From 6db81643fe41f51cf9f6fd10558b8b323d9198ba Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 25 Apr 2017 11:45:59 -0300 Subject: [PATCH 01/16] perf buildid: Move prototypes from util.h to build-id.h Where they belong. Link: http://lkml.kernel.org/n/tip-94m3dziejxgo7k0488q3mqjm@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/sdt.c | 1 - tools/perf/util/build-id.h | 4 ++++ tools/perf/util/util.h | 3 --- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/tools/perf/tests/sdt.c b/tools/perf/tests/sdt.c index f73b3c5e125d..06eda675ae2c 100644 --- a/tools/perf/tests/sdt.c +++ b/tools/perf/tests/sdt.c @@ -1,7 +1,6 @@ #include #include #include -#include #include #include #include "tests.h" diff --git a/tools/perf/util/build-id.h b/tools/perf/util/build-id.h index a96081121179..8a89b195c1fc 100644 --- a/tools/perf/util/build-id.h +++ b/tools/perf/util/build-id.h @@ -44,6 +44,10 @@ bool build_id_cache__cached(const char *sbuild_id); int build_id_cache__add_s(const char *sbuild_id, const char *name, bool is_kallsyms, bool is_vdso); int build_id_cache__remove_s(const char *sbuild_id); + +extern char buildid_dir[]; + +void set_buildid_dir(const char *dir); void disable_buildid_cache(void); #endif diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 3852b6d3270a..e7b6f36d8904 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -14,8 +14,6 @@ #include #include -extern char buildid_dir[]; - #ifdef __GNUC__ #define NORETURN __attribute__((__noreturn__)) #else @@ -36,7 +34,6 @@ void warning(const char *err, ...) __attribute__((format (printf, 1, 2))); void set_warning_routine(void (*routine)(const char *err, va_list params)); int prefixcmp(const char *str, const char *prefix); -void set_buildid_dir(const char *dir); static inline void *zalloc(size_t size) { From 5ab8c689f7c0c97e98b8014b7f0ede386bef5aaf Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 25 Apr 2017 15:30:47 -0300 Subject: [PATCH 02/16] perf tools: Move event prototypes from util.h to event.h More needs to be done to have the actual functions and variables in a smaller .c file that can then be included in the python binding, avoiding dragging more stuff into it. Link: http://lkml.kernel.org/n/tip-uecxz7cqkssouj7tlxrkqpl4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-c2c.c | 1 + tools/perf/builtin-top.c | 2 +- tools/perf/builtin-trace.c | 1 + tools/perf/perf.c | 1 + tools/perf/tests/hists_cumulate.c | 1 + tools/perf/tests/hists_filter.c | 1 + tools/perf/tests/hists_output.c | 1 + tools/perf/util/event.c | 1 + tools/perf/util/event.h | 8 ++++++++ tools/perf/util/evsel.c | 1 + tools/perf/util/util.h | 8 -------- 11 files changed, 17 insertions(+), 9 deletions(-) diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c index e33b4acece90..620a467ee304 100644 --- a/tools/perf/builtin-c2c.c +++ b/tools/perf/builtin-c2c.c @@ -27,6 +27,7 @@ #include "tool.h" #include "data.h" #include "sort.h" +#include "event.h" #include "evlist.h" #include "evsel.h" #include diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 7ab42b8311a1..10b6362ca0bf 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -27,13 +27,13 @@ #include "util/drv_configs.h" #include "util/evlist.h" #include "util/evsel.h" +#include "util/event.h" #include "util/machine.h" #include "util/session.h" #include "util/symbol.h" #include "util/thread.h" #include "util/thread_map.h" #include "util/top.h" -#include "util/util.h" #include #include #include "util/parse-events.h" diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index eaa66fb57347..d014350adc52 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -21,6 +21,7 @@ #include "builtin.h" #include "util/color.h" #include "util/debug.h" +#include "util/event.h" #include "util/evlist.h" #include #include "util/machine.h" diff --git a/tools/perf/perf.c b/tools/perf/perf.c index 4cc6960f6226..628a5e412cb1 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -17,6 +17,7 @@ #include #include "util/bpf-loader.h" #include "util/debug.h" +#include "util/event.h" #include #include #include diff --git a/tools/perf/tests/hists_cumulate.c b/tools/perf/tests/hists_cumulate.c index 70918b986568..d549a9f2c41b 100644 --- a/tools/perf/tests/hists_cumulate.c +++ b/tools/perf/tests/hists_cumulate.c @@ -1,5 +1,6 @@ #include "perf.h" #include "util/debug.h" +#include "util/event.h" #include "util/symbol.h" #include "util/sort.h" #include "util/evsel.h" diff --git a/tools/perf/tests/hists_filter.c b/tools/perf/tests/hists_filter.c index f171b2da4899..df9c91f49af1 100644 --- a/tools/perf/tests/hists_filter.c +++ b/tools/perf/tests/hists_filter.c @@ -3,6 +3,7 @@ #include "util/symbol.h" #include "util/sort.h" #include "util/evsel.h" +#include "util/event.h" #include "util/evlist.h" #include "util/machine.h" #include "util/thread.h" diff --git a/tools/perf/tests/hists_output.c b/tools/perf/tests/hists_output.c index cdf0dde5fe97..06e5080182d3 100644 --- a/tools/perf/tests/hists_output.c +++ b/tools/perf/tests/hists_output.c @@ -1,5 +1,6 @@ #include "perf.h" #include "util/debug.h" +#include "util/event.h" #include "util/symbol.h" #include "util/sort.h" #include "util/evsel.h" diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 142835c0ca0a..437194fe0434 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -8,6 +8,7 @@ #include #include /* To get things like MAP_HUGETLB even on older libc headers */ #include +#include #include "event.h" #include "debug.h" #include "hist.h" diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index db2de6413518..cfbe32bd7413 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -681,4 +681,12 @@ u64 kallsyms__get_function_start(const char *kallsyms_filename, void *cpu_map_data__alloc(struct cpu_map *map, size_t *size, u16 *type, int *max); void cpu_map_data__synthesize(struct cpu_map_data *data, struct cpu_map *map, u16 type, int max); + +void event_attr_init(struct perf_event_attr *attr); + +int perf_event_paranoid(void); + +extern int sysctl_perf_event_max_stack; +extern int sysctl_perf_event_max_contexts_per_stack; + #endif /* __PERF_RECORD_H */ diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index 0e879097adfb..e4f7902d5afa 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -21,6 +21,7 @@ #include "asm/bug.h" #include "callchain.h" #include "cgroup.h" +#include "event.h" #include "evsel.h" #include "evlist.h" #include "util.h" diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index e7b6f36d8904..f87b8948efdc 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -56,17 +56,11 @@ int copyfile_offset(int fromfd, loff_t from_ofs, int tofd, loff_t to_ofs, u64 si ssize_t readn(int fd, void *buf, size_t n); ssize_t writen(int fd, void *buf, size_t n); -struct perf_event_attr; - -void event_attr_init(struct perf_event_attr *attr); - size_t hex_width(u64 v); int hex2u64(const char *ptr, u64 *val); extern unsigned int page_size; extern int cacheline_size; -extern int sysctl_perf_event_max_stack; -extern int sysctl_perf_event_max_contexts_per_stack; struct parse_tag { char tag; @@ -75,8 +69,6 @@ struct parse_tag { unsigned long parse_tag_value(const char *str, struct parse_tag *tags); -int perf_event_paranoid(void); - void mem_bswap_64(void *src, int byte_size); void mem_bswap_32(void *src, int byte_size); From 98521b3869f8d6b4b9d2fdad8a56059e819ae002 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 25 Apr 2017 15:45:35 -0300 Subject: [PATCH 03/16] perf memswap: Split the byteswap memory range wrappers from util.[ch] Just one more step into splitting util.[ch] to reduce the includes hell. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-navarr9mijkgwgbzu464dwam@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/Build | 1 + tools/perf/util/header.c | 1 + tools/perf/util/intel-pt.c | 1 + tools/perf/util/memswap.c | 24 ++++++++++++++++++++++++ tools/perf/util/memswap.h | 7 +++++++ tools/perf/util/session.c | 1 + tools/perf/util/util.c | 22 ---------------------- tools/perf/util/util.h | 3 --- 8 files changed, 35 insertions(+), 25 deletions(-) create mode 100644 tools/perf/util/memswap.c create mode 100644 tools/perf/util/memswap.h diff --git a/tools/perf/util/Build b/tools/perf/util/Build index 069583bdc670..79dea95a7f68 100644 --- a/tools/perf/util/Build +++ b/tools/perf/util/Build @@ -13,6 +13,7 @@ libperf-y += find_bit.o libperf-y += kallsyms.o libperf-y += levenshtein.o libperf-y += llvm-utils.o +libperf-y += memswap.o libperf-y += parse-events.o libperf-y += perf_regs.o libperf-y += path.o diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 948b2c5efb65..314a07151fb7 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -19,6 +19,7 @@ #include "evlist.h" #include "evsel.h" #include "header.h" +#include "memswap.h" #include "../perf.h" #include "trace-event.h" #include "session.h" diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index bdd4a28c6cee..4c7718f87a08 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -23,6 +23,7 @@ #include "../perf.h" #include "session.h" #include "machine.h" +#include "memswap.h" #include "sort.h" #include "tool.h" #include "event.h" diff --git a/tools/perf/util/memswap.c b/tools/perf/util/memswap.c new file mode 100644 index 000000000000..55f7faa8d9ec --- /dev/null +++ b/tools/perf/util/memswap.c @@ -0,0 +1,24 @@ +#include +#include "memswap.h" +#include + +void mem_bswap_32(void *src, int byte_size) +{ + u32 *m = src; + while (byte_size > 0) { + *m = bswap_32(*m); + byte_size -= sizeof(u32); + ++m; + } +} + +void mem_bswap_64(void *src, int byte_size) +{ + u64 *m = src; + + while (byte_size > 0) { + *m = bswap_64(*m); + byte_size -= sizeof(u64); + ++m; + } +} diff --git a/tools/perf/util/memswap.h b/tools/perf/util/memswap.h new file mode 100644 index 000000000000..7d1b1c34bb57 --- /dev/null +++ b/tools/perf/util/memswap.h @@ -0,0 +1,7 @@ +#ifndef PERF_MEMSWAP_H_ +#define PERF_MEMSWAP_H_ + +void mem_bswap_64(void *src, int byte_size); +void mem_bswap_32(void *src, int byte_size); + +#endif /* PERF_MEMSWAP_H_ */ diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 3041c6b98191..7dc1096264c5 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -11,6 +11,7 @@ #include "evlist.h" #include "evsel.h" +#include "memswap.h" #include "session.h" #include "tool.h" #include "sort.h" diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index 6450c75a6f5b..b460f0db84d1 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include @@ -372,27 +371,6 @@ int perf_event_paranoid(void) return value; } -void mem_bswap_32(void *src, int byte_size) -{ - u32 *m = src; - while (byte_size > 0) { - *m = bswap_32(*m); - byte_size -= sizeof(u32); - ++m; - } -} - -void mem_bswap_64(void *src, int byte_size) -{ - u64 *m = src; - - while (byte_size > 0) { - *m = bswap_64(*m); - byte_size -= sizeof(u64); - ++m; - } -} - bool find_process(const char *name) { size_t len = strlen(name); diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index f87b8948efdc..6855c454e5bc 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -69,9 +69,6 @@ struct parse_tag { unsigned long parse_tag_value(const char *str, struct parse_tag *tags); -void mem_bswap_64(void *src, int byte_size); -void mem_bswap_32(void *src, int byte_size); - bool find_process(const char *name); int fetch_kernel_version(unsigned int *puint, From 7de96c3e759e49b1d176722ed26c14c057bac34f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 26 Apr 2017 15:27:52 -0300 Subject: [PATCH 04/16] perf tools: Move HAS_BOOL define to where perl headers are used This is a perl specific hack, so move it from util.h to where perl headers are used. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-4igctbinuom2sr6g4b03jqht@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/scripting-engines/trace-event-perl.c | 4 +++- tools/perf/util/util.h | 1 - 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c index 2b12bdb3ce33..7b79c413486b 100644 --- a/tools/perf/util/scripting-engines/trace-event-perl.c +++ b/tools/perf/util/scripting-engines/trace-event-perl.c @@ -28,7 +28,9 @@ #include #include -#include "../util.h" +#include +/* perl needs the following define, right after including stdbool.h */ +#define HAS_BOOL #include #include diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 6855c454e5bc..622c2d251ad3 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -5,7 +5,6 @@ #define _BSD_SOURCE 1 /* glibc 2.20 deprecates _BSD_SOURCE in favour of _DEFAULT_SOURCE */ #define _DEFAULT_SOURCE 1 -#define HAS_BOOL #include #include From 5068b52f732157385d9dccd205bc4043da3a3cce Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 26 Apr 2017 15:31:57 -0300 Subject: [PATCH 05/16] perf ui gtk: Move gtk .so name to the only place where it is used No need to pollute util.h with this. Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/n/tip-kec0chbdtgrd71o3oi2kz2zt@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/setup.c | 3 +++ tools/perf/util/util.h | 2 -- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/perf/ui/setup.c b/tools/perf/ui/setup.c index 5ea0b40c4fc2..caf1ce6f5152 100644 --- a/tools/perf/ui/setup.c +++ b/tools/perf/ui/setup.c @@ -10,7 +10,10 @@ pthread_mutex_t ui__lock = PTHREAD_MUTEX_INITIALIZER; void *perf_gtk_handle; int use_browser = -1; +#define PERF_GTK_DSO "libperf-gtk.so" + #ifdef HAVE_GTK2_SUPPORT + static int setup_gtk_browser(void) { int (*perf_ui_init)(void); diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 622c2d251ad3..dabdc810ffdc 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -22,8 +22,6 @@ #endif #endif -#define PERF_GTK_DSO "libperf-gtk.so" - /* General helper functions */ void usage(const char *err) NORETURN; void die(const char *err, ...) NORETURN __attribute__((format (printf, 1, 2))); From 3caeafce5392a8eba7b36d0d097d403cacc66e2d Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 26 Apr 2017 15:40:31 -0300 Subject: [PATCH 06/16] perf units: Move parse_tag_value() to units.[ch] Its basically to do units handling, so move to a more appropriately named object. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-90ob9vfepui24l8l2makhd9u@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/units.c | 29 +++++++++++++++++++++++++++++ tools/perf/util/units.h | 7 +++++++ tools/perf/util/util.c | 27 --------------------------- tools/perf/util/util.h | 7 ------- 4 files changed, 36 insertions(+), 34 deletions(-) diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c index f6a2a3d117d5..4767ec2c5ef6 100644 --- a/tools/perf/util/units.c +++ b/tools/perf/util/units.c @@ -1,8 +1,37 @@ #include "units.h" #include +#include +#include +#include #include #include +unsigned long parse_tag_value(const char *str, struct parse_tag *tags) +{ + struct parse_tag *i = tags; + + while (i->tag) { + char *s = strchr(str, i->tag); + + if (s) { + unsigned long int value; + char *endptr; + + value = strtoul(str, &endptr, 10); + if (s != endptr) + break; + + if (value > ULONG_MAX / i->mult) + break; + value *= i->mult; + return value; + } + i++; + } + + return (unsigned long) -1; +} + unsigned long convert_unit(unsigned long value, char *unit) { *unit = ' '; diff --git a/tools/perf/util/units.h b/tools/perf/util/units.h index 3ed7774afaa9..f02c87317150 100644 --- a/tools/perf/util/units.h +++ b/tools/perf/util/units.h @@ -4,6 +4,13 @@ #include #include +struct parse_tag { + char tag; + int mult; +}; + +unsigned long parse_tag_value(const char *str, struct parse_tag *tags); + unsigned long convert_unit(unsigned long value, char *unit); int unit_number__scnprintf(char *buf, size_t size, u64 n); diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index b460f0db84d1..28c9f335006c 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -334,33 +334,6 @@ int hex2u64(const char *ptr, u64 *long_val) return p - ptr; } -unsigned long parse_tag_value(const char *str, struct parse_tag *tags) -{ - struct parse_tag *i = tags; - - while (i->tag) { - char *s; - - s = strchr(str, i->tag); - if (s) { - unsigned long int value; - char *endptr; - - value = strtoul(str, &endptr, 10); - if (s != endptr) - break; - - if (value > ULONG_MAX / i->mult) - break; - value *= i->mult; - return value; - } - i++; - } - - return (unsigned long) -1; -} - int perf_event_paranoid(void) { int value; diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index dabdc810ffdc..d620719775a8 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -59,13 +59,6 @@ int hex2u64(const char *ptr, u64 *val); extern unsigned int page_size; extern int cacheline_size; -struct parse_tag { - char tag; - int mult; -}; - -unsigned long parse_tag_value(const char *str, struct parse_tag *tags); - bool find_process(const char *name); int fetch_kernel_version(unsigned int *puint, From 96395cbbc7e94cbbe4f76cf64cf122fabc19123d Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 26 Apr 2017 15:49:21 -0300 Subject: [PATCH 07/16] tools lib string: Adopt prefixcmp() from perf and subcmd Both had copies originating from git.git, move those to tools/lib/string.c, getting both tools/lib/subcmd/ and tools/perf/ to use it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Josh Poimboeuf Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-uidwtticro1qhttzd2rkrkg1@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/linux/string.h | 2 ++ tools/lib/string.c | 9 +++++++++ tools/lib/subcmd/help.c | 1 + tools/lib/subcmd/parse-options.c | 1 + tools/lib/subcmd/subcmd-util.h | 9 --------- tools/perf/util/strbuf.c | 9 --------- tools/perf/util/util.h | 2 -- 7 files changed, 13 insertions(+), 20 deletions(-) diff --git a/tools/include/linux/string.h b/tools/include/linux/string.h index f436d2420a18..d62b56cf8c12 100644 --- a/tools/include/linux/string.h +++ b/tools/include/linux/string.h @@ -18,4 +18,6 @@ extern size_t strlcpy(char *dest, const char *src, size_t size); char *str_error_r(int errnum, char *buf, size_t buflen); +int prefixcmp(const char *str, const char *prefix); + #endif /* _LINUX_STRING_H_ */ diff --git a/tools/lib/string.c b/tools/lib/string.c index bd239bc1d557..8e678af1c6ee 100644 --- a/tools/lib/string.c +++ b/tools/lib/string.c @@ -87,3 +87,12 @@ size_t __weak strlcpy(char *dest, const char *src, size_t size) } return ret; } + +int prefixcmp(const char *str, const char *prefix) +{ + for (; ; str++, prefix++) + if (!*prefix) + return 0; + else if (*str != *prefix) + return (unsigned char)*prefix - (unsigned char)*str; +} diff --git a/tools/lib/subcmd/help.c b/tools/lib/subcmd/help.c index e228c3cb3716..ba970a73d053 100644 --- a/tools/lib/subcmd/help.c +++ b/tools/lib/subcmd/help.c @@ -1,6 +1,7 @@ #include #include #include +#include #include #include #include diff --git a/tools/lib/subcmd/parse-options.c b/tools/lib/subcmd/parse-options.c index 6bc24025d054..359bfa77f39c 100644 --- a/tools/lib/subcmd/parse-options.c +++ b/tools/lib/subcmd/parse-options.c @@ -1,4 +1,5 @@ #include +#include #include #include #include diff --git a/tools/lib/subcmd/subcmd-util.h b/tools/lib/subcmd/subcmd-util.h index fc2e45d8aaf1..8fa5f036eff0 100644 --- a/tools/lib/subcmd/subcmd-util.h +++ b/tools/lib/subcmd/subcmd-util.h @@ -79,13 +79,4 @@ static inline void astrcat(char **out, const char *add) free(tmp); } -static inline int prefixcmp(const char *str, const char *prefix) -{ - for (; ; str++, prefix++) - if (!*prefix) - return 0; - else if (*str != *prefix) - return (unsigned char)*prefix - (unsigned char)*str; -} - #endif /* __SUBCMD_UTIL_H */ diff --git a/tools/perf/util/strbuf.c b/tools/perf/util/strbuf.c index e91b5e86f027..aafe908b82b5 100644 --- a/tools/perf/util/strbuf.c +++ b/tools/perf/util/strbuf.c @@ -3,15 +3,6 @@ #include #include -int prefixcmp(const char *str, const char *prefix) -{ - for (; ; str++, prefix++) - if (!*prefix) - return 0; - else if (*str != *prefix) - return (unsigned char)*prefix - (unsigned char)*str; -} - /* * Used as the default ->buf value, so that people can always assume * buf is non NULL and ->buf is NUL terminated even for a freshly diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index d620719775a8..5dfb9bb6482d 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -30,8 +30,6 @@ void warning(const char *err, ...) __attribute__((format (printf, 1, 2))); void set_warning_routine(void (*routine)(const char *err, va_list params)); -int prefixcmp(const char *str, const char *prefix); - static inline void *zalloc(size_t size) { return calloc(1, size); From b843f62ad942f3879c2d7acd27b44f931c609685 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 27 Apr 2017 21:21:09 -0300 Subject: [PATCH 08/16] perf symbols: Accept symbols starting at address 0 That is the case of _text on s390, and we have some functions that return an address, using address zero to report problems, oops. This would lead the symbol loading routines to not use "_text" as the reference relocation symbol, or the first symbol for the kernel, but use instead "_stext", that is at the same address on x86_64 and others, but not on s390: [acme@localhost perf-4.11.0-rc6]$ head -15 /proc/kallsyms 0000000000000000 T _text 0000000000000418 t iplstart 0000000000000800 T start 000000000000080a t .base 000000000000082e t .sk8x8 0000000000000834 t .gotr 0000000000000842 t .cmd 0000000000000846 t .parm 000000000000084a t .lowcase 0000000000010000 T startup 0000000000010010 T startup_kdump 0000000000010214 t startup_kdump_relocated 0000000000011000 T startup_continue 00000000000112a0 T _ehead 0000000000100000 T _stext [acme@localhost perf-4.11.0-rc6]$ Which in turn would make 'perf test vmlinux' to fail because it wouldn't find the symbols before "_stext" in kallsyms. Fix it by using the return value only for errors and storing the address, when the symbol is successfully found, in a provided pointer arg. Before this patch: After: [acme@localhost perf-4.11.0-rc6]$ tools/perf/perf test -v 1 1: vmlinux symtab matches kallsyms : --- start --- test child forked, pid 40693 Looking at the vmlinux_path (8 entries long) Using /usr/lib/debug/lib/modules/3.10.0-654.el7.s390x/vmlinux for symbols ERR : 0: _text not on kallsyms ERR : 0x418: iplstart not on kallsyms ERR : 0x800: start not on kallsyms ERR : 0x80a: .base not on kallsyms ERR : 0x82e: .sk8x8 not on kallsyms ERR : 0x834: .gotr not on kallsyms ERR : 0x842: .cmd not on kallsyms ERR : 0x846: .parm not on kallsyms ERR : 0x84a: .lowcase not on kallsyms ERR : 0x10000: startup not on kallsyms ERR : 0x10010: startup_kdump not on kallsyms ERR : 0x10214: startup_kdump_relocated not on kallsyms ERR : 0x11000: startup_continue not on kallsyms ERR : 0x112a0: _ehead not on kallsyms test child finished with -1 ---- end ---- vmlinux symtab matches kallsyms: FAILED! [acme@localhost perf-4.11.0-rc6]$ After: [acme@localhost perf-4.11.0-rc6]$ tools/perf/perf test -v 1 1: vmlinux symtab matches kallsyms : --- start --- test child forked, pid 47160 test child finished with 0 ---- end ---- vmlinux symtab matches kallsyms: Ok [acme@localhost perf-4.11.0-rc6]$ Reported-by: Michael Petlan Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-9x9bwgd3btwdk1u51xie93fz@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-buildid-cache.c | 13 ++++++++----- tools/perf/util/event.c | 9 +++++---- tools/perf/util/event.h | 4 ++-- tools/perf/util/machine.c | 28 +++++++++++++++++----------- tools/perf/util/symbol.c | 11 +++++------ 5 files changed, 37 insertions(+), 28 deletions(-) diff --git a/tools/perf/builtin-buildid-cache.c b/tools/perf/builtin-buildid-cache.c index 64b44e81c771..9eba7f1add1f 100644 --- a/tools/perf/builtin-buildid-cache.c +++ b/tools/perf/builtin-buildid-cache.c @@ -49,19 +49,22 @@ static bool same_kallsyms_reloc(const char *from_dir, char *to_dir) char to[PATH_MAX]; const char *name; u64 addr1 = 0, addr2 = 0; - int i; + int i, err = -1; scnprintf(from, sizeof(from), "%s/kallsyms", from_dir); scnprintf(to, sizeof(to), "%s/kallsyms", to_dir); for (i = 0; (name = ref_reloc_sym_names[i]) != NULL; i++) { - addr1 = kallsyms__get_function_start(from, name); - if (addr1) + err = kallsyms__get_function_start(from, name, &addr1); + if (!err) break; } - if (name) - addr2 = kallsyms__get_function_start(to, name); + if (err) + return false; + + if (kallsyms__get_function_start(to, name, &addr2)) + return false; return addr1 == addr2; } diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 437194fe0434..dc5c3bb69d73 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -768,15 +768,16 @@ static int find_symbol_cb(void *arg, const char *name, char type, return 1; } -u64 kallsyms__get_function_start(const char *kallsyms_filename, - const char *symbol_name) +int kallsyms__get_function_start(const char *kallsyms_filename, + const char *symbol_name, u64 *addr) { struct process_symbol_args args = { .name = symbol_name, }; if (kallsyms__parse(kallsyms_filename, &args, find_symbol_cb) <= 0) - return 0; + return -1; - return args.start; + *addr = args.start; + return 0; } int perf_event__synthesize_kernel_mmap(struct perf_tool *tool, diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index cfbe32bd7413..27ac047490c3 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -675,8 +675,8 @@ size_t perf_event__fprintf_cpu_map(union perf_event *event, FILE *fp); size_t perf_event__fprintf_namespaces(union perf_event *event, FILE *fp); size_t perf_event__fprintf(union perf_event *event, FILE *fp); -u64 kallsyms__get_function_start(const char *kallsyms_filename, - const char *symbol_name); +int kallsyms__get_function_start(const char *kallsyms_filename, + const char *symbol_name, u64 *addr); void *cpu_map_data__alloc(struct cpu_map *map, size_t *size, u16 *type, int *max); void cpu_map_data__synthesize(struct cpu_map_data *data, struct cpu_map *map, diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 7a47f52ccfcc..d97e014c3df3 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -796,11 +796,11 @@ const char *ref_reloc_sym_names[] = {"_text", "_stext", NULL}; * Returns the name of the start symbol in *symbol_name. Pass in NULL as * symbol_name if it's not that important. */ -static u64 machine__get_running_kernel_start(struct machine *machine, - const char **symbol_name) +static int machine__get_running_kernel_start(struct machine *machine, + const char **symbol_name, u64 *start) { char filename[PATH_MAX]; - int i; + int i, err = -1; const char *name; u64 addr = 0; @@ -810,21 +810,28 @@ static u64 machine__get_running_kernel_start(struct machine *machine, return 0; for (i = 0; (name = ref_reloc_sym_names[i]) != NULL; i++) { - addr = kallsyms__get_function_start(filename, name); - if (addr) + err = kallsyms__get_function_start(filename, name, &addr); + if (!err) break; } + if (err) + return -1; + if (symbol_name) *symbol_name = name; - return addr; + *start = addr; + return 0; } int __machine__create_kernel_maps(struct machine *machine, struct dso *kernel) { int type; - u64 start = machine__get_running_kernel_start(machine, NULL); + u64 start = 0; + + if (machine__get_running_kernel_start(machine, NULL, &start)) + return -1; /* In case of renewal the kernel map, destroy previous one */ machine__destroy_kernel_maps(machine); @@ -1185,8 +1192,8 @@ static int machine__create_modules(struct machine *machine) int machine__create_kernel_maps(struct machine *machine) { struct dso *kernel = machine__get_kernel(machine); - const char *name; - u64 addr; + const char *name = NULL; + u64 addr = 0; int ret; if (kernel == NULL) @@ -1211,8 +1218,7 @@ int machine__create_kernel_maps(struct machine *machine) */ map_groups__fixup_end(&machine->kmaps); - addr = machine__get_running_kernel_start(machine, &name); - if (!addr) { + if (machine__get_running_kernel_start(machine, &name, &addr)) { } else if (maps__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps, name, addr)) { machine__destroy_kernel_maps(machine); return -1; diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 2cb7665e9973..b349e8eda0e2 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -466,7 +466,7 @@ void dso__insert_symbol(struct dso *dso, enum map_type type, struct symbol *sym) struct symbol *dso__find_symbol(struct dso *dso, enum map_type type, u64 addr) { - if (dso->last_find_result[type].addr != addr) { + if (dso->last_find_result[type].addr != addr || dso->last_find_result[type].symbol == NULL) { dso->last_find_result[type].addr = addr; dso->last_find_result[type].symbol = symbols__find(&dso->symbols[type], addr); } @@ -1075,8 +1075,9 @@ static int validate_kcore_addresses(const char *kallsyms_filename, if (kmap->ref_reloc_sym && kmap->ref_reloc_sym->name) { u64 start; - start = kallsyms__get_function_start(kallsyms_filename, - kmap->ref_reloc_sym->name); + if (kallsyms__get_function_start(kallsyms_filename, + kmap->ref_reloc_sym->name, &start)) + return -ENOENT; if (start != kmap->ref_reloc_sym->addr) return -EINVAL; } @@ -1248,9 +1249,7 @@ static int kallsyms__delta(struct map *map, const char *filename, u64 *delta) if (!kmap->ref_reloc_sym || !kmap->ref_reloc_sym->name) return 0; - addr = kallsyms__get_function_start(filename, - kmap->ref_reloc_sym->name); - if (!addr) + if (kallsyms__get_function_start(filename, kmap->ref_reloc_sym->name, &addr)) return -1; *delta = addr - kmap->ref_reloc_sym->addr; From d80406453ad4a69932dc17a617d5b7bc7ae80b8f Mon Sep 17 00:00:00 2001 From: Paul Clarke Date: Tue, 25 Apr 2017 13:15:49 -0500 Subject: [PATCH 09/16] perf symbols: Allow user probes on versioned symbols Symbol versioning, as in glibc, results in symbols being defined as: @[@] (Note that "@@" identifies a default symbol, if the symbol name is repeated.) perf is currently unable to deal with this, and is unable to create user probes at such symbols: -- $ nm /lib/powerpc64le-linux-gnu/libpthread.so.0 | grep pthread_create 0000000000008d30 t __pthread_create_2_1 0000000000008d30 T pthread_create@@GLIBC_2.17 $ /usr/bin/sudo perf probe -v -x /lib/powerpc64le-linux-gnu/libpthread.so.0 pthread_create probe-definition(0): pthread_create symbol:pthread_create file:(null) line:0 offset:0 return:0 lazy:(null) 0 arguments Open Debuginfo file: /usr/lib/debug/lib/powerpc64le-linux-gnu/libpthread-2.19.so Try to find probe point from debuginfo. Probe point 'pthread_create' not found. Error: Failed to add events. Reason: No such file or directory (Code: -2) -- One is not able to specify the fully versioned symbol, either, due to syntactic conflicts with other uses of "@" by perf: -- $ /usr/bin/sudo perf probe -v -x /lib/powerpc64le-linux-gnu/libpthread.so.0 pthread_create@@GLIBC_2.17 probe-definition(0): pthread_create@@GLIBC_2.17 Semantic error :SRC@SRC is not allowed. 0 arguments Error: Command Parse Error. Reason: Invalid argument (Code: -22) -- This patch ignores versioning for default symbols, thus allowing probes to be created for these symbols: -- $ /usr/bin/sudo ./perf probe -x /lib/powerpc64le-linux-gnu/libpthread.so.0 pthread_create Added new event: probe_libpthread:pthread_create (on pthread_create in /lib/powerpc64le-linux-gnu/libpthread-2.19.so) You can now use it in all perf tools, such as: perf record -e probe_libpthread:pthread_create -aR sleep 1 $ /usr/bin/sudo ./perf record -e probe_libpthread:pthread_create -aR ./test 2 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.052 MB perf.data (2 samples) ] $ /usr/bin/sudo ./perf script test 2915 [000] 19124.260729: probe_libpthread:pthread_create: (3fff99248d38) test 2916 [000] 19124.260962: probe_libpthread:pthread_create: (3fff99248d38) $ /usr/bin/sudo ./perf probe --del=probe_libpthread:pthread_create Removed event: probe_libpthread:pthread_create -- Committer note: Change the variable storing the result of strlen() to 'int', to fix the build on debian:experimental-x-mipsel, fedora:24-x-ARC-uClibc, ubuntu:16.04-x-arm, etc: util/symbol.c: In function 'symbol__match_symbol_name': util/symbol.c:422:11: error: comparison between signed and unsigned integer expressions [-Werror=sign-compare] if (len < versioning - name) ^ Signed-off-by: Paul A. Clarke Tested-by: Arnaldo Carvalho de Melo Acked-by: Masami Hiramatsu Cc: David Ahern Link: http://lkml.kernel.org/r/c2b18d9c-17f8-9285-4868-f58b6359ccac@us.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/powerpc/util/sym-handling.c | 12 ++++ tools/perf/util/map.c | 5 -- tools/perf/util/map.h | 5 +- tools/perf/util/symbol.c | 61 ++++++++++++++++----- tools/perf/util/symbol.h | 11 ++++ 5 files changed, 74 insertions(+), 20 deletions(-) diff --git a/tools/perf/arch/powerpc/util/sym-handling.c b/tools/perf/arch/powerpc/util/sym-handling.c index 39dbe512b9fc..bf9a2594572c 100644 --- a/tools/perf/arch/powerpc/util/sym-handling.c +++ b/tools/perf/arch/powerpc/util/sym-handling.c @@ -52,6 +52,18 @@ int arch__compare_symbol_names(const char *namea, const char *nameb) return strcmp(namea, nameb); } + +int arch__compare_symbol_names_n(const char *namea, const char *nameb, + unsigned int n) +{ + /* Skip over initial dot */ + if (*namea == '.') + namea++; + if (*nameb == '.') + nameb++; + + return strncmp(namea, nameb, n); +} #endif #if defined(_CALL_ELF) && _CALL_ELF == 2 diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index ebfa5d92358a..2179b2deb730 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -325,11 +325,6 @@ int map__load(struct map *map) return 0; } -int __weak arch__compare_symbol_names(const char *namea, const char *nameb) -{ - return strcmp(namea, nameb); -} - struct symbol *map__find_symbol(struct map *map, u64 addr) { if (map__load(map) < 0) diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index c8a5a644c0a9..f9e8ac8a52cd 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -130,13 +130,14 @@ struct thread; */ #define __map__for_each_symbol_by_name(map, sym_name, pos) \ for (pos = map__find_symbol_by_name(map, sym_name); \ - pos && arch__compare_symbol_names(pos->name, sym_name) == 0; \ + pos && \ + !symbol__match_symbol_name(pos->name, sym_name, \ + SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); \ pos = symbol__next_by_name(pos)) #define map__for_each_symbol_by_name(map, sym_name, pos) \ __map__for_each_symbol_by_name(map, sym_name, (pos)) -int arch__compare_symbol_names(const char *namea, const char *nameb); void map__init(struct map *map, enum map_type type, u64 start, u64 end, u64 pgoff, struct dso *dso); struct map *map__new(struct machine *machine, u64 start, u64 len, diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index b349e8eda0e2..8f2b068ff756 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -90,6 +90,17 @@ static int prefix_underscores_count(const char *str) return tail - str; } +int __weak arch__compare_symbol_names(const char *namea, const char *nameb) +{ + return strcmp(namea, nameb); +} + +int __weak arch__compare_symbol_names_n(const char *namea, const char *nameb, + unsigned int n) +{ + return strncmp(namea, nameb, n); +} + int __weak arch__choose_best_symbol(struct symbol *syma, struct symbol *symb __maybe_unused) { @@ -399,8 +410,26 @@ static void symbols__sort_by_name(struct rb_root *symbols, } } +int symbol__match_symbol_name(const char *name, const char *str, + enum symbol_tag_include includes) +{ + const char *versioning; + + if (includes == SYMBOL_TAG_INCLUDE__DEFAULT_ONLY && + (versioning = strstr(name, "@@"))) { + int len = strlen(str); + + if (len < versioning - name) + len = versioning - name; + + return arch__compare_symbol_names_n(name, str, len); + } else + return arch__compare_symbol_names(name, str); +} + static struct symbol *symbols__find_by_name(struct rb_root *symbols, - const char *name) + const char *name, + enum symbol_tag_include includes) { struct rb_node *n; struct symbol_name_rb_node *s = NULL; @@ -414,11 +443,11 @@ static struct symbol *symbols__find_by_name(struct rb_root *symbols, int cmp; s = rb_entry(n, struct symbol_name_rb_node, rb_node); - cmp = arch__compare_symbol_names(name, s->sym.name); + cmp = symbol__match_symbol_name(s->sym.name, name, includes); - if (cmp < 0) + if (cmp > 0) n = n->rb_left; - else if (cmp > 0) + else if (cmp < 0) n = n->rb_right; else break; @@ -427,16 +456,17 @@ static struct symbol *symbols__find_by_name(struct rb_root *symbols, if (n == NULL) return NULL; - /* return first symbol that has same name (if any) */ - for (n = rb_prev(n); n; n = rb_prev(n)) { - struct symbol_name_rb_node *tmp; + if (includes != SYMBOL_TAG_INCLUDE__DEFAULT_ONLY) + /* return first symbol that has same name (if any) */ + for (n = rb_prev(n); n; n = rb_prev(n)) { + struct symbol_name_rb_node *tmp; - tmp = rb_entry(n, struct symbol_name_rb_node, rb_node); - if (arch__compare_symbol_names(tmp->sym.name, s->sym.name)) - break; + tmp = rb_entry(n, struct symbol_name_rb_node, rb_node); + if (arch__compare_symbol_names(tmp->sym.name, s->sym.name)) + break; - s = tmp; - } + s = tmp; + } return &s->sym; } @@ -503,7 +533,12 @@ struct symbol *symbol__next_by_name(struct symbol *sym) struct symbol *dso__find_symbol_by_name(struct dso *dso, enum map_type type, const char *name) { - return symbols__find_by_name(&dso->symbol_names[type], name); + struct symbol *s = symbols__find_by_name(&dso->symbol_names[type], name, + SYMBOL_TAG_INCLUDE__NONE); + if (!s) + s = symbols__find_by_name(&dso->symbol_names[type], name, + SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); + return s; } void dso__sort_by_name(struct dso *dso, enum map_type type) diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 7acd70fce68e..41ebba9a2eb2 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -348,8 +348,19 @@ void arch__sym_update(struct symbol *s, GElf_Sym *sym); #define SYMBOL_A 0 #define SYMBOL_B 1 +int arch__compare_symbol_names(const char *namea, const char *nameb); +int arch__compare_symbol_names_n(const char *namea, const char *nameb, + unsigned int n); int arch__choose_best_symbol(struct symbol *syma, struct symbol *symb); +enum symbol_tag_include { + SYMBOL_TAG_INCLUDE__NONE = 0, + SYMBOL_TAG_INCLUDE__DEFAULT_ONLY +}; + +int symbol__match_symbol_name(const char *namea, const char *nameb, + enum symbol_tag_include includes); + /* structure containing an SDT note's info */ struct sdt_note { char *name; /* name of the note*/ From 4341ec6b3db4c3e903d6c44958722918baec1e59 Mon Sep 17 00:00:00 2001 From: Taeung Song Date: Wed, 26 Apr 2017 21:21:02 +0900 Subject: [PATCH 10/16] perf config: Refactor a duplicated code for obtaining config file name We were doing the same sequence to figure out what is the config pathname to use, fix it by doing it before those two uses. Signed-off-by: Taeung Song Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/1493209268-5543-2-git-send-email-treeze.taeung@gmail.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-config.c | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/tools/perf/builtin-config.c b/tools/perf/builtin-config.c index 55f04f85b049..80668fa7556e 100644 --- a/tools/perf/builtin-config.c +++ b/tools/perf/builtin-config.c @@ -159,6 +159,7 @@ int cmd_config(int argc, const char **argv) int i, ret = 0; struct perf_config_set *set; char *user_config = mkpath("%s/.perfconfig", getenv("HOME")); + const char *config_filename; argc = parse_options(argc, argv, config_options, config_usage, PARSE_OPT_STOP_AT_NON_OPTION); @@ -175,6 +176,11 @@ int cmd_config(int argc, const char **argv) else if (use_user_config) config_exclusive_filename = user_config; + if (!config_exclusive_filename) + config_filename = user_config; + else + config_filename = config_exclusive_filename; + /* * At only 'config' sub-command, individually use the config set * because of reinitializing with options config file location. @@ -192,13 +198,9 @@ int cmd_config(int argc, const char **argv) parse_options_usage(config_usage, config_options, "l", 1); } else { ret = show_config(set); - if (ret < 0) { - const char * config_filename = config_exclusive_filename; - if (!config_exclusive_filename) - config_filename = user_config; + if (ret < 0) pr_err("Nothing configured, " "please check your %s \n", config_filename); - } } break; default: @@ -221,13 +223,8 @@ int cmd_config(int argc, const char **argv) if (value == NULL) ret = show_spec_config(set, var); - else { - const char *config_filename = config_exclusive_filename; - - if (!config_exclusive_filename) - config_filename = user_config; + else ret = set_config(set, config_filename, var, value); - } free(arg); } } else From 33b88e708e7dfa58dc896da2a98f5719d2eb315c Mon Sep 17 00:00:00 2001 From: Vince Weaver Date: Tue, 2 May 2017 14:08:50 -0400 Subject: [PATCH 11/16] perf/x86: Fix Broadwell-EP DRAM RAPL events It appears as though the Broadwell-EP DRAM units share the special units quirk with Haswell-EP/KNL. Without this patch, you get really high results (a single DRAM using 20W of power). The powercap driver in drivers/powercap/intel_rapl.c already has this change. Signed-off-by: Vince Weaver Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Kan Liang Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Stephane Eranian Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Cc: Signed-off-by: Ingo Molnar --- arch/x86/events/intel/rapl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/events/intel/rapl.c b/arch/x86/events/intel/rapl.c index 9d05c7e67f60..a45e2114a846 100644 --- a/arch/x86/events/intel/rapl.c +++ b/arch/x86/events/intel/rapl.c @@ -761,7 +761,7 @@ static const struct x86_cpu_id rapl_cpu_match[] __initconst = { X86_RAPL_MODEL_MATCH(INTEL_FAM6_BROADWELL_CORE, hsw_rapl_init), X86_RAPL_MODEL_MATCH(INTEL_FAM6_BROADWELL_GT3E, hsw_rapl_init), - X86_RAPL_MODEL_MATCH(INTEL_FAM6_BROADWELL_X, hsw_rapl_init), + X86_RAPL_MODEL_MATCH(INTEL_FAM6_BROADWELL_X, hsx_rapl_init), X86_RAPL_MODEL_MATCH(INTEL_FAM6_BROADWELL_XEON_D, hsw_rapl_init), X86_RAPL_MODEL_MATCH(INTEL_FAM6_XEON_PHI_KNL, knl_rapl_init), From 1291927a490832a9bbd0158a5d3c05c25a3c7159 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Wed, 3 May 2017 13:13:50 +0100 Subject: [PATCH 12/16] perf tools: Fix spelling mistakes Mostly in the documentation. Signed-off-by: Kim Phillips Cc: Alexander Shishkin Cc: Peter Zijlstra Link: http://lkml.kernel.org/r/20170503131350.cebeecd8bd0f2968417626ab@arm.com [ Fix spelling of "parameter" in one of the spell-checked lines ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Documentation/perf-c2c.txt | 4 ++-- tools/perf/Documentation/perf-record.txt | 2 +- tools/perf/Documentation/perf-report.txt | 6 +++--- tools/perf/Documentation/perf.data-file-format.txt | 4 ++-- tools/perf/Documentation/tips.txt | 2 +- tools/perf/util/event.h | 2 +- 6 files changed, 10 insertions(+), 10 deletions(-) diff --git a/tools/perf/Documentation/perf-c2c.txt b/tools/perf/Documentation/perf-c2c.txt index 2da07e51e119..822414235170 100644 --- a/tools/perf/Documentation/perf-c2c.txt +++ b/tools/perf/Documentation/perf-c2c.txt @@ -76,7 +76,7 @@ REPORT OPTIONS -c:: --coalesce:: - Specify sorintg fields for single cacheline display. + Specify sorting fields for single cacheline display. Following fields are available: tid,pid,iaddr,dso (see COALESCE) @@ -106,7 +106,7 @@ REPORT OPTIONS -d:: --display:: - Siwtch to HITM type (rmt, lcl) to display and sort on. Total HITMs as default. + Switch to HITM type (rmt, lcl) to display and sort on. Total HITMs as default. C2C RECORD ---------- diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt index ea3789d05e5e..b0e9e921d534 100644 --- a/tools/perf/Documentation/perf-record.txt +++ b/tools/perf/Documentation/perf-record.txt @@ -225,7 +225,7 @@ OPTIONS the libunwind or libdw library) should be used instead. Using the "lbr" method doesn't require any compiler options. It will produce call graphs from the hardware LBR registers. The - main limition is that it is only available on new Intel + main limitation is that it is only available on new Intel platforms, such as Haswell. It can only get user call chain. It doesn't work with branch stack sampling at the same time. diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt index 37a175914157..9fa84617181e 100644 --- a/tools/perf/Documentation/perf-report.txt +++ b/tools/perf/Documentation/perf-report.txt @@ -182,7 +182,7 @@ OPTIONS --parent=:: A regex filter to identify parent. The parent is a caller of this function and searched through the callchain, thus it requires callchain - information recorded. The pattern is in the exteneded regex format and + information recorded. The pattern is in the extended regex format and defaults to "\^sys_|^do_page_fault", see '--sort parent'. -x:: @@ -207,8 +207,8 @@ OPTIONS -g:: --call-graph=:: Display call chains using type, min percent threshold, print limit, - call order, sort key, optional branch and value. Note that ordering of - parameters is not fixed so any parement can be given in an arbitraty order. + call order, sort key, optional branch and value. Note that ordering + is not fixed so any parameter can be given in an arbitrary order. One exception is the print_limit which should be preceded by threshold. print_type can be either: diff --git a/tools/perf/Documentation/perf.data-file-format.txt b/tools/perf/Documentation/perf.data-file-format.txt index fa2a9132f0a9..de8b39dda7b8 100644 --- a/tools/perf/Documentation/perf.data-file-format.txt +++ b/tools/perf/Documentation/perf.data-file-format.txt @@ -270,7 +270,7 @@ When the event stream contains multiple events each event is identified by an ID. This can be either through the PERF_SAMPLE_ID or the PERF_SAMPLE_IDENTIFIER header. The PERF_SAMPLE_IDENTIFIER header is at a fixed offset from the event header, which allows reliable -parsing of the header. Relying on ID may be ambigious. +parsing of the header. Relying on ID may be ambiguous. IDENTIFIER is only supported by newer Linux kernels. Perf record specific events: @@ -288,7 +288,7 @@ struct attr_event { uint64_t id[]; }; - PERF_RECORD_HEADER_EVENT_TYPE = 65, /* depreceated */ + PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ #define MAX_EVENT_NAME 64 diff --git a/tools/perf/Documentation/tips.txt b/tools/perf/Documentation/tips.txt index 170b0289a7bc..db0ca3063eae 100644 --- a/tools/perf/Documentation/tips.txt +++ b/tools/perf/Documentation/tips.txt @@ -23,7 +23,7 @@ For memory address profiling, try: perf mem record / perf mem report For tracepoint events, try: perf report -s trace_fields To record callchains for each sample: perf record -g To record every process run by a user: perf record -u -Skip collecing build-id when recording: perf record -B +Skip collecting build-id when recording: perf record -B To change sampling frequency to 100 Hz: perf record -F 100 See assembly instructions with percentage: perf annotate If you prefer Intel style assembly, try: perf annotate -M intel diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h index 27ac047490c3..7c3fa1c8cbcd 100644 --- a/tools/perf/util/event.h +++ b/tools/perf/util/event.h @@ -229,7 +229,7 @@ struct build_id_event { enum perf_user_event_type { /* above any possible kernel type */ PERF_RECORD_USER_TYPE_START = 64, PERF_RECORD_HEADER_ATTR = 64, - PERF_RECORD_HEADER_EVENT_TYPE = 65, /* depreceated */ + PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ PERF_RECORD_HEADER_TRACING_DATA = 66, PERF_RECORD_HEADER_BUILD_ID = 67, PERF_RECORD_FINISHED_ROUND = 68, From d1f7b0234ec7743899321aaaba379b2f5f9d0500 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Wed, 3 May 2017 13:13:56 +0100 Subject: [PATCH 13/16] perf annotate: Fix AArch64 comment char The commit 0fcb1da4aba "perf annotate: AArch64 support" blindly copied the comment character from the original: https://lkml.org/lkml/2016/5/19/461 whereas that same commit shows objdump output utilizing the C++ style "//" as the comment delimeter. Since '/' doesn't occur elsewhere in objdump output, we retain the single character check, but fix it to be '/'. Signed-off-by: Kim Phillips Cc: Alexander Shishkin Cc: Chris Riyder Cc: Peter Zijlstra Fixes: 0fcb1da4aba6 ("perf annotate: AArch64 support") Link: http://lkml.kernel.org/r/20170503131356.be88f977094fb3fa0f49b99d@arm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/arm64/annotate/instructions.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/arch/arm64/annotate/instructions.c b/tools/perf/arch/arm64/annotate/instructions.c index 44eafd6f2d50..8f1908756cb6 100644 --- a/tools/perf/arch/arm64/annotate/instructions.c +++ b/tools/perf/arch/arm64/annotate/instructions.c @@ -50,7 +50,7 @@ static int arm64__annotate_init(struct arch *arch) arch->initialized = true; arch->priv = arm; arch->associate_instruction_ops = arm64__associate_instruction_ops; - arch->objdump.comment_char = ';'; + arch->objdump.comment_char = '/'; arch->objdump.skip_functions_char = '+'; return 0; From 805b151a1afd24414706a7f6ae275fbb9649be74 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Wed, 3 May 2017 13:14:02 +0100 Subject: [PATCH 14/16] perf tests kmod-path: Don't fail if compressed modules aren't supported __kmod_path__parse() uses is_supported_compression() to determine and parse out compressed module file extensions. On systems without zlib, this test fails and __kmod_path__parse() continues to strcmp "ko" with "gz". Don't do this on those systems. Signed-off-by: Kim Phillips Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Peter Zijlstra Fixes: 3c8a67f50a1e ("perf tools: Add kmod_path__parse function") Link: http://lkml.kernel.org/r/20170503131402.c66e314460026c80cd787b34@arm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/kmod-path.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/perf/tests/kmod-path.c b/tools/perf/tests/kmod-path.c index 76f41f249944..6cd9e5107f77 100644 --- a/tools/perf/tests/kmod-path.c +++ b/tools/perf/tests/kmod-path.c @@ -61,6 +61,7 @@ int test__kmod_path__parse(int subtest __maybe_unused) M("/xxxx/xxxx/x-x.ko", PERF_RECORD_MISC_KERNEL, true); M("/xxxx/xxxx/x-x.ko", PERF_RECORD_MISC_USER, false); +#ifdef HAVE_ZLIB_SUPPORT /* path alloc_name alloc_ext kmod comp name ext */ T("/xxxx/xxxx/x.ko.gz", true , true , true, true, "[x]", "gz"); T("/xxxx/xxxx/x.ko.gz", false , true , true, true, NULL , "gz"); @@ -96,6 +97,7 @@ int test__kmod_path__parse(int subtest __maybe_unused) M("x.ko.gz", PERF_RECORD_MISC_CPUMODE_UNKNOWN, true); M("x.ko.gz", PERF_RECORD_MISC_KERNEL, true); M("x.ko.gz", PERF_RECORD_MISC_USER, false); +#endif /* path alloc_name alloc_ext kmod comp name ext */ T("[test_module]", true , true , true, false, "[test_module]", NULL); From 1e6e7eae8a3a66f5fee4f44ef165ffcf31f0c40e Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 4 May 2017 10:30:40 -0300 Subject: [PATCH 15/16] tools build: Fixup sched_getcpu feature test We have tools/build/feature/test-all.c to speed up feature testing, doing all tests at once, but then all tests in this file should normally pass. That is not the case with the sched-getcpu one, that wasn't passing when included from test-all.c because it needs to have _GNU_SOURCE defined before including sched.h, but _GNU_SOURCE is defined by a header included from another feature test included earlier in test-all.d, test-libpython.c, resulting in: $ cat /tmp/build/perf/feature/test-all.make.output In file included from test-all.c:121:0: test-sched_getcpu.c:1:0: error: "_GNU_SOURCE" redefined [-Werror] #define _GNU_SOURCE In file included from /usr/include/python2.7/pyconfig.h:6:0, from /usr/include/python2.7/Python.h:8, from test-libpython.c:1, from test-all.c:13: /usr/include/python2.7/pyconfig-64.h:1177:0: note: this is the location of the previous definition #define _GNU_SOURCE 1 cc1: all warnings being treated as errors Which would trigger testing the tests individually, when that _GNU_SOURCE redefinition would not take place, and the whole process would continue, just slower... Fix it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Fixes: 120010cb1eea ("tools build: Add test for sched_getcpu()") Link: http://lkml.kernel.org/n/tip-3qp1it69xsc4w8gnuu1e9ayh@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/build/feature/test-sched_getcpu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/build/feature/test-sched_getcpu.c b/tools/build/feature/test-sched_getcpu.c index c4a148dd7104..9c6b4cbffb1c 100644 --- a/tools/build/feature/test-sched_getcpu.c +++ b/tools/build/feature/test-sched_getcpu.c @@ -1,4 +1,6 @@ +#ifndef _GNU_SOURCE #define _GNU_SOURCE +#endif #include int main(void) From 88b0193d9418c00340e45e0a913a0813bc6c8c96 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 9 May 2017 18:00:04 +0100 Subject: [PATCH 16/16] perf/callchain: Force USER_DS when invoking perf_callchain_user() Perf can generate and record a user callchain in response to a synchronous request, such as a tracepoint firing. If this happens under set_fs(KERNEL_DS), then we can end up walking the user stack (and dereferencing/saving whatever we find there) without the protections usually afforded by checks such as access_ok. Rather than play whack-a-mole with each architecture's stack unwinding implementation, fix the root of the problem by ensuring that we force USER_DS when invoking perf_callchain_user from the perf core. Reported-by: Al Viro Signed-off-by: Will Deacon Acked-by: Peter Zijlstra Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Thomas Gleixner Signed-off-by: Ingo Molnar --- kernel/events/callchain.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c index c04917cad1bf..1b2be63c8528 100644 --- a/kernel/events/callchain.c +++ b/kernel/events/callchain.c @@ -229,12 +229,18 @@ get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user, } if (regs) { + mm_segment_t fs; + if (crosstask) goto exit_put; if (add_mark) perf_callchain_store_context(&ctx, PERF_CONTEXT_USER); + + fs = get_fs(); + set_fs(USER_DS); perf_callchain_user(&ctx, regs); + set_fs(fs); } }