,11 +163,6 @@ all::
 #
 # Define NO_DWARF if you do not want debug-info analysis feature at all.
 
-$(shell sh -c 'mkdir -p $(OUTPUT)scripts/python/Perf-Trace-Util/' 2> /dev/null)
-$(shell sh -c 'mkdir -p $(OUTPUT)scripts/perl/Perf-Trace-Util/' 2> /dev/null)
-$(shell sh -c 'mkdir -p $(OUTPUT)util/scripting-engines/' 2> /dev/null)
-$(shell sh -c 'mkdir $(OUTPUT)bench' 2> /dev/null)
-
 $(OUTPUT)PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
         @$(SHELL_PATH) util/PERF-VERSION-GEN $(OUTPUT)
 -include $(OUTPUT)PERF-VERSION-FILE
@@ -187,8 +188,6 @@ ifeq ($(ARCH),x86_64)
         ARCH := x86
 endif
 
-$(shell sh -c 'mkdir -p $(OUTPUT)arch/$(ARCH)/util/' 2> /dev/null)
-
 # CFLAGS and LDFLAGS are for the users to override from the command line.
 
 #
@@ -269,6 +268,7 @@ export prefix bindir sharedir sysconfdir
 CC = $(CROSS_COMPILE)gcc
 AR = $(CROSS_COMPILE)ar
 RM = rm -f
+MKDIR = mkdir
 TAR = tar
 FIND = find
 INSTALL = install
@@ -313,6 +313,9 @@ TEST_PROGRAMS =
 
 SCRIPT_SH += perf-archive.sh
 
+grep-libs = $(filter -l%,$(1))
+strip-libs = $(filter-out -l%,$(1))
+
 #
 # No Perl scripts right now:
 #
@@ -568,21 +571,37 @@ else
                 # Fedora has /usr/include/slang/slang.h, but ubuntu /usr/include/slang.h
                 BASIC_CFLAGS += -I/usr/include/slang
                 EXTLIBS += -lnewt -lslang
-                LIB_OBJS += $(OUTPUT)util/newt.o
+                LIB_OBJS += $(OUTPUT)util/ui/setup.o
+                LIB_OBJS += $(OUTPUT)util/ui/browser.o
+                LIB_OBJS += $(OUTPUT)util/ui/browsers/annotate.o
+                LIB_OBJS += $(OUTPUT)util/ui/browsers/hists.o
+                LIB_OBJS += $(OUTPUT)util/ui/browsers/map.o
+                LIB_OBJS += $(OUTPUT)util/ui/helpline.o
+                LIB_OBJS += $(OUTPUT)util/ui/progress.o
+                LIB_OBJS += $(OUTPUT)util/ui/util.o
+                LIB_H += util/ui/browser.h
+                LIB_H += util/ui/browsers/map.h
+                LIB_H += util/ui/helpline.h
+                LIB_H += util/ui/libslang.h
+                LIB_H += util/ui/progress.h
+                LIB_H += util/ui/util.h
         endif
 endif
 
 ifdef NO_LIBPERL
         BASIC_CFLAGS += -DNO_LIBPERL
 else
-        PERL_EMBED_LDOPTS = `perl -MExtUtils::Embed -e ldopts 2>/dev/null`
+       PERL_EMBED_LDOPTS = $(shell perl -MExtUtils::Embed -e ldopts 2>/dev/null)
+       PERL_EMBED_LDFLAGS = $(call strip-libs,$(PERL_EMBED_LDOPTS))
+       PERL_EMBED_LIBADD = $(call grep-libs,$(PERL_EMBED_LDOPTS))
         PERL_EMBED_CCOPTS = `perl -MExtUtils::Embed -e ccopts 2>/dev/null`
         FLAGS_PERL_EMBED=$(PERL_EMBED_CCOPTS) $(PERL_EMBED_LDOPTS)
 
         ifneq ($(call try-cc,$(SOURCE_PERL_EMBED),$(FLAGS_PERL_EMBED)),y)
                 BASIC_CFLAGS += -DNO_LIBPERL
         else
-                ALL_LDFLAGS += $(PERL_EMBED_LDOPTS)
+               ALL_LDFLAGS += $(PERL_EMBED_LDFLAGS)
+               EXTLIBS += $(PERL_EMBED_LIBADD)
                 LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-perl.o
                 LIB_OBJS += $(OUTPUT)scripts/perl/Perf-Trace-Util/Context.o
         endif
@@ -591,13 +610,16 @@ endif
 ifdef NO_LIBPYTHON
         BASIC_CFLAGS += -DNO_LIBPYTHON
 else
-        PYTHON_EMBED_LDOPTS = `python-config --ldflags 2>/dev/null`
+       PYTHON_EMBED_LDOPTS = $(shell python-config --ldflags 2>/dev/null)
+       PYTHON_EMBED_LDFLAGS = $(call strip-libs,$(PYTHON_EMBED_LDOPTS))
+       PYTHON_EMBED_LIBADD = $(call grep-libs,$(PYTHON_EMBED_LDOPTS))
         PYTHON_EMBED_CCOPTS = `python-config --cflags 2>/dev/null`
         FLAGS_PYTHON_EMBED=$(PYTHON_EMBED_CCOPTS) $(PYTHON_EMBED_LDOPTS)
         ifneq ($(call try-cc,$(SOURCE_PYTHON_EMBED),$(FLAGS_PYTHON_EMBED)),y)
                 BASIC_CFLAGS += -DNO_LIBPYTHON
         else
-                ALL_LDFLAGS += $(PYTHON_EMBED_LDOPTS)
+               ALL_LDFLAGS += $(PYTHON_EMBED_LDFLAGS)
+               EXTLIBS += $(PYTHON_EMBED_LIBADD)
                 LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-python.o
                 LIB_OBJS += $(OUTPUT)scripts/python/Perf-Trace-Util/Context.o
         endif
@@ -640,6 +662,15 @@ else
         endif
 endif
 
+
+ifdef NO_STRLCPY
+        BASIC_CFLAGS += -DNO_STRLCPY
+else
+        ifneq ($(call try-cc,$(SOURCE_STRLCPY),),y)
+                BASIC_CFLAGS += -DNO_STRLCPY
+        endif
+endif
+
 ifndef CC_LD_DYNPATH
         ifdef NO_R_TO_GCC_LINKER
                 # Some gcc does not accept and pass -R to the linker to specify
@@ -826,6 +857,7 @@ ifndef V
         QUIET_CC       = @echo '   ' CC $@;
         QUIET_AR       = @echo '   ' AR $@;
         QUIET_LINK     = @echo '   ' LINK $@;
+        QUIET_MKDIR    = @echo '   ' MKDIR $@;
         QUIET_BUILT_IN = @echo '   ' BUILTIN $@;
         QUIET_GEN      = @echo '   ' GEN $@;
         QUIET_SUBDIR0  = +@subdir=
@@ -896,8 +928,8 @@ $(OUTPUT)perf.o: perf.c $(OUTPUT)common-cmds.h $(OUTPUT)PERF-CFLAGS
                 $(ALL_CFLAGS) -c $(filter %.c,$^) -o $@
 
 $(OUTPUT)perf$X: $(OUTPUT)perf.o $(BUILTIN_OBJS) $(PERFLIBS)
-        $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(OUTPUT)perf.o \
-                $(BUILTIN_OBJS) $(ALL_LDFLAGS) $(LIBS)
+        $(QUIET_LINK)$(CC) $(ALL_CFLAGS) $(ALL_LDFLAGS) $(OUTPUT)perf.o \
+               $(BUILTIN_OBJS) $(LIBS) -o $@
 
 $(OUTPUT)builtin-help.o: builtin-help.c $(OUTPUT)common-cmds.h $(OUTPUT)PERF-CFLAGS
         $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) \
@@ -923,15 +955,15 @@ $(OUTPUT)common-cmds.h: $(wildcard Documentation/perf-*.txt)
         $(QUIET_GEN). util/generate-cmdlist.sh > $@+ && mv $@+ $@
 
 $(patsubst %.sh,%,$(SCRIPT_SH)) : % : %.sh
-        $(QUIET_GEN)$(RM) $@ $@+ && \
+        $(QUIET_GEN)$(RM) $(OUTPUT)$@ $(OUTPUT)$@+ && \
         sed -e '1s|#!.*/sh|#!$(SHELL_PATH_SQ)|' \
             -e 's|@SHELL_PATH@|$(SHELL_PATH_SQ)|' \
             -e 's|@@PERL@@|$(PERL_PATH_SQ)|g' \
             -e 's/@@PERF_VERSION@@/$(PERF_VERSION)/g' \
             -e 's/@@NO_CURL@@/$(NO_CURL)/g' \
-            $@.sh >$@+ && \
-        chmod +x $@+ && \
-        mv $@+ $(OUTPUT)$@
+            $@.sh > $(OUTPUT)$@+ && \
+        chmod +x $(OUTPUT)$@+ && \
+        mv $(OUTPUT)$@+ $(OUTPUT)$@
 
 configure: configure.ac
         $(QUIET_GEN)$(RM) $@ $<+ && \
@@ -966,7 +998,16 @@ $(OUTPUT)builtin-init-db.o: builtin-init-db.c $(OUTPUT)PERF-CFLAGS
 $(OUTPUT)util/config.o: util/config.c $(OUTPUT)PERF-CFLAGS
         $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
 
-$(OUTPUT)util/newt.o: util/newt.c $(OUTPUT)PERF-CFLAGS
+$(OUTPUT)util/ui/browser.o: util/ui/browser.c $(OUTPUT)PERF-CFLAGS
+        $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $<
+
+$(OUTPUT)util/ui/browsers/annotate.o: util/ui/browsers/annotate.c $(OUTPUT)PERF-CFLAGS
+        $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $<
+
+$(OUTPUT)util/ui/browsers/hists.o: util/ui/browsers/hists.c $(OUTPUT)PERF-CFLAGS
+        $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $<
+
+$(OUTPUT)util/ui/browsers/map.o: util/ui/browsers/map.c $(OUTPUT)PERF-CFLAGS
         $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $<
 
 $(OUTPUT)util/rbtree.o: ../../lib/rbtree.c $(OUTPUT)PERF-CFLAGS
@@ -991,6 +1032,14 @@ $(LIB_OBJS) $(BUILTIN_OBJS): $(LIB_H)
 $(patsubst perf-%$X,%.o,$(PROGRAMS)): $(LIB_H) $(wildcard */*.h)
 builtin-revert.o wt-status.o: wt-status.h
 
+# we compile into subdirectories. if the target directory is not the source directory, they might not exists. So
+# we depend the various files onto their directories.
+DIRECTORY_DEPS = $(LIB_OBJS) $(BUILTIN_OBJS) $(OUTPUT)PERF-VERSION-FILE $(OUTPUT)common-cmds.h
+$(DIRECTORY_DEPS): | $(sort $(dir $(DIRECTORY_DEPS)))
+# In the second step, we make a rule to actually create these directories
+$(sort $(dir $(DIRECTORY_DEPS))):
+        $(QUIET_MKDIR)$(MKDIR) -p $@ 2>/dev/null
+
 $(LIB_FILE): $(LIB_OBJS)
         $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIB_OBJS)
 
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index fd20670ce986..6d5604d8df95 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -28,7 +28,7 @@
 
 static char             const *input_name = "perf.data";
 
-static bool             force;
+static bool             force, use_tui, use_stdio;
 
 static bool             full_paths;
 
@@ -285,7 +285,7 @@ static int hist_entry__tty_annotate(struct hist_entry *he)
         LIST_HEAD(head);
         struct objdump_line *pos, *n;
 
-        if (hist_entry__annotate(he, &head) < 0)
+        if (hist_entry__annotate(he, &head, 0) < 0)
                 return -1;
 
         if (full_paths)
@@ -321,7 +321,7 @@ static int hist_entry__tty_annotate(struct hist_entry *he)
 
 static void hists__find_annotations(struct hists *self)
 {
-        struct rb_node *first = rb_first(&self->entries), *nd = first;
+        struct rb_node *nd = rb_first(&self->entries), *next;
         int key = KEY_RIGHT;
 
         while (nd) {
@@ -343,20 +343,19 @@ find_next:
 
                 if (use_browser > 0) {
                         key = hist_entry__tui_annotate(he);
-                        if (is_exit_key(key))
-                                break;
                         switch (key) {
                         case KEY_RIGHT:
-                        case '\t':
-                                nd = rb_next(nd);
+                                next = rb_next(nd);
                                 break;
                         case KEY_LEFT:
-                                if (nd == first)
-                                        continue;
-                                nd = rb_prev(nd);
-                        default:
+                                next = rb_prev(nd);
                                 break;
+                        default:
+                                return;
                         }
+
+                        if (next != NULL)
+                                nd = next;
                 } else {
                         hist_entry__tty_annotate(he);
                         nd = rb_next(nd);
@@ -428,6 +427,8 @@ static const struct option options[] = {
                     "be more verbose (show symbol address, etc)"),
         OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
                     "dump raw trace in ASCII"),
+        OPT_BOOLEAN(0, "tui", &use_tui, "Use the TUI interface"),
+        OPT_BOOLEAN(0, "stdio", &use_stdio, "Use the stdio interface"),
         OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name,
                    "file", "vmlinux pathname"),
         OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules,
@@ -443,6 +444,11 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used)
 {
         argc = parse_options(argc, argv, options, annotate_usage, 0);
 
+        if (use_stdio)
+                use_browser = 0;
+        else if (use_tui)
+                use_browser = 1;
+
         setup_browser();
 
         symbol_conf.priv_size = sizeof(struct sym_priv);
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
index 44a47e13bd67..c49837de7d3f 100644
--- a/tools/perf/builtin-buildid-list.c
+++ b/tools/perf/builtin-buildid-list.c
@@ -36,7 +36,6 @@ static const struct option options[] = {
 
 static int __cmd_buildid_list(void)
 {
-        int err = -1;
         struct perf_session *session;
 
         session = perf_session__new(input_name, O_RDONLY, force, false);
@@ -49,7 +48,7 @@ static int __cmd_buildid_list(void)
         perf_session__fprintf_dsos_buildid(session, stdout, with_hits);
 
         perf_session__delete(session);
-        return err;
+        return 0;
 }
 
 int cmd_buildid_list(int argc, const char **argv, const char *prefix __used)
diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c
index 199d5e19554f..add163c9f0e7 100644
--- a/tools/perf/builtin-probe.c
+++ b/tools/perf/builtin-probe.c
@@ -50,14 +50,17 @@ static struct {
         bool list_events;
         bool force_add;
         bool show_lines;
+        bool show_vars;
+        bool show_ext_vars;
+        bool mod_events;
         int nevents;
         struct perf_probe_event events[MAX_PROBES];
         struct strlist *dellist;
         struct line_range line_range;
+        const char *target_module;
         int max_probe_points;
 } params;
 
-
 /* Parse an event definition. Note that any error must die. */
 static int parse_probe_event(const char *str)
 {
@@ -92,6 +95,7 @@ static int parse_probe_event_argv(int argc, const char **argv)
         len = 0;
         for (i = 0; i < argc; i++)
                 len += sprintf(&buf[len], "%s ", argv[i]);
+        params.mod_events = true;
         ret = parse_probe_event(buf);
         free(buf);
         return ret;
@@ -100,9 +104,10 @@ static int parse_probe_event_argv(int argc, const char **argv)
 static int opt_add_probe_event(const struct option *opt __used,
                               const char *str, int unset __used)
 {
-        if (str)
+        if (str) {
+                params.mod_events = true;
                 return parse_probe_event(str);
-        else
+        } else
                 return 0;
 }
 
@@ -110,6 +115,7 @@ static int opt_del_probe_event(const struct option *opt __used,
                                const char *str, int unset __used)
 {
         if (str) {
+                params.mod_events = true;
                 if (!params.dellist)
                         params.dellist = strlist__new(true, NULL);
                 strlist__add(params.dellist, str);
@@ -130,6 +136,25 @@ static int opt_show_lines(const struct option *opt __used,
 
         return ret;
 }
+
+static int opt_show_vars(const struct option *opt __used,
+                         const char *str, int unset __used)
+{
+        struct perf_probe_event *pev = &params.events[params.nevents];
+        int ret;
+
+        if (!str)
+                return 0;
+
+        ret = parse_probe_event(str);
+        if (!ret && pev->nargs != 0) {
+                pr_err("  Error: '--vars' doesn't accept arguments.\n");
+                return -EINVAL;
+        }
+        params.show_vars = true;
+
+        return ret;
+}
 #endif
 
 static const char * const probe_usage[] = {
@@ -138,7 +163,8 @@ static const char * const probe_usage[] = {
         "perf probe [<options>] --del '[GROUP:]EVENT' ...",
         "perf probe --list",
 #ifdef DWARF_SUPPORT
-        "perf probe --line 'LINEDESC'",
+        "perf probe [<options>] --line 'LINEDESC'",
+        "perf probe [<options>] --vars 'PROBEPOINT'",
 #endif
         NULL
 };
@@ -180,10 +206,17 @@ static const struct option options[] = {
         OPT_CALLBACK('L', "line", NULL,
                      "FUNC[:RLN[+NUM|-RLN2]]|SRC:ALN[+NUM|-ALN2]",
                      "Show source code lines.", opt_show_lines),
+        OPT_CALLBACK('V', "vars", NULL,
+                     "FUNC[@SRC][+OFF|%return|:RL|;PT]|SRC:AL|SRC;PT",
+                     "Show accessible variables on PROBEDEF", opt_show_vars),
+        OPT_BOOLEAN('\0', "externs", &params.show_ext_vars,
+                    "Show external variables too (with --vars only)"),
         OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name,
                    "file", "vmlinux pathname"),
         OPT_STRING('s', "source", &symbol_conf.source_prefix,
                    "directory", "path to kernel source"),
+        OPT_STRING('m', "module", &params.target_module,
+                   "modname", "target module name"),
 #endif
         OPT__DRY_RUN(&probe_event_dry_run),
         OPT_INTEGER('\0', "max-probes", &params.max_probe_points,
@@ -216,8 +249,13 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
              !params.show_lines))
                 usage_with_options(probe_usage, options);
 
+        /*
+         * Only consider the user's kernel image path if given.
+         */
+        symbol_conf.try_vmlinux_path = (symbol_conf.vmlinux_name == NULL);
+
         if (params.list_events) {
-                if (params.nevents != 0 || params.dellist) {
+                if (params.mod_events) {
                         pr_err("  Error: Don't use --list with --add/--del.\n");
                         usage_with_options(probe_usage, options);
                 }
@@ -225,6 +263,10 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
                         pr_err("  Error: Don't use --list with --line.\n");
                         usage_with_options(probe_usage, options);
                 }
+                if (params.show_vars) {
+                        pr_err(" Error: Don't use --list with --vars.\n");
+                        usage_with_options(probe_usage, options);
+                }
                 ret = show_perf_probe_events();
                 if (ret < 0)
                         pr_err("  Error: Failed to show event list. (%d)\n",
@@ -234,17 +276,35 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 
 #ifdef DWARF_SUPPORT
         if (params.show_lines) {
-                if (params.nevents != 0 || params.dellist) {
-                        pr_warning("  Error: Don't use --line with"
-                                   " --add/--del.\n");
+                if (params.mod_events) {
+                        pr_err("  Error: Don't use --line with"
+                               " --add/--del.\n");
+                        usage_with_options(probe_usage, options);
+                }
+                if (params.show_vars) {
+                        pr_err(" Error: Don't use --line with --vars.\n");
                         usage_with_options(probe_usage, options);
                 }
 
-                ret = show_line_range(&params.line_range);
+                ret = show_line_range(&params.line_range, params.target_module);
                 if (ret < 0)
                         pr_err("  Error: Failed to show lines. (%d)\n", ret);
                 return ret;
         }
+        if (params.show_vars) {
+                if (params.mod_events) {
+                        pr_err("  Error: Don't use --vars with"
+                               " --add/--del.\n");
+                        usage_with_options(probe_usage, options);
+                }
+                ret = show_available_vars(params.events, params.nevents,
+                                          params.max_probe_points,
+                                          params.target_module,
+                                          params.show_ext_vars);
+                if (ret < 0)
+                        pr_err("  Error: Failed to show vars. (%d)\n", ret);
+                return ret;
+        }
 #endif
 
         if (params.dellist) {
@@ -258,8 +318,9 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 
         if (params.nevents) {
                 ret = add_perf_probe_events(params.events, params.nevents,
-                                            params.force_add,
-                                            params.max_probe_points);
+                                            params.max_probe_points,
+                                            params.target_module,
+                                            params.force_add);
                 if (ret < 0) {
                         pr_err("  Error: Failed to add events. (%d)\n", ret);
                         return ret;
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index ff77b805de71..564491fa18b2 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -197,7 +197,7 @@ static void sig_atexit(void)
         if (child_pid > 0)
                 kill(child_pid, SIGTERM);
 
-        if (signr == -1)
+        if (signr == -1 || signr == SIGUSR1)
                 return;
 
         signal(signr, SIG_DFL);
@@ -353,7 +353,7 @@ try_again:
                 }
 
                 if (read(fd[nr_cpu][counter][thread_index], &read_data, sizeof(read_data)) == -1) {
-                        perror("Unable to read perf file descriptor\n");
+                        perror("Unable to read perf file descriptor");
                         exit(-1);
                 }
 
@@ -515,6 +515,7 @@ static int __cmd_record(int argc, const char **argv)
         atexit(sig_atexit);
         signal(SIGCHLD, sig_handler);
         signal(SIGINT, sig_handler);
+        signal(SIGUSR1, sig_handler);
 
         if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) {
                 perror("failed to create pipes");
@@ -606,6 +607,7 @@ static int __cmd_record(int argc, const char **argv)
                         execvp(argv[0], (char **)argv);
 
                         perror(argv[0]);
+                        kill(getppid(), SIGUSR1);
                         exit(-1);
                 }
 
@@ -626,7 +628,7 @@ static int __cmd_record(int argc, const char **argv)
 
         nr_cpus = read_cpu_map(cpu_list);
         if (nr_cpus < 1) {
-                perror("failed to collect number of CPUs\n");
+                perror("failed to collect number of CPUs");
                 return -1;
         }
 
@@ -697,17 +699,18 @@ static int __cmd_record(int argc, const char **argv)
         if (err < 0)
                 err = event__synthesize_kernel_mmap(process_synthesized_event,
                                                     session, machine, "_stext");
-        if (err < 0) {
-                pr_err("Couldn't record kernel reference relocation symbol.\n");
-                return err;
-        }
+        if (err < 0)
+                pr_err("Couldn't record kernel reference relocation symbol\n"
+                       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
+                       "Check /proc/kallsyms permission or run as root.\n");
 
         err = event__synthesize_modules(process_synthesized_event,
                                         session, machine);
-        if (err < 0) {
-                pr_err("Couldn't record kernel reference relocation symbol.\n");
-                return err;
-        }
+        if (err < 0)
+                pr_err("Couldn't record kernel module information.\n"
+                       "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n"
+                       "Check /proc/modules permission or run as root.\n");
+
         if (perf_guest)
                 perf_session__process_machines(session, event__synthesize_guest_os);
 
@@ -761,6 +764,9 @@ static int __cmd_record(int argc, const char **argv)
                 }
         }
 
+        if (quiet || signr == SIGUSR1)
+                return 0;
+
         fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking);
 
         /*
@@ -787,7 +793,7 @@ static const char * const record_usage[] = {
 
 static bool force, append_file;
 
-static const struct option options[] = {
+const struct option record_options[] = {
         OPT_CALLBACK('e', "event", NULL, "event",
                      "event selector. use 'perf list' to list available events",
                      parse_events),
@@ -820,6 +826,7 @@ static const struct option options[] = {
                     "do call-graph (stack chain/backtrace) recording"),
         OPT_INCR('v', "verbose", &verbose,
                     "be more verbose (show counter open errors, etc)"),
+        OPT_BOOLEAN('q', "quiet", &quiet, "don't print any message"),
         OPT_BOOLEAN('s', "stat", &inherit_stat,
                     "per thread counts"),
         OPT_BOOLEAN('d', "data", &sample_address,
@@ -835,16 +842,16 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
 {
         int i, j, err = -ENOMEM;
 
-        argc = parse_options(argc, argv, options, record_usage,
+        argc = parse_options(argc, argv, record_options, record_usage,
                             PARSE_OPT_STOP_AT_NON_OPTION);
         if (!argc && target_pid == -1 && target_tid == -1 &&
                 !system_wide && !cpu_list)
-                usage_with_options(record_usage, options);
+                usage_with_options(record_usage, record_options);
 
         if (force && append_file) {
                 fprintf(stderr, "Can't overwrite and append at the same time."
                                 " You need to choose between -f and -A");
-                usage_with_options(record_usage, options);
+                usage_with_options(record_usage, record_options);
         } else if (append_file) {
                 write_mode = WRITE_APPEND;
         } else {
@@ -867,7 +874,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
                 if (thread_num <= 0) {
                         fprintf(stderr, "Can't find all threads of pid %d\n",
                                         target_pid);
-                        usage_with_options(record_usage, options);
+                        usage_with_options(record_usage, record_options);
                 }
         } else {
                 all_tids=malloc(sizeof(pid_t));
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 2f4b92925b26..5de405d45230 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -32,7 +32,7 @@
 
 static char             const *input_name = "perf.data";
 
-static bool             force;
+static bool             force, use_tui, use_stdio;
 static bool             hide_unresolved;
 static bool             dont_use_callchains;
 
@@ -107,7 +107,8 @@ static int perf_session__add_hist_entry(struct perf_session *self,
                 goto out_free_syms;
         err = 0;
         if (symbol_conf.use_callchain) {
-                err = append_chain(he->callchain, data->callchain, syms, data->period);
+                err = callchain_append(he->callchain, data->callchain, syms,
+                                       data->period);
                 if (err)
                         goto out_free_syms;
         }
@@ -348,7 +349,18 @@ static int __cmd_report(void)
                 hists__tty_browse_tree(&session->hists_tree, help);
 
 out_delete:
-        perf_session__delete(session);
+        /*
+         * Speed up the exit process, for large files this can
+         * take quite a while.
+         *
+         * XXX Enable this when using valgrind or if we ever
+         * librarize this command.
+         *
+         * Also experiment with obstacks to see how much speed
+         * up we'll get here.
+         *
+         * perf_session__delete(session);
+         */
         return ret;
 }
 
@@ -439,6 +451,8 @@ static const struct option options[] = {
                     "Show per-thread event counters"),
         OPT_STRING(0, "pretty", &pretty_printing_style, "key",
                    "pretty printing style key: normal raw"),
+        OPT_BOOLEAN(0, "tui", &use_tui, "Use the TUI interface"),
+        OPT_BOOLEAN(0, "stdio", &use_stdio, "Use the stdio interface"),
         OPT_STRING('s', "sort", &sort_order, "key[,key2...]",
                    "sort by key(s): pid, comm, dso, symbol, parent"),
         OPT_BOOLEAN(0, "showcpuutilization", &symbol_conf.show_cpu_utilization,
@@ -471,15 +485,38 @@ int cmd_report(int argc, const char **argv, const char *prefix __used)
 {
         argc = parse_options(argc, argv, options, report_usage, 0);
 
+        if (use_stdio)
+                use_browser = 0;
+        else if (use_tui)
+                use_browser = 1;
+
         if (strcmp(input_name, "-") != 0)
                 setup_browser();
+        else
+                use_browser = 0;
         /*
          * Only in the newt browser we are doing integrated annotation,
          * so don't allocate extra space that won't be used in the stdio
          * implementation.
          */
-        if (use_browser > 0)
+        if (use_browser > 0) {
                 symbol_conf.priv_size = sizeof(struct sym_priv);
+                /*
+                 * For searching by name on the "Browse map details".
+                 * providing it only in verbose mode not to bloat too
+                 * much struct symbol.
+                 */
+                if (verbose) {
+                        /*
+                         * XXX: Need to provide a less kludgy way to ask for
+                         * more space per symbol, the u32 is for the index on
+                         * the ui browser.
+                         * See symbol__browser_index.
+                         */
+                        symbol_conf.priv_size += sizeof(u32);
+                        symbol_conf.sort_by_name = true;
+                }
+        }
 
         if (symbol__init() < 0)
                 return -1;
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 5161619d4714..9bcc38f0b706 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -455,8 +455,8 @@ static void sched_switch(int cpu, u64 timestamp, struct trace_entry *te)
                 if (p->current->state != TYPE_NONE)
                         pid_put_sample(sw->next_pid, p->current->state, cpu, p->current->state_since, timestamp);
 
-                        p->current->state_since = timestamp;
-                        p->current->state = TYPE_RUNNING;
+                p->current->state_since = timestamp;
+                p->current->state = TYPE_RUNNING;
         }
 
         if (prev_p->current) {
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index b513e40974f4..dd625808c2a5 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -69,7 +69,6 @@ static int			target_tid			=     -1;
 static pid_t                    *all_tids                       =      NULL;
 static int                      thread_num                      =      0;
 static bool                     inherit                         =  false;
-static int                      profile_cpu                     =     -1;
 static int                      nr_cpus                         =      0;
 static int                      realtime_prio                   =      0;
 static bool                     group                           =  false;
@@ -558,13 +557,13 @@ static void print_sym_table(void)
         else
                 printf(" (all");
 
-        if (profile_cpu != -1)
-                printf(", cpu: %d)\n", profile_cpu);
+        if (cpu_list)
+                printf(", CPU%s: %s)\n", nr_cpus > 1 ? "s" : "", cpu_list);
         else {
                 if (target_tid != -1)
                         printf(")\n");
                 else
-                        printf(", %d CPUs)\n", nr_cpus);
+                        printf(", %d CPU%s)\n", nr_cpus, nr_cpus > 1 ? "s" : "");
         }
 
         printf("%-*.*s\n", win_width, win_width, graph_dotted_line);
@@ -1187,11 +1186,10 @@ int group_fd;
 static void start_counter(int i, int counter)
 {
         struct perf_event_attr *attr;
-        int cpu;
+        int cpu = -1;
         int thread_index;
 
-        cpu = profile_cpu;
-        if (target_tid == -1 && profile_cpu == -1)
+        if (target_tid == -1)
                 cpu = cpumap[i];
 
         attr = attrs + counter;
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 294da725a57d..86cfe3800e6b 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -1,19 +1,24 @@
 #include "builtin.h"
 
-#include "util/util.h"
+#include "perf.h"
 #include "util/cache.h"
+#include "util/debug.h"
+#include "util/exec_cmd.h"
+#include "util/header.h"
+#include "util/parse-options.h"
+#include "util/session.h"
 #include "util/symbol.h"
 #include "util/thread.h"
-#include "util/header.h"
-#include "util/exec_cmd.h"
 #include "util/trace-event.h"
-#include "util/session.h"
+#include "util/parse-options.h"
+#include "util/util.h"
 
 static char const               *script_name;
 static char const               *generate_script_lang;
 static bool                     debug_mode;
 static u64                      last_timestamp;
 static u64                      nr_unordered;
+extern const struct option      record_options[];
 
 static int default_start_script(const char *script __unused,
                                 int argc __unused,
@@ -43,9 +48,6 @@ static struct scripting_ops	*scripting_ops;
 
 static void setup_scripting(void)
 {
-        /* make sure PERF_EXEC_PATH is set for scripts */
-        perf_set_argv_exec_path(perf_exec_path());
-
         setup_perl_scripting();
         setup_python_scripting();
 
@@ -59,14 +61,6 @@ static int cleanup_scripting(void)
         return scripting_ops->stop_script();
 }
 
-#include "util/parse-options.h"
-
-#include "perf.h"
-#include "util/debug.h"
-
-#include "util/trace-event.h"
-#include "util/exec_cmd.h"
-
 static char const               *input_name = "perf.data";
 
 static int process_sample_event(event_t *event, struct perf_session *session)
@@ -290,7 +284,7 @@ static int parse_scriptname(const struct option *opt __used,
                 script++;
         } else {
                 script = str;
-                ext = strchr(script, '.');
+                ext = strrchr(script, '.');
                 if (!ext) {
                         fprintf(stderr, "invalid script extension");
                         return -1;
@@ -336,7 +330,7 @@ static struct script_desc *script_desc__new(const char *name)
 {
         struct script_desc *s = zalloc(sizeof(*s));
 
-        if (s != NULL)
+        if (s != NULL && name)
                 s->name = strdup(name);
 
         return s;
@@ -345,6 +339,8 @@ static struct script_desc *script_desc__new(const char *name)
 static void script_desc__delete(struct script_desc *s)
 {
         free(s->name);
+        free(s->half_liner);
+        free(s->args);
         free(s);
 }
 
@@ -545,8 +541,40 @@ static char *get_script_path(const char *script_root, const char *suffix)
         return path;
 }
 
+static bool is_top_script(const char *script_path)
+{
+        return ends_with((char *)script_path, "top") == NULL ? false : true;
+}
+
+static int has_required_arg(char *script_path)
+{
+        struct script_desc *desc;
+        int n_args = 0;
+        char *p;
+
+        desc = script_desc__new(NULL);
+
+        if (read_script_info(desc, script_path))
+                goto out;
+
+        if (!desc->args)
+                goto out;
+
+        for (p = desc->args; *p; p++)
+                if (*p == '<')
+                        n_args++;
+out:
+        script_desc__delete(desc);
+
+        return n_args;
+}
+
 static const char * const trace_usage[] = {
-        "perf trace [<options>] <command>",
+        "perf trace [<options>]",
+        "perf trace [<options>] record <script> [<record-options>] <command>",
+        "perf trace [<options>] report <script> [script-args]",
+        "perf trace [<options>] <script> [<record-options>] <command>",
+        "perf trace [<options>] <top-script> [script-args]",
         NULL
 };
 
@@ -572,47 +600,81 @@ static const struct option options[] = {
         OPT_END()
 };
 
+static bool have_cmd(int argc, const char **argv)
+{
+        char **__argv = malloc(sizeof(const char *) * argc);
+
+        if (!__argv)
+                die("malloc");
+        memcpy(__argv, argv, sizeof(const char *) * argc);
+        argc = parse_options(argc, (const char **)__argv, record_options,
+                             NULL, PARSE_OPT_STOP_AT_NON_OPTION);
+        free(__argv);
+
+        return argc != 0;
+}
+
 int cmd_trace(int argc, const char **argv, const char *prefix __used)
 {
+        char *rec_script_path = NULL;
+        char *rep_script_path = NULL;
         struct perf_session *session;
-        const char *suffix = NULL;
+        char *script_path = NULL;
         const char **__argv;
-        char *script_path;
-        int i, err;
+        bool system_wide;
+        int i, j, err;
 
-        if (argc >= 2 && strncmp(argv[1], "rec", strlen("rec")) == 0) {
-                if (argc < 3) {
-                        fprintf(stderr,
-                                "Please specify a record script\n");
-                        return -1;
-                }
-                suffix = RECORD_SUFFIX;
+        setup_scripting();
+
+        argc = parse_options(argc, argv, options, trace_usage,
+                             PARSE_OPT_STOP_AT_NON_OPTION);
+
+        if (argc > 1 && !strncmp(argv[0], "rec", strlen("rec"))) {
+                rec_script_path = get_script_path(argv[1], RECORD_SUFFIX);
+                if (!rec_script_path)
+                        return cmd_record(argc, argv, NULL);
         }
 
-        if (argc >= 2 && strncmp(argv[1], "rep", strlen("rep")) == 0) {
-                if (argc < 3) {
+        if (argc > 1 && !strncmp(argv[0], "rep", strlen("rep"))) {
+                rep_script_path = get_script_path(argv[1], REPORT_SUFFIX);
+                if (!rep_script_path) {
                         fprintf(stderr,
-                                "Please specify a report script\n");
+                                "Please specify a valid report script"
+                                "(see 'perf trace -l' for listing)\n");
                         return -1;
                 }
-                suffix = REPORT_SUFFIX;
         }
 
-        if (!suffix && argc >= 2 && strncmp(argv[1], "-", strlen("-")) != 0) {
-                char *record_script_path, *report_script_path;
+        /* make sure PERF_EXEC_PATH is set for scripts */
+        perf_set_argv_exec_path(perf_exec_path());
+
+        if (argc && !script_name && !rec_script_path && !rep_script_path) {
                 int live_pipe[2];
+                int rep_args;
                 pid_t pid;
 
-                record_script_path = get_script_path(argv[1], RECORD_SUFFIX);
-                if (!record_script_path) {
-                        fprintf(stderr, "record script not found\n");
-                        return -1;
+                rec_script_path = get_script_path(argv[0], RECORD_SUFFIX);
+                rep_script_path = get_script_path(argv[0], REPORT_SUFFIX);
+
+                if (!rec_script_path && !rep_script_path) {
+                        fprintf(stderr, " Couldn't find script %s\n\n See perf"
+                                " trace -l for available scripts.\n", argv[0]);
+                        usage_with_options(trace_usage, options);
                 }
 
-                report_script_path = get_script_path(argv[1], REPORT_SUFFIX);
-                if (!report_script_path) {
-                        fprintf(stderr, "report script not found\n");
-                        return -1;
+                if (is_top_script(argv[0])) {
+                        rep_args = argc - 1;
+                } else {
+                        int rec_args;
+
+                        rep_args = has_required_arg(rep_script_path);
+                        rec_args = (argc - 1) - rep_args;
+                        if (rec_args < 0) {
+                                fprintf(stderr, " %s script requires options."
+                                        "\n\n See perf trace -l for available "
+                                        "scripts and options.\n", argv[0]);
+                                usage_with_options(trace_usage, options);
+                        }
                 }
 
                 if (pipe(live_pipe) < 0) {
@@ -627,59 +689,84 @@ int cmd_trace(int argc, const char **argv, const char *prefix __used)
                 }
 
                 if (!pid) {
+                        system_wide = true;
+                        j = 0;
+
                         dup2(live_pipe[1], 1);
                         close(live_pipe[0]);
 
-                        __argv = malloc(5 * sizeof(const char *));
-                        __argv[0] = "/bin/sh";
-                        __argv[1] = record_script_path;
-                        __argv[2] = "-o";
-                        __argv[3] = "-";
-                        __argv[4] = NULL;
+                        if (!is_top_script(argv[0]))
+                                system_wide = !have_cmd(argc - rep_args,
+                                                        &argv[rep_args]);
+
+                        __argv = malloc((argc + 6) * sizeof(const char *));
+                        if (!__argv)
+                                die("malloc");
+
+                        __argv[j++] = "/bin/sh";
+                        __argv[j++] = rec_script_path;
+                        if (system_wide)
+                                __argv[j++] = "-a";
+                        __argv[j++] = "-q";
+                        __argv[j++] = "-o";
+                        __argv[j++] = "-";
+                        for (i = rep_args + 1; i < argc; i++)
+                                __argv[j++] = argv[i];
+                        __argv[j++] = NULL;
 
                         execvp("/bin/sh", (char **)__argv);
+                        free(__argv);
                         exit(-1);
                 }
 
                 dup2(live_pipe[0], 0);
                 close(live_pipe[1]);
 
-                __argv = malloc((argc + 3) * sizeof(const char *));
-                __argv[0] = "/bin/sh";
-                __argv[1] = report_script_path;
-                for (i = 2; i < argc; i++)
-                        __argv[i] = argv[i];
-                __argv[i++] = "-i";
-                __argv[i++] = "-";
-                __argv[i++] = NULL;
+                __argv = malloc((argc + 4) * sizeof(const char *));
+                if (!__argv)
+                        die("malloc");
+                j = 0;
+                __argv[j++] = "/bin/sh";
+                __argv[j++] = rep_script_path;
+                for (i = 1; i < rep_args + 1; i++)
+                        __argv[j++] = argv[i];
+                __argv[j++] = "-i";
+                __argv[j++] = "-";
+                __argv[j++] = NULL;
 
                 execvp("/bin/sh", (char **)__argv);
+                free(__argv);
                 exit(-1);
         }
 
-        if (suffix) {
-                script_path = get_script_path(argv[2], suffix);
-                if (!script_path) {
-                        fprintf(stderr, "script not found\n");
-                        return -1;
-                }
-
-                __argv = malloc((argc + 1) * sizeof(const char *));
-                __argv[0] = "/bin/sh";
-                __argv[1] = script_path;
-                for (i = 3; i < argc; i++)
-                        __argv[i - 1] = argv[i];
-                __argv[argc - 1] = NULL;
+        if (rec_script_path)
+                script_path = rec_script_path;
+        if (rep_script_path)
+                script_path = rep_script_path;
+
+        if (script_path) {
+                system_wide = false;
+                j = 0;
+
+                if (rec_script_path)
+                        system_wide = !have_cmd(argc - 1, &argv[1]);
+
+                __argv = malloc((argc + 2) * sizeof(const char *));
+                if (!__argv)
+                        die("malloc");
+                __argv[j++] = "/bin/sh";
+                __argv[j++] = script_path;
+                if (system_wide)
+                        __argv[j++] = "-a";
+                for (i = 2; i < argc; i++)
+                        __argv[j++] = argv[i];
+                __argv[j++] = NULL;
 
                 execvp("/bin/sh", (char **)__argv);
+                free(__argv);
                 exit(-1);
         }
 
-        setup_scripting();
-
-        argc = parse_options(argc, argv, options, trace_usage,
-                             PARSE_OPT_STOP_AT_NON_OPTION);
-
         if (symbol__init() < 0)
                 return -1;
         if (!script_name)
diff --git a/tools/perf/feature-tests.mak b/tools/perf/feature-tests.mak
index ddb68e601f0e..b253db634f04 100644
--- a/tools/perf/feature-tests.mak
+++ b/tools/perf/feature-tests.mak
@@ -110,10 +110,21 @@ int main(void)
 }
 endef
 
+define SOURCE_STRLCPY
+#include <stdlib.h>
+extern size_t strlcpy(char *dest, const char *src, size_t size);
+
+int main(void)
+{
+        strlcpy(NULL, NULL, 0);
+        return 0;
+}
+endef
+
 # try-cc
 # Usage: option = $(call try-cc, source-to-build, cc-options)
 try-cc = $(shell sh -c                                            \
-        'TMP="$(TMPOUT).$$$$";                                    \
+        'TMP="$(OUTPUT)$(TMPOUT).$$$$";                           \
          echo "$(1)" |                                            \
          $(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
          rm -f "$$TMP"')
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index ef7aa0a0c526..95aaf565c704 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -73,6 +73,18 @@ void get_term_dimensions(struct winsize *ws);
 #define cpu_relax()     asm volatile("":::"memory")
 #endif
 
+#ifdef __mips__
+#include "../../arch/mips/include/asm/unistd.h"
+#define rmb()           asm volatile(                                   \
+                                ".set   mips2\n\t"                      \
+                                "sync\n\t"                              \
+                                ".set   mips0"                          \
+                                : /* no output */                       \
+                                : /* no input */                        \
+                                : "memory")
+#define cpu_relax()     asm volatile("" ::: "memory")
+#endif
+
 #include <time.h>
 #include <unistd.h>
 #include <sys/types.h>
diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-record b/tools/perf/scripts/perl/bin/failed-syscalls-record
index eb5846bcb565..8104895a7b67 100644
--- a/tools/perf/scripts/perl/bin/failed-syscalls-record
+++ b/tools/perf/scripts/perl/bin/failed-syscalls-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_exit $@
+perf record -e raw_syscalls:sys_exit $@
diff --git a/tools/perf/scripts/perl/bin/failed-syscalls-report b/tools/perf/scripts/perl/bin/failed-syscalls-report
index e3a5e55d54ff..4028d92dc4ae 100644
--- a/tools/perf/scripts/perl/bin/failed-syscalls-report
+++ b/tools/perf/scripts/perl/bin/failed-syscalls-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
         shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/failed-syscalls.pl $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/failed-syscalls.pl $comm
diff --git a/tools/perf/scripts/perl/bin/rw-by-file-record b/tools/perf/scripts/perl/bin/rw-by-file-record
index 5bfaae5a6cba..33efc8673aae 100644
--- a/tools/perf/scripts/perl/bin/rw-by-file-record
+++ b/tools/perf/scripts/perl/bin/rw-by-file-record
@@ -1,3 +1,3 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_enter_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_enter_write $@
 
diff --git a/tools/perf/scripts/perl/bin/rw-by-file-report b/tools/perf/scripts/perl/bin/rw-by-file-report
index d83070b7eeb5..ba25f4d41fb0 100644
--- a/tools/perf/scripts/perl/bin/rw-by-file-report
+++ b/tools/perf/scripts/perl/bin/rw-by-file-report
@@ -7,7 +7,7 @@ if [ $# -lt 1 ] ; then
 fi
 comm=$1
 shift
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rw-by-file.pl $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rw-by-file.pl $comm
 
 
 
diff --git a/tools/perf/scripts/perl/bin/rw-by-pid-record b/tools/perf/scripts/perl/bin/rw-by-pid-record
index 6e0b2f7755ac..7cb9db230448 100644
--- a/tools/perf/scripts/perl/bin/rw-by-pid-record
+++ b/tools/perf/scripts/perl/bin/rw-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
diff --git a/tools/perf/scripts/perl/bin/rw-by-pid-report b/tools/perf/scripts/perl/bin/rw-by-pid-report
index 7ef46983f62f..641a3f5d085c 100644
--- a/tools/perf/scripts/perl/bin/rw-by-pid-report
+++ b/tools/perf/scripts/perl/bin/rw-by-pid-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: system-wide r/w activity
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rw-by-pid.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rw-by-pid.pl
 
 
 
diff --git a/tools/perf/scripts/perl/bin/rwtop-record b/tools/perf/scripts/perl/bin/rwtop-record
index 6e0b2f7755ac..7cb9db230448 100644
--- a/tools/perf/scripts/perl/bin/rwtop-record
+++ b/tools/perf/scripts/perl/bin/rwtop-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
+perf record -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write $@
diff --git a/tools/perf/scripts/perl/bin/rwtop-report b/tools/perf/scripts/perl/bin/rwtop-report
index 93e698cd3f38..4918dba77021 100644
--- a/tools/perf/scripts/perl/bin/rwtop-report
+++ b/tools/perf/scripts/perl/bin/rwtop-report
@@ -17,7 +17,7 @@ if [ "$n_args" -gt 0 ] ; then
     interval=$1
     shift
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/rwtop.pl $interval
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/rwtop.pl $interval
 
 
 
diff --git a/tools/perf/scripts/perl/bin/wakeup-latency-record b/tools/perf/scripts/perl/bin/wakeup-latency-record
index 9f2acaaae9f0..464251a1bd7e 100644
--- a/tools/perf/scripts/perl/bin/wakeup-latency-record
+++ b/tools/perf/scripts/perl/bin/wakeup-latency-record
@@ -1,5 +1,5 @@
 #!/bin/bash
-perf record -a -e sched:sched_switch -e sched:sched_wakeup $@
+perf record -e sched:sched_switch -e sched:sched_wakeup $@
 
 
 
diff --git a/tools/perf/scripts/perl/bin/wakeup-latency-report b/tools/perf/scripts/perl/bin/wakeup-latency-report
index a0d898f9ca1d..49052ebcb632 100644
--- a/tools/perf/scripts/perl/bin/wakeup-latency-report
+++ b/tools/perf/scripts/perl/bin/wakeup-latency-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: system-wide min/max/avg wakeup latency
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/wakeup-latency.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/wakeup-latency.pl
 
 
 
diff --git a/tools/perf/scripts/perl/bin/workqueue-stats-record b/tools/perf/scripts/perl/bin/workqueue-stats-record
index 85301f2471ff..8edda9078d5d 100644
--- a/tools/perf/scripts/perl/bin/workqueue-stats-record
+++ b/tools/perf/scripts/perl/bin/workqueue-stats-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e workqueue:workqueue_creation -e workqueue:workqueue_destruction -e workqueue:workqueue_execution -e workqueue:workqueue_insertion $@
+perf record -e workqueue:workqueue_creation -e workqueue:workqueue_destruction -e workqueue:workqueue_execution -e workqueue:workqueue_insertion $@
diff --git a/tools/perf/scripts/perl/bin/workqueue-stats-report b/tools/perf/scripts/perl/bin/workqueue-stats-report
index 35081132ef97..df0c65f4ca93 100644
--- a/tools/perf/scripts/perl/bin/workqueue-stats-report
+++ b/tools/perf/scripts/perl/bin/workqueue-stats-report
@@ -1,6 +1,6 @@
 #!/bin/bash
 # description: workqueue stats (ins/exe/create/destroy)
-perf trace $@ -s ~/libexec/perf-core/scripts/perl/workqueue-stats.pl
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/perl/workqueue-stats.pl
 
 
 
diff --git a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
index 9689bc0acd9f..13cc02b5893a 100644
--- a/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
+++ b/tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py
@@ -6,6 +6,14 @@
 # Public License ("GPL") version 2 as published by the Free Software
 # Foundation.
 
+import errno, os
+
+FUTEX_WAIT = 0
+FUTEX_WAKE = 1
+FUTEX_PRIVATE_FLAG = 128
+FUTEX_CLOCK_REALTIME = 256
+FUTEX_CMD_MASK = ~(FUTEX_PRIVATE_FLAG | FUTEX_CLOCK_REALTIME)
+
 NSECS_PER_SEC    = 1000000000
 
 def avg(total, n):
@@ -24,5 +32,55 @@ def nsecs_str(nsecs):
     str = "%5u.%09u" % (nsecs_secs(nsecs), nsecs_nsecs(nsecs)),
     return str
 
+def add_stats(dict, key, value):
+        if not dict.has_key(key):
+                dict[key] = (value, value, value, 1)
+        else:
+                min, max, avg, count = dict[key]
+                if value < min:
+                        min = value
+                if value > max:
+                        max = value
+                avg = (avg + value) / 2
+                dict[key] = (min, max, avg, count + 1)
+
 def clear_term():
     print("\x1b[H\x1b[2J")
+
+audit_package_warned = False
+
+try:
+        import audit
+        machine_to_id = {
+                'x86_64': audit.MACH_86_64,
+                'alpha' : audit.MACH_ALPHA,
+                'ia64'  : audit.MACH_IA64,
+                'ppc'   : audit.MACH_PPC,
+                'ppc64' : audit.MACH_PPC64,
+                's390'  : audit.MACH_S390,
+                's390x' : audit.MACH_S390X,
+                'i386'  : audit.MACH_X86,
+                'i586'  : audit.MACH_X86,
+                'i686'  : audit.MACH_X86,
+        }
+        try:
+                machine_to_id['armeb'] = audit.MACH_ARMEB
+        except:
+                pass
+        machine_id = machine_to_id[os.uname()[4]]
+except:
+        if not audit_package_warned:
+                audit_package_warned = True
+                print "Install the audit-libs-python package to get syscall names"
+
+def syscall_name(id):
+        try:
+                return audit.audit_syscall_to_name(id, machine_id)
+        except:
+                return str(id)
+
+def strerror(nr):
+        try:
+                return errno.errorcode[abs(nr)]
+        except:
+                return "Unknown %d errno" % nr
diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
index eb5846bcb565..8104895a7b67 100644
--- a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
+++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_exit $@
+perf record -e raw_syscalls:sys_exit $@
diff --git a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
index 30293545fcc2..03587021463d 100644
--- a/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
+++ b/tools/perf/scripts/python/bin/failed-syscalls-by-pid-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
         shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/failed-syscalls-by-pid.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/failed-syscalls-by-pid.py $comm
diff --git a/tools/perf/scripts/python/bin/futex-contention-record b/tools/perf/scripts/python/bin/futex-contention-record
new file mode 100644
index 000000000000..b1495c9a9b20
--- /dev/null
+++ b/tools/perf/scripts/python/bin/futex-contention-record
@@ -0,0 +1,2 @@
+#!/bin/bash
+perf record -e syscalls:sys_enter_futex -e syscalls:sys_exit_futex $@
diff --git a/tools/perf/scripts/python/bin/futex-contention-report b/tools/perf/scripts/python/bin/futex-contention-report
new file mode 100644
index 000000000000..c8268138fb7e
--- /dev/null
+++ b/tools/perf/scripts/python/bin/futex-contention-report
@@ -0,0 +1,4 @@
+#!/bin/bash
+# description: futext contention measurement
+
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/futex-contention.py
diff --git a/tools/perf/scripts/python/bin/netdev-times-record b/tools/perf/scripts/python/bin/netdev-times-record
new file mode 100644
index 000000000000..558754b840a9
--- /dev/null
+++ b/tools/perf/scripts/python/bin/netdev-times-record
@@ -0,0 +1,8 @@
+#!/bin/bash
+perf record -e net:net_dev_xmit -e net:net_dev_queue            \
+                -e net:netif_receive_skb -e net:netif_rx                \
+                -e skb:consume_skb -e skb:kfree_skb                     \
+                -e skb:skb_copy_datagram_iovec -e napi:napi_poll        \
+                -e irq:irq_handler_entry -e irq:irq_handler_exit        \
+                -e irq:softirq_entry -e irq:softirq_exit                \
+                -e irq:softirq_raise $@
diff --git a/tools/perf/scripts/python/bin/netdev-times-report b/tools/perf/scripts/python/bin/netdev-times-report
new file mode 100644
index 000000000000..4ad361b31249
--- /dev/null
+++ b/tools/perf/scripts/python/bin/netdev-times-report
@@ -0,0 +1,5 @@
+#!/bin/bash
+# description: display a process of packet and processing time
+# args: [tx] [rx] [dev=] [debug]
+
+perf trace -s "$PERF_EXEC_PATH"/scripts/python/netdev-times.py $@
diff --git a/tools/perf/scripts/python/bin/sched-migration-record b/tools/perf/scripts/python/bin/sched-migration-record
index 17a3e9bd9e8f..7493fddbe995 100644
--- a/tools/perf/scripts/python/bin/sched-migration-record
+++ b/tools/perf/scripts/python/bin/sched-migration-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -m 16384 -a -e sched:sched_wakeup -e sched:sched_wakeup_new -e sched:sched_switch -e sched:sched_migrate_task $@
+perf record -m 16384 -e sched:sched_wakeup -e sched:sched_wakeup_new -e sched:sched_switch -e sched:sched_migrate_task $@
diff --git a/tools/perf/scripts/python/bin/sched-migration-report b/tools/perf/scripts/python/bin/sched-migration-report
index 61d05f72e443..df1791f07c24 100644
--- a/tools/perf/scripts/python/bin/sched-migration-report
+++ b/tools/perf/scripts/python/bin/sched-migration-report
@@ -1,3 +1,3 @@
 #!/bin/bash
 # description: sched migration overview
-perf trace $@ -s ~/libexec/perf-core/scripts/python/sched-migration.py
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/sched-migration.py
diff --git a/tools/perf/scripts/python/bin/sctop-record b/tools/perf/scripts/python/bin/sctop-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/sctop-record
+++ b/tools/perf/scripts/python/bin/sctop-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/sctop-report b/tools/perf/scripts/python/bin/sctop-report
index b01c842ae7b4..36b409c05e50 100644
--- a/tools/perf/scripts/python/bin/sctop-report
+++ b/tools/perf/scripts/python/bin/sctop-report
@@ -21,4 +21,4 @@ elif [ "$n_args" -gt 0 ] ; then
     interval=$1
     shift
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/sctop.py $comm $interval
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/sctop.py $comm $interval
diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-record b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
+++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/syscall-counts-by-pid-report b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
index 9e9d8ddd72ce..4eb88c9fc83c 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
+++ b/tools/perf/scripts/python/bin/syscall-counts-by-pid-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
         shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/syscall-counts-by-pid.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/syscall-counts-by-pid.py $comm
diff --git a/tools/perf/scripts/python/bin/syscall-counts-record b/tools/perf/scripts/python/bin/syscall-counts-record
index 1fc5998b721d..4efbfaa7f6a5 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-record
+++ b/tools/perf/scripts/python/bin/syscall-counts-record
@@ -1,2 +1,2 @@
 #!/bin/bash
-perf record -a -e raw_syscalls:sys_enter $@
+perf record -e raw_syscalls:sys_enter $@
diff --git a/tools/perf/scripts/python/bin/syscall-counts-report b/tools/perf/scripts/python/bin/syscall-counts-report
index dc076b618796..cb2f9c5cf17e 100644
--- a/tools/perf/scripts/python/bin/syscall-counts-report
+++ b/tools/perf/scripts/python/bin/syscall-counts-report
@@ -7,4 +7,4 @@ if [ $# -gt 0 ] ; then
         shift
     fi
 fi
-perf trace $@ -s ~/libexec/perf-core/scripts/python/syscall-counts.py $comm
+perf trace $@ -s "$PERF_EXEC_PATH"/scripts/python/syscall-counts.py $comm
diff --git a/tools/perf/scripts/python/failed-syscalls-by-pid.py b/tools/perf/scripts/python/failed-syscalls-by-pid.py
index 0ca02278fe69..acd7848717b3 100644
--- a/tools/perf/scripts/python/failed-syscalls-by-pid.py
+++ b/tools/perf/scripts/python/failed-syscalls-by-pid.py
@@ -13,21 +13,26 @@ sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 
 from perf_trace_context import *
 from Core import *
+from Util import *
 
-usage = "perf trace -s syscall-counts-by-pid.py [comm]\n";
+usage = "perf trace -s syscall-counts-by-pid.py [comm|pid]\n";
 
 for_comm = None
+for_pid = None
 
 if len(sys.argv) > 2:
         sys.exit(usage)
 
 if len(sys.argv) > 1:
-        for_comm = sys.argv[1]
+        try:
+                for_pid = int(sys.argv[1])
+        except:
+                for_comm = sys.argv[1]
 
 syscalls = autodict()
 
 def trace_begin():
-        pass
+        print "Press control+C to stop and show the summary"
 
 def trace_end():
         print_error_totals()
@@ -35,9 +40,9 @@ def trace_end():
 def raw_syscalls__sys_exit(event_name, context, common_cpu,
         common_secs, common_nsecs, common_pid, common_comm,
         id, ret):
-        if for_comm is not None:
-                if common_comm != for_comm:
-                        return
+        if (for_comm and common_comm != for_comm) or \
+           (for_pid  and common_pid  != for_pid ):
+                return
 
         if ret < 0:
                 try:
@@ -62,7 +67,7 @@ def print_error_totals():
                     print "\n%s [%d]\n" % (comm, pid),
                     id_keys = syscalls[comm][pid].keys()
                     for id in id_keys:
-                            print "  syscall: %-16d\n" % (id),
+                            print "  syscall: %-16s\n" % syscall_name(id),
                             ret_keys = syscalls[comm][pid][id].keys()
                             for ret, val in sorted(syscalls[comm][pid][id].iteritems(), key = lambda(k, v): (v, k),  reverse = True):
-                                    print "    err = %-20d  %10d\n" % (ret, val),
+                                    print "    err = %-20s  %10d\n" % (strerror(ret), val),
diff --git a/tools/perf/scripts/python/futex-contention.py b/tools/perf/scripts/python/futex-contention.py
new file mode 100644
index 000000000000..11e70a388d41
--- /dev/null
+++ b/tools/perf/scripts/python/futex-contention.py
@@ -0,0 +1,50 @@
+# futex contention
+# (c) 2010, Arnaldo Carvalho de Melo <acme@redhat.com>
+# Licensed under the terms of the GNU GPL License version 2
+#
+# Translation of:
+#
+# http://sourceware.org/systemtap/wiki/WSFutexContention
+#
+# to perf python scripting.
+#
+# Measures futex contention
+
+import os, sys
+sys.path.append(os.environ['PERF_EXEC_PATH'] + '/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
+from Util import *
+
+process_names = {}
+thread_thislock = {}
+thread_blocktime = {}
+
+lock_waits = {} # long-lived stats on (tid,lock) blockage elapsed time
+process_names = {} # long-lived pid-to-execname mapping
+
+def syscalls__sys_enter_futex(event, ctxt, cpu, s, ns, tid, comm,
+                              nr, uaddr, op, val, utime, uaddr2, val3):
+        cmd = op & FUTEX_CMD_MASK
+        if cmd != FUTEX_WAIT:
+                return # we don't care about originators of WAKE events
+
+        process_names[tid] = comm
+        thread_thislock[tid] = uaddr
+        thread_blocktime[tid] = nsecs(s, ns)
+
+def syscalls__sys_exit_futex(event, ctxt, cpu, s, ns, tid, comm,
+                             nr, ret):
+        if thread_blocktime.has_key(tid):
+                elapsed = nsecs(s, ns) - thread_blocktime[tid]
+                add_stats(lock_waits, (tid, thread_thislock[tid]), elapsed)
+                del thread_blocktime[tid]
+                del thread_thislock[tid]
+
+def trace_begin():
+        print "Press control+C to stop and show the summary"
+
+def trace_end():
+        for (tid, lock) in lock_waits:
+                min, max, avg, count = lock_waits[tid, lock]
+                print "%s[%d] lock %x contended %d times, %d avg ns" % \
+                      (process_names[tid], tid, lock, count, avg)
+
diff --git a/tools/perf/scripts/python/netdev-times.py b/tools/perf/scripts/python/netdev-times.py
new file mode 100644
index 000000000000..9aa0a32972e8
--- /dev/null
+++ b/tools/perf/scripts/python/netdev-times.py
@@ -0,0 +1,464 @@
+# Display a process of packets and processed time.
+# It helps us to investigate networking or network device.
+#
+# options
+# tx: show only tx chart
+# rx: show only rx chart
+# dev=: show only thing related to specified device
+# debug: work with debug mode. It shows buffer status.
+
+import os
+import sys
+
+sys.path.append(os.environ['PERF_EXEC_PATH'] + \
+        '/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
+
+from perf_trace_context import *
+from Core import *
+from Util import *
+
+all_event_list = []; # insert all tracepoint event related with this script
+irq_dic = {}; # key is cpu and value is a list which stacks irqs
+              # which raise NET_RX softirq
+net_rx_dic = {}; # key is cpu and value include time of NET_RX softirq-entry
+                 # and a list which stacks receive
+receive_hunk_list = []; # a list which include a sequence of receive events
+rx_skb_list = []; # received packet list for matching
+                       # skb_copy_datagram_iovec
+
+buffer_budget = 65536; # the budget of rx_skb_list, tx_queue_list and
+                       # tx_xmit_list
+of_count_rx_skb_list = 0; # overflow count
+
+tx_queue_list = []; # list of packets which pass through dev_queue_xmit
+of_count_tx_queue_list = 0; # overflow count
+
+tx_xmit_list = [];  # list of packets which pass through dev_hard_start_xmit
+of_count_tx_xmit_list = 0; # overflow count
+
+tx_free_list = [];  # list of packets which is freed
+
+# options
+show_tx = 0;
+show_rx = 0;
+dev = 0; # store a name of device specified by option "dev="
+debug = 0;
+
+# indices of event_info tuple
+EINFO_IDX_NAME=   0
+EINFO_IDX_CONTEXT=1
+EINFO_IDX_CPU=    2
+EINFO_IDX_TIME=   3
+EINFO_IDX_PID=    4
+EINFO_IDX_COMM=   5
+
+# Calculate a time interval(msec) from src(nsec) to dst(nsec)
+def diff_msec(src, dst):
+        return (dst - src) / 1000000.0
+
+# Display a process of transmitting a packet
+def print_transmit(hunk):
+        if dev != 0 and hunk['dev'].find(dev) < 0:
+                return
+        print "%7s %5d %6d.%06dsec %12.3fmsec      %12.3fmsec" % \
+                (hunk['dev'], hunk['len'],
+                nsecs_secs(hunk['queue_t']),
+                nsecs_nsecs(hunk['queue_t'])/1000,
+                diff_msec(hunk['queue_t'], hunk['xmit_t']),
+                diff_msec(hunk['xmit_t'], hunk['free_t']))
+
+# Format for displaying rx packet processing
+PF_IRQ_ENTRY= "  irq_entry(+%.3fmsec irq=%d:%s)"
+PF_SOFT_ENTRY="  softirq_entry(+%.3fmsec)"
+PF_NAPI_POLL= "  napi_poll_exit(+%.3fmsec %s)"
+PF_JOINT=     "         |"
+PF_WJOINT=    "         |            |"
+PF_NET_RECV=  "         |---netif_receive_skb(+%.3fmsec skb=%x len=%d)"
+PF_NET_RX=    "         |---netif_rx(+%.3fmsec skb=%x)"
+PF_CPY_DGRAM= "         |      skb_copy_datagram_iovec(+%.3fmsec %d:%s)"
+PF_KFREE_SKB= "         |      kfree_skb(+%.3fmsec location=%x)"
+PF_CONS_SKB=  "         |      consume_skb(+%.3fmsec)"
+
+# Display a process of received packets and interrputs associated with
+# a NET_RX softirq
+def print_receive(hunk):
+        show_hunk = 0
+        irq_list = hunk['irq_list']
+        cpu = irq_list[0]['cpu']
+        base_t = irq_list[0]['irq_ent_t']
+        # check if this hunk should be showed
+        if dev != 0:
+                for i in range(len(irq_list)):
+                        if irq_list[i]['name'].find(dev) >= 0:
+                                show_hunk = 1
+                                break
+        else:
+                show_hunk = 1
+        if show_hunk == 0:
+                return
+
+        print "%d.%06dsec cpu=%d" % \
+                (nsecs_secs(base_t), nsecs_nsecs(base_t)/1000, cpu)
+        for i in range(len(irq_list)):
+                print PF_IRQ_ENTRY % \
+                        (diff_msec(base_t, irq_list[i]['irq_ent_t']),
+                        irq_list[i]['irq'], irq_list[i]['name'])
+                print PF_JOINT
+                irq_event_list = irq_list[i]['event_list']
+                for j in range(len(irq_event_list)):
+                        irq_event = irq_event_list[j]
+                        if irq_event['event'] == 'netif_rx':
+                                print PF_NET_RX % \
+                                        (diff_msec(base_t, irq_event['time']),
+                                        irq_event['skbaddr'])
+                                print PF_JOINT
+        print PF_SOFT_ENTRY % \
+                diff_msec(base_t, hunk['sirq_ent_t'])
+        print PF_JOINT
+        event_list = hunk['event_list']
+        for i in range(len(event_list)):
+                event = event_list[i]
+                if event['event_name'] == 'napi_poll':
+                        print PF_NAPI_POLL % \
+                            (diff_msec(base_t, event['event_t']), event['dev'])
+                        if i == len(event_list) - 1:
+                                print ""
+                        else:
+                                print PF_JOINT
+                else:
+                        print PF_NET_RECV % \
+                            (diff_msec(base_t, event['event_t']), event['skbaddr'],
+                                event['len'])
+                        if 'comm' in event.keys():
+                                print PF_WJOINT
+                                print PF_CPY_DGRAM % \
+                                        (diff_msec(base_t, event['comm_t']),
+                                        event['pid'], event['comm'])
+                        elif 'handle' in event.keys():
+                                print PF_WJOINT
+                                if event['handle'] == "kfree_skb":
+                                        print PF_KFREE_SKB % \
+                                                (diff_msec(base_t,
+                                                event['comm_t']),
+                                                event['location'])
+                                elif event['handle'] == "consume_skb":
+                                        print PF_CONS_SKB % \
+                                                diff_msec(base_t,
+                                                        event['comm_t'])
+                        print PF_JOINT
+
+def trace_begin():
+        global show_tx
+        global show_rx
+        global dev
+        global debug
+
+        for i in range(len(sys.argv)):
+                if i == 0:
+                        continue
+                arg = sys.argv[i]
+                if arg == 'tx':
+                        show_tx = 1
+                elif arg =='rx':
+                        show_rx = 1
+                elif arg.find('dev=',0, 4) >= 0:
+                        dev = arg[4:]
+                elif arg == 'debug':
+                        debug = 1
+        if show_tx == 0  and show_rx == 0:
+                show_tx = 1
+                show_rx = 1
+
+def trace_end():
+        # order all events in time
+        all_event_list.sort(lambda a,b :cmp(a[EINFO_IDX_TIME],
+                                            b[EINFO_IDX_TIME]))
+        # process all events
+        for i in range(len(all_event_list)):
+                event_info = all_event_list[i]
+                name = event_info[EINFO_IDX_NAME]
+                if name == 'irq__softirq_exit':
+                        handle_irq_softirq_exit(event_info)
+                elif name == 'irq__softirq_entry':
+                        handle_irq_softirq_entry(event_info)
+                elif name == 'irq__softirq_raise':
+                        handle_irq_softirq_raise(event_info)
+                elif name == 'irq__irq_handler_entry':
+                        handle_irq_handler_entry(event_info)
+                elif name == 'irq__irq_handler_exit':
+                        handle_irq_handler_exit(event_info)
+                elif name == 'napi__napi_poll':
+                        handle_napi_poll(event_info)
+                elif name == 'net__netif_receive_skb':
+                        handle_netif_receive_skb(event_info)
+                elif name == 'net__netif_rx':
+                        handle_netif_rx(event_info)
+                elif name == 'skb__skb_copy_datagram_iovec':
+                        handle_skb_copy_datagram_iovec(event_info)
+                elif name == 'net__net_dev_queue':
+                        handle_net_dev_queue(event_info)
+                elif name == 'net__net_dev_xmit':
+                        handle_net_dev_xmit(event_info)
+                elif name == 'skb__kfree_skb':
+                        handle_kfree_skb(event_info)
+                elif name == 'skb__consume_skb':
+                        handle_consume_skb(event_info)
+        # display receive hunks
+        if show_rx:
+                for i in range(len(receive_hunk_list)):
+                        print_receive(receive_hunk_list[i])
+        # display transmit hunks
+        if show_tx:
+                print "   dev    len      Qdisc        " \
+                        "       netdevice             free"
+                for i in range(len(tx_free_list)):
+                        print_transmit(tx_free_list[i])
+        if debug:
+                print "debug buffer status"
+                print "----------------------------"
+                print "xmit Qdisc:remain:%d overflow:%d" % \
+                        (len(tx_queue_list), of_count_tx_queue_list)
+                print "xmit netdevice:remain:%d overflow:%d" % \
+                        (len(tx_xmit_list), of_count_tx_xmit_list)
+                print "receive:remain:%d overflow:%d" % \
+                        (len(rx_skb_list), of_count_rx_skb_list)
+
+# called from perf, when it finds a correspoinding event
+def irq__softirq_entry(name, context, cpu, sec, nsec, pid, comm, vec):
+        if symbol_str("irq__softirq_entry", "vec", vec) != "NET_RX":
+                return
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm, vec)
+        all_event_list.append(event_info)
+
+def irq__softirq_exit(name, context, cpu, sec, nsec, pid, comm, vec):
+        if symbol_str("irq__softirq_entry", "vec", vec) != "NET_RX":
+                return
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm, vec)
+        all_event_list.append(event_info)
+
+def irq__softirq_raise(name, context, cpu, sec, nsec, pid, comm, vec):
+        if symbol_str("irq__softirq_entry", "vec", vec) != "NET_RX":
+                return
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm, vec)
+        all_event_list.append(event_info)
+
+def irq__irq_handler_entry(name, context, cpu, sec, nsec, pid, comm,
+                        irq, irq_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        irq, irq_name)
+        all_event_list.append(event_info)
+
+def irq__irq_handler_exit(name, context, cpu, sec, nsec, pid, comm, irq, ret):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm, irq, ret)
+        all_event_list.append(event_info)
+
+def napi__napi_poll(name, context, cpu, sec, nsec, pid, comm, napi, dev_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        napi, dev_name)
+        all_event_list.append(event_info)
+
+def net__netif_receive_skb(name, context, cpu, sec, nsec, pid, comm, skbaddr,
+                        skblen, dev_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, skblen, dev_name)
+        all_event_list.append(event_info)
+
+def net__netif_rx(name, context, cpu, sec, nsec, pid, comm, skbaddr,
+                        skblen, dev_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, skblen, dev_name)
+        all_event_list.append(event_info)
+
+def net__net_dev_queue(name, context, cpu, sec, nsec, pid, comm,
+                        skbaddr, skblen, dev_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, skblen, dev_name)
+        all_event_list.append(event_info)
+
+def net__net_dev_xmit(name, context, cpu, sec, nsec, pid, comm,
+                        skbaddr, skblen, rc, dev_name):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, skblen, rc ,dev_name)
+        all_event_list.append(event_info)
+
+def skb__kfree_skb(name, context, cpu, sec, nsec, pid, comm,
+                        skbaddr, protocol, location):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, protocol, location)
+        all_event_list.append(event_info)
+
+def skb__consume_skb(name, context, cpu, sec, nsec, pid, comm, skbaddr):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr)
+        all_event_list.append(event_info)
+
+def skb__skb_copy_datagram_iovec(name, context, cpu, sec, nsec, pid, comm,
+        skbaddr, skblen):
+        event_info = (name, context, cpu, nsecs(sec, nsec), pid, comm,
+                        skbaddr, skblen)
+        all_event_list.append(event_info)
+
+def handle_irq_handler_entry(event_info):
+        (name, context, cpu, time, pid, comm, irq, irq_name) = event_info
+        if cpu not in irq_dic.keys():
+                irq_dic[cpu] = []
+        irq_record = {'irq':irq, 'name':irq_name, 'cpu':cpu, 'irq_ent_t':time}
+        irq_dic[cpu].append(irq_record)
+
+def handle_irq_handler_exit(event_info):
+        (name, context, cpu, time, pid, comm, irq, ret) = event_info
+        if cpu not in irq_dic.keys():
+                return
+        irq_record = irq_dic[cpu].pop()
+        if irq != irq_record['irq']:
+                return
+        irq_record.update({'irq_ext_t':time})
+        # if an irq doesn't include NET_RX softirq, drop.
+        if 'event_list' in irq_record.keys():
+                irq_dic[cpu].append(irq_record)
+
+def handle_irq_softirq_raise(event_info):
+        (name, context, cpu, time, pid, comm, vec) = event_info
+        if cpu not in irq_dic.keys() \
+        or len(irq_dic[cpu]) == 0:
+                return
+        irq_record = irq_dic[cpu].pop()
+        if 'event_list' in irq_record.keys():
+                irq_event_list = irq_record['event_list']
+        else:
+                irq_event_list = []
+        irq_event_list.append({'time':time, 'event':'sirq_raise'})
+        irq_record.update({'event_list':irq_event_list})
+        irq_dic[cpu].append(irq_record)
+
+def handle_irq_softirq_entry(event_info):
+        (name, context, cpu, time, pid, comm, vec) = event_info
+        net_rx_dic[cpu] = {'sirq_ent_t':time, 'event_list':[]}
+
+def handle_irq_softirq_exit(event_info):
+        (name, context, cpu, time, pid, comm, vec) = event_info
+        irq_list = []
+        event_list = 0
+        if cpu in irq_dic.keys():
+                irq_list = irq_dic[cpu]
+                del irq_dic[cpu]
+        if cpu in net_rx_dic.keys():
+                sirq_ent_t = net_rx_dic[cpu]['sirq_ent_t']
+                event_list = net_rx_dic[cpu]['event_list']
+                del net_rx_dic[cpu]
+        if irq_list == [] or event_list == 0:
+                return
+        rec_data = {'sirq_ent_t':sirq_ent_t, 'sirq_ext_t':time,
+                    'irq_list':irq_list, 'event_list':event_list}
+        # merge information realted to a NET_RX softirq
+        receive_hunk_list.append(rec_data)
+
+def handle_napi_poll(event_info):
+        (name, context, cpu, time, pid, comm, napi, dev_name) = event_info
+        if cpu in net_rx_dic.keys():
+                event_list = net_rx_dic[cpu]['event_list']
+                rec_data = {'event_name':'napi_poll',
+                                'dev':dev_name, 'event_t':time}
+                event_list.append(rec_data)
+
+def handle_netif_rx(event_info):
+        (name, context, cpu, time, pid, comm,
+                skbaddr, skblen, dev_name) = event_info
+        if cpu not in irq_dic.keys() \
+        or len(irq_dic[cpu]) == 0:
+                return
+        irq_record = irq_dic[cpu].pop()
+        if 'event_list' in irq_record.keys():
+                irq_event_list = irq_record['event_list']
+        else:
+                irq_event_list = []
+        irq_event_list.append({'time':time, 'event':'netif_rx',
+                'skbaddr':skbaddr, 'skblen':skblen, 'dev_name':dev_name})
+        irq_record.update({'event_list':irq_event_list})
+        irq_dic[cpu].append(irq_record)
+
+def handle_netif_receive_skb(event_info):
+        global of_count_rx_skb_list
+
+        (name, context, cpu, time, pid, comm,
+                skbaddr, skblen, dev_name) = event_info
+        if cpu in net_rx_dic.keys():
+                rec_data = {'event_name':'netif_receive_skb',
+                            'event_t':time, 'skbaddr':skbaddr, 'len':skblen}
+                event_list = net_rx_dic[cpu]['event_list']
+                event_list.append(rec_data)
+                rx_skb_list.insert(0, rec_data)
+                if len(rx_skb_list) > buffer_budget:
+                        rx_skb_list.pop()
+                        of_count_rx_skb_list += 1
+
+def handle_net_dev_queue(event_info):
+        global of_count_tx_queue_list
+
+        (name, context, cpu, time, pid, comm,
+                skbaddr, skblen, dev_name) = event_info
+        skb = {'dev':dev_name, 'skbaddr':skbaddr, 'len':skblen, 'queue_t':time}
+        tx_queue_list.insert(0, skb)
+        if len(tx_queue_list) > buffer_budget:
+                tx_queue_list.pop()
+                of_count_tx_queue_list += 1
+
+def handle_net_dev_xmit(event_info):
+        global of_count_tx_xmit_list
+
+        (name, context, cpu, time, pid, comm,
+                skbaddr, skblen, rc, dev_name) = event_info
+        if rc == 0: # NETDEV_TX_OK
+                for i in range(len(tx_queue_list)):
+                        skb = tx_queue_list[i]
+                        if skb['skbaddr'] == skbaddr:
+                                skb['xmit_t'] = time
+                                tx_xmit_list.insert(0, skb)
+                                del tx_queue_list[i]
+                                if len(tx_xmit_list) > buffer_budget:
+                                        tx_xmit_list.pop()
+                                        of_count_tx_xmit_list += 1
+                                return
+
+def handle_kfree_skb(event_info):
+        (name, context, cpu, time, pid, comm,
+                skbaddr, protocol, location) = event_info
+        for i in range(len(tx_queue_list)):
+                skb = tx_queue_list[i]
+                if skb['skbaddr'] == skbaddr:
+                        del tx_queue_list[i]
+                        return
+        for i in range(len(tx_xmit_list)):
+                skb = tx_xmit_list[i]
+                if skb['skbaddr'] == skbaddr:
+                        skb['free_t'] = time
+                        tx_free_list.append(skb)
+                        del tx_xmit_list[i]
+                        return
+        for i in range(len(rx_skb_list)):
+                rec_data = rx_skb_list[i]
+                if rec_data['skbaddr'] == skbaddr:
+                        rec_data.update({'handle':"kfree_skb",
+                                        'comm':comm, 'pid':pid, 'comm_t':time})
+                        del rx_skb_list[i]
+                        return
+
+def handle_consume_skb(event_info):
+        (name, context, cpu, time, pid, comm, skbaddr) = event_info
+        for i in range(len(tx_xmit_list)):
+                skb = tx_xmit_list[i]
+                if skb['skbaddr'] == skbaddr:
+                        skb['free_t'] = time
+                        tx_free_list.append(skb)
+                        del tx_xmit_list[i]
+                        return
+
+def handle_skb_copy_datagram_iovec(event_info):
+        (name, context, cpu, time, pid, comm, skbaddr, skblen) = event_info
+        for i in range(len(rx_skb_list)):
+                rec_data = rx_skb_list[i]
+                if skbaddr == rec_data['skbaddr']:
+                        rec_data.update({'handle':"skb_copy_datagram_iovec",
+                                        'comm':comm, 'pid':pid, 'comm_t':time})
+                        del rx_skb_list[i]
+                        return
diff --git a/tools/perf/scripts/python/sctop.py b/tools/perf/scripts/python/sctop.py
index 6cafad40c296..7a6ec2c7d8ab 100644
--- a/tools/perf/scripts/python/sctop.py
+++ b/tools/perf/scripts/python/sctop.py
@@ -8,10 +8,7 @@
 # will be refreshed every [interval] seconds.  The default interval is
 # 3 seconds.
 
-import thread
-import time
-import os
-import sys
+import os, sys, thread, time
 
 sys.path.append(os.environ['PERF_EXEC_PATH'] + \
         '/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
@@ -20,7 +17,7 @@ from perf_trace_context import *
 from Core import *
 from Util import *
 
-usage = "perf trace -s syscall-counts.py [comm] [interval]\n";
+usage = "perf trace -s sctop.py [comm] [interval]\n";
 
 for_comm = None
 default_interval = 3
@@ -71,7 +68,7 @@ def print_syscall_totals(interval):
                 for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \
                                               reverse = True):
                         try:
-                                print "%-40d  %10d\n" % (id, val),
+                                print "%-40s  %10d\n" % (syscall_name(id), val),
                         except TypeError:
                                 pass
                 syscalls.clear()
diff --git a/tools/perf/scripts/python/syscall-counts-by-pid.py b/tools/perf/scripts/python/syscall-counts-by-pid.py
index af722d6a4b3f..d1ee3ec10cf2 100644
--- a/tools/perf/scripts/python/syscall-counts-by-pid.py
+++ b/tools/perf/scripts/python/syscall-counts-by-pid.py
@@ -5,29 +5,33 @@
 # Displays system-wide system call totals, broken down by syscall.
 # If a [comm] arg is specified, only syscalls called by [comm] are displayed.
 
-import os
-import sys
+import os, sys
 
 sys.path.append(os.environ['PERF_EXEC_PATH'] + \
         '/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
 
 from perf_trace_context import *
 from Core import *
+from Util import syscall_name
 
 usage = "perf trace -s syscall-counts-by-pid.py [comm]\n";
 
 for_comm = None
+for_pid = None
 
 if len(sys.argv) > 2:
         sys.exit(usage)
 
 if len(sys.argv) > 1:
-        for_comm = sys.argv[1]
+        try:
+                for_pid = int(sys.argv[1])
+        except:
+                for_comm = sys.argv[1]
 
 syscalls = autodict()
 
 def trace_begin():
-        pass
+        print "Press control+C to stop and show the summary"
 
 def trace_end():
         print_syscall_totals()
@@ -35,9 +39,10 @@ def trace_end():
 def raw_syscalls__sys_enter(event_name, context, common_cpu,
         common_secs, common_nsecs, common_pid, common_comm,
         id, args):
-        if for_comm is not None:
-                if common_comm != for_comm:
-                        return
+
+        if (for_comm and common_comm != for_comm) or \
+           (for_pid  and common_pid  != for_pid ):
+                return
         try:
                 syscalls[common_comm][common_pid][id] += 1
         except TypeError:
@@ -61,4 +66,4 @@ def print_syscall_totals():
                     id_keys = syscalls[comm][pid].keys()
                     for id, val in sorted(syscalls[comm][pid].iteritems(), \
                                   key = lambda(k, v): (v, k),  reverse = True):
-                            print "  %-38d  %10d\n" % (id, val),
+                            print "  %-38s  %10d\n" % (syscall_name(id), val),
diff --git a/tools/perf/scripts/python/syscall-counts.py b/tools/perf/scripts/python/syscall-counts.py
index f977e85ff049..ea183dc82d29 100644
--- a/tools/perf/scripts/python/syscall-counts.py
+++ b/tools/perf/scripts/python/syscall-counts.py
@@ -13,6 +13,7 @@ sys.path.append(os.environ['PERF_EXEC_PATH'] + \
 
 from perf_trace_context import *
 from Core import *
+from Util import syscall_name
 
 usage = "perf trace -s syscall-counts.py [comm]\n";
 
@@ -27,7 +28,7 @@ if len(sys.argv) > 1:
 syscalls = autodict()
 
 def trace_begin():
-        pass
+        print "Press control+C to stop and show the summary"
 
 def trace_end():
         print_syscall_totals()
@@ -55,4 +56,4 @@ def print_syscall_totals():
 
     for id, val in sorted(syscalls.iteritems(), key = lambda(k, v): (v, k), \
                                   reverse = True):
-            print "%-40d  %10d\n" % (id, val),
+            print "%-40s  %10d\n" % (syscall_name(id), val),
diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h
index 27e9ebe4076e..a7729797fd96 100644
--- a/tools/perf/util/cache.h
+++ b/tools/perf/util/cache.h
@@ -82,6 +82,8 @@ extern char *perf_path(const char *fmt, ...) __attribute__((format (printf, 1, 2
 extern char *perf_pathdup(const char *fmt, ...)
         __attribute__((format (printf, 1, 2)));
 
+#ifdef NO_STRLCPY
 extern size_t strlcpy(char *dest, const char *src, size_t size);
+#endif
 
 #endif /* __PERF_CACHE_H */
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index f231f43424d2..e12d539417b2 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -28,6 +28,9 @@ bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event)
 #define chain_for_each_child(child, parent)     \
         list_for_each_entry(child, &parent->children, brothers)
 
+#define chain_for_each_child_safe(child, next, parent)  \
+        list_for_each_entry_safe(child, next, &parent->children, brothers)
+
 static void
 rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
                     enum chain_mode mode)
@@ -86,10 +89,10 @@ __sort_chain_flat(struct rb_root *rb_root, struct callchain_node *node,
  * sort them by hit
  */
 static void
-sort_chain_flat(struct rb_root *rb_root, struct callchain_node *node,
+sort_chain_flat(struct rb_root *rb_root, struct callchain_root *root,
                 u64 min_hit, struct callchain_param *param __used)
 {
-        __sort_chain_flat(rb_root, node, min_hit);
+        __sort_chain_flat(rb_root, &root->node, min_hit);
 }
 
 static void __sort_chain_graph_abs(struct callchain_node *node,
@@ -108,11 +111,11 @@ static void __sort_chain_graph_abs(struct callchain_node *node,
 }
 
 static void
-sort_chain_graph_abs(struct rb_root *rb_root, struct callchain_node *chain_root,
+sort_chain_graph_abs(struct rb_root *rb_root, struct callchain_root *chain_root,
                      u64 min_hit, struct callchain_param *param __used)
 {
-        __sort_chain_graph_abs(chain_root, min_hit);
-        rb_root->rb_node = chain_root->rb_root.rb_node;
+        __sort_chain_graph_abs(&chain_root->node, min_hit);
+        rb_root->rb_node = chain_root->node.rb_root.rb_node;
 }
 
 static void __sort_chain_graph_rel(struct callchain_node *node,
@@ -133,11 +136,11 @@ static void __sort_chain_graph_rel(struct callchain_node *node,
 }
 
 static void
-sort_chain_graph_rel(struct rb_root *rb_root, struct callchain_node *chain_root,
+sort_chain_graph_rel(struct rb_root *rb_root, struct callchain_root *chain_root,
                      u64 min_hit __used, struct callchain_param *param)
 {
-        __sort_chain_graph_rel(chain_root, param->min_percent / 100.0);
-        rb_root->rb_node = chain_root->rb_root.rb_node;
+        __sort_chain_graph_rel(&chain_root->node, param->min_percent / 100.0);
+        rb_root->rb_node = chain_root->node.rb_root.rb_node;
 }
 
 int register_callchain_param(struct callchain_param *param)
@@ -284,19 +287,18 @@ split_add_child(struct callchain_node *parent, struct resolved_chain *chain,
 }
 
 static int
-__append_chain(struct callchain_node *root, struct resolved_chain *chain,
-               unsigned int start, u64 period);
+append_chain(struct callchain_node *root, struct resolved_chain *chain,
+             unsigned int start, u64 period);
 
 static void
-__append_chain_children(struct callchain_node *root,
-                        struct resolved_chain *chain,
-                        unsigned int start, u64 period)
+append_chain_children(struct callchain_node *root, struct resolved_chain *chain,
+                      unsigned int start, u64 period)
 {
         struct callchain_node *rnode;
 
         /* lookup in childrens */
         chain_for_each_child(rnode, root) {
-                unsigned int ret = __append_chain(rnode, chain, start, period);
+                unsigned int ret = append_chain(rnode, chain, start, period);
 
                 if (!ret)
                         goto inc_children_hit;
@@ -309,8 +311,8 @@ inc_children_hit:
 }
 
 static int
-__append_chain(struct callchain_node *root, struct resolved_chain *chain,
-               unsigned int start, u64 period)
+append_chain(struct callchain_node *root, struct resolved_chain *chain,
+             unsigned int start, u64 period)
 {
         struct callchain_list *cnode;
         unsigned int i = start;
@@ -357,7 +359,7 @@ __append_chain(struct callchain_node *root, struct resolved_chain *chain,
         }
 
         /* We match the node and still have a part remaining */
-        __append_chain_children(root, chain, i, period);
+        append_chain_children(root, chain, i, period);
 
         return 0;
 }
@@ -380,8 +382,8 @@ static void filter_context(struct ip_callchain *old, struct resolved_chain *new,
 }
 
 
-int append_chain(struct callchain_node *root, struct ip_callchain *chain,
-                 struct map_symbol *syms, u64 period)
+int callchain_append(struct callchain_root *root, struct ip_callchain *chain,
+                     struct map_symbol *syms, u64 period)
 {
         struct resolved_chain *filtered;
 
@@ -398,9 +400,65 @@ int append_chain(struct callchain_node *root, struct ip_callchain *chain,
         if (!filtered->nr)
                 goto end;
 
-        __append_chain_children(root, filtered, 0, period);
+        append_chain_children(&root->node, filtered, 0, period);
+
+        if (filtered->nr > root->max_depth)
+                root->max_depth = filtered->nr;
 end:
         free(filtered);
 
         return 0;
 }
+
+static int
+merge_chain_branch(struct callchain_node *dst, struct callchain_node *src,
+                   struct resolved_chain *chain)
+{
+        struct callchain_node *child, *next_child;
+        struct callchain_list *list, *next_list;
+        int old_pos = chain->nr;
+        int err = 0;
+
+        list_for_each_entry_safe(list, next_list, &src->val, list) {
+                chain->ips[chain->nr].ip = list->ip;
+                chain->ips[chain->nr].ms = list->ms;
+                chain->nr++;
+                list_del(&list->list);
+                free(list);
+        }
+
+        if (src->hit)
+                append_chain_children(dst, chain, 0, src->hit);
+
+        chain_for_each_child_safe(child, next_child, src) {
+                err = merge_chain_branch(dst, child, chain);
+                if (err)
+                        break;
+
+                list_del(&child->brothers);
+                free(child);
+        }
+
+        chain->nr = old_pos;
+
+        return err;
+}
+
+int callchain_merge(struct callchain_root *dst, struct callchain_root *src)
+{
+        struct resolved_chain *chain;
+        int err;
+
+        chain = malloc(sizeof(*chain) +
+                       src->max_depth * sizeof(struct resolved_ip));
+        if (!chain)
+                return -ENOMEM;
+
+        chain->nr = 0;
+
+        err = merge_chain_branch(&dst->node, &src->node, chain);
+
+        free(chain);
+
+        return err;
+}
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
index 624a96c636fd..c15fb8c24ad2 100644
--- a/tools/perf/util/callchain.h
+++ b/tools/perf/util/callchain.h
@@ -26,9 +26,14 @@ struct callchain_node {
         u64                     children_hit;
 };
 
+struct callchain_root {
+        u64                     max_depth;
+        struct callchain_node   node;
+};
+
 struct callchain_param;
 
-typedef void (*sort_chain_func_t)(struct rb_root *, struct callchain_node *,
+typedef void (*sort_chain_func_t)(struct rb_root *, struct callchain_root *,
                                  u64, struct callchain_param *);
 
 struct callchain_param {
@@ -44,14 +49,16 @@ struct callchain_list {
         struct list_head        list;
 };
 
-static inline void callchain_init(struct callchain_node *node)
+static inline void callchain_init(struct callchain_root *root)
 {
-        INIT_LIST_HEAD(&node->brothers);
-        INIT_LIST_HEAD(&node->children);
-        INIT_LIST_HEAD(&node->val);
+        INIT_LIST_HEAD(&root->node.brothers);
+        INIT_LIST_HEAD(&root->node.children);
+        INIT_LIST_HEAD(&root->node.val);
 
-        node->parent = NULL;
-        node->hit = 0;
+        root->node.parent = NULL;
+        root->node.hit = 0;
+        root->node.children_hit = 0;
+        root->max_depth = 0;
 }
 
 static inline u64 cumul_hits(struct callchain_node *node)
@@ -60,8 +67,9 @@ static inline u64 cumul_hits(struct callchain_node *node)
 }
 
 int register_callchain_param(struct callchain_param *param);
-int append_chain(struct callchain_node *root, struct ip_callchain *chain,
-                 struct map_symbol *syms, u64 period);
+int callchain_append(struct callchain_root *root, struct ip_callchain *chain,
+                     struct map_symbol *syms, u64 period);
+int callchain_merge(struct callchain_root *dst, struct callchain_root *src);
 
 bool ip_callchain__valid(struct ip_callchain *chain, const event_t *event);
 #endif  /* __PERF_CALLCHAIN_H */
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index 318dab15d177..c8d81b00089d 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -12,8 +12,8 @@
 #include "debug.h"
 #include "util.h"
 
-int verbose = 0;
-bool dump_trace = false;
+int verbose;
+bool dump_trace = false, quiet = false;
 
 int eprintf(int level, const char *fmt, ...)
 {
@@ -23,7 +23,7 @@ int eprintf(int level, const char *fmt, ...)
         if (verbose >= level) {
                 va_start(args, fmt);
                 if (use_browser > 0)
-                        ret = browser__show_help(fmt, args);
+                        ret = ui_helpline__show_help(fmt, args);
                 else
                         ret = vfprintf(stderr, fmt, args);
                 va_end(args);
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index 047ac3324ebe..7b514082bbaf 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -6,7 +6,7 @@
 #include "event.h"
 
 extern int verbose;
-extern bool dump_trace;
+extern bool quiet, dump_trace;
 
 int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
 void trace_event(event_t *event);
@@ -14,7 +14,7 @@ void trace_event(event_t *event);
 struct ui_progress;
 
 #ifdef NO_NEWT_SUPPORT
-static inline int browser__show_help(const char *format __used, va_list ap __used)
+static inline int ui_helpline__show_help(const char *format __used, va_list ap __used)
 {
         return 0;
 }
@@ -30,10 +30,9 @@ static inline void ui_progress__update(struct ui_progress *self __used,
 
 static inline void ui_progress__delete(struct ui_progress *self __used) {}
 #else
-int browser__show_help(const char *format, va_list ap);
-struct ui_progress *ui_progress__new(const char *title, u64 total);
-void ui_progress__update(struct ui_progress *self, u64 curr);
-void ui_progress__delete(struct ui_progress *self);
+extern char ui_helpline__last_msg[];
+int ui_helpline__show_help(const char *format, va_list ap);
+#include "ui/progress.h"
 #endif
 
 #endif  /* __PERF_DEBUG_H */
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index d7e67b167ea3..7cba0551a565 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -265,15 +265,16 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
                           const char *name, bool is_kallsyms)
 {
         const size_t size = PATH_MAX;
-        char *filename = malloc(size),
+        char *realname = realpath(name, NULL),
+             *filename = malloc(size),
              *linkname = malloc(size), *targetname;
         int len, err = -1;
 
-        if (filename == NULL || linkname == NULL)
+        if (realname == NULL || filename == NULL || linkname == NULL)
                 goto out_free;
 
         len = snprintf(filename, size, "%s%s%s",
-                       debugdir, is_kallsyms ? "/" : "", name);
+                       debugdir, is_kallsyms ? "/" : "", realname);
         if (mkdir_p(filename, 0755))
                 goto out_free;
 
@@ -283,7 +284,7 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
                 if (is_kallsyms) {
                          if (copyfile("/proc/kallsyms", filename))
                                 goto out_free;
-                } else if (link(name, filename) && copyfile(name, filename))
+                } else if (link(realname, filename) && copyfile(name, filename))
                         goto out_free;
         }
 
@@ -300,6 +301,7 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir,
         if (symlink(targetname, linkname) == 0)
                 err = 0;
 out_free:
+        free(realname);
         free(filename);
         free(linkname);
         return err;
@@ -946,11 +948,16 @@ perf_header__find_attr(u64 id, struct perf_header *header)
 
         /*
          * We set id to -1 if the data file doesn't contain sample
-         * ids. Check for this and avoid walking through the entire
-         * list of ids which may be large.
+         * ids. This can happen when the data file contains one type
+         * of event and in that case, the header can still store the
+         * event attribute information. Check for this and avoid
+         * walking through the entire list of ids which may be large.
          */
-        if (id == -1ULL)
+        if (id == -1ULL) {
+                if (header->attrs > 0)
+                        return &header->attr[0]->attr;
                 return NULL;
+        }
 
         for (i = 0; i < header->attrs; i++) {
                 struct perf_header_attr *attr = header->attr[i];
diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c
index e7263d49bcf0..2022e8740994 100644
--- a/tools/perf/util/hist.c
+++ b/tools/perf/util/hist.c
@@ -87,7 +87,7 @@ static void hist_entry__add_cpumode_period(struct hist_entry *self,
 
 static struct hist_entry *hist_entry__new(struct hist_entry *template)
 {
-        size_t callchain_size = symbol_conf.use_callchain ? sizeof(struct callchain_node) : 0;
+        size_t callchain_size = symbol_conf.use_callchain ? sizeof(struct callchain_root) : 0;
         struct hist_entry *self = malloc(sizeof(*self) + callchain_size);
 
         if (self != NULL) {
@@ -226,6 +226,8 @@ static bool collapse__insert_entry(struct rb_root *root, struct hist_entry *he)
 
                 if (!cmp) {
                         iter->period += he->period;
+                        if (symbol_conf.use_callchain)
+                                callchain_merge(iter->callchain, he->callchain);
                         hist_entry__free(he);
                         return false;
                 }
@@ -876,6 +878,9 @@ unsigned int hists__sort_list_width(struct hists *self)
                 if (!se->elide)
                         ret += 2 + hists__col_len(self, se->se_width_idx);
 
+        if (verbose) /* Addr + origin */
+                ret += 3 + BITS_PER_LONG / 4;
+
         return ret;
 }
 
@@ -980,9 +985,9 @@ int hist_entry__inc_addr_samples(struct hist_entry *self, u64 ip)
         return 0;
 }
 
-static struct objdump_line *objdump_line__new(s64 offset, char *line)
+static struct objdump_line *objdump_line__new(s64 offset, char *line, size_t privsize)
 {
-        struct objdump_line *self = malloc(sizeof(*self));
+        struct objdump_line *self = malloc(sizeof(*self) + privsize);
 
         if (self != NULL) {
                 self->offset = offset;
@@ -1014,7 +1019,7 @@ struct objdump_line *objdump__get_next_ip_line(struct list_head *head,
 }
 
 static int hist_entry__parse_objdump_line(struct hist_entry *self, FILE *file,
-                                          struct list_head *head)
+                                          struct list_head *head, size_t privsize)
 {
         struct symbol *sym = self->ms.sym;
         struct objdump_line *objdump_line;
@@ -1065,7 +1070,7 @@ static int hist_entry__parse_objdump_line(struct hist_entry *self, FILE *file,
                         offset = -1;
         }
 
-        objdump_line = objdump_line__new(offset, line);
+        objdump_line = objdump_line__new(offset, line, privsize);
         if (objdump_line == NULL) {
                 free(line);
                 return -1;
@@ -1075,7 +1080,8 @@ static int hist_entry__parse_objdump_line(struct hist_entry *self, FILE *file,
         return 0;
 }
 
-int hist_entry__annotate(struct hist_entry *self, struct list_head *head)
+int hist_entry__annotate(struct hist_entry *self, struct list_head *head,
+                         size_t privsize)
 {
         struct symbol *sym = self->ms.sym;
         struct map *map = self->ms.map;
@@ -1140,7 +1146,7 @@ fallback:
                 goto out_free_filename;
 
         while (!feof(file))
-                if (hist_entry__parse_objdump_line(self, file, head) < 0)
+                if (hist_entry__parse_objdump_line(self, file, head, privsize) < 0)
                         break;
 
         pclose(file);
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index 65a48db46a29..587d375d3430 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -101,7 +101,8 @@ size_t hists__fprintf(struct hists *self, struct hists *pair,
                       bool show_displacement, FILE *fp);
 
 int hist_entry__inc_addr_samples(struct hist_entry *self, u64 ip);
-int hist_entry__annotate(struct hist_entry *self, struct list_head *head);
+int hist_entry__annotate(struct hist_entry *self, struct list_head *head,
+                         size_t privsize);
 
 void hists__filter_by_dso(struct hists *self, const struct dso *dso);
 void hists__filter_by_thread(struct hists *self, const struct thread *thread);
diff --git a/tools/perf/util/include/linux/list.h b/tools/perf/util/include/linux/list.h
index dbe4b814382a..f5ca26e53fbb 100644
--- a/tools/perf/util/include/linux/list.h
+++ b/tools/perf/util/include/linux/list.h
@@ -15,4 +15,12 @@ static inline void list_del_range(struct list_head *begin,
         begin->prev->next = end->next;
         end->next->prev = begin->prev;
 }
+
+/**
+ * list_for_each_from   -       iterate over a list from one of its nodes
+ * @pos:  the &struct list_head to use as a loop cursor, from where to start
+ * @head: the head for your list.
+ */
+#define list_for_each_from(pos, head) \
+        for (; prefetch(pos->next), pos != (head); pos = pos->next)
 #endif
diff --git a/tools/perf/util/include/linux/types.h b/tools/perf/util/include/linux/types.h
index 196862a81a21..12de3b8112f9 100644
--- a/tools/perf/util/include/linux/types.h
+++ b/tools/perf/util/include/linux/types.h
@@ -6,4 +6,16 @@
 #define DECLARE_BITMAP(name,bits) \
         unsigned long name[BITS_TO_LONGS(bits)]
 
+struct list_head {
+        struct list_head *next, *prev;
+};
+
+struct hlist_head {
+        struct hlist_node *first;
+};
+
+struct hlist_node {
+        struct hlist_node *next, **pprev;
+};
+
 #endif
diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h
index 78575796d5f3..b397c0383728 100644
--- a/tools/perf/util/map.h
+++ b/tools/perf/util/map.h
@@ -215,6 +215,16 @@ struct symbol *map_groups__find_function_by_name(struct map_groups *self,
         return map_groups__find_symbol_by_name(self, MAP__FUNCTION, name, mapp, filter);
 }
 
+static inline
+struct symbol *machine__find_kernel_function_by_name(struct machine *self,
+                                                     const char *name,
+                                                     struct map **mapp,
+                                                     symbol_filter_t filter)
+{
+        return map_groups__find_function_by_name(&self->kmaps, name, mapp,
+                                                 filter);
+}
+
 int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
                                    int verbose, FILE *fp);
 
diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c
deleted file mode 100644
index 91de99b58445..000000000000
--- a/tools/perf/util/newt.c
+++ /dev/null
@@ -1,1568 +0,0 @@
-#define _GNU_SOURCE
-#include <stdio.h>
-#undef _GNU_SOURCE
-/*
- * slang versions <= 2.0.6 have a "#if HAVE_LONG_LONG" that breaks
- * the build if it isn't defined. Use the equivalent one that glibc
- * has on features.h.
- */
-#include <features.h>
-#ifndef HAVE_LONG_LONG
-#define HAVE_LONG_LONG __GLIBC_HAVE_LONG_LONG
-#endif
-#include <slang.h>
-#include <signal.h>
-#include <stdlib.h>
-#include <newt.h>
-#include <sys/ttydefaults.h>
-
-#include "cache.h"
-#include "hist.h"
-#include "pstack.h"
-#include "session.h"
-#include "sort.h"
-#include "symbol.h"
-
-#if SLANG_VERSION < 20104
-#define slsmg_printf(msg, args...) SLsmg_printf((char *)msg, ##args)
-#define slsmg_write_nstring(msg, len) SLsmg_write_nstring((char *)msg, len)
-#define sltt_set_color(obj, name, fg, bg) SLtt_set_color(obj,(char *)name,\
-                                                         (char *)fg, (char *)bg)
-#else
-#define slsmg_printf SLsmg_printf
-#define slsmg_write_nstring SLsmg_write_nstring
-#define sltt_set_color SLtt_set_color
-#endif
-
-struct ui_progress {
-        newtComponent form, scale;
-};
-
-struct ui_progress *ui_progress__new(const char *title, u64 total)
-{
-        struct ui_progress *self = malloc(sizeof(*self));
-
-        if (self != NULL) {
-                int cols;
-
-                if (use_browser <= 0)   
-                        return self;
-                newtGetScreenSize(&cols, NULL);
-                cols -= 4;
-                newtCenteredWindow(cols, 1, title);
-                self->form  = newtForm(NULL, NULL, 0);
-                if (self->form == NULL)
-                        goto out_free_self;
-                self->scale = newtScale(0, 0, cols, total);
-                if (self->scale == NULL)
-                        goto out_free_form;
-                newtFormAddComponent(self->form, self->scale);
-                newtRefresh();
-        }
-
-        return self;
-
-out_free_form:
-        newtFormDestroy(self->form);
-out_free_self:
-        free(self);
-        return NULL;
-}
-
-void ui_progress__update(struct ui_progress *self, u64 curr)
-{
-        /*
-         * FIXME: We should have a per UI backend way of showing progress,
-         * stdio will just show a percentage as NN%, etc.
-         */
-        if (use_browser <= 0)
-                return;
-        newtScaleSet(self->scale, curr);
-        newtRefresh();
-}
-
-void ui_progress__delete(struct ui_progress *self)
-{
-        if (use_browser > 0) {
-                newtFormDestroy(self->form);
-                newtPopWindow();
-        }
-        free(self);
-}
-
-static void ui_helpline__pop(void)
-{
-        newtPopHelpLine();
-}
-
-static void ui_helpline__push(const char *msg)
-{
-        newtPushHelpLine(msg);
-}
-
-static void ui_helpline__vpush(const char *fmt, va_list ap)
-{
-        char *s;
-
-        if (vasprintf(&s, fmt, ap) < 0)
-                vfprintf(stderr, fmt, ap);
-        else {
-                ui_helpline__push(s);
-                free(s);
-        }
-}
-
-static void ui_helpline__fpush(const char *fmt, ...)
-{
-        va_list ap;
-
-        va_start(ap, fmt);
-        ui_helpline__vpush(fmt, ap);
-        va_end(ap);
-}
-
-static void ui_helpline__puts(const char *msg)
-{
-        ui_helpline__pop();
-        ui_helpline__push(msg);
-}
-
-static char browser__last_msg[1024];
-
-int browser__show_help(const char *format, va_list ap)
-{
-        int ret;
-        static int backlog;
-
-        ret = vsnprintf(browser__last_msg + backlog,
-                        sizeof(browser__last_msg) - backlog, format, ap);
-        backlog += ret;
-
-        if (browser__last_msg[backlog - 1] == '\n') {
-                ui_helpline__puts(browser__last_msg);
-                newtRefresh();
-                backlog = 0;
-        }
-
-        return ret;
-}
-
-static void newt_form__set_exit_keys(newtComponent self)
-{
-        newtFormAddHotKey(self, NEWT_KEY_LEFT);
-        newtFormAddHotKey(self, NEWT_KEY_ESCAPE);
-        newtFormAddHotKey(self, 'Q');
-        newtFormAddHotKey(self, 'q');
-        newtFormAddHotKey(self, CTRL('c'));
-}
-
-static newtComponent newt_form__new(void)
-{
-        newtComponent self = newtForm(NULL, NULL, 0);
-        if (self)
-                newt_form__set_exit_keys(self);
-        return self;
-}
-
-static int popup_menu(int argc, char * const argv[])
-{
-        struct newtExitStruct es;
-        int i, rc = -1, max_len = 5;
-        newtComponent listbox, form = newt_form__new();
-
-        if (form == NULL)
-                return -1;
-
-        listbox = newtListbox(0, 0, argc, NEWT_FLAG_RETURNEXIT);
-        if (listbox == NULL)
-                goto out_destroy_form;
-
-        newtFormAddComponent(form, listbox);
-
-        for (i = 0; i < argc; ++i) {
-                int len = strlen(argv[i]);
-                if (len > max_len)
-                        max_len = len;
-                if (newtListboxAddEntry(listbox, argv[i], (void *)(long)i))
-                        goto out_destroy_form;
-        }
-
-        newtCenteredWindow(max_len, argc, NULL);
-        newtFormRun(form, &es);
-        rc = newtListboxGetCurrent(listbox) - NULL;
-        if (es.reason == NEWT_EXIT_HOTKEY)
-                rc = -1;
-        newtPopWindow();
-out_destroy_form:
-        newtFormDestroy(form);
-        return rc;
-}
-
-static int ui__help_window(const char *text)
-{
-        struct newtExitStruct es;
-        newtComponent tb, form = newt_form__new();
-        int rc = -1;
-        int max_len = 0, nr_lines = 0;
-        const char *t;
-
-        if (form == NULL)
-                return -1;
-
-        t = text;
-        while (1) {
-                const char *sep = strchr(t, '\n');
-                int len;
-
-                if (sep == NULL)
-                        sep = strchr(t, '\0');
-                len = sep - t;
-                if (max_len < len)
-                        max_len = len;
-                ++nr_lines;
-                if (*sep == '\0')
-                        break;
-                t = sep + 1;
-        }
-
-        tb = newtTextbox(0, 0, max_len, nr_lines, 0);
-        if (tb == NULL)
-                goto out_destroy_form;
-
-        newtTextboxSetText(tb, text);
-        newtFormAddComponent(form, tb);
-        newtCenteredWindow(max_len, nr_lines, NULL);
-        newtFormRun(form, &es);
-        newtPopWindow();
-        rc = 0;
-out_destroy_form:
-        newtFormDestroy(form);
-        return rc;
-}
-
-static bool dialog_yesno(const char *msg)
-{
-        /* newtWinChoice should really be accepting const char pointers... */
-        char yes[] = "Yes", no[] = "No";
-        return newtWinChoice(NULL, yes, no, (char *)msg) == 1;
-}
-
-static void ui__error_window(const char *fmt, ...)
-{
-        va_list ap;
-
-        va_start(ap, fmt);
-        newtWinMessagev((char *)"Error", (char *)"Ok", (char *)fmt, ap);
-        va_end(ap);
-}
-
-#define HE_COLORSET_TOP         50
-#define HE_COLORSET_MEDIUM      51
-#define HE_COLORSET_NORMAL      52
-#define HE_COLORSET_SELECTED    53
-#define HE_COLORSET_CODE        54
-
-static int ui_browser__percent_color(double percent, bool current)
-{
-        if (current)
-                return HE_COLORSET_SELECTED;
-        if (percent >= MIN_RED)
-                return HE_COLORSET_TOP;
-        if (percent >= MIN_GREEN)
-                return HE_COLORSET_MEDIUM;
-        return HE_COLORSET_NORMAL;
-}
-
-struct ui_browser {
-        newtComponent   form, sb;
-        u64             index, first_visible_entry_idx;
-        void            *first_visible_entry, *entries;
-        u16             top, left, width, height;
-        void            *priv;
-        unsigned int    (*refresh_entries)(struct ui_browser *self);
-        void            (*seek)(struct ui_browser *self,
-                                off_t offset, int whence);
-        u32             nr_entries;
-};
-
-static void ui_browser__list_head_seek(struct ui_browser *self,
-                                       off_t offset, int whence)
-{
-        struct list_head *head = self->entries;
-        struct list_head *pos;
-
-        switch (whence) {
-        case SEEK_SET:
-                pos = head->next;
-                break;
-        case SEEK_CUR:
-                pos = self->first_visible_entry;
-                break;
-        case SEEK_END:
-                pos = head->prev;
-                break;
-        default:
-                return;
-        }
-
-        if (offset > 0) {
-                while (offset-- != 0)
-                        pos = pos->next;
-        } else {
-                while (offset++ != 0)
-                        pos = pos->prev;
-        }
-
-        self->first_visible_entry = pos;
-}
-
-static bool ui_browser__is_current_entry(struct ui_browser *self, unsigned row)
-{
-        return (self->first_visible_entry_idx + row) == self->index;
-}
-
-static void ui_browser__refresh_dimensions(struct ui_browser *self)
-{
-        int cols, rows;
-        newtGetScreenSize(&cols, &rows);
-
-        if (self->width > cols - 4)
-                self->width = cols - 4;
-        self->height = rows - 5;
-        if (self->height > self->nr_entries)
-                self->height = self->nr_entries;
-        self->top  = (rows - self->height) / 2;
-        self->left = (cols - self->width) / 2;
-}
-
-static void ui_browser__reset_index(struct ui_browser *self)
-{
-        self->index = self->first_visible_entry_idx = 0;
-        self->seek(self, 0, SEEK_SET);
-}
-
-static int ui_browser__show(struct ui_browser *self, const char *title)
-{
-        if (self->form != NULL) {
-                newtFormDestroy(self->form);
-                newtPopWindow();
-        }
-        ui_browser__refresh_dimensions(self);
-        newtCenteredWindow(self->width, self->height, title);
-        self->form = newt_form__new();
-        if (self->form == NULL)
-                return -1;
-
-        self->sb = newtVerticalScrollbar(self->width, 0, self->height,
-                                         HE_COLORSET_NORMAL,
-                                         HE_COLORSET_SELECTED);
-        if (self->sb == NULL)
-                return -1;
-
-        newtFormAddHotKey(self->form, NEWT_KEY_UP);
-        newtFormAddHotKey(self->form, NEWT_KEY_DOWN);
-        newtFormAddHotKey(self->form, NEWT_KEY_PGUP);
-        newtFormAddHotKey(self->form, NEWT_KEY_PGDN);
-        newtFormAddHotKey(self->form, NEWT_KEY_HOME);
-        newtFormAddHotKey(self->form, NEWT_KEY_END);
-        newtFormAddComponent(self->form, self->sb);
-        return 0;
-}
-
-static int objdump_line__show(struct objdump_line *self, struct list_head *head,
-                              int width, struct hist_entry *he, int len,
-                              bool current_entry)
-{
-        if (self->offset != -1) {
-                struct symbol *sym = he->ms.sym;
-                unsigned int hits = 0;
-                double percent = 0.0;
-                int color;
-                struct sym_priv *priv = symbol__priv(sym);
-                struct sym_ext *sym_ext = priv->ext;
-                struct sym_hist *h = priv->hist;
-                s64 offset = self->offset;
-                struct objdump_line *next = objdump__get_next_ip_line(head, self);
-
-                while (offset < (s64)len &&
-                       (next == NULL || offset < next->offset)) {
-                        if (sym_ext) {
-                                percent += sym_ext[offset].percent;
-                        } else
-                                hits += h->ip[offset];
-
-                        ++offset;
-                }
-
-                if (sym_ext == NULL && h->sum)
-                        percent = 100.0 * hits / h->sum;
-
-                color = ui_browser__percent_color(percent, current_entry);
-                SLsmg_set_color(color);
-                slsmg_printf(" %7.2f ", percent);
-                if (!current_entry)
-                        SLsmg_set_color(HE_COLORSET_CODE);
-        } else {
-                int color = ui_browser__percent_color(0, current_entry);
-                SLsmg_set_color(color);
-                slsmg_write_nstring(" ", 9);
-        }
-
-        SLsmg_write_char(':');
-        slsmg_write_nstring(" ", 8);
-        if (!*self->line)
-                slsmg_write_nstring(" ", width - 18);
-        else
-                slsmg_write_nstring(self->line, width - 18);
-
-        return 0;
-}
-
-static int ui_browser__refresh_entries(struct ui_browser *self)
-{
-        int row;
-
-        newtScrollbarSet(self->sb, self->index, self->nr_entries - 1);
-        row = self->refresh_entries(self);
-        SLsmg_set_color(HE_COLORSET_NORMAL);
-        SLsmg_fill_region(self->top + row, self->left,
-                          self->height - row, self->width, ' ');
-
-        return 0;
-}
-
-static int ui_browser__run(struct ui_browser *self, struct newtExitStruct *es)
-{
-        if (ui_browser__refresh_entries(self) < 0)
-                return -1;
-
-        while (1) {
-                off_t offset;
-
-                newtFormRun(self->form, es);
-
-                if (es->reason != NEWT_EXIT_HOTKEY)
-                        break;
-                if (is_exit_key(es->u.key))
-                        return es->u.key;
-                switch (es->u.key) {
-                case NEWT_KEY_DOWN:
-                        if (self->index == self->nr_entries - 1)
-                                break;
-                        ++self->index;
-                        if (self->index == self->first_visible_entry_idx + self->height) {
-                                ++self->first_visible_entry_idx;
-                                self->seek(self, +1, SEEK_CUR);
-                        }
-                        break;
-                case NEWT_KEY_UP:
-                        if (self->index == 0)
-                                break;
-                        --self->index;
-                        if (self->index < self->first_visible_entry_idx) {
-                                --self->first_visible_entry_idx;
-                                self->seek(self, -1, SEEK_CUR);
-                        }
-                        break;
-                case NEWT_KEY_PGDN:
-                case ' ':
-                        if (self->first_visible_entry_idx + self->height > self->nr_entries - 1)
-                                break;
-
-                        offset = self->height;
-                        if (self->index + offset > self->nr_entries - 1)
-                                offset = self->nr_entries - 1 - self->index;
-                        self->index += offset;
-                        self->first_visible_entry_idx += offset;
-                        self->seek(self, +offset, SEEK_CUR);
-                        break;
-                case NEWT_KEY_PGUP:
-                        if (self->first_visible_entry_idx == 0)
-                                break;
-
-                        if (self->first_visible_entry_idx < self->height)
-                                offset = self->first_visible_entry_idx;
-                        else
-                                offset = self->height;
-
-                        self->index -= offset;
-                        self->first_visible_entry_idx -= offset;
-                        self->seek(self, -offset, SEEK_CUR);
-                        break;
-                case NEWT_KEY_HOME:
-                        ui_browser__reset_index(self);
-                        break;
-                case NEWT_KEY_END:
-                        offset = self->height - 1;
-                        if (offset >= self->nr_entries)
-                                offset = self->nr_entries - 1;
-
-                        self->index = self->nr_entries - 1;
-                        self->first_visible_entry_idx = self->index - offset;
-                        self->seek(self, -offset, SEEK_END);
-                        break;
-                default:
-                        return es->u.key;
-                }
-                if (ui_browser__refresh_entries(self) < 0)
-                        return -1;
-        }
-        return 0;
-}
-
-static char *callchain_list__sym_name(struct callchain_list *self,
-                                      char *bf, size_t bfsize)
-{
-        if (self->ms.sym)
-                return self->ms.sym->name;
-
-        snprintf(bf, bfsize, "%#Lx", self->ip);
-        return bf;
-}
-
-static unsigned int hist_entry__annotate_browser_refresh(struct ui_browser *self)
-{
-        struct objdump_line *pos;
-        struct list_head *head = self->entries;
-        struct hist_entry *he = self->priv;
-        int row = 0;
-        int len = he->ms.sym->end - he->ms.sym->start;
-
-        if (self->first_visible_entry == NULL || self->first_visible_entry == self->entries)
-                self->first_visible_entry = head->next;
-
-        pos = list_entry(self->first_visible_entry, struct objdump_line, node);
-
-        list_for_each_entry_from(pos, head, node) {
-                bool current_entry = ui_browser__is_current_entry(self, row);
-                SLsmg_gotorc(self->top + row, self->left);
-                objdump_line__show(pos, head, self->width,
-                                   he, len, current_entry);
-                if (++row == self->height)
-                        break;
-        }
-
-        return row;
-}
-
-int hist_entry__tui_annotate(struct hist_entry *self)
-{
-        struct ui_browser browser;
-        struct newtExitStruct es;
-        struct objdump_line *pos, *n;
-        LIST_HEAD(head);
-        int ret;
-
-        if (self->ms.sym == NULL)
-                return -1;
-
-        if (self->ms.map->dso->annotate_warned)
-                return -1;
-
-        if (hist_entry__annotate(self, &head) < 0) {
-                ui__error_window(browser__last_msg);
-                return -1;
-        }
-
-        ui_helpline__push("Press <- or ESC to exit");
-
-        memset(&browser, 0, sizeof(browser));
-        browser.entries         = &head;
-        browser.refresh_entries = hist_entry__annotate_browser_refresh;
-        browser.seek            = ui_browser__list_head_seek;
-        browser.priv = self;
-        list_for_each_entry(pos, &head, node) {
-                size_t line_len = strlen(pos->line);
-                if (browser.width < line_len)
-                        browser.width = line_len;
-                ++browser.nr_entries;
-        }
-
-        browser.width += 18; /* Percentage */
-        ui_browser__show(&browser, self->ms.sym->name);
-        newtFormAddHotKey(browser.form, ' ');
-        ret = ui_browser__run(&browser, &es);
-        newtFormDestroy(browser.form);
-        newtPopWindow();
-        list_for_each_entry_safe(pos, n, &head, node) {
-                list_del(&pos->node);
-                objdump_line__free(pos);
-        }
-        ui_helpline__pop();
-        return ret;
-}
-
-struct hist_browser {
-        struct ui_browser   b;
-        struct hists        *hists;
-        struct hist_entry   *he_selection;
-        struct map_symbol   *selection;
-};
-
-static void hist_browser__reset(struct hist_browser *self);
-static int hist_browser__run(struct hist_browser *self, const char *title,
-                             struct newtExitStruct *es);
-static unsigned int hist_browser__refresh_entries(struct ui_browser *self);
-static void ui_browser__hists_seek(struct ui_browser *self,
-                                   off_t offset, int whence);
-
-static struct hist_browser *hist_browser__new(struct hists *hists)
-{
-        struct hist_browser *self = zalloc(sizeof(*self));
-
-        if (self) {
-                self->hists = hists;
-                self->b.refresh_entries = hist_browser__refresh_entries;
-                self->b.seek = ui_browser__hists_seek;
-        }
-
-        return self;
-}
-
-static void hist_browser__delete(struct hist_browser *self)
-{
-        newtFormDestroy(self->b.form);
-        newtPopWindow();
-        free(self);
-}
-
-static struct hist_entry *hist_browser__selected_entry(struct hist_browser *self)
-{
-        return self->he_selection;
-}
-
-static struct thread *hist_browser__selected_thread(struct hist_browser *self)
-{
-        return self->he_selection->thread;
-}
-
-static int hist_browser__title(char *bf, size_t size, const char *ev_name,
-                               const struct dso *dso, const struct thread *thread)
-{
-        int printed = 0;
-
-        if (thread)
-                printed += snprintf(bf + printed, size - printed,
-                                    "Thread: %s(%d)",
-                                    (thread->comm_set ?  thread->comm : ""),
-                                    thread->pid);
-        if (dso)
-                printed += snprintf(bf + printed, size - printed,
-                                    "%sDSO: %s", thread ? " " : "",
-                                    dso->short_name);
-        return printed ?: snprintf(bf, size, "Event: %s", ev_name);
-}
-
-int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
-{
-        struct hist_browser *browser = hist_browser__new(self);
-        struct pstack *fstack;
-        const struct thread *thread_filter = NULL;
-        const struct dso *dso_filter = NULL;
-        struct newtExitStruct es;
-        char msg[160];
-        int key = -1;
-
-        if (browser == NULL)
-                return -1;
-
-        fstack = pstack__new(2);
-        if (fstack == NULL)
-                goto out;
-
-        ui_helpline__push(helpline);
-
-        hist_browser__title(msg, sizeof(msg), ev_name,
-                            dso_filter, thread_filter);
-
-        while (1) {
-                const struct thread *thread;
-                const struct dso *dso;
-                char *options[16];
-                int nr_options = 0, choice = 0, i,
-                    annotate = -2, zoom_dso = -2, zoom_thread = -2;
-
-                if (hist_browser__run(browser, msg, &es))
-                        break;
-
-                thread = hist_browser__selected_thread(browser);
-                dso = browser->selection->map ? browser->selection->map->dso : NULL;
-
-                if (es.reason == NEWT_EXIT_HOTKEY) {
-                        key = es.u.key;
-
-                        switch (key) {
-                        case NEWT_KEY_F1:
-                                goto do_help;
-                        case NEWT_KEY_TAB:
-                        case NEWT_KEY_UNTAB:
-                                /*
-                                 * Exit the browser, let hists__browser_tree
-                                 * go to the next or previous
-                                 */
-                                goto out_free_stack;
-                        default:;
-                        }
-
-                        key = toupper(key);
-                        switch (key) {
-                        case 'A':
-                                if (browser->selection->map == NULL &&
-                                    browser->selection->map->dso->annotate_warned)
-                                        continue;
-                                goto do_annotate;
-                        case 'D':
-                                goto zoom_dso;
-                        case 'T':
-                                goto zoom_thread;
-                        case 'H':
-                        case '?':
-do_help:
-                                ui__help_window("->        Zoom into DSO/Threads & Annotate current symbol\n"
-                                                "<-        Zoom out\n"
-                                                "a         Annotate current symbol\n"
-                                                "h/?/F1    Show this window\n"
-                                                "d         Zoom into current DSO\n"
-                                                "t         Zoom into current Thread\n"
-                                                "q/CTRL+C  Exit browser");
-                                continue;
-                        default:;
-                        }
-                        if (is_exit_key(key)) {
-                                if (key == NEWT_KEY_ESCAPE) {
-                                        if (dialog_yesno("Do you really want to exit?"))
-                                                break;
-                                        else
-                                                continue;
-                                } else
-                                        break;
-                        }
-
-                        if (es.u.key == NEWT_KEY_LEFT) {
-                                const void *top;
-
-                                if (pstack__empty(fstack))
-                                        continue;
-                                top = pstack__pop(fstack);
-                                if (top == &dso_filter)
-                                        goto zoom_out_dso;
-                                if (top == &thread_filter)
-                                        goto zoom_out_thread;
-                                continue;
-                        }
-                }
-
-                if (browser->selection->sym != NULL &&
-                    !browser->selection->map->dso->annotate_warned &&
-                    asprintf(&options[nr_options], "Annotate %s",
-                             browser->selection->sym->name) > 0)
-                        annotate = nr_options++;
-
-                if (thread != NULL &&
-                    asprintf(&options[nr_options], "Zoom %s %s(%d) thread",
-                             (thread_filter ? "out of" : "into"),
-                             (thread->comm_set ? thread->comm : ""),
-                             thread->pid) > 0)
-                        zoom_thread = nr_options++;
-
-                if (dso != NULL &&
-                    asprintf(&options[nr_options], "Zoom %s %s DSO",
-                             (dso_filter ? "out of" : "into"),
-                             (dso->kernel ? "the Kernel" : dso->short_name)) > 0)
-                        zoom_dso = nr_options++;
-
-                options[nr_options++] = (char *)"Exit";
-
-                choice = popup_menu(nr_options, options);
-
-                for (i = 0; i < nr_options - 1; ++i)
-                        free(options[i]);
-
-                if (choice == nr_options - 1)
-                        break;
-
-                if (choice == -1)
-                        continue;
-
-                if (choice == annotate) {
-                        struct hist_entry *he;
-do_annotate:
-                        if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) {
-                                browser->selection->map->dso->annotate_warned = 1;
-                                ui_helpline__puts("No vmlinux file found, can't "
-                                                 "annotate with just a "
-                                                 "kallsyms file");
-                                continue;
-                        }
-
-                        he = hist_browser__selected_entry(browser);
-                        if (he == NULL)
-                                continue;
-
-                        hist_entry__tui_annotate(he);
-                } else if (choice == zoom_dso) {
-zoom_dso:
-                        if (dso_filter) {
-                                pstack__remove(fstack, &dso_filter);
-zoom_out_dso:
-                                ui_helpline__pop();
-                                dso_filter = NULL;
-                        } else {
-                                if (dso == NULL)
-                                        continue;
-                                ui_helpline__fpush("To zoom out press <- or -> + \"Zoom out of %s DSO\"",
-                                                   dso->kernel ? "the Kernel" : dso->short_name);
-                                dso_filter = dso;
-                                pstack__push(fstack, &dso_filter);
-                        }
-                        hists__filter_by_dso(self, dso_filter);
-                        hist_browser__title(msg, sizeof(msg), ev_name,
-                                            dso_filter, thread_filter);
-                        hist_browser__reset(browser);
-                } else if (choice == zoom_thread) {
-zoom_thread:
-                        if (thread_filter) {
-                                pstack__remove(fstack, &thread_filter);
-zoom_out_thread:
-                                ui_helpline__pop();
-                                thread_filter = NULL;
-                        } else {
-                                ui_helpline__fpush("To zoom out press <- or -> + \"Zoom out of %s(%d) thread\"",
-                                                   thread->comm_set ? thread->comm : "",
-                                                   thread->pid);
-                                thread_filter = thread;
-                                pstack__push(fstack, &thread_filter);
-                        }
-                        hists__filter_by_thread(self, thread_filter);
-                        hist_browser__title(msg, sizeof(msg), ev_name,
-                                            dso_filter, thread_filter);
-                        hist_browser__reset(browser);
-                }
-        }
-out_free_stack:
-        pstack__delete(fstack);
-out:
-        hist_browser__delete(browser);
-        return key;
-}
-
-int hists__tui_browse_tree(struct rb_root *self, const char *help)
-{
-        struct rb_node *first = rb_first(self), *nd = first, *next;
-        int key = 0;
-
-        while (nd) {
-                struct hists *hists = rb_entry(nd, struct hists, rb_node);
-                const char *ev_name = __event_name(hists->type, hists->config);
-
-                key = hists__browse(hists, help, ev_name);
-
-                if (is_exit_key(key))
-                        break;
-
-                switch (key) {
-                case NEWT_KEY_TAB:
-                        next = rb_next(nd);
-                        if (next)
-                                nd = next;
-                        break;
-                case NEWT_KEY_UNTAB:
-                        if (nd == first)
-                                continue;
-                        nd = rb_prev(nd);
-                default:
-                        break;
-                }
-        }
-
-        return key;
-}
-
-static struct newtPercentTreeColors {
-        const char *topColorFg, *topColorBg;
-        const char *mediumColorFg, *mediumColorBg;
-        const char *normalColorFg, *normalColorBg;
-        const char *selColorFg, *selColorBg;
-        const char *codeColorFg, *codeColorBg;
-} defaultPercentTreeColors = {
-        "red",       "lightgray",
-        "green",     "lightgray",
-        "black",     "lightgray",
-        "lightgray", "magenta",
-        "blue",      "lightgray",
-};
-
-static void newt_suspend(void *d __used)
-{
-        newtSuspend();
-        raise(SIGTSTP);
-        newtResume();
-}
-
-void setup_browser(void)
-{
-        struct newtPercentTreeColors *c = &defaultPercentTreeColors;
-
-        if (!isatty(1) || !use_browser || dump_trace) {
-                use_browser = 0;
-                setup_pager();
-                return;
-        }
-
-        use_browser = 1;
-        newtInit();
-        newtCls();
-        newtSetSuspendCallback(newt_suspend, NULL);
-        ui_helpline__puts(" ");
-        sltt_set_color(HE_COLORSET_TOP, NULL, c->topColorFg, c->topColorBg);
-        sltt_set_color(HE_COLORSET_MEDIUM, NULL, c->mediumColorFg, c->mediumColorBg);
-        sltt_set_color(HE_COLORSET_NORMAL, NULL, c->normalColorFg, c->normalColorBg);
-        sltt_set_color(HE_COLORSET_SELECTED, NULL, c->selColorFg, c->selColorBg);
-        sltt_set_color(HE_COLORSET_CODE, NULL, c->codeColorFg, c->codeColorBg);
-}
-
-void exit_browser(bool wait_for_ok)
-{
-        if (use_browser > 0) {
-                if (wait_for_ok) {
-                        char title[] = "Fatal Error", ok[] = "Ok";
-                        newtWinMessage(title, ok, browser__last_msg);
-                }
-                newtFinished();
-        }
-}
-
-static void hist_browser__refresh_dimensions(struct hist_browser *self)
-{
-        /* 3 == +/- toggle symbol before actual hist_entry rendering */
-        self->b.width = 3 + (hists__sort_list_width(self->hists) +
-                             sizeof("[k]"));
-}
-
-static void hist_browser__reset(struct hist_browser *self)
-{
-        self->b.nr_entries = self->hists->nr_entries;
-        hist_browser__refresh_dimensions(self);
-        ui_browser__reset_index(&self->b);
-}
-
-static char tree__folded_sign(bool unfolded)
-{
-        return unfolded ? '-' : '+';
-}
-
-static char map_symbol__folded(const struct map_symbol *self)
-{
-        return self->has_children ? tree__folded_sign(self->unfolded) : ' ';
-}
-
-static char hist_entry__folded(const struct hist_entry *self)
-{
-        return map_symbol__folded(&self->ms);
-}
-
-static char callchain_list__folded(const struct callchain_list *self)
-{
-        return map_symbol__folded(&self->ms);
-}
-
-static bool map_symbol__toggle_fold(struct map_symbol *self)
-{
-        if (!self->has_children)
-                return false;
-
-        self->unfolded = !self->unfolded;
-        return true;
-}
-
-#define LEVEL_OFFSET_STEP 3
-
-static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self,
-                                                     struct callchain_node *chain_node,
-                                                     u64 total, int level,
-                                                     unsigned short row,
-                                                     off_t *row_offset,
-                                                     bool *is_current_entry)
-{
-        struct rb_node *node;
-        int first_row = row, width, offset = level * LEVEL_OFFSET_STEP;
-        u64 new_total, remaining;
-
-        if (callchain_param.mode == CHAIN_GRAPH_REL)
-                new_total = chain_node->children_hit;
-        else
-                new_total = total;
-
-        remaining = new_total;
-        node = rb_first(&chain_node->rb_root);
-        while (node) {
-                struct callchain_node *child = rb_entry(node, struct callchain_node, rb_node);
-                struct rb_node *next = rb_next(node);
-                u64 cumul = cumul_hits(child);
-                struct callchain_list *chain;
-                char folded_sign = ' ';
-                int first = true;
-                int extra_offset = 0;
-
-                remaining -= cumul;
-
-                list_for_each_entry(chain, &child->val, list) {
-                        char ipstr[BITS_PER_LONG / 4 + 1], *alloc_str;
-                        const char *str;
-                        int color;
-                        bool was_first = first;
-
-                        if (first) {
-                                first = false;
-                                chain->ms.has_children = chain->list.next != &child->val ||
-                                                         rb_first(&child->rb_root) != NULL;
-                        } else {
-                                extra_offset = LEVEL_OFFSET_STEP;
-                                chain->ms.has_children = chain->list.next == &child->val &&
-                                                         rb_first(&child->rb_root) != NULL;
-                        }
-
-                        folded_sign = callchain_list__folded(chain);
-                        if (*row_offset != 0) {
-                                --*row_offset;
-                                goto do_next;
-                        }
-
-                        alloc_str = NULL;
-                        str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr));
-                        if (was_first) {
-                                double percent = cumul * 100.0 / new_total;
-
-                                if (asprintf(&alloc_str, "%2.2f%% %s", percent, str) < 0)
-                                        str = "Not enough memory!";
-                                else
-                                        str = alloc_str;
-                        }
-
-                        color = HE_COLORSET_NORMAL;
-                        width = self->b.width - (offset + extra_offset + 2);
-                        if (ui_browser__is_current_entry(&self->b, row)) {
-                                self->selection = &chain->ms;
-                                color = HE_COLORSET_SELECTED;
-                                *is_current_entry = true;
-                        }
-
-                        SLsmg_set_color(color);
-                        SLsmg_gotorc(self->b.top + row, self->b.left);
-                        slsmg_write_nstring(" ", offset + extra_offset);
-                        slsmg_printf("%c ", folded_sign);
-                        slsmg_write_nstring(str, width);
-                        free(alloc_str);
-
-                        if (++row == self->b.height)
-                                goto out;
-do_next:
-                        if (folded_sign == '+')
-                                break;
-                }
-
-                if (folded_sign == '-') {
-                        const int new_level = level + (extra_offset ? 2 : 1);
-                        row += hist_browser__show_callchain_node_rb_tree(self, child, new_total,
-                                                                         new_level, row, row_offset,
-                                                                         is_current_entry);
-                }
-                if (row == self->b.height)
-                        goto out;
-                node = next;
-        }
-out:
-        return row - first_row;
-}
-
-static int hist_browser__show_callchain_node(struct hist_browser *self,
-                                             struct callchain_node *node,
-                                             int level, unsigned short row,
-                                             off_t *row_offset,
-                                             bool *is_current_entry)
-{
-        struct callchain_list *chain;
-        int first_row = row,
-             offset = level * LEVEL_OFFSET_STEP,
-             width = self->b.width - offset;
-        char folded_sign = ' ';
-
-        list_for_each_entry(chain, &node->val, list) {
-                char ipstr[BITS_PER_LONG / 4 + 1], *s;
-                int color;
-                /*
-                 * FIXME: This should be moved to somewhere else,
-                 * probably when the callchain is created, so as not to
-                 * traverse it all over again
-                 */
-                chain->ms.has_children = rb_first(&node->rb_root) != NULL;
-                folded_sign = callchain_list__folded(chain);
-
-                if (*row_offset != 0) {
-                        --*row_offset;
-                        continue;
-                }
-
-                color = HE_COLORSET_NORMAL;
-                if (ui_browser__is_current_entry(&self->b, row)) {
-                        self->selection = &chain->ms;
-                        color = HE_COLORSET_SELECTED;
-                        *is_current_entry = true;
-                }
-
-                s = callchain_list__sym_name(chain, ipstr, sizeof(ipstr));
-                SLsmg_gotorc(self->b.top + row, self->b.left);
-                SLsmg_set_color(color);
-                slsmg_write_nstring(" ", offset);
-                slsmg_printf("%c ", folded_sign);
-                slsmg_write_nstring(s, width - 2);
-
-                if (++row == self->b.height)
-                        goto out;
-        }
-
-        if (folded_sign == '-')
-                row += hist_browser__show_callchain_node_rb_tree(self, node,
-                                                                 self->hists->stats.total_period,
-                                                                 level + 1, row,
-                                                                 row_offset,
-                                                                 is_current_entry);
-out:
-        return row - first_row;
-}
-
-static int hist_browser__show_callchain(struct hist_browser *self,
-                                        struct rb_root *chain,
-                                        int level, unsigned short row,
-                                        off_t *row_offset,
-                                        bool *is_current_entry)
-{
-        struct rb_node *nd;
-        int first_row = row;
-
-        for (nd = rb_first(chain); nd; nd = rb_next(nd)) {
-                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
-
-                row += hist_browser__show_callchain_node(self, node, level,
-                                                         row, row_offset,
-                                                         is_current_entry);
-                if (row == self->b.height)
-                        break;
-        }
-
-        return row - first_row;
-}
-
-static int hist_browser__show_entry(struct hist_browser *self,
-                                    struct hist_entry *entry,
-                                    unsigned short row)
-{
-        char s[256];
-        double percent;
-        int printed = 0;
-        int color, width = self->b.width;
-        char folded_sign = ' ';
-        bool current_entry = ui_browser__is_current_entry(&self->b, row);
-        off_t row_offset = entry->row_offset;
-
-        if (current_entry) {
-                self->he_selection = entry;
-                self->selection = &entry->ms;
-        }
-
-        if (symbol_conf.use_callchain) {
-                entry->ms.has_children = !RB_EMPTY_ROOT(&entry->sorted_chain);
-                folded_sign = hist_entry__folded(entry);
-        }
-
-        if (row_offset == 0) {
-                hist_entry__snprintf(entry, s, sizeof(s), self->hists, NULL, false,
-                                     0, false, self->hists->stats.total_period);
-                percent = (entry->period * 100.0) / self->hists->stats.total_period;
-
-                color = HE_COLORSET_SELECTED;
-                if (!current_entry) {
-                        if (percent >= MIN_RED)
-                                color = HE_COLORSET_TOP;
-                        else if (percent >= MIN_GREEN)
-                                color = HE_COLORSET_MEDIUM;
-                        else
-                                color = HE_COLORSET_NORMAL;
-                }
-
-                SLsmg_set_color(color);
-                SLsmg_gotorc(self->b.top + row, self->b.left);
-                if (symbol_conf.use_callchain) {
-                        slsmg_printf("%c ", folded_sign);
-                        width -= 2;
-                }
-                slsmg_write_nstring(s, width);
-                ++row;
-                ++printed;
-        } else
-                --row_offset;
-
-        if (folded_sign == '-' && row != self->b.height) {
-                printed += hist_browser__show_callchain(self, &entry->sorted_chain,
-                                                        1, row, &row_offset,
-                                                        &current_entry);
-                if (current_entry)
-                        self->he_selection = entry;
-        }
-
-        return printed;
-}
-
-static unsigned int hist_browser__refresh_entries(struct ui_browser *self)
-{
-        unsigned row = 0;
-        struct rb_node *nd;
-        struct hist_browser *hb = container_of(self, struct hist_browser, b);
-
-        if (self->first_visible_entry == NULL)
-                self->first_visible_entry = rb_first(&hb->hists->entries);
-
-        for (nd = self->first_visible_entry; nd; nd = rb_next(nd)) {
-                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
-
-                if (h->filtered)
-                        continue;
-
-                row += hist_browser__show_entry(hb, h, row);
-                if (row == self->height)
-                        break;
-        }
-
-        return row;
-}
-
-static void callchain_node__init_have_children_rb_tree(struct callchain_node *self)
-{
-        struct rb_node *nd = rb_first(&self->rb_root);
-
-        for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) {
-                struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node);
-                struct callchain_list *chain;
-                int first = true;
-
-                list_for_each_entry(chain, &child->val, list) {
-                        if (first) {
-                                first = false;
-                                chain->ms.has_children = chain->list.next != &child->val ||
-                                                         rb_first(&child->rb_root) != NULL;
-                        } else
-                                chain->ms.has_children = chain->list.next == &child->val &&
-                                                         rb_first(&child->rb_root) != NULL;
-                }
-
-                callchain_node__init_have_children_rb_tree(child);
-        }
-}
-
-static void callchain_node__init_have_children(struct callchain_node *self)
-{
-        struct callchain_list *chain;
-
-        list_for_each_entry(chain, &self->val, list)
-                chain->ms.has_children = rb_first(&self->rb_root) != NULL;
-
-        callchain_node__init_have_children_rb_tree(self);
-}
-
-static void callchain__init_have_children(struct rb_root *self)
-{
-        struct rb_node *nd;
-
-        for (nd = rb_first(self); nd; nd = rb_next(nd)) {
-                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
-                callchain_node__init_have_children(node);
-        }
-}
-
-static void hist_entry__init_have_children(struct hist_entry *self)
-{
-        if (!self->init_have_children) {
-                callchain__init_have_children(&self->sorted_chain);
-                self->init_have_children = true;
-        }
-}
-
-static struct rb_node *hists__filter_entries(struct rb_node *nd)
-{
-        while (nd != NULL) {
-                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
-                if (!h->filtered)
-                        return nd;
-
-                nd = rb_next(nd);
-        }
-
-        return NULL;
-}
-
-static struct rb_node *hists__filter_prev_entries(struct rb_node *nd)
-{
-        while (nd != NULL) {
-                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
-                if (!h->filtered)
-                        return nd;
-
-                nd = rb_prev(nd);
-        }
-
-        return NULL;
-}
-
-static void ui_browser__hists_seek(struct ui_browser *self,
-                                   off_t offset, int whence)
-{
-        struct hist_entry *h;
-        struct rb_node *nd;
-        bool first = true;
-
-        switch (whence) {
-        case SEEK_SET:
-                nd = hists__filter_entries(rb_first(self->entries));
-                break;
-        case SEEK_CUR:
-                nd = self->first_visible_entry;
-                goto do_offset;
-        case SEEK_END:
-                nd = hists__filter_prev_entries(rb_last(self->entries));
-                first = false;
-                break;
-        default:
-                return;
-        }
-
-        /*
-         * Moves not relative to the first visible entry invalidates its
-         * row_offset:
-         */
-        h = rb_entry(self->first_visible_entry, struct hist_entry, rb_node);
-        h->row_offset = 0;
-
-        /*
-         * Here we have to check if nd is expanded (+), if it is we can't go
-         * the next top level hist_entry, instead we must compute an offset of
-         * what _not_ to show and not change the first visible entry.
-         *
-         * This offset increments when we are going from top to bottom and
-         * decreases when we're going from bottom to top.
-         *
-         * As we don't have backpointers to the top level in the callchains
-         * structure, we need to always print the whole hist_entry callchain,
-         * skipping the first ones that are before the first visible entry
-         * and stop when we printed enough lines to fill the screen.
-         */
-do_offset:
-        if (offset > 0) {
-                do {
-                        h = rb_entry(nd, struct hist_entry, rb_node);
-                        if (h->ms.unfolded) {
-                                u16 remaining = h->nr_rows - h->row_offset;
-                                if (offset > remaining) {
-                                        offset -= remaining;
-                                        h->row_offset = 0;
-                                } else {
-                                        h->row_offset += offset;
-                                        offset = 0;
-                                        self->first_visible_entry = nd;
-                                        break;
-                                }
-                        }
-                        nd = hists__filter_entries(rb_next(nd));
-                        if (nd == NULL)
-                                break;
-                        --offset;
-                        self->first_visible_entry = nd;
-                } while (offset != 0);
-        } else if (offset < 0) {
-                while (1) {
-                        h = rb_entry(nd, struct hist_entry, rb_node);
-                        if (h->ms.unfolded) {
-                                if (first) {
-                                        if (-offset > h->row_offset) {
-                                                offset += h->row_offset;
-                                                h->row_offset = 0;
-                                        } else {
-                                                h->row_offset += offset;
-                                                offset = 0;
-                                                self->first_visible_entry = nd;
-                                                break;
-                                        }
-                                } else {
-                                        if (-offset > h->nr_rows) {
-                                                offset += h->nr_rows;
-                                                h->row_offset = 0;
-                                        } else {
-                                                h->row_offset = h->nr_rows + offset;
-                                                offset = 0;
-                                                self->first_visible_entry = nd;
-                                                break;
-                                        }
-                                }
-                        }
-
-                        nd = hists__filter_prev_entries(rb_prev(nd));
-                        if (nd == NULL)
-                                break;
-                        ++offset;
-                        self->first_visible_entry = nd;
-                        if (offset == 0) {
-                                /*
-                                 * Last unfiltered hist_entry, check if it is
-                                 * unfolded, if it is then we should have
-                                 * row_offset at its last entry.
-                                 */
-                                h = rb_entry(nd, struct hist_entry, rb_node);
-                                if (h->ms.unfolded)
-                                        h->row_offset = h->nr_rows;
-                                break;
-                        }
-                        first = false;
-                }
-        } else {
-                self->first_visible_entry = nd;
-                h = rb_entry(nd, struct hist_entry, rb_node);
-                h->row_offset = 0;
-        }
-}
-
-static int callchain_node__count_rows_rb_tree(struct callchain_node *self)
-{
-        int n = 0;
-        struct rb_node *nd;
-
-        for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) {
-                struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node);
-                struct callchain_list *chain;
-                char folded_sign = ' '; /* No children */
-
-                list_for_each_entry(chain, &child->val, list) {
-                        ++n;
-                        /* We need this because we may not have children */
-                        folded_sign = callchain_list__folded(chain);
-                        if (folded_sign == '+')
-                                break;
-                }
-
-                if (folded_sign == '-') /* Have children and they're unfolded */
-                        n += callchain_node__count_rows_rb_tree(child);
-        }
-
-        return n;
-}
-
-static int callchain_node__count_rows(struct callchain_node *node)
-{
-        struct callchain_list *chain;
-        bool unfolded = false;
-        int n = 0;
-
-        list_for_each_entry(chain, &node->val, list) {
-                ++n;
-                unfolded = chain->ms.unfolded;
-        }
-
-        if (unfolded)
-                n += callchain_node__count_rows_rb_tree(node);
-
-        return n;
-}
-
-static int callchain__count_rows(struct rb_root *chain)
-{
-        struct rb_node *nd;
-        int n = 0;
-
-        for (nd = rb_first(chain); nd; nd = rb_next(nd)) {
-                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
-                n += callchain_node__count_rows(node);
-        }
-
-        return n;
-}
-
-static bool hist_browser__toggle_fold(struct hist_browser *self)
-{
-        if (map_symbol__toggle_fold(self->selection)) {
-                struct hist_entry *he = self->he_selection;
-
-                hist_entry__init_have_children(he);
-                self->hists->nr_entries -= he->nr_rows;
-
-                if (he->ms.unfolded)
-                        he->nr_rows = callchain__count_rows(&he->sorted_chain);
-                else
-                        he->nr_rows = 0;
-                self->hists->nr_entries += he->nr_rows;
-                self->b.nr_entries = self->hists->nr_entries;
-
-                return true;
-        }
-
-        /* If it doesn't have children, no toggling performed */
-        return false;
-}
-
-static int hist_browser__run(struct hist_browser *self, const char *title,
-                             struct newtExitStruct *es)
-{
-        char str[256], unit;
-        unsigned long nr_events = self->hists->stats.nr_events[PERF_RECORD_SAMPLE];
-
-        self->b.entries = &self->hists->entries;
-        self->b.nr_entries = self->hists->nr_entries;
-
-        hist_browser__refresh_dimensions(self);
-
-        nr_events = convert_unit(nr_events, &unit);
-        snprintf(str, sizeof(str), "Events: %lu%c                            ",
-                 nr_events, unit);
-        newtDrawRootText(0, 0, str);
-
-        if (ui_browser__show(&self->b, title) < 0)
-                return -1;
-
-        newtFormAddHotKey(self->b.form, 'A');
-        newtFormAddHotKey(self->b.form, 'a');
-        newtFormAddHotKey(self->b.form, '?');
-        newtFormAddHotKey(self->b.form, 'h');
-        newtFormAddHotKey(self->b.form, 'H');
-        newtFormAddHotKey(self->b.form, 'd');
-
-        newtFormAddHotKey(self->b.form, NEWT_KEY_LEFT);
-        newtFormAddHotKey(self->b.form, NEWT_KEY_RIGHT);
-        newtFormAddHotKey(self->b.form, NEWT_KEY_ENTER);
-
-        while (1) {
-                ui_browser__run(&self->b, es);
-
-                if (es->reason != NEWT_EXIT_HOTKEY)
-                        break;
-                switch (es->u.key) {
-                case 'd': { /* Debug */
-                        static int seq;
-                        struct hist_entry *h = rb_entry(self->b.first_visible_entry,
-                                                        struct hist_entry, rb_node);
-                        ui_helpline__pop();
-                        ui_helpline__fpush("%d: nr_ent=(%d,%d), height=%d, idx=%d, fve: idx=%d, row_off=%d, nrows=%d",
-                                           seq++, self->b.nr_entries,
-                                           self->hists->nr_entries,
-                                           self->b.height,
-                                           self->b.index,
-                                           self->b.first_visible_entry_idx,
-                                           h->row_offset, h->nr_rows);
-                }
-                        continue;
-                case NEWT_KEY_ENTER:
-                        if (hist_browser__toggle_fold(self))
-                                break;
-                        /* fall thru */
-                default:
-                        return 0;
-                }
-        }
-        return 0;
-}
diff --git a/tools/perf/util/path.c b/tools/perf/util/path.c
index 58a470d036dd..bd7497711424 100644
--- a/tools/perf/util/path.c
+++ b/tools/perf/util/path.c
@@ -22,6 +22,7 @@ static const char *get_perf_dir(void)
         return ".";
 }
 
+#ifdef NO_STRLCPY
 size_t strlcpy(char *dest, const char *src, size_t size)
 {
         size_t ret = strlen(src);
@@ -33,7 +34,7 @@ size_t strlcpy(char *dest, const char *src, size_t size)
         }
         return ret;
 }
-
+#endif
 
 static char *get_pathname(void)
 {
diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
index 2e665cb84055..61191c6cbe7a 100644
--- a/tools/perf/util/probe-event.c
+++ b/tools/perf/util/probe-event.c
@@ -74,10 +74,9 @@ static int e_snprintf(char *str, size_t size, const char *format, ...)
 static char *synthesize_perf_probe_point(struct perf_probe_point *pp);
 static struct machine machine;
 
-/* Initialize symbol maps and path of vmlinux */
+/* Initialize symbol maps and path of vmlinux/modules */
 static int init_vmlinux(void)
 {
-        struct dso *kernel;
         int ret;
 
         symbol_conf.sort_by_name = true;
@@ -91,33 +90,70 @@ static int init_vmlinux(void)
                 goto out;
         }
 
-        ret = machine__init(&machine, "/", 0);
+        ret = machine__init(&machine, "", HOST_KERNEL_ID);
         if (ret < 0)
                 goto out;
 
-        kernel = dso__new_kernel(symbol_conf.vmlinux_name);
-        if (kernel == NULL)
-                die("Failed to create kernel dso.");
-
-        ret = __machine__create_kernel_maps(&machine, kernel);
-        if (ret < 0)
-                pr_debug("Failed to create kernel maps.\n");
-
+        if (machine__create_kernel_maps(&machine) < 0) {
+                pr_debug("machine__create_kernel_maps ");
+                goto out;
+        }
 out:
         if (ret < 0)
                 pr_warning("Failed to init vmlinux path.\n");
         return ret;
 }
 
+static struct symbol *__find_kernel_function_by_name(const char *name,
+                                                     struct map **mapp)
+{
+        return machine__find_kernel_function_by_name(&machine, name, mapp,
+                                                     NULL);
+}
+
+const char *kernel_get_module_path(const char *module)
+{
+        struct dso *dso;
+        struct map *map;
+        const char *vmlinux_name;
+
+        if (module) {
+                list_for_each_entry(dso, &machine.kernel_dsos, node) {
+                        if (strncmp(dso->short_name + 1, module,
+                                    dso->short_name_len - 2) == 0)
+                                goto found;
+                }
+                pr_debug("Failed to find module %s.\n", module);
+                return NULL;
+        }
+
+        map = machine.vmlinux_maps[MAP__FUNCTION];
+        dso = map->dso;
+
+        vmlinux_name = symbol_conf.vmlinux_name;
+        if (vmlinux_name) {
+                if (dso__load_vmlinux(dso, map, vmlinux_name, NULL) <= 0)
+                        return NULL;
+        } else {
+                if (dso__load_vmlinux_path(dso, map, NULL) <= 0) {
+                        pr_debug("Failed to load kernel map.\n");
+                        return NULL;
+                }
+        }
+found:
+        return dso->long_name;
+}
+
 #ifdef DWARF_SUPPORT
-static int open_vmlinux(void)
+static int open_vmlinux(const char *module)
 {
-        if (map__load(machine.vmlinux_maps[MAP__FUNCTION], NULL) < 0) {
-                pr_debug("Failed to load kernel map.\n");
-                return -EINVAL;
+        const char *path = kernel_get_module_path(module);
+        if (!path) {
+                pr_err("Failed to find path of %s module", module ?: "kernel");
+                return -ENOENT;
         }
-        pr_debug("Try to open %s\n", machine.vmlinux_maps[MAP__FUNCTION]->dso->long_name);
-        return open(machine.vmlinux_maps[MAP__FUNCTION]->dso->long_name, O_RDONLY);
+        pr_debug("Try to open %s\n", path);
+        return open(path, O_RDONLY);
 }
 
 /*
@@ -125,20 +161,19 @@ static int open_vmlinux(void)
  * Currently only handles kprobes.
  */
 static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
-                                       struct perf_probe_point *pp)
+                                        struct perf_probe_point *pp)
 {
         struct symbol *sym;
-        int fd, ret = -ENOENT;
+        struct map *map;
+        u64 addr;
+        int ret = -ENOENT;
 
-        sym = map__find_symbol_by_name(machine.vmlinux_maps[MAP__FUNCTION],
-                                       tp->symbol, NULL);
+        sym = __find_kernel_function_by_name(tp->symbol, &map);
         if (sym) {
-                fd = open_vmlinux();
-                if (fd >= 0) {
-                        ret = find_perf_probe_point(fd,
-                                                 sym->start + tp->offset, pp);
-                        close(fd);
-                }
+                addr = map->unmap_ip(map, sym->start + tp->offset);
+                pr_debug("try to find %s+%ld@%llx\n", tp->symbol,
+                         tp->offset, addr);
+                ret = find_perf_probe_point((unsigned long)addr, pp);
         }
         if (ret <= 0) {
                 pr_debug("Failed to find corresponding probes from "
@@ -156,12 +191,12 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
 /* Try to find perf_probe_event with debuginfo */
 static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
                                            struct probe_trace_event **tevs,
-                                           int max_tevs)
+                                           int max_tevs, const char *module)
 {
         bool need_dwarf = perf_probe_event_need_dwarf(pev);
         int fd, ntevs;
 
-        fd = open_vmlinux();
+        fd = open_vmlinux(module);
         if (fd < 0) {
                 if (need_dwarf) {
                         pr_warning("Failed to open debuginfo file.\n");
@@ -300,7 +335,7 @@ error:
  * Show line-range always requires debuginfo to find source file and
  * line number.
  */
-int show_line_range(struct line_range *lr)
+int show_line_range(struct line_range *lr, const char *module)
 {
         int l = 1;
         struct line_node *ln;
@@ -313,7 +348,7 @@ int show_line_range(struct line_range *lr)
         if (ret < 0)
                 return ret;
 
-        fd = open_vmlinux();
+        fd = open_vmlinux(module);
         if (fd < 0) {
                 pr_warning("Failed to open debuginfo file.\n");
                 return fd;
@@ -378,11 +413,84 @@ end:
         return ret;
 }
 
+static int show_available_vars_at(int fd, struct perf_probe_event *pev,
+                                  int max_vls, bool externs)
+{
+        char *buf;
+        int ret, i;
+        struct str_node *node;
+        struct variable_list *vls = NULL, *vl;
+
+        buf = synthesize_perf_probe_point(&pev->point);
+        if (!buf)
+                return -EINVAL;
+        pr_debug("Searching variables at %s\n", buf);
+
+        ret = find_available_vars_at(fd, pev, &vls, max_vls, externs);
+        if (ret > 0) {
+                /* Some variables were found */
+                fprintf(stdout, "Available variables at %s\n", buf);
+                for (i = 0; i < ret; i++) {
+                        vl = &vls[i];
+                        /*
+                         * A probe point might be converted to
+                         * several trace points.
+                         */
+                        fprintf(stdout, "\t@<%s+%lu>\n", vl->point.symbol,
+                                vl->point.offset);
+                        free(vl->point.symbol);
+                        if (vl->vars) {
+                                strlist__for_each(node, vl->vars)
+                                        fprintf(stdout, "\t\t%s\n", node->s);
+                                strlist__delete(vl->vars);
+                        } else
+                                fprintf(stdout, "(No variables)\n");
+                }
+                free(vls);
+        } else
+                pr_err("Failed to find variables at %s (%d)\n", buf, ret);
+
+        free(buf);
+        return ret;
+}
+
+/* Show available variables on given probe point */
+int show_available_vars(struct perf_probe_event *pevs, int npevs,
+                        int max_vls, const char *module, bool externs)
+{
+        int i, fd, ret = 0;
+
+        ret = init_vmlinux();
+        if (ret < 0)
+                return ret;
+
+        fd = open_vmlinux(module);
+        if (fd < 0) {
+                pr_warning("Failed to open debuginfo file.\n");
+                return fd;
+        }
+
+        setup_pager();
+
+        for (i = 0; i < npevs && ret >= 0; i++)
+                ret = show_available_vars_at(fd, &pevs[i], max_vls, externs);
+
+        close(fd);
+        return ret;
+}
+
 #else   /* !DWARF_SUPPORT */
 
 static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
-                                       struct perf_probe_point *pp)
+                                        struct perf_probe_point *pp)
 {
+        struct symbol *sym;
+
+        sym = __find_kernel_function_by_name(tp->symbol, NULL);
+        if (!sym) {
+                pr_err("Failed to find symbol %s in kernel.\n", tp->symbol);
+                return -ENOENT;
+        }
         pp->function = strdup(tp->symbol);
         if (pp->function == NULL)
                 return -ENOMEM;
@@ -394,7 +502,7 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp,
 
 static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
                                 struct probe_trace_event **tevs __unused,
-                                int max_tevs __unused)
+                                int max_tevs __unused, const char *mod __unused)
 {
         if (perf_probe_event_need_dwarf(pev)) {
                 pr_warning("Debuginfo-analysis is not supported.\n");
@@ -403,12 +511,19 @@ static int try_to_find_probe_trace_events(struct perf_probe_event *pev,
         return 0;
 }
 
-int show_line_range(struct line_range *lr __unused)
+int show_line_range(struct line_range *lr __unused, const char *module __unused)
 {
         pr_warning("Debuginfo-analysis is not supported.\n");
         return -ENOSYS;
 }
 
+int show_available_vars(struct perf_probe_event *pevs __unused,
+                        int npevs __unused, int max_vls __unused,
+                        const char *module __unused, bool externs __unused)
+{
+        pr_warning("Debuginfo-analysis is not supported.\n");
+        return -ENOSYS;
+}
 #endif
 
 int parse_line_range_desc(const char *arg, struct line_range *lr)
@@ -1087,7 +1202,7 @@ error:
 }
 
 static int convert_to_perf_probe_event(struct probe_trace_event *tev,
-                                struct perf_probe_event *pev)
+                                       struct perf_probe_event *pev)
 {
         char buf[64] = "";
         int i, ret;
@@ -1516,14 +1631,14 @@ static int __add_probe_trace_events(struct perf_probe_event *pev,
 
 static int convert_to_probe_trace_events(struct perf_probe_event *pev,
                                           struct probe_trace_event **tevs,
-                                          int max_tevs)
+                                          int max_tevs, const char *module)
 {
         struct symbol *sym;
         int ret = 0, i;
         struct probe_trace_event *tev;
 
         /* Convert perf_probe_event with debuginfo */
-        ret = try_to_find_probe_trace_events(pev, tevs, max_tevs);
+        ret = try_to_find_probe_trace_events(pev, tevs, max_tevs, module);
         if (ret != 0)
                 return ret;
 
@@ -1539,6 +1654,7 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev,
                 goto error;
         }
         tev->point.offset = pev->point.offset;
+        tev->point.retprobe = pev->point.retprobe;
         tev->nargs = pev->nargs;
         if (tev->nargs) {
                 tev->args = zalloc(sizeof(struct probe_trace_arg)
@@ -1571,8 +1687,7 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev,
         }
 
         /* Currently just checking function name from symbol map */
-        sym = map__find_symbol_by_name(machine.vmlinux_maps[MAP__FUNCTION],
-                                       tev->point.symbol, NULL);
+        sym = __find_kernel_function_by_name(tev->point.symbol, NULL);
         if (!sym) {
                 pr_warning("Kernel symbol \'%s\' not found.\n",
                            tev->point.symbol);
@@ -1595,7 +1710,7 @@ struct __event_package {
 };
 
 int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
-                          bool force_add, int max_tevs)
+                          int max_tevs, const char *module, bool force_add)
 {
         int i, j, ret;
         struct __event_package *pkgs;
@@ -1606,15 +1721,19 @@ int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
 
         /* Init vmlinux path */
         ret = init_vmlinux();
-        if (ret < 0)
+        if (ret < 0) {
+                free(pkgs);
                 return ret;
+        }
 
         /* Loop 1: convert all events */
         for (i = 0; i < npevs; i++) {
                 pkgs[i].pev = &pevs[i];
                 /* Convert with or without debuginfo */
                 ret  = convert_to_probe_trace_events(pkgs[i].pev,
-                                                      &pkgs[i].tevs, max_tevs);
+                                                     &pkgs[i].tevs,
+                                                     max_tevs,
+                                                     module);
                 if (ret < 0)
                         goto end;
                 pkgs[i].ntevs = ret;
@@ -1625,10 +1744,13 @@ int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
                 ret = __add_probe_trace_events(pkgs[i].pev, pkgs[i].tevs,
                                                 pkgs[i].ntevs, force_add);
 end:
-        /* Loop 3: cleanup trace events  */
-        for (i = 0; i < npevs; i++)
+        /* Loop 3: cleanup and free trace events  */
+        for (i = 0; i < npevs; i++) {
                 for (j = 0; j < pkgs[i].ntevs; j++)
                         clear_probe_trace_event(&pkgs[i].tevs[j]);
+                free(pkgs[i].tevs);
+        }
+        free(pkgs);
 
         return ret;
 }
diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h
index 5af39243a25b..5accbedfea37 100644
--- a/tools/perf/util/probe-event.h
+++ b/tools/perf/util/probe-event.h
@@ -90,6 +90,12 @@ struct line_range {
         struct list_head        line_list;      /* Visible lines */
 };
 
+/* List of variables */
+struct variable_list {
+        struct probe_trace_point        point;  /* Actual probepoint */
+        struct strlist                  *vars;  /* Available variables */
+};
+
 /* Command string to events */
 extern int parse_perf_probe_command(const char *cmd,
                                     struct perf_probe_event *pev);
@@ -109,12 +115,18 @@ extern void clear_perf_probe_event(struct perf_probe_event *pev);
 /* Command string to line-range */
 extern int parse_line_range_desc(const char *cmd, struct line_range *lr);
 
+/* Internal use: Return kernel/module path */
+extern const char *kernel_get_module_path(const char *module);
 
 extern int add_perf_probe_events(struct perf_probe_event *pevs, int npevs,
-                                 bool force_add, int max_probe_points);
+                                 int max_probe_points, const char *module,
+                                 bool force_add);
 extern int del_perf_probe_events(struct strlist *dellist);
 extern int show_perf_probe_events(void);
-extern int show_line_range(struct line_range *lr);
+extern int show_line_range(struct line_range *lr, const char *module);
+extern int show_available_vars(struct perf_probe_event *pevs, int npevs,
+                               int max_probe_points, const char *module,
+                               bool externs);
 
 
 /* Maximum index number of event-name postfix */
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index 840f1aabbb74..ddf4d4556321 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -33,7 +33,6 @@
 #include <ctype.h>
 #include <dwarf-regs.h>
 
-#include "string.h"
 #include "event.h"
 #include "debug.h"
 #include "util.h"
@@ -117,6 +116,126 @@ static void line_list__free(struct list_head *head)
         }
 }
 
+/* Dwarf FL wrappers */
+static char *debuginfo_path;    /* Currently dummy */
+
+static const Dwfl_Callbacks offline_callbacks = {
+        .find_debuginfo = dwfl_standard_find_debuginfo,
+        .debuginfo_path = &debuginfo_path,
+
+        .section_address = dwfl_offline_section_address,
+
+        /* We use this table for core files too.  */
+        .find_elf = dwfl_build_id_find_elf,
+};
+
+/* Get a Dwarf from offline image */
+static Dwarf *dwfl_init_offline_dwarf(int fd, Dwfl **dwflp, Dwarf_Addr *bias)
+{
+        Dwfl_Module *mod;
+        Dwarf *dbg = NULL;
+
+        if (!dwflp)
+                return NULL;
+
+        *dwflp = dwfl_begin(&offline_callbacks);
+        if (!*dwflp)
+                return NULL;
+
+        mod = dwfl_report_offline(*dwflp, "", "", fd);
+        if (!mod)
+                goto error;
+
+        dbg = dwfl_module_getdwarf(mod, bias);
+        if (!dbg) {
+error:
+                dwfl_end(*dwflp);
+                *dwflp = NULL;
+        }
+        return dbg;
+}
+
+#if _ELFUTILS_PREREQ(0, 148)
+/* This method is buggy if elfutils is older than 0.148 */
+static int __linux_kernel_find_elf(Dwfl_Module *mod,
+                                   void **userdata,
+                                   const char *module_name,
+                                   Dwarf_Addr base,
+                                   char **file_name, Elf **elfp)
+{
+        int fd;
+        const char *path = kernel_get_module_path(module_name);
+
+        pr_debug2("Use file %s for %s\n", path, module_name);
+        if (path) {
+                fd = open(path, O_RDONLY);
+                if (fd >= 0) {
+                        *file_name = strdup(path);
+                        return fd;
+                }
+        }
+        /* If failed, try to call standard method */
+        return dwfl_linux_kernel_find_elf(mod, userdata, module_name, base,
+                                          file_name, elfp);
+}
+
+static const Dwfl_Callbacks kernel_callbacks = {
+        .find_debuginfo = dwfl_standard_find_debuginfo,
+        .debuginfo_path = &debuginfo_path,
+
+        .find_elf = __linux_kernel_find_elf,
+        .section_address = dwfl_linux_kernel_module_section_address,
+};
+
+/* Get a Dwarf from live kernel image */
+static Dwarf *dwfl_init_live_kernel_dwarf(Dwarf_Addr addr, Dwfl **dwflp,
+                                          Dwarf_Addr *bias)
+{
+        Dwarf *dbg;
+
+        if (!dwflp)
+                return NULL;
+
+        *dwflp = dwfl_begin(&kernel_callbacks);
+        if (!*dwflp)
+                return NULL;
+
+        /* Load the kernel dwarves: Don't care the result here */
+        dwfl_linux_kernel_report_kernel(*dwflp);
+        dwfl_linux_kernel_report_modules(*dwflp);
+
+        dbg = dwfl_addrdwarf(*dwflp, addr, bias);
+        /* Here, check whether we could get a real dwarf */
+        if (!dbg) {
+                pr_debug("Failed to find kernel dwarf at %lx\n",
+                         (unsigned long)addr);
+                dwfl_end(*dwflp);
+                *dwflp = NULL;
+        }
+        return dbg;
+}
+#else
+/* With older elfutils, this just support kernel module... */
+static Dwarf *dwfl_init_live_kernel_dwarf(Dwarf_Addr addr __used, Dwfl **dwflp,
+                                          Dwarf_Addr *bias)
+{
+        int fd;
+        const char *path = kernel_get_module_path("kernel");
+
+        if (!path) {
+                pr_err("Failed to find vmlinux path\n");
+                return NULL;
+        }
+
+        pr_debug2("Use file %s for debuginfo\n", path);
+        fd = open(path, O_RDONLY);
+        if (fd < 0)
+                return NULL;
+
+        return dwfl_init_offline_dwarf(fd, dwflp, bias);
+}
+#endif
+
 /* Dwarf wrappers */
 
 /* Find the realpath of the target file. */
@@ -161,26 +280,44 @@ static bool die_compare_name(Dwarf_Die *dw_die, const char *tname)
         return name ? (strcmp(tname, name) == 0) : false;
 }
 
-/* Get type die, but skip qualifiers and typedef */
-static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+/* Get type die */
+static Dwarf_Die *die_get_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
 {
         Dwarf_Attribute attr;
+
+        if (dwarf_attr_integrate(vr_die, DW_AT_type, &attr) &&
+            dwarf_formref_die(&attr, die_mem))
+                return die_mem;
+        else
+                return NULL;
+}
+
+/* Get a type die, but skip qualifiers */
+static Dwarf_Die *__die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+{
         int tag;
 
         do {
-                if (dwarf_attr(vr_die, DW_AT_type, &attr) == NULL ||
-                    dwarf_formref_die(&attr, die_mem) == NULL)
-                        return NULL;
-
-                tag = dwarf_tag(die_mem);
-                vr_die = die_mem;
+                vr_die = die_get_type(vr_die, die_mem);
+                if (!vr_die)
+                        break;
+                tag = dwarf_tag(vr_die);
         } while (tag == DW_TAG_const_type ||
                  tag == DW_TAG_restrict_type ||
                  tag == DW_TAG_volatile_type ||
-                 tag == DW_TAG_shared_type ||
-                 tag == DW_TAG_typedef);
+                 tag == DW_TAG_shared_type);
 
-        return die_mem;
+        return vr_die;
+}
+
+/* Get a type die, but skip qualifiers and typedef */
+static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem)
+{
+        do {
+                vr_die = __die_get_real_type(vr_die, die_mem);
+        } while (vr_die && dwarf_tag(vr_die) == DW_TAG_typedef);
+
+        return vr_die;
 }
 
 static bool die_is_signed_type(Dwarf_Die *tp_die)
@@ -321,25 +458,35 @@ static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr,
         return die_find_child(sp_die, __die_find_inline_cb, &addr, die_mem);
 }
 
+struct __find_variable_param {
+        const char *name;
+        Dwarf_Addr addr;
+};
+
 static int __die_find_variable_cb(Dwarf_Die *die_mem, void *data)
 {
-        const char *name = data;
+        struct __find_variable_param *fvp = data;
         int tag;
 
         tag = dwarf_tag(die_mem);
         if ((tag == DW_TAG_formal_parameter ||
              tag == DW_TAG_variable) &&
-            die_compare_name(die_mem, name))
+            die_compare_name(die_mem, fvp->name))
                 return DIE_FIND_CB_FOUND;
 
-        return DIE_FIND_CB_CONTINUE;
+        if (dwarf_haspc(die_mem, fvp->addr))
+                return DIE_FIND_CB_CONTINUE;
+        else
+                return DIE_FIND_CB_SIBLING;
 }
 
-/* Find a variable called 'name' */
-static Dwarf_Die *die_find_variable(Dwarf_Die *sp_die, const char *name,
-                                    Dwarf_Die *die_mem)
+/* Find a variable called 'name' at given address */
+static Dwarf_Die *die_find_variable_at(Dwarf_Die *sp_die, const char *name,
+                                       Dwarf_Addr addr, Dwarf_Die *die_mem)
 {
-        return die_find_child(sp_die, __die_find_variable_cb, (void *)name,
+        struct __find_variable_param fvp = { .name = name, .addr = addr};
+
+        return die_find_child(sp_die, __die_find_variable_cb, (void *)&fvp,
                               die_mem);
 }
 
@@ -362,6 +509,60 @@ static Dwarf_Die *die_find_member(Dwarf_Die *st_die, const char *name,
                               die_mem);
 }
 
+/* Get the name of given variable DIE */
+static int die_get_typename(Dwarf_Die *vr_die, char *buf, int len)
+{
+        Dwarf_Die type;
+        int tag, ret, ret2;
+        const char *tmp = "";
+
+        if (__die_get_real_type(vr_die, &type) == NULL)
+                return -ENOENT;
+
+        tag = dwarf_tag(&type);
+        if (tag == DW_TAG_array_type || tag == DW_TAG_pointer_type)
+                tmp = "*";
+        else if (tag == DW_TAG_subroutine_type) {
+                /* Function pointer */
+                ret = snprintf(buf, len, "(function_type)");
+                return (ret >= len) ? -E2BIG : ret;
+        } else {
+                if (!dwarf_diename(&type))
+                        return -ENOENT;
+                if (tag == DW_TAG_union_type)
+                        tmp = "union ";
+                else if (tag == DW_TAG_structure_type)
+                        tmp = "struct ";
+                /* Write a base name */
+                ret = snprintf(buf, len, "%s%s", tmp, dwarf_diename(&type));
+                return (ret >= len) ? -E2BIG : ret;
+        }
+        ret = die_get_typename(&type, buf, len);
+        if (ret > 0) {
+                ret2 = snprintf(buf + ret, len - ret, "%s", tmp);
+                ret = (ret2 >= len - ret) ? -E2BIG : ret2 + ret;
+        }
+        return ret;
+}
+
+/* Get the name and type of given variable DIE, stored as "type\tname" */
+static int die_get_varname(Dwarf_Die *vr_die, char *buf, int len)
+{
+        int ret, ret2;
+
+        ret = die_get_typename(vr_die, buf, len);
+        if (ret < 0) {
+                pr_debug("Failed to get type, make it unknown.\n");
+                ret = snprintf(buf, len, "(unknown_type)");
+        }
+        if (ret > 0) {
+                ret2 = snprintf(buf + ret, len - ret, "\t%s",
+                                dwarf_diename(vr_die));
+                ret = (ret2 >= len - ret) ? -E2BIG : ret2 + ret;
+        }
+        return ret;
+}
+
 /*
  * Probe finder related functions
  */
@@ -375,8 +576,13 @@ static struct probe_trace_arg_ref *alloc_trace_arg_ref(long offs)
         return ref;
 }
 
-/* Show a location */
-static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
+/*
+ * Convert a location into trace_arg.
+ * If tvar == NULL, this just checks variable can be converted.
+ */
+static int convert_variable_location(Dwarf_Die *vr_die, Dwarf_Addr addr,
+                                     Dwarf_Op *fb_ops,
+                                     struct probe_trace_arg *tvar)
 {
         Dwarf_Attribute attr;
         Dwarf_Op *op;
@@ -385,20 +591,23 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
         Dwarf_Word offs = 0;
         bool ref = false;
         const char *regs;
-        struct probe_trace_arg *tvar = pf->tvar;
         int ret;
 
+        if (dwarf_attr(vr_die, DW_AT_external, &attr) != NULL)
+                goto static_var;
+
         /* TODO: handle more than 1 exprs */
         if (dwarf_attr(vr_die, DW_AT_location, &attr) == NULL ||
-            dwarf_getlocation_addr(&attr, pf->addr, &op, &nops, 1) <= 0 ||
+            dwarf_getlocation_addr(&attr, addr, &op, &nops, 1) <= 0 ||
             nops == 0) {
                 /* TODO: Support const_value */
-                pr_err("Failed to find the location of %s at this address.\n"
-                       " Perhaps, it has been optimized out.\n", pf->pvar->var);
                 return -ENOENT;
         }
 
         if (op->atom == DW_OP_addr) {
+static_var:
+                if (!tvar)
+                        return 0;
                 /* Static variables on memory (not stack), make @varname */
                 ret = strlen(dwarf_diename(vr_die));
                 tvar->value = zalloc(ret + 2);
@@ -413,14 +622,11 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
 
         /* If this is based on frame buffer, set the offset */
         if (op->atom == DW_OP_fbreg) {
-                if (pf->fb_ops == NULL) {
-                        pr_warning("The attribute of frame base is not "
-                                   "supported.\n");
+                if (fb_ops == NULL)
                         return -ENOTSUP;
-                }
                 ref = true;
                 offs = op->number;
-                op = &pf->fb_ops[0];
+                op = &fb_ops[0];
         }
 
         if (op->atom >= DW_OP_breg0 && op->atom <= DW_OP_breg31) {
@@ -436,13 +642,18 @@ static int convert_variable_location(Dwarf_Die *vr_die, struct probe_finder *pf)
         } else if (op->atom == DW_OP_regx) {
                 regn = op->number;
         } else {
-                pr_warning("DW_OP %x is not supported.\n", op->atom);
+                pr_debug("DW_OP %x is not supported.\n", op->atom);
                 return -ENOTSUP;
         }
 
+        if (!tvar)
+                return 0;
+
         regs = get_arch_regstr(regn);
         if (!regs) {
-                pr_warning("Mapping for DWARF register number %u missing on this architecture.", regn);
+                /* This should be a bug in DWARF or this tool */
+                pr_warning("Mapping for DWARF register number %u "
+                           "missing on this architecture.", regn);
                 return -ERANGE;
         }
 
@@ -667,8 +878,14 @@ static int convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf)
         pr_debug("Converting variable %s into trace event.\n",
                  dwarf_diename(vr_die));
 
-        ret = convert_variable_location(vr_die, pf);
-        if (ret == 0 && pf->pvar->field) {
+        ret = convert_variable_location(vr_die, pf->addr, pf->fb_ops,
+                                        pf->tvar);
+        if (ret == -ENOENT)
+                pr_err("Failed to find the location of %s at this address.\n"
+                       " Perhaps, it has been optimized out.\n", pf->pvar->var);
+        else if (ret == -ENOTSUP)
+                pr_err("Sorry, we don't support this variable location yet.\n");
+        else if (pf->pvar->field) {
                 ret = convert_variable_fields(vr_die, pf->pvar->var,
                                               pf->pvar->field, &pf->tvar->ref,
                                               &die_mem);
@@ -687,6 +904,25 @@ static int find_variable(Dwarf_Die *sp_die, struct probe_finder *pf)
         char buf[32], *ptr;
         int ret, nscopes;
 
+        if (!is_c_varname(pf->pvar->var)) {
+                /* Copy raw parameters */
+                pf->tvar->value = strdup(pf->pvar->var);
+                if (pf->tvar->value == NULL)
+                        return -ENOMEM;
+                if (pf->pvar->type) {
+                        pf->tvar->type = strdup(pf->pvar->type);
+                        if (pf->tvar->type == NULL)
+                                return -ENOMEM;
+                }
+                if (pf->pvar->name) {
+                        pf->tvar->name = strdup(pf->pvar->name);
+                        if (pf->tvar->name == NULL)
+                                return -ENOMEM;
+                } else
+                        pf->tvar->name = NULL;
+                return 0;
+        }
+
         if (pf->pvar->name)
                 pf->tvar->name = strdup(pf->pvar->name);
         else {
@@ -701,68 +937,42 @@ static int find_variable(Dwarf_Die *sp_die, struct probe_finder *pf)
         if (pf->tvar->name == NULL)
                 return -ENOMEM;
 
-        if (!is_c_varname(pf->pvar->var)) {
-                /* Copy raw parameters */
-                pf->tvar->value = strdup(pf->pvar->var);
-                if (pf->tvar->value == NULL)
-                        return -ENOMEM;
-                else
-                        return 0;
-        }
-
         pr_debug("Searching '%s' variable in context.\n",
                  pf->pvar->var);
         /* Search child die for local variables and parameters. */
-        if (die_find_variable(sp_die, pf->pvar->var, &vr_die))
+        if (die_find_variable_at(sp_die, pf->pvar->var, pf->addr, &vr_die))
                 ret = convert_variable(&vr_die, pf);
         else {
                 /* Search upper class */
                 nscopes = dwarf_getscopes_die(sp_die, &scopes);
-                if (nscopes > 0) {
-                        ret = dwarf_getscopevar(scopes, nscopes, pf->pvar->var,
-                                                0, NULL, 0, 0, &vr_die);
-                        if (ret >= 0)
+                while (nscopes-- > 1) {
+                        pr_debug("Searching variables in %s\n",
+                                 dwarf_diename(&scopes[nscopes]));
+                        /* We should check this scope, so give dummy address */
+                        if (die_find_variable_at(&scopes[nscopes],
+                                                 pf->pvar->var, 0,
+                                                 &vr_die)) {
                                 ret = convert_variable(&vr_die, pf);
-                        else
-                                ret = -ENOENT;
+                                goto found;
+                        }
+                }
+                if (scopes)
                         free(scopes);
-                } else
-                        ret = -ENOENT;
+                ret = -ENOENT;
         }
+found:
         if (ret < 0)
                 pr_warning("Failed to find '%s' in this function.\n",
                            pf->pvar->var);
         return ret;
 }
 
-/* Show a probe point to output buffer */
-static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
+/* Convert subprogram DIE to trace point */
+static int convert_to_trace_point(Dwarf_Die *sp_die, Dwarf_Addr paddr,
+                                  bool retprobe, struct probe_trace_point *tp)
 {
-        struct probe_trace_event *tev;
         Dwarf_Addr eaddr;
-        Dwarf_Die die_mem;
         const char *name;
-        int ret, i;
-        Dwarf_Attribute fb_attr;
-        size_t nops;
-
-        if (pf->ntevs == pf->max_tevs) {
-                pr_warning("Too many( > %d) probe point found.\n",
-                           pf->max_tevs);
-                return -ERANGE;
-        }
-        tev = &pf->tevs[pf->ntevs++];
-
-        /* If no real subprogram, find a real one */
-        if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) {
-                sp_die = die_find_real_subprogram(&pf->cu_die,
-                                                 pf->addr, &die_mem);
-                if (!sp_die) {
-                        pr_warning("Failed to find probe point in any "
-                                   "functions.\n");
-                        return -ENOENT;
-                }
-        }
 
         /* Copy the name of probe point */
         name = dwarf_diename(sp_die);
@@ -772,16 +982,45 @@ static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
                                    dwarf_diename(sp_die));
                         return -ENOENT;
                 }
-                tev->point.symbol = strdup(name);
-                if (tev->point.symbol == NULL)
+                tp->symbol = strdup(name);
+                if (tp->symbol == NULL)
                         return -ENOMEM;
-                tev->point.offset = (unsigned long)(pf->addr - eaddr);
+                tp->offset = (unsigned long)(paddr - eaddr);
         } else
                 /* This function has no name. */
-                tev->point.offset = (unsigned long)pf->addr;
+                tp->offset = (unsigned long)paddr;
 
-        pr_debug("Probe point found: %s+%lu\n", tev->point.symbol,
-                 tev->point.offset);
+        /* Return probe must be on the head of a subprogram */
+        if (retprobe) {
+                if (eaddr != paddr) {
+                        pr_warning("Return probe must be on the head of"
+                                   " a real function\n");
+                        return -EINVAL;
+                }
+                tp->retprobe = true;
+        }
+
+        return 0;
+}
+
+/* Call probe_finder callback with real subprogram DIE */
+static int call_probe_finder(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+        Dwarf_Die die_mem;
+        Dwarf_Attribute fb_attr;
+        size_t nops;
+        int ret;
+
+        /* If no real subprogram, find a real one */
+        if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) {
+                sp_die = die_find_real_subprogram(&pf->cu_die,
+                                                  pf->addr, &die_mem);
+                if (!sp_die) {
+                        pr_warning("Failed to find probe point in any "
+                                   "functions.\n");
+                        return -ENOENT;
+                }
+        }
 
         /* Get the frame base attribute/ops */
         dwarf_attr(sp_die, DW_AT_frame_base, &fb_attr);
@@ -801,22 +1040,13 @@ static int convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf)
 #endif
         }
 
-        /* Find each argument */
-        tev->nargs = pf->pev->nargs;
-        tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
-        if (tev->args == NULL)
-                return -ENOMEM;
-        for (i = 0; i < pf->pev->nargs; i++) {
-                pf->pvar = &pf->pev->args[i];
-                pf->tvar = &tev->args[i];
-                ret = find_variable(sp_die, pf);
-                if (ret != 0)
-                        return ret;
-        }
+        /* Call finder's callback handler */
+        ret = pf->callback(sp_die, pf);
 
         /* *pf->fb_ops will be cached in libdw. Don't free it. */
         pf->fb_ops = NULL;
-        return 0;
+
+        return ret;
 }
 
 /* Find probe point from its line number */
@@ -852,7 +1082,7 @@ static int find_probe_point_by_line(struct probe_finder *pf)
                          (int)i, lineno, (uintmax_t)addr);
                 pf->addr = addr;
 
-                ret = convert_probe_point(NULL, pf);
+                ret = call_probe_finder(NULL, pf);
                 /* Continuing, because target line might be inlined. */
         }
         return ret;
@@ -965,7 +1195,7 @@ static int find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf)
                          (int)i, lineno, (unsigned long long)addr);
                 pf->addr = addr;
 
-                ret = convert_probe_point(sp_die, pf);
+                ret = call_probe_finder(sp_die, pf);
                 /* Continuing, because target line might be inlined. */
         }
         /* TODO: deallocate lines, but how? */
@@ -1000,7 +1230,7 @@ static int probe_point_inline_cb(Dwarf_Die *in_die, void *data)
                 pr_debug("found inline addr: 0x%jx\n",
                          (uintmax_t)pf->addr);
 
-                param->retval = convert_probe_point(in_die, pf);
+                param->retval = call_probe_finder(in_die, pf);
                 if (param->retval < 0)
                         return DWARF_CB_ABORT;
         }
@@ -1038,7 +1268,7 @@ static int probe_point_search_cb(Dwarf_Die *sp_die, void *data)
                         }
                         pf->addr += pp->offset;
                         /* TODO: Check the address in this function */
-                        param->retval = convert_probe_point(sp_die, pf);
+                        param->retval = call_probe_finder(sp_die, pf);
                 }
         } else {
                 struct dwarf_callback_param _param = {.data = (void *)pf,
@@ -1060,90 +1290,276 @@ static int find_probe_point_by_func(struct probe_finder *pf)
         return _param.retval;
 }
 
-/* Find probe_trace_events specified by perf_probe_event from debuginfo */
-int find_probe_trace_events(int fd, struct perf_probe_event *pev,
-                             struct probe_trace_event **tevs, int max_tevs)
+/* Find probe points from debuginfo */
+static int find_probes(int fd, struct probe_finder *pf)
 {
-        struct probe_finder pf = {.pev = pev, .max_tevs = max_tevs};
-        struct perf_probe_point *pp = &pev->point;
+        struct perf_probe_point *pp = &pf->pev->point;
         Dwarf_Off off, noff;
         size_t cuhl;
         Dwarf_Die *diep;
-        Dwarf *dbg;
+        Dwarf *dbg = NULL;
+        Dwfl *dwfl;
+        Dwarf_Addr bias;        /* Currently ignored */
         int ret = 0;
 
-        pf.tevs = zalloc(sizeof(struct probe_trace_event) * max_tevs);
-        if (pf.tevs == NULL)
-                return -ENOMEM;
-        *tevs = pf.tevs;
-        pf.ntevs = 0;
-
-        dbg = dwarf_begin(fd, DWARF_C_READ);
+        dbg = dwfl_init_offline_dwarf(fd, &dwfl, &bias);
         if (!dbg) {
                 pr_warning("No dwarf info found in the vmlinux - "
                         "please rebuild with CONFIG_DEBUG_INFO=y.\n");
-                free(pf.tevs);
-                *tevs = NULL;
                 return -EBADF;
         }
 
 #if _ELFUTILS_PREREQ(0, 142)
         /* Get the call frame information from this dwarf */
-        pf.cfi = dwarf_getcfi(dbg);
+        pf->cfi = dwarf_getcfi(dbg);
 #endif
 
         off = 0;
-        line_list__init(&pf.lcache);
+        line_list__init(&pf->lcache);
         /* Loop on CUs (Compilation Unit) */
         while (!dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL) &&
                ret >= 0) {
                 /* Get the DIE(Debugging Information Entry) of this CU */
-                diep = dwarf_offdie(dbg, off + cuhl, &pf.cu_die);
+                diep = dwarf_offdie(dbg, off + cuhl, &pf->cu_die);
                 if (!diep)
                         continue;
 
                 /* Check if target file is included. */
                 if (pp->file)
-                        pf.fname = cu_find_realpath(&pf.cu_die, pp->file);
+                        pf->fname = cu_find_realpath(&pf->cu_die, pp->file);
                 else
-                        pf.fname = NULL;
+                        pf->fname = NULL;
 
-                if (!pp->file || pf.fname) {
+                if (!pp->file || pf->fname) {
                         if (pp->function)
-                                ret = find_probe_point_by_func(&pf);
+                                ret = find_probe_point_by_func(pf);
                         else if (pp->lazy_line)
-                                ret = find_probe_point_lazy(NULL, &pf);
+                                ret = find_probe_point_lazy(NULL, pf);
                         else {
-                                pf.lno = pp->line;
-                                ret = find_probe_point_by_line(&pf);
+                                pf->lno = pp->line;
+                                ret = find_probe_point_by_line(pf);
                         }
                 }
                 off = noff;
         }
-        line_list__free(&pf.lcache);
-        dwarf_end(dbg);
+        line_list__free(&pf->lcache);
+        if (dwfl)
+                dwfl_end(dwfl);
+
+        return ret;
+}
+
+/* Add a found probe point into trace event list */
+static int add_probe_trace_event(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+        struct trace_event_finder *tf =
+                        container_of(pf, struct trace_event_finder, pf);
+        struct probe_trace_event *tev;
+        int ret, i;
+
+        /* Check number of tevs */
+        if (tf->ntevs == tf->max_tevs) {
+                pr_warning("Too many( > %d) probe point found.\n",
+                           tf->max_tevs);
+                return -ERANGE;
+        }
+        tev = &tf->tevs[tf->ntevs++];
+
+        ret = convert_to_trace_point(sp_die, pf->addr, pf->pev->point.retprobe,
+                                     &tev->point);
+        if (ret < 0)
+                return ret;
+
+        pr_debug("Probe point found: %s+%lu\n", tev->point.symbol,
+                 tev->point.offset);
+
+        /* Find each argument */
+        tev->nargs = pf->pev->nargs;
+        tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs);
+        if (tev->args == NULL)
+                return -ENOMEM;
+        for (i = 0; i < pf->pev->nargs; i++) {
+                pf->pvar = &pf->pev->args[i];
+                pf->tvar = &tev->args[i];
+                ret = find_variable(sp_die, pf);
+                if (ret != 0)
+                        return ret;
+        }
 
-        return (ret < 0) ? ret : pf.ntevs;
+        return 0;
+}
+
+/* Find probe_trace_events specified by perf_probe_event from debuginfo */
+int find_probe_trace_events(int fd, struct perf_probe_event *pev,
+                            struct probe_trace_event **tevs, int max_tevs)
+{
+        struct trace_event_finder tf = {
+                        .pf = {.pev = pev, .callback = add_probe_trace_event},
+                        .max_tevs = max_tevs};
+        int ret;
+
+        /* Allocate result tevs array */
+        *tevs = zalloc(sizeof(struct probe_trace_event) * max_tevs);
+        if (*tevs == NULL)
+                return -ENOMEM;
+
+        tf.tevs = *tevs;
+        tf.ntevs = 0;
+
+        ret = find_probes(fd, &tf.pf);
+        if (ret < 0) {
+                free(*tevs);
+                *tevs = NULL;
+                return ret;
+        }
+
+        return (ret < 0) ? ret : tf.ntevs;
+}
+
+#define MAX_VAR_LEN 64
+
+/* Collect available variables in this scope */
+static int collect_variables_cb(Dwarf_Die *die_mem, void *data)
+{
+        struct available_var_finder *af = data;
+        struct variable_list *vl;
+        char buf[MAX_VAR_LEN];
+        int tag, ret;
+
+        vl = &af->vls[af->nvls - 1];
+
+        tag = dwarf_tag(die_mem);
+        if (tag == DW_TAG_formal_parameter ||
+            tag == DW_TAG_variable) {
+                ret = convert_variable_location(die_mem, af->pf.addr,
+                                                af->pf.fb_ops, NULL);
+                if (ret == 0) {
+                        ret = die_get_varname(die_mem, buf, MAX_VAR_LEN);
+                        pr_debug2("Add new var: %s\n", buf);
+                        if (ret > 0)
+                                strlist__add(vl->vars, buf);
+                }
+        }
+
+        if (af->child && dwarf_haspc(die_mem, af->pf.addr))
+                return DIE_FIND_CB_CONTINUE;
+        else
+                return DIE_FIND_CB_SIBLING;
+}
+
+/* Add a found vars into available variables list */
+static int add_available_vars(Dwarf_Die *sp_die, struct probe_finder *pf)
+{
+        struct available_var_finder *af =
+                        container_of(pf, struct available_var_finder, pf);
+        struct variable_list *vl;
+        Dwarf_Die die_mem, *scopes = NULL;
+        int ret, nscopes;
+
+        /* Check number of tevs */
+        if (af->nvls == af->max_vls) {
+                pr_warning("Too many( > %d) probe point found.\n", af->max_vls);
+                return -ERANGE;
+        }
+        vl = &af->vls[af->nvls++];
+
+        ret = convert_to_trace_point(sp_die, pf->addr, pf->pev->point.retprobe,
+                                     &vl->point);
+        if (ret < 0)
+                return ret;
+
+        pr_debug("Probe point found: %s+%lu\n", vl->point.symbol,
+                 vl->point.offset);
+
+        /* Find local variables */
+        vl->vars = strlist__new(true, NULL);
+        if (vl->vars == NULL)
+                return -ENOMEM;
+        af->child = true;
+        die_find_child(sp_die, collect_variables_cb, (void *)af, &die_mem);
+
+        /* Find external variables */
+        if (!af->externs)
+                goto out;
+        /* Don't need to search child DIE for externs. */
+        af->child = false;
+        nscopes = dwarf_getscopes_die(sp_die, &scopes);
+        while (nscopes-- > 1)
+                die_find_child(&scopes[nscopes], collect_variables_cb,
+                               (void *)af, &die_mem);
+        if (scopes)
+                free(scopes);
+
+out:
+        if (strlist__empty(vl->vars)) {
+                strlist__delete(vl->vars);
+                vl->vars = NULL;
+        }
+
+        return ret;
+}
+
+/* Find available variables at given probe point */
+int find_available_vars_at(int fd, struct perf_probe_event *pev,
+                           struct variable_list **vls, int max_vls,
+                           bool externs)
+{
+        struct available_var_finder af = {
+                        .pf = {.pev = pev, .callback = add_available_vars},
+                        .max_vls = max_vls, .externs = externs};
+        int ret;
+
+        /* Allocate result vls array */
+        *vls = zalloc(sizeof(struct variable_list) * max_vls);
+        if (*vls == NULL)
+                return -ENOMEM;
+
+        af.vls = *vls;
+        af.nvls = 0;
+
+        ret = find_probes(fd, &af.pf);
+        if (ret < 0) {
+                /* Free vlist for error */
+                while (af.nvls--) {
+                        if (af.vls[af.nvls].point.symbol)
+                                free(af.vls[af.nvls].point.symbol);
+                        if (af.vls[af.nvls].vars)
+                                strlist__delete(af.vls[af.nvls].vars);
+                }
+                free(af.vls);
+                *vls = NULL;
+                return ret;
+        }
+
+        return (ret < 0) ? ret : af.nvls;
 }
 
 /* Reverse search */
-int find_perf_probe_point(int fd, unsigned long addr,
-                          struct perf_probe_point *ppt)
+int find_perf_probe_point(unsigned long addr, struct perf_probe_point *ppt)
 {
         Dwarf_Die cudie, spdie, indie;
-        Dwarf *dbg;
+        Dwarf *dbg = NULL;
+        Dwfl *dwfl = NULL;
         Dwarf_Line *line;
-        Dwarf_Addr laddr, eaddr;
+        Dwarf_Addr laddr, eaddr, bias = 0;
         const char *tmp;
         int lineno, ret = 0;
         bool found = false;
 
-        dbg = dwarf_begin(fd, DWARF_C_READ);
-        if (!dbg)
-                return -EBADF;
+        /* Open the live linux kernel */
+        dbg = dwfl_init_live_kernel_dwarf(addr, &dwfl, &bias);
+        if (!dbg) {
+                pr_warning("No dwarf info found in the vmlinux - "
+                        "please rebuild with CONFIG_DEBUG_INFO=y.\n");
+                ret = -EINVAL;
+                goto end;
+        }
 
+        /* Adjust address with bias */
+        addr += bias;
         /* Find cu die */
-        if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr, &cudie)) {
+        if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr - bias, &cudie)) {
+                pr_warning("No CU DIE is found at %lx\n", addr);
                 ret = -EINVAL;
                 goto end;
         }
@@ -1206,7 +1622,8 @@ found:
         }
 
 end:
-        dwarf_end(dbg);
+        if (dwfl)
+                dwfl_end(dwfl);
         if (ret >= 0)
                 ret = found ? 1 : 0;
         return ret;
@@ -1339,6 +1756,9 @@ static int line_range_search_cb(Dwarf_Die *sp_die, void *data)
         struct line_finder *lf = param->data;
         struct line_range *lr = lf->lr;
 
+        pr_debug("find (%llx) %s\n",
+                 (unsigned long long)dwarf_dieoffset(sp_die),
+                 dwarf_diename(sp_die));
         if (dwarf_tag(sp_die) == DW_TAG_subprogram &&
             die_compare_name(sp_die, lr->function)) {
                 lf->fname = dwarf_decl_file(sp_die);
@@ -1382,10 +1802,12 @@ int find_line_range(int fd, struct line_range *lr)
         Dwarf_Off off = 0, noff;
         size_t cuhl;
         Dwarf_Die *diep;
-        Dwarf *dbg;
+        Dwarf *dbg = NULL;
+        Dwfl *dwfl;
+        Dwarf_Addr bias;        /* Currently ignored */
         const char *comp_dir;
 
-        dbg = dwarf_begin(fd, DWARF_C_READ);
+        dbg = dwfl_init_offline_dwarf(fd, &dwfl, &bias);
         if (!dbg) {
                 pr_warning("No dwarf info found in the vmlinux - "
                         "please rebuild with CONFIG_DEBUG_INFO=y.\n");
@@ -1431,8 +1853,7 @@ int find_line_range(int fd, struct line_range *lr)
         }
 
         pr_debug("path: %s\n", lr->path);
-        dwarf_end(dbg);
-
+        dwfl_end(dwfl);
         return (ret < 0) ? ret : lf.found;
 }
 
diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h
index 4507d519f183..bba69d455699 100644
--- a/tools/perf/util/probe-finder.h
+++ b/tools/perf/util/probe-finder.h
@@ -22,20 +22,27 @@ extern int find_probe_trace_events(int fd, struct perf_probe_event *pev,
                                     int max_tevs);
 
 /* Find a perf_probe_point from debuginfo */
-extern int find_perf_probe_point(int fd, unsigned long addr,
+extern int find_perf_probe_point(unsigned long addr,
                                  struct perf_probe_point *ppt);
 
+/* Find a line range */
 extern int find_line_range(int fd, struct line_range *lr);
 
+/* Find available variables */
+extern int find_available_vars_at(int fd, struct perf_probe_event *pev,
+                                  struct variable_list **vls, int max_points,
+                                  bool externs);
+
 #include <dwarf.h>
 #include <libdw.h>
+#include <libdwfl.h>
 #include <version.h>
 
 struct probe_finder {
         struct perf_probe_event *pev;           /* Target probe event */
-        struct probe_trace_event *tevs;         /* Result trace events */
-        int                     ntevs;          /* Number of trace events */
-        int                     max_tevs;       /* Max number of trace events */
+
+        /* Callback when a probe point is found */
+        int (*callback)(Dwarf_Die *sp_die, struct probe_finder *pf);
 
         /* For function searching */
         int                     lno;            /* Line number */
@@ -53,6 +60,22 @@ struct probe_finder {
         struct probe_trace_arg  *tvar;          /* Current result variable */
 };
 
+struct trace_event_finder {
+        struct probe_finder     pf;
+        struct probe_trace_event *tevs;         /* Found trace events */
+        int                     ntevs;          /* Number of trace events */
+        int                     max_tevs;       /* Max number of trace events */
+};
+
+struct available_var_finder {
+        struct probe_finder     pf;
+        struct variable_list    *vls;           /* Found variable lists */
+        int                     nvls;           /* Number of variable lists */
+        int                     max_vls;        /* Max no. of variable lists */
+        bool                    externs;        /* Find external vars too */
+        bool                    child;          /* Search child scopes */
+};
+
 struct line_finder {
         struct line_range       *lr;            /* Target line range */
 
diff --git a/tools/perf/util/pstack.h b/tools/perf/util/pstack.h
index 5ad07023504b..4cedea59f518 100644
--- a/tools/perf/util/pstack.h
+++ b/tools/perf/util/pstack.h
@@ -1,6 +1,8 @@
 #ifndef _PERF_PSTACK_
 #define _PERF_PSTACK_
 
+#include <stdbool.h>
+
 struct pstack;
 struct pstack *pstack__new(unsigned short max_nr_entries);
 void pstack__delete(struct pstack *self);
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 1c61a4f4aa8a..b62a553cc67d 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -196,7 +196,8 @@ static int hist_entry__sym_snprintf(struct hist_entry *self, char *bf,
 
         if (verbose) {
                 char o = self->ms.map ? dso__symtab_origin(self->ms.map->dso) : '!';
-                ret += repsep_snprintf(bf, size, "%#018llx %c ", self->ip, o);
+                ret += repsep_snprintf(bf, size, "%*Lx %c ",
+                                       BITS_PER_LONG / 4, self->ip, o);
         }
 
         ret += repsep_snprintf(bf + ret, size - ret, "[%c] ", self->level);
@@ -204,7 +205,8 @@ static int hist_entry__sym_snprintf(struct hist_entry *self, char *bf,
                 ret += repsep_snprintf(bf + ret, size - ret, "%s",
                                        self->ms.sym->name);
         else
-                ret += repsep_snprintf(bf + ret, size - ret, "%#016llx", self->ip);
+                ret += repsep_snprintf(bf + ret, size - ret, "%*Lx",
+                                       BITS_PER_LONG / 4, self->ip);
 
         return ret;
 }
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index 46e531d09e8b..0b91053a7d11 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -70,7 +70,7 @@ struct hist_entry {
                 struct hist_entry *pair;
                 struct rb_root    sorted_chain;
         };
-        struct callchain_node   callchain[0];
+        struct callchain_root   callchain[0];
 };
 
 enum sort_type {
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index 0409fc7c0058..8fc0bd3a3a4a 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -259,7 +259,7 @@ static bool __match_glob(const char *str, const char *pat, bool ignore_space)
                 if (!*pat)      /* Tail wild card matches all */
                         return true;
                 while (*str)
-                        if (strglobmatch(str++, pat))
+                        if (__match_glob(str++, pat, ignore_space))
                                 return true;
         }
         return !*str && !*pat;
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 6f0dd90c36ce..439ab947daf4 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -131,7 +131,8 @@ static void map_groups__fixup_end(struct map_groups *self)
                 __map_groups__fixup_end(self, i);
 }
 
-static struct symbol *symbol__new(u64 start, u64 len, const char *name)
+static struct symbol *symbol__new(u64 start, u64 len, u8 binding,
+                                  const char *name)
 {
         size_t namelen = strlen(name) + 1;
         struct symbol *self = calloc(1, (symbol_conf.priv_size +
@@ -144,6 +145,7 @@ static struct symbol *symbol__new(u64 start, u64 len, const char *name)
 
         self->start   = start;
         self->end     = len ? start + len - 1 : start;
+        self->binding = binding;
         self->namelen = namelen - 1;
 
         pr_debug4("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end);
@@ -160,8 +162,11 @@ void symbol__delete(struct symbol *self)
 
 static size_t symbol__fprintf(struct symbol *self, FILE *fp)
 {
-        return fprintf(fp, " %llx-%llx %s\n",
-                       self->start, self->end, self->name);
+        return fprintf(fp, " %llx-%llx %c %s\n",
+                       self->start, self->end,
+                       self->binding == STB_GLOBAL ? 'g' :
+                       self->binding == STB_LOCAL  ? 'l' : 'w',
+                       self->name);
 }
 
 void dso__set_long_name(struct dso *self, char *name)
@@ -290,7 +295,9 @@ static void symbols__insert_by_name(struct rb_root *self, struct symbol *sym)
 {
         struct rb_node **p = &self->rb_node;
         struct rb_node *parent = NULL;
-        struct symbol_name_rb_node *symn = ((void *)sym) - sizeof(*parent), *s;
+        struct symbol_name_rb_node *symn, *s;
+
+        symn = container_of(sym, struct symbol_name_rb_node, sym);
 
         while (*p != NULL) {
                 parent = *p;
@@ -383,6 +390,20 @@ size_t dso__fprintf_buildid(struct dso *self, FILE *fp)
         return fprintf(fp, "%s", sbuild_id);
 }
 
+size_t dso__fprintf_symbols_by_name(struct dso *self, enum map_type type, FILE *fp)
+{
+        size_t ret = 0;
+        struct rb_node *nd;
+        struct symbol_name_rb_node *pos;
+
+        for (nd = rb_first(&self->symbol_names[type]); nd; nd = rb_next(nd)) {
+                pos = rb_entry(nd, struct symbol_name_rb_node, rb_node);
+                fprintf(fp, "%s\n", pos->sym.name);
+        }
+
+        return ret;
+}
+
 size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp)
 {
         struct rb_node *nd;
@@ -453,6 +474,14 @@ struct process_kallsyms_args {
         struct dso *dso;
 };
 
+static u8 kallsyms2elf_type(char type)
+{
+        if (type == 'W')
+                return STB_WEAK;
+
+        return isupper(type) ? STB_GLOBAL : STB_LOCAL;
+}
+
 static int map__process_kallsym_symbol(void *arg, const char *name,
                                        char type, u64 start)
 {
@@ -466,7 +495,7 @@ static int map__process_kallsym_symbol(void *arg, const char *name,
         /*
          * Will fix up the end later, when we have all symbols sorted.
          */
-        sym = symbol__new(start, 0, name);
+        sym = symbol__new(start, 0, kallsyms2elf_type(type), name);
 
         if (sym == NULL)
                 return -ENOMEM;
@@ -503,7 +532,7 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
         struct machine *machine = kmaps->machine;
         struct map *curr_map = map;
         struct symbol *pos;
-        int count = 0;
+        int count = 0, moved = 0;       
         struct rb_root *root = &self->symbols[map->type];
         struct rb_node *next = rb_first(root);
         int kernel_range = 0;
@@ -561,6 +590,11 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
                         char dso_name[PATH_MAX];
                         struct dso *dso;
 
+                        if (count == 0) {
+                                curr_map = map;
+                                goto filter_symbol;
+                        }
+
                         if (self->kernel == DSO_TYPE_GUEST_KERNEL)
                                 snprintf(dso_name, sizeof(dso_name),
                                         "[guest.kernel].%d",
@@ -586,7 +620,7 @@ static int dso__split_kallsyms(struct dso *self, struct map *map,
                         map_groups__insert(kmaps, curr_map);
                         ++kernel_range;
                 }
-
+filter_symbol:
                 if (filter && filter(curr_map, pos)) {
 discard_symbol:         rb_erase(&pos->rb_node, root);
                         symbol__delete(pos);
@@ -594,8 +628,9 @@ discard_symbol:		rb_erase(&pos->rb_node, root);
                         if (curr_map != map) {
                                 rb_erase(&pos->rb_node, root);
                                 symbols__insert(&curr_map->dso->symbols[curr_map->type], pos);
-                        }
-                        count++;
+                                ++moved;
+                        } else
+                                ++count;
                 }
         }
 
@@ -605,7 +640,7 @@ discard_symbol:		rb_erase(&pos->rb_node, root);
                 dso__set_loaded(curr_map->dso, curr_map->type);
         }
 
-        return count;
+        return count + moved;
 }
 
 int dso__load_kallsyms(struct dso *self, const char *filename,
@@ -661,7 +696,7 @@ static int dso__load_perf_map(struct dso *self, struct map *map,
                 if (len + 2 >= line_len)
                         continue;
 
-                sym = symbol__new(start, size, line + len);
+                sym = symbol__new(start, size, STB_GLOBAL, line + len);
 
                 if (sym == NULL)
                         goto out_delete_line;
@@ -873,7 +908,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, struct map *map,
                                  "%s@plt", elf_sym__name(&sym, symstrs));
 
                         f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-                                        sympltname);
+                                        STB_GLOBAL, sympltname);
                         if (!f)
                                 goto out_elf_end;
 
@@ -895,7 +930,7 @@ static int dso__synthesize_plt_symbols(struct  dso *self, struct map *map,
                                  "%s@plt", elf_sym__name(&sym, symstrs));
 
                         f = symbol__new(plt_offset, shdr_plt.sh_entsize,
-                                        sympltname);
+                                        STB_GLOBAL, sympltname);
                         if (!f)
                                 goto out_elf_end;
 
@@ -1066,6 +1101,16 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
                 if (!is_label && !elf_sym__is_a(&sym, map->type))
                         continue;
 
+                /* Reject ARM ELF "mapping symbols": these aren't unique and
+                 * don't identify functions, so will confuse the profile
+                 * output: */
+                if (ehdr.e_machine == EM_ARM) {
+                        if (!strcmp(elf_name, "$a") ||
+                            !strcmp(elf_name, "$d") ||
+                            !strcmp(elf_name, "$t"))
+                                continue;
+                }
+
                 if (opdsec && sym.st_shndx == opdidx) {
                         u32 offset = sym.st_value - opdshdr.sh_addr;
                         u64 *opd = opddata->d_buf + offset;
@@ -1146,7 +1191,8 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
                 if (demangled != NULL)
                         elf_name = demangled;
 new_symbol:
-                f = symbol__new(sym.st_value, sym.st_size, elf_name);
+                f = symbol__new(sym.st_value, sym.st_size,
+                                GELF_ST_BIND(sym.st_info), elf_name);
                 free(demangled);
                 if (!f)
                         goto out_elf_end;
@@ -1734,8 +1780,8 @@ out_failure:
         return -1;
 }
 
-static int dso__load_vmlinux(struct dso *self, struct map *map,
-                             const char *vmlinux, symbol_filter_t filter)
+int dso__load_vmlinux(struct dso *self, struct map *map,
+                      const char *vmlinux, symbol_filter_t filter)
 {
         int err = -1, fd;
 
@@ -2085,14 +2131,55 @@ static struct dso *machine__create_kernel(struct machine *self)
         return kernel;
 }
 
+struct process_args {
+        u64 start;
+};
+
+static int symbol__in_kernel(void *arg, const char *name,
+                             char type __used, u64 start)
+{
+        struct process_args *args = arg;
+
+        if (strchr(name, '['))
+                return 0;
+
+        args->start = start;
+        return 1;
+}
+
+/* Figure out the start address of kernel map from /proc/kallsyms */
+static u64 machine__get_kernel_start_addr(struct machine *machine)
+{
+        const char *filename;
+        char path[PATH_MAX];
+        struct process_args args;
+
+        if (machine__is_host(machine)) {
+                filename = "/proc/kallsyms";
+        } else {
+                if (machine__is_default_guest(machine))
+                        filename = (char *)symbol_conf.default_guest_kallsyms;
+                else {
+                        sprintf(path, "%s/proc/kallsyms", machine->root_dir);
+                        filename = path;
+                }
+        }
+
+        if (kallsyms__parse(filename, &args, symbol__in_kernel) <= 0)
+                return 0;
+
+        return args.start;
+}
+
 int __machine__create_kernel_maps(struct machine *self, struct dso *kernel)
 {
         enum map_type type;
+        u64 start = machine__get_kernel_start_addr(self);
 
         for (type = 0; type < MAP__NR_TYPES; ++type) {
                 struct kmap *kmap;
 
-                self->vmlinux_maps[type] = map__new2(0, kernel, type);
+                self->vmlinux_maps[type] = map__new2(start, kernel, type);
                 if (self->vmlinux_maps[type] == NULL)
                         return -1;
 
@@ -2244,6 +2331,9 @@ static int setup_list(struct strlist **list, const char *list_str,
 
 int symbol__init(void)
 {
+        if (symbol_conf.initialized)
+                return 0;
+
         elf_version(EV_CURRENT);
         if (symbol_conf.sort_by_name)
                 symbol_conf.priv_size += (sizeof(struct symbol_name_rb_node) -
@@ -2269,6 +2359,7 @@ int symbol__init(void)
                        symbol_conf.sym_list_str, "symbol") < 0)
                 goto out_free_comm_list;
 
+        symbol_conf.initialized = true;
         return 0;
 
 out_free_dso_list:
@@ -2280,11 +2371,14 @@ out_free_comm_list:
 
 void symbol__exit(void)
 {
+        if (!symbol_conf.initialized)
+                return;
         strlist__delete(symbol_conf.sym_list);
         strlist__delete(symbol_conf.dso_list);
         strlist__delete(symbol_conf.comm_list);
         vmlinux_path__exit();
         symbol_conf.sym_list = symbol_conf.dso_list = symbol_conf.comm_list = NULL;
+        symbol_conf.initialized = false;
 }
 
 int machines__create_kernel_maps(struct rb_root *self, pid_t pid)
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 906be20011d9..6c6eafdb932d 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -53,6 +53,7 @@ struct symbol {
         u64             start;
         u64             end;
         u16             namelen;
+        u8              binding;
         char            name[0];
 };
 
@@ -68,7 +69,8 @@ struct symbol_conf {
                         show_nr_samples,
                         use_callchain,
                         exclude_other,
-                        show_cpu_utilization;
+                        show_cpu_utilization,
+                        initialized;
         const char      *vmlinux_name,
                         *source_prefix,
                         *field_sep;
@@ -164,6 +166,8 @@ void dso__sort_by_name(struct dso *self, enum map_type type);
 struct dso *__dsos__findnew(struct list_head *head, const char *name);
 
 int dso__load(struct dso *self, struct map *map, symbol_filter_t filter);
+int dso__load_vmlinux(struct dso *self, struct map *map,
+                      const char *vmlinux, symbol_filter_t filter);
 int dso__load_vmlinux_path(struct dso *self, struct map *map,
                            symbol_filter_t filter);
 int dso__load_kallsyms(struct dso *self, const char *filename, struct map *map,
@@ -180,6 +184,7 @@ size_t machines__fprintf_dsos(struct rb_root *self, FILE *fp);
 size_t machines__fprintf_dsos_buildid(struct rb_root *self, FILE *fp, bool with_hits);
 
 size_t dso__fprintf_buildid(struct dso *self, FILE *fp);
+size_t dso__fprintf_symbols_by_name(struct dso *self, enum map_type type, FILE *fp);
 size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp);
 
 enum dso_origin {
diff --git a/tools/perf/util/trace-event-scripting.c b/tools/perf/util/trace-event-scripting.c
index 7ea983acfaea..f7af2fca965d 100644
--- a/tools/perf/util/trace-event-scripting.c
+++ b/tools/perf/util/trace-event-scripting.c
@@ -97,7 +97,7 @@ void setup_python_scripting(void)
         register_python_scripting(&python_scripting_unsupported_ops);
 }
 #else
-struct scripting_ops python_scripting_ops;
+extern struct scripting_ops python_scripting_ops;
 
 void setup_python_scripting(void)
 {
@@ -158,7 +158,7 @@ void setup_perl_scripting(void)
         register_perl_scripting(&perl_scripting_unsupported_ops);
 }
 #else
-struct scripting_ops perl_scripting_ops;
+extern struct scripting_ops perl_scripting_ops;
 
 void setup_perl_scripting(void)
 {
diff --git a/tools/perf/util/ui/browser.c b/tools/perf/util/ui/browser.c
new file mode 100644
index 000000000000..8bc010edca25
--- /dev/null
+++ b/tools/perf/util/ui/browser.c
@@ -0,0 +1,337 @@
+#include "libslang.h"
+#include <linux/compiler.h>
+#include <linux/list.h>
+#include <linux/rbtree.h>
+#include <stdlib.h>
+#include <sys/ttydefaults.h>
+#include "browser.h"
+#include "helpline.h"
+#include "../color.h"
+#include "../util.h"
+#include <stdio.h>
+
+static int ui_browser__percent_color(double percent, bool current)
+{
+        if (current)
+                return HE_COLORSET_SELECTED;
+        if (percent >= MIN_RED)
+                return HE_COLORSET_TOP;
+        if (percent >= MIN_GREEN)
+                return HE_COLORSET_MEDIUM;
+        return HE_COLORSET_NORMAL;
+}
+
+void ui_browser__set_color(struct ui_browser *self __used, int color)
+{
+        SLsmg_set_color(color);
+}
+
+void ui_browser__set_percent_color(struct ui_browser *self,
+                                   double percent, bool current)
+{
+         int color = ui_browser__percent_color(percent, current);
+         ui_browser__set_color(self, color);
+}
+
+void ui_browser__gotorc(struct ui_browser *self, int y, int x)
+{
+        SLsmg_gotorc(self->y + y, self->x + x);
+}
+
+void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whence)
+{
+        struct list_head *head = self->entries;
+        struct list_head *pos;
+
+        switch (whence) {
+        case SEEK_SET:
+                pos = head->next;
+                break;
+        case SEEK_CUR:
+                pos = self->top;
+                break;
+        case SEEK_END:
+                pos = head->prev;
+                break;
+        default:
+                return;
+        }
+
+        if (offset > 0) {
+                while (offset-- != 0)
+                        pos = pos->next;
+        } else {
+                while (offset++ != 0)
+                        pos = pos->prev;
+        }
+
+        self->top = pos;
+}
+
+void ui_browser__rb_tree_seek(struct ui_browser *self, off_t offset, int whence)
+{
+        struct rb_root *root = self->entries;
+        struct rb_node *nd;
+
+        switch (whence) {
+        case SEEK_SET:
+                nd = rb_first(root);
+                break;
+        case SEEK_CUR:
+                nd = self->top;
+                break;
+        case SEEK_END:
+                nd = rb_last(root);
+                break;
+        default:
+                return;
+        }
+
+        if (offset > 0) {
+                while (offset-- != 0)
+                        nd = rb_next(nd);
+        } else {
+                while (offset++ != 0)
+                        nd = rb_prev(nd);
+        }
+
+        self->top = nd;
+}
+
+unsigned int ui_browser__rb_tree_refresh(struct ui_browser *self)
+{
+        struct rb_node *nd;
+        int row = 0;
+
+        if (self->top == NULL)
+                self->top = rb_first(self->entries);
+
+        nd = self->top;
+
+        while (nd != NULL) {
+                ui_browser__gotorc(self, row, 0);
+                self->write(self, nd, row);
+                if (++row == self->height)
+                        break;
+                nd = rb_next(nd);
+        }
+
+        return row;
+}
+
+bool ui_browser__is_current_entry(struct ui_browser *self, unsigned row)
+{
+        return self->top_idx + row == self->index;
+}
+
+void ui_browser__refresh_dimensions(struct ui_browser *self)
+{
+        int cols, rows;
+        newtGetScreenSize(&cols, &rows);
+
+        self->width = cols - 1;
+        self->height = rows - 2;
+        self->y = 1;
+        self->x = 0;
+}
+
+void ui_browser__reset_index(struct ui_browser *self)
+{
+        self->index = self->top_idx = 0;
+        self->seek(self, 0, SEEK_SET);
+}
+
+void ui_browser__add_exit_key(struct ui_browser *self, int key)
+{
+        newtFormAddHotKey(self->form, key);
+}
+
+void ui_browser__add_exit_keys(struct ui_browser *self, int keys[])
+{
+        int i = 0;
+
+        while (keys[i] && i < 64) {
+                ui_browser__add_exit_key(self, keys[i]);
+                ++i;
+        }
+}
+
+int ui_browser__show(struct ui_browser *self, const char *title,
+                     const char *helpline, ...)
+{
+        va_list ap;
+        int keys[] = { NEWT_KEY_UP, NEWT_KEY_DOWN, NEWT_KEY_PGUP,
+                       NEWT_KEY_PGDN, NEWT_KEY_HOME, NEWT_KEY_END, ' ',
+                       NEWT_KEY_LEFT, NEWT_KEY_ESCAPE, 'q', CTRL('c'), 0 };
+
+        if (self->form != NULL)
+                newtFormDestroy(self->form);
+
+        ui_browser__refresh_dimensions(self);
+        self->form = newtForm(NULL, NULL, 0);
+        if (self->form == NULL)
+                return -1;
+
+        self->sb = newtVerticalScrollbar(self->width, 1, self->height,
+                                         HE_COLORSET_NORMAL,
+                                         HE_COLORSET_SELECTED);
+        if (self->sb == NULL)
+                return -1;
+
+        SLsmg_gotorc(0, 0);
+        ui_browser__set_color(self, NEWT_COLORSET_ROOT);
+        slsmg_write_nstring(title, self->width);
+
+        ui_browser__add_exit_keys(self, keys);
+        newtFormAddComponent(self->form, self->sb);
+
+        va_start(ap, helpline);
+        ui_helpline__vpush(helpline, ap);
+        va_end(ap);
+        return 0;
+}
+
+void ui_browser__hide(struct ui_browser *self)
+{
+        newtFormDestroy(self->form);
+        self->form = NULL;
+        ui_helpline__pop();
+}
+
+int ui_browser__refresh(struct ui_browser *self)
+{
+        int row;
+
+        newtScrollbarSet(self->sb, self->index, self->nr_entries - 1);
+        row = self->refresh(self);
+        ui_browser__set_color(self, HE_COLORSET_NORMAL);
+        SLsmg_fill_region(self->y + row, self->x,
+                          self->height - row, self->width, ' ');
+
+        return 0;
+}
+
+int ui_browser__run(struct ui_browser *self)
+{
+        struct newtExitStruct es;
+
+        if (ui_browser__refresh(self) < 0)
+                return -1;
+
+        while (1) {
+                off_t offset;
+
+                newtFormRun(self->form, &es);
+
+                if (es.reason != NEWT_EXIT_HOTKEY)
+                        break;
+                switch (es.u.key) {
+                case NEWT_KEY_DOWN:
+                        if (self->index == self->nr_entries - 1)
+                                break;
+                        ++self->index;
+                        if (self->index == self->top_idx + self->height) {
+                                ++self->top_idx;
+                                self->seek(self, +1, SEEK_CUR);
+                        }
+                        break;
+                case NEWT_KEY_UP:
+                        if (self->index == 0)
+                                break;
+                        --self->index;
+                        if (self->index < self->top_idx) {
+                                --self->top_idx;
+                                self->seek(self, -1, SEEK_CUR);
+                        }
+                        break;
+                case NEWT_KEY_PGDN:
+                case ' ':
+                        if (self->top_idx + self->height > self->nr_entries - 1)
+                                break;
+
+                        offset = self->height;
+                        if (self->index + offset > self->nr_entries - 1)
+                                offset = self->nr_entries - 1 - self->index;
+                        self->index += offset;
+                        self->top_idx += offset;
+                        self->seek(self, +offset, SEEK_CUR);
+                        break;
+                case NEWT_KEY_PGUP:
+                        if (self->top_idx == 0)
+                                break;
+
+                        if (self->top_idx < self->height)
+                                offset = self->top_idx;
+                        else
+                                offset = self->height;
+
+                        self->index -= offset;
+                        self->top_idx -= offset;
+                        self->seek(self, -offset, SEEK_CUR);
+                        break;
+                case NEWT_KEY_HOME:
+                        ui_browser__reset_index(self);
+                        break;
+                case NEWT_KEY_END:
+                        offset = self->height - 1;
+                        if (offset >= self->nr_entries)
+                                offset = self->nr_entries - 1;
+
+                        self->index = self->nr_entries - 1;
+                        self->top_idx = self->index - offset;
+                        self->seek(self, -offset, SEEK_END);
+                        break;
+                default:
+                        return es.u.key;
+                }
+                if (ui_browser__refresh(self) < 0)
+                        return -1;
+        }
+        return -1;
+}
+
+unsigned int ui_browser__list_head_refresh(struct ui_browser *self)
+{
+        struct list_head *pos;
+        struct list_head *head = self->entries;
+        int row = 0;
+
+        if (self->top == NULL || self->top == self->entries)
+                self->top = head->next;
+
+        pos = self->top;
+
+        list_for_each_from(pos, head) {
+                ui_browser__gotorc(self, row, 0);
+                self->write(self, pos, row);
+                if (++row == self->height)
+                        break;
+        }
+
+        return row;
+}
+
+static struct newtPercentTreeColors {
+        const char *topColorFg, *topColorBg;
+        const char *mediumColorFg, *mediumColorBg;
+        const char *normalColorFg, *normalColorBg;
+        const char *selColorFg, *selColorBg;
+        const char *codeColorFg, *codeColorBg;
+} defaultPercentTreeColors = {
+        "red",       "lightgray",
+        "green",     "lightgray",
+        "black",     "lightgray",
+        "lightgray", "magenta",
+        "blue",      "lightgray",
+};
+
+void ui_browser__init(void)
+{
+        struct newtPercentTreeColors *c = &defaultPercentTreeColors;
+
+        sltt_set_color(HE_COLORSET_TOP, NULL, c->topColorFg, c->topColorBg);
+        sltt_set_color(HE_COLORSET_MEDIUM, NULL, c->mediumColorFg, c->mediumColorBg);
+        sltt_set_color(HE_COLORSET_NORMAL, NULL, c->normalColorFg, c->normalColorBg);
+        sltt_set_color(HE_COLORSET_SELECTED, NULL, c->selColorFg, c->selColorBg);
+        sltt_set_color(HE_COLORSET_CODE, NULL, c->codeColorFg, c->codeColorBg);
+}
diff --git a/tools/perf/util/ui/browser.h b/tools/perf/util/ui/browser.h
new file mode 100644
index 000000000000..0dc7e4da36f5
--- /dev/null
+++ b/tools/perf/util/ui/browser.h
@@ -0,0 +1,51 @@
+#ifndef _PERF_UI_BROWSER_H_
+#define _PERF_UI_BROWSER_H_ 1
+
+#include <stdbool.h>
+#include <newt.h>
+#include <sys/types.h>
+#include "../types.h"
+
+#define HE_COLORSET_TOP         50
+#define HE_COLORSET_MEDIUM      51
+#define HE_COLORSET_NORMAL      52
+#define HE_COLORSET_SELECTED    53
+#define HE_COLORSET_CODE        54
+
+struct ui_browser {
+        newtComponent form, sb;
+        u64           index, top_idx;
+        void          *top, *entries;
+        u16           y, x, width, height;
+        void          *priv;
+        unsigned int  (*refresh)(struct ui_browser *self);
+        void          (*write)(struct ui_browser *self, void *entry, int row);
+        void          (*seek)(struct ui_browser *self, off_t offset, int whence);
+        u32           nr_entries;
+};
+
+
+void ui_browser__set_color(struct ui_browser *self, int color);
+void ui_browser__set_percent_color(struct ui_browser *self,
+                                   double percent, bool current);
+bool ui_browser__is_current_entry(struct ui_browser *self, unsigned row);
+void ui_browser__refresh_dimensions(struct ui_browser *self);
+void ui_browser__reset_index(struct ui_browser *self);
+
+void ui_browser__gotorc(struct ui_browser *self, int y, int x);
+void ui_browser__add_exit_key(struct ui_browser *self, int key);
+void ui_browser__add_exit_keys(struct ui_browser *self, int keys[]);
+int ui_browser__show(struct ui_browser *self, const char *title,
+                     const char *helpline, ...);
+void ui_browser__hide(struct ui_browser *self);
+int ui_browser__refresh(struct ui_browser *self);
+int ui_browser__run(struct ui_browser *self);
+
+void ui_browser__rb_tree_seek(struct ui_browser *self, off_t offset, int whence);
+unsigned int ui_browser__rb_tree_refresh(struct ui_browser *self);
+
+void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whence);
+unsigned int ui_browser__list_head_refresh(struct ui_browser *self);
+
+void ui_browser__init(void);
+#endif /* _PERF_UI_BROWSER_H_ */
diff --git a/tools/perf/util/ui/browsers/annotate.c b/tools/perf/util/ui/browsers/annotate.c
new file mode 100644
index 000000000000..82b78f99251b
--- /dev/null
+++ b/tools/perf/util/ui/browsers/annotate.c
@@ -0,0 +1,237 @@
+#include "../browser.h"
+#include "../helpline.h"
+#include "../libslang.h"
+#include "../../hist.h"
+#include "../../sort.h"
+#include "../../symbol.h"
+
+static void ui__error_window(const char *fmt, ...)
+{
+        va_list ap;
+
+        va_start(ap, fmt);
+        newtWinMessagev((char *)"Error", (char *)"Ok", (char *)fmt, ap);
+        va_end(ap);
+}
+
+struct annotate_browser {
+        struct ui_browser b;
+        struct rb_root    entries;
+        struct rb_node    *curr_hot;
+};
+
+struct objdump_line_rb_node {
+        struct rb_node  rb_node;
+        double          percent;
+        u32             idx;
+};
+
+static inline
+struct objdump_line_rb_node *objdump_line__rb(struct objdump_line *self)
+{
+        return (struct objdump_line_rb_node *)(self + 1);
+}
+
+static void annotate_browser__write(struct ui_browser *self, void *entry, int row)
+{
+        struct objdump_line *ol = rb_entry(entry, struct objdump_line, node);
+        bool current_entry = ui_browser__is_current_entry(self, row);
+        int width = self->width;
+
+        if (ol->offset != -1) {
+                struct objdump_line_rb_node *olrb = objdump_line__rb(ol);
+                ui_browser__set_percent_color(self, olrb->percent, current_entry);
+                slsmg_printf(" %7.2f ", olrb->percent);
+                if (!current_entry)
+                        ui_browser__set_color(self, HE_COLORSET_CODE);
+        } else {
+                ui_browser__set_percent_color(self, 0, current_entry);
+                slsmg_write_nstring(" ", 9);
+        }
+
+        SLsmg_write_char(':');
+        slsmg_write_nstring(" ", 8);
+        if (!*ol->line)
+                slsmg_write_nstring(" ", width - 18);
+        else
+                slsmg_write_nstring(ol->line, width - 18);
+}
+
+static double objdump_line__calc_percent(struct objdump_line *self,
+                                         struct list_head *head,
+                                         struct symbol *sym)
+{
+        double percent = 0.0;
+
+        if (self->offset != -1) {
+                int len = sym->end - sym->start;
+                unsigned int hits = 0;
+                struct sym_priv *priv = symbol__priv(sym);
+                struct sym_ext *sym_ext = priv->ext;
+                struct sym_hist *h = priv->hist;
+                s64 offset = self->offset;
+                struct objdump_line *next = objdump__get_next_ip_line(head, self);
+
+
+                while (offset < (s64)len &&
+                       (next == NULL || offset < next->offset)) {
+                        if (sym_ext) {
+                                percent += sym_ext[offset].percent;
+                        } else
+                                hits += h->ip[offset];
+
+                        ++offset;
+                }
+
+                if (sym_ext == NULL && h->sum)
+                        percent = 100.0 * hits / h->sum;
+        }
+
+        return percent;
+}
+
+static void objdump__insert_line(struct rb_root *self,
+                                 struct objdump_line_rb_node *line)
+{
+        struct rb_node **p = &self->rb_node;
+        struct rb_node *parent = NULL;
+        struct objdump_line_rb_node *l;
+
+        while (*p != NULL) {
+                parent = *p;
+                l = rb_entry(parent, struct objdump_line_rb_node, rb_node);
+                if (line->percent < l->percent)
+                        p = &(*p)->rb_left;
+                else
+                        p = &(*p)->rb_right;
+        }
+        rb_link_node(&line->rb_node, parent, p);
+        rb_insert_color(&line->rb_node, self);
+}
+
+static void annotate_browser__set_top(struct annotate_browser *self,
+                                      struct rb_node *nd)
+{
+        struct objdump_line_rb_node *rbpos;
+        struct objdump_line *pos;
+        unsigned back;
+
+        ui_browser__refresh_dimensions(&self->b);
+        back = self->b.height / 2;
+        rbpos = rb_entry(nd, struct objdump_line_rb_node, rb_node);
+        pos = ((struct objdump_line *)rbpos) - 1;
+        self->b.top_idx = self->b.index = rbpos->idx;
+
+        while (self->b.top_idx != 0 && back != 0) {
+                pos = list_entry(pos->node.prev, struct objdump_line, node);
+
+                --self->b.top_idx;
+                --back;
+        }
+
+        self->b.top = pos;
+        self->curr_hot = nd;
+}
+
+static int annotate_browser__run(struct annotate_browser *self)
+{
+        struct rb_node *nd;
+        struct hist_entry *he = self->b.priv;
+        int key;
+
+        if (ui_browser__show(&self->b, he->ms.sym->name,
+                             "<-, -> or ESC: exit, TAB/shift+TAB: cycle thru samples") < 0)
+                return -1;
+        /*
+         * To allow builtin-annotate to cycle thru multiple symbols by
+         * examining the exit key for this function.
+         */
+        ui_browser__add_exit_key(&self->b, NEWT_KEY_RIGHT);
+
+        nd = self->curr_hot;
+        if (nd) {
+                int tabs[] = { NEWT_KEY_TAB, NEWT_KEY_UNTAB, 0 };
+                ui_browser__add_exit_keys(&self->b, tabs);
+        }
+
+        while (1) {
+                key = ui_browser__run(&self->b);
+
+                switch (key) {
+                case NEWT_KEY_TAB:
+                        nd = rb_prev(nd);
+                        if (nd == NULL)
+                                nd = rb_last(&self->entries);
+                        annotate_browser__set_top(self, nd);
+                        break;
+                case NEWT_KEY_UNTAB:
+                        nd = rb_next(nd);
+                        if (nd == NULL)
+                                nd = rb_first(&self->entries);
+                        annotate_browser__set_top(self, nd);
+                        break;
+                default:
+                        goto out;
+                }
+        }
+out:
+        ui_browser__hide(&self->b);
+        return key;
+}
+
+int hist_entry__tui_annotate(struct hist_entry *self)
+{
+        struct objdump_line *pos, *n;
+        struct objdump_line_rb_node *rbpos;
+        LIST_HEAD(head);
+        struct annotate_browser browser = {
+                .b = {
+                        .entries = &head,
+                        .refresh = ui_browser__list_head_refresh,
+                        .seek    = ui_browser__list_head_seek,
+                        .write   = annotate_browser__write,
+                        .priv    = self,
+                },
+        };
+        int ret;
+
+        if (self->ms.sym == NULL)
+                return -1;
+
+        if (self->ms.map->dso->annotate_warned)
+                return -1;
+
+        if (hist_entry__annotate(self, &head, sizeof(*rbpos)) < 0) {
+                ui__error_window(ui_helpline__last_msg);
+                return -1;
+        }
+
+        ui_helpline__push("Press <- or ESC to exit");
+
+        list_for_each_entry(pos, &head, node) {
+                size_t line_len = strlen(pos->line);
+                if (browser.b.width < line_len)
+                        browser.b.width = line_len;
+                rbpos = objdump_line__rb(pos);
+                rbpos->idx = browser.b.nr_entries++;
+                rbpos->percent = objdump_line__calc_percent(pos, &head, self->ms.sym);
+                if (rbpos->percent < 0.01)
+                        continue;
+                objdump__insert_line(&browser.entries, rbpos);
+        }
+
+        /*
+         * Position the browser at the hottest line.
+         */
+        browser.curr_hot = rb_last(&browser.entries);
+        if (browser.curr_hot)
+                annotate_browser__set_top(&browser, browser.curr_hot);
+
+        browser.b.width += 18; /* Percentage */
+        ret = annotate_browser__run(&browser);
+        list_for_each_entry_safe(pos, n, &head, node) {
+                list_del(&pos->node);
+                objdump_line__free(pos);
+        }
+        return ret;
+}
diff --git a/tools/perf/util/ui/browsers/hists.c b/tools/perf/util/ui/browsers/hists.c
new file mode 100644
index 000000000000..ebda8c3fde9e
--- /dev/null
+++ b/tools/perf/util/ui/browsers/hists.c
@@ -0,0 +1,1013 @@
+#define _GNU_SOURCE
+#include <stdio.h>
+#undef _GNU_SOURCE
+#include "../libslang.h"
+#include <stdlib.h>
+#include <string.h>
+#include <newt.h>
+#include <linux/rbtree.h>
+
+#include "../../hist.h"
+#include "../../pstack.h"
+#include "../../sort.h"
+#include "../../util.h"
+
+#include "../browser.h"
+#include "../helpline.h"
+#include "../util.h"
+#include "map.h"
+
+struct hist_browser {
+        struct ui_browser   b;
+        struct hists        *hists;
+        struct hist_entry   *he_selection;
+        struct map_symbol   *selection;
+};
+
+static void hist_browser__refresh_dimensions(struct hist_browser *self)
+{
+        /* 3 == +/- toggle symbol before actual hist_entry rendering */
+        self->b.width = 3 + (hists__sort_list_width(self->hists) +
+                             sizeof("[k]"));
+}
+
+static void hist_browser__reset(struct hist_browser *self)
+{
+        self->b.nr_entries = self->hists->nr_entries;
+        hist_browser__refresh_dimensions(self);
+        ui_browser__reset_index(&self->b);
+}
+
+static char tree__folded_sign(bool unfolded)
+{
+        return unfolded ? '-' : '+';
+}
+
+static char map_symbol__folded(const struct map_symbol *self)
+{
+        return self->has_children ? tree__folded_sign(self->unfolded) : ' ';
+}
+
+static char hist_entry__folded(const struct hist_entry *self)
+{
+        return map_symbol__folded(&self->ms);
+}
+
+static char callchain_list__folded(const struct callchain_list *self)
+{
+        return map_symbol__folded(&self->ms);
+}
+
+static void map_symbol__set_folding(struct map_symbol *self, bool unfold)
+{
+        self->unfolded = unfold ? self->has_children : false;
+}
+
+static int callchain_node__count_rows_rb_tree(struct callchain_node *self)
+{
+        int n = 0;
+        struct rb_node *nd;
+
+        for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) {
+                struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node);
+                struct callchain_list *chain;
+                char folded_sign = ' '; /* No children */
+
+                list_for_each_entry(chain, &child->val, list) {
+                        ++n;
+                        /* We need this because we may not have children */
+                        folded_sign = callchain_list__folded(chain);
+                        if (folded_sign == '+')
+                                break;
+                }
+
+                if (folded_sign == '-') /* Have children and they're unfolded */
+                        n += callchain_node__count_rows_rb_tree(child);
+        }
+
+        return n;
+}
+
+static int callchain_node__count_rows(struct callchain_node *node)
+{
+        struct callchain_list *chain;
+        bool unfolded = false;
+        int n = 0;
+
+        list_for_each_entry(chain, &node->val, list) {
+                ++n;
+                unfolded = chain->ms.unfolded;
+        }
+
+        if (unfolded)
+                n += callchain_node__count_rows_rb_tree(node);
+
+        return n;
+}
+
+static int callchain__count_rows(struct rb_root *chain)
+{
+        struct rb_node *nd;
+        int n = 0;
+
+        for (nd = rb_first(chain); nd; nd = rb_next(nd)) {
+                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
+                n += callchain_node__count_rows(node);
+        }
+
+        return n;
+}
+
+static bool map_symbol__toggle_fold(struct map_symbol *self)
+{
+        if (!self->has_children)
+                return false;
+
+        self->unfolded = !self->unfolded;
+        return true;
+}
+
+static void callchain_node__init_have_children_rb_tree(struct callchain_node *self)
+{
+        struct rb_node *nd = rb_first(&self->rb_root);
+
+        for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) {
+                struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node);
+                struct callchain_list *chain;
+                bool first = true;
+
+                list_for_each_entry(chain, &child->val, list) {
+                        if (first) {
+                                first = false;
+                                chain->ms.has_children = chain->list.next != &child->val ||
+                                                         !RB_EMPTY_ROOT(&child->rb_root);
+                        } else
+                                chain->ms.has_children = chain->list.next == &child->val &&
+                                                         !RB_EMPTY_ROOT(&child->rb_root);
+                }
+
+                callchain_node__init_have_children_rb_tree(child);
+        }
+}
+
+static void callchain_node__init_have_children(struct callchain_node *self)
+{
+        struct callchain_list *chain;
+
+        list_for_each_entry(chain, &self->val, list)
+                chain->ms.has_children = !RB_EMPTY_ROOT(&self->rb_root);
+
+        callchain_node__init_have_children_rb_tree(self);
+}
+
+static void callchain__init_have_children(struct rb_root *self)
+{
+        struct rb_node *nd;
+
+        for (nd = rb_first(self); nd; nd = rb_next(nd)) {
+                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
+                callchain_node__init_have_children(node);
+        }
+}
+
+static void hist_entry__init_have_children(struct hist_entry *self)
+{
+        if (!self->init_have_children) {
+                self->ms.has_children = !RB_EMPTY_ROOT(&self->sorted_chain);
+                callchain__init_have_children(&self->sorted_chain);
+                self->init_have_children = true;
+        }
+}
+
+static bool hist_browser__toggle_fold(struct hist_browser *self)
+{
+        if (map_symbol__toggle_fold(self->selection)) {
+                struct hist_entry *he = self->he_selection;
+
+                hist_entry__init_have_children(he);
+                self->hists->nr_entries -= he->nr_rows;
+
+                if (he->ms.unfolded)
+                        he->nr_rows = callchain__count_rows(&he->sorted_chain);
+                else
+                        he->nr_rows = 0;
+                self->hists->nr_entries += he->nr_rows;
+                self->b.nr_entries = self->hists->nr_entries;
+
+                return true;
+        }
+
+        /* If it doesn't have children, no toggling performed */
+        return false;
+}
+
+static int callchain_node__set_folding_rb_tree(struct callchain_node *self, bool unfold)
+{
+        int n = 0;
+        struct rb_node *nd;
+
+        for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) {
+                struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node);
+                struct callchain_list *chain;
+                bool has_children = false;
+
+                list_for_each_entry(chain, &child->val, list) {
+                        ++n;
+                        map_symbol__set_folding(&chain->ms, unfold);
+                        has_children = chain->ms.has_children;
+                }
+
+                if (has_children)
+                        n += callchain_node__set_folding_rb_tree(child, unfold);
+        }
+
+        return n;
+}
+
+static int callchain_node__set_folding(struct callchain_node *node, bool unfold)
+{
+        struct callchain_list *chain;
+        bool has_children = false;
+        int n = 0;
+
+        list_for_each_entry(chain, &node->val, list) {
+                ++n;
+                map_symbol__set_folding(&chain->ms, unfold);
+                has_children = chain->ms.has_children;
+        }
+
+        if (has_children)
+                n += callchain_node__set_folding_rb_tree(node, unfold);
+
+        return n;
+}
+
+static int callchain__set_folding(struct rb_root *chain, bool unfold)
+{
+        struct rb_node *nd;
+        int n = 0;
+
+        for (nd = rb_first(chain); nd; nd = rb_next(nd)) {
+                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
+                n += callchain_node__set_folding(node, unfold);
+        }
+
+        return n;
+}
+
+static void hist_entry__set_folding(struct hist_entry *self, bool unfold)
+{
+        hist_entry__init_have_children(self);
+        map_symbol__set_folding(&self->ms, unfold);
+
+        if (self->ms.has_children) {
+                int n = callchain__set_folding(&self->sorted_chain, unfold);
+                self->nr_rows = unfold ? n : 0;
+        } else
+                self->nr_rows = 0;
+}
+
+static void hists__set_folding(struct hists *self, bool unfold)
+{
+        struct rb_node *nd;
+
+        self->nr_entries = 0;
+
+        for (nd = rb_first(&self->entries); nd; nd = rb_next(nd)) {
+                struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node);
+                hist_entry__set_folding(he, unfold);
+                self->nr_entries += 1 + he->nr_rows;
+        }
+}
+
+static void hist_browser__set_folding(struct hist_browser *self, bool unfold)
+{
+        hists__set_folding(self->hists, unfold);
+        self->b.nr_entries = self->hists->nr_entries;
+        /* Go to the start, we may be way after valid entries after a collapse */
+        ui_browser__reset_index(&self->b);
+}
+
+static int hist_browser__run(struct hist_browser *self, const char *title)
+{
+        int key;
+        int exit_keys[] = { 'a', '?', 'h', 'C', 'd', 'D', 'E', 't',
+                            NEWT_KEY_ENTER, NEWT_KEY_RIGHT, NEWT_KEY_LEFT, 0, };
+
+        self->b.entries = &self->hists->entries;
+        self->b.nr_entries = self->hists->nr_entries;
+
+        hist_browser__refresh_dimensions(self);
+
+        if (ui_browser__show(&self->b, title,
+                             "Press '?' for help on key bindings") < 0)
+                return -1;
+
+        ui_browser__add_exit_keys(&self->b, exit_keys);
+
+        while (1) {
+                key = ui_browser__run(&self->b);
+
+                switch (key) {
+                case 'D': { /* Debug */
+                        static int seq;
+                        struct hist_entry *h = rb_entry(self->b.top,
+                                                        struct hist_entry, rb_node);
+                        ui_helpline__pop();
+                        ui_helpline__fpush("%d: nr_ent=(%d,%d), height=%d, idx=%d, fve: idx=%d, row_off=%d, nrows=%d",
+                                           seq++, self->b.nr_entries,
+                                           self->hists->nr_entries,
+                                           self->b.height,
+                                           self->b.index,
+                                           self->b.top_idx,
+                                           h->row_offset, h->nr_rows);
+                }
+                        break;
+                case 'C':
+                        /* Collapse the whole world. */
+                        hist_browser__set_folding(self, false);
+                        break;
+                case 'E':
+                        /* Expand the whole world. */
+                        hist_browser__set_folding(self, true);
+                        break;
+                case NEWT_KEY_ENTER:
+                        if (hist_browser__toggle_fold(self))
+                                break;
+                        /* fall thru */
+                default:
+                        goto out;
+                }
+        }
+out:
+        ui_browser__hide(&self->b);
+        return key;
+}
+
+static char *callchain_list__sym_name(struct callchain_list *self,
+                                      char *bf, size_t bfsize)
+{
+        if (self->ms.sym)
+                return self->ms.sym->name;
+
+        snprintf(bf, bfsize, "%#Lx", self->ip);
+        return bf;
+}
+
+#define LEVEL_OFFSET_STEP 3
+
+static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self,
+                                                     struct callchain_node *chain_node,
+                                                     u64 total, int level,
+                                                     unsigned short row,
+                                                     off_t *row_offset,
+                                                     bool *is_current_entry)
+{
+        struct rb_node *node;
+        int first_row = row, width, offset = level * LEVEL_OFFSET_STEP;
+        u64 new_total, remaining;
+
+        if (callchain_param.mode == CHAIN_GRAPH_REL)
+                new_total = chain_node->children_hit;
+        else
+                new_total = total;
+
+        remaining = new_total;
+        node = rb_first(&chain_node->rb_root);
+        while (node) {
+                struct callchain_node *child = rb_entry(node, struct callchain_node, rb_node);
+                struct rb_node *next = rb_next(node);
+                u64 cumul = cumul_hits(child);
+                struct callchain_list *chain;
+                char folded_sign = ' ';
+                int first = true;
+                int extra_offset = 0;
+
+                remaining -= cumul;
+
+                list_for_each_entry(chain, &child->val, list) {
+                        char ipstr[BITS_PER_LONG / 4 + 1], *alloc_str;
+                        const char *str;
+                        int color;
+                        bool was_first = first;
+
+                        if (first)
+                                first = false;
+                        else
+                                extra_offset = LEVEL_OFFSET_STEP;
+
+                        folded_sign = callchain_list__folded(chain);
+                        if (*row_offset != 0) {
+                                --*row_offset;
+                                goto do_next;
+                        }
+
+                        alloc_str = NULL;
+                        str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr));
+                        if (was_first) {
+                                double percent = cumul * 100.0 / new_total;
+
+                                if (asprintf(&alloc_str, "%2.2f%% %s", percent, str) < 0)
+                                        str = "Not enough memory!";
+                                else
+                                        str = alloc_str;
+                        }
+
+                        color = HE_COLORSET_NORMAL;
+                        width = self->b.width - (offset + extra_offset + 2);
+                        if (ui_browser__is_current_entry(&self->b, row)) {
+                                self->selection = &chain->ms;
+                                color = HE_COLORSET_SELECTED;
+                                *is_current_entry = true;
+                        }
+
+                        ui_browser__set_color(&self->b, color);
+                        ui_browser__gotorc(&self->b, row, 0);
+                        slsmg_write_nstring(" ", offset + extra_offset);
+                        slsmg_printf("%c ", folded_sign);
+                        slsmg_write_nstring(str, width);
+                        free(alloc_str);
+
+                        if (++row == self->b.height)
+                                goto out;
+do_next:
+                        if (folded_sign == '+')
+                                break;
+                }
+
+                if (folded_sign == '-') {
+                        const int new_level = level + (extra_offset ? 2 : 1);
+                        row += hist_browser__show_callchain_node_rb_tree(self, child, new_total,
+                                                                         new_level, row, row_offset,
+                                                                         is_current_entry);
+                }
+                if (row == self->b.height)
+                        goto out;
+                node = next;
+        }
+out:
+        return row - first_row;
+}
+
+static int hist_browser__show_callchain_node(struct hist_browser *self,
+                                             struct callchain_node *node,
+                                             int level, unsigned short row,
+                                             off_t *row_offset,
+                                             bool *is_current_entry)
+{
+        struct callchain_list *chain;
+        int first_row = row,
+             offset = level * LEVEL_OFFSET_STEP,
+             width = self->b.width - offset;
+        char folded_sign = ' ';
+
+        list_for_each_entry(chain, &node->val, list) {
+                char ipstr[BITS_PER_LONG / 4 + 1], *s;
+                int color;
+
+                folded_sign = callchain_list__folded(chain);
+
+                if (*row_offset != 0) {
+                        --*row_offset;
+                        continue;
+                }
+
+                color = HE_COLORSET_NORMAL;
+                if (ui_browser__is_current_entry(&self->b, row)) {
+                        self->selection = &chain->ms;
+                        color = HE_COLORSET_SELECTED;
+                        *is_current_entry = true;
+                }
+
+                s = callchain_list__sym_name(chain, ipstr, sizeof(ipstr));
+                ui_browser__gotorc(&self->b, row, 0);
+                ui_browser__set_color(&self->b, color);
+                slsmg_write_nstring(" ", offset);
+                slsmg_printf("%c ", folded_sign);
+                slsmg_write_nstring(s, width - 2);
+
+                if (++row == self->b.height)
+                        goto out;
+        }
+
+        if (folded_sign == '-')
+                row += hist_browser__show_callchain_node_rb_tree(self, node,
+                                                                 self->hists->stats.total_period,
+                                                                 level + 1, row,
+                                                                 row_offset,
+                                                                 is_current_entry);
+out:
+        return row - first_row;
+}
+
+static int hist_browser__show_callchain(struct hist_browser *self,
+                                        struct rb_root *chain,
+                                        int level, unsigned short row,
+                                        off_t *row_offset,
+                                        bool *is_current_entry)
+{
+        struct rb_node *nd;
+        int first_row = row;
+
+        for (nd = rb_first(chain); nd; nd = rb_next(nd)) {
+                struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node);
+
+                row += hist_browser__show_callchain_node(self, node, level,
+                                                         row, row_offset,
+                                                         is_current_entry);
+                if (row == self->b.height)
+                        break;
+        }
+
+        return row - first_row;
+}
+
+static int hist_browser__show_entry(struct hist_browser *self,
+                                    struct hist_entry *entry,
+                                    unsigned short row)
+{
+        char s[256];
+        double percent;
+        int printed = 0;
+        int color, width = self->b.width;
+        char folded_sign = ' ';
+        bool current_entry = ui_browser__is_current_entry(&self->b, row);
+        off_t row_offset = entry->row_offset;
+
+        if (current_entry) {
+                self->he_selection = entry;
+                self->selection = &entry->ms;
+        }
+
+        if (symbol_conf.use_callchain) {
+                hist_entry__init_have_children(entry);
+                folded_sign = hist_entry__folded(entry);
+        }
+
+        if (row_offset == 0) {
+                hist_entry__snprintf(entry, s, sizeof(s), self->hists, NULL, false,
+                                     0, false, self->hists->stats.total_period);
+                percent = (entry->period * 100.0) / self->hists->stats.total_period;
+
+                color = HE_COLORSET_SELECTED;
+                if (!current_entry) {
+                        if (percent >= MIN_RED)
+                                color = HE_COLORSET_TOP;
+                        else if (percent >= MIN_GREEN)
+                                color = HE_COLORSET_MEDIUM;
+                        else
+                                color = HE_COLORSET_NORMAL;
+                }
+
+                ui_browser__set_color(&self->b, color);
+                ui_browser__gotorc(&self->b, row, 0);
+                if (symbol_conf.use_callchain) {
+                        slsmg_printf("%c ", folded_sign);
+                        width -= 2;
+                }
+                slsmg_write_nstring(s, width);
+                ++row;
+                ++printed;
+        } else
+                --row_offset;
+
+        if (folded_sign == '-' && row != self->b.height) {
+                printed += hist_browser__show_callchain(self, &entry->sorted_chain,
+                                                        1, row, &row_offset,
+                                                        &current_entry);
+                if (current_entry)
+                        self->he_selection = entry;
+        }
+
+        return printed;
+}
+
+static unsigned int hist_browser__refresh(struct ui_browser *self)
+{
+        unsigned row = 0;
+        struct rb_node *nd;
+        struct hist_browser *hb = container_of(self, struct hist_browser, b);
+
+        if (self->top == NULL)
+                self->top = rb_first(&hb->hists->entries);
+
+        for (nd = self->top; nd; nd = rb_next(nd)) {
+                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
+
+                if (h->filtered)
+                        continue;
+
+                row += hist_browser__show_entry(hb, h, row);
+                if (row == self->height)
+                        break;
+        }
+
+        return row;
+}
+
+static struct rb_node *hists__filter_entries(struct rb_node *nd)
+{
+        while (nd != NULL) {
+                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
+                if (!h->filtered)
+                        return nd;
+
+                nd = rb_next(nd);
+        }
+
+        return NULL;
+}
+
+static struct rb_node *hists__filter_prev_entries(struct rb_node *nd)
+{
+        while (nd != NULL) {
+                struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node);
+                if (!h->filtered)
+                        return nd;
+
+                nd = rb_prev(nd);
+        }
+
+        return NULL;
+}
+
+static void ui_browser__hists_seek(struct ui_browser *self,
+                                   off_t offset, int whence)
+{
+        struct hist_entry *h;
+        struct rb_node *nd;
+        bool first = true;
+
+        switch (whence) {
+        case SEEK_SET:
+                nd = hists__filter_entries(rb_first(self->entries));
+                break;
+        case SEEK_CUR:
+                nd = self->top;
+                goto do_offset;
+        case SEEK_END:
+                nd = hists__filter_prev_entries(rb_last(self->entries));
+                first = false;
+                break;
+        default:
+                return;
+        }
+
+        /*
+         * Moves not relative to the first visible entry invalidates its
+         * row_offset:
+         */
+        h = rb_entry(self->top, struct hist_entry, rb_node);
+        h->row_offset = 0;
+
+        /*
+         * Here we have to check if nd is expanded (+), if it is we can't go
+         * the next top level hist_entry, instead we must compute an offset of
+         * what _not_ to show and not change the first visible entry.
+         *
+         * This offset increments when we are going from top to bottom and
+         * decreases when we're going from bottom to top.
+         *
+         * As we don't have backpointers to the top level in the callchains
+         * structure, we need to always print the whole hist_entry callchain,
+         * skipping the first ones that are before the first visible entry
+         * and stop when we printed enough lines to fill the screen.
+         */
+do_offset:
+        if (offset > 0) {
+                do {
+                        h = rb_entry(nd, struct hist_entry, rb_node);
+                        if (h->ms.unfolded) {
+                                u16 remaining = h->nr_rows - h->row_offset;
+                                if (offset > remaining) {
+                                        offset -= remaining;
+                                        h->row_offset = 0;
+                                } else {
+                                        h->row_offset += offset;
+                                        offset = 0;
+                                        self->top = nd;
+                                        break;
+                                }
+                        }
+                        nd = hists__filter_entries(rb_next(nd));
+                        if (nd == NULL)
+                                break;
+                        --offset;
+                        self->top = nd;
+                } while (offset != 0);
+        } else if (offset < 0) {
+                while (1) {
+                        h = rb_entry(nd, struct hist_entry, rb_node);
+                        if (h->ms.unfolded) {
+                                if (first) {
+                                        if (-offset > h->row_offset) {
+                                                offset += h->row_offset;
+                                                h->row_offset = 0;
+                                        } else {
+                                                h->row_offset += offset;
+                                                offset = 0;
+                                                self->top = nd;
+                                                break;
+                                        }
+                                } else {
+                                        if (-offset > h->nr_rows) {
+                                                offset += h->nr_rows;
+                                                h->row_offset = 0;
+                                        } else {
+                                                h->row_offset = h->nr_rows + offset;
+                                                offset = 0;
+                                                self->top = nd;
+                                                break;
+                                        }
+                                }
+                        }
+
+                        nd = hists__filter_prev_entries(rb_prev(nd));
+                        if (nd == NULL)
+                                break;
+                        ++offset;
+                        self->top = nd;
+                        if (offset == 0) {
+                                /*
+                                 * Last unfiltered hist_entry, check if it is
+                                 * unfolded, if it is then we should have
+                                 * row_offset at its last entry.
+                                 */
+                                h = rb_entry(nd, struct hist_entry, rb_node);
+                                if (h->ms.unfolded)
+                                        h->row_offset = h->nr_rows;
+                                break;
+                        }
+                        first = false;
+                }
+        } else {
+                self->top = nd;
+                h = rb_entry(nd, struct hist_entry, rb_node);
+                h->row_offset = 0;
+        }
+}
+
+static struct hist_browser *hist_browser__new(struct hists *hists)
+{
+        struct hist_browser *self = zalloc(sizeof(*self));
+
+        if (self) {
+                self->hists = hists;
+                self->b.refresh = hist_browser__refresh;
+                self->b.seek = ui_browser__hists_seek;
+        }
+
+        return self;
+}
+
+static void hist_browser__delete(struct hist_browser *self)
+{
+        free(self);
+}
+
+static struct hist_entry *hist_browser__selected_entry(struct hist_browser *self)
+{
+        return self->he_selection;
+}
+
+static struct thread *hist_browser__selected_thread(struct hist_browser *self)
+{
+        return self->he_selection->thread;
+}
+
+static int hists__browser_title(struct hists *self, char *bf, size_t size,
+                                const char *ev_name, const struct dso *dso,
+                                const struct thread *thread)
+{
+        char unit;
+        int printed;
+        unsigned long nr_events = self->stats.nr_events[PERF_RECORD_SAMPLE];
+
+        nr_events = convert_unit(nr_events, &unit);
+        printed = snprintf(bf, size, "Events: %lu%c %s", nr_events, unit, ev_name);
+
+        if (thread)
+                printed += snprintf(bf + printed, size - printed,
+                                    ", Thread: %s(%d)",
+                                    (thread->comm_set ? thread->comm : ""),
+                                    thread->pid);
+        if (dso)
+                printed += snprintf(bf + printed, size - printed,
+                                    ", DSO: %s", dso->short_name);
+        return printed;
+}
+
+int hists__browse(struct hists *self, const char *helpline, const char *ev_name)
+{
+        struct hist_browser *browser = hist_browser__new(self);
+        struct pstack *fstack;
+        const struct thread *thread_filter = NULL;
+        const struct dso *dso_filter = NULL;
+        char msg[160];
+        int key = -1;
+
+        if (browser == NULL)
+                return -1;
+
+        fstack = pstack__new(2);
+        if (fstack == NULL)
+                goto out;
+
+        ui_helpline__push(helpline);
+
+        hists__browser_title(self, msg, sizeof(msg), ev_name,
+                             dso_filter, thread_filter);
+        while (1) {
+                const struct thread *thread;
+                const struct dso *dso;
+                char *options[16];
+                int nr_options = 0, choice = 0, i,
+                    annotate = -2, zoom_dso = -2, zoom_thread = -2,
+                    browse_map = -2;
+
+                key = hist_browser__run(browser, msg);
+
+                thread = hist_browser__selected_thread(browser);
+                dso = browser->selection->map ? browser->selection->map->dso : NULL;
+
+                switch (key) {
+                case NEWT_KEY_TAB:
+                case NEWT_KEY_UNTAB:
+                        /*
+                         * Exit the browser, let hists__browser_tree
+                         * go to the next or previous
+                         */
+                        goto out_free_stack;
+                case 'a':
+                        if (browser->selection->map == NULL &&
+                            browser->selection->map->dso->annotate_warned)
+                                continue;
+                        goto do_annotate;
+                case 'd':
+                        goto zoom_dso;
+                case 't':
+                        goto zoom_thread;
+                case NEWT_KEY_F1:
+                case 'h':
+                case '?':
+                        ui__help_window("->        Zoom into DSO/Threads & Annotate current symbol\n"
+                                        "<-        Zoom out\n"
+                                        "a         Annotate current symbol\n"
+                                        "h/?/F1    Show this window\n"
+                                        "C         Collapse all callchains\n"
+                                        "E         Expand all callchains\n"
+                                        "d         Zoom into current DSO\n"
+                                        "t         Zoom into current Thread\n"
+                                        "q/CTRL+C  Exit browser");
+                        continue;
+                case NEWT_KEY_ENTER:
+                case NEWT_KEY_RIGHT:
+                        /* menu */
+                        break;
+                case NEWT_KEY_LEFT: {
+                        const void *top;
+
+                        if (pstack__empty(fstack))
+                                continue;
+                        top = pstack__pop(fstack);
+                        if (top == &dso_filter)
+                                goto zoom_out_dso;
+                        if (top == &thread_filter)
+                                goto zoom_out_thread;
+                        continue;
+                }
+                case NEWT_KEY_ESCAPE:
+                        if (!ui__dialog_yesno("Do you really want to exit?"))
+                                continue;
+                        /* Fall thru */
+                default:
+                        goto out_free_stack;
+                }
+
+                if (browser->selection->sym != NULL &&
+                    !browser->selection->map->dso->annotate_warned &&
+                    asprintf(&options[nr_options], "Annotate %s",
+                             browser->selection->sym->name) > 0)
+                        annotate = nr_options++;
+
+                if (thread != NULL &&
+                    asprintf(&options[nr_options], "Zoom %s %s(%d) thread",
+                             (thread_filter ? "out of" : "into"),
+                             (thread->comm_set ? thread->comm : ""),
+                             thread->pid) > 0)
+                        zoom_thread = nr_options++;
+
+                if (dso != NULL &&
+                    asprintf(&options[nr_options], "Zoom %s %s DSO",
+                             (dso_filter ? "out of" : "into"),
+                             (dso->kernel ? "the Kernel" : dso->short_name)) > 0)
+                        zoom_dso = nr_options++;
+
+                if (browser->selection->map != NULL &&
+                    asprintf(&options[nr_options], "Browse map details") > 0)
+                        browse_map = nr_options++;
+
+                options[nr_options++] = (char *)"Exit";
+
+                choice = ui__popup_menu(nr_options, options);
+
+                for (i = 0; i < nr_options - 1; ++i)
+                        free(options[i]);
+
+                if (choice == nr_options - 1)
+                        break;
+
+                if (choice == -1)
+                        continue;
+
+                if (choice == annotate) {
+                        struct hist_entry *he;
+do_annotate:
+                        if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) {
+                                browser->selection->map->dso->annotate_warned = 1;
+                                ui_helpline__puts("No vmlinux file found, can't "
+                                                 "annotate with just a "
+                                                 "kallsyms file");
+                                continue;
+                        }
+
+                        he = hist_browser__selected_entry(browser);
+                        if (he == NULL)
+                                continue;
+
+                        hist_entry__tui_annotate(he);
+                } else if (choice == browse_map)
+                        map__browse(browser->selection->map);
+                else if (choice == zoom_dso) {
+zoom_dso:
+                        if (dso_filter) {
+                                pstack__remove(fstack, &dso_filter);
+zoom_out_dso:
+                                ui_helpline__pop();
+                                dso_filter = NULL;
+                        } else {
+                                if (dso == NULL)
+                                        continue;
+                                ui_helpline__fpush("To zoom out press <- or -> + \"Zoom out of %s DSO\"",
+                                                   dso->kernel ? "the Kernel" : dso->short_name);
+                                dso_filter = dso;
+                                pstack__push(fstack, &dso_filter);
+                        }
+                        hists__filter_by_dso(self, dso_filter);
+                        hists__browser_title(self, msg, sizeof(msg), ev_name,
+                                             dso_filter, thread_filter);
+                        hist_browser__reset(browser);
+                } else if (choice == zoom_thread) {
+zoom_thread:
+                        if (thread_filter) {
+                                pstack__remove(fstack, &thread_filter);
+zoom_out_thread:
+                                ui_helpline__pop();
+                                thread_filter = NULL;
+                        } else {
+                                ui_helpline__fpush("To zoom out press <- or -> + \"Zoom out of %s(%d) thread\"",
+                                                   thread->comm_set ? thread->comm : "",
+                                                   thread->pid);
+                                thread_filter = thread;
+                                pstack__push(fstack, &thread_filter);
+                        }
+                        hists__filter_by_thread(self, thread_filter);
+                        hists__browser_title(self, msg, sizeof(msg), ev_name,
+                                             dso_filter, thread_filter);
+                        hist_browser__reset(browser);
+                }
+        }
+out_free_stack:
+        pstack__delete(fstack);
+out:
+        hist_browser__delete(browser);
+        return key;
+}
+
+int hists__tui_browse_tree(struct rb_root *self, const char *help)
+{
+        struct rb_node *first = rb_first(self), *nd = first, *next;
+        int key = 0;
+
+        while (nd) {
+                struct hists *hists = rb_entry(nd, struct hists, rb_node);
+                const char *ev_name = __event_name(hists->type, hists->config);
+
+                key = hists__browse(hists, help, ev_name);
+                switch (key) {
+                case NEWT_KEY_TAB:
+                        next = rb_next(nd);
+                        if (next)
+                                nd = next;
+                        break;
+                case NEWT_KEY_UNTAB:
+                        if (nd == first)
+                                continue;
+                        nd = rb_prev(nd);
+                default:
+                        return key;
+                }
+        }
+
+        return key;
+}
diff --git a/tools/perf/util/ui/browsers/map.c b/tools/perf/util/ui/browsers/map.c
new file mode 100644
index 000000000000..e35437dfa5b4
--- /dev/null
+++ b/tools/perf/util/ui/browsers/map.c
@@ -0,0 +1,155 @@
+#include "../libslang.h"
+#include <elf.h>
+#include <sys/ttydefaults.h>
+#include <ctype.h>
+#include <string.h>
+#include <linux/bitops.h>
+#include "../../debug.h"
+#include "../../symbol.h"
+#include "../browser.h"
+#include "../helpline.h"
+#include "map.h"
+
+static int ui_entry__read(const char *title, char *bf, size_t size, int width)
+{
+        struct newtExitStruct es;
+        newtComponent form, entry;
+        const char *result;
+        int err = -1;
+
+        newtCenteredWindow(width, 1, title);
+        form = newtForm(NULL, NULL, 0);
+        if (form == NULL)
+                return -1;
+
+        entry = newtEntry(0, 0, "0x", width, &result, NEWT_FLAG_SCROLL);
+        if (entry == NULL)
+                goto out_free_form;
+
+        newtFormAddComponent(form, entry);
+        newtFormAddHotKey(form, NEWT_KEY_ENTER);
+        newtFormAddHotKey(form, NEWT_KEY_ESCAPE);
+        newtFormAddHotKey(form, NEWT_KEY_LEFT);
+        newtFormAddHotKey(form, CTRL('c'));
+        newtFormRun(form, &es);
+
+        if (result != NULL) {
+                strncpy(bf, result, size);
+                err = 0;
+        }
+out_free_form:
+        newtPopWindow();
+        newtFormDestroy(form);
+        return 0;
+}
+
+struct map_browser {
+        struct ui_browser b;
+        struct map        *map;
+        u8                addrlen;
+};
+
+static void map_browser__write(struct ui_browser *self, void *nd, int row)
+{
+        struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+        struct map_browser *mb = container_of(self, struct map_browser, b);
+        bool current_entry = ui_browser__is_current_entry(self, row);
+        int width;
+
+        ui_browser__set_percent_color(self, 0, current_entry);
+        slsmg_printf("%*llx %*llx %c ",
+                     mb->addrlen, sym->start, mb->addrlen, sym->end,
+                     sym->binding == STB_GLOBAL ? 'g' :
+                     sym->binding == STB_LOCAL  ? 'l' : 'w');
+        width = self->width - ((mb->addrlen * 2) + 4);
+        if (width > 0)
+                slsmg_write_nstring(sym->name, width);
+}
+
+/* FIXME uber-kludgy, see comment on cmd_report... */
+static u32 *symbol__browser_index(struct symbol *self)
+{
+        return ((void *)self) - sizeof(struct rb_node) - sizeof(u32);
+}
+
+static int map_browser__search(struct map_browser *self)
+{
+        char target[512];
+        struct symbol *sym;
+        int err = ui_entry__read("Search by name/addr", target, sizeof(target), 40);
+
+        if (err)
+                return err;
+
+        if (target[0] == '0' && tolower(target[1]) == 'x') {
+                u64 addr = strtoull(target, NULL, 16);
+                sym = map__find_symbol(self->map, addr, NULL);
+        } else
+                sym = map__find_symbol_by_name(self->map, target, NULL);
+
+        if (sym != NULL) {
+                u32 *idx = symbol__browser_index(sym);
+
+                self->b.top = &sym->rb_node;
+                self->b.index = self->b.top_idx = *idx;
+        } else
+                ui_helpline__fpush("%s not found!", target);
+
+        return 0;
+}
+
+static int map_browser__run(struct map_browser *self)
+{
+        int key;
+
+        if (ui_browser__show(&self->b, self->map->dso->long_name,
+                             "Press <- or ESC to exit, %s / to search",
+                             verbose ? "" : "restart with -v to use") < 0)
+                return -1;
+
+        if (verbose)
+                ui_browser__add_exit_key(&self->b, '/');
+
+        while (1) {
+                key = ui_browser__run(&self->b);
+
+                if (verbose && key == '/')
+                        map_browser__search(self);
+                else
+                        break;
+        }
+
+        ui_browser__hide(&self->b);
+        return key;
+}
+
+int map__browse(struct map *self)
+{
+        struct map_browser mb = {
+                .b = {
+                        .entries = &self->dso->symbols[self->type],
+                        .refresh = ui_browser__rb_tree_refresh,
+                        .seek    = ui_browser__rb_tree_seek,
+                        .write   = map_browser__write,
+                },
+                .map = self,
+        };
+        struct rb_node *nd;
+        char tmp[BITS_PER_LONG / 4];
+        u64 maxaddr = 0;
+
+        for (nd = rb_first(mb.b.entries); nd; nd = rb_next(nd)) {
+                struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
+
+                if (maxaddr < pos->end)
+                        maxaddr = pos->end;
+                if (verbose) {
+                        u32 *idx = symbol__browser_index(pos);
+                        *idx = mb.b.nr_entries;
+                }
+                ++mb.b.nr_entries;
+        }
+
+        mb.addrlen = snprintf(tmp, sizeof(tmp), "%llx", maxaddr);
+        return map_browser__run(&mb);
+}
diff --git a/tools/perf/util/ui/browsers/map.h b/tools/perf/util/ui/browsers/map.h
new file mode 100644
index 000000000000..df8581a43e17
--- /dev/null
+++ b/tools/perf/util/ui/browsers/map.h
@@ -0,0 +1,6 @@
+#ifndef _PERF_UI_MAP_BROWSER_H_
+#define _PERF_UI_MAP_BROWSER_H_ 1
+struct map;
+
+int map__browse(struct map *self);
+#endif /* _PERF_UI_MAP_BROWSER_H_ */
diff --git a/tools/perf/util/ui/helpline.c b/tools/perf/util/ui/helpline.c
new file mode 100644
index 000000000000..8d79daa4458a
--- /dev/null
+++ b/tools/perf/util/ui/helpline.c
@@ -0,0 +1,69 @@
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <stdlib.h>
+#include <newt.h>
+
+#include "../debug.h"
+#include "helpline.h"
+
+void ui_helpline__pop(void)
+{
+        newtPopHelpLine();
+}
+
+void ui_helpline__push(const char *msg)
+{
+        newtPushHelpLine(msg);
+}
+
+void ui_helpline__vpush(const char *fmt, va_list ap)
+{
+        char *s;
+
+        if (vasprintf(&s, fmt, ap) < 0)
+                vfprintf(stderr, fmt, ap);
+        else {
+                ui_helpline__push(s);
+                free(s);
+        }
+}
+
+void ui_helpline__fpush(const char *fmt, ...)
+{
+        va_list ap;
+
+        va_start(ap, fmt);
+        ui_helpline__vpush(fmt, ap);
+        va_end(ap);
+}
+
+void ui_helpline__puts(const char *msg)
+{
+        ui_helpline__pop();
+        ui_helpline__push(msg);
+}
+
+void ui_helpline__init(void)
+{
+        ui_helpline__puts(" ");
+}
+
+char ui_helpline__last_msg[1024];
+
+int ui_helpline__show_help(const char *format, va_list ap)
+{
+        int ret;
+        static int backlog;
+
+        ret = vsnprintf(ui_helpline__last_msg + backlog,
+                        sizeof(ui_helpline__last_msg) - backlog, format, ap);
+        backlog += ret;
+
+        if (ui_helpline__last_msg[backlog - 1] == '\n') {
+                ui_helpline__puts(ui_helpline__last_msg);
+                newtRefresh();
+                backlog = 0;
+        }
+
+        return ret;
+}
diff --git a/tools/perf/util/ui/helpline.h b/tools/perf/util/ui/helpline.h
new file mode 100644
index 000000000000..ab6028d0c401
--- /dev/null
+++ b/tools/perf/util/ui/helpline.h
@@ -0,0 +1,11 @@
+#ifndef _PERF_UI_HELPLINE_H_
+#define _PERF_UI_HELPLINE_H_ 1
+
+void ui_helpline__init(void);
+void ui_helpline__pop(void);
+void ui_helpline__push(const char *msg);
+void ui_helpline__vpush(const char *fmt, va_list ap);
+void ui_helpline__fpush(const char *fmt, ...);
+void ui_helpline__puts(const char *msg);
+
+#endif /* _PERF_UI_HELPLINE_H_ */
diff --git a/tools/perf/util/ui/libslang.h b/tools/perf/util/ui/libslang.h
new file mode 100644
index 000000000000..5623da8e8080
--- /dev/null
+++ b/tools/perf/util/ui/libslang.h
@@ -0,0 +1,27 @@
+#ifndef _PERF_UI_SLANG_H_
+#define _PERF_UI_SLANG_H_ 1
+/*
+ * slang versions <= 2.0.6 have a "#if HAVE_LONG_LONG" that breaks
+ * the build if it isn't defined. Use the equivalent one that glibc
+ * has on features.h.
+ */
+#include <features.h>
+#ifndef HAVE_LONG_LONG
+#define HAVE_LONG_LONG __GLIBC_HAVE_LONG_LONG
+#endif
+#include <slang.h>
+
+#if SLANG_VERSION < 20104
+#define slsmg_printf(msg, args...) \
+        SLsmg_printf((char *)msg, ##args)
+#define slsmg_write_nstring(msg, len) \
+        SLsmg_write_nstring((char *)msg, len)
+#define sltt_set_color(obj, name, fg, bg) \
+        SLtt_set_color(obj,(char *)name, (char *)fg, (char *)bg)
+#else
+#define slsmg_printf SLsmg_printf
+#define slsmg_write_nstring SLsmg_write_nstring
+#define sltt_set_color SLtt_set_color
+#endif
+
+#endif /* _PERF_UI_SLANG_H_ */
diff --git a/tools/perf/util/ui/progress.c b/tools/perf/util/ui/progress.c
new file mode 100644
index 000000000000..d7fc399d36b3
--- /dev/null
+++ b/tools/perf/util/ui/progress.c
@@ -0,0 +1,60 @@
+#include <stdlib.h>
+#include <newt.h>
+#include "../cache.h"
+#include "progress.h"
+
+struct ui_progress {
+        newtComponent form, scale;
+};
+
+struct ui_progress *ui_progress__new(const char *title, u64 total)
+{
+        struct ui_progress *self = malloc(sizeof(*self));
+
+        if (self != NULL) {
+                int cols;
+
+                if (use_browser <= 0)
+                        return self;
+                newtGetScreenSize(&cols, NULL);
+                cols -= 4;
+                newtCenteredWindow(cols, 1, title);
+                self->form  = newtForm(NULL, NULL, 0);
+                if (self->form == NULL)
+                        goto out_free_self;
+                self->scale = newtScale(0, 0, cols, total);
+                if (self->scale == NULL)
+                        goto out_free_form;
+                newtFormAddComponent(self->form, self->scale);
+                newtRefresh();
+        }
+
+        return self;
+
+out_free_form:
+        newtFormDestroy(self->form);
+out_free_self:
+        free(self);
+        return NULL;
+}
+
+void ui_progress__update(struct ui_progress *self, u64 curr)
+{
+        /*
+         * FIXME: We should have a per UI backend way of showing progress,
+         * stdio will just show a percentage as NN%, etc.
+         */
+        if (use_browser <= 0)
+                return;
+        newtScaleSet(self->scale, curr);
+        newtRefresh();
+}
+
+void ui_progress__delete(struct ui_progress *self)
+{
+        if (use_browser > 0) {
+                newtFormDestroy(self->form);
+                newtPopWindow();
+        }
+        free(self);
+}
diff --git a/tools/perf/util/ui/progress.h b/tools/perf/util/ui/progress.h
new file mode 100644
index 000000000000..a3820a0beb5b
--- /dev/null
+++ b/tools/perf/util/ui/progress.h
@@ -0,0 +1,11 @@
+#ifndef _PERF_UI_PROGRESS_H_
+#define _PERF_UI_PROGRESS_H_ 1
+
+struct ui_progress;
+
+struct ui_progress *ui_progress__new(const char *title, u64 total);
+void ui_progress__delete(struct ui_progress *self);
+
+void ui_progress__update(struct ui_progress *self, u64 curr);
+
+#endif
diff --git a/tools/perf/util/ui/setup.c b/tools/perf/util/ui/setup.c
new file mode 100644
index 000000000000..662085032eb7
--- /dev/null
+++ b/tools/perf/util/ui/setup.c
@@ -0,0 +1,42 @@
+#include <newt.h>
+#include <signal.h>
+#include <stdbool.h>
+
+#include "../cache.h"
+#include "../debug.h"
+#include "browser.h"
+#include "helpline.h"
+
+static void newt_suspend(void *d __used)
+{
+        newtSuspend();
+        raise(SIGTSTP);
+        newtResume();
+}
+
+void setup_browser(void)
+{
+        if (!isatty(1) || !use_browser || dump_trace) {
+                use_browser = 0;
+                setup_pager();
+                return;
+        }
+
+        use_browser = 1;
+        newtInit();
+        newtCls();
+        newtSetSuspendCallback(newt_suspend, NULL);
+        ui_helpline__init();
+        ui_browser__init();
+}
+
+void exit_browser(bool wait_for_ok)
+{
+        if (use_browser > 0) {
+                if (wait_for_ok) {
+                        char title[] = "Fatal Error", ok[] = "Ok";
+                        newtWinMessage(title, ok, ui_helpline__last_msg);
+                }
+                newtFinished();
+        }
+}
diff --git a/tools/perf/util/ui/util.c b/tools/perf/util/ui/util.c
new file mode 100644
index 000000000000..056c69521a38
--- /dev/null
+++ b/tools/perf/util/ui/util.c
@@ -0,0 +1,113 @@
+#include <newt.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdbool.h>
+#include <string.h>
+#include <sys/ttydefaults.h>
+
+#include "../cache.h"
+#include "../debug.h"
+#include "browser.h"
+#include "helpline.h"
+#include "util.h"
+
+static void newt_form__set_exit_keys(newtComponent self)
+{
+        newtFormAddHotKey(self, NEWT_KEY_LEFT);
+        newtFormAddHotKey(self, NEWT_KEY_ESCAPE);
+        newtFormAddHotKey(self, 'Q');
+        newtFormAddHotKey(self, 'q');
+        newtFormAddHotKey(self, CTRL('c'));
+}
+
+static newtComponent newt_form__new(void)
+{
+        newtComponent self = newtForm(NULL, NULL, 0);
+        if (self)
+                newt_form__set_exit_keys(self);
+        return self;
+}
+
+int ui__popup_menu(int argc, char * const argv[])
+{
+        struct newtExitStruct es;
+        int i, rc = -1, max_len = 5;
+        newtComponent listbox, form = newt_form__new();
+
+        if (form == NULL)
+                return -1;
+
+        listbox = newtListbox(0, 0, argc, NEWT_FLAG_RETURNEXIT);
+        if (listbox == NULL)
+                goto out_destroy_form;
+
+        newtFormAddComponent(form, listbox);
+
+        for (i = 0; i < argc; ++i) {
+                int len = strlen(argv[i]);
+                if (len > max_len)
+                        max_len = len;
+                if (newtListboxAddEntry(listbox, argv[i], (void *)(long)i))
+                        goto out_destroy_form;
+        }
+
+        newtCenteredWindow(max_len, argc, NULL);
+        newtFormRun(form, &es);
+        rc = newtListboxGetCurrent(listbox) - NULL;
+        if (es.reason == NEWT_EXIT_HOTKEY)
+                rc = -1;
+        newtPopWindow();
+out_destroy_form:
+        newtFormDestroy(form);
+        return rc;
+}
+
+int ui__help_window(const char *text)
+{
+        struct newtExitStruct es;
+        newtComponent tb, form = newt_form__new();
+        int rc = -1;
+        int max_len = 0, nr_lines = 0;
+        const char *t;
+
+        if (form == NULL)
+                return -1;
+
+        t = text;
+        while (1) {
+                const char *sep = strchr(t, '\n');
+                int len;
+
+                if (sep == NULL)
+                        sep = strchr(t, '\0');
+                len = sep - t;
+                if (max_len < len)
+                        max_len = len;
+                ++nr_lines;
+                if (*sep == '\0')
+                        break;
+                t = sep + 1;
+        }
+
+        tb = newtTextbox(0, 0, max_len, nr_lines, 0);
+        if (tb == NULL)
+                goto out_destroy_form;
+
+        newtTextboxSetText(tb, text);
+        newtFormAddComponent(form, tb);
+        newtCenteredWindow(max_len, nr_lines, NULL);
+        newtFormRun(form, &es);
+        newtPopWindow();
+        rc = 0;
+out_destroy_form:
+        newtFormDestroy(form);
+        return rc;
+}
+
+static const char yes[] = "Yes", no[] = "No";
+
+bool ui__dialog_yesno(const char *msg)
+{
+        /* newtWinChoice should really be accepting const char pointers... */
+        return newtWinChoice(NULL, (char *)yes, (char *)no, (char *)msg) == 1;
+}
diff --git a/tools/perf/util/ui/util.h b/tools/perf/util/ui/util.h
new file mode 100644
index 000000000000..afcbc1d99531
--- /dev/null
+++ b/tools/perf/util/ui/util.h
@@ -0,0 +1,10 @@
+#ifndef _PERF_UI_UTIL_H_
+#define _PERF_UI_UTIL_H_ 1
+
+#include <stdbool.h>
+
+int ui__popup_menu(int argc, char * const argv[]);
+int ui__help_window(const char *text);
+bool ui__dialog_yesno(const char *msg);
+
+#endif /* _PERF_UI_UTIL_H_ */
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index f380fed74359..7562707ddd1c 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -266,19 +266,6 @@ bool strglobmatch(const char *str, const char *pat);
 bool strlazymatch(const char *str, const char *pat);
 unsigned long convert_unit(unsigned long value, char *unit);
 
-#ifndef ESC
-#define ESC 27
-#endif
-
-static inline bool is_exit_key(int key)
-{
-        char up;
-        if (key == CTRL('c') || key == ESC)
-                return true;
-        up = toupper(key);
-        return up == 'Q';
-}
-
 #define _STR(x) #x
 #define STR(x) _STR(x)