summaryrefslogtreecommitdiff
path: root/tools
diff options
context:
space:
mode:
Diffstat (limited to 'tools')
-rw-r--r--tools/perf/Documentation/perf-bench.txt120
-rw-r--r--tools/perf/Documentation/perf-buildid-list.txt34
-rw-r--r--tools/perf/Documentation/perf-probe.txt49
-rw-r--r--tools/perf/Documentation/perf-report.txt8
-rw-r--r--tools/perf/Documentation/perf-timechart.txt5
-rw-r--r--tools/perf/Makefile101
-rw-r--r--tools/perf/bench/bench.h16
-rw-r--r--tools/perf/bench/sched-messaging.c336
-rw-r--r--tools/perf/bench/sched-pipe.c124
-rw-r--r--tools/perf/builtin-annotate.c158
-rw-r--r--tools/perf/builtin-bench.c183
-rw-r--r--tools/perf/builtin-buildid-list.c116
-rw-r--r--tools/perf/builtin-help.c12
-rw-r--r--tools/perf/builtin-kmem.c578
-rw-r--r--tools/perf/builtin-probe.c435
-rw-r--r--tools/perf/builtin-record.c259
-rw-r--r--tools/perf/builtin-report.c115
-rw-r--r--tools/perf/builtin-sched.c8
-rw-r--r--tools/perf/builtin-stat.c34
-rw-r--r--tools/perf/builtin-timechart.c165
-rw-r--r--tools/perf/builtin-top.c411
-rw-r--r--tools/perf/builtin-trace.c6
-rw-r--r--tools/perf/builtin.h4
-rw-r--r--tools/perf/command-list.txt3
-rw-r--r--tools/perf/design.txt2
-rw-r--r--tools/perf/perf.c70
-rw-r--r--tools/perf/perf.h12
-rwxr-xr-xtools/perf/util/PERF-VERSION-GEN2
-rw-r--r--tools/perf/util/callchain.c2
-rw-r--r--tools/perf/util/data_map.c113
-rw-r--r--tools/perf/util/data_map.h1
-rw-r--r--tools/perf/util/debug.c4
-rw-r--r--tools/perf/util/debug.h5
-rw-r--r--tools/perf/util/debugfs.c241
-rw-r--r--tools/perf/util/debugfs.h25
-rw-r--r--tools/perf/util/event.c177
-rw-r--r--tools/perf/util/event.h25
-rw-r--r--tools/perf/util/header.c431
-rw-r--r--tools/perf/util/header.h72
-rw-r--r--tools/perf/util/include/asm/asm-offsets.h1
-rw-r--r--tools/perf/util/include/asm/bitops.h18
-rw-r--r--tools/perf/util/include/asm/byteorder.h2
-rw-r--r--tools/perf/util/include/asm/swab.h1
-rw-r--r--tools/perf/util/include/asm/uaccess.h14
-rw-r--r--tools/perf/util/include/linux/bitmap.h3
-rw-r--r--tools/perf/util/include/linux/bitops.h27
-rw-r--r--tools/perf/util/include/linux/compiler.h10
-rw-r--r--tools/perf/util/include/linux/ctype.h1
-rw-r--r--tools/perf/util/include/linux/kernel.h76
-rw-r--r--tools/perf/util/include/linux/string.h1
-rw-r--r--tools/perf/util/include/linux/types.h9
-rw-r--r--tools/perf/util/map.c72
-rw-r--r--tools/perf/util/parse-events.c43
-rw-r--r--tools/perf/util/probe-finder.c732
-rw-r--r--tools/perf/util/probe-finder.h57
-rw-r--r--tools/perf/util/sort.c20
-rw-r--r--tools/perf/util/sort.h9
-rw-r--r--tools/perf/util/string.c84
-rw-r--r--tools/perf/util/string.h1
-rw-r--r--tools/perf/util/svghelper.c2
-rw-r--r--tools/perf/util/symbol.c505
-rw-r--r--tools/perf/util/symbol.h61
-rw-r--r--tools/perf/util/thread.c17
-rw-r--r--tools/perf/util/thread.h2
-rw-r--r--tools/perf/util/trace-event-parse.c9
-rw-r--r--tools/perf/util/util.h12
66 files changed, 5352 insertions, 899 deletions
diff --git a/tools/perf/Documentation/perf-bench.txt b/tools/perf/Documentation/perf-bench.txt
new file mode 100644
index 000000000000..ae525ac5a2ce
--- /dev/null
+++ b/tools/perf/Documentation/perf-bench.txt
@@ -0,0 +1,120 @@
+perf-bench(1)
+============
+
+NAME
+----
+perf-bench - General framework for benchmark suites
+
+SYNOPSIS
+--------
+[verse]
+'perf bench' [<common options>] <subsystem> <suite> [<options>]
+
+DESCRIPTION
+-----------
+This 'perf bench' command is general framework for benchmark suites.
+
+COMMON OPTIONS
+--------------
+-f::
+--format=::
+Specify format style.
+Current available format styles are,
+
+'default'::
+Default style. This is mainly for human reading.
+---------------------
+% perf bench sched pipe # with no style specify
+(executing 1000000 pipe operations between two tasks)
+ Total time:5.855 sec
+ 5.855061 usecs/op
+ 170792 ops/sec
+---------------------
+
+'simple'::
+This simple style is friendly for automated
+processing by scripts.
+---------------------
+% perf bench --format=simple sched pipe # specified simple
+5.988
+---------------------
+
+SUBSYSTEM
+---------
+
+'sched'::
+ Scheduler and IPC mechanisms.
+
+SUITES FOR 'sched'
+~~~~~~~~~~~~~~~~~~
+*messaging*::
+Suite for evaluating performance of scheduler and IPC mechanisms.
+Based on hackbench by Rusty Russell.
+
+Options of *pipe*
+^^^^^^^^^^^^^^^^^
+-p::
+--pipe::
+Use pipe() instead of socketpair()
+
+-t::
+--thread::
+Be multi thread instead of multi process
+
+-g::
+--group=::
+Specify number of groups
+
+-l::
+--loop=::
+Specify number of loops
+
+Example of *messaging*
+^^^^^^^^^^^^^^^^^^^^^^
+
+---------------------
+% perf bench sched messaging # run with default
+options (20 sender and receiver processes per group)
+(10 groups == 400 processes run)
+
+ Total time:0.308 sec
+
+% perf bench sched messaging -t -g 20 # be multi-thread,with 20 groups
+(20 sender and receiver threads per group)
+(20 groups == 800 threads run)
+
+ Total time:0.582 sec
+---------------------
+
+*pipe*::
+Suite for pipe() system call.
+Based on pipe-test-1m.c by Ingo Molnar.
+
+Options of *pipe*
+^^^^^^^^^^^^^^^^^
+-l::
+--loop=::
+Specify number of loops.
+
+Example of *pipe*
+^^^^^^^^^^^^^^^^^
+
+---------------------
+% perf bench sched pipe
+(executing 1000000 pipe operations between two tasks)
+
+ Total time:8.091 sec
+ 8.091833 usecs/op
+ 123581 ops/sec
+
+% perf bench sched pipe -l 1000 # loop 1000
+(executing 1000 pipe operations between two tasks)
+
+ Total time:0.016 sec
+ 16.948000 usecs/op
+ 59004 ops/sec
+---------------------
+
+SEE ALSO
+--------
+linkperf:perf[1]
diff --git a/tools/perf/Documentation/perf-buildid-list.txt b/tools/perf/Documentation/perf-buildid-list.txt
new file mode 100644
index 000000000000..01b642c0bf8f
--- /dev/null
+++ b/tools/perf/Documentation/perf-buildid-list.txt
@@ -0,0 +1,34 @@
+perf-buildid-list(1)
+====================
+
+NAME
+----
+perf-buildid-list - List the buildids in a perf.data file
+
+SYNOPSIS
+--------
+[verse]
+'perf buildid-list <options>'
+
+DESCRIPTION
+-----------
+This command displays the buildids found in a perf.data file, so that other
+tools can be used to fetch packages with matching symbol tables for use by
+perf report.
+
+OPTIONS
+-------
+-i::
+--input=::
+ Input file name. (default: perf.data)
+-f::
+--force::
+ Don't do ownership validation.
+-v::
+--verbose::
+ Be more verbose.
+
+SEE ALSO
+--------
+linkperf:perf-record[1], linkperf:perf-top[1],
+linkperf:perf-report[1]
diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt
new file mode 100644
index 000000000000..9270594e6dfd
--- /dev/null
+++ b/tools/perf/Documentation/perf-probe.txt
@@ -0,0 +1,49 @@
+perf-probe(1)
+=============
+
+NAME
+----
+perf-probe - Define new dynamic tracepoints
+
+SYNOPSIS
+--------
+[verse]
+'perf probe' [options] --add 'PROBE' [--add 'PROBE' ...]
+or
+'perf probe' [options] 'PROBE' ['PROBE' ...]
+
+
+DESCRIPTION
+-----------
+This command defines dynamic tracepoint events, by symbol and registers
+without debuginfo, or by C expressions (C line numbers, C function names,
+and C local variables) with debuginfo.
+
+
+OPTIONS
+-------
+-k::
+--vmlinux=PATH::
+ Specify vmlinux path which has debuginfo (Dwarf binary).
+
+-v::
+--verbose::
+ Be more verbose (show parsed arguments, etc).
+
+-a::
+--add::
+ Define a probe point (see PROBE SYNTAX for detail)
+
+PROBE SYNTAX
+------------
+Probe points are defined by following syntax.
+
+ "FUNC[+OFFS|:RLN|%return][@SRC]|SRC:ALN [ARG ...]"
+
+'FUNC' specifies a probed function name, and it may have one of the following options; '+OFFS' is the offset from function entry address in bytes, 'RLN' is the relative-line number from function entry line, and '%return' means that it probes function return. In addition, 'SRC' specifies a source file which has that function.
+It is also possible to specify a probe point by the source line number by using 'SRC:ALN' syntax, where 'SRC' is the source file path and 'ALN' is the line number.
+'ARG' specifies the arguments of this probe point. You can use the name of local variable, or kprobe-tracer argument format (e.g. $retval, %ax, etc).
+
+SEE ALSO
+--------
+linkperf:perf-trace[1], linkperf:perf-record[1]
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index 59f0b846cd71..9dccb180b7af 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -24,11 +24,11 @@ OPTIONS
--dsos=::
Only consider symbols in these dsos. CSV that understands
file://filename entries.
--n
---show-nr-samples
+-n::
+--show-nr-samples::
Show the number of samples for each symbol
--T
---threads
+-T::
+--threads::
Show per-thread event counters
-C::
--comms=::
diff --git a/tools/perf/Documentation/perf-timechart.txt b/tools/perf/Documentation/perf-timechart.txt
index a7910099d6fd..4b1788355eca 100644
--- a/tools/perf/Documentation/perf-timechart.txt
+++ b/tools/perf/Documentation/perf-timechart.txt
@@ -31,9 +31,12 @@ OPTIONS
-w::
--width=::
Select the width of the SVG file (default: 1000)
--p::
+-P::
--power-only::
Only output the CPU power section of the diagram
+-p::
+--process::
+ Select the processes to display, by name or PID
SEE ALSO
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 106c15055b50..d7198c54bb69 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -145,6 +145,8 @@ all::
# Define NO_EXTERNAL_GREP if you don't want "perf grep" to ever call
# your external grep (e.g., if your system lacks grep, if its grep is
# broken, or spawning external process is slower than built-in grep perf has).
+#
+# Define LDFLAGS=-static to build a static binary.
PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
@$(SHELL_PATH) util/PERF-VERSION-GEN
@@ -164,10 +166,12 @@ ifdef NO_64BIT
MBITS := -m32
else
#
- # If we're on a 64-bit kernel, use -m64:
+ # If we're on a 64-bit kernel (except ia64), use -m64:
#
- ifneq ($(patsubst %64,%,$(uname_M)),$(uname_M))
- MBITS := -m64
+ ifneq ($(uname_M),ia64)
+ ifneq ($(patsubst %64,%,$(uname_M)),$(uname_M))
+ MBITS := -m64
+ endif
endif
endif
@@ -177,8 +181,7 @@ endif
# Include saner warnings here, which can catch bugs:
#
-EXTRA_WARNINGS := -Wcast-align
-EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat
+EXTRA_WARNINGS := -Wformat
EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-security
EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-y2k
EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wshadow
@@ -201,8 +204,15 @@ EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition
EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes
EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement
-CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
-LDFLAGS = -lpthread -lrt -lelf -lm
+ifeq ("$(origin DEBUG)", "command line")
+ PERF_DEBUG = $(DEBUG)
+endif
+ifndef PERF_DEBUG
+ CFLAGS_OPTIMIZE = -O6
+endif
+
+CFLAGS = $(MBITS) -ggdb3 -Wall -Wextra -std=gnu99 -Werror $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
+EXTLIBS = -lpthread -lrt -lelf -lm
ALL_CFLAGS = $(CFLAGS)
ALL_LDFLAGS = $(LDFLAGS)
STRIP ?= strip
@@ -253,6 +263,9 @@ PTHREAD_LIBS = -lpthread
# explicitly what architecture to check for. Fix this up for yours..
SPARSE_FLAGS = -D__BIG_ENDIAN__ -D__powerpc__
+ifeq ($(shell sh -c "echo 'int foo(void) {char X[2]; return 3;}' | $(CC) -x c -c -Werror -fstack-protector-all - -o /dev/null >/dev/null 2>&1 && echo y"), y)
+ CFLAGS := $(CFLAGS) -fstack-protector-all
+endif
### --- END CONFIGURATION SECTION ---
@@ -328,8 +341,27 @@ LIB_FILE=libperf.a
LIB_H += ../../include/linux/perf_event.h
LIB_H += ../../include/linux/rbtree.h
LIB_H += ../../include/linux/list.h
+LIB_H += ../../include/linux/stringify.h
+LIB_H += util/include/linux/bitmap.h
+LIB_H += util/include/linux/bitops.h
+LIB_H += util/include/linux/compiler.h
+LIB_H += util/include/linux/ctype.h
+LIB_H += util/include/linux/kernel.h
LIB_H += util/include/linux/list.h
+LIB_H += util/include/linux/module.h
+LIB_H += util/include/linux/poison.h
+LIB_H += util/include/linux/prefetch.h
+LIB_H += util/include/linux/rbtree.h
+LIB_H += util/include/linux/string.h
+LIB_H += util/include/linux/types.h
+LIB_H += util/include/asm/asm-offsets.h
+LIB_H += util/include/asm/bitops.h
+LIB_H += util/include/asm/byteorder.h
+LIB_H += util/include/asm/swab.h
+LIB_H += util/include/asm/system.h
+LIB_H += util/include/asm/uaccess.h
LIB_H += perf.h
+LIB_H += util/debugfs.h
LIB_H += util/event.h
LIB_H += util/types.h
LIB_H += util/levenshtein.h
@@ -355,7 +387,9 @@ LIB_OBJS += util/abspath.o
LIB_OBJS += util/alias.o
LIB_OBJS += util/config.o
LIB_OBJS += util/ctype.o
+LIB_OBJS += util/debugfs.o
LIB_OBJS += util/environment.o
+LIB_OBJS += util/event.o
LIB_OBJS += util/exec_cmd.o
LIB_OBJS += util/help.o
LIB_OBJS += util/levenshtein.o
@@ -363,6 +397,9 @@ LIB_OBJS += util/parse-options.o
LIB_OBJS += util/parse-events.o
LIB_OBJS += util/path.o
LIB_OBJS += util/rbtree.o
+LIB_OBJS += util/bitmap.o
+LIB_OBJS += util/hweight.o
+LIB_OBJS += util/find_next_bit.o
LIB_OBJS += util/run-command.o
LIB_OBJS += util/quote.o
LIB_OBJS += util/strbuf.o
@@ -389,8 +426,16 @@ LIB_OBJS += util/hist.o
LIB_OBJS += util/data_map.o
BUILTIN_OBJS += builtin-annotate.o
+
+BUILTIN_OBJS += builtin-bench.o
+
+# Benchmark modules
+BUILTIN_OBJS += bench/sched-messaging.o
+BUILTIN_OBJS += bench/sched-pipe.o
+
BUILTIN_OBJS += builtin-help.o
BUILTIN_OBJS += builtin-sched.o
+BUILTIN_OBJS += builtin-buildid-list.o
BUILTIN_OBJS += builtin-list.o
BUILTIN_OBJS += builtin-record.o
BUILTIN_OBJS += builtin-report.o
@@ -398,6 +443,8 @@ BUILTIN_OBJS += builtin-stat.o
BUILTIN_OBJS += builtin-timechart.o
BUILTIN_OBJS += builtin-top.o
BUILTIN_OBJS += builtin-trace.o
+BUILTIN_OBJS += builtin-probe.o
+BUILTIN_OBJS += builtin-kmem.o
PERFLIBS = $(LIB_FILE)
@@ -428,27 +475,44 @@ ifeq ($(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 $(ALL_LDFLAGS) > /dev/null 2>&1 && echo y"), y)
+ifeq ($(shell sh -c "(echo '\#include <libelf.h>'; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) > /dev/null 2>&1 && echo y"), y)
+ifneq ($(shell sh -c "(echo '\#include <gnu/libc-version.h>'; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) > /dev/null 2>&1 && echo y"), y)
+ msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]);
+endif
+
+ ifneq ($(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) $(EXTLIBS) > /dev/null 2>&1 && echo y"), y)
+ BASIC_CFLAGS += -DLIBELF_NO_MMAP
+ endif
+else
msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]);
endif
+ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) > /dev/null 2>&1 && echo y"), y)
+ msg := $(warning No libdwarf.h found or old libdwarf.h found, disables dwarf support. Please install libdwarf-dev/libdwarf-devel >= 20081231);
+ BASIC_CFLAGS += -DNO_LIBDWARF
+else
+ EXTLIBS += -lelf -ldwarf
+ LIB_H += util/probe-finder.h
+ LIB_OBJS += util/probe-finder.o
+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")
+ 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) $(EXTLIBS) -lbfd > /dev/null 2>&1 && echo y")
ifeq ($(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")
+ 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) $(EXTLIBS) -lbfd -liberty > /dev/null 2>&1 && echo y")
ifeq ($(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")
+ 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) $(EXTLIBS) -lbfd -liberty -lz > /dev/null 2>&1 && echo y")
ifeq ($(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")
+ 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) $(EXTLIBS) -liberty > /dev/null 2>&1 && echo y")
ifeq ($(has_cplus_demangle),y)
EXTLIBS += -liberty
BASIC_CFLAGS += -DHAVE_CPLUS_DEMANGLE
@@ -790,6 +854,19 @@ util/config.o: util/config.c PERF-CFLAGS
util/rbtree.o: ../../lib/rbtree.c PERF-CFLAGS
$(QUIET_CC)$(CC) -o util/rbtree.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+# some perf warning policies can't fit to lib/bitmap.c, eg: it warns about variable shadowing
+# from <string.h> that comes from kernel headers wrapping.
+KBITMAP_FLAGS=`echo $(ALL_CFLAGS) | sed s/-Wshadow// | sed s/-Wswitch-default// | sed s/-Wextra//`
+
+util/bitmap.o: ../../lib/bitmap.c PERF-CFLAGS
+ $(QUIET_CC)$(CC) -o util/bitmap.o -c $(KBITMAP_FLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/hweight.o: ../../lib/hweight.c PERF-CFLAGS
+ $(QUIET_CC)$(CC) -o util/hweight.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
+util/find_next_bit.o: ../../lib/find_next_bit.c PERF-CFLAGS
+ $(QUIET_CC)$(CC) -o util/find_next_bit.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $<
+
perf-%$X: %.o $(PERFLIBS)
$(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS)
diff --git a/tools/perf/bench/bench.h b/tools/perf/bench/bench.h
new file mode 100644
index 000000000000..9fbd8d745fa1
--- /dev/null
+++ b/tools/perf/bench/bench.h
@@ -0,0 +1,16 @@
+#ifndef BENCH_H
+#define BENCH_H
+
+extern int bench_sched_messaging(int argc, const char **argv, const char *prefix);
+extern int bench_sched_pipe(int argc, const char **argv, const char *prefix);
+
+#define BENCH_FORMAT_DEFAULT_STR "default"
+#define BENCH_FORMAT_DEFAULT 0
+#define BENCH_FORMAT_SIMPLE_STR "simple"
+#define BENCH_FORMAT_SIMPLE 1
+
+#define BENCH_FORMAT_UNKNOWN -1
+
+extern int bench_format;
+
+#endif
diff --git a/tools/perf/bench/sched-messaging.c b/tools/perf/bench/sched-messaging.c
new file mode 100644
index 000000000000..605a2a959aa8
--- /dev/null
+++ b/tools/perf/bench/sched-messaging.c
@@ -0,0 +1,336 @@
+/*
+ *
+ * builtin-bench-messaging.c
+ *
+ * messaging: Benchmark for scheduler and IPC mechanisms
+ *
+ * Based on hackbench by Rusty Russell <rusty@rustcorp.com.au>
+ * Ported to perf by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/parse-options.h"
+#include "../builtin.h"
+#include "bench.h"
+
+/* Test groups of 20 processes spraying to 20 receivers */
+#include <pthread.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <sys/time.h>
+#include <sys/poll.h>
+#include <limits.h>
+
+#define DATASIZE 100
+
+static int use_pipes = 0;
+static unsigned int loops = 100;
+static unsigned int thread_mode = 0;
+static unsigned int num_groups = 10;
+
+struct sender_context {
+ unsigned int num_fds;
+ int ready_out;
+ int wakefd;
+ int out_fds[0];
+};
+
+struct receiver_context {
+ unsigned int num_packets;
+ int in_fds[2];
+ int ready_out;
+ int wakefd;
+};
+
+static void barf(const char *msg)
+{
+ fprintf(stderr, "%s (error: %s)\n", msg, strerror(errno));
+ exit(1);
+}
+
+static void fdpair(int fds[2])
+{
+ if (use_pipes) {
+ if (pipe(fds) == 0)
+ return;
+ } else {
+ if (socketpair(AF_UNIX, SOCK_STREAM, 0, fds) == 0)
+ return;
+ }
+
+ barf(use_pipes ? "pipe()" : "socketpair()");
+}
+
+/* Block until we're ready to go */
+static void ready(int ready_out, int wakefd)
+{
+ char dummy;
+ struct pollfd pollfd = { .fd = wakefd, .events = POLLIN };
+
+ /* Tell them we're ready. */
+ if (write(ready_out, &dummy, 1) != 1)
+ barf("CLIENT: ready write");
+
+ /* Wait for "GO" signal */
+ if (poll(&pollfd, 1, -1) != 1)
+ barf("poll");
+}
+
+/* Sender sprays loops messages down each file descriptor */
+static void *sender(struct sender_context *ctx)
+{
+ char data[DATASIZE];
+ unsigned int i, j;
+
+ ready(ctx->ready_out, ctx->wakefd);
+
+ /* Now pump to every receiver. */
+ for (i = 0; i < loops; i++) {
+ for (j = 0; j < ctx->num_fds; j++) {
+ int ret, done = 0;
+
+again:
+ ret = write(ctx->out_fds[j], data + done,
+ sizeof(data)-done);
+ if (ret < 0)
+ barf("SENDER: write");
+ done += ret;
+ if (done < DATASIZE)
+ goto again;
+ }
+ }
+
+ return NULL;
+}
+
+
+/* One receiver per fd */
+static void *receiver(struct receiver_context* ctx)
+{
+ unsigned int i;
+
+ if (!thread_mode)
+ close(ctx->in_fds[1]);
+
+ /* Wait for start... */
+ ready(ctx->ready_out, ctx->wakefd);
+
+ /* Receive them all */
+ for (i = 0; i < ctx->num_packets; i++) {
+ char data[DATASIZE];
+ int ret, done = 0;
+
+again:
+ ret = read(ctx->in_fds[0], data + done, DATASIZE - done);
+ if (ret < 0)
+ barf("SERVER: read");
+ done += ret;
+ if (done < DATASIZE)
+ goto again;
+ }
+
+ return NULL;
+}
+
+static pthread_t create_worker(void *ctx, void *(*func)(void *))
+{
+ pthread_attr_t attr;
+ pthread_t childid;
+ int err;
+
+ if (!thread_mode) {
+ /* process mode */
+ /* Fork the receiver. */
+ switch (fork()) {
+ case -1:
+ barf("fork()");
+ break;
+ case 0:
+ (*func) (ctx);
+ exit(0);
+ break;
+ default:
+ break;
+ }
+
+ return (pthread_t)0;
+ }
+
+ if (pthread_attr_init(&attr) != 0)
+ barf("pthread_attr_init:");
+
+#ifndef __ia64__
+ if (pthread_attr_setstacksize(&attr, PTHREAD_STACK_MIN) != 0)
+ barf("pthread_attr_setstacksize");
+#endif
+
+ err = pthread_create(&childid, &attr, func, ctx);
+ if (err != 0) {
+ fprintf(stderr, "pthread_create failed: %s (%d)\n",
+ strerror(err), err);
+ exit(-1);
+ }
+ return childid;
+}
+
+static void reap_worker(pthread_t id)
+{
+ int proc_status;
+ void *thread_status;
+
+ if (!thread_mode) {
+ /* process mode */
+ wait(&proc_status);
+ if (!WIFEXITED(proc_status))
+ exit(1);
+ } else {
+ pthread_join(id, &thread_status);
+ }
+}
+
+/* One group of senders and receivers */
+static unsigned int group(pthread_t *pth,
+ unsigned int num_fds,
+ int ready_out,
+ int wakefd)
+{
+ unsigned int i;
+ struct sender_context *snd_ctx = malloc(sizeof(struct sender_context)
+ + num_fds * sizeof(int));
+
+ if (!snd_ctx)
+ barf("malloc()");
+
+ for (i = 0; i < num_fds; i++) {
+ int fds[2];
+ struct receiver_context *ctx = malloc(sizeof(*ctx));
+
+ if (!ctx)
+ barf("malloc()");
+
+
+ /* Create the pipe between client and server */
+ fdpair(fds);
+
+ ctx->num_packets = num_fds * loops;
+ ctx->in_fds[0] = fds[0];
+ ctx->in_fds[1] = fds[1];
+ ctx->ready_out = ready_out;
+ ctx->wakefd = wakefd;
+
+ pth[i] = create_worker(ctx, (void *)receiver);
+
+ snd_ctx->out_fds[i] = fds[1];
+ if (!thread_mode)
+ close(fds[0]);
+ }
+
+ /* Now we have all the fds, fork the senders */
+ for (i = 0; i < num_fds; i++) {
+ snd_ctx->ready_out = ready_out;
+ snd_ctx->wakefd = wakefd;
+ snd_ctx->num_fds = num_fds;
+
+ pth[num_fds+i] = create_worker(snd_ctx, (void *)sender);
+ }
+
+ /* Close the fds we have left */
+ if (!thread_mode)
+ for (i = 0; i < num_fds; i++)
+ close(snd_ctx->out_fds[i]);
+
+ /* Return number of children to reap */
+ return num_fds * 2;
+}
+
+static const struct option options[] = {
+ OPT_BOOLEAN('p', "pipe", &use_pipes,
+ "Use pipe() instead of socketpair()"),
+ OPT_BOOLEAN('t', "thread", &thread_mode,
+ "Be multi thread instead of multi process"),
+ OPT_INTEGER('g', "group", &num_groups,
+ "Specify number of groups"),
+ OPT_INTEGER('l', "loop", &loops,
+ "Specify number of loops"),
+ OPT_END()
+};
+
+static const char * const bench_sched_message_usage[] = {
+ "perf bench sched messaging <options>",
+ NULL
+};
+
+int bench_sched_messaging(int argc, const char **argv,
+ const char *prefix __used)
+{
+ unsigned int i, total_children;
+ struct timeval start, stop, diff;
+ unsigned int num_fds = 20;
+ int readyfds[2], wakefds[2];
+ char dummy;
+ pthread_t *pth_tab;
+
+ argc = parse_options(argc, argv, options,
+ bench_sched_message_usage, 0);
+
+ pth_tab = malloc(num_fds * 2 * num_groups * sizeof(pthread_t));
+ if (!pth_tab)
+ barf("main:malloc()");
+
+ fdpair(readyfds);
+ fdpair(wakefds);
+
+ total_children = 0;
+ for (i = 0; i < num_groups; i++)
+ total_children += group(pth_tab+total_children, num_fds,
+ readyfds[1], wakefds[0]);
+
+ /* Wait for everyone to be ready */
+ for (i = 0; i < total_children; i++)
+ if (read(readyfds[0], &dummy, 1) != 1)
+ barf("Reading for readyfds");
+
+ gettimeofday(&start, NULL);
+
+ /* Kick them off */
+ if (write(wakefds[1], &dummy, 1) != 1)
+ barf("Writing to start them");
+
+ /* Reap them all */
+ for (i = 0; i < total_children; i++)
+ reap_worker(pth_tab[i]);
+
+ gettimeofday(&stop, NULL);
+
+ timersub(&stop, &start, &diff);
+
+ switch (bench_format) {
+ case BENCH_FORMAT_DEFAULT:
+ printf("# %d sender and receiver %s per group\n",
+ num_fds, thread_mode ? "threads" : "processes");
+ printf("# %d groups == %d %s run\n\n",
+ num_groups, num_groups * 2 * num_fds,
+ thread_mode ? "threads" : "processes");
+ printf(" %14s: %lu.%03lu [sec]\n", "Total time",
+ diff.tv_sec, diff.tv_usec/1000);
+ break;
+ case BENCH_FORMAT_SIMPLE:
+ printf("%lu.%03lu\n", diff.tv_sec, diff.tv_usec/1000);
+ break;
+ default:
+ /* reaching here is something disaster */
+ fprintf(stderr, "Unknown format:%d\n", bench_format);
+ exit(1);
+ break;
+ }
+
+ return 0;
+}
diff --git a/tools/perf/bench/sched-pipe.c b/tools/perf/bench/sched-pipe.c
new file mode 100644
index 000000000000..238185f97977
--- /dev/null
+++ b/tools/perf/bench/sched-pipe.c
@@ -0,0 +1,124 @@
+/*
+ *
+ * builtin-bench-pipe.c
+ *
+ * pipe: Benchmark for pipe()
+ *
+ * Based on pipe-test-1m.c by Ingo Molnar <mingo@redhat.com>
+ * http://people.redhat.com/mingo/cfs-scheduler/tools/pipe-test-1m.c
+ * Ported to perf by Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+#include "../perf.h"
+#include "../util/util.h"
+#include "../util/parse-options.h"
+#include "../builtin.h"
+#include "bench.h"
+
+#include <unistd.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <sys/wait.h>
+#include <linux/unistd.h>
+#include <string.h>
+#include <errno.h>
+#include <assert.h>
+#include <sys/time.h>
+#include <sys/types.h>
+
+#define LOOPS_DEFAULT 1000000
+static int loops = LOOPS_DEFAULT;
+
+static const struct option options[] = {
+ OPT_INTEGER('l', "loop", &loops,
+ "Specify number of loops"),
+ OPT_END()
+};
+
+static const char * const bench_sched_pipe_usage[] = {
+ "perf bench sched pipe <options>",
+ NULL
+};
+
+int bench_sched_pipe(int argc, const char **argv,
+ const char *prefix __used)
+{
+ int pipe_1[2], pipe_2[2];
+ int m = 0, i;
+ struct timeval start, stop, diff;
+ unsigned long long result_usec = 0;
+
+ /*
+ * why does "ret" exist?
+ * discarding returned value of read(), write()
+ * causes error in building environment for perf
+ */
+ int ret, wait_stat;
+ pid_t pid, retpid;
+
+ argc = parse_options(argc, argv, options,
+ bench_sched_pipe_usage, 0);
+
+ assert(!pipe(pipe_1));
+ assert(!pipe(pipe_2));
+
+ pid = fork();
+ assert(pid >= 0);
+
+ gettimeofday(&start, NULL);
+
+ if (!pid) {
+ for (i = 0; i < loops; i++) {
+ ret = read(pipe_1[0], &m, sizeof(int));
+ ret = write(pipe_2[1], &m, sizeof(int));
+ }
+ } else {
+ for (i = 0; i < loops; i++) {
+ ret = write(pipe_1[1], &m, sizeof(int));
+ ret = read(pipe_2[0], &m, sizeof(int));
+ }
+ }
+
+ gettimeofday(&stop, NULL);
+ timersub(&stop, &start, &diff);
+
+ if (pid) {
+ retpid = waitpid(pid, &wait_stat, 0);
+ assert((retpid == pid) && WIFEXITED(wait_stat));
+ return 0;
+ }
+
+ switch (bench_format) {
+ case BENCH_FORMAT_DEFAULT:
+ printf("# Extecuted %d pipe operations between two tasks\n\n",
+ loops);
+
+ result_usec = diff.tv_sec * 1000000;
+ result_usec += diff.tv_usec;
+
+ printf(" %14s: %lu.%03lu [sec]\n\n", "Total time",
+ diff.tv_sec, diff.tv_usec/1000);
+
+ printf(" %14lf usecs/op\n",
+ (double)result_usec / (double)loops);
+ printf(" %14d ops/sec\n",
+ (int)((double)loops /
+ ((double)result_usec / (double)1000000)));
+ break;
+
+ case BENCH_FORMAT_SIMPLE:
+ printf("%lu.%03lu\n",
+ diff.tv_sec, diff.tv_usec / 1000);
+ break;
+
+ default:
+ /* reaching here is something disaster */
+ fprintf(stderr, "Unknown format:%d\n", bench_format);
+ exit(1);
+ break;
+ }
+
+ return 0;
+}
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 56ba71658d70..77d50a6d6802 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -37,12 +37,45 @@ static int print_line;
static unsigned long page_size;
static unsigned long mmap_window = 32;
+struct sym_hist {
+ u64 sum;
+ u64 ip[0];
+};
+
struct sym_ext {
struct rb_node node;
double percent;
char *path;
};
+struct sym_priv {
+ struct sym_hist *hist;
+ struct sym_ext *ext;
+};
+
+static const char *sym_hist_filter;
+
+static int symbol_filter(struct map *map __used, struct symbol *sym)
+{
+ if (sym_hist_filter == NULL ||
+ strcmp(sym->name, sym_hist_filter) == 0) {
+ struct sym_priv *priv = symbol__priv(sym);
+ const int size = (sizeof(*priv->hist) +
+ (sym->end - sym->start) * sizeof(u64));
+
+ priv->hist = malloc(size);
+ if (priv->hist)
+ memset(priv->hist, 0, size);
+ return 0;
+ }
+ /*
+ * FIXME: We should really filter it out, as we don't want to go thru symbols
+ * we're not interested, and if a DSO ends up with no symbols, delete it too,
+ * but right now the kernel loading routines in symbol.c bail out if no symbols
+ * are found, fix it later.
+ */
+ return 0;
+}
/*
* collect histogram counts
@@ -51,28 +84,38 @@ static void hist_hit(struct hist_entry *he, u64 ip)
{
unsigned int sym_size, offset;
struct symbol *sym = he->sym;
+ struct sym_priv *priv;
+ struct sym_hist *h;
he->count++;
- if (!sym || !sym->hist)
+ if (!sym || !he->map)
+ return;
+
+ priv = symbol__priv(sym);
+ if (!priv->hist)
return;
sym_size = sym->end - sym->start;
- ip = he->map->map_ip(he->map, ip);
offset = ip - sym->start;
+ if (verbose)
+ fprintf(stderr, "%s: ip=%Lx\n", __func__,
+ he->map->unmap_ip(he->map, ip));
+
if (offset >= sym_size)
return;
- sym->hist_sum++;
- sym->hist[offset]++;
+ h = priv->hist;
+ h->sum++;
+ h->ip[offset]++;
if (verbose >= 3)
printf("%p %s: count++ [ip: %p, %08Lx] => %Ld\n",
(void *)(unsigned long)he->sym->start,
he->sym->name,
(void *)(unsigned long)ip, ip - he->sym->start,
- sym->hist[offset]);
+ h->ip[offset]);
}
static int hist_entry__add(struct thread *thread, struct map *map,
@@ -83,8 +126,7 @@ static int hist_entry__add(struct thread *thread, struct map *map,
count, level, &hit);
if (he == NULL)
return -ENOMEM;
- if (hit)
- hist_hit(he, ip);
+ hist_hit(he, ip);
return 0;
}
@@ -123,7 +165,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
if (map != NULL) {
got_map:
ip = map->map_ip(map, ip);
- sym = map->dso->find_symbol(map->dso, ip);
+ sym = map__find_symbol(map, ip, symbol_filter);
} else {
/*
* If this is outside of all known maps,
@@ -260,14 +302,15 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
return 0;
}
-static int
-parse_line(FILE *file, struct symbol *sym, u64 len)
+static int parse_line(FILE *file, struct hist_entry *he, u64 len)
{
+ struct symbol *sym = he->sym;
char *line = NULL, *tmp, *tmp2;
static const char *prev_line;
static const char *prev_color;
unsigned int offset;
size_t line_len;
+ u64 start;
s64 line_ip;
int ret;
char *c;
@@ -304,22 +347,26 @@ parse_line(FILE *file, struct symbol *sym, u64 len)
line_ip = -1;
}
+ start = he->map->unmap_ip(he->map, sym->start);
+
if (line_ip != -1) {
const char *path = NULL;
unsigned int hits = 0;
double percent = 0.0;
const char *color;
- struct sym_ext *sym_ext = sym->priv;
+ struct sym_priv *priv = symbol__priv(sym);
+ struct sym_ext *sym_ext = priv->ext;
+ struct sym_hist *h = priv->hist;
- offset = line_ip - sym->start;
+ offset = line_ip - start;
if (offset < len)
- hits = sym->hist[offset];
+ hits = h->ip[offset];
if (offset < len && sym_ext) {
path = sym_ext[offset].path;
percent = sym_ext[offset].percent;
- } else if (sym->hist_sum)
- percent = 100.0 * hits / sym->hist_sum;
+ } else if (h->sum)
+ percent = 100.0 * hits / h->sum;
color = get_percent_color(percent);
@@ -372,9 +419,10 @@ static void insert_source_line(struct sym_ext *sym_ext)
rb_insert_color(&sym_ext->node, &root_sym_ext);
}
-static void free_source_line(struct symbol *sym, int len)
+static void free_source_line(struct hist_entry *he, int len)
{
- struct sym_ext *sym_ext = sym->priv;
+ struct sym_priv *priv = symbol__priv(he->sym);
+ struct sym_ext *sym_ext = priv->ext;
int i;
if (!sym_ext)
@@ -384,26 +432,30 @@ static void free_source_line(struct symbol *sym, int len)
free(sym_ext[i].path);
free(sym_ext);
- sym->priv = NULL;
+ priv->ext = NULL;
root_sym_ext = RB_ROOT;
}
/* Get the filename:line for the colored entries */
static void
-get_source_line(struct symbol *sym, int len, const char *filename)
+get_source_line(struct hist_entry *he, int len, const char *filename)
{
+ struct symbol *sym = he->sym;
+ u64 start;
int i;
char cmd[PATH_MAX * 2];
struct sym_ext *sym_ext;
+ struct sym_priv *priv = symbol__priv(sym);
+ struct sym_hist *h = priv->hist;
- if (!sym->hist_sum)
+ if (!h->sum)
return;
- sym->priv = calloc(len, sizeof(struct sym_ext));
- if (!sym->priv)
+ sym_ext = priv->ext = calloc(len, sizeof(struct sym_ext));
+ if (!priv->ext)
return;
- sym_ext = sym->priv;
+ start = he->map->unmap_ip(he->map, sym->start);
for (i = 0; i < len; i++) {
char *path = NULL;
@@ -411,11 +463,11 @@ get_source_line(struct symbol *sym, int len, const char *filename)
u64 offset;
FILE *fp;
- sym_ext[i].percent = 100.0 * sym->hist[i] / sym->hist_sum;
+ sym_ext[i].percent = 100.0 * h->ip[i] / h->sum;
if (sym_ext[i].percent <= 0.5)
continue;
- offset = sym->start + i;
+ offset = start + i;
sprintf(cmd, "addr2line -e %s %016llx", filename, offset);
fp = popen(cmd, "r");
if (!fp)
@@ -465,8 +517,11 @@ static void print_summary(const char *filename)
}
}
-static void annotate_sym(struct dso *dso, struct symbol *sym)
+static void annotate_sym(struct hist_entry *he)
{
+ struct map *map = he->map;
+ struct dso *dso = map->dso;
+ struct symbol *sym = he->sym;
const char *filename = dso->long_name, *d_filename;
u64 len;
char command[PATH_MAX*2];
@@ -475,6 +530,12 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
if (!filename)
return;
+ if (verbose)
+ fprintf(stderr, "%s: filename=%s, sym=%s, start=%Lx, end=%Lx\n",
+ __func__, filename, sym->name,
+ map->unmap_ip(map, sym->start),
+ map->unmap_ip(map, sym->end));
+
if (full_paths)
d_filename = filename;
else
@@ -483,7 +544,7 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
len = sym->end - sym->start;
if (print_line) {
- get_source_line(sym, len, filename);
+ get_source_line(he, len, filename);
print_summary(filename);
}
@@ -496,7 +557,8 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
dso, dso->long_name, sym, sym->name);
sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s|grep -v %s",
- sym->start, sym->end, filename, filename);
+ map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end),
+ filename, filename);
if (verbose >= 3)
printf("doing: %s\n", command);
@@ -506,35 +568,38 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
return;
while (!feof(file)) {
- if (parse_line(file, sym, len) < 0)
+ if (parse_line(file, he, len) < 0)
break;
}
pclose(file);
if (print_line)
- free_source_line(sym, len);
+ free_source_line(he, len);
}
static void find_annotations(void)
{
struct rb_node *nd;
- struct dso *dso;
- int count = 0;
- list_for_each_entry(dso, &dsos, node) {
+ for (nd = rb_first(&output_hists); nd; nd = rb_next(nd)) {
+ struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node);
+ struct sym_priv *priv;
- for (nd = rb_first(&dso->syms); nd; nd = rb_next(nd)) {
- struct symbol *sym = rb_entry(nd, struct symbol, rb_node);
+ if (he->sym == NULL)
+ continue;
- if (sym->hist) {
- annotate_sym(dso, sym);
- count++;
- }
- }
- }
+ priv = symbol__priv(he->sym);
+ if (priv->hist == NULL)
+ continue;
- if (!count)
- printf(" Error: symbol '%s' not present amongst the samples.\n", sym_hist_filter);
+ annotate_sym(he);
+ /*
+ * Since we have a hist_entry per IP for the same symbol, free
+ * he->sym->hist to signal we already processed this symbol.
+ */
+ free(priv->hist);
+ priv->hist = NULL;
+ }
}
static int __cmd_annotate(void)
@@ -571,7 +636,7 @@ static int __cmd_annotate(void)
exit(0);
}
- if (load_kernel() < 0) {
+ if (load_kernel(symbol_filter) < 0) {
perror("failed to load kernel symbols");
return EXIT_FAILURE;
}
@@ -703,7 +768,7 @@ static void setup_sorting(void)
int cmd_annotate(int argc, const char **argv, const char *prefix __used)
{
- symbol__init();
+ symbol__init(sizeof(struct sym_priv));
page_size = getpagesize();
@@ -722,9 +787,6 @@ int cmd_annotate(int argc, const char **argv, const char *prefix __used)
sym_hist_filter = argv[0];
}
- if (!sym_hist_filter)
- usage_with_options(annotate_usage, options);
-
setup_pager();
if (field_sep && *field_sep == '.') {
diff --git a/tools/perf/builtin-bench.c b/tools/perf/builtin-bench.c
new file mode 100644
index 000000000000..90c39baae0de
--- /dev/null
+++ b/tools/perf/builtin-bench.c
@@ -0,0 +1,183 @@
+/*
+ *
+ * builtin-bench.c
+ *
+ * General benchmarking subsystem provided by perf
+ *
+ * Copyright (C) 2009, Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
+ *
+ */
+
+/*
+ *
+ * Available subsystem list:
+ * sched ... scheduler and IPC mechanism
+ *
+ */
+
+#include "perf.h"
+#include "util/util.h"
+#include "util/parse-options.h"
+#include "builtin.h"
+#include "bench/bench.h"
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+struct bench_suite {
+ const char *name;
+ const char *summary;
+ int (*fn)(int, const char **, const char *);
+};
+
+static struct bench_suite sched_suites[] = {
+ { "messaging",
+ "Benchmark for scheduler and IPC mechanisms",
+ bench_sched_messaging },
+ { "pipe",
+ "Flood of communication over pipe() between two processes",
+ bench_sched_pipe },
+ { NULL,
+ NULL,
+ NULL }
+};
+
+struct bench_subsys {
+ const char *name;
+ const char *summary;
+ struct bench_suite *suites;
+};
+
+static struct bench_subsys subsystems[] = {
+ { "sched",
+ "scheduler and IPC mechanism",
+ sched_suites },
+ { NULL,
+ NULL,
+ NULL }
+};
+
+static void dump_suites(int subsys_index)
+{
+ int i;
+
+ printf("List of available suites for %s...\n\n",
+ subsystems[subsys_index].name);
+
+ for (i = 0; subsystems[subsys_index].suites[i].name; i++)
+ printf("\t%s: %s\n",
+ subsystems[subsys_index].suites[i].name,
+ subsystems[subsys_index].suites[i].summary);
+
+ printf("\n");
+ return;
+}
+
+static char *bench_format_str;
+int bench_format = BENCH_FORMAT_DEFAULT;
+
+static const struct option bench_options[] = {
+ OPT_STRING('f', "format", &bench_format_str, "default",
+ "Specify format style"),
+ OPT_END()
+};
+
+static const char * const bench_usage[] = {
+ "perf bench [<common options>] <subsystem> <suite> [<options>]",
+ NULL
+};
+
+static void print_usage(void)
+{
+ int i;
+
+ printf("Usage: \n");
+ for (i = 0; bench_usage[i]; i++)
+ printf("\t%s\n", bench_usage[i]);
+ printf("\n");
+
+ printf("List of available subsystems...\n\n");
+
+ for (i = 0; subsystems[i].name; i++)
+ printf("\t%s: %s\n",
+ subsystems[i].name, subsystems[i].summary);
+ printf("\n");
+}
+
+static int bench_str2int(char *str)
+{
+ if (!str)
+ return BENCH_FORMAT_DEFAULT;
+
+ if (!strcmp(str, BENCH_FORMAT_DEFAULT_STR))
+ return BENCH_FORMAT_DEFAULT;
+ else if (!strcmp(str, BENCH_FORMAT_SIMPLE_STR))
+ return BENCH_FORMAT_SIMPLE;
+
+ return BENCH_FORMAT_UNKNOWN;
+}
+
+int cmd_bench(int argc, const char **argv, const char *prefix __used)
+{
+ int i, j, status = 0;
+
+ if (argc < 2) {
+ /* No subsystem specified. */
+ print_usage();
+ goto end;
+ }
+
+ argc = parse_options(argc, argv, bench_options, bench_usage,
+ PARSE_OPT_STOP_AT_NON_OPTION);
+
+ bench_format = bench_str2int(bench_format_str);
+ if (bench_format == BENCH_FORMAT_UNKNOWN) {
+ printf("Unknown format descriptor:%s\n", bench_format_str);
+ goto end;
+ }
+
+ if (argc < 1) {
+ print_usage();
+ goto end;
+ }
+
+ for (i = 0; subsystems[i].name; i++) {
+ if (strcmp(subsystems[i].name, argv[0]))
+ continue;
+
+ if (argc < 2) {
+ /* No suite specified. */
+ dump_suites(i);
+ goto end;
+ }
+
+ for (j = 0; subsystems[i].suites[j].name; j++) {
+ if (strcmp(subsystems[i].suites[j].name, argv[1]))
+ continue;
+
+ if (bench_format == BENCH_FORMAT_DEFAULT)
+ printf("# Running %s/%s benchmark...\n",
+ subsystems[i].name,
+ subsystems[i].suites[j].name);
+ status = subsystems[i].suites[j].fn(argc - 1,
+ argv + 1, prefix);
+ goto end;
+ }
+
+ if (!strcmp(argv[1], "-h") || !strcmp(argv[1], "--help")) {
+ dump_suites(i);
+ goto end;
+ }
+
+ printf("Unknown suite:%s for %s\n", argv[1], argv[0]);
+ status = 1;
+ goto end;
+ }
+
+ printf("Unknown subsystem:%s\n", argv[0]);
+ status = 1;
+
+end:
+ return status;
+}
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
new file mode 100644
index 000000000000..7dee9d19ab7a
--- /dev/null
+++ b/tools/perf/builtin-buildid-list.c
@@ -0,0 +1,116 @@
+/*
+ * builtin-buildid-list.c
+ *
+ * Builtin buildid-list command: list buildids in perf.data
+ *
+ * Copyright (C) 2009, Red Hat Inc.
+ * Copyright (C) 2009, Arnaldo Carvalho de Melo <acme@redhat.com>
+ */
+#include "builtin.h"
+#include "perf.h"
+#include "util/cache.h"
+#include "util/data_map.h"
+#include "util/debug.h"
+#include "util/header.h"
+#include "util/parse-options.h"
+#include "util/symbol.h"
+
+static char const *input_name = "perf.data";
+static int force;
+
+static const char *const buildid_list_usage[] = {
+ "perf report [<options>]",
+ NULL
+};
+
+static const struct option options[] = {
+ OPT_STRING('i', "input", &input_name, "file",
+ "input file name"),
+ OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
+ OPT_BOOLEAN('v', "verbose", &verbose,
+ "be more verbose"),
+ OPT_END()
+};
+
+static int perf_file_section__process_buildids(struct perf_file_section *self,
+ int feat, int fd)
+{
+ if (feat != HEADER_BUILD_ID)
+ return 0;
+
+ if (lseek(fd, self->offset, SEEK_SET) < 0) {
+ pr_warning("Failed to lseek to %Ld offset for buildids!\n",
+ self->offset);
+ return -1;
+ }
+
+ if (perf_header__read_build_ids(fd, self->offset, self->size)) {
+ pr_warning("Failed to read buildids!\n");
+ return -1;
+ }
+
+ return 0;
+}
+
+static int __cmd_buildid_list(void)
+{
+ int err = -1;
+ struct perf_header *header;
+ struct perf_file_header f_header;
+ struct stat input_stat;
+ int input = open(input_name, O_RDONLY);
+
+ if (input < 0) {
+ pr_err("failed to open file: %s", input_name);
+ if (!strcmp(input_name, "perf.data"))
+ pr_err(" (try 'perf record' first)");
+ pr_err("\n");
+ goto out;
+ }
+
+ err = fstat(input, &input_stat);
+ if (err < 0) {
+ perror("failed to stat file");
+ goto out_close;
+ }
+
+ if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
+ pr_err("file %s not owned by current user or root\n",
+ input_name);
+ goto out_close;
+ }
+
+ if (!input_stat.st_size) {
+ pr_info("zero-sized file, nothing to do!\n");
+ goto out_close;
+ }
+
+ err = -1;
+ header = perf_header__new();
+ if (header == NULL)
+ goto out_close;
+
+ if (perf_file_header__read(&f_header, header, input) < 0) {
+ pr_warning("incompatible file format");
+ goto out_close;
+ }
+
+ err = perf_header__process_sections(header, input,
+ perf_file_section__process_buildids);
+
+ if (err < 0)
+ goto out_close;
+
+ dsos__fprintf_buildid(stdout);
+out_close:
+ close(input);
+out:
+ return err;
+}
+
+int cmd_buildid_list(int argc, const char **argv, const char *prefix __used)
+{
+ argc = parse_options(argc, argv, options, buildid_list_usage, 0);
+ setup_pager();
+ return __cmd_buildid_list();
+}
diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
index 4fb8734a796e..768f9c826312 100644
--- a/tools/perf/builtin-help.c
+++ b/tools/perf/builtin-help.c
@@ -61,8 +61,7 @@ static const char *get_man_viewer_info(const char *name)
{
struct man_viewer_info_list *viewer;
- for (viewer = man_viewer_info_list; viewer; viewer = viewer->next)
- {
+ for (viewer = man_viewer_info_list; viewer; viewer = viewer->next) {
if (!strcasecmp(name, viewer->name))
return viewer->info;
}
@@ -115,7 +114,7 @@ static int check_emacsclient_version(void)
return 0;
}
-static void exec_woman_emacs(const char* path, const char *page)
+static void exec_woman_emacs(const char *path, const char *page)
{
if (!check_emacsclient_version()) {
/* This works only with emacsclient version >= 22. */
@@ -129,7 +128,7 @@ static void exec_woman_emacs(const char* path, const char *page)
}
}
-static void exec_man_konqueror(const char* path, const char *page)
+static void exec_man_konqueror(const char *path, const char *page)
{
const char *display = getenv("DISPLAY");
if (display && *display) {
@@ -157,7 +156,7 @@ static void exec_man_konqueror(const char* path, const char *page)
}
}
-static void exec_man_man(const char* path, const char *page)
+static void exec_man_man(const char *path, const char *page)
{
if (!path)
path = "man";
@@ -364,9 +363,8 @@ static void show_man_page(const char *perf_cmd)
setup_man_path();
for (viewer = man_viewer_list; viewer; viewer = viewer->next)
- {
exec_viewer(viewer->name, page); /* will return when unable */
- }
+
if (fallback)
exec_viewer(fallback, page);
exec_viewer("man", page);
diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c
new file mode 100644
index 000000000000..f315b052f819
--- /dev/null
+++ b/tools/perf/builtin-kmem.c
@@ -0,0 +1,578 @@
+#include "builtin.h"
+#include "perf.h"
+
+#include "util/util.h"
+#include "util/cache.h"
+#include "util/symbol.h"
+#include "util/thread.h"
+#include "util/header.h"
+
+#include "util/parse-options.h"
+#include "util/trace-event.h"
+
+#include "util/debug.h"
+#include "util/data_map.h"
+
+#include <linux/rbtree.h>
+
+struct alloc_stat;
+typedef int (*sort_fn_t)(struct alloc_stat *, struct alloc_stat *);
+
+static char const *input_name = "perf.data";
+
+static struct perf_header *header;
+static u64 sample_type;
+
+static int alloc_flag;
+static int caller_flag;
+
+sort_fn_t alloc_sort_fn;
+sort_fn_t caller_sort_fn;
+
+static int alloc_lines = -1;
+static int caller_lines = -1;
+
+static char *cwd;
+static int cwdlen;
+
+struct alloc_stat {
+ union {
+ struct {
+ char *name;
+ u64 call_site;
+ };
+ u64 ptr;
+ };
+ u64 bytes_req;
+ u64 bytes_alloc;
+ u32 hit;
+
+ struct rb_node node;
+};
+
+static struct rb_root root_alloc_stat;
+static struct rb_root root_alloc_sorted;
+static struct rb_root root_caller_stat;
+static struct rb_root root_caller_sorted;
+
+static unsigned long total_requested, total_allocated;
+
+struct raw_event_sample {
+ u32 size;
+ char data[0];
+};
+
+static int
+process_comm_event(event_t *event, unsigned long offset, unsigned long head)
+{
+ struct thread *thread = threads__findnew(event->comm.pid);
+
+ dump_printf("%p [%p]: PERF_RECORD_COMM: %s:%d\n",
+ (void *)(offset + head),
+ (void *)(long)(event->header.size),
+ event->comm.comm, event->comm.pid);
+
+ if (thread == NULL ||
+ thread__set_comm(thread, event->comm.comm)) {
+ dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
+ return -1;
+ }
+
+ return 0;
+}
+
+static void insert_alloc_stat(unsigned long ptr,
+ int bytes_req, int bytes_alloc)
+{
+ struct rb_node **node = &root_alloc_stat.rb_node;
+ struct rb_node *parent = NULL;
+ struct alloc_stat *data = NULL;
+
+ if (!alloc_flag)
+ return;
+
+ while (*node) {
+ parent = *node;
+ data = rb_entry(*node, struct alloc_stat, node);
+
+ if (ptr > data->ptr)
+ node = &(*node)->rb_right;
+ else if (ptr < data->ptr)
+ node = &(*node)->rb_left;
+ else
+ break;
+ }
+
+ if (data && data->ptr == ptr) {
+ data->hit++;
+ data->bytes_req += bytes_req;
+ data->bytes_alloc += bytes_req;
+ } else {
+ data = malloc(sizeof(*data));
+ data->ptr = ptr;
+ data->hit = 1;
+ data->bytes_req = bytes_req;
+ data->bytes_alloc = bytes_alloc;
+
+ rb_link_node(&data->node, parent, node);
+ rb_insert_color(&data->node, &root_alloc_stat);
+ }
+}
+
+static void insert_caller_stat(unsigned long call_site,
+ int bytes_req, int bytes_alloc)
+{
+ struct rb_node **node = &root_caller_stat.rb_node;
+ struct rb_node *parent = NULL;
+ struct alloc_stat *data = NULL;
+
+ if (!caller_flag)
+ return;
+
+ while (*node) {
+ parent = *node;
+ data = rb_entry(*node, struct alloc_stat, node);
+
+ if (call_site > data->call_site)
+ node = &(*node)->rb_right;
+ else if (call_site < data->call_site)
+ node = &(*node)->rb_left;
+ else
+ break;
+ }
+
+ if (data && data->call_site == call_site) {
+ data->hit++;
+ data->bytes_req += bytes_req;
+ data->bytes_alloc += bytes_req;
+ } else {
+ data = malloc(sizeof(*data));
+ data->call_site = call_site;
+ data->hit = 1;
+ data->bytes_req = bytes_req;
+ data->bytes_alloc = bytes_alloc;
+
+ rb_link_node(&data->node, parent, node);
+ rb_insert_color(&data->node, &root_caller_stat);
+ }
+}
+
+static void process_alloc_event(struct raw_event_sample *raw,
+ struct event *event,
+ int cpu __used,
+ u64 timestamp __used,
+ struct thread *thread __used,
+ int node __used)
+{
+ unsigned long call_site;
+ unsigned long ptr;
+ int bytes_req;
+ int bytes_alloc;
+
+ ptr = raw_field_value(event, "ptr", raw->data);
+ call_site = raw_field_value(event, "call_site", raw->data);
+ bytes_req = raw_field_value(event, "bytes_req", raw->data);
+ bytes_alloc = raw_field_value(event, "bytes_alloc", raw->data);
+
+ insert_alloc_stat(ptr, bytes_req, bytes_alloc);
+ insert_caller_stat(call_site, bytes_req, bytes_alloc);
+
+ total_requested += bytes_req;
+ total_allocated += bytes_alloc;
+}
+
+static void process_free_event(struct raw_event_sample *raw __used,
+ struct event *event __used,
+ int cpu __used,
+ u64 timestamp __used,
+ struct thread *thread __used)
+{
+}
+
+static void
+process_raw_event(event_t *raw_event __used, void *more_data,
+ int cpu, u64 timestamp, struct thread *thread)
+{
+ struct raw_event_sample *raw = more_data;
+ struct event *event;
+ int type;
+
+ type = trace_parse_common_type(raw->data);
+ event = trace_find_event(type);
+
+ if (!strcmp(event->name, "kmalloc") ||
+ !strcmp(event->name, "kmem_cache_alloc")) {
+ process_alloc_event(raw, event, cpu, timestamp, thread, 0);
+ return;
+ }
+
+ if (!strcmp(event->name, "kmalloc_node") ||
+ !strcmp(event->name, "kmem_cache_alloc_node")) {
+ process_alloc_event(raw, event, cpu, timestamp, thread, 1);
+ return;
+ }
+
+ if (!strcmp(event->name, "kfree") ||
+ !strcmp(event->name, "kmem_cache_free")) {
+ process_free_event(raw, event, cpu, timestamp, thread);
+ return;
+ }
+}
+
+static int
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
+{
+ u64 ip = event->ip.ip;
+ u64 timestamp = -1;
+ u32 cpu = -1;
+ u64 period = 1;
+ void *more_data = event->ip.__more_data;
+ struct thread *thread = threads__findnew(event->ip.pid);
+
+ if (sample_type & PERF_SAMPLE_TIME) {
+ timestamp = *(u64 *)more_data;
+ more_data += sizeof(u64);
+ }
+
+ if (sample_type & PERF_SAMPLE_CPU) {
+ cpu = *(u32 *)more_data;
+ more_data += sizeof(u32);
+ more_data += sizeof(u32); /* reserved */
+ }
+
+ if (sample_type & PERF_SAMPLE_PERIOD) {
+ period = *(u64 *)more_data;
+ more_data += sizeof(u64);
+ }
+
+ dump_printf("%p [%p]: PERF_RECORD_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.tid,
+ (void *)(long)ip,
+ (long long)period);
+
+ if (thread == NULL) {
+ pr_debug("problem processing %d event, skipping it.\n",
+ event->header.type);
+ return -1;
+ }
+
+ dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+
+ process_raw_event(event, more_data, cpu, timestamp, thread);
+
+ return 0;
+}
+
+static int sample_type_check(u64 type)
+{
+ sample_type = type;
+
+ if (!(sample_type & PERF_SAMPLE_RAW)) {
+ fprintf(stderr,
+ "No trace sample to read. Did you call perf record "
+ "without -R?");
+ return -1;
+ }
+
+ return 0;
+}
+
+static struct perf_file_handler file_handler = {
+ .process_sample_event = process_sample_event,
+ .process_comm_event = process_comm_event,
+ .sample_type_check = sample_type_check,
+};
+
+static int read_events(void)
+{
+ register_idle_thread();
+ register_perf_file_handler(&file_handler);
+
+ return mmap_dispatch_perf_file(&header, input_name, 0, 0,
+ &cwdlen, &cwd);
+}
+
+static double fragmentation(unsigned long n_req, unsigned long n_alloc)
+{
+ if (n_alloc == 0)
+ return 0.0;
+ else
+ return 100.0 - (100.0 * n_req / n_alloc);
+}
+
+static void __print_result(struct rb_root *root, int n_lines, int is_caller)
+{
+ struct rb_node *next;
+
+ printf("\n ------------------------------------------------------------------------------\n");
+ if (is_caller)
+ printf(" Callsite |");
+ else
+ printf(" Alloc Ptr |");
+ printf(" Total_alloc/Per | Total_req/Per | Hit | Fragmentation\n");
+ printf(" ------------------------------------------------------------------------------\n");
+
+ next = rb_first(root);
+
+ while (next && n_lines--) {
+ struct alloc_stat *data;
+
+ data = rb_entry(next, struct alloc_stat, node);
+
+ printf(" %-16p | %8llu/%-6lu | %8llu/%-6lu | %6lu | %8.3f%%\n",
+ is_caller ? (void *)(unsigned long)data->call_site :
+ (void *)(unsigned long)data->ptr,
+ (unsigned long long)data->bytes_alloc,
+ (unsigned long)data->bytes_alloc / data->hit,
+ (unsigned long long)data->bytes_req,
+ (unsigned long)data->bytes_req / data->hit,
+ (unsigned long)data->hit,
+ fragmentation(data->bytes_req, data->bytes_alloc));
+
+ next = rb_next(next);
+ }
+
+ if (n_lines == -1)
+ printf(" ... | ... | ... | ... | ... \n");
+
+ printf(" ------------------------------------------------------------------------------\n");
+}
+
+static void print_summary(void)
+{
+ printf("\nSUMMARY\n=======\n");
+ printf("Total bytes requested: %lu\n", total_requested);
+ printf("Total bytes allocated: %lu\n", total_allocated);
+ printf("Total bytes wasted on internal fragmentation: %lu\n",
+ total_allocated - total_requested);
+ printf("Internal fragmentation: %f%%\n",
+ fragmentation(total_requested, total_allocated));
+}
+
+static void print_result(void)
+{
+ if (caller_flag)
+ __print_result(&root_caller_sorted, caller_lines, 1);
+ if (alloc_flag)
+ __print_result(&root_alloc_sorted, alloc_lines, 0);
+ print_summary();
+}
+
+static void sort_insert(struct rb_root *root, struct alloc_stat *data,
+ sort_fn_t sort_fn)
+{
+ struct rb_node **new = &(root->rb_node);
+ struct rb_node *parent = NULL;
+
+ while (*new) {
+ struct alloc_stat *this;
+ int cmp;
+
+ this = rb_entry(*new, struct alloc_stat, node);
+ parent = *new;
+
+ cmp = sort_fn(data, this);
+
+ if (cmp > 0)
+ new = &((*new)->rb_left);
+ else
+ new = &((*new)->rb_right);
+ }
+
+ rb_link_node(&data->node, parent, new);
+ rb_insert_color(&data->node, root);
+}
+
+static void __sort_result(struct rb_root *root, struct rb_root *root_sorted,
+ sort_fn_t sort_fn)
+{
+ struct rb_node *node;
+ struct alloc_stat *data;
+
+ for (;;) {
+ node = rb_first(root);
+ if (!node)
+ break;
+
+ rb_erase(node, root);
+ data = rb_entry(node, struct alloc_stat, node);
+ sort_insert(root_sorted, data, sort_fn);
+ }
+}
+
+static void sort_result(void)
+{
+ __sort_result(&root_alloc_stat, &root_alloc_sorted, alloc_sort_fn);
+ __sort_result(&root_caller_stat, &root_caller_sorted, caller_sort_fn);
+}
+
+static int __cmd_kmem(void)
+{
+ setup_pager();
+ read_events();
+ sort_result();
+ print_result();
+
+ return 0;
+}
+
+static const char * const kmem_usage[] = {
+ "perf kmem [<options>] {record}",
+ NULL
+};
+
+
+static int ptr_cmp(struct alloc_stat *l, struct alloc_stat *r)
+{
+ if (l->ptr < r->ptr)
+ return -1;
+ else if (l->ptr > r->ptr)
+ return 1;
+ return 0;
+}
+
+static int callsite_cmp(struct alloc_stat *l, struct alloc_stat *r)
+{
+ if (l->call_site < r->call_site)
+ return -1;
+ else if (l->call_site > r->call_site)
+ return 1;
+ return 0;
+}
+
+static int bytes_cmp(struct alloc_stat *l, struct alloc_stat *r)
+{
+ if (l->bytes_alloc < r->bytes_alloc)
+ return -1;
+ else if (l->bytes_alloc > r->bytes_alloc)
+ return 1;
+ return 0;
+}
+
+static int parse_sort_opt(const struct option *opt __used,
+ const char *arg, int unset __used)
+{
+ sort_fn_t sort_fn;
+
+ if (!arg)
+ return -1;
+
+ if (strcmp(arg, "ptr") == 0)
+ sort_fn = ptr_cmp;
+ else if (strcmp(arg, "call_site") == 0)
+ sort_fn = callsite_cmp;
+ else if (strcmp(arg, "bytes") == 0)
+ sort_fn = bytes_cmp;
+ else
+ return -1;
+
+ if (caller_flag > alloc_flag)
+ caller_sort_fn = sort_fn;
+ else
+ alloc_sort_fn = sort_fn;
+
+ return 0;
+}
+
+static int parse_stat_opt(const struct option *opt __used,
+ const char *arg, int unset __used)
+{
+ if (!arg)
+ return -1;
+
+ if (strcmp(arg, "alloc") == 0)
+ alloc_flag = (caller_flag + 1);
+ else if (strcmp(arg, "caller") == 0)
+ caller_flag = (alloc_flag + 1);
+ else
+ return -1;
+ return 0;
+}
+
+static int parse_line_opt(const struct option *opt __used,
+ const char *arg, int unset __used)
+{
+ int lines;
+
+ if (!arg)
+ return -1;
+
+ lines = strtoul(arg, NULL, 10);
+
+ if (caller_flag > alloc_flag)
+ caller_lines = lines;
+ else
+ alloc_lines = lines;
+
+ return 0;
+}
+
+static const struct option kmem_options[] = {
+ OPT_STRING('i', "input", &input_name, "file",
+ "input file name"),
+ OPT_CALLBACK(0, "stat", NULL, "<alloc>|<caller>",
+ "stat selector, Pass 'alloc' or 'caller'.",
+ parse_stat_opt),
+ OPT_CALLBACK('s', "sort", NULL, "key",
+ "sort by key: ptr, call_site, hit, bytes",
+ parse_sort_opt),
+ OPT_CALLBACK('l', "line", NULL, "num",
+ "show n lins",
+ parse_line_opt),
+ OPT_END()
+};
+
+static const char *record_args[] = {
+ "record",
+ "-a",
+ "-R",
+ "-M",
+ "-f",
+ "-c", "1",
+ "-e", "kmem:kmalloc",
+ "-e", "kmem:kmalloc_node",
+ "-e", "kmem:kfree",
+ "-e", "kmem:kmem_cache_alloc",
+ "-e", "kmem:kmem_cache_alloc_node",
+ "-e", "kmem:kmem_cache_free",
+};
+
+static int __cmd_record(int argc, const char **argv)
+{
+ unsigned int rec_argc, i, j;
+ const char **rec_argv;
+
+ rec_argc = ARRAY_SIZE(record_args) + argc - 1;
+ rec_argv = calloc(rec_argc + 1, sizeof(char *));
+
+ for (i = 0; i < ARRAY_SIZE(record_args); i++)
+ rec_argv[i] = strdup(record_args[i]);
+
+ for (j = 1; j < (unsigned int)argc; j++, i++)
+ rec_argv[i] = argv[j];
+
+ return cmd_record(i, rec_argv, NULL);
+}
+
+int cmd_kmem(int argc, const char **argv, const char *prefix __used)
+{
+ symbol__init(0);
+
+ argc = parse_options(argc, argv, kmem_options, kmem_usage, 0);
+
+ if (argc && !strncmp(argv[0], "rec", 3))
+ return __cmd_record(argc, argv);
+ else if (argc)
+ usage_with_options(kmem_usage, kmem_options);
+
+ if (!alloc_sort_fn)
+ alloc_sort_fn = bytes_cmp;
+ if (!caller_sort_fn)
+ caller_sort_fn = bytes_cmp;
+
+ return __cmd_kmem();
+}
+
diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c
new file mode 100644
index 000000000000..d78a3d945492
--- /dev/null
+++ b/tools/perf/builtin-probe.c
@@ -0,0 +1,435 @@
+/*
+ * builtin-probe.c
+ *
+ * Builtin probe command: Set up probe events by C expression
+ *
+ * Written by Masami Hiramatsu <mhiramat@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ */
+#define _GNU_SOURCE
+#include <sys/utsname.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <string.h>
+
+#undef _GNU_SOURCE
+#include "perf.h"
+#include "builtin.h"
+#include "util/util.h"
+#include "util/event.h"
+#include "util/debug.h"
+#include "util/parse-options.h"
+#include "util/parse-events.h" /* For debugfs_path */
+#include "util/probe-finder.h"
+
+/* Default vmlinux search paths */
+#define NR_SEARCH_PATH 3
+const char *default_search_path[NR_SEARCH_PATH] = {
+"/lib/modules/%s/build/vmlinux", /* Custom build kernel */
+"/usr/lib/debug/lib/modules/%s/vmlinux", /* Red Hat debuginfo */
+"/boot/vmlinux-debug-%s", /* Ubuntu */
+};
+
+#define MAX_PATH_LEN 256
+#define MAX_PROBES 128
+#define MAX_PROBE_ARGS 128
+#define PERFPROBE_GROUP "probe"
+
+/* Session management structure */
+static struct {
+ char *vmlinux;
+ char *release;
+ int need_dwarf;
+ int nr_probe;
+ struct probe_point probes[MAX_PROBES];
+} session;
+
+#define semantic_error(msg ...) die("Semantic error :" msg)
+
+/* Parse probe point. Return 1 if return probe */
+static void parse_probe_point(char *arg, struct probe_point *pp)
+{
+ char *ptr, *tmp;
+ char c, nc = 0;
+ /*
+ * <Syntax>
+ * perf probe SRC:LN
+ * perf probe FUNC[+OFFS|%return][@SRC]
+ */
+
+ ptr = strpbrk(arg, ":+@%");
+ if (ptr) {
+ nc = *ptr;
+ *ptr++ = '\0';
+ }
+
+ /* Check arg is function or file and copy it */
+ if (strchr(arg, '.')) /* File */
+ pp->file = strdup(arg);
+ else /* Function */
+ pp->function = strdup(arg);
+ DIE_IF(pp->file == NULL && pp->function == NULL);
+
+ /* Parse other options */
+ while (ptr) {
+ arg = ptr;
+ c = nc;
+ ptr = strpbrk(arg, ":+@%");
+ if (ptr) {
+ nc = *ptr;
+ *ptr++ = '\0';
+ }
+ switch (c) {
+ case ':': /* Line number */
+ pp->line = strtoul(arg, &tmp, 0);
+ if (*tmp != '\0')
+ semantic_error("There is non-digit charactor"
+ " in line number.");
+ break;
+ case '+': /* Byte offset from a symbol */
+ pp->offset = strtoul(arg, &tmp, 0);
+ if (*tmp != '\0')
+ semantic_error("There is non-digit charactor"
+ " in offset.");
+ break;
+ case '@': /* File name */
+ if (pp->file)
+ semantic_error("SRC@SRC is not allowed.");
+ pp->file = strdup(arg);
+ DIE_IF(pp->file == NULL);
+ if (ptr)
+ semantic_error("@SRC must be the last "
+ "option.");
+ break;
+ case '%': /* Probe places */
+ if (strcmp(arg, "return") == 0) {
+ pp->retprobe = 1;
+ } else /* Others not supported yet */
+ semantic_error("%%%s is not supported.", arg);
+ break;
+ default:
+ DIE_IF("Program has a bug.");
+ break;
+ }
+ }
+
+ /* Exclusion check */
+ if (pp->line && pp->offset)
+ semantic_error("Offset can't be used with line number.");
+ if (!pp->line && pp->file && !pp->function)
+ semantic_error("File always requires line number.");
+ if (pp->offset && !pp->function)
+ semantic_error("Offset requires an entry function.");
+ if (pp->retprobe && !pp->function)
+ semantic_error("Return probe requires an entry function.");
+ if ((pp->offset || pp->line) && pp->retprobe)
+ semantic_error("Offset/Line can't be used with return probe.");
+
+ pr_debug("symbol:%s file:%s line:%d offset:%d, return:%d\n",
+ pp->function, pp->file, pp->line, pp->offset, pp->retprobe);
+}
+
+/* Parse an event definition. Note that any error must die. */
+static void parse_probe_event(const char *str)
+{
+ char *argv[MAX_PROBE_ARGS + 2]; /* Event + probe + args */
+ int argc, i;
+ struct probe_point *pp = &session.probes[session.nr_probe];
+
+ pr_debug("probe-definition(%d): %s\n", session.nr_probe, str);
+ if (++session.nr_probe == MAX_PROBES)
+ semantic_error("Too many probes");
+
+ /* Separate arguments, similar to argv_split */
+ argc = 0;
+ do {
+ /* Skip separators */
+ while (isspace(*str))
+ str++;
+
+ /* Add an argument */
+ if (*str != '\0') {
+ const char *s = str;
+
+ /* Skip the argument */
+ while (!isspace(*str) && *str != '\0')
+ str++;
+
+ /* Duplicate the argument */
+ argv[argc] = strndup(s, str - s);
+ if (argv[argc] == NULL)
+ die("strndup");
+ if (++argc == MAX_PROBE_ARGS)
+ semantic_error("Too many arguments");
+ pr_debug("argv[%d]=%s\n", argc, argv[argc - 1]);
+ }
+ } while (*str != '\0');
+ if (!argc)
+ semantic_error("An empty argument.");
+
+ /* Parse probe point */
+ parse_probe_point(argv[0], pp);
+ free(argv[0]);
+ if (pp->file || pp->line)
+ session.need_dwarf = 1;
+
+ /* Copy arguments */
+ pp->nr_args = argc - 1;
+ if (pp->nr_args > 0) {
+ pp->args = (char **)malloc(sizeof(char *) * pp->nr_args);
+ if (!pp->args)
+ die("malloc");
+ memcpy(pp->args, &argv[1], sizeof(char *) * pp->nr_args);
+ }
+
+ /* Ensure return probe has no C argument */
+ for (i = 0; i < pp->nr_args; i++)
+ if (is_c_varname(pp->args[i])) {
+ if (pp->retprobe)
+ semantic_error("You can't specify local"
+ " variable for kretprobe");
+ session.need_dwarf = 1;
+ }
+
+ pr_debug("%d arguments\n", pp->nr_args);
+}
+
+static int opt_add_probe_event(const struct option *opt __used,
+ const char *str, int unset __used)
+{
+ if (str)
+ parse_probe_event(str);
+ return 0;
+}
+
+#ifndef NO_LIBDWARF
+static int open_default_vmlinux(void)
+{
+ struct utsname uts;
+ char fname[MAX_PATH_LEN];
+ int fd, ret, i;
+
+ ret = uname(&uts);
+ if (ret) {
+ pr_debug("uname() failed.\n");
+ return -errno;
+ }
+ session.release = uts.release;
+ for (i = 0; i < NR_SEARCH_PATH; i++) {
+ ret = snprintf(fname, MAX_PATH_LEN,
+ default_search_path[i], session.release);
+ if (ret >= MAX_PATH_LEN || ret < 0) {
+ pr_debug("Filename(%d,%s) is too long.\n", i,
+ uts.release);
+ errno = E2BIG;
+ return -E2BIG;
+ }
+ pr_debug("try to open %s\n", fname);
+ fd = open(fname, O_RDONLY);
+ if (fd >= 0)
+ break;
+ }
+ return fd;
+}
+#endif
+
+static const char * const probe_usage[] = {
+ "perf probe [<options>] 'PROBEDEF' ['PROBEDEF' ...]",
+ "perf probe [<options>] --add 'PROBEDEF' [--add 'PROBEDEF' ...]",
+ NULL
+};
+
+static const struct option options[] = {
+ OPT_BOOLEAN('v', "verbose", &verbose,
+ "be more verbose (show parsed arguments, etc)"),
+#ifndef NO_LIBDWARF
+ OPT_STRING('k', "vmlinux", &session.vmlinux, "file",
+ "vmlinux/module pathname"),
+#endif
+ OPT_CALLBACK('a', "add", NULL,
+#ifdef NO_LIBDWARF
+ "FUNC[+OFFS|%return] [ARG ...]",
+#else
+ "FUNC[+OFFS|%return|:RLN][@SRC]|SRC:ALN [ARG ...]",
+#endif
+ "probe point definition, where\n"
+ "\t\tGRP:\tGroup name (optional)\n"
+ "\t\tNAME:\tEvent name\n"
+ "\t\tFUNC:\tFunction name\n"
+ "\t\tOFFS:\tOffset from function entry (in byte)\n"
+ "\t\t%return:\tPut the probe at function return\n"
+#ifdef NO_LIBDWARF
+ "\t\tARG:\tProbe argument (only \n"
+#else
+ "\t\tSRC:\tSource code path\n"
+ "\t\tRLN:\tRelative line number from function entry.\n"
+ "\t\tALN:\tAbsolute line number in file.\n"
+ "\t\tARG:\tProbe argument (local variable name or\n"
+#endif
+ "\t\t\tkprobe-tracer argument format is supported.)\n",
+ opt_add_probe_event),
+ OPT_END()
+};
+
+static int write_new_event(int fd, const char *buf)
+{
+ int ret;
+
+ ret = write(fd, buf, strlen(buf));
+ if (ret <= 0)
+ die("Failed to create event.");
+ else
+ printf("Added new event: %s\n", buf);
+
+ return ret;
+}
+
+#define MAX_CMDLEN 256
+
+static int synthesize_probe_event(struct probe_point *pp)
+{
+ char *buf;
+ int i, len, ret;
+ pp->probes[0] = buf = (char *)calloc(MAX_CMDLEN, sizeof(char));
+ if (!buf)
+ die("Failed to allocate memory by calloc.");
+ ret = snprintf(buf, MAX_CMDLEN, "%s+%d", pp->function, pp->offset);
+ if (ret <= 0 || ret >= MAX_CMDLEN)
+ goto error;
+ len = ret;
+
+ for (i = 0; i < pp->nr_args; i++) {
+ ret = snprintf(&buf[len], MAX_CMDLEN - len, " %s",
+ pp->args[i]);
+ if (ret <= 0 || ret >= MAX_CMDLEN - len)
+ goto error;
+ len += ret;
+ }
+ pp->found = 1;
+ return pp->found;
+error:
+ free(pp->probes[0]);
+ if (ret > 0)
+ ret = -E2BIG;
+ return ret;
+}
+
+int cmd_probe(int argc, const char **argv, const char *prefix __used)
+{
+ int i, j, fd, ret;
+ struct probe_point *pp;
+ char buf[MAX_CMDLEN];
+
+ argc = parse_options(argc, argv, options, probe_usage,
+ PARSE_OPT_STOP_AT_NON_OPTION);
+ for (i = 0; i < argc; i++)
+ parse_probe_event(argv[i]);
+
+ if (session.nr_probe == 0)
+ usage_with_options(probe_usage, options);
+
+ if (session.need_dwarf)
+#ifdef NO_LIBDWARF
+ semantic_error("Debuginfo-analysis is not supported");
+#else /* !NO_LIBDWARF */
+ pr_info("Some probes require debuginfo.\n");
+
+ if (session.vmlinux)
+ fd = open(session.vmlinux, O_RDONLY);
+ else
+ fd = open_default_vmlinux();
+ if (fd < 0) {
+ if (session.need_dwarf)
+ die("Could not open vmlinux/module file.");
+
+ pr_warning("Could not open vmlinux/module file."
+ " Try to use symbols.\n");
+ goto end_dwarf;
+ }
+
+ /* Searching probe points */
+ for (j = 0; j < session.nr_probe; j++) {
+ pp = &session.probes[j];
+ if (pp->found)
+ continue;
+
+ lseek(fd, SEEK_SET, 0);
+ ret = find_probepoint(fd, pp);
+ if (ret < 0) {
+ if (session.need_dwarf)
+ die("Could not analyze debuginfo.");
+
+ pr_warning("An error occurred in debuginfo analysis. Try to use symbols.\n");
+ break;
+ }
+ if (ret == 0) /* No error but failed to find probe point. */
+ die("No probe point found.");
+ }
+ close(fd);
+
+end_dwarf:
+#endif /* !NO_LIBDWARF */
+
+ /* Synthesize probes without dwarf */
+ for (j = 0; j < session.nr_probe; j++) {
+ pp = &session.probes[j];
+ if (pp->found) /* This probe is already found. */
+ continue;
+
+ ret = synthesize_probe_event(pp);
+ if (ret == -E2BIG)
+ semantic_error("probe point is too long.");
+ else if (ret < 0)
+ die("Failed to synthesize a probe point.");
+ }
+
+ /* Settng up probe points */
+ snprintf(buf, MAX_CMDLEN, "%s/../kprobe_events", debugfs_path);
+ fd = open(buf, O_WRONLY, O_APPEND);
+ if (fd < 0) {
+ if (errno == ENOENT)
+ die("kprobe_events file does not exist - please rebuild with CONFIG_KPROBE_TRACER.");
+ else
+ die("Could not open kprobe_events file: %s",
+ strerror(errno));
+ }
+ for (j = 0; j < session.nr_probe; j++) {
+ pp = &session.probes[j];
+ if (pp->found == 1) {
+ snprintf(buf, MAX_CMDLEN, "%c:%s/%s_%x %s\n",
+ pp->retprobe ? 'r' : 'p', PERFPROBE_GROUP,
+ pp->function, pp->offset, pp->probes[0]);
+ write_new_event(fd, buf);
+ } else
+ for (i = 0; i < pp->found; i++) {
+ snprintf(buf, MAX_CMDLEN, "%c:%s/%s_%x_%d %s\n",
+ pp->retprobe ? 'r' : 'p',
+ PERFPROBE_GROUP,
+ pp->function, pp->offset, i,
+ pp->probes[0]);
+ write_new_event(fd, buf);
+ }
+ }
+ close(fd);
+ return 0;
+}
+
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 8b2c860c49a2..87f98fdb0513 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -17,13 +17,11 @@
#include "util/header.h"
#include "util/event.h"
#include "util/debug.h"
+#include "util/symbol.h"
#include <unistd.h>
#include <sched.h>
-#define ALIGN(x, a) __ALIGN_MASK(x, (typeof(x))(a)-1)
-#define __ALIGN_MASK(x, mask) (((x)+(mask))&~(mask))
-
static int fd[MAX_NR_CPUS][MAX_COUNTERS];
static long default_interval = 0;
@@ -112,6 +110,24 @@ static void write_output(void *buf, size_t size)
}
}
+static void write_event(event_t *buf, size_t size)
+{
+ /*
+ * Add it to the list of DSOs, so that when we finish this
+ * record session we can pick the available build-ids.
+ */
+ if (buf->header.type == PERF_RECORD_MMAP)
+ dsos__findnew(buf->mmap.filename);
+
+ write_output(buf, size);
+}
+
+static int process_synthesized_event(event_t *event)
+{
+ write_event(event, event->header.size);
+ return 0;
+}
+
static void mmap_read(struct mmap_data *md)
{
unsigned int head = mmap_read_head(md);
@@ -160,14 +176,14 @@ static void mmap_read(struct mmap_data *md)
size = md->mask + 1 - (old & md->mask);
old += size;
- write_output(buf, size);
+ write_event(buf, size);
}
buf = &data[old & md->mask];
size = head - old;
old += size;
- write_output(buf, size);
+ write_event(buf, size);
md->prev = old;
mmap_write_tail(md, old);
@@ -194,168 +210,6 @@ static void sig_atexit(void)
kill(getpid(), signr);
}
-static pid_t pid_synthesize_comm_event(pid_t pid, int full)
-{
- struct comm_event comm_ev;
- char filename[PATH_MAX];
- char bf[BUFSIZ];
- FILE *fp;
- size_t size = 0;
- DIR *tasks;
- struct dirent dirent, *next;
- pid_t tgid = 0;
-
- snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
-
- 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 0;
- }
-
- memset(&comm_ev, 0, sizeof(comm_ev));
- 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_RECORD_COMM;
- size = ALIGN(size, sizeof(u64));
- comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size);
-
- if (!full) {
- comm_ev.tid = pid;
-
- write_output(&comm_ev, comm_ev.header.size);
- goto out_fclose;
- }
-
- snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
-
- tasks = opendir(filename);
- while (!readdir_r(tasks, &dirent, &next) && next) {
- char *end;
- pid = strtol(dirent.d_name, &end, 10);
- if (*end)
- continue;
-
- comm_ev.tid = pid;
-
- write_output(&comm_ev, comm_ev.header.size);
- }
- closedir(tasks);
-
-out_fclose:
- fclose(fp);
- return tgid;
-
-out_failure:
- fprintf(stderr, "couldn't get COMM and pgid, malformed %s\n",
- filename);
- exit(EXIT_FAILURE);
-}
-
-static void pid_synthesize_mmap_samples(pid_t pid, pid_t tgid)
-{
- char filename[PATH_MAX];
- FILE *fp;
-
- snprintf(filename, sizeof(filename), "/proc/%d/maps", pid);
-
- 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;
- }
- while (1) {
- char bf[BUFSIZ], *pbf = bf;
- struct mmap_event mmap_ev = {
- .header = { .type = PERF_RECORD_MMAP },
- };
- int n;
- size_t size;
- if (fgets(bf, sizeof(bf), fp) == NULL)
- break;
-
- /* 00400000-0040c000 r-xp 00000000 fd:01 41038 /bin/cat */
- n = hex2u64(pbf, &mmap_ev.start);
- if (n < 0)
- continue;
- pbf += n + 1;
- n = hex2u64(pbf, &mmap_ev.len);
- if (n < 0)
- continue;
- pbf += n + 3;
- if (*pbf == 'x') { /* vm_exec */
- char *execname = strchr(bf, '/');
-
- /* Catch VDSO */
- if (execname == NULL)
- execname = strstr(bf, "[vdso]");
-
- if (execname == NULL)
- continue;
-
- size = strlen(execname);
- execname[size - 1] = '\0'; /* Remove \n */
- memcpy(mmap_ev.filename, execname, size);
- size = ALIGN(size, sizeof(u64));
- mmap_ev.len -= mmap_ev.start;
- mmap_ev.header.size = (sizeof(mmap_ev) -
- (sizeof(mmap_ev.filename) - size));
- mmap_ev.pid = tgid;
- mmap_ev.tid = pid;
-
- write_output(&mmap_ev, mmap_ev.header.size);
- }
- }
-
- fclose(fp);
-}
-
-static void synthesize_all(void)
-{
- DIR *proc;
- struct dirent dirent, *next;
-
- proc = opendir("/proc");
-
- while (!readdir_r(proc, &dirent, &next) && next) {
- char *end;
- pid_t pid, tgid;
-
- pid = strtol(dirent.d_name, &end, 10);
- if (*end) /* only interested in proper numerical dirents */
- continue;
-
- tgid = pid_synthesize_comm_event(pid, 1);
- pid_synthesize_mmap_samples(pid, tgid);
- }
-
- closedir(proc);
-}
-
static int group_fd;
static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int nr)
@@ -366,7 +220,11 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n
h_attr = header->attr[nr];
} else {
h_attr = perf_header_attr__new(a);
- perf_header__add_attr(header, h_attr);
+ if (h_attr != NULL)
+ if (perf_header__add_attr(header, h_attr) < 0) {
+ perf_header_attr__delete(h_attr);
+ h_attr = NULL;
+ }
}
return h_attr;
@@ -427,7 +285,7 @@ try_again:
if (fd[nr_cpu][counter] < 0) {
int err = errno;
- if (err == EPERM)
+ if (err == EPERM || err == EACCES)
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");
@@ -454,6 +312,8 @@ try_again:
}
h_attr = get_header_attr(attr, counter);
+ if (h_attr == NULL)
+ die("nomem\n");
if (!file_new) {
if (memcmp(&h_attr->attr, attr, sizeof(*attr))) {
@@ -467,7 +327,10 @@ try_again:
exit(-1);
}
- perf_header_attr__add_id(h_attr, read_data.id);
+ if (perf_header_attr__add_id(h_attr, read_data.id) < 0) {
+ pr_warning("Not enough memory to add id\n");
+ exit(-1);
+ }
assert(fd[nr_cpu][counter] >= 0);
fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK);
@@ -528,7 +391,7 @@ static void atexit_header(void)
{
header->data_size += bytes_written;
- perf_header__write(header, output);
+ perf_header__write(header, output, true);
}
static int __cmd_record(int argc, const char **argv)
@@ -537,7 +400,7 @@ static int __cmd_record(int argc, const char **argv)
struct stat st;
pid_t pid = 0;
int flags;
- int ret;
+ int err;
unsigned long waking = 0;
page_size = sysconf(_SC_PAGE_SIZE);
@@ -571,17 +434,24 @@ static int __cmd_record(int argc, const char **argv)
exit(-1);
}
- if (!file_new)
- header = perf_header__read(output);
- else
- header = perf_header__new();
+ header = perf_header__new();
+ if (header == NULL) {
+ pr_err("Not enough memory for reading perf file header\n");
+ return -1;
+ }
+
+ if (!file_new) {
+ err = perf_header__read(header, output);
+ if (err < 0)
+ return err;
+ }
if (raw_samples) {
- perf_header__set_trace_info();
+ perf_header__set_feat(header, HEADER_TRACE_INFO);
} else {
for (i = 0; i < nr_counters; i++) {
if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
- perf_header__set_trace_info();
+ perf_header__set_feat(header, HEADER_TRACE_INFO);
break;
}
}
@@ -604,25 +474,36 @@ static int __cmd_record(int argc, const char **argv)
}
}
- if (file_new)
- perf_header__write(header, output);
+ if (file_new) {
+ err = perf_header__write(header, output, false);
+ if (err < 0)
+ return err;
+ }
- if (!system_wide) {
- pid_t tgid = pid_synthesize_comm_event(pid, 0);
- pid_synthesize_mmap_samples(pid, tgid);
- } else
- synthesize_all();
+ if (!system_wide)
+ event__synthesize_thread(pid, process_synthesized_event);
+ else
+ event__synthesize_threads(process_synthesized_event);
if (target_pid == -1 && argc) {
pid = fork();
if (pid < 0)
- perror("failed to fork");
+ die("failed to fork");
if (!pid) {
if (execvp(argv[0], (char **)argv)) {
perror(argv[0]);
exit(-1);
}
+ } else {
+ /*
+ * Wait a bit for the execv'ed child to appear
+ * and be updated in /proc
+ * FIXME: Do you know a less heuristical solution?
+ */
+ usleep(1000);
+ event__synthesize_thread(pid,
+ process_synthesized_event);
}
child_pid = pid;
@@ -633,7 +514,7 @@ static int __cmd_record(int argc, const char **argv)
param.sched_priority = realtime_prio;
if (sched_setscheduler(0, SCHED_FIFO, &param)) {
- printf("Could not set realtime priority.\n");
+ pr_err("Could not set realtime priority.\n");
exit(-1);
}
}
@@ -651,7 +532,7 @@ static int __cmd_record(int argc, const char **argv)
if (hits == samples) {
if (done)
break;
- ret = poll(event_array, nr_poll, -1);
+ err = poll(event_array, nr_poll, -1);
waking++;
}
@@ -732,6 +613,8 @@ int cmd_record(int argc, const char **argv, const char *prefix __used)
{
int counter;
+ symbol__init(0);
+
argc = parse_options(argc, argv, options, record_usage,
PARSE_OPT_STOP_AT_NON_OPTION);
if (!argc && target_pid == -1 && !system_wide)
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index a4f8cc209151..1a806d5f05cf 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -59,12 +59,28 @@ static struct perf_header *header;
static u64 sample_type;
-static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask)
+
+static size_t
+callchain__fprintf_left_margin(FILE *fp, int left_margin)
+{
+ int i;
+ int ret;
+
+ ret = fprintf(fp, " ");
+
+ for (i = 0; i < left_margin; i++)
+ ret += fprintf(fp, " ");
+
+ return ret;
+}
+
+static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask,
+ int left_margin)
{
int i;
size_t ret = 0;
- ret += fprintf(fp, "%s", " ");
+ ret += callchain__fprintf_left_margin(fp, left_margin);
for (i = 0; i < depth; i++)
if (depth_mask & (1 << i))
@@ -79,12 +95,12 @@ static size_t ipchain__fprintf_graph_line(FILE *fp, int depth, int depth_mask)
static size_t
ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, int depth,
int depth_mask, int count, u64 total_samples,
- int hits)
+ int hits, int left_margin)
{
int i;
size_t ret = 0;
- ret += fprintf(fp, "%s", " ");
+ ret += callchain__fprintf_left_margin(fp, left_margin);
for (i = 0; i < depth; i++) {
if (depth_mask & (1 << i))
ret += fprintf(fp, "|");
@@ -122,8 +138,9 @@ static void init_rem_hits(void)
}
static size_t
-callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
- u64 total_samples, int depth, int depth_mask)
+__callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
+ u64 total_samples, int depth, int depth_mask,
+ int left_margin)
{
struct rb_node *node, *next;
struct callchain_node *child;
@@ -164,7 +181,8 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
* But we keep the older depth mask for the line seperator
* to keep the level link until we reach the last child
*/
- ret += ipchain__fprintf_graph_line(fp, depth, depth_mask);
+ ret += ipchain__fprintf_graph_line(fp, depth, depth_mask,
+ left_margin);
i = 0;
list_for_each_entry(chain, &child->val, list) {
if (chain->ip >= PERF_CONTEXT_MAX)
@@ -172,11 +190,13 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
ret += ipchain__fprintf_graph(fp, chain, depth,
new_depth_mask, i++,
new_total,
- cumul);
+ cumul,
+ left_margin);
}
- ret += callchain__fprintf_graph(fp, child, new_total,
- depth + 1,
- new_depth_mask | (1 << depth));
+ ret += __callchain__fprintf_graph(fp, child, new_total,
+ depth + 1,
+ new_depth_mask | (1 << depth),
+ left_margin);
node = next;
}
@@ -190,9 +210,48 @@ callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
ret += ipchain__fprintf_graph(fp, &rem_hits, depth,
new_depth_mask, 0, new_total,
- remaining);
+ remaining, left_margin);
+ }
+
+ return ret;
+}
+
+
+static size_t
+callchain__fprintf_graph(FILE *fp, struct callchain_node *self,
+ u64 total_samples, int left_margin)
+{
+ struct callchain_list *chain;
+ bool printed = false;
+ int i = 0;
+ int ret = 0;
+
+ list_for_each_entry(chain, &self->val, list) {
+ if (chain->ip >= PERF_CONTEXT_MAX)
+ continue;
+
+ if (!i++ && sort__first_dimension == SORT_SYM)
+ continue;
+
+ if (!printed) {
+ ret += callchain__fprintf_left_margin(fp, left_margin);
+ ret += fprintf(fp, "|\n");
+ ret += callchain__fprintf_left_margin(fp, left_margin);
+ ret += fprintf(fp, "---");
+
+ left_margin += 3;
+ printed = true;
+ } else
+ ret += callchain__fprintf_left_margin(fp, left_margin);
+
+ if (chain->sym)
+ ret += fprintf(fp, " %s\n", chain->sym->name);
+ else
+ ret += fprintf(fp, " %p\n", (void *)(long)chain->ip);
}
+ ret += __callchain__fprintf_graph(fp, self, total_samples, 1, 1, left_margin);
+
return ret;
}
@@ -224,7 +283,7 @@ callchain__fprintf_flat(FILE *fp, struct callchain_node *self,
static size_t
hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
- u64 total_samples)
+ u64 total_samples, int left_margin)
{
struct rb_node *rb_node;
struct callchain_node *chain;
@@ -244,8 +303,8 @@ hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
break;
case CHAIN_GRAPH_ABS: /* Falldown */
case CHAIN_GRAPH_REL:
- ret += callchain__fprintf_graph(fp, chain,
- total_samples, 1, 1);
+ ret += callchain__fprintf_graph(fp, chain, total_samples,
+ left_margin);
case CHAIN_NONE:
default:
break;
@@ -290,8 +349,19 @@ hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
ret += fprintf(fp, "\n");
- if (callchain)
- hist_entry_callchain__fprintf(fp, self, total_samples);
+ if (callchain) {
+ int left_margin = 0;
+
+ if (sort__first_dimension == SORT_COMM) {
+ se = list_first_entry(&hist_entry__sort_list, typeof(*se),
+ list);
+ left_margin = se->width ? *se->width : 0;
+ left_margin -= thread__comm_len(self->thread);
+ }
+
+ hist_entry_callchain__fprintf(fp, self, total_samples,
+ left_margin);
+ }
return ret;
}
@@ -385,7 +455,7 @@ got_map:
dump_printf(" ...... map: %Lx -> %Lx\n", *ipp, ip);
*ipp = ip;
- return map ? map->dso->find_symbol(map->dso, ip) : NULL;
+ return map ? map__find_symbol(map, ip, NULL) : NULL;
}
static int call__match(struct symbol *sym)
@@ -619,7 +689,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
dump_printf("... chain: nr:%Lu\n", chain->nr);
if (validate_chain(chain, event) < 0) {
- eprintf("call-chain problem with event, skipping it.\n");
+ pr_debug("call-chain problem with event, "
+ "skipping it.\n");
return 0;
}
@@ -630,7 +701,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
}
if (thread == NULL) {
- eprintf("problem processing %d event, skipping it.\n",
+ pr_debug("problem processing %d event, skipping it.\n",
event->header.type);
return -1;
}
@@ -668,7 +739,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
if (hist_entry__add(thread, map, sym, ip,
chain, level, period)) {
- eprintf("problem incrementing symbol count, skipping event\n");
+ pr_debug("problem incrementing symbol count, skipping event\n");
return -1;
}
@@ -1022,7 +1093,7 @@ static void setup_list(struct strlist **list, const char *list_str,
int cmd_report(int argc, const char **argv, const char *prefix __used)
{
- symbol__init();
+ symbol__init(0);
argc = parse_options(argc, argv, options, report_usage, 0);
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 57ad3f458ef5..df44b756cecc 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -38,8 +38,6 @@ static int cwdlen;
#define PR_SET_NAME 15 /* Set process name */
#define MAX_CPUS 4096
-#define BUG_ON(x) assert(!(x))
-
static u64 run_measurement_overhead;
static u64 sleep_measurement_overhead;
@@ -1668,8 +1666,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
(long long)period);
if (thread == NULL) {
- eprintf("problem processing %d event, skipping it.\n",
- event->header.type);
+ pr_debug("problem processing %d event, skipping it.\n",
+ event->header.type);
return -1;
}
@@ -1939,7 +1937,7 @@ static int __cmd_record(int argc, const char **argv)
int cmd_sched(int argc, const char **argv, const char *prefix __used)
{
- symbol__init();
+ symbol__init(0);
argc = parse_options(argc, argv, sched_options, sched_usage,
PARSE_OPT_STOP_AT_NON_OPTION);
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 3db31e7bf173..c70d72003557 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -50,15 +50,17 @@
static struct perf_event_attr default_attrs[] = {
- { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK },
- { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES},
- { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS },
- { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS },
-
- { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES },
- { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS },
- { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES},
- { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS },
+ { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS },
+
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES },
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS },
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES },
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES },
+ { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES },
};
@@ -125,6 +127,7 @@ struct stats event_res_stats[MAX_COUNTERS][3];
struct stats runtime_nsecs_stats;
struct stats walltime_nsecs_stats;
struct stats runtime_cycles_stats;
+struct stats runtime_branches_stats;
#define MATCH_EVENT(t, c, counter) \
(attrs[counter].type == PERF_TYPE_##t && \
@@ -235,6 +238,8 @@ static void read_counter(int counter)
update_stats(&runtime_nsecs_stats, count[0]);
if (MATCH_EVENT(HARDWARE, HW_CPU_CYCLES, counter))
update_stats(&runtime_cycles_stats, count[0]);
+ if (MATCH_EVENT(HARDWARE, HW_BRANCH_INSTRUCTIONS, counter))
+ update_stats(&runtime_branches_stats, count[0]);
}
static int run_perf_stat(int argc __used, const char **argv)
@@ -352,7 +357,16 @@ static void abs_printout(int counter, double avg)
ratio = avg / total;
fprintf(stderr, " # %10.3f IPC ", ratio);
- } else {
+ } else if (MATCH_EVENT(HARDWARE, HW_BRANCH_MISSES, counter) &&
+ runtime_branches_stats.n != 0) {
+ total = avg_stats(&runtime_branches_stats);
+
+ if (total)
+ ratio = avg * 100 / total;
+
+ fprintf(stderr, " # %10.3f %% ", ratio);
+
+ } else if (runtime_nsecs_stats.n != 0) {
total = avg_stats(&runtime_nsecs_stats);
if (total)
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 702d8fe58fbc..dd4d82ac7aa4 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -153,6 +153,17 @@ static struct wake_event *wake_events;
struct sample_wrapper *all_samples;
+
+struct process_filter;
+struct process_filter {
+ char *name;
+ int pid;
+ struct process_filter *next;
+};
+
+static struct process_filter *process_filter;
+
+
static struct per_pid *find_create_pid(int pid)
{
struct per_pid *cursor = all_data;
@@ -763,21 +774,42 @@ static void draw_wakeups(void)
c = p->all;
while (c) {
if (c->Y && c->start_time <= we->time && c->end_time >= we->time) {
- if (p->pid == we->waker) {
+ if (p->pid == we->waker && !from) {
from = c->Y;
- task_from = c->comm;
+ task_from = strdup(c->comm);
}
- if (p->pid == we->wakee) {
+ if (p->pid == we->wakee && !to) {
to = c->Y;
- task_to = c->comm;
+ task_to = strdup(c->comm);
}
}
c = c->next;
}
+ c = p->all;
+ while (c) {
+ if (p->pid == we->waker && !from) {
+ from = c->Y;
+ task_from = strdup(c->comm);
+ }
+ if (p->pid == we->wakee && !to) {
+ to = c->Y;
+ task_to = strdup(c->comm);
+ }
+ c = c->next;
+ }
}
p = p->next;
}
+ if (!task_from) {
+ task_from = malloc(40);
+ sprintf(task_from, "[%i]", we->waker);
+ }
+ if (!task_to) {
+ task_to = malloc(40);
+ sprintf(task_to, "[%i]", we->wakee);
+ }
+
if (we->waker == -1)
svg_interrupt(we->time, to);
else if (from && to && abs(from - to) == 1)
@@ -785,6 +817,9 @@ static void draw_wakeups(void)
else
svg_partial_wakeline(we->time, from, task_from, to, task_to);
we = we->next;
+
+ free(task_from);
+ free(task_to);
}
}
@@ -858,12 +893,89 @@ static void draw_process_bars(void)
}
}
+static void add_process_filter(const char *string)
+{
+ struct process_filter *filt;
+ int pid;
+
+ pid = strtoull(string, NULL, 10);
+ filt = malloc(sizeof(struct process_filter));
+ if (!filt)
+ return;
+
+ filt->name = strdup(string);
+ filt->pid = pid;
+ filt->next = process_filter;
+
+ process_filter = filt;
+}
+
+static int passes_filter(struct per_pid *p, struct per_pidcomm *c)
+{
+ struct process_filter *filt;
+ if (!process_filter)
+ return 1;
+
+ filt = process_filter;
+ while (filt) {
+ if (filt->pid && p->pid == filt->pid)
+ return 1;
+ if (strcmp(filt->name, c->comm) == 0)
+ return 1;
+ filt = filt->next;
+ }
+ return 0;
+}
+
+static int determine_display_tasks_filtered(void)
+{
+ struct per_pid *p;
+ struct per_pidcomm *c;
+ int count = 0;
+
+ p = all_data;
+ while (p) {
+ p->display = 0;
+ if (p->start_time == 1)
+ p->start_time = first_time;
+
+ /* no exit marker, task kept running to the end */
+ if (p->end_time == 0)
+ p->end_time = last_time;
+
+ c = p->all;
+
+ while (c) {
+ c->display = 0;
+
+ if (c->start_time == 1)
+ c->start_time = first_time;
+
+ if (passes_filter(p, c)) {
+ c->display = 1;
+ p->display = 1;
+ count++;
+ }
+
+ if (c->end_time == 0)
+ c->end_time = last_time;
+
+ c = c->next;
+ }
+ p = p->next;
+ }
+ return count;
+}
+
static int determine_display_tasks(u64 threshold)
{
struct per_pid *p;
struct per_pidcomm *c;
int count = 0;
+ if (process_filter)
+ return determine_display_tasks_filtered();
+
p = all_data;
while (p) {
p->display = 0;
@@ -981,7 +1093,7 @@ static void process_samples(void)
static int __cmd_timechart(void)
{
- int ret, rc = EXIT_FAILURE;
+ int err, rc = EXIT_FAILURE;
unsigned long offset = 0;
unsigned long head, shift;
struct stat statbuf;
@@ -999,8 +1111,8 @@ static int __cmd_timechart(void)
exit(-1);
}
- ret = fstat(input, &statbuf);
- if (ret < 0) {
+ err = fstat(input, &statbuf);
+ if (err < 0) {
perror("failed to stat file");
exit(-1);
}
@@ -1010,7 +1122,16 @@ static int __cmd_timechart(void)
exit(0);
}
- header = perf_header__read(input);
+ header = perf_header__new();
+ if (header == NULL)
+ return -ENOMEM;
+
+ err = perf_header__read(header, input);
+ if (err < 0) {
+ perf_header__delete(header);
+ return err;
+ }
+
head = header->data_offset;
sample_type = perf_header__sample_type(header);
@@ -1050,12 +1171,10 @@ more:
size = event->header.size;
if (!size || process_event(event) < 0) {
-
- printf("%p [%p]: skipping unknown header type: %d\n",
- (void *)(offset + head),
- (void *)(long)(event->header.size),
- event->header.type);
-
+ pr_warning("%p [%p]: skipping unknown header type: %d\n",
+ (void *)(offset + head),
+ (void *)(long)(event->header.size),
+ event->header.type);
/*
* assume we lost track of the stream, check alignment, and
* increment a single u64 in the hope to catch on again 'soon'.
@@ -1088,7 +1207,8 @@ done:
write_svg_file(output_name);
- printf("Written %2.1f seconds of trace to %s.\n", (last_time - first_time) / 1000000000.0, output_name);
+ pr_info("Written %2.1f seconds of trace to %s.\n",
+ (last_time - first_time) / 1000000000.0, output_name);
return rc;
}
@@ -1129,6 +1249,14 @@ static int __cmd_record(int argc, const char **argv)
return cmd_record(i, rec_argv, NULL);
}
+static int
+parse_process(const struct option *opt __used, const char *arg, int __used unset)
+{
+ if (arg)
+ add_process_filter(arg);
+ return 0;
+}
+
static const struct option options[] = {
OPT_STRING('i', "input", &input_name, "file",
"input file name"),
@@ -1136,15 +1264,18 @@ static const struct option options[] = {
"output file name"),
OPT_INTEGER('w', "width", &svg_page_width,
"page width"),
- OPT_BOOLEAN('p', "power-only", &power_only,
+ OPT_BOOLEAN('P', "power-only", &power_only,
"output power data only"),
+ OPT_CALLBACK('p', "process", NULL, "process",
+ "process selector. Pass a pid or process name.",
+ parse_process),
OPT_END()
};
int cmd_timechart(int argc, const char **argv, const char *prefix __used)
{
- symbol__init();
+ symbol__init(0);
page_size = getpagesize();
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index 2d8806bac258..6d770ac7be0b 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -60,7 +60,7 @@ static int system_wide = 0;
static int default_interval = 0;
static int count_filter = 5;
-static int print_entries = 15;
+static int print_entries;
static int target_pid = -1;
static int inherit = 0;
@@ -76,6 +76,17 @@ static int delay_secs = 2;
static int zero = 0;
static int dump_symtab = 0;
+static bool hide_kernel_symbols = false;
+static bool hide_user_symbols = false;
+static struct winsize winsize;
+static const char *graph_line =
+ "_____________________________________________________________________"
+ "_____________________________________________________________________";
+static const char *graph_dotted_line =
+ "---------------------------------------------------------------------"
+ "---------------------------------------------------------------------"
+ "---------------------------------------------------------------------";
+
/*
* Source
*/
@@ -97,27 +108,75 @@ static int display_weighted = -1;
* Symbols
*/
+struct sym_entry_source {
+ struct source_line *source;
+ struct source_line *lines;
+ struct source_line **lines_tail;
+ pthread_mutex_t lock;
+};
+
struct sym_entry {
struct rb_node rb_node;
struct list_head node;
- unsigned long count[MAX_COUNTERS];
unsigned long snap_count;
double weight;
int skip;
+ u16 name_len;
+ u8 origin;
struct map *map;
- struct source_line *source;
- struct source_line *lines;
- struct source_line **lines_tail;
- pthread_mutex_t source_lock;
+ struct sym_entry_source *src;
+ unsigned long count[0];
};
/*
* Source functions
*/
+static inline struct symbol *sym_entry__symbol(struct sym_entry *self)
+{
+ return ((void *)self) + symbol__priv_size;
+}
+
+static void get_term_dimensions(struct winsize *ws)
+{
+ char *s = getenv("LINES");
+
+ if (s != NULL) {
+ ws->ws_row = atoi(s);
+ s = getenv("COLUMNS");
+ if (s != NULL) {
+ ws->ws_col = atoi(s);
+ if (ws->ws_row && ws->ws_col)
+ return;
+ }
+ }
+#ifdef TIOCGWINSZ
+ if (ioctl(1, TIOCGWINSZ, ws) == 0 &&
+ ws->ws_row && ws->ws_col)
+ return;
+#endif
+ ws->ws_row = 25;
+ ws->ws_col = 80;
+}
+
+static void update_print_entries(struct winsize *ws)
+{
+ print_entries = ws->ws_row;
+
+ if (print_entries > 9)
+ print_entries -= 9;
+}
+
+static void sig_winch_handler(int sig __used)
+{
+ get_term_dimensions(&winsize);
+ update_print_entries(&winsize);
+}
+
static void parse_source(struct sym_entry *syme)
{
struct symbol *sym;
+ struct sym_entry_source *source;
struct map *map;
FILE *file;
char command[PATH_MAX*2];
@@ -127,12 +186,21 @@ static void parse_source(struct sym_entry *syme)
if (!syme)
return;
- if (syme->lines) {
- pthread_mutex_lock(&syme->source_lock);
+ if (syme->src == NULL) {
+ syme->src = calloc(1, sizeof(*source));
+ if (syme->src == NULL)
+ return;
+ pthread_mutex_init(&syme->src->lock, NULL);
+ }
+
+ source = syme->src;
+
+ if (source->lines) {
+ pthread_mutex_lock(&source->lock);
goto out_assign;
}
- sym = (struct symbol *)(syme + 1);
+ sym = sym_entry__symbol(syme);
map = syme->map;
path = map->dso->long_name;
@@ -141,14 +209,15 @@ static void parse_source(struct sym_entry *syme)
sprintf(command,
"objdump --start-address=0x%016Lx "
"--stop-address=0x%016Lx -dS %s",
- sym->start, sym->end, path);
+ map->unmap_ip(map, sym->start),
+ map->unmap_ip(map, sym->end), path);
file = popen(command, "r");
if (!file)
return;
- pthread_mutex_lock(&syme->source_lock);
- syme->lines_tail = &syme->lines;
+ pthread_mutex_lock(&source->lock);
+ source->lines_tail = &source->lines;
while (!feof(file)) {
struct source_line *src;
size_t dummy = 0;
@@ -168,22 +237,22 @@ static void parse_source(struct sym_entry *syme)
*c = 0;
src->next = NULL;
- *syme->lines_tail = src;
- syme->lines_tail = &src->next;
+ *source->lines_tail = src;
+ source->lines_tail = &src->next;
if (strlen(src->line)>8 && src->line[8] == ':') {
src->eip = strtoull(src->line, NULL, 16);
- src->eip += map->start;
+ src->eip = map->unmap_ip(map, src->eip);
}
if (strlen(src->line)>8 && src->line[16] == ':') {
src->eip = strtoull(src->line, NULL, 16);
- src->eip += map->start;
+ src->eip = map->unmap_ip(map, src->eip);
}
}
pclose(file);
out_assign:
sym_filter_entry = syme;
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&source->lock);
}
static void __zero_source_counters(struct sym_entry *syme)
@@ -191,7 +260,7 @@ static void __zero_source_counters(struct sym_entry *syme)
int i;
struct source_line *line;
- line = syme->lines;
+ line = syme->src->lines;
while (line) {
for (i = 0; i < nr_counters; i++)
line->count[i] = 0;
@@ -206,13 +275,13 @@ static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip)
if (syme != sym_filter_entry)
return;
- if (pthread_mutex_trylock(&syme->source_lock))
+ if (pthread_mutex_trylock(&syme->src->lock))
return;
- if (!syme->source)
+ if (syme->src == NULL || syme->src->source == NULL)
goto out_unlock;
- for (line = syme->lines; line; line = line->next) {
+ for (line = syme->src->lines; line; line = line->next) {
if (line->eip == ip) {
line->count[counter]++;
break;
@@ -221,25 +290,25 @@ static void record_precise_ip(struct sym_entry *syme, int counter, u64 ip)
break;
}
out_unlock:
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&syme->src->lock);
}
static void lookup_sym_source(struct sym_entry *syme)
{
- struct symbol *symbol = (struct symbol *)(syme + 1);
+ struct symbol *symbol = sym_entry__symbol(syme);
struct source_line *line;
char pattern[PATH_MAX];
sprintf(pattern, "<%s>:", symbol->name);
- pthread_mutex_lock(&syme->source_lock);
- for (line = syme->lines; line; line = line->next) {
+ pthread_mutex_lock(&syme->src->lock);
+ for (line = syme->src->lines; line; line = line->next) {
if (strstr(line->line, pattern)) {
- syme->source = line;
+ syme->src->source = line;
break;
}
}
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&syme->src->lock);
}
static void show_lines(struct source_line *queue, int count, int total)
@@ -269,24 +338,24 @@ static void show_details(struct sym_entry *syme)
if (!syme)
return;
- if (!syme->source)
+ if (!syme->src->source)
lookup_sym_source(syme);
- if (!syme->source)
+ if (!syme->src->source)
return;
- symbol = (struct symbol *)(syme + 1);
+ symbol = sym_entry__symbol(syme);
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;
+ pthread_mutex_lock(&syme->src->lock);
+ line = syme->src->source;
while (line) {
total += line->count[sym_counter];
line = line->next;
}
- line = syme->source;
+ line = syme->src->source;
while (line) {
float pcnt = 0.0;
@@ -311,13 +380,13 @@ static void show_details(struct sym_entry *syme)
line->count[sym_counter] = zero ? 0 : line->count[sym_counter] * 7 / 8;
line = line->next;
}
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&syme->src->lock);
if (more)
printf("%d lines not displayed, maybe increase display entries [e]\n", more);
}
/*
- * Symbols will be added here in record_ip and will get out
+ * Symbols will be added here in event__process_sample and will get out
* after decayed.
*/
static LIST_HEAD(active_symbols);
@@ -388,6 +457,9 @@ static void print_sym_table(void)
struct sym_entry *syme, *n;
struct rb_root tmp = RB_ROOT;
struct rb_node *nd;
+ int sym_width = 0, dso_width = 0;
+ const int win_width = winsize.ws_col - 1;
+ struct dso *unique_dso = NULL, *first_dso = NULL;
samples = userspace_samples = 0;
@@ -399,6 +471,14 @@ static void print_sym_table(void)
list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
syme->snap_count = syme->count[snap];
if (syme->snap_count != 0) {
+
+ if ((hide_user_symbols &&
+ syme->origin == PERF_RECORD_MISC_USER) ||
+ (hide_kernel_symbols &&
+ syme->origin == PERF_RECORD_MISC_KERNEL)) {
+ list_remove_active_sym(syme);
+ continue;
+ }
syme->weight = sym_weight(syme);
rb_insert_active_sym(&tmp, syme);
sum_ksamples += syme->snap_count;
@@ -411,8 +491,7 @@ static void print_sym_table(void)
puts(CONSOLE_CLEAR);
- printf(
-"------------------------------------------------------------------------------\n");
+ printf("%-*.*s\n", win_width, win_width, graph_dotted_line);
printf( " PerfTop:%8.0f irqs/sec kernel:%4.1f%% [",
samples_per_sec,
100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec)));
@@ -450,33 +529,70 @@ static void print_sym_table(void)
printf(", %d CPUs)\n", nr_cpus);
}
- printf("------------------------------------------------------------------------------\n\n");
+ printf("%-*.*s\n", win_width, win_width, graph_dotted_line);
if (sym_filter_entry) {
show_details(sym_filter_entry);
return;
}
+ /*
+ * Find the longest symbol name that will be displayed
+ */
+ for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) {
+ syme = rb_entry(nd, struct sym_entry, rb_node);
+ if (++printed > print_entries ||
+ (int)syme->snap_count < count_filter)
+ continue;
+
+ if (first_dso == NULL)
+ unique_dso = first_dso = syme->map->dso;
+ else if (syme->map->dso != first_dso)
+ unique_dso = NULL;
+
+ if (syme->map->dso->long_name_len > dso_width)
+ dso_width = syme->map->dso->long_name_len;
+
+ if (syme->name_len > sym_width)
+ sym_width = syme->name_len;
+ }
+
+ printed = 0;
+
+ if (unique_dso)
+ printf("DSO: %s\n", unique_dso->long_name);
+ else {
+ int max_dso_width = winsize.ws_col - sym_width - 29;
+ if (dso_width > max_dso_width)
+ dso_width = max_dso_width;
+ putchar('\n');
+ }
if (nr_counters == 1)
- printf(" samples pcnt");
+ printf(" samples pcnt");
else
- printf(" weight samples pcnt");
+ printf(" weight samples pcnt");
if (verbose)
printf(" RIP ");
- printf(" kernel function\n");
- printf(" %s _______ _____",
+ printf(" %-*.*s", sym_width, sym_width, "function");
+ if (!unique_dso)
+ printf(" DSO");
+ putchar('\n');
+ printf(" %s _______ _____",
nr_counters == 1 ? " " : "______");
if (verbose)
- printf(" ________________");
- printf(" _______________\n\n");
+ printf(" ________________");
+ printf(" %-*.*s", sym_width, sym_width, graph_line);
+ if (!unique_dso)
+ printf(" %-*.*s", dso_width, dso_width, graph_line);
+ puts("\n");
for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) {
struct symbol *sym;
double pcnt;
syme = rb_entry(nd, struct sym_entry, rb_node);
- sym = (struct symbol *)(syme + 1);
+ sym = sym_entry__symbol(syme);
if (++printed > print_entries || (int)syme->snap_count < count_filter)
continue;
@@ -485,16 +601,19 @@ static void print_sym_table(void)
sum_ksamples));
if (nr_counters == 1 || !display_weighted)
- printf("%20.2f - ", syme->weight);
+ printf("%20.2f ", syme->weight);
else
- printf("%9.1f %10ld - ", syme->weight, syme->snap_count);
+ printf("%9.1f %10ld ", syme->weight, syme->snap_count);
percent_color_fprintf(stdout, "%4.1f%%", pcnt);
if (verbose)
- printf(" - %016llx", sym->start);
- printf(" : %s", sym->name);
- if (syme->map->dso->name[0] == '[')
- printf(" \t%s", syme->map->dso->name);
+ printf(" %016llx", sym->start);
+ printf(" %-*.*s", sym_width, sym_width, sym->name);
+ if (!unique_dso)
+ printf(" %-*.*s", dso_width, dso_width,
+ dso_width >= syme->map->dso->long_name_len ?
+ syme->map->dso->long_name :
+ syme->map->dso->short_name);
printf("\n");
}
}
@@ -542,10 +661,10 @@ static void prompt_symbol(struct sym_entry **target, const char *msg)
/* zero counters of active symbol */
if (syme) {
- pthread_mutex_lock(&syme->source_lock);
+ pthread_mutex_lock(&syme->src->lock);
__zero_source_counters(syme);
*target = NULL;
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&syme->src->lock);
}
fprintf(stdout, "\n%s: ", msg);
@@ -561,7 +680,7 @@ static void prompt_symbol(struct sym_entry **target, const char *msg)
pthread_mutex_unlock(&active_symbols_lock);
list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
- struct symbol *sym = (struct symbol *)(syme + 1);
+ struct symbol *sym = sym_entry__symbol(syme);
if (!strcmp(buf, sym->name)) {
found = syme;
@@ -585,7 +704,7 @@ static void print_mapped_keys(void)
char *name = NULL;
if (sym_filter_entry) {
- struct symbol *sym = (struct symbol *)(sym_filter_entry+1);
+ struct symbol *sym = sym_entry__symbol(sym_filter_entry);
name = sym->name;
}
@@ -607,6 +726,12 @@ static void print_mapped_keys(void)
if (nr_counters > 1)
fprintf(stdout, "\t[w] toggle display weighted/count[E]r. \t(%d)\n", display_weighted ? 1 : 0);
+ fprintf(stdout,
+ "\t[K] hide kernel_symbols symbols. \t(%s)\n",
+ hide_kernel_symbols ? "yes" : "no");
+ fprintf(stdout,
+ "\t[U] hide user symbols. \t(%s)\n",
+ hide_user_symbols ? "yes" : "no");
fprintf(stdout, "\t[z] toggle sample zeroing. \t(%d)\n", zero ? 1 : 0);
fprintf(stdout, "\t[qQ] quit.\n");
}
@@ -620,6 +745,8 @@ static int key_mapped(int c)
case 'z':
case 'q':
case 'Q':
+ case 'K':
+ case 'U':
return 1;
case 'E':
case 'w':
@@ -663,9 +790,16 @@ static void handle_keypress(int c)
switch (c) {
case 'd':
prompt_integer(&delay_secs, "Enter display delay");
+ if (delay_secs < 1)
+ delay_secs = 1;
break;
case 'e':
prompt_integer(&print_entries, "Enter display entries (lines)");
+ if (print_entries == 0) {
+ sig_winch_handler(SIGWINCH);
+ signal(SIGWINCH, sig_winch_handler);
+ } else
+ signal(SIGWINCH, SIG_DFL);
break;
case 'E':
if (nr_counters > 1) {
@@ -690,6 +824,9 @@ static void handle_keypress(int c)
case 'F':
prompt_percent(&sym_pcnt_filter, "Enter details display event filter (percent)");
break;
+ case 'K':
+ hide_kernel_symbols = !hide_kernel_symbols;
+ break;
case 'q':
case 'Q':
printf("exiting.\n");
@@ -703,12 +840,15 @@ static void handle_keypress(int c)
else {
struct sym_entry *syme = sym_filter_entry;
- pthread_mutex_lock(&syme->source_lock);
+ pthread_mutex_lock(&syme->src->lock);
sym_filter_entry = NULL;
__zero_source_counters(syme);
- pthread_mutex_unlock(&syme->source_lock);
+ pthread_mutex_unlock(&syme->src->lock);
}
break;
+ case 'U':
+ hide_user_symbols = !hide_user_symbols;
+ break;
case 'w':
display_weighted = ~display_weighted;
break;
@@ -787,9 +927,9 @@ static int symbol_filter(struct map *map, struct symbol *sym)
strstr(name, "_text_end"))
return 1;
- syme = dso__sym_priv(map->dso, sym);
+ syme = symbol__priv(sym);
syme->map = map;
- pthread_mutex_init(&syme->source_lock, NULL);
+ syme->src = NULL;
if (!sym_filter_entry && sym_filter && !strcmp(name, sym_filter))
sym_filter_entry = syme;
@@ -800,13 +940,20 @@ static int symbol_filter(struct map *map, struct symbol *sym)
}
}
+ if (!syme->skip)
+ syme->name_len = strlen(sym->name);
+
return 0;
}
static int parse_symbols(void)
{
- if (dsos__load_kernel(vmlinux_name, sizeof(struct sym_entry),
- symbol_filter, verbose, 1) <= 0)
+ struct dso *kernel = dsos__load_kernel();
+
+ if (kernel == NULL)
+ return -1;
+
+ if (dso__load_kernel_sym(kernel, symbol_filter, 1) <= 0)
return -1;
if (dump_symtab)
@@ -815,41 +962,104 @@ static int parse_symbols(void)
return 0;
}
-/*
- * Binary search in the histogram table and record the hit:
- */
-static void record_ip(u64 ip, int counter)
+static void event__process_sample(const event_t *self, int counter)
{
+ u64 ip = self->ip.ip;
struct map *map;
- struct symbol *sym = kernel_maps__find_symbol(ip, &map);
-
- if (sym != NULL) {
- struct sym_entry *syme = dso__sym_priv(map->dso, sym);
-
- 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);
- pthread_mutex_unlock(&active_symbols_lock);
+ struct sym_entry *syme;
+ struct symbol *sym;
+ u8 origin = self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
+
+ switch (origin) {
+ case PERF_RECORD_MISC_USER: {
+ struct thread *thread;
+
+ if (hide_user_symbols)
return;
+
+ thread = threads__findnew(self->ip.pid);
+ if (thread == NULL)
+ return;
+
+ map = thread__find_map(thread, ip);
+ if (map != NULL) {
+ ip = map->map_ip(map, ip);
+ sym = map__find_symbol(map, ip, symbol_filter);
+ if (sym == NULL)
+ return;
+ userspace_samples++;
+ break;
}
}
+ /*
+ * If this is outside of all known maps,
+ * and is a negative address, try to look it
+ * up in the kernel dso, as it might be a
+ * vsyscall or vdso (which executes in user-mode).
+ */
+ if ((long long)ip >= 0)
+ return;
+ /* Fall thru */
+ case PERF_RECORD_MISC_KERNEL:
+ if (hide_kernel_symbols)
+ return;
+
+ sym = kernel_maps__find_symbol(ip, &map);
+ if (sym == NULL)
+ return;
+ break;
+ default:
+ return;
+ }
- samples--;
+ syme = symbol__priv(sym);
+
+ if (!syme->skip) {
+ syme->count[counter]++;
+ syme->origin = origin;
+ 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);
+ pthread_mutex_unlock(&active_symbols_lock);
+ ++samples;
+ return;
+ }
}
-static void process_event(u64 ip, int counter, int user)
+static void event__process_mmap(event_t *self)
{
- samples++;
+ struct thread *thread = threads__findnew(self->mmap.pid);
- if (user) {
- userspace_samples++;
- return;
+ if (thread != NULL) {
+ struct map *map = map__new(&self->mmap, NULL, 0);
+ if (map != NULL)
+ thread__insert_map(thread, map);
+ }
+}
+
+static void event__process_comm(event_t *self)
+{
+ struct thread *thread = threads__findnew(self->comm.pid);
+
+ if (thread != NULL)
+ thread__set_comm(thread, self->comm.comm);
+}
+
+static int event__process(event_t *event)
+{
+ switch (event->header.type) {
+ case PERF_RECORD_COMM:
+ event__process_comm(event);
+ break;
+ case PERF_RECORD_MMAP:
+ event__process_mmap(event);
+ break;
+ default:
+ break;
}
- record_ip(ip, counter);
+ return 0;
}
struct mmap_data {
@@ -922,13 +1132,11 @@ static void mmap_read_counter(struct mmap_data *md)
event = &event_copy;
}
+ if (event->header.type == PERF_RECORD_SAMPLE)
+ event__process_sample(event, md->counter);
+ else
+ event__process(event);
old += size;
-
- if (event->header.type == PERF_RECORD_SAMPLE) {
- int user =
- (event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK) == PERF_RECORD_MISC_USER;
- process_event(event->ip.ip, md->counter, user);
- }
}
md->prev = old;
@@ -970,6 +1178,7 @@ static void start_counter(int i, int counter)
}
attr->inherit = (cpu < 0) && inherit;
+ attr->mmap = 1;
try_again:
fd[i][counter] = sys_perf_event_open(attr, target_pid, cpu, group_fd, 0);
@@ -977,7 +1186,7 @@ try_again:
if (fd[i][counter] < 0) {
int err = errno;
- if (err == EPERM)
+ if (err == EPERM || err == EACCES)
die("No permission - are you root?\n");
/*
* If it's cycles then fall back to hrtimer
@@ -1028,6 +1237,11 @@ static int __cmd_top(void)
int i, counter;
int ret;
+ if (target_pid != -1)
+ event__synthesize_thread(target_pid, event__process);
+ else
+ event__synthesize_threads(event__process);
+
for (i = 0; i < nr_cpus; i++) {
group_fd = -1;
for (counter = 0; counter < nr_counters; counter++)
@@ -1084,6 +1298,8 @@ static const struct option options[] = {
OPT_INTEGER('C', "CPU", &profile_cpu,
"CPU to profile on"),
OPT_STRING('k', "vmlinux", &vmlinux_name, "file", "vmlinux pathname"),
+ OPT_BOOLEAN('K', "hide_kernel_symbols", &hide_kernel_symbols,
+ "hide kernel symbols"),
OPT_INTEGER('m', "mmap-pages", &mmap_pages,
"number of mmap data pages"),
OPT_INTEGER('r', "realtime", &realtime_prio,
@@ -1106,6 +1322,8 @@ static const struct option options[] = {
"profile at this frequency"),
OPT_INTEGER('E', "entries", &print_entries,
"display this many functions"),
+ OPT_BOOLEAN('U', "hide_user_symbols", &hide_user_symbols,
+ "hide user symbols"),
OPT_BOOLEAN('v', "verbose", &verbose,
"be more verbose (show counter open errors, etc)"),
OPT_END()
@@ -1115,8 +1333,6 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
{
int counter;
- symbol__init();
-
page_size = sysconf(_SC_PAGE_SIZE);
argc = parse_options(argc, argv, options, top_usage, 0);
@@ -1133,6 +1349,9 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
if (!nr_counters)
nr_counters = 1;
+ symbol__init(sizeof(struct sym_entry) +
+ (nr_counters + 1) * sizeof(unsigned long));
+
if (delay_secs < 1)
delay_secs = 1;
@@ -1169,5 +1388,11 @@ int cmd_top(int argc, const char **argv, const char *prefix __used)
if (target_pid != -1 || profile_cpu != -1)
nr_cpus = 1;
+ get_term_dimensions(&winsize);
+ if (print_entries == 0) {
+ update_print_entries(&winsize);
+ signal(SIGWINCH, sig_winch_handler);
+ }
+
return __cmd_top();
}
diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 4c129ff0bb16..d042d656c561 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -81,8 +81,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
(long long)period);
if (thread == NULL) {
- eprintf("problem processing %d event, skipping it.\n",
- event->header.type);
+ pr_debug("problem processing %d event, skipping it.\n",
+ event->header.type);
return -1;
}
@@ -151,7 +151,7 @@ static const struct option options[] = {
int cmd_trace(int argc, const char **argv, const char *prefix __used)
{
- symbol__init();
+ symbol__init(0);
argc = parse_options(argc, argv, options, annotate_usage, 0);
if (argc) {
diff --git a/tools/perf/builtin.h b/tools/perf/builtin.h
index e11d8d231c3b..a3d8bf65f26c 100644
--- a/tools/perf/builtin.h
+++ b/tools/perf/builtin.h
@@ -15,6 +15,8 @@ extern int read_line_with_nul(char *buf, int size, FILE *file);
extern int check_pager_config(const char *cmd);
extern int cmd_annotate(int argc, const char **argv, const char *prefix);
+extern int cmd_bench(int argc, const char **argv, const char *prefix);
+extern int cmd_buildid_list(int argc, const char **argv, const char *prefix);
extern int cmd_help(int argc, const char **argv, const char *prefix);
extern int cmd_sched(int argc, const char **argv, const char *prefix);
extern int cmd_list(int argc, const char **argv, const char *prefix);
@@ -25,5 +27,7 @@ extern int cmd_timechart(int argc, const char **argv, const char *prefix);
extern int cmd_top(int argc, const char **argv, const char *prefix);
extern int cmd_trace(int argc, const char **argv, const char *prefix);
extern int cmd_version(int argc, const char **argv, const char *prefix);
+extern int cmd_probe(int argc, const char **argv, const char *prefix);
+extern int cmd_kmem(int argc, const char **argv, const char *prefix);
#endif
diff --git a/tools/perf/command-list.txt b/tools/perf/command-list.txt
index 00326e230d87..d3a6e18e4a5e 100644
--- a/tools/perf/command-list.txt
+++ b/tools/perf/command-list.txt
@@ -3,6 +3,8 @@
# command name category [deprecated] [common]
#
perf-annotate mainporcelain common
+perf-bench mainporcelain common
+perf-buildid-list mainporcelain common
perf-list mainporcelain common
perf-sched mainporcelain common
perf-record mainporcelain common
@@ -11,3 +13,4 @@ perf-stat mainporcelain common
perf-timechart mainporcelain common
perf-top mainporcelain common
perf-trace mainporcelain common
+perf-probe mainporcelain common
diff --git a/tools/perf/design.txt b/tools/perf/design.txt
index fdd42a824c98..f000c30877ac 100644
--- a/tools/perf/design.txt
+++ b/tools/perf/design.txt
@@ -137,6 +137,8 @@ enum sw_event_ids {
PERF_COUNT_SW_CPU_MIGRATIONS = 4,
PERF_COUNT_SW_PAGE_FAULTS_MIN = 5,
PERF_COUNT_SW_PAGE_FAULTS_MAJ = 6,
+ PERF_COUNT_SW_ALIGNMENT_FAULTS = 7,
+ PERF_COUNT_SW_EMULATION_FAULTS = 8,
};
Counters of the type PERF_TYPE_TRACEPOINT are available when the ftrace event
diff --git a/tools/perf/perf.c b/tools/perf/perf.c
index 624e62d9d1e0..cf64049bc9bd 100644
--- a/tools/perf/perf.c
+++ b/tools/perf/perf.c
@@ -14,6 +14,7 @@
#include "util/run-command.h"
#include "util/parse-events.h"
#include "util/string.h"
+#include "util/debugfs.h"
const char perf_usage_string[] =
"perf [--version] [--help] COMMAND [ARGS]";
@@ -284,17 +285,21 @@ static void handle_internal_command(int argc, const char **argv)
{
const char *cmd = argv[0];
static struct cmd_struct commands[] = {
- { "help", cmd_help, 0 },
- { "list", cmd_list, 0 },
- { "record", cmd_record, 0 },
- { "report", cmd_report, 0 },
- { "stat", cmd_stat, 0 },
- { "timechart", cmd_timechart, 0 },
- { "top", cmd_top, 0 },
- { "annotate", cmd_annotate, 0 },
- { "version", cmd_version, 0 },
- { "trace", cmd_trace, 0 },
- { "sched", cmd_sched, 0 },
+ { "buildid-list", cmd_buildid_list, 0 },
+ { "help", cmd_help, 0 },
+ { "list", cmd_list, 0 },
+ { "record", cmd_record, 0 },
+ { "report", cmd_report, 0 },
+ { "bench", cmd_bench, 0 },
+ { "stat", cmd_stat, 0 },
+ { "timechart", cmd_timechart, 0 },
+ { "top", cmd_top, 0 },
+ { "annotate", cmd_annotate, 0 },
+ { "version", cmd_version, 0 },
+ { "trace", cmd_trace, 0 },
+ { "sched", cmd_sched, 0 },
+ { "probe", cmd_probe, 0 },
+ { "kmem", cmd_kmem, 0 },
};
unsigned int i;
static const char ext[] = STRIP_EXTENSION;
@@ -382,45 +387,12 @@ static int run_argv(int *argcp, const char ***argv)
/* mini /proc/mounts parser: searching for "^blah /mount/point debugfs" */
static void get_debugfs_mntpt(void)
{
- FILE *file;
- char fs_type[100];
- char debugfs[MAXPATHLEN];
+ const char *path = debugfs_find_mountpoint();
- /*
- * try the standard location
- */
- if (valid_debugfs_mount("/sys/kernel/debug/") == 0) {
- strcpy(debugfs_mntpt, "/sys/kernel/debug/");
- return;
- }
-
- /*
- * try the sane location
- */
- if (valid_debugfs_mount("/debug/") == 0) {
- strcpy(debugfs_mntpt, "/debug/");
- return;
- }
-
- /*
- * give up and parse /proc/mounts
- */
- file = fopen("/proc/mounts", "r");
- if (file == NULL)
- return;
-
- while (fscanf(file, "%*s %"
- STR(MAXPATHLEN)
- "s %99s %*s %*d %*d\n",
- debugfs, fs_type) == 2) {
- if (strcmp(fs_type, "debugfs") == 0)
- break;
- }
- fclose(file);
- if (strcmp(fs_type, "debugfs") == 0) {
- strncpy(debugfs_mntpt, debugfs, MAXPATHLEN);
- debugfs_mntpt[MAXPATHLEN - 1] = '\0';
- }
+ if (path)
+ strncpy(debugfs_mntpt, path, sizeof(debugfs_mntpt));
+ else
+ debugfs_mntpt[0] = '\0';
}
int main(int argc, const char **argv)
diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 8cc4623afd6f..454d5d55f32d 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -47,6 +47,18 @@
#define cpu_relax() asm volatile("":::"memory")
#endif
+#ifdef __alpha__
+#include "../../arch/alpha/include/asm/unistd.h"
+#define rmb() asm volatile("mb" ::: "memory")
+#define cpu_relax() asm volatile("" ::: "memory")
+#endif
+
+#ifdef __ia64__
+#include "../../arch/ia64/include/asm/unistd.h"
+#define rmb() asm volatile ("mf" ::: "memory")
+#define cpu_relax() asm volatile ("hint @pause" ::: "memory")
+#endif
+
#include <time.h>
#include <unistd.h>
#include <sys/types.h>
diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN
index c561d1538c03..54552a00a117 100755
--- a/tools/perf/util/PERF-VERSION-GEN
+++ b/tools/perf/util/PERF-VERSION-GEN
@@ -1,7 +1,7 @@
#!/bin/sh
GVF=PERF-VERSION-FILE
-DEF_VER=v0.0.1.PERF
+DEF_VER=v0.0.2.PERF
LF='
'
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index 3b8380f1b478..b3b71258272a 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -206,7 +206,7 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain,
}
node->val_nr = chain->nr - start;
if (!node->val_nr)
- printf("Warning: empty node in callchain tree\n");
+ pr_warning("Warning: empty node in callchain tree\n");
}
static void
diff --git a/tools/perf/util/data_map.c b/tools/perf/util/data_map.c
index 242b0555ab91..5543e7d0487d 100644
--- a/tools/perf/util/data_map.c
+++ b/tools/perf/util/data_map.c
@@ -70,6 +70,35 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
}
}
+int perf_header__read_build_ids(int input, off_t offset, off_t size)
+{
+ struct build_id_event bev;
+ char filename[PATH_MAX];
+ off_t limit = offset + size;
+ int err = -1;
+
+ while (offset < limit) {
+ struct dso *dso;
+ ssize_t len;
+
+ if (read(input, &bev, sizeof(bev)) != sizeof(bev))
+ goto out;
+
+ len = bev.header.size - sizeof(bev);
+ if (read(input, filename, len) != len)
+ goto out;
+
+ dso = dsos__findnew(filename);
+ if (dso != NULL)
+ dso__set_build_id(dso, &bev.build_id);
+
+ offset += bev.header.size;
+ }
+ err = 0;
+out:
+ return err;
+}
+
int mmap_dispatch_perf_file(struct perf_header **pheader,
const char *input_name,
int force,
@@ -77,7 +106,7 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
int *cwdlen,
char **cwd)
{
- int ret, rc = EXIT_FAILURE;
+ int err;
struct perf_header *header;
unsigned long head, shift;
unsigned long offset = 0;
@@ -89,56 +118,69 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
int input;
char *buf;
- if (!curr_handler)
- die("Forgot to register perf file handler");
+ if (curr_handler == NULL) {
+ pr_debug("Forgot to register perf file handler\n");
+ return -EINVAL;
+ }
page_size = getpagesize();
input = open(input_name, O_RDONLY);
if (input < 0) {
- fprintf(stderr, " failed to open file: %s", input_name);
+ pr_err("Failed to open file: %s", input_name);
if (!strcmp(input_name, "perf.data"))
- fprintf(stderr, " (try 'perf record' first)");
- fprintf(stderr, "\n");
- exit(-1);
+ pr_err(" (try 'perf record' first)");
+ pr_err("\n");
+ return -errno;
}
- ret = fstat(input, &input_stat);
- if (ret < 0) {
- perror("failed to stat file");
- exit(-1);
+ if (fstat(input, &input_stat) < 0) {
+ pr_err("failed to stat file");
+ err = -errno;
+ goto out_close;
}
+ err = -EACCES;
if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
- fprintf(stderr, "file: %s not owned by current user or root\n",
+ pr_err("file: %s not owned by current user or root\n",
input_name);
- exit(-1);
+ goto out_close;
}
- if (!input_stat.st_size) {
- fprintf(stderr, "zero-sized file, nothing to do!\n");
- exit(0);
+ if (input_stat.st_size == 0) {
+ pr_info("zero-sized file, nothing to do!\n");
+ goto done;
}
- *pheader = perf_header__read(input);
- header = *pheader;
+ err = -ENOMEM;
+ header = perf_header__new();
+ if (header == NULL)
+ goto out_close;
+
+ err = perf_header__read(header, input);
+ if (err < 0)
+ goto out_delete;
+ *pheader = header;
head = header->data_offset;
sample_type = perf_header__sample_type(header);
- if (curr_handler->sample_type_check)
- if (curr_handler->sample_type_check(sample_type) < 0)
- exit(-1);
+ err = -EINVAL;
+ if (curr_handler->sample_type_check &&
+ curr_handler->sample_type_check(sample_type) < 0)
+ goto out_delete;
- if (load_kernel() < 0) {
- perror("failed to load kernel symbols");
- return EXIT_FAILURE;
+ err = -ENOMEM;
+ if (load_kernel(NULL) < 0) {
+ pr_err("failed to load kernel symbols\n");
+ goto out_delete;
}
if (!full_paths) {
if (getcwd(__cwd, sizeof(__cwd)) == NULL) {
- perror("failed to get the current directory");
- return EXIT_FAILURE;
+ pr_err("failed to get the current directory\n");
+ err = -errno;
+ goto out_delete;
}
*cwd = __cwd;
*cwdlen = strlen(*cwd);
@@ -152,11 +194,12 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
head -= shift;
remap:
- buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
- MAP_SHARED, input, offset);
+ buf = mmap(NULL, page_size * mmap_window, PROT_READ,
+ MAP_SHARED, input, offset);
if (buf == MAP_FAILED) {
- perror("failed to mmap file");
- exit(-1);
+ pr_err("failed to mmap file\n");
+ err = -errno;
+ goto out_delete;
}
more:
@@ -213,10 +256,12 @@ more:
goto more;
done:
- rc = EXIT_SUCCESS;
+ err = 0;
+out_close:
close(input);
- return rc;
+ return err;
+out_delete:
+ perf_header__delete(header);
+ goto out_close;
}
-
-
diff --git a/tools/perf/util/data_map.h b/tools/perf/util/data_map.h
index 716d1053b074..ae036ecd7625 100644
--- a/tools/perf/util/data_map.h
+++ b/tools/perf/util/data_map.h
@@ -27,5 +27,6 @@ int mmap_dispatch_perf_file(struct perf_header **pheader,
int full_paths,
int *cwdlen,
char **cwd);
+int perf_header__read_build_ids(int input, off_t offset, off_t file_size);
#endif
diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c
index e8ca98fe0bd4..28d520d5a1fb 100644
--- a/tools/perf/util/debug.c
+++ b/tools/perf/util/debug.c
@@ -13,12 +13,12 @@
int verbose = 0;
int dump_trace = 0;
-int eprintf(const char *fmt, ...)
+int eprintf(int level, const char *fmt, ...)
{
va_list args;
int ret = 0;
- if (verbose) {
+ if (verbose >= level) {
va_start(args, fmt);
ret = vfprintf(stderr, fmt, args);
va_end(args);
diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h
index 02d1fa1c2465..c6c24c522dea 100644
--- a/tools/perf/util/debug.h
+++ b/tools/perf/util/debug.h
@@ -2,10 +2,13 @@
#ifndef __PERF_DEBUG_H
#define __PERF_DEBUG_H
+#include "event.h"
+
extern int verbose;
extern int dump_trace;
-int eprintf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+int eprintf(int level,
+ const char *fmt, ...) __attribute__((format(printf, 2, 3)));
int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
void trace_event(event_t *event);
diff --git a/tools/perf/util/debugfs.c b/tools/perf/util/debugfs.c
new file mode 100644
index 000000000000..06b73ee02c49
--- /dev/null
+++ b/tools/perf/util/debugfs.c
@@ -0,0 +1,241 @@
+#include "util.h"
+#include "debugfs.h"
+#include "cache.h"
+
+static int debugfs_premounted;
+static char debugfs_mountpoint[MAX_PATH+1];
+
+static const char *debugfs_known_mountpoints[] = {
+ "/sys/kernel/debug/",
+ "/debug/",
+ 0,
+};
+
+/* use this to force a umount */
+void debugfs_force_cleanup(void)
+{
+ debugfs_find_mountpoint();
+ debugfs_premounted = 0;
+ debugfs_umount();
+}
+
+/* construct a full path to a debugfs element */
+int debugfs_make_path(const char *element, char *buffer, int size)
+{
+ int len;
+
+ if (strlen(debugfs_mountpoint) == 0) {
+ buffer[0] = '\0';
+ return -1;
+ }
+
+ len = strlen(debugfs_mountpoint) + strlen(element) + 1;
+ if (len >= size)
+ return len+1;
+
+ snprintf(buffer, size-1, "%s/%s", debugfs_mountpoint, element);
+ return 0;
+}
+
+static int debugfs_found;
+
+/* find the path to the mounted debugfs */
+const char *debugfs_find_mountpoint(void)
+{
+ const char **ptr;
+ char type[100];
+ FILE *fp;
+
+ if (debugfs_found)
+ return (const char *) debugfs_mountpoint;
+
+ ptr = debugfs_known_mountpoints;
+ while (*ptr) {
+ if (debugfs_valid_mountpoint(*ptr) == 0) {
+ debugfs_found = 1;
+ strcpy(debugfs_mountpoint, *ptr);
+ return debugfs_mountpoint;
+ }
+ ptr++;
+ }
+
+ /* give up and parse /proc/mounts */
+ fp = fopen("/proc/mounts", "r");
+ if (fp == NULL)
+ die("Can't open /proc/mounts for read");
+
+ while (fscanf(fp, "%*s %"
+ STR(MAX_PATH)
+ "s %99s %*s %*d %*d\n",
+ debugfs_mountpoint, type) == 2) {
+ if (strcmp(type, "debugfs") == 0)
+ break;
+ }
+ fclose(fp);
+
+ if (strcmp(type, "debugfs") != 0)
+ return NULL;
+
+ debugfs_found = 1;
+
+ return debugfs_mountpoint;
+}
+
+/* verify that a mountpoint is actually a debugfs instance */
+
+int debugfs_valid_mountpoint(const char *debugfs)
+{
+ struct statfs st_fs;
+
+ if (statfs(debugfs, &st_fs) < 0)
+ return -ENOENT;
+ else if (st_fs.f_type != (long) DEBUGFS_MAGIC)
+ return -ENOENT;
+
+ return 0;
+}
+
+
+int debugfs_valid_entry(const char *path)
+{
+ struct stat st;
+
+ if (stat(path, &st))
+ return -errno;
+
+ return 0;
+}
+
+/* mount the debugfs somewhere */
+
+int debugfs_mount(const char *mountpoint)
+{
+ char mountcmd[128];
+
+ /* see if it's already mounted */
+ if (debugfs_find_mountpoint()) {
+ debugfs_premounted = 1;
+ return 0;
+ }
+
+ /* if not mounted and no argument */
+ if (mountpoint == NULL) {
+ /* see if environment variable set */
+ mountpoint = getenv(PERF_DEBUGFS_ENVIRONMENT);
+ /* if no environment variable, use default */
+ if (mountpoint == NULL)
+ mountpoint = "/sys/kernel/debug";
+ }
+
+ /* save the mountpoint */
+ strncpy(debugfs_mountpoint, mountpoint, sizeof(debugfs_mountpoint));
+
+ /* mount it */
+ snprintf(mountcmd, sizeof(mountcmd),
+ "/bin/mount -t debugfs debugfs %s", mountpoint);
+ return system(mountcmd);
+}
+
+/* umount the debugfs */
+
+int debugfs_umount(void)
+{
+ char umountcmd[128];
+ int ret;
+
+ /* if it was already mounted, leave it */
+ if (debugfs_premounted)
+ return 0;
+
+ /* make sure it's a valid mount point */
+ ret = debugfs_valid_mountpoint(debugfs_mountpoint);
+ if (ret)
+ return ret;
+
+ snprintf(umountcmd, sizeof(umountcmd),
+ "/bin/umount %s", debugfs_mountpoint);
+ return system(umountcmd);
+}
+
+int debugfs_write(const char *entry, const char *value)
+{
+ char path[MAX_PATH+1];
+ int ret, count;
+ int fd;
+
+ /* construct the path */
+ snprintf(path, sizeof(path), "%s/%s", debugfs_mountpoint, entry);
+
+ /* verify that it exists */
+ ret = debugfs_valid_entry(path);
+ if (ret)
+ return ret;
+
+ /* get how many chars we're going to write */
+ count = strlen(value);
+
+ /* open the debugfs entry */
+ fd = open(path, O_RDWR);
+ if (fd < 0)
+ return -errno;
+
+ while (count > 0) {
+ /* write it */
+ ret = write(fd, value, count);
+ if (ret <= 0) {
+ if (ret == EAGAIN)
+ continue;
+ close(fd);
+ return -errno;
+ }
+ count -= ret;
+ }
+
+ /* close it */
+ close(fd);
+
+ /* return success */
+ return 0;
+}
+
+/*
+ * read a debugfs entry
+ * returns the number of chars read or a negative errno
+ */
+int debugfs_read(const char *entry, char *buffer, size_t size)
+{
+ char path[MAX_PATH+1];
+ int ret;
+ int fd;
+
+ /* construct the path */
+ snprintf(path, sizeof(path), "%s/%s", debugfs_mountpoint, entry);
+
+ /* verify that it exists */
+ ret = debugfs_valid_entry(path);
+ if (ret)
+ return ret;
+
+ /* open the debugfs entry */
+ fd = open(path, O_RDONLY);
+ if (fd < 0)
+ return -errno;
+
+ do {
+ /* read it */
+ ret = read(fd, buffer, size);
+ if (ret == 0) {
+ close(fd);
+ return EOF;
+ }
+ } while (ret < 0 && errno == EAGAIN);
+
+ /* close it */
+ close(fd);
+
+ /* make *sure* there's a null character at the end */
+ buffer[ret] = '\0';
+
+ /* return the number of chars read */
+ return ret;
+}
diff --git a/tools/perf/util/debugfs.h b/tools/perf/util/debugfs.h
new file mode 100644
index 000000000000..3cd14f9ae784
--- /dev/null
+++ b/tools/perf/util/debugfs.h
@@ -0,0 +1,25 @@
+#ifndef __DEBUGFS_H__
+#define __DEBUGFS_H__
+
+#include <sys/mount.h>
+
+#ifndef MAX_PATH
+# define MAX_PATH 256
+#endif
+
+#ifndef STR
+# define _STR(x) #x
+# define STR(x) _STR(x)
+#endif
+
+extern const char *debugfs_find_mountpoint(void);
+extern int debugfs_valid_mountpoint(const char *debugfs);
+extern int debugfs_valid_entry(const char *path);
+extern int debugfs_mount(const char *mountpoint);
+extern int debugfs_umount(void);
+extern int debugfs_write(const char *entry, const char *value);
+extern int debugfs_read(const char *entry, char *buffer, size_t size);
+extern void debugfs_force_cleanup(void);
+extern int debugfs_make_path(const char *element, char *buffer, int size);
+
+#endif /* __DEBUGFS_H__ */
diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
new file mode 100644
index 000000000000..1dae7e3b400d
--- /dev/null
+++ b/tools/perf/util/event.c
@@ -0,0 +1,177 @@
+#include <linux/types.h>
+#include "event.h"
+#include "debug.h"
+#include "string.h"
+
+static pid_t event__synthesize_comm(pid_t pid, int full,
+ int (*process)(event_t *event))
+{
+ event_t ev;
+ char filename[PATH_MAX];
+ char bf[BUFSIZ];
+ FILE *fp;
+ size_t size = 0;
+ DIR *tasks;
+ struct dirent dirent, *next;
+ pid_t tgid = 0;
+
+ snprintf(filename, sizeof(filename), "/proc/%d/status", pid);
+
+ fp = fopen(filename, "r");
+ if (fp == NULL) {
+out_race:
+ /*
+ * We raced with a task exiting - just return:
+ */
+ pr_debug("couldn't open %s\n", filename);
+ return 0;
+ }
+
+ memset(&ev.comm, 0, sizeof(ev.comm));
+ while (!ev.comm.comm[0] || !ev.comm.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(ev.comm.comm, name, size++);
+ } else if (memcmp(bf, "Tgid:", 5) == 0) {
+ char *tgids = bf + 5;
+ while (*tgids && isspace(*tgids))
+ ++tgids;
+ tgid = ev.comm.pid = atoi(tgids);
+ }
+ }
+
+ ev.comm.header.type = PERF_RECORD_COMM;
+ size = ALIGN(size, sizeof(u64));
+ ev.comm.header.size = sizeof(ev.comm) - (sizeof(ev.comm.comm) - size);
+
+ if (!full) {
+ ev.comm.tid = pid;
+
+ process(&ev);
+ goto out_fclose;
+ }
+
+ snprintf(filename, sizeof(filename), "/proc/%d/task", pid);
+
+ tasks = opendir(filename);
+ if (tasks == NULL)
+ goto out_race;
+
+ while (!readdir_r(tasks, &dirent, &next) && next) {
+ char *end;
+ pid = strtol(dirent.d_name, &end, 10);
+ if (*end)
+ continue;
+
+ ev.comm.tid = pid;
+
+ process(&ev);
+ }
+ closedir(tasks);
+
+out_fclose:
+ fclose(fp);
+ return tgid;
+
+out_failure:
+ pr_warning("couldn't get COMM and pgid, malformed %s\n", filename);
+ return -1;
+}
+
+static int event__synthesize_mmap_events(pid_t pid, pid_t tgid,
+ int (*process)(event_t *event))
+{
+ char filename[PATH_MAX];
+ FILE *fp;
+
+ snprintf(filename, sizeof(filename), "/proc/%d/maps", pid);
+
+ fp = fopen(filename, "r");
+ if (fp == NULL) {
+ /*
+ * We raced with a task exiting - just return:
+ */
+ pr_debug("couldn't open %s\n", filename);
+ return -1;
+ }
+
+ while (1) {
+ char bf[BUFSIZ], *pbf = bf;
+ event_t ev = {
+ .header = { .type = PERF_RECORD_MMAP },
+ };
+ int n;
+ size_t size;
+ if (fgets(bf, sizeof(bf), fp) == NULL)
+ break;
+
+ /* 00400000-0040c000 r-xp 00000000 fd:01 41038 /bin/cat */
+ n = hex2u64(pbf, &ev.mmap.start);
+ if (n < 0)
+ continue;
+ pbf += n + 1;
+ n = hex2u64(pbf, &ev.mmap.len);
+ if (n < 0)
+ continue;
+ pbf += n + 3;
+ if (*pbf == 'x') { /* vm_exec */
+ char *execname = strchr(bf, '/');
+
+ /* Catch VDSO */
+ if (execname == NULL)
+ execname = strstr(bf, "[vdso]");
+
+ if (execname == NULL)
+ continue;
+
+ size = strlen(execname);
+ execname[size - 1] = '\0'; /* Remove \n */
+ memcpy(ev.mmap.filename, execname, size);
+ size = ALIGN(size, sizeof(u64));
+ ev.mmap.len -= ev.mmap.start;
+ ev.mmap.header.size = (sizeof(ev.mmap) -
+ (sizeof(ev.mmap.filename) - size));
+ ev.mmap.pid = tgid;
+ ev.mmap.tid = pid;
+
+ process(&ev);
+ }
+ }
+
+ fclose(fp);
+ return 0;
+}
+
+int event__synthesize_thread(pid_t pid, int (*process)(event_t *event))
+{
+ pid_t tgid = event__synthesize_comm(pid, 1, process);
+ if (tgid == -1)
+ return -1;
+ return event__synthesize_mmap_events(pid, tgid, process);
+}
+
+void event__synthesize_threads(int (*process)(event_t *event))
+{
+ DIR *proc;
+ struct dirent dirent, *next;
+
+ proc = opendir("/proc");
+
+ while (!readdir_r(proc, &dirent, &next) && next) {
+ char *end;
+ pid_t pid = strtol(dirent.d_name, &end, 10);
+
+ if (*end) /* only interested in proper numerical dirents */
+ continue;
+
+ event__synthesize_thread(pid, process);
+ }
+
+ closedir(proc);
+}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index c2e62be62798..34c6fcb82d92 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -61,6 +61,13 @@ struct sample_event{
u64 array[];
};
+#define BUILD_ID_SIZE 20
+
+struct build_id_event {
+ struct perf_event_header header;
+ u8 build_id[ALIGN(BUILD_ID_SIZE, sizeof(u64))];
+ char filename[];
+};
typedef union event_union {
struct perf_event_header header;
@@ -82,6 +89,7 @@ struct map {
u64 end;
u64 pgoff;
u64 (*map_ip)(struct map *, u64);
+ u64 (*unmap_ip)(struct map *, u64);
struct dso *dso;
};
@@ -90,14 +98,29 @@ static inline u64 map__map_ip(struct map *map, u64 ip)
return ip - map->start + map->pgoff;
}
-static inline u64 vdso__map_ip(struct map *map __used, u64 ip)
+static inline u64 map__unmap_ip(struct map *map, u64 ip)
+{
+ return ip + map->start - map->pgoff;
+}
+
+static inline u64 identity__map_ip(struct map *map __used, u64 ip)
{
return ip;
}
+struct symbol;
+
+typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym);
+
+void map__init(struct map *self, u64 start, u64 end, u64 pgoff,
+ struct dso *dso);
struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen);
struct map *map__clone(struct map *self);
int map__overlap(struct map *l, struct map *r);
size_t map__fprintf(struct map *self, FILE *fp);
+struct symbol *map__find_symbol(struct map *self, u64 ip, symbol_filter_t filter);
+
+int event__synthesize_thread(pid_t pid, int (*process)(event_t *event));
+void event__synthesize_threads(int (*process)(event_t *event));
#endif /* __PERF_RECORD_H */
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 9aae360c0f28..d5c81ebc0a84 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -2,11 +2,15 @@
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>
+#include <linux/list.h>
#include "util.h"
#include "header.h"
#include "../perf.h"
#include "trace-event.h"
+#include "symbol.h"
+#include "data_map.h"
+#include "debug.h"
/*
* Create new perf.data header attribute:
@@ -15,32 +19,43 @@ struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr)
{
struct perf_header_attr *self = malloc(sizeof(*self));
- if (!self)
- die("nomem");
-
- self->attr = *attr;
- self->ids = 0;
- self->size = 1;
- self->id = malloc(sizeof(u64));
-
- if (!self->id)
- die("nomem");
+ if (self != NULL) {
+ self->attr = *attr;
+ self->ids = 0;
+ self->size = 1;
+ self->id = malloc(sizeof(u64));
+ if (self->id == NULL) {
+ free(self);
+ self = NULL;
+ }
+ }
return self;
}
-void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
+void perf_header_attr__delete(struct perf_header_attr *self)
+{
+ free(self->id);
+ free(self);
+}
+
+int perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
{
int pos = self->ids;
self->ids++;
if (self->ids > self->size) {
- self->size *= 2;
- self->id = realloc(self->id, self->size * sizeof(u64));
- if (!self->id)
- die("nomem");
+ int nsize = self->size * 2;
+ u64 *nid = realloc(self->id, nsize * sizeof(u64));
+
+ if (nid == NULL)
+ return -1;
+
+ self->size = nsize;
+ self->id = nid;
}
self->id[pos] = id;
+ return 0;
}
/*
@@ -48,44 +63,52 @@ void perf_header_attr__add_id(struct perf_header_attr *self, u64 id)
*/
struct perf_header *perf_header__new(void)
{
- struct perf_header *self = malloc(sizeof(*self));
+ struct perf_header *self = calloc(sizeof(*self), 1);
- if (!self)
- die("nomem");
+ if (self != NULL) {
+ self->size = 1;
+ self->attr = malloc(sizeof(void *));
- self->frozen = 0;
+ if (self->attr == NULL) {
+ free(self);
+ self = NULL;
+ }
+ }
- self->attrs = 0;
- self->size = 1;
- self->attr = malloc(sizeof(void *));
+ return self;
+}
- if (!self->attr)
- die("nomem");
+void perf_header__delete(struct perf_header *self)
+{
+ int i;
- self->data_offset = 0;
- self->data_size = 0;
- self->trace_info_offset = 0;
- self->trace_info_size = 0;
+ for (i = 0; i < self->attrs; ++i)
+ perf_header_attr__delete(self->attr[i]);
- return self;
+ free(self->attr);
+ free(self);
}
-void perf_header__add_attr(struct perf_header *self,
- struct perf_header_attr *attr)
+int perf_header__add_attr(struct perf_header *self,
+ struct perf_header_attr *attr)
{
- int pos = self->attrs;
-
if (self->frozen)
- die("frozen");
+ return -1;
- self->attrs++;
- if (self->attrs > self->size) {
- self->size *= 2;
- self->attr = realloc(self->attr, self->size * sizeof(void *));
- if (!self->attr)
- die("nomem");
+ if (self->attrs == self->size) {
+ int nsize = self->size * 2;
+ struct perf_header_attr **nattr;
+
+ nattr = realloc(self->attr, nsize * sizeof(void *));
+ if (nattr == NULL)
+ return -1;
+
+ self->size = nsize;
+ self->attr = nattr;
}
- self->attr[pos] = attr;
+
+ self->attr[self->attrs++] = attr;
+ return 0;
}
#define MAX_EVENT_NAME 64
@@ -101,7 +124,7 @@ static struct perf_trace_event_type *events;
void perf_header__push_event(u64 id, const char *name)
{
if (strlen(name) > MAX_EVENT_NAME)
- printf("Event %s will be truncated\n", name);
+ pr_warning("Event %s will be truncated\n", name);
if (!events) {
events = malloc(sizeof(struct perf_trace_event_type));
@@ -132,52 +155,135 @@ static const char *__perf_magic = "PERFFILE";
#define PERF_MAGIC (*(u64 *)__perf_magic)
-struct perf_file_section {
- u64 offset;
- u64 size;
-};
-
struct perf_file_attr {
struct perf_event_attr attr;
struct perf_file_section ids;
};
-struct perf_file_header {
- u64 magic;
- u64 size;
- u64 attr_size;
- struct perf_file_section attrs;
- struct perf_file_section data;
- struct perf_file_section event_types;
- struct perf_file_section trace_info;
-};
-
-static int trace_info;
+void perf_header__set_feat(struct perf_header *self, int feat)
+{
+ set_bit(feat, self->adds_features);
+}
-void perf_header__set_trace_info(void)
+bool perf_header__has_feat(const struct perf_header *self, int feat)
{
- trace_info = 1;
+ return test_bit(feat, self->adds_features);
}
-static void do_write(int fd, void *buf, size_t size)
+static int do_write(int fd, const void *buf, size_t size)
{
while (size) {
int ret = write(fd, buf, size);
if (ret < 0)
- die("failed to write");
+ return -errno;
size -= ret;
buf += ret;
}
+
+ return 0;
+}
+
+static int dsos__write_buildid_table(int fd)
+{
+ struct dso *pos;
+
+ list_for_each_entry(pos, &dsos, node) {
+ int err;
+ struct build_id_event b;
+ size_t len;
+
+ if (!pos->has_build_id)
+ continue;
+ len = pos->long_name_len + 1;
+ len = ALIGN(len, 64);
+ memset(&b, 0, sizeof(b));
+ memcpy(&b.build_id, pos->build_id, sizeof(pos->build_id));
+ b.header.size = sizeof(b) + len;
+ err = do_write(fd, &b, sizeof(b));
+ if (err < 0)
+ return err;
+ err = do_write(fd, pos->long_name, len);
+ if (err < 0)
+ return err;
+ }
+
+ return 0;
}
-void perf_header__write(struct perf_header *self, int fd)
+static int perf_header__adds_write(struct perf_header *self, int fd)
+{
+ int nr_sections;
+ struct perf_file_section *feat_sec;
+ int sec_size;
+ u64 sec_start;
+ int idx = 0, err;
+
+ if (dsos__read_build_ids())
+ perf_header__set_feat(self, HEADER_BUILD_ID);
+
+ nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+ if (!nr_sections)
+ return 0;
+
+ feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+ if (feat_sec == NULL)
+ return -ENOMEM;
+
+ sec_size = sizeof(*feat_sec) * nr_sections;
+
+ sec_start = self->data_offset + self->data_size;
+ lseek(fd, sec_start + sec_size, SEEK_SET);
+
+ if (perf_header__has_feat(self, HEADER_TRACE_INFO)) {
+ struct perf_file_section *trace_sec;
+
+ trace_sec = &feat_sec[idx++];
+
+ /* Write trace info */
+ trace_sec->offset = lseek(fd, 0, SEEK_CUR);
+ read_tracing_data(fd, attrs, nr_counters);
+ trace_sec->size = lseek(fd, 0, SEEK_CUR) - trace_sec->offset;
+ }
+
+
+ if (perf_header__has_feat(self, HEADER_BUILD_ID)) {
+ struct perf_file_section *buildid_sec;
+
+ buildid_sec = &feat_sec[idx++];
+
+ dsos__load_kernel();
+ /*
+ * Read the list of loaded modules with its build_ids
+ */
+ dsos__load_modules();
+
+ /* Write build-ids */
+ buildid_sec->offset = lseek(fd, 0, SEEK_CUR);
+ err = dsos__write_buildid_table(fd);
+ if (err < 0) {
+ pr_debug("failed to write buildid table\n");
+ goto out_free;
+ }
+ buildid_sec->size = lseek(fd, 0, SEEK_CUR) - buildid_sec->offset;
+ }
+
+ lseek(fd, sec_start, SEEK_SET);
+ err = do_write(fd, feat_sec, sec_size);
+ if (err < 0)
+ pr_debug("failed to write feature section\n");
+out_free:
+ free(feat_sec);
+ return err;
+}
+
+int perf_header__write(struct perf_header *self, int fd, bool at_exit)
{
struct perf_file_header f_header;
struct perf_file_attr f_attr;
struct perf_header_attr *attr;
- int i;
+ int i, err;
lseek(fd, sizeof(f_header), SEEK_SET);
@@ -186,7 +292,11 @@ void perf_header__write(struct perf_header *self, int fd)
attr = self->attr[i];
attr->id_offset = lseek(fd, 0, SEEK_CUR);
- do_write(fd, attr->id, attr->ids * sizeof(u64));
+ err = do_write(fd, attr->id, attr->ids * sizeof(u64));
+ if (err < 0) {
+ pr_debug("failed to write perf header\n");
+ return err;
+ }
}
@@ -202,34 +312,31 @@ void perf_header__write(struct perf_header *self, int fd)
.size = attr->ids * sizeof(u64),
}
};
- do_write(fd, &f_attr, sizeof(f_attr));
+ err = do_write(fd, &f_attr, sizeof(f_attr));
+ if (err < 0) {
+ pr_debug("failed to write perf header attribute\n");
+ return err;
+ }
}
self->event_offset = lseek(fd, 0, SEEK_CUR);
self->event_size = event_count * sizeof(struct perf_trace_event_type);
- if (events)
- do_write(fd, events, self->event_size);
-
- if (trace_info) {
- static int trace_info_written;
-
- /*
- * Write it only once
- */
- if (!trace_info_written) {
- self->trace_info_offset = lseek(fd, 0, SEEK_CUR);
- read_tracing_data(fd, attrs, nr_counters);
- self->trace_info_size = lseek(fd, 0, SEEK_CUR) -
- self->trace_info_offset;
- trace_info_written = 1;
- } else {
- lseek(fd, self->trace_info_offset +
- self->trace_info_size, SEEK_SET);
+ if (events) {
+ err = do_write(fd, events, self->event_size);
+ if (err < 0) {
+ pr_debug("failed to write perf header events\n");
+ return err;
}
}
self->data_offset = lseek(fd, 0, SEEK_CUR);
+ if (at_exit) {
+ err = perf_header__adds_write(self, fd);
+ if (err < 0)
+ return err;
+ }
+
f_header = (struct perf_file_header){
.magic = PERF_MAGIC,
.size = sizeof(f_header),
@@ -246,17 +353,20 @@ void perf_header__write(struct perf_header *self, int fd)
.offset = self->event_offset,
.size = self->event_size,
},
- .trace_info = {
- .offset = self->trace_info_offset,
- .size = self->trace_info_size,
- },
};
+ memcpy(&f_header.adds_features, &self->adds_features, sizeof(self->adds_features));
+
lseek(fd, 0, SEEK_SET);
- do_write(fd, &f_header, sizeof(f_header));
+ err = do_write(fd, &f_header, sizeof(f_header));
+ if (err < 0) {
+ pr_debug("failed to write perf header\n");
+ return err;
+ }
lseek(fd, self->data_offset + self->data_size, SEEK_SET);
self->frozen = 1;
+ return 0;
}
static void do_read(int fd, void *buf, size_t size)
@@ -274,29 +384,110 @@ static void do_read(int fd, void *buf, size_t size)
}
}
-struct perf_header *perf_header__read(int fd)
+int perf_header__process_sections(struct perf_header *self, int fd,
+ int (*process)(struct perf_file_section *self,
+ int feat, int fd))
{
- struct perf_header *self = perf_header__new();
- struct perf_file_header f_header;
- struct perf_file_attr f_attr;
- u64 f_id;
+ struct perf_file_section *feat_sec;
+ int nr_sections;
+ int sec_size;
+ int idx = 0;
+ int err = 0, feat = 1;
- int nr_attrs, nr_ids, i, j;
+ nr_sections = bitmap_weight(self->adds_features, HEADER_FEAT_BITS);
+ if (!nr_sections)
+ return 0;
+
+ feat_sec = calloc(sizeof(*feat_sec), nr_sections);
+ if (!feat_sec)
+ return -1;
+
+ sec_size = sizeof(*feat_sec) * nr_sections;
+ lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+
+ do_read(fd, feat_sec, sec_size);
+
+ while (idx < nr_sections && feat < HEADER_LAST_FEATURE) {
+ if (perf_header__has_feat(self, feat)) {
+ struct perf_file_section *sec = &feat_sec[idx++];
+
+ err = process(sec, feat, fd);
+ if (err < 0)
+ break;
+ }
+ ++feat;
+ }
+
+ free(feat_sec);
+ return err;
+};
+
+int perf_file_header__read(struct perf_file_header *self,
+ struct perf_header *ph, int fd)
+{
lseek(fd, 0, SEEK_SET);
- do_read(fd, &f_header, sizeof(f_header));
+ do_read(fd, self, sizeof(*self));
- if (f_header.magic != PERF_MAGIC ||
- f_header.attr_size != sizeof(f_attr))
- die("incompatible file format");
+ if (self->magic != PERF_MAGIC ||
+ self->attr_size != sizeof(struct perf_file_attr))
+ return -1;
- if (f_header.size != sizeof(f_header)) {
+ if (self->size != sizeof(*self)) {
/* Support the previous format */
- if (f_header.size == offsetof(typeof(f_header), trace_info))
- f_header.trace_info.size = 0;
+ if (self->size == offsetof(typeof(*self), adds_features))
+ bitmap_zero(self->adds_features, HEADER_FEAT_BITS);
else
- die("incompatible file format");
+ return -1;
}
+
+ memcpy(&ph->adds_features, &self->adds_features,
+ sizeof(self->adds_features));
+
+ ph->event_offset = self->event_types.offset;
+ ph->event_size = self->event_types.size;
+ ph->data_offset = self->data.offset;
+ ph->data_size = self->data.size;
+ return 0;
+}
+
+static int perf_file_section__process(struct perf_file_section *self,
+ int feat, int fd)
+{
+ if (lseek(fd, self->offset, SEEK_SET) < 0) {
+ pr_debug("Failed to lseek to %Ld offset for feature %d, "
+ "continuing...\n", self->offset, feat);
+ return 0;
+ }
+
+ switch (feat) {
+ case HEADER_TRACE_INFO:
+ trace_report(fd);
+ break;
+
+ case HEADER_BUILD_ID:
+ if (perf_header__read_build_ids(fd, self->offset, self->size))
+ pr_debug("Failed to read buildids, continuing...\n");
+ break;
+ default:
+ pr_debug("unknown feature %d, continuing...\n", feat);
+ }
+
+ return 0;
+}
+
+int perf_header__read(struct perf_header *self, int fd)
+{
+ struct perf_file_header f_header;
+ struct perf_file_attr f_attr;
+ u64 f_id;
+ int nr_attrs, nr_ids, i, j;
+
+ if (perf_file_header__read(&f_header, self, fd) < 0) {
+ pr_debug("incompatible file format\n");
+ return -EINVAL;
+ }
+
nr_attrs = f_header.attrs.size / sizeof(f_attr);
lseek(fd, f_header.attrs.offset, SEEK_SET);
@@ -308,6 +499,8 @@ struct perf_header *perf_header__read(int fd)
tmp = lseek(fd, 0, SEEK_CUR);
attr = perf_header_attr__new(&f_attr.attr);
+ if (attr == NULL)
+ return -ENOMEM;
nr_ids = f_attr.ids.size / sizeof(u64);
lseek(fd, f_attr.ids.offset, SEEK_SET);
@@ -315,40 +508,34 @@ struct perf_header *perf_header__read(int fd)
for (j = 0; j < nr_ids; j++) {
do_read(fd, &f_id, sizeof(f_id));
- perf_header_attr__add_id(attr, f_id);
+ if (perf_header_attr__add_id(attr, f_id) < 0) {
+ perf_header_attr__delete(attr);
+ return -ENOMEM;
+ }
+ }
+ if (perf_header__add_attr(self, attr) < 0) {
+ perf_header_attr__delete(attr);
+ return -ENOMEM;
}
- perf_header__add_attr(self, attr);
+
lseek(fd, tmp, SEEK_SET);
}
if (f_header.event_types.size) {
lseek(fd, f_header.event_types.offset, SEEK_SET);
events = malloc(f_header.event_types.size);
- if (!events)
- die("nomem");
+ if (events == NULL)
+ return -ENOMEM;
do_read(fd, events, f_header.event_types.size);
event_count = f_header.event_types.size / sizeof(struct perf_trace_event_type);
}
- self->trace_info_offset = f_header.trace_info.offset;
- self->trace_info_size = f_header.trace_info.size;
-
- if (self->trace_info_size) {
- lseek(fd, self->trace_info_offset, SEEK_SET);
- trace_report(fd);
- }
-
- self->event_offset = f_header.event_types.offset;
- self->event_size = f_header.event_types.size;
-
- self->data_offset = f_header.data.offset;
- self->data_size = f_header.data.size;
+ perf_header__process_sections(self, fd, perf_file_section__process);
lseek(fd, self->data_offset, SEEK_SET);
self->frozen = 1;
-
- return self;
+ return 0;
}
u64 perf_header__sample_type(struct perf_header *header)
diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h
index 30aee5160dc0..d1dbe2b79c42 100644
--- a/tools/perf/util/header.h
+++ b/tools/perf/util/header.h
@@ -3,8 +3,11 @@
#include "../../../include/linux/perf_event.h"
#include <sys/types.h>
+#include <stdbool.h>
#include "types.h"
+#include <linux/bitmap.h>
+
struct perf_header_attr {
struct perf_event_attr attr;
int ids, size;
@@ -12,38 +15,71 @@ struct perf_header_attr {
off_t id_offset;
};
+enum {
+ HEADER_TRACE_INFO = 1,
+ HEADER_BUILD_ID,
+ HEADER_LAST_FEATURE,
+};
+
+#define HEADER_FEAT_BITS 256
+
+struct perf_file_section {
+ u64 offset;
+ u64 size;
+};
+
+struct perf_file_header {
+ u64 magic;
+ u64 size;
+ u64 attr_size;
+ struct perf_file_section attrs;
+ struct perf_file_section data;
+ struct perf_file_section event_types;
+ DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
+};
+
+struct perf_header;
+
+int perf_file_header__read(struct perf_file_header *self,
+ struct perf_header *ph, int fd);
+
struct perf_header {
- int frozen;
- int attrs, size;
+ int frozen;
+ int attrs, size;
struct perf_header_attr **attr;
- s64 attr_offset;
- u64 data_offset;
- u64 data_size;
- u64 event_offset;
- u64 event_size;
- u64 trace_info_offset;
- u64 trace_info_size;
+ s64 attr_offset;
+ u64 data_offset;
+ u64 data_size;
+ u64 event_offset;
+ u64 event_size;
+ DECLARE_BITMAP(adds_features, HEADER_FEAT_BITS);
};
-struct perf_header *perf_header__read(int fd);
-void perf_header__write(struct perf_header *self, int fd);
+struct perf_header *perf_header__new(void);
+void perf_header__delete(struct perf_header *self);
+
+int perf_header__read(struct perf_header *self, int fd);
+int perf_header__write(struct perf_header *self, int fd, bool at_exit);
-void perf_header__add_attr(struct perf_header *self,
- struct perf_header_attr *attr);
+int perf_header__add_attr(struct perf_header *self,
+ struct perf_header_attr *attr);
void perf_header__push_event(u64 id, const char *name);
char *perf_header__find_event(u64 id);
+struct perf_header_attr *perf_header_attr__new(struct perf_event_attr *attr);
+void perf_header_attr__delete(struct perf_header_attr *self);
-struct perf_header_attr *
-perf_header_attr__new(struct perf_event_attr *attr);
-void perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
+int perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
u64 perf_header__sample_type(struct perf_header *header);
struct perf_event_attr *
perf_header__find_attr(u64 id, struct perf_header *header);
-void perf_header__set_trace_info(void);
+void perf_header__set_feat(struct perf_header *self, int feat);
+bool perf_header__has_feat(const struct perf_header *self, int feat);
-struct perf_header *perf_header__new(void);
+int perf_header__process_sections(struct perf_header *self, int fd,
+ int (*process)(struct perf_file_section *self,
+ int feat, int fd));
#endif /* __PERF_HEADER_H */
diff --git a/tools/perf/util/include/asm/asm-offsets.h b/tools/perf/util/include/asm/asm-offsets.h
new file mode 100644
index 000000000000..ed538942523d
--- /dev/null
+++ b/tools/perf/util/include/asm/asm-offsets.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/bitops.h b/tools/perf/util/include/asm/bitops.h
new file mode 100644
index 000000000000..58e9817ffae0
--- /dev/null
+++ b/tools/perf/util/include/asm/bitops.h
@@ -0,0 +1,18 @@
+#ifndef _PERF_ASM_BITOPS_H_
+#define _PERF_ASM_BITOPS_H_
+
+#include <sys/types.h>
+#include "../../types.h"
+#include <linux/compiler.h>
+
+/* CHECKME: Not sure both always match */
+#define BITS_PER_LONG __WORDSIZE
+
+#include "../../../../include/asm-generic/bitops/__fls.h"
+#include "../../../../include/asm-generic/bitops/fls.h"
+#include "../../../../include/asm-generic/bitops/fls64.h"
+#include "../../../../include/asm-generic/bitops/__ffs.h"
+#include "../../../../include/asm-generic/bitops/ffz.h"
+#include "../../../../include/asm-generic/bitops/hweight.h"
+
+#endif
diff --git a/tools/perf/util/include/asm/byteorder.h b/tools/perf/util/include/asm/byteorder.h
new file mode 100644
index 000000000000..b722abe3a626
--- /dev/null
+++ b/tools/perf/util/include/asm/byteorder.h
@@ -0,0 +1,2 @@
+#include <asm/types.h>
+#include "../../../../include/linux/swab.h"
diff --git a/tools/perf/util/include/asm/swab.h b/tools/perf/util/include/asm/swab.h
new file mode 100644
index 000000000000..ed538942523d
--- /dev/null
+++ b/tools/perf/util/include/asm/swab.h
@@ -0,0 +1 @@
+/* stub */
diff --git a/tools/perf/util/include/asm/uaccess.h b/tools/perf/util/include/asm/uaccess.h
new file mode 100644
index 000000000000..d0f72b8fcc35
--- /dev/null
+++ b/tools/perf/util/include/asm/uaccess.h
@@ -0,0 +1,14 @@
+#ifndef _PERF_ASM_UACCESS_H_
+#define _PERF_ASM_UACCESS_H_
+
+#define __get_user(src, dest) \
+({ \
+ (src) = *dest; \
+ 0; \
+})
+
+#define get_user __get_user
+
+#define access_ok(type, addr, size) 1
+
+#endif
diff --git a/tools/perf/util/include/linux/bitmap.h b/tools/perf/util/include/linux/bitmap.h
new file mode 100644
index 000000000000..94507639a8c4
--- /dev/null
+++ b/tools/perf/util/include/linux/bitmap.h
@@ -0,0 +1,3 @@
+#include "../../../../include/linux/bitmap.h"
+#include "../../../../include/asm-generic/bitops/find.h"
+#include <linux/errno.h>
diff --git a/tools/perf/util/include/linux/bitops.h b/tools/perf/util/include/linux/bitops.h
new file mode 100644
index 000000000000..ace57c36d1d0
--- /dev/null
+++ b/tools/perf/util/include/linux/bitops.h
@@ -0,0 +1,27 @@
+#ifndef _PERF_LINUX_BITOPS_H_
+#define _PERF_LINUX_BITOPS_H_
+
+#define __KERNEL__
+
+#define CONFIG_GENERIC_FIND_NEXT_BIT
+#define CONFIG_GENERIC_FIND_FIRST_BIT
+#include "../../../../include/linux/bitops.h"
+
+static inline void set_bit(int nr, unsigned long *addr)
+{
+ addr[nr / BITS_PER_LONG] |= 1UL << (nr % BITS_PER_LONG);
+}
+
+static __always_inline int test_bit(unsigned int nr, const unsigned long *addr)
+{
+ return ((1UL << (nr % BITS_PER_LONG)) &
+ (((unsigned long *)addr)[nr / BITS_PER_LONG])) != 0;
+}
+
+unsigned long generic_find_next_zero_le_bit(const unsigned long *addr, unsigned
+ long size, unsigned long offset);
+
+unsigned long generic_find_next_le_bit(const unsigned long *addr, unsigned
+ long size, unsigned long offset);
+
+#endif
diff --git a/tools/perf/util/include/linux/compiler.h b/tools/perf/util/include/linux/compiler.h
new file mode 100644
index 000000000000..dfb0713ed47f
--- /dev/null
+++ b/tools/perf/util/include/linux/compiler.h
@@ -0,0 +1,10 @@
+#ifndef _PERF_LINUX_COMPILER_H_
+#define _PERF_LINUX_COMPILER_H_
+
+#ifndef __always_inline
+#define __always_inline inline
+#endif
+#define __user
+#define __attribute_const__
+
+#endif
diff --git a/tools/perf/util/include/linux/ctype.h b/tools/perf/util/include/linux/ctype.h
new file mode 100644
index 000000000000..a53d4ee1e0b7
--- /dev/null
+++ b/tools/perf/util/include/linux/ctype.h
@@ -0,0 +1 @@
+#include "../util.h"
diff --git a/tools/perf/util/include/linux/kernel.h b/tools/perf/util/include/linux/kernel.h
index a6b87390cb52..21c0274c02fa 100644
--- a/tools/perf/util/include/linux/kernel.h
+++ b/tools/perf/util/include/linux/kernel.h
@@ -1,6 +1,16 @@
#ifndef PERF_LINUX_KERNEL_H_
#define PERF_LINUX_KERNEL_H_
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <assert.h>
+
+#define DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d))
+
+#define ALIGN(x,a) __ALIGN_MASK(x,(typeof(x))(a)-1)
+#define __ALIGN_MASK(x,mask) (((x)+(mask))&~(mask))
+
#ifndef offsetof
#define offsetof(TYPE, MEMBER) ((size_t) &((TYPE *)0)->MEMBER)
#endif
@@ -26,4 +36,70 @@
_max1 > _max2 ? _max1 : _max2; })
#endif
+#ifndef min
+#define min(x, y) ({ \
+ typeof(x) _min1 = (x); \
+ typeof(y) _min2 = (y); \
+ (void) (&_min1 == &_min2); \
+ _min1 < _min2 ? _min1 : _min2; })
+#endif
+
+#ifndef BUG_ON
+#define BUG_ON(cond) assert(!(cond))
+#endif
+
+/*
+ * Both need more care to handle endianness
+ * (Don't use bitmap_copy_le() for now)
+ */
+#define cpu_to_le64(x) (x)
+#define cpu_to_le32(x) (x)
+
+static inline int
+vscnprintf(char *buf, size_t size, const char *fmt, va_list args)
+{
+ int i;
+ ssize_t ssize = size;
+
+ i = vsnprintf(buf, size, fmt, args);
+
+ return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline int scnprintf(char * buf, size_t size, const char * fmt, ...)
+{
+ va_list args;
+ ssize_t ssize = size;
+ int i;
+
+ va_start(args, fmt);
+ i = vsnprintf(buf, size, fmt, args);
+ va_end(args);
+
+ return (i >= ssize) ? (ssize - 1) : i;
+}
+
+static inline unsigned long
+simple_strtoul(const char *nptr, char **endptr, int base)
+{
+ return strtoul(nptr, endptr, base);
+}
+
+#ifndef pr_fmt
+#define pr_fmt(fmt) fmt
+#endif
+
+#define pr_err(fmt, ...) \
+ do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_warning(fmt, ...) \
+ do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_info(fmt, ...) \
+ do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0)
+#define pr_debug(fmt, ...) \
+ eprintf(1, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debugN(n, fmt, ...) \
+ eprintf(n, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_debug3(fmt, ...) pr_debugN(3, pr_fmt(fmt), ##__VA_ARGS__)
+
#endif
diff --git a/tools/perf/util/include/linux/string.h b/tools/perf/util/include/linux/string.h
new file mode 100644
index 000000000000..3b2f5900276f
--- /dev/null
+++ b/tools/perf/util/include/linux/string.h
@@ -0,0 +1 @@
+#include <string.h>
diff --git a/tools/perf/util/include/linux/types.h b/tools/perf/util/include/linux/types.h
new file mode 100644
index 000000000000..196862a81a21
--- /dev/null
+++ b/tools/perf/util/include/linux/types.h
@@ -0,0 +1,9 @@
+#ifndef _PERF_LINUX_TYPES_H_
+#define _PERF_LINUX_TYPES_H_
+
+#include <asm/types.h>
+
+#define DECLARE_BITMAP(name,bits) \
+ unsigned long name[BITS_TO_LONGS(bits)]
+
+#endif
diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index 804e02382739..94ca95073c40 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -3,6 +3,7 @@
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
+#include "debug.h"
static inline int is_anon_memory(const char *filename)
{
@@ -19,13 +20,26 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
return n;
}
- struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen)
+void map__init(struct map *self, u64 start, u64 end, u64 pgoff,
+ struct dso *dso)
+{
+ self->start = start;
+ self->end = end;
+ self->pgoff = pgoff;
+ self->dso = dso;
+ self->map_ip = map__map_ip;
+ self->unmap_ip = map__unmap_ip;
+ RB_CLEAR_NODE(&self->rb_node);
+}
+
+struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen)
{
struct map *self = malloc(sizeof(*self));
if (self != NULL) {
const char *filename = event->filename;
char newfilename[PATH_MAX];
+ struct dso *dso;
int anon;
if (cwd) {
@@ -45,18 +59,15 @@ static int strcommon(const char *pathname, char *cwd, int cwdlen)
filename = newfilename;
}
- self->start = event->start;
- self->end = event->start + event->len;
- self->pgoff = event->pgoff;
-
- self->dso = dsos__findnew(filename);
- if (self->dso == NULL)
+ dso = dsos__findnew(filename);
+ if (dso == NULL)
goto out_delete;
+ map__init(self, event->start, event->start + event->len,
+ event->pgoff, dso);
+
if (self->dso == vdso || anon)
- self->map_ip = vdso__map_ip;
- else
- self->map_ip = map__map_ip;
+ self->map_ip = self->unmap_ip = identity__map_ip;
}
return self;
out_delete:
@@ -64,6 +75,47 @@ out_delete:
return NULL;
}
+#define DSO__DELETED "(deleted)"
+
+struct symbol *
+map__find_symbol(struct map *self, u64 ip, symbol_filter_t filter)
+{
+ if (!self->dso->loaded) {
+ int nr = dso__load(self->dso, self, filter);
+
+ if (nr < 0) {
+ if (self->dso->has_build_id) {
+ char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+ build_id__sprintf(self->dso->build_id,
+ sizeof(self->dso->build_id),
+ sbuild_id);
+ pr_warning("%s with build id %s not found",
+ self->dso->long_name, sbuild_id);
+ } else
+ pr_warning("Failed to open %s",
+ self->dso->long_name);
+ pr_warning(", continuing without symbols\n");
+ return NULL;
+ } else if (nr == 0) {
+ const char *name = self->dso->long_name;
+ const size_t len = strlen(name);
+ const size_t real_len = len - sizeof(DSO__DELETED);
+
+ if (len > sizeof(DSO__DELETED) &&
+ strcmp(name + real_len + 1, DSO__DELETED) == 0) {
+ pr_warning("%.*s was updated, restart the long running apps that use it!\n",
+ (int)real_len, name);
+ } else {
+ pr_warning("no symbols found in %s, maybe install a debug package?\n", name);
+ }
+ return NULL;
+ }
+ }
+
+ return self->dso->find_symbol(self->dso, ip);
+}
+
struct map *map__clone(struct map *self)
{
struct map *map = malloc(sizeof(*self));
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index b097570e9623..0faf4f2bb5ca 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -7,6 +7,7 @@
#include "string.h"
#include "cache.h"
#include "header.h"
+#include "debugfs.h"
int nr_counters;
@@ -47,6 +48,8 @@ static struct event_symbol event_symbols[] = {
{ CSW(PAGE_FAULTS_MAJ), "major-faults", "" },
{ CSW(CONTEXT_SWITCHES), "context-switches", "cs" },
{ CSW(CPU_MIGRATIONS), "cpu-migrations", "migrations" },
+ { CSW(ALIGNMENT_FAULTS), "alignment-faults", "" },
+ { CSW(EMULATION_FAULTS), "emulation-faults", "" },
};
#define __PERF_EVENT_FIELD(config, name) \
@@ -75,6 +78,8 @@ static const char *sw_event_names[] = {
"CPU-migrations",
"minor-faults",
"major-faults",
+ "alignment-faults",
+ "emulation-faults",
};
#define MAX_ALIASES 8
@@ -149,16 +154,6 @@ static int tp_event_has_id(struct dirent *sys_dir, struct dirent *evt_dir)
#define MAX_EVENT_LENGTH 512
-int valid_debugfs_mount(const char *debugfs)
-{
- struct statfs st_fs;
-
- if (statfs(debugfs, &st_fs) < 0)
- return -ENOENT;
- else if (st_fs.f_type != (long) DEBUGFS_MAGIC)
- return -ENOENT;
- return 0;
-}
struct tracepoint_path *tracepoint_id_to_path(u64 config)
{
@@ -171,7 +166,7 @@ struct tracepoint_path *tracepoint_id_to_path(u64 config)
char evt_path[MAXPATHLEN];
char dir_path[MAXPATHLEN];
- if (valid_debugfs_mount(debugfs_path))
+ if (debugfs_valid_mountpoint(debugfs_path))
return NULL;
sys_dir = opendir(debugfs_path);
@@ -510,7 +505,7 @@ static enum event_result parse_tracepoint_event(const char **strp,
char sys_name[MAX_EVENT_LENGTH];
unsigned int sys_length, evt_length;
- if (valid_debugfs_mount(debugfs_path))
+ if (debugfs_valid_mountpoint(debugfs_path))
return 0;
evt_name = strchr(*strp, ':');
@@ -678,6 +673,8 @@ parse_event_symbols(const char **str, struct perf_event_attr *attr)
if (ret != EVT_FAILED)
goto modifier;
+ fprintf(stderr, "invalid or unsupported event: '%s'\n", *str);
+ fprintf(stderr, "Run 'perf list' for a list of valid events\n");
return EVT_FAILED;
modifier:
@@ -786,7 +783,7 @@ static void print_tracepoint_events(void)
char evt_path[MAXPATHLEN];
char dir_path[MAXPATHLEN];
- if (valid_debugfs_mount(debugfs_path))
+ if (debugfs_valid_mountpoint(debugfs_path))
return;
sys_dir = opendir(debugfs_path);
@@ -804,7 +801,7 @@ static void print_tracepoint_events(void)
for_each_event(sys_dirent, evt_dir, evt_dirent, evt_next) {
snprintf(evt_path, MAXPATHLEN, "%s:%s",
sys_dirent.d_name, evt_dirent.d_name);
- fprintf(stderr, " %-42s [%s]\n", evt_path,
+ printf(" %-42s [%s]\n", evt_path,
event_type_descriptors[PERF_TYPE_TRACEPOINT+1]);
}
closedir(evt_dir);
@@ -821,8 +818,8 @@ void print_events(void)
unsigned int i, type, op, prev_type = -1;
char name[40];
- fprintf(stderr, "\n");
- fprintf(stderr, "List of pre-defined events (to be used in -e):\n");
+ printf("\n");
+ printf("List of pre-defined events (to be used in -e):\n");
for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) {
type = syms->type + 1;
@@ -830,19 +827,19 @@ void print_events(void)
type = 0;
if (type != prev_type)
- fprintf(stderr, "\n");
+ printf("\n");
if (strlen(syms->alias))
sprintf(name, "%s OR %s", syms->symbol, syms->alias);
else
strcpy(name, syms->symbol);
- fprintf(stderr, " %-42s [%s]\n", name,
+ printf(" %-42s [%s]\n", name,
event_type_descriptors[type]);
prev_type = type;
}
- fprintf(stderr, "\n");
+ printf("\n");
for (type = 0; type < PERF_COUNT_HW_CACHE_MAX; type++) {
for (op = 0; op < PERF_COUNT_HW_CACHE_OP_MAX; op++) {
/* skip invalid cache type */
@@ -850,17 +847,17 @@ void print_events(void)
continue;
for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
- fprintf(stderr, " %-42s [%s]\n",
+ printf(" %-42s [%s]\n",
event_cache_name(type, op, i),
event_type_descriptors[4]);
}
}
}
- fprintf(stderr, "\n");
- fprintf(stderr, " %-42s [raw hardware event descriptor]\n",
+ printf("\n");
+ printf(" %-42s [raw hardware event descriptor]\n",
"rNNN");
- fprintf(stderr, "\n");
+ printf("\n");
print_tracepoint_events();
diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
new file mode 100644
index 000000000000..293cdfc1b8ca
--- /dev/null
+++ b/tools/perf/util/probe-finder.c
@@ -0,0 +1,732 @@
+/*
+ * probe-finder.c : C expression to kprobe event converter
+ *
+ * Written by Masami Hiramatsu <mhiramat@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ */
+
+#include <sys/utsname.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <getopt.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+#include <ctype.h>
+
+#include "event.h"
+#include "debug.h"
+#include "util.h"
+#include "probe-finder.h"
+
+
+/* Dwarf_Die Linkage to parent Die */
+struct die_link {
+ struct die_link *parent; /* Parent die */
+ Dwarf_Die die; /* Current die */
+};
+
+static Dwarf_Debug __dw_debug;
+static Dwarf_Error __dw_error;
+
+/*
+ * Generic dwarf analysis helpers
+ */
+
+#define X86_32_MAX_REGS 8
+const char *x86_32_regs_table[X86_32_MAX_REGS] = {
+ "%ax",
+ "%cx",
+ "%dx",
+ "%bx",
+ "$stack", /* Stack address instead of %sp */
+ "%bp",
+ "%si",
+ "%di",
+};
+
+#define X86_64_MAX_REGS 16
+const char *x86_64_regs_table[X86_64_MAX_REGS] = {
+ "%ax",
+ "%dx",
+ "%cx",
+ "%bx",
+ "%si",
+ "%di",
+ "%bp",
+ "%sp",
+ "%r8",
+ "%r9",
+ "%r10",
+ "%r11",
+ "%r12",
+ "%r13",
+ "%r14",
+ "%r15",
+};
+
+/* TODO: switching by dwarf address size */
+#ifdef __x86_64__
+#define ARCH_MAX_REGS X86_64_MAX_REGS
+#define arch_regs_table x86_64_regs_table
+#else
+#define ARCH_MAX_REGS X86_32_MAX_REGS
+#define arch_regs_table x86_32_regs_table
+#endif
+
+/* Return architecture dependent register string (for kprobe-tracer) */
+static const char *get_arch_regstr(unsigned int n)
+{
+ return (n <= ARCH_MAX_REGS) ? arch_regs_table[n] : NULL;
+}
+
+/*
+ * Compare the tail of two strings.
+ * Return 0 if whole of either string is same as another's tail part.
+ */
+static int strtailcmp(const char *s1, const char *s2)
+{
+ int i1 = strlen(s1);
+ int i2 = strlen(s2);
+ while (--i1 > 0 && --i2 > 0) {
+ if (s1[i1] != s2[i2])
+ return s1[i1] - s2[i2];
+ }
+ return 0;
+}
+
+/* Find the fileno of the target file. */
+static Dwarf_Unsigned cu_find_fileno(Dwarf_Die cu_die, const char *fname)
+{
+ Dwarf_Signed cnt, i;
+ Dwarf_Unsigned found = 0;
+ char **srcs;
+ int ret;
+
+ if (!fname)
+ return 0;
+
+ ret = dwarf_srcfiles(cu_die, &srcs, &cnt, &__dw_error);
+ if (ret == DW_DLV_OK) {
+ for (i = 0; i < cnt && !found; i++) {
+ if (strtailcmp(srcs[i], fname) == 0)
+ found = i + 1;
+ dwarf_dealloc(__dw_debug, srcs[i], DW_DLA_STRING);
+ }
+ for (; i < cnt; i++)
+ dwarf_dealloc(__dw_debug, srcs[i], DW_DLA_STRING);
+ dwarf_dealloc(__dw_debug, srcs, DW_DLA_LIST);
+ }
+ if (found)
+ pr_debug("found fno: %d\n", (int)found);
+ return found;
+}
+
+/* Compare diename and tname */
+static int die_compare_name(Dwarf_Die dw_die, const char *tname)
+{
+ char *name;
+ int ret;
+ ret = dwarf_diename(dw_die, &name, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK) {
+ ret = strcmp(tname, name);
+ dwarf_dealloc(__dw_debug, name, DW_DLA_STRING);
+ } else
+ ret = -1;
+ return ret;
+}
+
+/* Check the address is in the subprogram(function). */
+static int die_within_subprogram(Dwarf_Die sp_die, Dwarf_Addr addr,
+ Dwarf_Signed *offs)
+{
+ Dwarf_Addr lopc, hipc;
+ int ret;
+
+ /* TODO: check ranges */
+ ret = dwarf_lowpc(sp_die, &lopc, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_NO_ENTRY)
+ return 0;
+ ret = dwarf_highpc(sp_die, &hipc, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ if (lopc <= addr && addr < hipc) {
+ *offs = addr - lopc;
+ return 1;
+ } else
+ return 0;
+}
+
+/* Check the die is inlined function */
+static Dwarf_Bool die_inlined_subprogram(Dwarf_Die dw_die)
+{
+ /* TODO: check strictly */
+ Dwarf_Bool inl;
+ int ret;
+
+ ret = dwarf_hasattr(dw_die, DW_AT_inline, &inl, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ return inl;
+}
+
+/* Get the offset of abstruct_origin */
+static Dwarf_Off die_get_abstract_origin(Dwarf_Die dw_die)
+{
+ Dwarf_Attribute attr;
+ Dwarf_Off cu_offs;
+ int ret;
+
+ ret = dwarf_attr(dw_die, DW_AT_abstract_origin, &attr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ ret = dwarf_formref(attr, &cu_offs, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+ return cu_offs;
+}
+
+/* Get entry pc(or low pc, 1st entry of ranges) of the die */
+static Dwarf_Addr die_get_entrypc(Dwarf_Die dw_die)
+{
+ Dwarf_Attribute attr;
+ Dwarf_Addr addr;
+ Dwarf_Off offs;
+ Dwarf_Ranges *ranges;
+ Dwarf_Signed cnt;
+ int ret;
+
+ /* Try to get entry pc */
+ ret = dwarf_attr(dw_die, DW_AT_entry_pc, &attr, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK) {
+ ret = dwarf_formaddr(attr, &addr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+ return addr;
+ }
+
+ /* Try to get low pc */
+ ret = dwarf_lowpc(dw_die, &addr, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK)
+ return addr;
+
+ /* Try to get ranges */
+ ret = dwarf_attr(dw_die, DW_AT_ranges, &attr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ ret = dwarf_formref(attr, &offs, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ ret = dwarf_get_ranges(__dw_debug, offs, &ranges, &cnt, NULL,
+ &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ addr = ranges[0].dwr_addr1;
+ dwarf_ranges_dealloc(__dw_debug, ranges, cnt);
+ return addr;
+}
+
+/*
+ * Search a Die from Die tree.
+ * Note: cur_link->die should be deallocated in this function.
+ */
+static int __search_die_tree(struct die_link *cur_link,
+ int (*die_cb)(struct die_link *, void *),
+ void *data)
+{
+ Dwarf_Die new_die;
+ struct die_link new_link;
+ int ret;
+
+ if (!die_cb)
+ return 0;
+
+ /* Check current die */
+ while (!(ret = die_cb(cur_link, data))) {
+ /* Check child die */
+ ret = dwarf_child(cur_link->die, &new_die, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK) {
+ new_link.parent = cur_link;
+ new_link.die = new_die;
+ ret = __search_die_tree(&new_link, die_cb, data);
+ if (ret)
+ break;
+ }
+
+ /* Move to next sibling */
+ ret = dwarf_siblingof(__dw_debug, cur_link->die, &new_die,
+ &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ dwarf_dealloc(__dw_debug, cur_link->die, DW_DLA_DIE);
+ cur_link->die = new_die;
+ if (ret == DW_DLV_NO_ENTRY)
+ return 0;
+ }
+ dwarf_dealloc(__dw_debug, cur_link->die, DW_DLA_DIE);
+ return ret;
+}
+
+/* Search a die in its children's die tree */
+static int search_die_from_children(Dwarf_Die parent_die,
+ int (*die_cb)(struct die_link *, void *),
+ void *data)
+{
+ struct die_link new_link;
+ int ret;
+
+ new_link.parent = NULL;
+ ret = dwarf_child(parent_die, &new_link.die, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK)
+ return __search_die_tree(&new_link, die_cb, data);
+ else
+ return 0;
+}
+
+/* Find a locdesc corresponding to the address */
+static int attr_get_locdesc(Dwarf_Attribute attr, Dwarf_Locdesc *desc,
+ Dwarf_Addr addr)
+{
+ Dwarf_Signed lcnt;
+ Dwarf_Locdesc **llbuf;
+ int ret, i;
+
+ ret = dwarf_loclist_n(attr, &llbuf, &lcnt, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ ret = DW_DLV_NO_ENTRY;
+ for (i = 0; i < lcnt; ++i) {
+ if (llbuf[i]->ld_lopc <= addr &&
+ llbuf[i]->ld_hipc > addr) {
+ memcpy(desc, llbuf[i], sizeof(Dwarf_Locdesc));
+ desc->ld_s =
+ malloc(sizeof(Dwarf_Loc) * llbuf[i]->ld_cents);
+ DIE_IF(desc->ld_s == NULL);
+ memcpy(desc->ld_s, llbuf[i]->ld_s,
+ sizeof(Dwarf_Loc) * llbuf[i]->ld_cents);
+ ret = DW_DLV_OK;
+ break;
+ }
+ dwarf_dealloc(__dw_debug, llbuf[i]->ld_s, DW_DLA_LOC_BLOCK);
+ dwarf_dealloc(__dw_debug, llbuf[i], DW_DLA_LOCDESC);
+ }
+ /* Releasing loop */
+ for (; i < lcnt; ++i) {
+ dwarf_dealloc(__dw_debug, llbuf[i]->ld_s, DW_DLA_LOC_BLOCK);
+ dwarf_dealloc(__dw_debug, llbuf[i], DW_DLA_LOCDESC);
+ }
+ dwarf_dealloc(__dw_debug, llbuf, DW_DLA_LIST);
+ return ret;
+}
+
+/* Get decl_file attribute value (file number) */
+static Dwarf_Unsigned die_get_decl_file(Dwarf_Die sp_die)
+{
+ Dwarf_Attribute attr;
+ Dwarf_Unsigned fno;
+ int ret;
+
+ ret = dwarf_attr(sp_die, DW_AT_decl_file, &attr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_formudata(attr, &fno, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+ return fno;
+}
+
+/* Get decl_line attribute value (line number) */
+static Dwarf_Unsigned die_get_decl_line(Dwarf_Die sp_die)
+{
+ Dwarf_Attribute attr;
+ Dwarf_Unsigned lno;
+ int ret;
+
+ ret = dwarf_attr(sp_die, DW_AT_decl_line, &attr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_formudata(attr, &lno, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+ return lno;
+}
+
+/*
+ * Probe finder related functions
+ */
+
+/* Show a location */
+static void show_location(Dwarf_Loc *loc, struct probe_finder *pf)
+{
+ Dwarf_Small op;
+ Dwarf_Unsigned regn;
+ Dwarf_Signed offs;
+ int deref = 0, ret;
+ const char *regs;
+
+ op = loc->lr_atom;
+
+ /* If this is based on frame buffer, set the offset */
+ if (op == DW_OP_fbreg) {
+ deref = 1;
+ offs = (Dwarf_Signed)loc->lr_number;
+ op = pf->fbloc.ld_s[0].lr_atom;
+ loc = &pf->fbloc.ld_s[0];
+ } else
+ offs = 0;
+
+ if (op >= DW_OP_breg0 && op <= DW_OP_breg31) {
+ regn = op - DW_OP_breg0;
+ offs += (Dwarf_Signed)loc->lr_number;
+ deref = 1;
+ } else if (op >= DW_OP_reg0 && op <= DW_OP_reg31) {
+ regn = op - DW_OP_reg0;
+ } else if (op == DW_OP_bregx) {
+ regn = loc->lr_number;
+ offs += (Dwarf_Signed)loc->lr_number2;
+ deref = 1;
+ } else if (op == DW_OP_regx) {
+ regn = loc->lr_number;
+ } else
+ die("Dwarf_OP %d is not supported.\n", op);
+
+ regs = get_arch_regstr(regn);
+ if (!regs)
+ die("%lld exceeds max register number.\n", regn);
+
+ if (deref)
+ ret = snprintf(pf->buf, pf->len,
+ " %s=%+lld(%s)", pf->var, offs, regs);
+ else
+ ret = snprintf(pf->buf, pf->len, " %s=%s", pf->var, regs);
+ DIE_IF(ret < 0);
+ DIE_IF(ret >= pf->len);
+}
+
+/* Show a variables in kprobe event format */
+static void show_variable(Dwarf_Die vr_die, struct probe_finder *pf)
+{
+ Dwarf_Attribute attr;
+ Dwarf_Locdesc ld;
+ int ret;
+
+ ret = dwarf_attr(vr_die, DW_AT_location, &attr, &__dw_error);
+ if (ret != DW_DLV_OK)
+ goto error;
+ ret = attr_get_locdesc(attr, &ld, (pf->addr - pf->cu_base));
+ if (ret != DW_DLV_OK)
+ goto error;
+ /* TODO? */
+ DIE_IF(ld.ld_cents != 1);
+ show_location(&ld.ld_s[0], pf);
+ free(ld.ld_s);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+ return ;
+error:
+ die("Failed to find the location of %s at this address.\n"
+ " Perhaps, it has been optimized out.\n", pf->var);
+}
+
+static int variable_callback(struct die_link *dlink, void *data)
+{
+ struct probe_finder *pf = (struct probe_finder *)data;
+ Dwarf_Half tag;
+ int ret;
+
+ ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if ((tag == DW_TAG_formal_parameter ||
+ tag == DW_TAG_variable) &&
+ (die_compare_name(dlink->die, pf->var) == 0)) {
+ show_variable(dlink->die, pf);
+ return 1;
+ }
+ /* TODO: Support struct members and arrays */
+ return 0;
+}
+
+/* Find a variable in a subprogram die */
+static void find_variable(Dwarf_Die sp_die, struct probe_finder *pf)
+{
+ int ret;
+
+ if (!is_c_varname(pf->var)) {
+ /* Output raw parameters */
+ ret = snprintf(pf->buf, pf->len, " %s", pf->var);
+ DIE_IF(ret < 0);
+ DIE_IF(ret >= pf->len);
+ return ;
+ }
+
+ pr_debug("Searching '%s' variable in context.\n", pf->var);
+ /* Search child die for local variables and parameters. */
+ ret = search_die_from_children(sp_die, variable_callback, pf);
+ if (!ret)
+ die("Failed to find '%s' in this function.\n", pf->var);
+}
+
+/* Get a frame base on the address */
+static void get_current_frame_base(Dwarf_Die sp_die, struct probe_finder *pf)
+{
+ Dwarf_Attribute attr;
+ int ret;
+
+ ret = dwarf_attr(sp_die, DW_AT_frame_base, &attr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ ret = attr_get_locdesc(attr, &pf->fbloc, (pf->addr - pf->cu_base));
+ DIE_IF(ret != DW_DLV_OK);
+ dwarf_dealloc(__dw_debug, attr, DW_DLA_ATTR);
+}
+
+static void free_current_frame_base(struct probe_finder *pf)
+{
+ free(pf->fbloc.ld_s);
+ memset(&pf->fbloc, 0, sizeof(Dwarf_Locdesc));
+}
+
+/* Show a probe point to output buffer */
+static void show_probepoint(Dwarf_Die sp_die, Dwarf_Signed offs,
+ struct probe_finder *pf)
+{
+ struct probe_point *pp = pf->pp;
+ char *name;
+ char tmp[MAX_PROBE_BUFFER];
+ int ret, i, len;
+
+ /* Output name of probe point */
+ ret = dwarf_diename(sp_die, &name, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_OK) {
+ ret = snprintf(tmp, MAX_PROBE_BUFFER, "%s+%u", name,
+ (unsigned int)offs);
+ /* Copy the function name if possible */
+ if (!pp->function) {
+ pp->function = strdup(name);
+ pp->offset = offs;
+ }
+ dwarf_dealloc(__dw_debug, name, DW_DLA_STRING);
+ } else {
+ /* This function has no name. */
+ ret = snprintf(tmp, MAX_PROBE_BUFFER, "0x%llx", pf->addr);
+ if (!pp->function) {
+ /* TODO: Use _stext */
+ pp->function = strdup("");
+ pp->offset = (int)pf->addr;
+ }
+ }
+ DIE_IF(ret < 0);
+ DIE_IF(ret >= MAX_PROBE_BUFFER);
+ len = ret;
+ pr_debug("Probe point found: %s\n", tmp);
+
+ /* Find each argument */
+ get_current_frame_base(sp_die, pf);
+ for (i = 0; i < pp->nr_args; i++) {
+ pf->var = pp->args[i];
+ pf->buf = &tmp[len];
+ pf->len = MAX_PROBE_BUFFER - len;
+ find_variable(sp_die, pf);
+ len += strlen(pf->buf);
+ }
+ free_current_frame_base(pf);
+
+ pp->probes[pp->found] = strdup(tmp);
+ pp->found++;
+}
+
+static int probeaddr_callback(struct die_link *dlink, void *data)
+{
+ struct probe_finder *pf = (struct probe_finder *)data;
+ Dwarf_Half tag;
+ Dwarf_Signed offs;
+ int ret;
+
+ ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ /* Check the address is in this subprogram */
+ if (tag == DW_TAG_subprogram &&
+ die_within_subprogram(dlink->die, pf->addr, &offs)) {
+ show_probepoint(dlink->die, offs, pf);
+ return 1;
+ }
+ return 0;
+}
+
+/* Find probe point from its line number */
+static void find_by_line(struct probe_finder *pf)
+{
+ Dwarf_Signed cnt, i, clm;
+ Dwarf_Line *lines;
+ Dwarf_Unsigned lineno = 0;
+ Dwarf_Addr addr;
+ Dwarf_Unsigned fno;
+ int ret;
+
+ ret = dwarf_srclines(pf->cu_die, &lines, &cnt, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+
+ for (i = 0; i < cnt; i++) {
+ ret = dwarf_line_srcfileno(lines[i], &fno, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ if (fno != pf->fno)
+ continue;
+
+ ret = dwarf_lineno(lines[i], &lineno, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ if (lineno != pf->lno)
+ continue;
+
+ ret = dwarf_lineoff(lines[i], &clm, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+
+ ret = dwarf_lineaddr(lines[i], &addr, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ pr_debug("Probe line found: line[%d]:%u,%d addr:0x%llx\n",
+ (int)i, (unsigned)lineno, (int)clm, addr);
+ pf->addr = addr;
+ /* Search a real subprogram including this line, */
+ ret = search_die_from_children(pf->cu_die,
+ probeaddr_callback, pf);
+ if (ret == 0)
+ die("Probe point is not found in subprograms.\n");
+ /* Continuing, because target line might be inlined. */
+ }
+ dwarf_srclines_dealloc(__dw_debug, lines, cnt);
+}
+
+/* Search function from function name */
+static int probefunc_callback(struct die_link *dlink, void *data)
+{
+ struct probe_finder *pf = (struct probe_finder *)data;
+ struct probe_point *pp = pf->pp;
+ struct die_link *lk;
+ Dwarf_Signed offs;
+ Dwarf_Half tag;
+ int ret;
+
+ ret = dwarf_tag(dlink->die, &tag, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (tag == DW_TAG_subprogram) {
+ if (die_compare_name(dlink->die, pp->function) == 0) {
+ if (pp->line) { /* Function relative line */
+ pf->fno = die_get_decl_file(dlink->die);
+ pf->lno = die_get_decl_line(dlink->die)
+ + pp->line;
+ find_by_line(pf);
+ return 1;
+ }
+ if (die_inlined_subprogram(dlink->die)) {
+ /* Inlined function, save it. */
+ ret = dwarf_die_CU_offset(dlink->die,
+ &pf->inl_offs,
+ &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+ pr_debug("inline definition offset %lld\n",
+ pf->inl_offs);
+ return 0; /* Continue to search */
+ }
+ /* Get probe address */
+ pf->addr = die_get_entrypc(dlink->die);
+ pf->addr += pp->offset;
+ /* TODO: Check the address in this function */
+ show_probepoint(dlink->die, pp->offset, pf);
+ return 1; /* Exit; no same symbol in this CU. */
+ }
+ } else if (tag == DW_TAG_inlined_subroutine && pf->inl_offs) {
+ if (die_get_abstract_origin(dlink->die) == pf->inl_offs) {
+ /* Get probe address */
+ pf->addr = die_get_entrypc(dlink->die);
+ pf->addr += pp->offset;
+ pr_debug("found inline addr: 0x%llx\n", pf->addr);
+ /* Inlined function. Get a real subprogram */
+ for (lk = dlink->parent; lk != NULL; lk = lk->parent) {
+ tag = 0;
+ dwarf_tag(lk->die, &tag, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (tag == DW_TAG_subprogram &&
+ !die_inlined_subprogram(lk->die))
+ goto found;
+ }
+ die("Failed to find real subprogram.\n");
+found:
+ /* Get offset from subprogram */
+ ret = die_within_subprogram(lk->die, pf->addr, &offs);
+ DIE_IF(!ret);
+ show_probepoint(lk->die, offs, pf);
+ /* Continue to search */
+ }
+ }
+ return 0;
+}
+
+static void find_by_func(struct probe_finder *pf)
+{
+ search_die_from_children(pf->cu_die, probefunc_callback, pf);
+}
+
+/* Find a probe point */
+int find_probepoint(int fd, struct probe_point *pp)
+{
+ Dwarf_Half addr_size = 0;
+ Dwarf_Unsigned next_cuh = 0;
+ int cu_number = 0, ret;
+ struct probe_finder pf = {.pp = pp};
+
+ ret = dwarf_init(fd, DW_DLC_READ, 0, 0, &__dw_debug, &__dw_error);
+ if (ret != DW_DLV_OK) {
+ pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO.\n");
+ return -ENOENT;
+ }
+
+ pp->found = 0;
+ while (++cu_number) {
+ /* Search CU (Compilation Unit) */
+ ret = dwarf_next_cu_header(__dw_debug, NULL, NULL, NULL,
+ &addr_size, &next_cuh, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_NO_ENTRY)
+ break;
+
+ /* Get the DIE(Debugging Information Entry) of this CU */
+ ret = dwarf_siblingof(__dw_debug, 0, &pf.cu_die, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+
+ /* Check if target file is included. */
+ if (pp->file)
+ pf.fno = cu_find_fileno(pf.cu_die, pp->file);
+
+ if (!pp->file || pf.fno) {
+ /* Save CU base address (for frame_base) */
+ ret = dwarf_lowpc(pf.cu_die, &pf.cu_base, &__dw_error);
+ DIE_IF(ret == DW_DLV_ERROR);
+ if (ret == DW_DLV_NO_ENTRY)
+ pf.cu_base = 0;
+ if (pp->function)
+ find_by_func(&pf);
+ else {
+ pf.lno = pp->line;
+ find_by_line(&pf);
+ }
+ }
+ dwarf_dealloc(__dw_debug, pf.cu_die, DW_DLA_DIE);
+ }
+ ret = dwarf_finish(__dw_debug, &__dw_error);
+ DIE_IF(ret != DW_DLV_OK);
+
+ return pp->found;
+}
+
diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h
new file mode 100644
index 000000000000..bdebca6697d2
--- /dev/null
+++ b/tools/perf/util/probe-finder.h
@@ -0,0 +1,57 @@
+#ifndef _PROBE_FINDER_H
+#define _PROBE_FINDER_H
+
+#define MAX_PATH_LEN 256
+#define MAX_PROBE_BUFFER 1024
+#define MAX_PROBES 128
+
+static inline int is_c_varname(const char *name)
+{
+ /* TODO */
+ return isalpha(name[0]) || name[0] == '_';
+}
+
+struct probe_point {
+ /* Inputs */
+ char *file; /* File name */
+ int line; /* Line number */
+
+ char *function; /* Function name */
+ int offset; /* Offset bytes */
+
+ int nr_args; /* Number of arguments */
+ char **args; /* Arguments */
+
+ int retprobe; /* Return probe */
+
+ /* Output */
+ int found; /* Number of found probe points */
+ char *probes[MAX_PROBES]; /* Output buffers (will be allocated)*/
+};
+
+#ifndef NO_LIBDWARF
+extern int find_probepoint(int fd, struct probe_point *pp);
+
+#include <libdwarf/dwarf.h>
+#include <libdwarf/libdwarf.h>
+
+struct probe_finder {
+ struct probe_point *pp; /* Target probe point */
+
+ /* For function searching */
+ Dwarf_Addr addr; /* Address */
+ Dwarf_Unsigned fno; /* File number */
+ Dwarf_Unsigned lno; /* Line number */
+ Dwarf_Off inl_offs; /* Inline offset */
+ Dwarf_Die cu_die; /* Current CU */
+
+ /* For variable searching */
+ Dwarf_Addr cu_base; /* Current CU base address */
+ Dwarf_Locdesc fbloc; /* Location of Current Frame Base */
+ const char *var; /* Current variable name */
+ char *buf; /* Current output buffer */
+ int len; /* Length of output buffer */
+};
+#endif /* NO_LIBDWARF */
+
+#endif /*_PROBE_FINDER_H */
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 40c9acd41cad..b490354d1b23 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -5,8 +5,10 @@ char default_parent_pattern[] = "^sys_|^do_page_fault";
char *parent_pattern = default_parent_pattern;
char default_sort_order[] = "comm,dso,symbol";
char *sort_order = default_sort_order;
-int sort__need_collapse = 0;
-int sort__has_parent = 0;
+int sort__need_collapse = 0;
+int sort__has_parent = 0;
+
+enum sort_type sort__first_dimension;
unsigned int dsos__col_width;
unsigned int comms__col_width;
@@ -265,6 +267,19 @@ int sort_dimension__add(const char *tok)
sort__has_parent = 1;
}
+ if (list_empty(&hist_entry__sort_list)) {
+ if (!strcmp(sd->name, "pid"))
+ sort__first_dimension = SORT_PID;
+ else if (!strcmp(sd->name, "comm"))
+ sort__first_dimension = SORT_COMM;
+ else if (!strcmp(sd->name, "dso"))
+ sort__first_dimension = SORT_DSO;
+ else if (!strcmp(sd->name, "symbol"))
+ sort__first_dimension = SORT_SYM;
+ else if (!strcmp(sd->name, "parent"))
+ sort__first_dimension = SORT_PARENT;
+ }
+
list_add_tail(&sd->entry->list, &hist_entry__sort_list);
sd->taken = 1;
@@ -273,4 +288,3 @@ int sort_dimension__add(const char *tok)
return -ESRCH;
}
-
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index 13806d782af6..333e664ff45f 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -39,6 +39,7 @@ extern struct sort_entry sort_parent;
extern unsigned int dsos__col_width;
extern unsigned int comms__col_width;
extern unsigned int threads__col_width;
+extern enum sort_type sort__first_dimension;
struct hist_entry {
struct rb_node rb_node;
@@ -53,6 +54,14 @@ struct hist_entry {
struct rb_root sorted_chain;
};
+enum sort_type {
+ SORT_PID,
+ SORT_COMM,
+ SORT_DSO,
+ SORT_SYM,
+ SORT_PARENT
+};
+
/*
* configurable sorting bits
*/
diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c
index 04743d3e9039..227043577e06 100644
--- a/tools/perf/util/string.c
+++ b/tools/perf/util/string.c
@@ -1,5 +1,7 @@
#include <string.h>
+#include <stdlib.h>
#include "string.h"
+#include "util.h"
static int hex(char ch)
{
@@ -43,3 +45,85 @@ char *strxfrchar(char *s, char from, char to)
return s;
}
+
+#define K 1024LL
+/*
+ * perf_atoll()
+ * Parse (\d+)(b|B|kb|KB|mb|MB|gb|GB|tb|TB) (e.g. "256MB")
+ * and return its numeric value
+ */
+s64 perf_atoll(const char *str)
+{
+ unsigned int i;
+ s64 length = -1, unit = 1;
+
+ if (!isdigit(str[0]))
+ goto out_err;
+
+ for (i = 1; i < strlen(str); i++) {
+ switch (str[i]) {
+ case 'B':
+ case 'b':
+ break;
+ case 'K':
+ if (str[i + 1] != 'B')
+ goto out_err;
+ else
+ goto kilo;
+ case 'k':
+ if (str[i + 1] != 'b')
+ goto out_err;
+kilo:
+ unit = K;
+ break;
+ case 'M':
+ if (str[i + 1] != 'B')
+ goto out_err;
+ else
+ goto mega;
+ case 'm':
+ if (str[i + 1] != 'b')
+ goto out_err;
+mega:
+ unit = K * K;
+ break;
+ case 'G':
+ if (str[i + 1] != 'B')
+ goto out_err;
+ else
+ goto giga;
+ case 'g':
+ if (str[i + 1] != 'b')
+ goto out_err;
+giga:
+ unit = K * K * K;
+ break;
+ case 'T':
+ if (str[i + 1] != 'B')
+ goto out_err;
+ else
+ goto tera;
+ case 't':
+ if (str[i + 1] != 'b')
+ goto out_err;
+tera:
+ unit = K * K * K * K;
+ break;
+ case '\0': /* only specified figures */
+ unit = 1;
+ break;
+ default:
+ if (!isdigit(str[i]))
+ goto out_err;
+ break;
+ }
+ }
+
+ length = atoll(str) * unit;
+ goto out;
+
+out_err:
+ length = -1;
+out:
+ return length;
+}
diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h
index 2c84bf65ba0f..e50b07f80827 100644
--- a/tools/perf/util/string.h
+++ b/tools/perf/util/string.h
@@ -5,6 +5,7 @@
int hex2u64(const char *ptr, u64 *val);
char *strxfrchar(char *s, char from, char to);
+s64 perf_atoll(const char *str);
#define _STR(x) #x
#define STR(x) _STR(x)
diff --git a/tools/perf/util/svghelper.c b/tools/perf/util/svghelper.c
index 856655d8b0b8..b3637db025a2 100644
--- a/tools/perf/util/svghelper.c
+++ b/tools/perf/util/svghelper.c
@@ -103,7 +103,7 @@ void open_svg(const char *filename, int cpus, int rows, u64 start, u64 end)
fprintf(svgfile, " rect.process2 { fill:rgb(180,180,180); fill-opacity:0.9; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
fprintf(svgfile, " rect.sample { fill:rgb( 0, 0,255); fill-opacity:0.8; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
fprintf(svgfile, " rect.blocked { fill:rgb(255, 0, 0); fill-opacity:0.5; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
- fprintf(svgfile, " rect.waiting { fill:rgb(214,214, 0); fill-opacity:0.3; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
+ fprintf(svgfile, " rect.waiting { fill:rgb(224,214, 0); fill-opacity:0.8; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
fprintf(svgfile, " rect.WAITING { fill:rgb(255,214, 48); fill-opacity:0.6; stroke-width:0; stroke:rgb( 0, 0, 0); } \n");
fprintf(svgfile, " rect.cpu { fill:rgb(192,192,192); fill-opacity:0.2; stroke-width:0.5; stroke:rgb(128,128,128); } \n");
fprintf(svgfile, " rect.pstate { fill:rgb(128,128,128); fill-opacity:0.8; stroke-width:0; } \n");
diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index faa84f5d4f54..4d75e745288f 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -9,10 +9,9 @@
#include <libelf.h>
#include <gelf.h>
#include <elf.h>
+#include <limits.h>
#include <sys/utsname.h>
-const char *sym_hist_filter;
-
enum dso_origin {
DSO__ORIG_KERNEL = 0,
DSO__ORIG_JAVA_JIT,
@@ -28,6 +27,7 @@ static void dsos__add(struct dso *dso);
static struct dso *dsos__find(const char *name);
static struct map *map__new2(u64 start, struct dso *dso);
static void kernel_maps__insert(struct map *map);
+unsigned int symbol__priv_size;
static struct rb_root kernel_maps;
@@ -77,39 +77,31 @@ static void kernel_maps__fixup_end(void)
}
}
-static struct symbol *symbol__new(u64 start, u64 len, const char *name,
- unsigned int priv_size, int v)
+static struct symbol *symbol__new(u64 start, u64 len, const char *name)
{
size_t namelen = strlen(name) + 1;
- struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen);
-
+ struct symbol *self = calloc(1, (symbol__priv_size +
+ sizeof(*self) + namelen));
if (!self)
return NULL;
- if (v > 2)
- printf("new symbol: %016Lx [%08lx]: %s, hist: %p\n",
- start, (unsigned long)len, name, self->hist);
-
- self->hist = NULL;
- self->hist_sum = 0;
-
- if (sym_hist_filter && !strcmp(name, sym_hist_filter))
- self->hist = calloc(sizeof(u64), len);
-
- if (priv_size) {
- memset(self, 0, priv_size);
- self = ((void *)self) + priv_size;
+ if (symbol__priv_size) {
+ memset(self, 0, symbol__priv_size);
+ self = ((void *)self) + symbol__priv_size;
}
self->start = start;
self->end = len ? start + len - 1 : start;
+
+ pr_debug3("%s: %s %#Lx-%#Lx\n", __func__, name, start, self->end);
+
memcpy(self->name, name, namelen);
return self;
}
-static void symbol__delete(struct symbol *self, unsigned int priv_size)
+static void symbol__delete(struct symbol *self)
{
- free(((void *)self) - priv_size);
+ free(((void *)self) - symbol__priv_size);
}
static size_t symbol__fprintf(struct symbol *self, FILE *fp)
@@ -118,19 +110,31 @@ static size_t symbol__fprintf(struct symbol *self, FILE *fp)
self->start, self->end, self->name);
}
-struct dso *dso__new(const char *name, unsigned int sym_priv_size)
+static void dso__set_long_name(struct dso *self, char *name)
+{
+ self->long_name = name;
+ self->long_name_len = strlen(name);
+}
+
+static void dso__set_basename(struct dso *self)
+{
+ self->short_name = basename(self->long_name);
+}
+
+struct dso *dso__new(const char *name)
{
struct dso *self = malloc(sizeof(*self) + strlen(name) + 1);
if (self != NULL) {
strcpy(self->name, name);
- self->long_name = self->name;
+ dso__set_long_name(self, self->name);
self->short_name = self->name;
self->syms = RB_ROOT;
- self->sym_priv_size = sym_priv_size;
self->find_symbol = dso__find_symbol;
self->slen_calculated = 0;
self->origin = DSO__ORIG_NOT_FOUND;
+ self->loaded = 0;
+ self->has_build_id = 0;
}
return self;
@@ -145,7 +149,7 @@ static void dso__delete_symbols(struct dso *self)
pos = rb_entry(next, struct symbol, rb_node);
next = rb_next(&pos->rb_node);
rb_erase(&pos->rb_node, &self->syms);
- symbol__delete(pos, self->sym_priv_size);
+ symbol__delete(pos);
}
}
@@ -157,6 +161,12 @@ void dso__delete(struct dso *self)
free(self);
}
+void dso__set_build_id(struct dso *self, void *build_id)
+{
+ memcpy(self->build_id, build_id, sizeof(self->build_id));
+ self->has_build_id = 1;
+}
+
static void dso__insert_symbol(struct dso *self, struct symbol *sym)
{
struct rb_node **p = &self->syms.rb_node;
@@ -199,11 +209,37 @@ struct symbol *dso__find_symbol(struct dso *self, u64 ip)
return NULL;
}
-size_t dso__fprintf(struct dso *self, FILE *fp)
+int build_id__sprintf(u8 *self, int len, char *bf)
{
- size_t ret = fprintf(fp, "dso: %s\n", self->short_name);
+ char *bid = bf;
+ u8 *raw = self;
+ int i;
+
+ for (i = 0; i < len; ++i) {
+ sprintf(bid, "%02x", *raw);
+ ++raw;
+ bid += 2;
+ }
+ return raw - self;
+}
+
+size_t dso__fprintf_buildid(struct dso *self, FILE *fp)
+{
+ char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+ build_id__sprintf(self->build_id, sizeof(self->build_id), sbuild_id);
+ return fprintf(fp, "%s", sbuild_id);
+}
+
+size_t dso__fprintf(struct dso *self, FILE *fp)
+{
struct rb_node *nd;
+ size_t ret = fprintf(fp, "dso: %s (", self->short_name);
+
+ ret += dso__fprintf_buildid(self, fp);
+ ret += fprintf(fp, ")\n");
+
for (nd = rb_first(&self->syms); nd; nd = rb_next(nd)) {
struct symbol *pos = rb_entry(nd, struct symbol, rb_node);
ret += symbol__fprintf(pos, fp);
@@ -217,7 +253,7 @@ size_t dso__fprintf(struct dso *self, FILE *fp)
* so that we can in the next step set the symbol ->end address and then
* call kernel_maps__split_kallsyms.
*/
-static int kernel_maps__load_all_kallsyms(int v)
+static int kernel_maps__load_all_kallsyms(void)
{
char *line = NULL;
size_t n;
@@ -259,12 +295,16 @@ static int kernel_maps__load_all_kallsyms(int v)
/*
* Will fix up the end later, when we have all symbols sorted.
*/
- sym = symbol__new(start, 0, symbol_name,
- kernel_map->dso->sym_priv_size, v);
+ sym = symbol__new(start, 0, symbol_name);
if (sym == NULL)
goto out_delete_line;
+ /*
+ * We will pass the symbols to the filter later, in
+ * kernel_maps__split_kallsyms, when we have split the
+ * maps per module
+ */
dso__insert_symbol(kernel_map->dso, sym);
}
@@ -308,8 +348,8 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
if (strcmp(map->dso->name, module)) {
map = kernel_maps__find_by_dso_name(module);
if (!map) {
- fputs("/proc/{kallsyms,modules} "
- "inconsistency!\n", stderr);
+ pr_err("/proc/{kallsyms,modules} "
+ "inconsistency!\n");
return -1;
}
}
@@ -326,8 +366,7 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
snprintf(dso_name, sizeof(dso_name), "[kernel].%d",
kernel_range++);
- dso = dso__new(dso_name,
- kernel_map->dso->sym_priv_size);
+ dso = dso__new(dso_name);
if (dso == NULL)
return -1;
@@ -337,7 +376,7 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
return -1;
}
- map->map_ip = vdso__map_ip;
+ map->map_ip = map->unmap_ip = identity__map_ip;
kernel_maps__insert(map);
++kernel_range;
}
@@ -345,7 +384,7 @@ static int kernel_maps__split_kallsyms(symbol_filter_t filter, int use_modules)
if (filter && filter(map, pos)) {
delete_symbol:
rb_erase(&pos->rb_node, &kernel_map->dso->syms);
- symbol__delete(pos, kernel_map->dso->sym_priv_size);
+ symbol__delete(pos);
} else {
if (map != kernel_map) {
rb_erase(&pos->rb_node, &kernel_map->dso->syms);
@@ -359,10 +398,9 @@ delete_symbol:
}
-static int kernel_maps__load_kallsyms(symbol_filter_t filter,
- int use_modules, int v)
+static int kernel_maps__load_kallsyms(symbol_filter_t filter, int use_modules)
{
- if (kernel_maps__load_all_kallsyms(v))
+ if (kernel_maps__load_all_kallsyms())
return -1;
dso__fixup_sym_end(kernel_map->dso);
@@ -370,9 +408,9 @@ static int kernel_maps__load_kallsyms(symbol_filter_t filter,
return kernel_maps__split_kallsyms(filter, use_modules);
}
-static size_t kernel_maps__fprintf(FILE *fp, int v)
+static size_t kernel_maps__fprintf(FILE *fp)
{
- size_t printed = fprintf(stderr, "Kernel maps:\n");
+ size_t printed = fprintf(fp, "Kernel maps:\n");
struct rb_node *nd;
for (nd = rb_first(&kernel_maps); nd; nd = rb_next(nd)) {
@@ -380,17 +418,17 @@ static size_t kernel_maps__fprintf(FILE *fp, int v)
printed += fprintf(fp, "Map:");
printed += map__fprintf(pos, fp);
- if (v > 1) {
+ if (verbose > 1) {
printed += dso__fprintf(pos->dso, fp);
printed += fprintf(fp, "--\n");
}
}
- return printed + fprintf(stderr, "END kernel maps\n");
+ return printed + fprintf(fp, "END kernel maps\n");
}
static int dso__load_perf_map(struct dso *self, struct map *map,
- symbol_filter_t filter, int v)
+ symbol_filter_t filter)
{
char *line = NULL;
size_t n;
@@ -427,14 +465,13 @@ 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,
- self->sym_priv_size, v);
+ sym = symbol__new(start, size, line + len);
if (sym == NULL)
goto out_delete_line;
if (filter && filter(map, sym))
- symbol__delete(sym, self->sym_priv_size);
+ symbol__delete(sym);
else {
dso__insert_symbol(self, sym);
nr_syms++;
@@ -542,7 +579,8 @@ static Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep,
* And always look at the original dso, not at debuginfo packages, that
* have the PLT data stripped out (shdr_rel_plt.sh_type == SHT_NOBITS).
*/
-static int dso__synthesize_plt_symbols(struct dso *self, int v)
+static int dso__synthesize_plt_symbols(struct dso *self, struct map *map,
+ symbol_filter_t filter)
{
uint32_t nr_rel_entries, idx;
GElf_Sym sym;
@@ -562,7 +600,7 @@ static int dso__synthesize_plt_symbols(struct dso *self, int v)
if (fd < 0)
goto out;
- elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+ elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
if (elf == NULL)
goto out_close;
@@ -626,12 +664,16 @@ static int dso__synthesize_plt_symbols(struct dso *self, int v)
"%s@plt", elf_sym__name(&sym, symstrs));
f = symbol__new(plt_offset, shdr_plt.sh_entsize,
- sympltname, self->sym_priv_size, v);
+ sympltname);
if (!f)
goto out_elf_end;
- dso__insert_symbol(self, f);
- ++nr;
+ if (filter && filter(map, f))
+ symbol__delete(f);
+ else {
+ dso__insert_symbol(self, f);
+ ++nr;
+ }
}
} else if (shdr_rel_plt.sh_type == SHT_REL) {
GElf_Rel pos_mem, *pos;
@@ -644,12 +686,16 @@ static int dso__synthesize_plt_symbols(struct dso *self, int v)
"%s@plt", elf_sym__name(&sym, symstrs));
f = symbol__new(plt_offset, shdr_plt.sh_entsize,
- sympltname, self->sym_priv_size, v);
+ sympltname);
if (!f)
goto out_elf_end;
- dso__insert_symbol(self, f);
- ++nr;
+ if (filter && filter(map, f))
+ symbol__delete(f);
+ else {
+ dso__insert_symbol(self, f);
+ ++nr;
+ }
}
}
@@ -662,14 +708,14 @@ out_close:
if (err == 0)
return nr;
out:
- fprintf(stderr, "%s: problems reading %s PLT info.\n",
- __func__, self->long_name);
+ pr_warning("%s: problems reading %s PLT info.\n",
+ __func__, self->long_name);
return 0;
}
static int dso__load_sym(struct dso *self, struct map *map, const char *name,
int fd, symbol_filter_t filter, int kernel,
- int kmodule, int v)
+ int kmodule)
{
struct map *curr_map = map;
struct dso *curr_dso = self;
@@ -686,17 +732,14 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
Elf *elf;
int nr = 0;
- elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+ elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
if (elf == NULL) {
- if (v)
- fprintf(stderr, "%s: cannot read %s ELF file.\n",
- __func__, name);
+ pr_err("%s: cannot read %s ELF file.\n", __func__, name);
goto out_close;
}
if (gelf_getehdr(elf, &ehdr) == NULL) {
- if (v)
- fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+ pr_err("%s: cannot get elf header.\n", __func__);
goto out_elf_end;
}
@@ -782,7 +825,7 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
if (kmodule)
start += map->start + shdr.sh_offset;
- curr_dso = dso__new(dso_name, self->sym_priv_size);
+ curr_dso = dso__new(dso_name);
if (curr_dso == NULL)
goto out_elf_end;
curr_map = map__new2(start, curr_dso);
@@ -790,7 +833,8 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
dso__delete(curr_dso);
goto out_elf_end;
}
- curr_map->map_ip = vdso__map_ip;
+ curr_map->map_ip = identity__map_ip;
+ curr_map->unmap_ip = identity__map_ip;
curr_dso->origin = DSO__ORIG_KERNEL;
kernel_maps__insert(curr_map);
dsos__add(curr_dso);
@@ -801,10 +845,9 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name,
}
if (curr_dso->adjust_symbols) {
- if (v > 2)
- printf("adjusting symbol: st_value: %Lx sh_addr: %Lx sh_offset: %Lx\n",
- (u64)sym.st_value, (u64)shdr.sh_addr, (u64)shdr.sh_offset);
-
+ pr_debug2("adjusting symbol: st_value: %Lx sh_addr: "
+ "%Lx sh_offset: %Lx\n", (u64)sym.st_value,
+ (u64)shdr.sh_addr, (u64)shdr.sh_offset);
sym.st_value -= shdr.sh_addr - shdr.sh_offset;
}
/*
@@ -816,14 +859,13 @@ 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,
- curr_dso->sym_priv_size, v);
+ f = symbol__new(sym.st_value, sym.st_size, elf_name);
free(demangled);
if (!f)
goto out_elf_end;
if (filter && filter(curr_map, f))
- symbol__delete(f, curr_dso->sym_priv_size);
+ symbol__delete(f);
else {
dso__insert_symbol(curr_dso, f);
nr++;
@@ -842,63 +884,108 @@ out_close:
return err;
}
-#define BUILD_ID_SIZE 128
+bool dsos__read_build_ids(void)
+{
+ bool have_build_id = false;
+ struct dso *pos;
+
+ list_for_each_entry(pos, &dsos, node)
+ if (filename__read_build_id(pos->long_name, pos->build_id,
+ sizeof(pos->build_id)) > 0) {
+ have_build_id = true;
+ pos->has_build_id = true;
+ }
+
+ return have_build_id;
+}
-static char *dso__read_build_id(struct dso *self, int v)
+int filename__read_build_id(const char *filename, void *bf, size_t size)
{
- int i;
+ int fd, err = -1;
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->long_name, O_RDONLY);
+ if (size < BUILD_ID_SIZE)
+ goto out;
+
+ fd = open(filename, O_RDONLY);
if (fd < 0)
goto out;
- elf = elf_begin(fd, ELF_C_READ_MMAP, NULL);
+ elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
if (elf == NULL) {
- if (v)
- fprintf(stderr, "%s: cannot read %s ELF file.\n",
- __func__, self->long_name);
+ pr_debug2("%s: cannot read %s ELF file.\n", __func__, filename);
goto out_close;
}
if (gelf_getehdr(elf, &ehdr) == NULL) {
- if (v)
- fprintf(stderr, "%s: cannot get elf header.\n", __func__);
+ pr_err("%s: cannot get elf header.\n", __func__);
goto out_elf_end;
}
- sec = elf_section_by_name(elf, &ehdr, &shdr, ".note.gnu.build-id", NULL);
+ 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 (v >= 2)
- printf("%s(%s): %s\n", __func__, self->long_name, build_id);
+ memcpy(bf, build_id_data->d_buf + 16, BUILD_ID_SIZE);
+ err = BUILD_ID_SIZE;
out_elf_end:
elf_end(elf);
out_close:
close(fd);
out:
- return build_id;
+ return err;
+}
+
+int sysfs__read_build_id(const char *filename, void *build_id, size_t size)
+{
+ int fd, err = -1;
+
+ if (size < BUILD_ID_SIZE)
+ goto out;
+
+ fd = open(filename, O_RDONLY);
+ if (fd < 0)
+ goto out;
+
+ while (1) {
+ char bf[BUFSIZ];
+ GElf_Nhdr nhdr;
+ int namesz, descsz;
+
+ if (read(fd, &nhdr, sizeof(nhdr)) != sizeof(nhdr))
+ break;
+
+ namesz = (nhdr.n_namesz + 3) & -4U;
+ descsz = (nhdr.n_descsz + 3) & -4U;
+ if (nhdr.n_type == NT_GNU_BUILD_ID &&
+ nhdr.n_namesz == sizeof("GNU")) {
+ if (read(fd, bf, namesz) != namesz)
+ break;
+ if (memcmp(bf, "GNU", sizeof("GNU")) == 0) {
+ if (read(fd, build_id,
+ BUILD_ID_SIZE) == BUILD_ID_SIZE) {
+ err = 0;
+ break;
+ }
+ } else if (read(fd, bf, descsz) != descsz)
+ break;
+ } else {
+ int n = namesz + descsz;
+ if (read(fd, bf, n) != n)
+ break;
+ }
+ }
+ close(fd);
+out:
+ return err;
}
char dso__symtab_origin(const struct dso *self)
@@ -918,20 +1005,23 @@ char dso__symtab_origin(const struct dso *self)
return origin[self->origin];
}
-int dso__load(struct dso *self, struct map *map, symbol_filter_t filter, int v)
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
{
int size = PATH_MAX;
- char *name = malloc(size), *build_id = NULL;
+ char *name = malloc(size);
+ u8 build_id[BUILD_ID_SIZE];
int ret = -1;
int fd;
+ self->loaded = 1;
+
if (!name)
return -1;
self->adjust_symbols = 0;
if (strncmp(self->name, "/tmp/perf-", 10) == 0) {
- ret = dso__load_perf_map(self, map, filter, v);
+ ret = dso__load_perf_map(self, map, filter);
self->origin = ret > 0 ? DSO__ORIG_JAVA_JIT :
DSO__ORIG_NOT_FOUND;
return ret;
@@ -952,12 +1042,17 @@ more:
self->long_name);
break;
case DSO__ORIG_BUILDID:
- build_id = dso__read_build_id(self, v);
- if (build_id != NULL) {
+ if (filename__read_build_id(self->long_name, build_id,
+ sizeof(build_id))) {
+ char build_id_hex[BUILD_ID_SIZE * 2 + 1];
+
+ build_id__sprintf(build_id, sizeof(build_id),
+ build_id_hex);
snprintf(name, size,
"/usr/lib/debug/.build-id/%.2s/%s.debug",
- build_id, build_id + 2);
- free(build_id);
+ build_id_hex, build_id_hex + 2);
+ if (self->has_build_id)
+ goto compare_build_id;
break;
}
self->origin++;
@@ -970,10 +1065,20 @@ more:
goto out;
}
+ if (self->has_build_id) {
+ if (filename__read_build_id(name, build_id,
+ sizeof(build_id)) < 0)
+ goto more;
+compare_build_id:
+ if (memcmp(build_id, self->build_id,
+ sizeof(self->build_id)) != 0)
+ goto more;
+ }
+
fd = open(name, O_RDONLY);
} while (fd < 0);
- ret = dso__load_sym(self, map, name, fd, filter, 0, 0, v);
+ ret = dso__load_sym(self, map, name, fd, filter, 0, 0);
close(fd);
/*
@@ -983,7 +1088,7 @@ more:
goto more;
if (ret > 0) {
- int nr_plt = dso__synthesize_plt_symbols(self, v);
+ int nr_plt = dso__synthesize_plt_symbols(self, map, filter);
if (nr_plt > 0)
ret += nr_plt;
}
@@ -1031,34 +1136,31 @@ struct map *kernel_maps__find_by_dso_name(const char *name)
}
static int dso__load_module_sym(struct dso *self, struct map *map,
- symbol_filter_t filter, int v)
+ symbol_filter_t filter)
{
int err = 0, fd = open(self->long_name, O_RDONLY);
+ self->loaded = 1;
+
if (fd < 0) {
- if (v)
- fprintf(stderr, "%s: cannot open %s\n",
- __func__, self->long_name);
+ pr_err("%s: cannot open %s\n", __func__, self->long_name);
return err;
}
- err = dso__load_sym(self, map, self->long_name, fd, filter, 0, 1, v);
+ err = dso__load_sym(self, map, self->long_name, fd, filter, 0, 1);
close(fd);
return err;
}
-static int dsos__load_modules_sym_dir(char *dirname,
- symbol_filter_t filter, int v)
+static int dsos__load_modules_sym_dir(char *dirname, symbol_filter_t filter)
{
struct dirent *dent;
int nr_symbols = 0, err;
DIR *dir = opendir(dirname);
if (!dir) {
- if (v)
- fprintf(stderr, "%s: cannot open %s dir\n", __func__,
- dirname);
+ pr_err("%s: cannot open %s dir\n", __func__, dirname);
return -1;
}
@@ -1072,7 +1174,7 @@ static int dsos__load_modules_sym_dir(char *dirname,
snprintf(path, sizeof(path), "%s/%s",
dirname, dent->d_name);
- err = dsos__load_modules_sym_dir(path, filter, v);
+ err = dsos__load_modules_sym_dir(path, filter);
if (err < 0)
goto failure;
} else {
@@ -1080,6 +1182,7 @@ static int dsos__load_modules_sym_dir(char *dirname,
dso_name[PATH_MAX];
struct map *map;
struct rb_node *last;
+ char *long_name;
if (dot == NULL || strcmp(dot, ".ko"))
continue;
@@ -1094,11 +1197,13 @@ static int dsos__load_modules_sym_dir(char *dirname,
snprintf(path, sizeof(path), "%s/%s",
dirname, dent->d_name);
- map->dso->long_name = strdup(path);
- if (map->dso->long_name == NULL)
+ long_name = strdup(path);
+ if (long_name == NULL)
goto failure;
+ dso__set_long_name(map->dso, long_name);
+ dso__set_basename(map->dso);
- err = dso__load_module_sym(map->dso, map, filter, v);
+ err = dso__load_module_sym(map->dso, map, filter);
if (err < 0)
goto failure;
last = rb_last(&map->dso->syms);
@@ -1125,7 +1230,7 @@ failure:
return -1;
}
-static int dsos__load_modules_sym(symbol_filter_t filter, int v)
+static int dsos__load_modules_sym(symbol_filter_t filter)
{
struct utsname uts;
char modules_path[PATH_MAX];
@@ -1136,7 +1241,7 @@ static int dsos__load_modules_sym(symbol_filter_t filter, int v)
snprintf(modules_path, sizeof(modules_path), "/lib/modules/%s/kernel",
uts.release);
- return dsos__load_modules_sym_dir(modules_path, filter, v);
+ return dsos__load_modules_sym_dir(modules_path, filter);
}
/*
@@ -1149,21 +1254,16 @@ static struct map *map__new2(u64 start, struct dso *dso)
struct map *self = malloc(sizeof(*self));
if (self != NULL) {
- self->start = start;
/*
- * Will be filled after we load all the symbols
+ * ->end will be filled after we load all the symbols
*/
- self->end = 0;
-
- self->pgoff = 0;
- self->dso = dso;
- self->map_ip = map__map_ip;
- RB_CLEAR_NODE(&self->rb_node);
+ map__init(self, start, 0, 0, dso);
}
+
return self;
}
-static int dsos__load_modules(unsigned int sym_priv_size)
+int dsos__load_modules(void)
{
char *line = NULL;
size_t n;
@@ -1202,7 +1302,7 @@ static int dsos__load_modules(unsigned int sym_priv_size)
*sep = '\0';
snprintf(name, sizeof(name), "[%s]", line);
- dso = dso__new(name, sym_priv_size);
+ dso = dso__new(name);
if (dso == NULL)
goto out_delete_line;
@@ -1213,6 +1313,12 @@ static int dsos__load_modules(unsigned int sym_priv_size)
goto out_delete_line;
}
+ snprintf(name, sizeof(name),
+ "/sys/module/%s/notes/.note.gnu.build-id", line);
+ if (sysfs__read_build_id(name, dso->build_id,
+ sizeof(dso->build_id)) == 0)
+ dso->has_build_id = true;
+
dso->origin = DSO__ORIG_KMODULE;
kernel_maps__insert(map);
dsos__add(dso);
@@ -1230,85 +1336,77 @@ out_failure:
}
static int dso__load_vmlinux(struct dso *self, struct map *map,
- const char *vmlinux,
- symbol_filter_t filter, int v)
+ const char *vmlinux, symbol_filter_t filter)
{
int err, fd = open(vmlinux, O_RDONLY);
+ self->loaded = 1;
+
if (fd < 0)
return -1;
- err = dso__load_sym(self, map, self->long_name, fd, filter, 1, 0, v);
+ err = dso__load_sym(self, map, self->long_name, fd, filter, 1, 0);
close(fd);
return err;
}
-int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
- symbol_filter_t filter, int v, int use_modules)
+int dso__load_kernel_sym(struct dso *self, symbol_filter_t filter, int use_modules)
{
int err = -1;
- struct dso *dso = dso__new(vmlinux, sym_priv_size);
- if (dso == NULL)
- return -1;
-
- dso->short_name = "[kernel]";
- kernel_map = map__new2(0, dso);
+ kernel_map = map__new2(0, self);
if (kernel_map == NULL)
goto out_delete_dso;
- kernel_map->map_ip = vdso__map_ip;
+ kernel_map->map_ip = kernel_map->unmap_ip = identity__map_ip;
- if (use_modules && dsos__load_modules(sym_priv_size) < 0) {
- fprintf(stderr, "Failed to load list of modules in use! "
- "Continuing...\n");
+ if (use_modules && dsos__load_modules() < 0) {
+ pr_warning("Failed to load list of modules in use! "
+ "Continuing...\n");
use_modules = 0;
}
- if (vmlinux) {
- err = dso__load_vmlinux(dso, kernel_map, vmlinux, filter, v);
- if (err > 0 && use_modules) {
- int syms = dsos__load_modules_sym(filter, v);
+ err = dso__load_vmlinux(self, kernel_map, self->name, filter);
+ if (err > 0 && use_modules) {
+ int syms = dsos__load_modules_sym(filter);
- if (syms < 0)
- fprintf(stderr, "Failed to read module symbols!"
- " Continuing...\n");
- else
- err += syms;
- }
+ if (syms < 0)
+ pr_warning("Failed to read module symbols!"
+ " Continuing...\n");
+ else
+ err += syms;
}
if (err <= 0)
- err = kernel_maps__load_kallsyms(filter, use_modules, v);
+ err = kernel_maps__load_kallsyms(filter, use_modules);
if (err > 0) {
- struct rb_node *node = rb_first(&dso->syms);
+ struct rb_node *node = rb_first(&self->syms);
struct symbol *sym = rb_entry(node, struct symbol, rb_node);
kernel_map->start = sym->start;
- node = rb_last(&dso->syms);
+ node = rb_last(&self->syms);
sym = rb_entry(node, struct symbol, rb_node);
kernel_map->end = sym->end;
- dso->origin = DSO__ORIG_KERNEL;
+ self->origin = DSO__ORIG_KERNEL;
kernel_maps__insert(kernel_map);
/*
* Now that we have all sorted out, just set the ->end of all
* maps:
*/
kernel_maps__fixup_end();
- dsos__add(dso);
- if (v > 0)
- kernel_maps__fprintf(stderr, v);
+ if (verbose)
+ kernel_maps__fprintf(stderr);
}
return err;
out_delete_dso:
- dso__delete(dso);
+ dso__delete(self);
return -1;
}
@@ -1336,30 +1434,16 @@ static struct dso *dsos__find(const char *name)
struct dso *dsos__findnew(const char *name)
{
struct dso *dso = dsos__find(name);
- int nr;
-
- if (dso)
- return dso;
-
- dso = dso__new(name, 0);
- if (!dso)
- goto out_delete_dso;
- nr = dso__load(dso, NULL, NULL, verbose);
- if (nr < 0) {
- eprintf("Failed to open: %s\n", name);
- goto out_delete_dso;
+ if (!dso) {
+ dso = dso__new(name);
+ if (dso != NULL) {
+ dsos__add(dso);
+ dso__set_basename(dso);
+ }
}
- if (!nr)
- eprintf("No symbols found in: %s, maybe install a debug package?\n", name);
-
- dsos__add(dso);
return dso;
-
-out_delete_dso:
- dso__delete(dso);
- return NULL;
}
void dsos__fprintf(FILE *fp)
@@ -1370,21 +1454,52 @@ void dsos__fprintf(FILE *fp)
dso__fprintf(pos, fp);
}
-int load_kernel(void)
+size_t dsos__fprintf_buildid(FILE *fp)
{
- if (dsos__load_kernel(vmlinux_name, 0, NULL, verbose, modules) <= 0)
- return -1;
+ struct dso *pos;
+ size_t ret = 0;
+
+ list_for_each_entry(pos, &dsos, node) {
+ ret += dso__fprintf_buildid(pos, fp);
+ ret += fprintf(fp, " %s\n", pos->long_name);
+ }
+ return ret;
+}
+
+struct dso *dsos__load_kernel(void)
+{
+ struct dso *kernel = dso__new(vmlinux_name);
+
+ if (kernel == NULL)
+ return NULL;
- vdso = dso__new("[vdso]", 0);
+ kernel->short_name = "[kernel]";
+ vdso = dso__new("[vdso]");
if (!vdso)
- return -1;
+ return NULL;
+
+ if (sysfs__read_build_id("/sys/kernel/notes", kernel->build_id,
+ sizeof(kernel->build_id)) == 0)
+ kernel->has_build_id = true;
+ dsos__add(kernel);
dsos__add(vdso);
- return 0;
+ return kernel;
+}
+
+int load_kernel(symbol_filter_t filter)
+{
+ struct dso *kernel = dsos__load_kernel();
+
+ if (kernel == NULL)
+ return -1;
+
+ return dso__load_kernel_sym(kernel, filter, modules);
}
-void symbol__init(void)
+void symbol__init(unsigned int priv_size)
{
elf_version(EV_CURRENT);
+ symbol__priv_size = priv_size;
}
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index 2e4522edeb07..f0593a649c3d 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -2,6 +2,7 @@
#define __PERF_SYMBOL 1
#include <linux/types.h>
+#include <stdbool.h>
#include "types.h"
#include <linux/list.h>
#include <linux/rbtree.h>
@@ -26,6 +27,16 @@ static inline char *bfd_demangle(void __used *v, const char __used *c,
#endif
#endif
+/*
+ * libelf 0.8.x and earlier do not support ELF_C_READ_MMAP;
+ * for newer versions we can use mmap to reduce memory usage:
+ */
+#ifdef LIBELF_NO_MMAP
+# define PERF_ELF_C_READ_MMAP ELF_C_READ
+#else
+# define PERF_ELF_C_READ_MMAP ELF_C_READ_MMAP
+#endif
+
#ifndef DMGL_PARAMS
#define DMGL_PARAMS (1 << 0) /* Include function args */
#define DMGL_ANSI (1 << 1) /* Include const, volatile, etc */
@@ -35,52 +46,58 @@ struct symbol {
struct rb_node rb_node;
u64 start;
u64 end;
- u64 hist_sum;
- u64 *hist;
- void *priv;
char name[0];
};
+extern unsigned int symbol__priv_size;
+
+static inline void *symbol__priv(struct symbol *self)
+{
+ return ((void *)self) - symbol__priv_size;
+}
+
struct dso {
struct list_head node;
struct rb_root syms;
struct symbol *(*find_symbol)(struct dso *, u64 ip);
- unsigned int sym_priv_size;
- unsigned char adjust_symbols;
- unsigned char slen_calculated;
+ u8 adjust_symbols:1;
+ u8 slen_calculated:1;
+ u8 loaded:1;
+ u8 has_build_id:1;
unsigned char origin;
+ u8 build_id[BUILD_ID_SIZE];
+ u16 long_name_len;
const char *short_name;
char *long_name;
char name[0];
};
-extern const char *sym_hist_filter;
-
-typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym);
-
-struct dso *dso__new(const char *name, unsigned int sym_priv_size);
+struct dso *dso__new(const char *name);
void dso__delete(struct dso *self);
-static inline void *dso__sym_priv(struct dso *self, struct symbol *sym)
-{
- return ((void *)sym) - self->sym_priv_size;
-}
-
struct symbol *dso__find_symbol(struct dso *self, u64 ip);
-int dsos__load_kernel(const char *vmlinux, unsigned int sym_priv_size,
- symbol_filter_t filter, int verbose, int modules);
-int dso__load(struct dso *self, struct map *map, symbol_filter_t filter,
- int verbose);
+int dsos__load_modules(void);
struct dso *dsos__findnew(const char *name);
+int dso__load(struct dso *self, struct map *map, symbol_filter_t filter);
+int dso__load_kernel_sym(struct dso *self, symbol_filter_t filter, int modules);
void dsos__fprintf(FILE *fp);
+size_t dsos__fprintf_buildid(FILE *fp);
+size_t dso__fprintf_buildid(struct dso *self, FILE *fp);
size_t dso__fprintf(struct dso *self, FILE *fp);
char dso__symtab_origin(const struct dso *self);
+void dso__set_build_id(struct dso *self, void *build_id);
+
+int filename__read_build_id(const char *filename, void *bf, size_t size);
+int sysfs__read_build_id(const char *filename, void *bf, size_t size);
+bool dsos__read_build_ids(void);
+int build_id__sprintf(u8 *self, int len, char *bf);
-int load_kernel(void);
+struct dso *dsos__load_kernel(void);
+int load_kernel(symbol_filter_t filter);
-void symbol__init(void);
+void symbol__init(unsigned int priv_size);
extern struct list_head dsos;
extern struct map *kernel_map;
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
index f53fad7c0a8d..0f6d78c9863a 100644
--- a/tools/perf/util/thread.c
+++ b/tools/perf/util/thread.c
@@ -33,6 +33,17 @@ int thread__set_comm(struct thread *self, const char *comm)
return self->comm ? 0 : -ENOMEM;
}
+int thread__comm_len(struct thread *self)
+{
+ if (!self->comm_len) {
+ if (!self->comm)
+ return 0;
+ self->comm_len = strlen(self->comm);
+ }
+
+ return self->comm_len;
+}
+
static size_t thread__fprintf(struct thread *self, FILE *fp)
{
struct rb_node *nd;
@@ -116,9 +127,9 @@ static void thread__remove_overlappings(struct thread *self, struct map *map)
continue;
if (verbose >= 2) {
- printf("overlapping maps:\n");
- map__fprintf(map, stdout);
- map__fprintf(pos, stdout);
+ fputs("overlapping maps:\n", stderr);
+ map__fprintf(map, stderr);
+ map__fprintf(pos, stderr);
}
rb_erase(&pos->rb_node, &self->maps);
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
index 1abef3b7455d..53addd77ce8f 100644
--- a/tools/perf/util/thread.h
+++ b/tools/perf/util/thread.h
@@ -12,9 +12,11 @@ struct thread {
pid_t pid;
char shortname[3];
char *comm;
+ int comm_len;
};
int thread__set_comm(struct thread *self, const char *comm);
+int thread__comm_len(struct thread *self);
struct thread *threads__findnew(pid_t pid);
struct thread *register_idle_thread(void);
void thread__insert_map(struct thread *self, struct map *map);
diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c
index 4b61b497040e..eae560503086 100644
--- a/tools/perf/util/trace-event-parse.c
+++ b/tools/perf/util/trace-event-parse.c
@@ -286,16 +286,19 @@ void parse_ftrace_printk(char *file, unsigned int size __unused)
char *line;
char *next = NULL;
char *addr_str;
- char *fmt;
int i;
line = strtok_r(file, "\n", &next);
while (line) {
+ addr_str = strsep(&line, ":");
+ if (!line) {
+ warning("error parsing print strings");
+ break;
+ }
item = malloc_or_die(sizeof(*item));
- addr_str = strtok_r(line, ":", &fmt);
item->addr = strtoull(addr_str, NULL, 16);
/* fmt still has a space, skip it */
- item->printk = strdup(fmt+1);
+ item->printk = strdup(line+1);
item->next = list;
list = item;
line = strtok_r(NULL, "\n", &next);
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 9de2329dd44d..f2203a0946bc 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -134,6 +134,15 @@ extern void die(const char *err, ...) NORETURN __attribute__((format (printf, 1,
extern int error(const char *err, ...) __attribute__((format (printf, 1, 2)));
extern void warning(const char *err, ...) __attribute__((format (printf, 1, 2)));
+#include "../../../include/linux/stringify.h"
+
+#define DIE_IF(cnd) \
+ do { if (cnd) \
+ die(" at (" __FILE__ ":" __stringify(__LINE__) "): " \
+ __stringify(cnd) "\n"); \
+ } while (0)
+
+
extern void set_die_routine(void (*routine)(const char *err, va_list params) NORETURN);
extern int prefixcmp(const char *str, const char *prefix);
@@ -306,6 +315,7 @@ static inline int has_extension(const char *filename, const char *ext)
#undef isascii
#undef isspace
#undef isdigit
+#undef isxdigit
#undef isalpha
#undef isprint
#undef isalnum
@@ -323,6 +333,8 @@ extern unsigned char sane_ctype[256];
#define isascii(x) (((x) & ~0x7f) == 0)
#define isspace(x) sane_istest(x,GIT_SPACE)
#define isdigit(x) sane_istest(x,GIT_DIGIT)
+#define isxdigit(x) \
+ (sane_istest(toupper(x), GIT_ALPHA | GIT_DIGIT) && toupper(x) < 'G')
#define isalpha(x) sane_istest(x,GIT_ALPHA)
#define isalnum(x) sane_istest(x,GIT_ALPHA | GIT_DIGIT)
#define isprint(x) sane_istest(x,GIT_PRINT)