Messages in this thread Patch in this message | | | From | Kirill Smelkov <> | Subject | [PATCH 4/6] perf annotate: fix it for non-prelinked *.so | Date | Fri, 8 Jan 2010 15:23:07 +0300 |
| |
The problem was we were incorrectly calculating objdump addresses for sym->start and sym->end, look:
For simple ET_DYN type DSO (*.so) with one function, objdump -dS output is something like this:
000004ac <my_strlen>: int my_strlen(const char *s) 4ac: 55 push %ebp 4ad: 89 e5 mov %esp,%ebp 4af: 83 ec 10 sub $0x10,%esp {
i.e. we have relative-to-dso-mapping IPs (=RIP) there.
For ET_EXEC type and probably for prelinked libs as well (sorry can't test - I don't use prelink) objdump outputs absolute IPs, e.g.
08048604 <zz_strlen>: extern "C" int zz_strlen(const char *s) 8048604: 55 push %ebp 8048605: 89 e5 mov %esp,%ebp 8048607: 83 ec 10 sub $0x10,%esp {
So, if sym->start is always relative to dso mapping(*), we'll have to unmap it for ET_EXEC like cases, and leave as is for ET_DYN cases.
(*) and it is - we've explicitely made it relative. Look for adjust_symbols handling in dso__load_sym()
Previously we were always unmapping sym->start and for ET_DYN dsos resulting addresses were wrong, and so objdump output was empty.
The end result was that perf annotate output for symbols from non-prelinked *.so had always 0.00% percents only, which is wrong.
----
To fix it, let's introduce a helper for converting rip to objdump address, and also let's document what map_ip() and unmap_ip() do -- I had to study sources for several hours to understand it.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Mike Galbraith <efault@gmx.de> --- tools/perf/builtin-annotate.c | 4 ++-- tools/perf/util/map.c | 14 ++++++++++++++ tools/perf/util/map.h | 9 +++++++++ 3 files changed, 25 insertions(+), 2 deletions(-)
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 117bbae..117301a 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -195,7 +195,7 @@ static int parse_line(FILE *file, struct hist_entry *he, u64 len) line_ip = -1; } - start = he->map->unmap_ip(he->map, sym->start); + start = map__rip_2objdump(he->map, sym->start); if (line_ip != -1) { const char *path = NULL; @@ -405,7 +405,7 @@ static void annotate_sym(struct hist_entry *he) dso, dso->long_name, sym, sym->name); sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s|grep -v %s", - map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end), + map__rip_2objdump(map, sym->start), map__rip_2objdump(map, sym->end), filename, filename); if (verbose >= 3) diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index c4d55a0..d6da969 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -201,3 +201,17 @@ size_t map__fprintf(struct map *self, FILE *fp) return fprintf(fp, " %Lx-%Lx %Lx %s\n", self->start, self->end, self->pgoff, self->dso->name); } + + + +/* + * objdump wants/reports absolute IPs for ET_EXEC, and RIPs for ET_DYN. + * map->dso->adjust_symbols==1 for ET_EXEC-like cases. + */ +u64 map__rip_2objdump(struct map *map, u64 rip) +{ + u64 addr = map->dso->adjust_symbols ? + map->unmap_ip(map, rip) : /* RIP -> IP */ + rip; + return addr; +} diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index 72f0b6a..39fa478 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -24,8 +24,12 @@ struct map { u64 end; enum map_type type; u64 pgoff; + + /* ip -> dso rip */ u64 (*map_ip)(struct map *, u64); + /* dso rip -> ip */ u64 (*unmap_ip)(struct map *, u64); + struct dso *dso; }; @@ -44,6 +48,11 @@ static inline u64 identity__map_ip(struct map *map __used, u64 ip) return ip; } + +/* rip -> addr suitable for passing to `objdump --start-address=` */ +u64 map__rip_2objdump(struct map *map, u64 rip); + + struct symbol; struct mmap_event; -- 1.6.6.79.gd514e.dirty
| |