Merge commit 'v2.6.31-rc8' into core/locking

Merge reason: we were on -rc4, move to -rc8 before applying
              a new batch of locking infrastructure changes.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

20 files changed, 1347 insertions, 163 deletions

diff --git a/tools/perf/Documentation/Makefile b/tools/perf/Documentation/Makefile
index 5457192e1b41..bdd3b7ecad0a 100644
--- a/tools/perf/Documentation/Makefile
+++ b/tools/perf/Documentation/Makefile
@@ -35,7 +35,7 @@ man7dir=$(mandir)/man7
 # DESTDIR=
 
 ASCIIDOC=asciidoc
-ASCIIDOC_EXTRA =
+ASCIIDOC_EXTRA = --unsafe
 MANPAGE_XSL = manpage-normal.xsl
 XMLTO_EXTRA =
 INSTALL?=install
diff --git a/tools/perf/Documentation/examples.txt b/tools/perf/Documentation/examples.txt
new file mode 100644
index 000000000000..8eb6c489fb15
--- /dev/null
+++ b/tools/perf/Documentation/examples.txt
@@ -0,0 +1,225 @@
+
+		------------------------------
+		****** perf by examples ******
+		------------------------------
+
+[ From an e-mail by Ingo Molnar, http://lkml.org/lkml/2009/8/4/346 ]
+
+
+First, discovery/enumeration of available counters can be done via
+'perf list':
+
+titan:~> perf list
+  [...]
+  kmem:kmalloc                             [Tracepoint event]
+  kmem:kmem_cache_alloc                    [Tracepoint event]
+  kmem:kmalloc_node                        [Tracepoint event]
+  kmem:kmem_cache_alloc_node               [Tracepoint event]
+  kmem:kfree                               [Tracepoint event]
+  kmem:kmem_cache_free                     [Tracepoint event]
+  kmem:mm_page_free_direct                 [Tracepoint event]
+  kmem:mm_pagevec_free                     [Tracepoint event]
+  kmem:mm_page_alloc                       [Tracepoint event]
+  kmem:mm_page_alloc_zone_locked           [Tracepoint event]
+  kmem:mm_page_pcpu_drain                  [Tracepoint event]
+  kmem:mm_page_alloc_extfrag               [Tracepoint event]
+
+Then any (or all) of the above event sources can be activated and
+measured. For example the page alloc/free properties of a 'hackbench
+run' are:
+
+ titan:~> perf stat -e kmem:mm_page_pcpu_drain -e kmem:mm_page_alloc
+ -e kmem:mm_pagevec_free -e kmem:mm_page_free_direct ./hackbench 10
+ Time: 0.575
+
+ Performance counter stats for './hackbench 10':
+
+          13857  kmem:mm_page_pcpu_drain
+          27576  kmem:mm_page_alloc
+           6025  kmem:mm_pagevec_free
+          20934  kmem:mm_page_free_direct
+
+    0.613972165  seconds time elapsed
+
+You can observe the statistical properties as well, by using the
+'repeat the workload N times' feature of perf stat:
+
+ titan:~> perf stat --repeat 5 -e kmem:mm_page_pcpu_drain -e
+   kmem:mm_page_alloc -e kmem:mm_pagevec_free -e
+   kmem:mm_page_free_direct ./hackbench 10
+ Time: 0.627
+ Time: 0.644
+ Time: 0.564
+ Time: 0.559
+ Time: 0.626
+
+ Performance counter stats for './hackbench 10' (5 runs):
+
+          12920  kmem:mm_page_pcpu_drain    ( +-   3.359% )
+          25035  kmem:mm_page_alloc         ( +-   3.783% )
+           6104  kmem:mm_pagevec_free       ( +-   0.934% )
+          18376  kmem:mm_page_free_direct   ( +-   4.941% )
+
+    0.643954516  seconds time elapsed   ( +-   2.363% )
+
+Furthermore, these tracepoints can be used to sample the workload as
+well. For example the page allocations done by a 'git gc' can be
+captured the following way:
+
+ titan:~/git> perf record -f -e kmem:mm_page_alloc -c 1 ./git gc
+ Counting objects: 1148, done.
+ Delta compression using up to 2 threads.
+ Compressing objects: 100% (450/450), done.
+ Writing objects: 100% (1148/1148), done.
+ Total 1148 (delta 690), reused 1148 (delta 690)
+ [ perf record: Captured and wrote 0.267 MB perf.data (~11679 samples) ]
+
+To check which functions generated page allocations:
+
+ titan:~/git> perf report
+ # Samples: 10646
+ #
+ # Overhead          Command               Shared Object
+ # ........  ...............  ..........................
+ #
+    23.57%       git-repack  /lib64/libc-2.5.so
+    21.81%              git  /lib64/libc-2.5.so
+    14.59%              git  ./git
+    11.79%       git-repack  ./git
+     7.12%              git  /lib64/ld-2.5.so
+     3.16%       git-repack  /lib64/libpthread-2.5.so
+     2.09%       git-repack  /bin/bash
+     1.97%               rm  /lib64/libc-2.5.so
+     1.39%               mv  /lib64/ld-2.5.so
+     1.37%               mv  /lib64/libc-2.5.so
+     1.12%       git-repack  /lib64/ld-2.5.so
+     0.95%               rm  /lib64/ld-2.5.so
+     0.90%  git-update-serv  /lib64/libc-2.5.so
+     0.73%  git-update-serv  /lib64/ld-2.5.so
+     0.68%             perf  /lib64/libpthread-2.5.so
+     0.64%       git-repack  /usr/lib64/libz.so.1.2.3
+
+Or to see it on a more finegrained level:
+
+titan:~/git> perf report --sort comm,dso,symbol
+# Samples: 10646
+#
+# Overhead          Command               Shared Object  Symbol
+# ........  ...............  ..........................  ......
+#
+     9.35%       git-repack  ./git                       [.] insert_obj_hash
+     9.12%              git  ./git                       [.] insert_obj_hash
+     7.31%              git  /lib64/libc-2.5.so          [.] memcpy
+     6.34%       git-repack  /lib64/libc-2.5.so          [.] _int_malloc
+     6.24%       git-repack  /lib64/libc-2.5.so          [.] memcpy
+     5.82%       git-repack  /lib64/libc-2.5.so          [.] __GI___fork
+     5.47%              git  /lib64/libc-2.5.so          [.] _int_malloc
+     2.99%              git  /lib64/libc-2.5.so          [.] memset
+
+Furthermore, call-graph sampling can be done too, of page
+allocations - to see precisely what kind of page allocations there
+are:
+
+ titan:~/git> perf record -f -g -e kmem:mm_page_alloc -c 1 ./git gc
+ Counting objects: 1148, done.
+ Delta compression using up to 2 threads.
+ Compressing objects: 100% (450/450), done.
+ Writing objects: 100% (1148/1148), done.
+ Total 1148 (delta 690), reused 1148 (delta 690)
+ [ perf record: Captured and wrote 0.963 MB perf.data (~42069 samples) ]
+
+ titan:~/git> perf report -g
+ # Samples: 10686
+ #
+ # Overhead          Command               Shared Object
+ # ........  ...............  ..........................
+ #
+    23.25%       git-repack  /lib64/libc-2.5.so
+                |
+                |--50.00%-- _int_free
+                |
+                |--37.50%-- __GI___fork
+                |          make_child
+                |
+                |--12.50%-- ptmalloc_unlock_all2
+                |          make_child
+                |
+                 --6.25%-- __GI_strcpy
+    21.61%              git  /lib64/libc-2.5.so
+                |
+                |--30.00%-- __GI_read
+                |          |
+                |           --83.33%-- git_config_from_file
+                |                     git_config
+                |                     |
+   [...]
+
+Or you can observe the whole system's page allocations for 10
+seconds:
+
+titan:~/git> perf stat -a -e kmem:mm_page_pcpu_drain -e
+kmem:mm_page_alloc -e kmem:mm_pagevec_free -e
+kmem:mm_page_free_direct sleep 10
+
+ Performance counter stats for 'sleep 10':
+
+         171585  kmem:mm_page_pcpu_drain
+         322114  kmem:mm_page_alloc
+          73623  kmem:mm_pagevec_free
+         254115  kmem:mm_page_free_direct
+
+   10.000591410  seconds time elapsed
+
+Or observe how fluctuating the page allocations are, via statistical
+analysis done over ten 1-second intervals:
+
+ titan:~/git> perf stat --repeat 10 -a -e kmem:mm_page_pcpu_drain -e
+   kmem:mm_page_alloc -e kmem:mm_pagevec_free -e
+   kmem:mm_page_free_direct sleep 1
+
+ Performance counter stats for 'sleep 1' (10 runs):
+
+          17254  kmem:mm_page_pcpu_drain    ( +-   3.709% )
+          34394  kmem:mm_page_alloc         ( +-   4.617% )
+           7509  kmem:mm_pagevec_free       ( +-   4.820% )
+          25653  kmem:mm_page_free_direct   ( +-   3.672% )
+
+    1.058135029  seconds time elapsed   ( +-   3.089% )
+
+Or you can annotate the recorded 'git gc' run on a per symbol basis
+and check which instructions/source-code generated page allocations:
+
+ titan:~/git> perf annotate __GI___fork
+ ------------------------------------------------
+  Percent |      Source code & Disassembly of libc-2.5.so
+ ------------------------------------------------
+          :
+          :
+          :      Disassembly of section .plt:
+          :      Disassembly of section .text:
+          :
+          :      00000031a2e95560 <__fork>:
+ [...]
+     0.00 :        31a2e95602:   b8 38 00 00 00          mov    $0x38,%eax
+     0.00 :        31a2e95607:   0f 05                   syscall
+    83.42 :        31a2e95609:   48 3d 00 f0 ff ff       cmp    $0xfffffffffffff000,%rax
+     0.00 :        31a2e9560f:   0f 87 4d 01 00 00       ja     31a2e95762 <__fork+0x202>
+     0.00 :        31a2e95615:   85 c0                   test   %eax,%eax
+
+( this shows that 83.42% of __GI___fork's page allocations come from
+  the 0x38 system call it performs. )
+
+etc. etc. - a lot more is possible. I could list a dozen of
+other different usecases straight away - neither of which is
+possible via /proc/vmstat.
+
+/proc/vmstat is not in the same league really, in terms of
+expressive power of system analysis and performance
+analysis.
+
+All that the above results needed were those new tracepoints
+in include/tracing/events/kmem.h.
+
+	Ingo
+
+
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index 1dbc1eeb4c01..6be696b0a2bb 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -29,13 +29,67 @@ OPTIONS
 	Select the PMU event. Selection can be a symbolic event name
 	(use 'perf list' to list all events) or a raw PMU
 	event (eventsel+umask) in the form of rNNN where NNN is a
-	 hexadecimal event descriptor.
+	hexadecimal event descriptor.
 
 -a::
-        system-wide collection
+        System-wide collection.
 
 -l::
-        scale counter values
+        Scale counter values.
+
+-p::
+--pid=::
+	Record events on existing pid.
+
+-r::
+--realtime=::
+	Collect data with this RT SCHED_FIFO priority.
+-A::
+--append::
+	Append to the output file to do incremental profiling.
+
+-f::
+--force::
+	Overwrite existing data file.
+
+-c::
+--count=::
+	Event period to sample.
+
+-o::
+--output=::
+	Output file name.
+
+-i::
+--inherit::
+	Child tasks inherit counters.
+-F::
+--freq=::
+	Profile at this frequency.
+
+-m::
+--mmap-pages=::
+	Number of mmap data pages.
+
+-g::
+--call-graph::
+	Do call-graph (stack chain/backtrace) recording.
+
+-v::
+--verbose::
+	Be more verbose (show counter open errors, etc).
+
+-s::
+--stat::
+	Per thread counts.
+
+-d::
+--data::
+	Sample addresses.
+
+-n::
+--no-samples::
+	Don't sample.
 
 SEE ALSO
 --------
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 0d74346d21ab..484080dd5b6f 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -40,7 +40,7 @@ OPTIONS
 -a::
         system-wide collection
 
--S::
+-c::
         scale counter values
 
 EXAMPLES
diff --git a/tools/perf/Documentation/perf-top.txt b/tools/perf/Documentation/perf-top.txt
index 539d01289725..4a7d558dc309 100644
--- a/tools/perf/Documentation/perf-top.txt
+++ b/tools/perf/Documentation/perf-top.txt
@@ -3,36 +3,122 @@ perf-top(1)
 
 NAME
 ----
-perf-top - Run a command and profile it
+perf-top - System profiling tool.
 
 SYNOPSIS
 --------
 [verse]
-'perf top' [-e <EVENT> | --event=EVENT] [-l] [-a] <command>
+'perf top' [-e <EVENT> | --event=EVENT] [<options>]
 
 DESCRIPTION
 -----------
-This command runs a command and gathers a performance counter profile
-from it.
+This command generates and displays a performance counter profile in realtime.
 
 
 OPTIONS
 -------
-<command>...::
-	Any command you can specify in a shell.
+-a::
+--all-cpus::
+        System-wide collection.  (default)
+
+-c <count>::
+--count=<count>::
+	Event period to sample.
+
+-C <cpu>::
+--CPU=<cpu>::
+	CPU to profile.
+
+-d <seconds>::
+--delay=<seconds>::
+	Number of seconds to delay between refreshes.
 
--e::
---event=::
+-e <event>::
+--event=<event>::
 	Select the PMU event. Selection can be a symbolic event name
 	(use 'perf list' to list all events) or a raw PMU
 	event (eventsel+umask) in the form of rNNN where NNN is a
-	 hexadecimal event descriptor.
+	hexadecimal event descriptor.
 
--a::
-        system-wide collection
+-E <entries>::
+--entries=<entries>::
+	Display this many functions.
+
+-f <count>::
+--count-filter=<count>::
+	Only display functions with more events than this.
+
+-F <freq>::
+--freq=<freq>::
+	Profile at this frequency.
+
+-i::
+--inherit::
+	Child tasks inherit counters, only makes sens with -p option.
+
+-k <path>::
+--vmlinux=<path>::
+	Path to vmlinux.  Required for annotation functionality.
+
+-m <pages>::
+--mmap-pages=<pages>::
+	Number of mmapped data pages.
+
+-p <pid>::
+--pid=<pid>::
+	Profile events on existing pid.
+
+-r <priority>::
+--realtime=<priority>::
+	Collect data with this RT SCHED_FIFO priority.
+
+-s <symbol>::
+--sym-annotate=<symbol>::
+        Annotate this symbol.  Requires -k option.
+
+-v::
+--verbose::
+	Be more verbose (show counter open errors, etc).
+
+-z::
+--zero::
+	Zero history across display updates.
+
+INTERACTIVE PROMPTING KEYS
+--------------------------
+
+[d]::
+	Display refresh delay.
+
+[e]::
+	Number of entries to display.
+
+[E]::
+	Event to display when multiple counters are active.
+
+[f]::
+	Profile display filter (>= hit count).
+
+[F]::
+	Annotation display filter (>= % of total).
+
+[s]::
+	Annotate symbol.
+
+[S]::
+	Stop annotation, return to full profile display.
+
+[w]::
+	Toggle between weighted sum and individual count[E]r profile.
+
+[z]::
+	Toggle event count zeroing across display updates.
+
+[qQ]::
+	Quit.
+
+Pressing any unmapped key displays a menu, and prompts for input.
 
--l::
-        scale counter values
 
 SEE ALSO
 --------
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index a5e9b876ca09..c045b4271e57 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -158,8 +158,10 @@ uname_P := $(shell sh -c 'uname -p 2>/dev/null || echo not')
 uname_V := $(shell sh -c 'uname -v 2>/dev/null || echo not')
 
 # If we're on a 64-bit kernel, use -m64
-ifneq ($(patsubst %64,%,$(uname_M)),$(uname_M))
-  M64 := -m64
+ifndef NO_64BIT
+	ifneq ($(patsubst %64,%,$(uname_M)),$(uname_M))
+	  M64 := -m64
+	endif
 endif
 
 # CFLAGS and LDFLAGS are for the users to override from the command line.
@@ -345,7 +347,6 @@ BUILTIN_OBJS += builtin-stat.o
 BUILTIN_OBJS += builtin-top.o
 
 PERFLIBS = $(LIB_FILE)
-EXTLIBS = -lbfd
 
 #
 # Platform specific tweaks
@@ -374,6 +375,39 @@ ifeq ($(uname_S),Darwin)
 	PTHREAD_LIBS =
 endif
 
+ifneq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ_MMAP, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
+	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel);
+endif
+
+ifdef NO_DEMANGLE
+	BASIC_CFLAGS += -DNO_DEMANGLE
+else
+	has_bfd := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd > /dev/null 2>&1 && echo y")
+
+	ifeq ($(has_bfd),y)
+		EXTLIBS += -lbfd
+	else
+		has_bfd_iberty := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty > /dev/null 2>&1 && echo y")
+		ifeq ($(has_bfd_iberty),y)
+			EXTLIBS += -lbfd -liberty
+		else
+			has_bfd_iberty_z := $(shell sh -c "(echo '\#include <bfd.h>'; echo 'int main(void) { bfd_demangle(0, 0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -lbfd -liberty -lz > /dev/null 2>&1 && echo y")
+			ifeq ($(has_bfd_iberty_z),y)
+				EXTLIBS += -lbfd -liberty -lz
+			else
+				has_cplus_demangle := $(shell sh -c "(echo 'extern char *cplus_demangle(const char *, int);'; echo 'int main(void) { cplus_demangle(0, 0); return 0; }') | $(CC) -x c - $(ALL_CFLAGS) -o /dev/null $(ALL_LDFLAGS) -liberty > /dev/null 2>&1 && echo y")
+				ifeq ($(has_cplus_demangle),y)
+					EXTLIBS += -liberty
+					BASIC_CFLAGS += -DHAVE_CPLUS_DEMANGLE
+				else
+					msg := $(warning No bfd.h/libbfd found, install binutils-dev[el] to gain symbol demangling)
+					BASIC_CFLAGS += -DNO_DEMANGLE
+				endif
+			endif
+		endif
+	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
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 1dba568e1941..5e17de984dc8 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -31,6 +31,7 @@ static char		*vmlinux = "vmlinux";
 static char		default_sort_order[] = "comm,symbol";
 static char		*sort_order = default_sort_order;
 
+static int		force;
 static int		input;
 static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
@@ -980,6 +981,13 @@ process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 		(void *)(long)(event->header.size),
 		event->fork.pid, event->fork.ppid);
 
+	/*
+	 * A thread clone will have the same PID for both
+	 * parent and child.
+	 */
+	if (thread == parent)
+		return 0;
+
 	if (!thread || !parent || thread__fork(thread, parent)) {
 		dprintf("problem processing PERF_EVENT_FORK, skipping event.\n");
 		return -1;
@@ -1327,6 +1335,11 @@ static int __cmd_annotate(void)
 		exit(-1);
 	}
 
+	if (!force && (stat.st_uid != geteuid())) {
+		fprintf(stderr, "file: %s not owned by current user\n", input_name);
+		exit(-1);
+	}
+
 	if (!stat.st_size) {
 		fprintf(stderr, "zero-sized file, nothing to do!\n");
 		exit(0);
@@ -1432,6 +1445,7 @@ static const struct option options[] = {
 		    "input file name"),
 	OPT_STRING('s', "symbol", &sym_hist_filter, "symbol",
 		    "symbol to annotate"),
+	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show symbol address, etc)"),
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
diff --git a/tools/perf/builtin-list.c b/tools/perf/builtin-list.c
index f990fa8a35c9..d88c6961274c 100644
--- a/tools/perf/builtin-list.c
+++ b/tools/perf/builtin-list.c
@@ -10,11 +10,12 @@
 
 #include "perf.h"
 
-#include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/cache.h"
 
 int cmd_list(int argc __used, const char **argv __used, const char *prefix __used)
 {
+	setup_pager();
 	print_events();
 	return 0;
 }
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 6da09928130f..89a5ddcd1ded 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -34,7 +34,9 @@ static int			output;
 static const char		*output_name			= "perf.data";
 static int			group				= 0;
 static unsigned int		realtime_prio			= 0;
+static int			raw_samples			= 0;
 static int			system_wide			= 0;
+static int			profile_cpu			= -1;
 static pid_t			target_pid			= -1;
 static int			inherit				= 1;
 static int			force				= 0;
@@ -203,46 +205,48 @@ static void sig_atexit(void)
 	kill(getpid(), signr);
 }
 
-static void pid_synthesize_comm_event(pid_t pid, int full)
+static pid_t pid_synthesize_comm_event(pid_t pid, int full)
 {
 	struct comm_event comm_ev;
 	char filename[PATH_MAX];
 	char bf[BUFSIZ];
-	int fd;
-	size_t size;
-	char *field, *sep;
+	FILE *fp;
+	size_t size = 0;
 	DIR *tasks;
 	struct dirent dirent, *next;
+	pid_t tgid = 0;
 
-	snprintf(filename, sizeof(filename), "/proc/%d/stat", pid);
+	snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
 
-	fd = open(filename, O_RDONLY);
-	if (fd < 0) {
+	fp = fopen(filename, "r");
+	if (fp == NULL) {
 		/*
 		 * We raced with a task exiting - just return:
 		 */
 		if (verbose)
 			fprintf(stderr, "couldn't open %s\n", filename);
-		return;
-	}
-	if (read(fd, bf, sizeof(bf)) < 0) {
-		fprintf(stderr, "couldn't read %s\n", filename);
-		exit(EXIT_FAILURE);
+		return 0;
 	}
-	close(fd);
 
-	/* 9027 (cat) R 6747 9027 6747 34816 9027 ... */
 	memset(&comm_ev, 0, sizeof(comm_ev));
-	field = strchr(bf, '(');
-	if (field == NULL)
-		goto out_failure;
-	sep = strchr(++field, ')');
-	if (sep == NULL)
-		goto out_failure;
-	size = sep - field;
-	memcpy(comm_ev.comm, field, size++);
-
-	comm_ev.pid = pid;
+	while (!comm_ev.comm[0] || !comm_ev.pid) {
+		if (fgets(bf, sizeof(bf), fp) == NULL)
+			goto out_failure;
+
+		if (memcmp(bf, "Name:", 5) == 0) {
+			char *name = bf + 5;
+			while (*name && isspace(*name))
+				++name;
+			size = strlen(name) - 1;
+			memcpy(comm_ev.comm, name, size++);
+		} else if (memcmp(bf, "Tgid:", 5) == 0) {
+			char *tgids = bf + 5;
+			while (*tgids && isspace(*tgids))
+				++tgids;
+			tgid = comm_ev.pid = atoi(tgids);
+		}
+	}
+
 	comm_ev.header.type = PERF_EVENT_COMM;
 	size = ALIGN(size, sizeof(u64));
 	comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size);
@@ -251,7 +255,7 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 		comm_ev.tid = pid;
 
 		write_output(&comm_ev, comm_ev.header.size);
-		return;
+		goto out_fclose;
 	}
 
 	snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
@@ -268,7 +272,10 @@ static void pid_synthesize_comm_event(pid_t pid, int full)
 		write_output(&comm_ev, comm_ev.header.size);
 	}
 	closedir(tasks);
-	return;
+
+out_fclose:
+	fclose(fp);
+	return tgid;
 
 out_failure:
 	fprintf(stderr, "couldn't get COMM and pgid, malformed %s\n",
@@ -276,7 +283,7 @@ out_failure:
 	exit(EXIT_FAILURE);
 }
 
-static void pid_synthesize_mmap_samples(pid_t pid)
+static void pid_synthesize_mmap_samples(pid_t pid, pid_t tgid)
 {
 	char filename[PATH_MAX];
 	FILE *fp;
@@ -328,7 +335,7 @@ static void pid_synthesize_mmap_samples(pid_t pid)
 			mmap_ev.len -= mmap_ev.start;
 			mmap_ev.header.size = (sizeof(mmap_ev) -
 					       (sizeof(mmap_ev.filename) - size));
-			mmap_ev.pid = pid;
+			mmap_ev.pid = tgid;
 			mmap_ev.tid = pid;
 
 			write_output(&mmap_ev, mmap_ev.header.size);
@@ -347,14 +354,14 @@ static void synthesize_all(void)
 
 	while (!readdir_r(proc, &dirent, &next) && next) {
 		char *end;
-		pid_t pid;
+		pid_t pid, tgid;
 
 		pid = strtol(dirent.d_name, &end, 10);
 		if (*end) /* only interested in proper numerical dirents */
 			continue;
 
-		pid_synthesize_comm_event(pid, 1);
-		pid_synthesize_mmap_samples(pid);
+		tgid = pid_synthesize_comm_event(pid, 1);
+		pid_synthesize_mmap_samples(pid, tgid);
 	}
 
 	closedir(proc);
@@ -392,7 +399,7 @@ static void create_counter(int counter, int cpu, pid_t pid)
 				  PERF_FORMAT_TOTAL_TIME_RUNNING |
 				  PERF_FORMAT_ID;
 
-	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
 
 	if (freq) {
 		attr->sample_type	|= PERF_SAMPLE_PERIOD;
@@ -412,6 +419,9 @@ static void create_counter(int counter, int cpu, pid_t pid)
 	if (call_graph)
 		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
 
+	if (raw_samples)
+		attr->sample_type	|= PERF_SAMPLE_RAW;
+
 	attr->mmap		= track;
 	attr->comm		= track;
 	attr->inherit		= (cpu < 0) && inherit;
@@ -425,6 +435,8 @@ try_again:
 
 		if (err == EPERM)
 			die("Permission error - are you root?\n");
+		else if (err ==  ENODEV && profile_cpu != -1)
+			die("No such device - did you specify an out-of-range profile CPU?\n");
 
 		/*
 		 * If it's cycles then fall back to hrtimer
@@ -524,10 +536,14 @@ static int __cmd_record(int argc, const char **argv)
 	signal(SIGCHLD, sig_handler);
 	signal(SIGINT, sig_handler);
 
-	if (!stat(output_name, &st) && !force && !append_file) {
-		fprintf(stderr, "Error, output file %s exists, use -A to append or -f to overwrite.\n",
-				output_name);
-		exit(-1);
+	if (!stat(output_name, &st) && st.st_size) {
+		if (!force && !append_file) {
+			fprintf(stderr, "Error, output file %s exists, use -A to append or -f to overwrite.\n",
+					output_name);
+			exit(-1);
+		}
+	} else {
+		append_file = 0;
 	}
 
 	flags = O_CREAT|O_RDWR;
@@ -554,16 +570,22 @@ static int __cmd_record(int argc, const char **argv)
 		if (pid == -1)
 			pid = getpid();
 
-		open_counters(-1, pid);
-	} else for (i = 0; i < nr_cpus; i++)
-		open_counters(i, target_pid);
+		open_counters(profile_cpu, pid);
+	} else {
+		if (profile_cpu != -1) {
+			open_counters(profile_cpu, target_pid);
+		} else {
+			for (i = 0; i < nr_cpus; i++)
+				open_counters(i, target_pid);
+		}
+	}
 
 	if (file_new)
 		perf_header__write(header, output);
 
 	if (!system_wide) {
-		pid_synthesize_comm_event(pid, 0);
-		pid_synthesize_mmap_samples(pid);
+		pid_t tgid = pid_synthesize_comm_event(pid, 0);
+		pid_synthesize_mmap_samples(pid, tgid);
 	} else
 		synthesize_all();
 
@@ -631,10 +653,14 @@ static const struct option options[] = {
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
 		    "collect data with this RT SCHED_FIFO priority"),
+	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
+		    "collect raw sample records from all opened counters"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 			    "system-wide collection from all CPUs"),
 	OPT_BOOLEAN('A', "append", &append_file,
 			    "append to the output file to do incremental profiling"),
+	OPT_INTEGER('C', "profile_cpu", &profile_cpu,
+			    "CPU to profile on"),
 	OPT_BOOLEAN('f', "force", &force,
 			"overwrite existing data file"),
 	OPT_LONG('c', "count", &default_interval,
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index b20a4b6e31b7..8b2ec882e6e0 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -31,13 +31,14 @@
 static char		const *input_name = "perf.data";
 static char		*vmlinux = NULL;
 
-static char		default_sort_order[] = "comm,dso";
+static char		default_sort_order[] = "comm,dso,symbol";
 static char		*sort_order = default_sort_order;
 static char		*dso_list_str, *comm_list_str, *sym_list_str,
 			*col_width_list_str;
 static struct strlist	*dso_list, *comm_list, *sym_list;
 static char		*field_sep;
 
+static int		force;
 static int		input;
 static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
@@ -68,7 +69,7 @@ static int		callchain;
 
 static
 struct callchain_param	callchain_param = {
-	.mode	= CHAIN_GRAPH_ABS,
+	.mode	= CHAIN_GRAPH_REL,
 	.min_percent = 0.5
 };
 
@@ -99,6 +100,7 @@ struct comm_event {
 struct fork_event {
 	struct perf_event_header header;
 	u32 pid, ppid;
+	u32 tid, ptid;
 };
 
 struct lost_event {
@@ -111,7 +113,9 @@ struct read_event {
 	struct perf_event_header header;
 	u32 pid,tid;
 	u64 value;
-	u64 format[3];
+	u64 time_enabled;
+	u64 time_running;
+	u64 id;
 };
 
 typedef union event_union {
@@ -252,7 +256,7 @@ static int strcommon(const char *pathname)
 {
 	int n = 0;
 
-	while (pathname[n] == cwd[n] && n < cwdlen)
+	while (n < cwdlen && pathname[n] == cwd[n])
 		++n;
 
 	return n;
@@ -697,7 +701,8 @@ sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used)
 	size_t ret = 0;
 
 	if (verbose)
-		ret += repsep_fprintf(fp, "%#018llx  ", (u64)self->ip);
+		ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip,
+				      dso__symtab_origin(self->dso));
 
 	ret += repsep_fprintf(fp, "[%c] ", self->level);
 	if (self->sym) {
@@ -887,6 +892,21 @@ ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, int depth,
 	return ret;
 }
 
+static struct symbol *rem_sq_bracket;
+static struct callchain_list rem_hits;
+
+static void init_rem_hits(void)
+{
+	rem_sq_bracket = malloc(sizeof(*rem_sq_bracket) + 6);
+	if (!rem_sq_bracket) {
+		fprintf(stderr, "Not enough memory to display remaining hits\n");
+		return;
+	}
+
+	strcpy(rem_sq_bracket->name, "[...]");
+	rem_hits.sym = rem_sq_bracket;
+}
+
 static size_t
 callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 			u64 total_samples, int depth, int depth_mask)
@@ -896,25 +916,34 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 	struct callchain_list *chain;
 	int new_depth_mask = depth_mask;
 	u64 new_total;
+	u64 remaining;
 	size_t ret = 0;
 	int i;
 
 	if (callchain_param.mode == CHAIN_GRAPH_REL)
-		new_total = self->cumul_hit;
+		new_total = self->children_hit;
 	else
 		new_total = total_samples;
 
+	remaining = new_total;
+
 	node = rb_first(&self->rb_root);
 	while (node) {
+		u64 cumul;
+
 		child = rb_entry(node, struct callchain_node, rb_node);
+		cumul = cumul_hits(child);
+		remaining -= cumul;
 
 		/*
 		 * The depth mask manages the output of pipes that show
 		 * the depth. We don't want to keep the pipes of the current
-		 * level for the last child of this depth
+		 * level for the last child of this depth.
+		 * Except if we have remaining filtered hits. They will
+		 * supersede the last child
 		 */
 		next = rb_next(node);
-		if (!next)
+		if (!next && (callchain_param.mode != CHAIN_GRAPH_REL || !remaining))
 			new_depth_mask &= ~(1 << (depth - 1));
 
 		/*
@@ -929,7 +958,7 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 			ret += ipchain__fprintf_graph(fp, chain, depth,
 						      new_depth_mask, i++,
 						      new_total,
-						      child->cumul_hit);
+						      cumul);
 		}
 		ret += callchain__fprintf_graph(fp, child, new_total,
 						depth + 1,
@@ -937,6 +966,19 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
 		node = next;
 	}
 
+	if (callchain_param.mode == CHAIN_GRAPH_REL &&
+		remaining && remaining != new_total) {
+
+		if (!rem_sq_bracket)
+			return ret;
+
+		new_depth_mask &= ~(1 << (depth - 1));
+
+		ret += ipchain__fprintf_graph(fp, &rem_hits, depth,
+					      new_depth_mask, 0, new_total,
+					      remaining);
+	}
+
 	return ret;
 }
 
@@ -1357,6 +1399,8 @@ static size_t output__fprintf(FILE *fp, u64 total_samples)
 	unsigned int width;
 	char *col_width = col_width_list_str;
 
+	init_rem_hits();
+
 	fprintf(fp, "# Samples: %Ld\n", (u64)total_samples);
 	fprintf(fp, "#\n");
 
@@ -1423,11 +1467,13 @@ print_entries:
 	if (sort_order == default_sort_order &&
 			parent_pattern == default_parent_pattern) {
 		fprintf(fp, "#\n");
-		fprintf(fp, "# (For more details, try: perf report --sort comm,dso,symbol)\n");
+		fprintf(fp, "# (For a higher level overview, try: perf report --sort comm,dso)\n");
 		fprintf(fp, "#\n");
 	}
 	fprintf(fp, "\n");
 
+	free(rem_sq_bracket);
+
 	return ret;
 }
 
@@ -1481,11 +1527,11 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 		more_data += sizeof(u64);
 	}
 
-	dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d: %p period: %Ld\n",
+	dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d/%d: %p period: %Ld\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->header.misc,
-		event->ip.pid,
+		event->ip.pid, event->ip.tid,
 		(void *)(long)ip,
 		(long long)period);
 
@@ -1545,10 +1591,11 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	if (show & show_mask) {
 		struct symbol *sym = resolve_symbol(thread, &map, &dso, &ip);
 
-		if (dso_list && dso && dso->name && !strlist__has_entry(dso_list, dso->name))
+		if (dso_list && (!dso || !dso->name ||
+				 !strlist__has_entry(dso_list, dso->name)))
 			return 0;
 
-		if (sym_list && sym && !strlist__has_entry(sym_list, sym->name))
+		if (sym_list && (!sym || !strlist__has_entry(sym_list, sym->name)))
 			return 0;
 
 		if (hist_entry__add(thread, map, dso, sym, ip, chain, level, period)) {
@@ -1567,10 +1614,11 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 	struct thread *thread = threads__findnew(event->mmap.pid);
 	struct map *map = map__new(&event->mmap);
 
-	dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n",
+	dprintf("%p [%p]: PERF_EVENT_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->mmap.pid,
+		event->mmap.tid,
 		(void *)(long)event->mmap.start,
 		(void *)(long)event->mmap.len,
 		(void *)(long)event->mmap.pgoff,
@@ -1608,15 +1656,27 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 }
 
 static int
-process_fork_event(event_t *event, unsigned long offset, unsigned long head)
+process_task_event(event_t *event, unsigned long offset, unsigned long head)
 {
 	struct thread *thread = threads__findnew(event->fork.pid);
 	struct thread *parent = threads__findnew(event->fork.ppid);
 
-	dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
+	dprintf("%p [%p]: PERF_EVENT_%s: (%d:%d):(%d:%d)\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
-		event->fork.pid, event->fork.ppid);
+		event->header.type == PERF_EVENT_FORK ? "FORK" : "EXIT",
+		event->fork.pid, event->fork.tid,
+		event->fork.ppid, event->fork.ptid);
+
+	/*
+	 * A thread clone will have the same PID for both
+	 * parent and child.
+	 */
+	if (thread == parent)
+		return 0;
+
+	if (event->header.type == PERF_EVENT_EXIT)
+		return 0;
 
 	if (!thread || !parent || thread__fork(thread, parent)) {
 		dprintf("problem processing PERF_EVENT_FORK, skipping event.\n");
@@ -1677,14 +1737,37 @@ static void trace_event(event_t *event)
 	dprintf(".\n");
 }
 
+static struct perf_header	*header;
+
+static struct perf_counter_attr *perf_header__find_attr(u64 id)
+{
+	int i;
+
+	for (i = 0; i < header->attrs; i++) {
+		struct perf_header_attr *attr = header->attr[i];
+		int j;
+
+		for (j = 0; j < attr->ids; j++) {
+			if (attr->id[j] == id)
+				return &attr->attr;
+		}
+	}
+
+	return NULL;
+}
+
 static int
 process_read_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	dprintf("%p [%p]: PERF_EVENT_READ: %d %d %Lu\n",
+	struct perf_counter_attr *attr = perf_header__find_attr(event->read.id);
+
+	dprintf("%p [%p]: PERF_EVENT_READ: %d %d %s %Lu\n",
 			(void *)(offset + head),
 			(void *)(long)(event->header.size),
 			event->read.pid,
 			event->read.tid,
+			attr ? __event_name(attr->type, attr->config)
+			     : "FAIL",
 			event->read.value);
 
 	return 0;
@@ -1706,7 +1789,8 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 		return process_comm_event(event, offset, head);
 
 	case PERF_EVENT_FORK:
-		return process_fork_event(event, offset, head);
+	case PERF_EVENT_EXIT:
+		return process_task_event(event, offset, head);
 
 	case PERF_EVENT_LOST:
 		return process_lost_event(event, offset, head);
@@ -1729,8 +1813,6 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
 	return 0;
 }
 
-static struct perf_header	*header;
-
 static u64 perf_header__sample_type(void)
 {
 	u64 sample_type = 0;
@@ -1775,6 +1857,11 @@ static int __cmd_report(void)
 		exit(-1);
 	}
 
+	if (!force && (stat.st_uid != geteuid())) {
+		fprintf(stderr, "file: %s not owned by current user\n", input_name);
+		exit(-1);
+	}
+
 	if (!stat.st_size) {
 		fprintf(stderr, "zero-sized file, nothing to do!\n");
 		exit(0);
@@ -1798,6 +1885,13 @@ static int __cmd_report(void)
 					" -g?\n");
 			exit(-1);
 		}
+	} else if (callchain_param.mode != CHAIN_NONE && !callchain) {
+			callchain = 1;
+			if (register_callchain_param(&callchain_param) < 0) {
+				fprintf(stderr, "Can't register callchain"
+						" params\n");
+				exit(-1);
+			}
 	}
 
 	if (load_kernel() < 0) {
@@ -1936,6 +2030,13 @@ parse_callchain_opt(const struct option *opt __used, const char *arg,
 	else if (!strncmp(tok, "fractal", strlen(arg)))
 		callchain_param.mode = CHAIN_GRAPH_REL;
 
+	else if (!strncmp(tok, "none", strlen(arg))) {
+		callchain_param.mode = CHAIN_NONE;
+		callchain = 0;
+
+		return 0;
+	}
+
 	else
 		return -1;
 
@@ -1969,6 +2070,7 @@ static const struct option options[] = {
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
 	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
+	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
 	OPT_BOOLEAN('m', "modules", &modules,
 		    "load module symbols - WARNING: use only with -k and LIVE kernel"),
 	OPT_BOOLEAN('n', "show-nr-samples", &show_nr_samples,
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index f9510eeeb6c7..b4b06c7903e1 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -496,7 +496,7 @@ static const struct option options[] = {
 		    "stat events on existing pid"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 		    "system-wide collection from all CPUs"),
-	OPT_BOOLEAN('S', "scale", &scale,
+	OPT_BOOLEAN('c', "scale", &scale,
 		    "scale/normalize counters"),
 	OPT_BOOLEAN('v', "verbose", &verbose,
 		    "be more verbose (show counter open errors, etc)"),
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index c0a423004e15..7de28ce9ca26 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -31,6 +31,8 @@
 #include <fcntl.h>
 
 #include <stdio.h>
+#include <termios.h>
+#include <unistd.h>
 
 #include <errno.h>
 #include <time.h>
@@ -54,7 +56,7 @@ static int			system_wide			=  0;
 
 static int			default_interval		= 100000;
 
-static u64			count_filter			=  5;
+static int			count_filter			=  5;
 static int			print_entries			= 15;
 
 static int			target_pid			= -1;
@@ -69,15 +71,28 @@ static int			freq				=  0;
 static int			verbose				=  0;
 static char			*vmlinux			=  NULL;
 
-static char			*sym_filter;
-static unsigned long		filter_start;
-static unsigned long		filter_end;
-
 static int			delay_secs			=  2;
 static int			zero;
 static int			dump_symtab;
 
 /*
+ * Source
+ */
+
+struct source_line {
+	u64			eip;
+	unsigned long		count[MAX_COUNTERS];
+	char			*line;
+	struct source_line	*next;
+};
+
+static char			*sym_filter			=  NULL;
+struct sym_entry		*sym_filter_entry		=  NULL;
+static int			sym_pcnt_filter			=  5;
+static int			sym_counter			=  0;
+static int			display_weighted		= -1;
+
+/*
  * Symbols
  */
 
@@ -91,9 +106,237 @@ struct sym_entry {
 	unsigned long		snap_count;
 	double			weight;
 	int			skip;
+	struct source_line	*source;
+	struct source_line	*lines;
+	struct source_line	**lines_tail;
+	pthread_mutex_t		source_lock;
 };
 
-struct sym_entry		*sym_filter_entry;
+/*
+ * Source functions
+ */
+
+static void parse_source(struct sym_entry *syme)
+{
+	struct symbol *sym;
+	struct module *module;
+	struct section *section = NULL;
+	FILE *file;
+	char command[PATH_MAX*2], *path = vmlinux;
+	u64 start, end, len;
+
+	if (!syme)
+		return;
+
+	if (syme->lines) {
+		pthread_mutex_lock(&syme->source_lock);
+		goto out_assign;
+	}
+
+	sym = (struct symbol *)(syme + 1);
+	module = sym->module;
+
+	if (module)
+		path = module->path;
+	if (!path)
+		return;
+
+	start = sym->obj_start;
+	if (!start)
+		start = sym->start;
+
+	if (module) {
+		section = module->sections->find_section(module->sections, ".text");
+		if (section)
+			start -= section->vma;
+	}
+
+	end = start + sym->end - sym->start + 1;
+	len = sym->end - sym->start;
+
+	sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s", start, end, path);
+
+	file = popen(command, "r");
+	if (!file)
+		return;
+
+	pthread_mutex_lock(&syme->source_lock);
+	syme->lines_tail = &syme->lines;
+	while (!feof(file)) {
+		struct source_line *src;
+		size_t dummy = 0;
+		char *c;
+
+		src = malloc(sizeof(struct source_line));
+		assert(src != NULL);
+		memset(src, 0, sizeof(struct source_line));
+
+		if (getline(&src->line, &dummy, file) < 0)
+			break;
+		if (!src->line)
+			break;
+
+		c = strchr(src->line, '\n');
+		if (c)
+			*c = 0;
+
+		src->next = NULL;
+		*syme->lines_tail = src;
+		syme->lines_tail = &src->next;
+
+		if (strlen(src->line)>8 && src->line[8] == ':') {
+			src->eip = strtoull(src->line, NULL, 16);
+			if (section)
+				src->eip += section->vma;
+		}
+		if (strlen(src->line)>8 && src->line[16] == ':') {
+			src->eip = strtoull(src->line, NULL, 16);
+			if (section)
+				src->eip += section->vma;
+		}
+	}
+	pclose(file);
+out_assign:
+	sym_filter_entry = syme;
+	pthread_mutex_unlock(&syme->source_lock);
+}
+
+static void __zero_source_counters(struct sym_entry *syme)
+{
+	int i;
+	struct source_line *line;
+
+	line = syme->lines;
+	while (line) {
+		for (i = 0; i < nr_counters; i++)
+			line->count[i] = 0;
+		line = line->next;
+	}
+}
+
+static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip)
+{
+	struct source_line *line;
+
+	if (syme != sym_filter_entry)
+		return;
+
+	if (pthread_mutex_trylock(&syme->source_lock))
+		return;
+
+	if (!syme->source)
+		goto out_unlock;
+
+	for (line = syme->lines; line; line = line->next) {
+		if (line->eip == ip) {
+			line->count[counter]++;
+			break;
+		}
+		if (line->eip > ip)
+			break;
+	}
+out_unlock:
+	pthread_mutex_unlock(&syme->source_lock);
+}
+
+static void lookup_sym_source(struct sym_entry *syme)
+{
+	struct symbol *symbol = (struct symbol *)(syme + 1);
+	struct source_line *line;
+	char pattern[PATH_MAX];
+	char *idx;
+
+	sprintf(pattern, "<%s>:", symbol->name);
+
+	if (symbol->module) {
+		idx = strstr(pattern, "\t");
+		if (idx)
+			*idx = 0;
+	}
+
+	pthread_mutex_lock(&syme->source_lock);
+	for (line = syme->lines; line; line = line->next) {
+		if (strstr(line->line, pattern)) {
+			syme->source = line;
+			break;
+		}
+	}
+	pthread_mutex_unlock(&syme->source_lock);
+}
+
+static void show_lines(struct source_line *queue, int count, int total)
+{
+	int i;
+	struct source_line *line;
+
+	line = queue;
+	for (i = 0; i < count; i++) {
+		float pcnt = 100.0*(float)line->count[sym_counter]/(float)total;
+
+		printf("%8li %4.1f%%\t%s\n", line->count[sym_counter], pcnt, line->line);
+		line = line->next;
+	}
+}
+
+#define TRACE_COUNT     3
+
+static void show_details(struct sym_entry *syme)
+{
+	struct symbol *symbol;
+	struct source_line *line;
+	struct source_line *line_queue = NULL;
+	int displayed = 0;
+	int line_queue_count = 0, total = 0, more = 0;
+
+	if (!syme)
+		return;
+
+	if (!syme->source)
+		lookup_sym_source(syme);
+
+	if (!syme->source)
+		return;
+
+	symbol = (struct symbol *)(syme + 1);
+	printf("Showing %s for %s\n", event_name(sym_counter), symbol->name);
+	printf("  Events  Pcnt (>=%d%%)\n", sym_pcnt_filter);
+
+	pthread_mutex_lock(&syme->source_lock);
+	line = syme->source;
+	while (line) {
+		total += line->count[sym_counter];
+		line = line->next;
+	}
+
+	line = syme->source;
+	while (line) {
+		float pcnt = 0.0;
+
+		if (!line_queue_count)
+			line_queue = line;
+		line_queue_count++;
+
+		if (line->count[sym_counter])
+			pcnt = 100.0 * line->count[sym_counter] / (float)total;
+		if (pcnt >= (float)sym_pcnt_filter) {
+			if (displayed <= print_entries)
+				show_lines(line_queue, line_queue_count, total);
+			else more++;
+			displayed += line_queue_count;
+			line_queue_count = 0;
+			line_queue = NULL;
+		} else if (line_queue_count > TRACE_COUNT) {
+			line_queue = line_queue->next;
+			line_queue_count--;
+		}
+
+		line->count[sym_counter] = zero ? 0 : line->count[sym_counter] * 7 / 8;
+		line = line->next;
+	}
+	pthread_mutex_unlock(&syme->source_lock);
+	if (more)
+		printf("%d lines not displayed, maybe increase display entries [e]\n", more);
+}
 
 struct dso			*kernel_dso;
 
@@ -112,6 +355,9 @@ static double sym_weight(const struct sym_entry *sym)
 	double weight = sym->snap_count;
 	int counter;
 
+	if (!display_weighted)
+		return weight;
+
 	for (counter = 1; counter < nr_counters-1; counter++)
 		weight *= sym->count[counter];
 
@@ -159,7 +405,7 @@ static void rb_insert_active_sym(struct rb_root *tree, struct sym_entry *se)
 static void print_sym_table(void)
 {
 	int printed = 0, j;
-	int counter;
+	int counter, snap = !display_weighted ? sym_counter : 0;
 	float samples_per_sec = samples/delay_secs;
 	float ksamples_per_sec = (samples-userspace_samples)/delay_secs;
 	float sum_ksamples = 0.0;
@@ -175,7 +421,7 @@ static void print_sym_table(void)
 	pthread_mutex_unlock(&active_symbols_lock);
 
 	list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
-		syme->snap_count = syme->count[0];
+		syme->snap_count = syme->count[snap];
 		if (syme->snap_count != 0) {
 			syme->weight = sym_weight(syme);
 			rb_insert_active_sym(&tmp, syme);
@@ -195,7 +441,7 @@ static void print_sym_table(void)
 		samples_per_sec,
 		100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec)));
 
-	if (nr_counters == 1) {
+	if (nr_counters == 1 || !display_weighted) {
 		printf("%Ld", (u64)attrs[0].sample_period);
 		if (freq)
 			printf("Hz ");
@@ -203,7 +449,9 @@ static void print_sym_table(void)
 			printf(" ");
 	}
 
-	for (counter = 0; counter < nr_counters; counter++) {
+	if (!display_weighted)
+		printf("%s", event_name(sym_counter));
+	else for (counter = 0; counter < nr_counters; counter++) {
 		if (counter)
 			printf("/");
 
@@ -228,6 +476,11 @@ static void print_sym_table(void)
 
 	printf("------------------------------------------------------------------------------\n\n");
 
+	if (sym_filter_entry) {
+		show_details(sym_filter_entry);
+		return;
+	}
+
 	if (nr_counters == 1)
 		printf("             samples    pcnt");
 	else
@@ -242,13 +495,13 @@ static void print_sym_table(void)
 		struct symbol *sym = (struct symbol *)(syme + 1);
 		double pcnt;
 
-		if (++printed > print_entries || syme->snap_count < count_filter)
+		if (++printed > print_entries || (int)syme->snap_count < count_filter)
 			continue;
 
 		pcnt = 100.0 - (100.0 * ((sum_ksamples - syme->snap_count) /
 					 sum_ksamples));
 
-		if (nr_counters == 1)
+		if (nr_counters == 1 || !display_weighted)
 			printf("%20.2f - ", syme->weight);
 		else
 			printf("%9.1f %10ld - ", syme->weight, syme->snap_count);
@@ -261,19 +514,250 @@ static void print_sym_table(void)
 	}
 }
 
+static void prompt_integer(int *target, const char *msg)
+{
+	char *buf = malloc(0), *p;
+	size_t dummy = 0;
+	int tmp;
+
+	fprintf(stdout, "\n%s: ", msg);
+	if (getline(&buf, &dummy, stdin) < 0)
+		return;
+
+	p = strchr(buf, '\n');
+	if (p)
+		*p = 0;
+
+	p = buf;
+	while(*p) {
+		if (!isdigit(*p))
+			goto out_free;
+		p++;
+	}
+	tmp = strtoul(buf, NULL, 10);
+	*target = tmp;
+out_free:
+	free(buf);
+}
+
+static void prompt_percent(int *target, const char *msg)
+{
+	int tmp = 0;
+
+	prompt_integer(&tmp, msg);
+	if (tmp >= 0 && tmp <= 100)
+		*target = tmp;
+}
+
+static void prompt_symbol(struct sym_entry **target, const char *msg)
+{
+	char *buf = malloc(0), *p;
+	struct sym_entry *syme = *target, *n, *found = NULL;
+	size_t dummy = 0;
+
+	/* zero counters of active symbol */
+	if (syme) {
+		pthread_mutex_lock(&syme->source_lock);
+		__zero_source_counters(syme);
+		*target = NULL;
+		pthread_mutex_unlock(&syme->source_lock);
+	}
+
+	fprintf(stdout, "\n%s: ", msg);
+	if (getline(&buf, &dummy, stdin) < 0)
+		goto out_free;
+
+	p = strchr(buf, '\n');
+	if (p)
+		*p = 0;
+
+	pthread_mutex_lock(&active_symbols_lock);
+	syme = list_entry(active_symbols.next, struct sym_entry, node);
+	pthread_mutex_unlock(&active_symbols_lock);
+
+	list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
+		struct symbol *sym = (struct symbol *)(syme + 1);
+
+		if (!strcmp(buf, sym->name)) {
+			found = syme;
+			break;
+		}
+	}
+
+	if (!found) {
+		fprintf(stderr, "Sorry, %s is not active.\n", sym_filter);
+		sleep(1);
+		return;
+	} else
+		parse_source(found);
+
+out_free:
+	free(buf);
+}
+
+static void print_mapped_keys(void)
+{
+	char *name = NULL;
+
+	if (sym_filter_entry) {
+		struct symbol *sym = (struct symbol *)(sym_filter_entry+1);
+		name = sym->name;
+	}
+
+	fprintf(stdout, "\nMapped keys:\n");
+	fprintf(stdout, "\t[d]     display refresh delay.             \t(%d)\n", delay_secs);
+	fprintf(stdout, "\t[e]     display entries (lines).           \t(%d)\n", print_entries);
+
+	if (nr_counters > 1)
+		fprintf(stdout, "\t[E]     active event counter.              \t(%s)\n", event_name(sym_counter));
+
+	fprintf(stdout, "\t[f]     profile display filter (count).    \t(%d)\n", count_filter);
+
+	if (vmlinux) {
+		fprintf(stdout, "\t[F]     annotate display filter (percent). \t(%d%%)\n", sym_pcnt_filter);
+		fprintf(stdout, "\t[s]     annotate symbol.                   \t(%s)\n", name?: "NULL");
+		fprintf(stdout, "\t[S]     stop annotation.\n");
+	}
+
+	if (nr_counters > 1)
+		fprintf(stdout, "\t[w]     toggle display weighted/count[E]r. \t(%d)\n", display_weighted ? 1 : 0);
+
+	fprintf(stdout, "\t[z]     toggle sample zeroing.             \t(%d)\n", zero ? 1 : 0);
+	fprintf(stdout, "\t[qQ]    quit.\n");
+}
+
+static int key_mapped(int c)
+{
+	switch (c) {
+		case 'd':
+		case 'e':
+		case 'f':
+		case 'z':
+		case 'q':
+		case 'Q':
+			return 1;
+		case 'E':
+		case 'w':
+			return nr_counters > 1 ? 1 : 0;
+		case 'F':
+		case 's':
+		case 'S':
+			return vmlinux ? 1 : 0;
+	}
+
+	return 0;
+}
+
+static void handle_keypress(int c)
+{
+	if (!key_mapped(c)) {
+		struct pollfd stdin_poll = { .fd = 0, .events = POLLIN };
+		struct termios tc, save;
+
+		print_mapped_keys();
+		fprintf(stdout, "\nEnter selection, or unmapped key to continue: ");
+		fflush(stdout);
+
+		tcgetattr(0, &save);
+		tc = save;
+		tc.c_lflag &= ~(ICANON | ECHO);
+		tc.c_cc[VMIN] = 0;
+		tc.c_cc[VTIME] = 0;
+		tcsetattr(0, TCSANOW, &tc);
+
+		poll(&stdin_poll, 1, -1);
+		c = getc(stdin);
+
+		tcsetattr(0, TCSAFLUSH, &save);
+		if (!key_mapped(c))
+			return;
+	}
+
+	switch (c) {
+		case 'd':
+			prompt_integer(&delay_secs, "Enter display delay");
+			break;
+		case 'e':
+			prompt_integer(&print_entries, "Enter display entries (lines)");
+			break;
+		case 'E':
+			if (nr_counters > 1) {
+				int i;
+
+				fprintf(stderr, "\nAvailable events:");
+				for (i = 0; i < nr_counters; i++)
+					fprintf(stderr, "\n\t%d %s", i, event_name(i));
+
+				prompt_integer(&sym_counter, "Enter details event counter");
+
+				if (sym_counter >= nr_counters) {
+					fprintf(stderr, "Sorry, no such event, using %s.\n", event_name(0));
+					sym_counter = 0;
+					sleep(1);
+				}
+			} else sym_counter = 0;
+			break;
+		case 'f':
+			prompt_integer(&count_filter, "Enter display event count filter");
+			break;
+		case 'F':
+			prompt_percent(&sym_pcnt_filter, "Enter details display event filter (percent)");
+			break;
+		case 'q':
+		case 'Q':
+			printf("exiting.\n");
+			exit(0);
+		case 's':
+			prompt_symbol(&sym_filter_entry, "Enter details symbol");
+			break;
+		case 'S':
+			if (!sym_filter_entry)
+				break;
+			else {
+				struct sym_entry *syme = sym_filter_entry;
+
+				pthread_mutex_lock(&syme->source_lock);
+				sym_filter_entry = NULL;
+				__zero_source_counters(syme);
+				pthread_mutex_unlock(&syme->source_lock);
+			}
+			break;
+		case 'w':
+			display_weighted = ~display_weighted;
+			break;
+		case 'z':
+			zero = ~zero;
+			break;
+	}
+}
+
 static void *display_thread(void *arg __used)
 {
 	struct pollfd stdin_poll = { .fd = 0, .events = POLLIN };
-	int delay_msecs = delay_secs * 1000;
+	struct termios tc, save;
+	int delay_msecs, c;
+
+	tcgetattr(0, &save);
+	tc = save;
+	tc.c_lflag &= ~(ICANON | ECHO);
+	tc.c_cc[VMIN] = 0;
+	tc.c_cc[VTIME] = 0;
 
-	printf("PerfTop refresh period: %d seconds\n", delay_secs);
+repeat:
+	delay_msecs = delay_secs * 1000;
+	tcsetattr(0, TCSANOW, &tc);
+	/* trash return*/
+	getc(stdin);
 
 	do {
 		print_sym_table();
 	} while (!poll(&stdin_poll, 1, delay_msecs) == 1);
 
-	printf("key pressed - exiting.\n");
-	exit(0);
+	c = getc(stdin);
+	tcsetattr(0, TCSAFLUSH, &save);
+
+	handle_keypress(c);
+	goto repeat;
 
 	return NULL;
 }
@@ -285,6 +769,7 @@ static const char *skip_symbols[] = {
 	"enter_idle",
 	"exit_idle",
 	"mwait_idle",
+	"mwait_idle_with_hints",
 	"ppc64_runlatch_off",
 	"pseries_dedicated_idle_sleep",
 	NULL
@@ -292,7 +777,6 @@ static const char *skip_symbols[] = {
 
 static int symbol_filter(struct dso *self, struct symbol *sym)
 {
-	static int filter_match;
 	struct sym_entry *syme;
 	const char *name = sym->name;
 	int i;
@@ -314,6 +798,10 @@ static int symbol_filter(struct dso *self, struct symbol *sym)
 		return 1;
 
 	syme = dso__sym_priv(self, sym);
+	pthread_mutex_init(&syme->source_lock, NULL);
+	if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter))
+		sym_filter_entry = syme;
+
 	for (i = 0; skip_symbols[i]; i++) {
 		if (!strcmp(skip_symbols[i], name)) {
 			syme->skip = 1;
@@ -321,29 +809,6 @@ static int symbol_filter(struct dso *self, struct symbol *sym)
 		}
 	}
 
-	if (filter_match == 1) {
-		filter_end = sym->start;
-		filter_match = -1;
-		if (filter_end - filter_start > 10000) {
-			fprintf(stderr,
-				"hm, too large filter symbol <%s> - skipping.\n",
-				sym_filter);
-			fprintf(stderr, "symbol filter start: %016lx\n",
-				filter_start);
-			fprintf(stderr, "                end: %016lx\n",
-				filter_end);
-			filter_end = filter_start = 0;
-			sym_filter = NULL;
-			sleep(1);
-		}
-	}
-
-	if (filter_match == 0 && sym_filter && !strcmp(name, sym_filter)) {
-		filter_match = 1;
-		filter_start = sym->start;
-	}
-
-
 	return 0;
 }
 
@@ -379,8 +844,6 @@ out_delete_dso:
 	return -1;
 }
 
-#define TRACE_COUNT     3
-
 /*
  * Binary search in the histogram table and record the hit:
  */
@@ -393,6 +856,7 @@ static void record_ip(u64 ip, int counter)
 
 		if (!syme->skip) {
 			syme->count[counter]++;
+			record_precise_ip(syme, counter, ip);
 			pthread_mutex_lock(&active_symbols_lock);
 			if (list_empty(&syme->node) || !syme->node.next)
 				__list_insert_active_sym(syme);
@@ -689,8 +1153,8 @@ static const struct option options[] = {
 			    "put the counters into a counter group"),
 	OPT_BOOLEAN('i', "inherit", &inherit,
 		    "child tasks inherit counters"),
-	OPT_STRING('s', "sym-filter", &sym_filter, "pattern",
-		    "only display symbols matchig this pattern"),
+	OPT_STRING('s', "sym-annotate", &sym_filter, "symbol name",
+		    "symbol to annotate - requires -k option"),
 	OPT_BOOLEAN('z', "zero", &zero,
 		    "zero history across updates"),
 	OPT_INTEGER('F', "freq", &freq,
@@ -733,6 +1197,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
 		delay_secs = 1;
 
 	parse_symbols();
+	parse_source(sym_filter_entry);
 
 	/*
 	 * Fill in the ones not specifically initialized via -c:
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index 9d3c8141b8c1..011473411642 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -13,6 +13,7 @@
 #include <stdio.h>
 #include <stdbool.h>
 #include <errno.h>
+#include <math.h>
 
 #include "callchain.h"
 
@@ -26,10 +27,14 @@ rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
 	struct callchain_node *rnode;
+	u64 chain_cumul = cumul_hits(chain);
 
 	while (*p) {
+		u64 rnode_cumul;
+
 		parent = *p;
 		rnode = rb_entry(parent, struct callchain_node, rb_node);
+		rnode_cumul = cumul_hits(rnode);
 
 		switch (mode) {
 		case CHAIN_FLAT:
@@ -40,7 +45,7 @@ rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
 			break;
 		case CHAIN_GRAPH_ABS: /* Falldown */
 		case CHAIN_GRAPH_REL:
-			if (rnode->cumul_hit < chain->cumul_hit)
+			if (rnode_cumul < chain_cumul)
 				p = &(*p)->rb_left;
 			else
 				p = &(*p)->rb_right;
@@ -87,7 +92,7 @@ static void __sort_chain_graph_abs(struct callchain_node *node,
 
 	chain_for_each_child(child, node) {
 		__sort_chain_graph_abs(child, min_hit);
-		if (child->cumul_hit >= min_hit)
+		if (cumul_hits(child) >= min_hit)
 			rb_insert_callchain(&node->rb_root, child,
 					    CHAIN_GRAPH_ABS);
 	}
@@ -108,11 +113,11 @@ static void __sort_chain_graph_rel(struct callchain_node *node,
 	u64 min_hit;
 
 	node->rb_root = RB_ROOT;
-	min_hit = node->cumul_hit * min_percent / 100.0;
+	min_hit = ceil(node->children_hit * min_percent);
 
 	chain_for_each_child(child, node) {
 		__sort_chain_graph_rel(child, min_percent);
-		if (child->cumul_hit >= min_hit)
+		if (cumul_hits(child) >= min_hit)
 			rb_insert_callchain(&node->rb_root, child,
 					    CHAIN_GRAPH_REL);
 	}
@@ -122,7 +127,7 @@ static void
 sort_chain_graph_rel(struct rb_root *rb_root, struct callchain_node *chain_root,
 		     u64 min_hit __used, struct callchain_param *param)
 {
-	__sort_chain_graph_rel(chain_root, param->min_percent);
+	__sort_chain_graph_rel(chain_root, param->min_percent / 100.0);
 	rb_root->rb_node = chain_root->rb_root.rb_node;
 }
 
@@ -211,7 +216,8 @@ add_child(struct callchain_node *parent, struct ip_callchain *chain,
 	new = create_child(parent, false);
 	fill_node(new, chain, start, syms);
 
-	new->cumul_hit = new->hit = 1;
+	new->children_hit = 0;
+	new->hit = 1;
 }
 
 /*
@@ -241,7 +247,8 @@ split_add_child(struct callchain_node *parent, struct ip_callchain *chain,
 
 	/* split the hits */
 	new->hit = parent->hit;
-	new->cumul_hit = parent->cumul_hit;
+	new->children_hit = parent->children_hit;
+	parent->children_hit = cumul_hits(new);
 	new->val_nr = parent->val_nr - idx_local;
 	parent->val_nr = idx_local;
 
@@ -249,6 +256,7 @@ split_add_child(struct callchain_node *parent, struct ip_callchain *chain,
 	if (idx_total < chain->nr) {
 		parent->hit = 0;
 		add_child(parent, chain, idx_total, syms);
+		parent->children_hit++;
 	} else {
 		parent->hit = 1;
 	}
@@ -269,13 +277,13 @@ __append_chain_children(struct callchain_node *root, struct ip_callchain *chain,
 		unsigned int ret = __append_chain(rnode, chain, start, syms);
 
 		if (!ret)
-			goto cumul;
+			goto inc_children_hit;
 	}
 	/* nothing in children, add to the current node */
 	add_child(root, chain, start, syms);
 
-cumul:
-	root->cumul_hit++;
+inc_children_hit:
+	root->children_hit++;
 }
 
 static int
@@ -317,8 +325,6 @@ __append_chain(struct callchain_node *root, struct ip_callchain *chain,
 	/* we match 100% of the path, increment the hit */
 	if (i - start == root->val_nr && i == chain->nr) {
 		root->hit++;
-		root->cumul_hit++;
-
 		return 0;
 	}
 
@@ -331,5 +337,7 @@ __append_chain(struct callchain_node *root, struct ip_callchain *chain,
 void append_chain(struct callchain_node *root, struct ip_callchain *chain,
 		  struct symbol **syms)
 {
+	if (!chain->nr)
+		return;
 	__append_chain_children(root, chain, syms, 0);
 }
diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h
index 7812122bea1d..a926ae4f5a16 100644
--- a/tools/perf/util/callchain.h
+++ b/tools/perf/util/callchain.h
@@ -7,6 +7,7 @@
 #include "symbol.h"
 
 enum chain_mode {
+	CHAIN_NONE,
 	CHAIN_FLAT,
 	CHAIN_GRAPH_ABS,
 	CHAIN_GRAPH_REL
@@ -21,7 +22,7 @@ struct callchain_node {
 	struct rb_root		rb_root; /* sorted tree of children */
 	unsigned int		val_nr;
 	u64			hit;
-	u64			cumul_hit; /* hit + hits of children */
+	u64			children_hit;
 };
 
 struct callchain_param;
@@ -48,6 +49,11 @@ static inline void callchain_init(struct callchain_node *node)
 	INIT_LIST_HEAD(&node->val);
 }
 
+static inline u64 cumul_hits(struct callchain_node *node)
+{
+	return node->hit + node->children_hit;
+}
+
 int register_callchain_param(struct callchain_param *param);
 void append_chain(struct callchain_node *root, struct ip_callchain *chain,
 		  struct symbol **syms);
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 450384b3bbe5..b92a457ca32e 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -185,6 +185,8 @@ static void do_read(int fd, void *buf, size_t size)
 
 		if (ret < 0)
 			die("failed to read");
+		if (ret == 0)
+			die("failed to read: missing data");
 
 		size -= ret;
 		buf += ret;
@@ -213,9 +215,10 @@ struct perf_header *perf_header__read(int fd)
 
 	for (i = 0; i < nr_attrs; i++) {
 		struct perf_header_attr *attr;
-		off_t tmp = lseek(fd, 0, SEEK_CUR);
+		off_t tmp;
 
 		do_read(fd, &f_attr, sizeof(f_attr));
+		tmp = lseek(fd, 0, SEEK_CUR);
 
 		attr = perf_header_attr__new(&f_attr.attr);
 
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 7bdad8df22a6..044178408783 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -121,13 +121,29 @@ static unsigned long hw_cache_stat[C(MAX)] = {
 	   (strcmp(sys_dirent.d_name, ".")) &&				       \
 	   (strcmp(sys_dirent.d_name, "..")))
 
+static int tp_event_has_id(struct dirent *sys_dir, struct dirent *evt_dir)
+{
+	char evt_path[MAXPATHLEN];
+	int fd;
+
+	snprintf(evt_path, MAXPATHLEN, "%s/%s/%s/id", debugfs_path,
+			sys_dir->d_name, evt_dir->d_name);
+	fd = open(evt_path, O_RDONLY);
+	if (fd < 0)
+		return -EINVAL;
+	close(fd);
+
+	return 0;
+}
+
 #define for_each_event(sys_dirent, evt_dir, evt_dirent, evt_next, file, st)    \
 	while (!readdir_r(evt_dir, &evt_dirent, &evt_next) && evt_next)        \
 	if (snprintf(file, MAXPATHLEN, "%s/%s/%s", debugfs_path,	       \
 		     sys_dirent.d_name, evt_dirent.d_name) &&		       \
 	   (!stat(file, &st)) && (S_ISDIR(st.st_mode)) &&		       \
 	   (strcmp(evt_dirent.d_name, ".")) &&				       \
-	   (strcmp(evt_dirent.d_name, "..")))
+	   (strcmp(evt_dirent.d_name, "..")) &&				       \
+	   (!tp_event_has_id(&sys_dirent, &evt_dirent)))
 
 #define MAX_EVENT_LENGTH 30
 
@@ -223,9 +239,15 @@ char *event_name(int counter)
 {
 	u64 config = attrs[counter].config;
 	int type = attrs[counter].type;
+
+	return __event_name(type, config);
+}
+
+char *__event_name(int type, u64 config)
+{
 	static char buf[32];
 
-	if (attrs[counter].type == PERF_TYPE_RAW) {
+	if (type == PERF_TYPE_RAW) {
 		sprintf(buf, "raw 0x%llx", config);
 		return buf;
 	}
@@ -357,6 +379,7 @@ static int parse_tracepoint_event(const char **strp,
 				    struct perf_counter_attr *attr)
 {
 	const char *evt_name;
+	char *flags;
 	char sys_name[MAX_EVENT_LENGTH];
 	char id_buf[4];
 	int fd;
@@ -378,6 +401,15 @@ static int parse_tracepoint_event(const char **strp,
 	strncpy(sys_name, *strp, sys_length);
 	sys_name[sys_length] = '\0';
 	evt_name = evt_name + 1;
+
+	flags = strchr(evt_name, ':');
+	if (flags) {
+		*flags = '\0';
+		flags++;
+		if (!strncmp(flags, "record", strlen(flags)))
+			attr->sample_type |= PERF_SAMPLE_RAW;
+	}
+
 	evt_length = strlen(evt_name);
 	if (evt_length >= MAX_EVENT_LENGTH)
 		return 0;
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index 1ea5d09b6eb1..192a962e3a0f 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -10,6 +10,7 @@ extern int			nr_counters;
 extern struct perf_counter_attr attrs[MAX_COUNTERS];
 
 extern char *event_name(int ctr);
+extern char *__event_name(int type, u64 config);
 
 extern int parse_events(const struct option *opt, const char *str, int unset);
 
diff --git a/tools/perf/util/quote.c b/tools/perf/util/quote.c
index c6e5dc0dc82f..2726fe40eb5d 100644
--- a/tools/perf/util/quote.c
+++ b/tools/perf/util/quote.c
@@ -318,7 +318,7 @@ char *quote_path_relative(const char *in, int len,
 		strbuf_addch(out, '"');
 	if (prefix) {
 		int off = 0;
-		while (prefix[off] && off < len && prefix[off] == in[off])
+		while (off < len && prefix[off] && prefix[off] == in[off])
 			if (prefix[off] == '/') {
 				prefix += off + 1;
 				in += off + 1;
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 28106059bf12..5c0f42e6b33b 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -6,14 +6,18 @@
 #include <libelf.h>
 #include <gelf.h>
 #include <elf.h>
-#include <bfd.h>
 
 const char *sym_hist_filter;
 
-#ifndef DMGL_PARAMS
-#define DMGL_PARAMS      (1 << 0)       /* Include function args */
-#define DMGL_ANSI        (1 << 1)       /* Include const, volatile, etc */
-#endif
+enum dso_origin {
+	DSO__ORIG_KERNEL = 0,
+	DSO__ORIG_JAVA_JIT,
+	DSO__ORIG_FEDORA,
+	DSO__ORIG_UBUNTU,
+	DSO__ORIG_BUILDID,
+	DSO__ORIG_DSO,
+	DSO__ORIG_NOT_FOUND,
+};
 
 static struct symbol *symbol__new(u64 start, u64 len,
 				  const char *name, unsigned int priv_size,
@@ -72,6 +76,7 @@ struct dso *dso__new(const char *name, unsigned int sym_priv_size)
 		self->sym_priv_size = sym_priv_size;
 		self->find_symbol = dso__find_symbol;
 		self->slen_calculated = 0;
+		self->origin = DSO__ORIG_NOT_FOUND;
 	}
 
 	return self;
@@ -565,7 +570,7 @@ static int dso__load_sym(struct dso *self, int fd, const char *name,
 		goto out_elf_end;
 
 	secstrs = elf_getdata(sec_strndx, NULL);
-	if (symstrs == NULL)
+	if (secstrs == NULL)
 		goto out_elf_end;
 
 	nr_syms = shdr.sh_size / shdr.sh_entsize;
@@ -652,11 +657,85 @@ out_close:
 	return err;
 }
 
+#define BUILD_ID_SIZE 128
+
+static char *dso__read_build_id(struct dso *self, int verbose)
+{
+	int i;
+	GElf_Ehdr ehdr;
+	GElf_Shdr shdr;
+	Elf_Data *build_id_data;
+	Elf_Scn *sec;
+	char *build_id = NULL, *bid;
+	unsigned char *raw;
+	Elf *elf;
+	int fd = open(self->name, O_RDONLY);
+
+	if (fd < 0)
+		goto out;
+
+	elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+	if (elf == NULL) {
+		if (verbose)
+			fprintf(stderr, "%s: cannot read %s ELF file.\n",
+				__func__, self->name);
+		goto out_close;
+	}
+
+	if (gelf_getehdr(elf, &ehdr) == NULL) {
+		if (verbose)
+			fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+		goto out_elf_end;
+	}
+
+	sec = elf_section_by_name(elf, &ehdr, &shdr, ".note.gnu.build-id", NULL);
+	if (sec == NULL)
+		goto out_elf_end;
+
+	build_id_data = elf_getdata(sec, NULL);
+	if (build_id_data == NULL)
+		goto out_elf_end;
+	build_id = malloc(BUILD_ID_SIZE);
+	if (build_id == NULL)
+		goto out_elf_end;
+	raw = build_id_data->d_buf + 16;
+	bid = build_id;
+
+	for (i = 0; i < 20; ++i) {
+		sprintf(bid, "%02x", *raw);
+		++raw;
+		bid += 2;
+	}
+	if (verbose >= 2)
+		printf("%s(%s): %s\n", __func__, self->name, build_id);
+out_elf_end:
+	elf_end(elf);
+out_close:
+	close(fd);
+out:
+	return build_id;
+}
+
+char dso__symtab_origin(const struct dso *self)
+{
+	static const char origin[] = {
+		[DSO__ORIG_KERNEL] =   'k',
+		[DSO__ORIG_JAVA_JIT] = 'j',
+		[DSO__ORIG_FEDORA] =   'f',
+		[DSO__ORIG_UBUNTU] =   'u',
+		[DSO__ORIG_BUILDID] =  'b',
+		[DSO__ORIG_DSO] =      'd',
+	};
+
+	if (self == NULL || self->origin == DSO__ORIG_NOT_FOUND)
+		return '!';
+	return origin[self->origin];
+}
+
 int dso__load(struct dso *self, symbol_filter_t filter, int verbose)
 {
-	int size = strlen(self->name) + sizeof("/usr/lib/debug%s.debug");
-	char *name = malloc(size);
-	int variant = 0;
+	int size = PATH_MAX;
+	char *name = malloc(size), *build_id = NULL;
 	int ret = -1;
 	int fd;
 
@@ -665,26 +744,43 @@ int dso__load(struct dso *self, symbol_filter_t filter, int verbose)
 
 	self->adjust_symbols = 0;
 
-	if (strncmp(self->name, "/tmp/perf-", 10) == 0)
-		return dso__load_perf_map(self, filter, verbose);
+	if (strncmp(self->name, "/tmp/perf-", 10) == 0) {
+		ret = dso__load_perf_map(self, filter, verbose);
+		self->origin = ret > 0 ? DSO__ORIG_JAVA_JIT :
+					 DSO__ORIG_NOT_FOUND;
+		return ret;
+	}
+
+	self->origin = DSO__ORIG_FEDORA - 1;
 
 more:
 	do {
-		switch (variant) {
-		case 0: /* Fedora */
+		self->origin++;
+		switch (self->origin) {
+		case DSO__ORIG_FEDORA:
 			snprintf(name, size, "/usr/lib/debug%s.debug", self->name);
 			break;
-		case 1: /* Ubuntu */
+		case DSO__ORIG_UBUNTU:
 			snprintf(name, size, "/usr/lib/debug%s", self->name);
 			break;
-		case 2: /* Sane people */
+		case DSO__ORIG_BUILDID:
+			build_id = dso__read_build_id(self, verbose);
+			if (build_id != NULL) {
+				snprintf(name, size,
+					 "/usr/lib/debug/.build-id/%.2s/%s.debug",
+					build_id, build_id + 2);
+				free(build_id);
+				break;
+			}
+			self->origin++;
+			/* Fall thru */
+		case DSO__ORIG_DSO:
 			snprintf(name, size, "%s", self->name);
 			break;
 
 		default:
 			goto out;
 		}
-		variant++;
 
 		fd = open(name, O_RDONLY);
 	} while (fd < 0);
@@ -705,6 +801,8 @@ more:
 	}
 out:
 	free(name);
+	if (ret < 0 && strstr(self->name, " (deleted)") != NULL)
+		return 0;
 	return ret;
 }
 
@@ -820,6 +918,9 @@ int dso__load_kernel(struct dso *self, const char *vmlinux,
 	if (err <= 0)
 		err = dso__load_kallsyms(self, filter, verbose);
 
+	if (err > 0)
+		self->origin = DSO__ORIG_KERNEL;
+
 	return err;
 }
 
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 2f92b21c712d..b53bf0125c1b 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -7,6 +7,30 @@
 #include <linux/rbtree.h>
 #include "module.h"
 
+#ifdef HAVE_CPLUS_DEMANGLE
+extern char *cplus_demangle(const char *, int);
+
+static inline char *bfd_demangle(void __used *v, const char *c, int i)
+{
+	return cplus_demangle(c, i);
+}
+#else
+#ifdef NO_DEMANGLE
+static inline char *bfd_demangle(void __used *v, const char __used *c,
+				 int __used i)
+{
+	return NULL;
+}
+#else
+#include <bfd.h>
+#endif
+#endif
+
+#ifndef DMGL_PARAMS
+#define DMGL_PARAMS      (1 << 0)       /* Include function args */
+#define DMGL_ANSI        (1 << 1)       /* Include const, volatile, etc */
+#endif
+
 struct symbol {
 	struct rb_node	rb_node;
 	u64		start;
@@ -26,6 +50,7 @@ struct dso {
 	unsigned int	 sym_priv_size;
 	unsigned char	 adjust_symbols;
 	unsigned char	 slen_calculated;
+	unsigned char	 origin;
 	char		 name[0];
 };
 
@@ -49,6 +74,7 @@ int dso__load_modules(struct dso *self, symbol_filter_t filter, int verbose);
 int dso__load(struct dso *self, symbol_filter_t filter, int verbose);
 
 size_t dso__fprintf(struct dso *self, FILE *fp);
+char dso__symtab_origin(const struct dso *self);
 
 void symbol__init(void);
 #endif /* _PERF_SYMBOL_ */