perf symbol: Reference counting, flat array storage, and LRU shrinking
Ian Rogers
reviewercover perf symbol: Reference counting, flat array storage, and LRU shrinking
This series modernizes struct symbol memory management and backing storage in tools/perf to support reference counting (DECLARE_RC_STRUCT(symbol)) and periodic LRU memory shrinking during long or large profiling sessions:
1. Field accessors (Patch 1): Introduces inline getters/setters for struct symbol fields across tools/perf in preparation for wrapping struct symbol with DECLARE_RC_STRUCT(symbol).
2. Remove negative-offset private allocations (Patch 2): Replaces symbol_conf.priv_size and symbol__priv() negative-offset allocations in util/annotate.c and util/db-export.c with hashmaps keyed by symbol pointer (struct annotated_symbol protected by an rw_semaphore, and sym_db_ids on struct db_export).
3. Flat sorted array backing storage (Patch 3): Replaces the dual address-sorted and name-sorted red-black trees (rb_node) in struct dso with a struct symbols container backed by dynamically sorted flat arrays protected by struct rw_semaphore (util/rwsem.h) with Clang thread-safety annotations. Saves 48 bytes per struct symbol (two rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.
4. Reference counting and DECLARE_RC_STRUCT(symbol) (Patch 4): Wraps struct symbol with DECLARE_RC_STRUCT(symbol) and adds refcount_t refcnt along with symbol__get(), symbol__put(), and symbol__zput(). Lookup functions return a counted reference acquired under symbols->lock, and lifetime holders (addr_location, map_symbol, hist_entry, call_path, block_info, disasm_line, thread_stack, db_export, etc.) properly manage symbol references.
5. LRU memory shrinking (Patch 5): Adds an accessed bit to struct dso and implements symbols__shrink(), dso__shrink_symbols(), dsos__shrink(), machine__shrink(), and machines__shrink(). Unreferenced symbols (refcnt == 1) in cold DSOs are reclaimed, and shrunken DSOs transparently reload on demand via dso__load() while preserving live symbols and their RC_STRUCT(symbol) identity.
6. Periodic session shrinking (Patch 6): Hooks machines__shrink() into perf_session event processing so long-running sessions periodically reclaim cold symbols and DSOs.
7. Unit tests (Patch 7): Adds unit tests in tools/perf/tests/symbols.c covering symbol reference counting, symbols__shrink() retention of active symbols, dsos__shrink() two-pass LRU clock behavior, and on-demand reload preserving live symbol identity.
Ian Rogers (7):
perf symbol: Add accessor functions for struct symbol fields perf symbol: Remove symbol_conf.priv_size and negative-offset
allocations
perf symbol: Switch backing storage from rbtree to struct symbols
array
perf symbol: Add reference counting and DECLARE_RC_STRUCT(symbol) perf symbol: Add LRU memory shrinking for symbols, DSOs, and machines perf session: Periodically shrink symbols and DSOs during event
processing
perf test symbols: Add tests for symbol and DSO LRU shrinking
.../arch/powerpc/util/skip-callchain-idx.c | 2 +- tools/perf/arch/powerpc/util/sym-handling.c | 40 +- tools/perf/builtin-annotate.c | 15 +- tools/perf/builtin-c2c.c | 14 - tools/perf/builtin-diff.c | 5 +- tools/perf/builtin-kallsyms.c | 10 +- tools/perf/builtin-kmem.c | 66 +- tools/perf/builtin-kwork.c | 4 +- tools/perf/builtin-lock.c | 27 +- tools/perf/builtin-mem.c | 2 +- tools/perf/builtin-report.c | 14 - tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 18 +- tools/perf/builtin-timechart.c | 2 +- tools/perf/builtin-top.c | 29 +- tools/perf/builtin-trace.c | 12 +- tools/perf/tests/builtin-test.c | 1 - tools/perf/tests/dlfilter-test.c | 12 +- tools/perf/tests/dwarf-unwind.c | 2 +- tools/perf/tests/hists_common.c | 4 +- tools/perf/tests/hists_cumulate.c | 8 +- tools/perf/tests/hists_filter.c | 10 +- tools/perf/tests/hists_link.c | 16 +- tools/perf/tests/hists_output.c | 9 +- tools/perf/tests/kallsyms-split.c | 5 +- tools/perf/tests/pe-file-parsing.c | 1 + tools/perf/tests/symbols.c | 289 +++++- tools/perf/tests/vmlinux-kallsyms.c | 191 ++-- tools/perf/ui/browsers/annotate.c | 11 +- tools/perf/ui/browsers/hists.c | 17 +- tools/perf/ui/browsers/map.c | 87 +- tools/perf/ui/gtk/annotate.c | 7 +- tools/perf/ui/stdio/hist.c | 6 +- tools/perf/util/addr_location.c | 4 + .../perf/util/annotate-arch/annotate-alpha.c | 2 +- .../util/annotate-arch/annotate-loongarch.c | 8 +- tools/perf/util/annotate-arch/annotate-s390.c | 2 +- tools/perf/util/annotate-arch/annotate-x86.c | 14 +- tools/perf/util/annotate-data.c | 14 +- tools/perf/util/annotate.c | 211 ++++- tools/perf/util/annotate.h | 6 +- tools/perf/util/auxtrace.c | 111 ++- tools/perf/util/block-info.c | 13 +- tools/perf/util/bpf_lock_contention.c | 39 +- tools/perf/util/c2c-function.c | 25 +- tools/perf/util/call-path.c | 15 +- tools/perf/util/callchain.c | 19 +- tools/perf/util/capstone.c | 21 +- tools/perf/util/data-convert-json.c | 4 +- tools/perf/util/db-export.c | 76 +- tools/perf/util/db-export.h | 6 + tools/perf/util/debug.c | 3 +- tools/perf/util/disasm.c | 57 +- tools/perf/util/dlfilter.c | 12 +- tools/perf/util/dso.c | 236 +++-- tools/perf/util/dso.h | 85 +- tools/perf/util/dsos.c | 100 +++ tools/perf/util/dsos.h | 12 +- tools/perf/util/event.c | 12 +- tools/perf/util/evsel_fprintf.c | 2 +- tools/perf/util/hist.c | 30 +- tools/perf/util/intel-pt.c | 74 +- tools/perf/util/libbfd.c | 5 +- tools/perf/util/libdw.c | 5 +- tools/perf/util/llvm.c | 6 +- tools/perf/util/lock-contention.c | 8 +- tools/perf/util/machine.c | 143 ++- tools/perf/util/machine.h | 2 + tools/perf/util/map.c | 53 +- tools/perf/util/map.h | 18 +- tools/perf/util/map_symbol.c | 4 +- tools/perf/util/maps.c | 3 +- tools/perf/util/probe-event.c | 135 +-- tools/perf/util/python.c | 20 +- .../util/scripting-engines/trace-event-perl.c | 8 +- .../scripting-engines/trace-event-python.c | 38 +- tools/perf/util/session.c | 31 + tools/perf/util/session.h | 13 + tools/perf/util/sort.c | 44 +- tools/perf/util/srcline.c | 19 +- tools/perf/util/symbol-elf.c | 22 +- tools/perf/util/symbol.c | 845 ++++++++++++------ tools/perf/util/symbol.h | 203 ++++- tools/perf/util/symbol_conf.h | 1 - tools/perf/util/symbol_fprintf.c | 14 +- tools/perf/util/thread-stack.c | 18 +- tools/perf/util/threads.c | 17 +- tools/perf/util/unwind-libdw.c | 4 +- tools/perf/util/unwind-libunwind.c | 2 +- 89 files changed, 2644 insertions(+), 1192 deletions(-)
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 1/7 perf symbol: Add accessor functions for struct symbol fields
To prepare for adding reference count checking (DECLARE_RC_STRUCT) to struct symbol, add inline accessor functions for struct symbol fields (symbol__start, symbol__set_start, symbol__end, symbol__set_end, symbol__name, symbol__namelen, symbol__arch_sym, symbol__set_arch_sym) and convert direct member accesses across tools/perf.
--- .../arch/powerpc/util/skip-callchain-idx.c | 2 +- tools/perf/arch/powerpc/util/sym-handling.c | 8 +- tools/perf/builtin-annotate.c | 4 +- tools/perf/builtin-diff.c | 5 +- tools/perf/builtin-kallsyms.c | 8 +- tools/perf/builtin-kmem.c | 25 ++-- tools/perf/builtin-kwork.c | 4 +- tools/perf/builtin-lock.c | 12 +- tools/perf/builtin-mem.c | 2 +- tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 14 +-- tools/perf/builtin-timechart.c | 2 +- tools/perf/builtin-top.c | 25 ++-- tools/perf/builtin-trace.c | 11 +- tools/perf/tests/dlfilter-test.c | 4 +- tools/perf/tests/dwarf-unwind.c | 2 +- tools/perf/tests/hists_common.c | 4 +- tools/perf/tests/hists_cumulate.c | 4 +- tools/perf/tests/hists_output.c | 2 +- tools/perf/tests/symbols.c | 4 +- tools/perf/tests/vmlinux-kallsyms.c | 34 +++--- tools/perf/ui/browsers/annotate.c | 11 +- tools/perf/ui/browsers/hists.c | 10 +- tools/perf/ui/browsers/map.c | 9 +- tools/perf/ui/gtk/annotate.c | 7 +- tools/perf/ui/stdio/hist.c | 3 +- .../util/annotate-arch/annotate-loongarch.c | 4 +- tools/perf/util/annotate-arch/annotate-x86.c | 14 +-- tools/perf/util/annotate-data.c | 14 ++- tools/perf/util/annotate.c | 73 +++++++----- tools/perf/util/auxtrace.c | 14 +-- tools/perf/util/block-info.c | 7 +- tools/perf/util/bpf_lock_contention.c | 20 ++-- tools/perf/util/c2c-function.c | 12 +- tools/perf/util/callchain.c | 15 ++- tools/perf/util/capstone.c | 20 ++-- tools/perf/util/data-convert-json.c | 4 +- tools/perf/util/db-export.c | 2 +- tools/perf/util/debug.c | 3 +- tools/perf/util/disasm.c | 45 ++++---- tools/perf/util/dlfilter.c | 12 +- tools/perf/util/dso.c | 4 +- tools/perf/util/event.c | 4 +- tools/perf/util/evsel_fprintf.c | 2 +- tools/perf/util/hist.c | 17 +-- tools/perf/util/intel-pt.c | 8 +- tools/perf/util/libbfd.c | 5 +- tools/perf/util/llvm.c | 6 +- tools/perf/util/lock-contention.c | 2 +- tools/perf/util/machine.c | 34 +++--- tools/perf/util/map.c | 6 +- tools/perf/util/map.h | 2 +- tools/perf/util/probe-event.c | 40 +++---- tools/perf/util/python.c | 8 +- .../util/scripting-engines/trace-event-perl.c | 8 +- .../scripting-engines/trace-event-python.c | 28 ++--- tools/perf/util/sort.c | 42 +++---- tools/perf/util/srcline.c | 13 ++- tools/perf/util/symbol-elf.c | 11 +- tools/perf/util/symbol.c | 107 ++++++++++-------- tools/perf/util/symbol.h | 42 ++++++- tools/perf/util/symbol_fprintf.c | 14 +-- tools/perf/util/thread-stack.c | 8 +- tools/perf/util/unwind-libdw.c | 2 +- 64 files changed, 504 insertions(+), 390 deletions(-)
tools/perf/arch/powerpc/util/skip-callchain-idx.c +1 -1
@@ -239,7 +239,7 @@ int arch_skip_callchain_idx(struct thread *thread, struct ip_callchain *chain) rc =check_return_addr (dso,map__map_ip (al.map, ip));pr_debug ("[DSO %s, sym %s, ip 0x%" PRIx64"] rc %d\n" , -dso__long_name (dso), al.sym->name, ip, rc); +dso__long_name (dso),symbol__name (al.sym), ip, rc);if (rc ==0 ) {/*
tools/perf/arch/powerpc/util/sym-handling.c +4 -4
@@ -13,7 +13,7 @@int arch__choose_best_symbol (struct symbol *syma,struct symbol *symb__maybe_unused ) { -char *sym = syma->name; +const char *sym =symbol__name (syma);#if !defined(_CALL_ELF) || _CALL_ELF !=2 /* Skip over any initial dot */
@@ -69,7 +69,7 @@ const char *arch__normalize_symbol_name(const char *name)#ifdef HAVE_LIBELF_SUPPORT void arch__sym_update (struct symbol *s,GElf_Sym *sym) { - s->arch_sym = sym->st_other; +symbol__set_arch_sym (s, sym->st_other); }#endif
@@ -102,7 +102,7 @@ void arch__fix_tev_from_maps(struct perf_probe_event *pev,return ; } - lep_offset =PPC64_LOCAL_ENTRY_OFFSET (sym->arch_sym); + lep_offset =PPC64_LOCAL_ENTRY_OFFSET (symbol__arch_sym (sym));if (map__dso (map)->symtab_type ==DSO_BINARY_TYPE__KALLSYMS ) tev->point.offset +=PPC64LE_LEP_OFFSET ;
@@ -131,7 +131,7 @@ void arch__post_process_probe_trace_events(struct perf_probe_event *pev,for (i =0 ; i < ntevs; i++) { tev = &pev->tevs[i];map__for_each_symbol (map, sym, tmp) { -if (map__unmap_ip (map, sym->start) == tev->point.address) { +if (map__unmap_ip (map,symbol__start (sym)) == tev->point.address) {arch__fix_tev_from_maps (pev, tev, map, sym);break ; }
tools/perf/builtin-annotate.c +2 -2
@@ -252,7 +252,7 @@ static int add_sample(struct perf_sample *sample,if ((!ann->has_br_stack || !has_annotation (ann)) && ann->sym_hist_filter !=NULL && (al->sym ==NULL || -strcmp (ann->sym_hist_filter, al->sym->name) !=0 )) { +strcmp (ann->sym_hist_filter,symbol__name (al->sym)) !=0 )) {/* We're only interested in a symbol named sym_hist_filter */ /* * FIXME: why isn't this done in the symbol_filter when loading
@@ -423,7 +423,7 @@ static void hists__find_annotations(struct hists *hists, goto find_next;if (ann->sym_hist_filter && - (strcmp (he->ms.sym->name, ann->sym_hist_filter) !=0 )) + (strcmp (symbol__name (he->ms.sym), ann->sym_hist_filter) !=0 )) goto find_next;if (ann->min_percent) {
tools/perf/builtin-diff.c +3 -2
@@ -1358,10 +1358,11 @@ static int cycles_printf(struct hist_entry *he, struct hist_entry *pair, bi = block_he->block_info; - start_line =map__srcline (he->ms.map, bi->sym->start + bi->start, + start_line =map__srcline (he->ms.map, +symbol__start (bi->sym) + bi->start, he->ms.sym); - end_line =map__srcline (he->ms.map, bi->sym->start + bi->end, + end_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->end, he->ms.sym); if (start_line != SRCLINE_UNKNOWN &&
tools/perf/builtin-kallsyms.c +5 -3
@@ -48,9 +48,11 @@ static int __cmd_kallsyms(int argc, const char **argv) dso =map__dso (map);printf ("%s: %s %s %#" PRIx64"-%#" PRIx64" (%#" PRIx64"-%#" PRIx64")\n" , - symbol->name,dso__short_name (dso),dso__long_name (dso), -map__unmap_ip (map, symbol->start),map__unmap_ip (map, symbol->end), - symbol->start, symbol->end); +symbol__name (symbol),dso__short_name (dso), +dso__long_name (dso), +map__unmap_ip (map,symbol__start (symbol)), +map__unmap_ip (map,symbol__end (symbol)), +symbol__start (symbol),symbol__end (symbol)); } out:machine__delete (machine);
tools/perf/builtin-kmem.c +13 -12
@@ -310,7 +310,7 @@ static struct rb_root page_caller_sorted;struct alloc_func {u64 start;u64 end; -char *name; +const char *name; };static int nr_alloc_funcs;
@@ -368,7 +368,7 @@ static int build_alloc_func_list(void) }map__for_each_symbol (kernel_map, sym, node) { -if (regexec (&alloc_func_regex, sym->name,0 ,NULL ,0 )) +if (regexec (&alloc_func_regex,symbol__name (sym),0 ,NULL ,0 ))continue ; func = realloc(alloc_func_list,
@@ -376,10 +376,10 @@ static int build_alloc_func_list(void)if (func ==NULL )return -ENOMEM ; -pr_debug ("alloc func: %s\n" , sym->name); - func[nr_alloc_funcs].start = sym->start; - func[nr_alloc_funcs].end = sym->end; - func[nr_alloc_funcs].name = sym->name; +pr_debug ("alloc func: %s\n" ,symbol__name (sym)); + func[nr_alloc_funcs].start =symbol__start (sym); + func[nr_alloc_funcs].end =symbol__end (sym); + func[nr_alloc_funcs].name =symbol__name (sym); alloc_func_list = func; nr_alloc_funcs++;
@@ -1049,8 +1049,9 @@ static void __print_slab_result(struct rb_root *root, addr = data->ptr;if (sym !=NULL ) -snprintf (buf,sizeof (buf),"%s+%" PRIx64"" , sym->name, - addr -map__unmap_ip (map, sym->start)); +snprintf (buf,sizeof (buf),"%s+%" PRIx64"" , +symbol__name (sym), + addr -map__unmap_ip (map,symbol__start (sym)));else snprintf (buf,sizeof (buf),"%#" PRIx64"" , addr);printf (" %-34s |" , buf);
@@ -1105,12 +1106,12 @@ static void __print_page_alloc_result(struct perf_session *session, int n_lines)struct symbol *sym;struct map *map;char buf[32 ]; -char *caller = buf; +const char *caller = buf; data = rb_entry(next, struct page_stat, node); sym =machine__find_kernel_symbol (machine, data->callsite, &map);if (sym) - caller = sym->name; + caller =symbol__name (sym);else scnprintf (buf,sizeof (buf),"%" PRIx64, data->callsite);
@@ -1147,12 +1148,12 @@ static void __print_page_caller_result(struct perf_session *session, int n_linesstruct symbol *sym;struct map *map;char buf[32 ]; -char *caller = buf; +const char *caller = buf; data = rb_entry(next, struct page_stat, node); sym =machine__find_kernel_symbol (machine, data->callsite, &map);if (sym) - caller = sym->name; + caller =symbol__name (sym);else scnprintf (buf,sizeof (buf),"%" PRIx64, data->callsite);
tools/perf/builtin-kwork.c +2 -2
@@ -768,8 +768,8 @@ static void timehist_save_callchain(struct perf_kwork *kwork, sym = node->ms.sym;if (sym) { -if (!strcmp (sym->name,"__softirqentry_text_start" ) || - !strcmp (sym->name,"__do_softirq" )) +if (!strcmp (symbol__name (sym),"__softirqentry_text_start" ) || + !strcmp (symbol__name (sym),"__do_softirq" ))symbol__set_ignore (sym, true); }
tools/perf/builtin-lock.c +7 -5
@@ -826,12 +826,13 @@ static int get_symbol_name_offset(struct map *map, struct symbol *sym, u64 ip,return 0 ; } - offset =map__map_ip (map, ip) - sym->start; + offset =map__map_ip (map, ip) -symbol__start (sym);if (offset) -return scnprintf (buf, size,"%s+%#lx" , sym->name, offset); +return scnprintf (buf, size,"%s+%#lx" ,symbol__name (sym), + offset);else -return strlcpy (buf, sym->name, size); +return strlcpy (buf,symbol__name (sym), size); }static int lock_contention_caller (struct perf_sample *sample,char *buf,int size)
@@ -1005,7 +1006,8 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)return -ENOMEM ; } - addrs[filters.nr_addrs++] =map__unmap_ip (kmap, sym->start); + addrs[filters.nr_addrs++] =map__unmap_ip (kmap, +symbol__start (sym)); filters.addrs = addrs; } }
@@ -1019,7 +1021,7 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)case LOCK_AGGR_ADDR : sym =machine__find_kernel_symbol (machine, key, &kmap);if (sym) - name = sym->name; + name =symbol__name (sym);break ;case LOCK_AGGR_CALLER : name = buf;
tools/perf/builtin-mem.c +1 -1
@@ -249,7 +249,7 @@ dump_raw_samples(const struct perf_tool *tool, sample->data_src, symbol_conf.field_sep, dso ?dso__long_name (dso) :"???" , - al.sym ? al.sym->name : "???"); + al.sym ? symbol__name(al.sym) : "???"); out_put:addr_location__exit (&al);return 0 ;
tools/perf/builtin-sched.c +3 -3
@@ -2629,9 +2629,9 @@ static void save_task_callchain(struct perf_sched *sched, sym = node->ms.sym;if (sym) { -if (!strcmp (sym->name,"schedule" ) || - !strcmp (sym->name,"__schedule" ) || - !strcmp (sym->name,"preempt_schedule" )) +if (!strcmp (symbol__name (sym),"schedule" ) || + !strcmp (symbol__name (sym),"__schedule" ) || + !strcmp (symbol__name (sym),"preempt_schedule" ))symbol__set_ignore (sym, true); }
tools/perf/builtin-script.c +7 -7
@@ -1404,7 +1404,7 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,addr_location__init (&al);thread__find_map (thread, cpumode, addr, &al); -if ((*lastsym) && al.addr >= (*lastsym)->start && al.addr < (*lastsym)->end) +if ((*lastsym) && al.addr >=symbol__start ((*lastsym)) && al.addr <symbol__end ((*lastsym))) goto out; al.cpu = cpu;
@@ -1415,11 +1415,11 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,if (!al.sym) goto out; -if (al.addr < al.sym->end) - off = al.addr - al.sym->start; +if (al.addr <symbol__end (al.sym)) + off = al.addr -symbol__start (al.sym);else - off = al.addr -map__start (al.map) - al.sym->start; - printed +=fprintf (fp,"\t%s" , al.sym->name); + off = al.addr -map__start (al.map) -symbol__start (al.sym); + printed +=fprintf (fp,"\t%s" ,symbol__name (al.sym));if (off) printed +=fprintf (fp,"%+d" , off); printed +=fprintf (fp,":" );
@@ -1638,7 +1638,7 @@ static const char *resolve_branch_sym(struct perf_sample *sample,if (!addr_al->thread)thread__resolve (thread, addr_al, sample);if (addr_al->sym) - name = addr_al->sym->name; + name =symbol__name (addr_al->sym);else *ip = sample->addr; }else {
@@ -1646,7 +1646,7 @@ static const char *resolve_branch_sym(struct perf_sample *sample, } }else if (sample->flags & (PERF_IP_FLAG_RETURN |PERF_IP_FLAG_TRACE_END )) {if (al->sym) - name = al->sym->name; + name =symbol__name (al->sym);else *ip = sample->ip; }
tools/perf/builtin-timechart.c +1 -1
@@ -565,7 +565,7 @@ static char *cat_backtrace(struct perf_sample *sample,addr_location__init (&tal); tal.filtered =0 ;if (thread__find_symbol (al.thread, cpumode, ip, &tal)) -fprintf (f,"..... %016" PRIx64" %s\n" , ip, tal.sym->name); +fprintf (f,"..... %016" PRIx64" %s\n" , ip,symbol__name (tal.sym));else fprintf (f,"..... %016" PRIx64"\n" , ip);
tools/perf/builtin-top.c +14 -11
@@ -133,8 +133,8 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he)* We can't annotate with just /proc/kallsyms */ if (dso__symtab_type (dso) ==DSO_BINARY_TYPE__KALLSYMS && !dso__is_kcore (dso)) { -pr_err ("Can't annotate %s: No vmlinux file was found in the " -"path\n" , sym->name); +pr_err ("Can't annotate %s: No vmlinux file was found in the path\n" , +symbol__name (sym));sleep (1 );return -1 ; }
@@ -145,7 +145,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he)if (!symbol__hists (sym,evlist__nr_entries (top->evlist))) {annotation__unlock (notes);pr_err ("Not enough memory for annotating '%s' symbol!\n" , - sym->name); +symbol__name (sym));sleep (1 );return err; }
@@ -156,7 +156,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he) }else {char msg[BUFSIZ ];symbol__strerror_disassemble (&he->ms, err, msg,sizeof (msg)); -pr_err ("Couldn't annotate %s: %s\n" , sym->name, msg); +pr_err ("Couldn't annotate %s: %s\n" ,symbol__name (sym), msg); }annotation__unlock (notes);
@@ -186,9 +186,11 @@ static void ui__warn_map_erange(struct map *map, struct symbol *sym, u64 ip)"Not all samples will be on the annotation output.\n\n" "Please report to [email protected]\n" , ip,dso__long_name (dso),dso__symtab_origin (dso), -map__start (map),map__end (map), sym->start, sym->end, +map__start (map),map__end (map),symbol__start (sym), +symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' : -symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , sym->name, +symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , +symbol__name (sym), err ?"[unknown]" : uts.machine, err ? "[unknown]" : uts.release, perf_version_string);if (use_browser <=0 )
@@ -232,7 +234,7 @@ static void perf_top__record_precise_ip(struct perf_top *top,ui__warn_map_erange (he->ms.map, sym, ip); else if (err == -ENOMEM) {pr_err ("Not enough memory for annotating '%s' symbol!\n" , - sym->name); +symbol__name (sym));sleep (1 ); }
@@ -263,7 +265,8 @@ static void perf_top__show_details(struct perf_top *top)if (notes->src ==NULL ) goto out_unlock; -printf ("Showing %s for %s\n" ,evsel__name (top->sym_evsel), symbol->name); +printf ("Showing %s for %s\n" ,evsel__name (top->sym_evsel), +symbol__name (symbol));printf (" Events Pcnt (>=%d%%)\n" , annotate_opts.min_pcnt); more =hist_entry__annotate_printf (he, top->sym_evsel);
@@ -418,7 +421,7 @@ static void perf_top__prompt_symbol(struct perf_top *top, const char *msg) next =rb_first_cached (&hists->entries);while (next) { n = rb_entry(next, struct hist_entry, rb_node); -if (n->ms.sym && !strcmp (buf, n->ms.sym->name)) { +if (n->ms.sym && !strcmp (buf,symbol__name (n->ms.sym))) { found = n;break ; }
@@ -437,11 +440,11 @@ static void perf_top__prompt_symbol(struct perf_top *top, const char *msg)static void perf_top__print_mapped_keys (struct perf_top *top) { -char *name =NULL ; +const char *name =NULL ;if (top->sym_filter_entry) {struct symbol *sym = top->sym_filter_entry->ms.sym; - name = sym->name; + name =symbol__name (sym); }fprintf (stdout,"\nMapped keys:\n" );
tools/perf/builtin-trace.c +6 -5
@@ -768,15 +768,16 @@ size_t syscall_arg__scnprintf_ksym(char *bf, size_t size, struct syscall_arg *ar arg->val, &map);if (sym) { -u64 start =map__unmap_ip (map, sym->start); +u64 start =map__unmap_ip (map,symbol__start (sym));u64 offset = arg->val - start;size_t printed;if (offset ==0 ) - printed =scnprintf (bf, size,"%s" , sym->name); + printed =scnprintf (bf, size,"%s" , +symbol__name (sym));else printed =scnprintf (bf, size,"%s+0x%" PRIx64, - sym->name, offset); +symbol__name (sym), offset);map__put (map);return printed; }
@@ -3652,8 +3653,8 @@ static void print_location(FILE *f, struct perf_sample *sample,fprintf (f,"%s@" ,dso__long_name (map__dso (al->map)));if ((verbose >0 || print_sym) && al->sym) -fprintf (f,"%s+0x%" PRIx64, al->sym->name, - al->addr - al->sym->start); +fprintf (f,"%s+0x%" PRIx64,symbol__name (al->sym), + al->addr -symbol__start (al->sym));else if (al->map)fprintf (f,"0x%" PRIx64, al->addr);else
tools/perf/tests/dlfilter-test.c +2 -2
@@ -274,11 +274,11 @@ static int get_ip_addr(struct test_data *td) sym =map__find_symbol_by_name (map,"foo" );if (sym) - td->foo = sym->start; + td->foo =symbol__start (sym); sym =map__find_symbol_by_name (map,"bar" );if (sym) - td->bar = sym->start; + td->bar =symbol__start (sym);map__put (map);
tools/perf/tests/dwarf-unwind.c +1 -1
@@ -56,7 +56,7 @@ int test__dwarf_unwind(struct test_suite *test, int subtest);static int unwind_entry (struct unwind_entry *entry,void *arg) {unsigned long *cnt = (unsigned long *) arg; -char *symbol = entry->ms.sym ? entry->ms.sym->name :NULL ; +const char *symbol = entry->ms.sym ?symbol__name (entry->ms.sym) :NULL ;static const char *funcs[MAX_STACK ] = {"test__arch_unwind_sample" ,"test_dwarf_unwind__thread" ,
tools/perf/tests/hists_common.c +2 -2
@@ -184,7 +184,7 @@ void print_hists_in(struct hists *hists)pr_info ("%2d: entry: %-8s [%-8s] %20s: period = %" PRIu64"\n" , i,thread__comm_str (he->thread),dso__short_name (dso), - he->ms.sym->name, he->stat.period); +symbol__name (he->ms.sym), he->stat.period); } i++;
@@ -213,7 +213,7 @@ void print_hists_out(struct hists *hists)pr_info ("%2d: entry: %8s:%5d [%-8s] %20s: period = %" PRIu64"/%" PRIu64"\n" , i,thread__comm_str (he->thread),thread__tid (he->thread),dso__short_name (dso), - he->ms.sym->name, he->stat.period, +symbol__name (he->ms.sym), he->stat.period, he->stat_acc ? he->stat_acc->period :0 ); }
tools/perf/tests/hists_cumulate.c +2 -2
@@ -164,11 +164,11 @@ typedef int (*test_fn_t)(struct evsel *, struct machine *);#define COMM (he) (thread__comm_str(he->thread))#define DSO (he) (dso__short_name(map__dso(he->ms.map))) -#define SYM (he) (he->ms.sym->name) +#define SYM (he) (symbol__name(he->ms.sym))#define CPU (he) (he->cpu)#define DEPTH (he) (he->callchain->max_depth)#define CDSO (cl) (dso__short_name(map__dso(cl->ms.map))) -#define CSYM (cl) (cl->ms.sym->name) +#define CSYM (cl) (symbol__name(cl->ms.sym))struct result {u64 children;
tools/perf/tests/hists_output.c +1 -1
@@ -129,7 +129,7 @@ typedef int (*test_fn_t)(struct evsel *, struct machine *);#define COMM (he) (thread__comm_str(he->thread))#define DSO (he) (dso__short_name(map__dso(he->ms.map))) -#define SYM (he) (he->ms.sym->name) +#define SYM (he) (symbol__name(he->ms.sym))#define CPU (he) (he->cpu)#define PID (he) (thread__tid(he->thread))
tools/perf/tests/symbols.c +2 -2
@@ -129,14 +129,14 @@ static int test_dso(struct dso *dso)continue ;/* Check for overlapping function symbols */ -if (last_sym && sym->start < last_sym->end) { +if (last_sym &&symbol__start (sym) <symbol__end (last_sym)) {pr_debug ("Overlapping symbols:\n" );symbol__fprintf (last_sym, stderr);symbol__fprintf (sym, stderr); ret =TEST_FAIL ; }/* Check for zero-length function symbol */ -if (sym->start == sym->end) { +if (symbol__start (sym) ==symbol__end (sym)) {pr_debug ("Zero-length symbol:\n" );symbol__fprintf (sym, stderr); ret =TEST_FAIL ;
tools/perf/tests/vmlinux-kallsyms.c +20 -14
@@ -292,18 +292,20 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused sym = rb_entry(nd, struct symbol, rb_node); -if (sym->start == sym->end) +if (symbol__start (sym) ==symbol__end (sym))continue ; - mem_start =map__unmap_ip (args.vmlinux_map, sym->start); - mem_end =map__unmap_ip (args.vmlinux_map, sym->end); + mem_start =map__unmap_ip (args.vmlinux_map, +symbol__start (sym)); + mem_end =map__unmap_ip (args.vmlinux_map,symbol__end (sym)); first_pair =machine__find_kernel_symbol (&args.kallsyms, mem_start,NULL ); pair = first_pair; -if (pair &&UM (pair->start) == mem_start) { +if (pair &&UM (symbol__start (pair)) == mem_start) { next_pair: -if (arch__compare_symbol_names (sym->name, pair->name) ==0 ) { +if (arch__compare_symbol_names (symbol__name (sym), +symbol__name (pair)) ==0 ) {/* * kallsyms don't have the symbol end, so we * set that by using the next symbol start - 1,
@@ -313,11 +315,12 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused* off the real size. More than that and we * _really_ have a problem. */ -s64 skew = mem_end -UM (pair->end); +s64 skew = mem_end -UM (symbol__end (pair));if (llabs (skew) >= page_size)pr_debug ("WARN: %#" PRIx64": diff end addr for %s v: %#" PRIx64" k: %#" PRIx64"\n" , - mem_start, sym->name, mem_end, -UM (pair->end)); + mem_start,symbol__name (sym), + mem_end, +UM (symbol__end (pair)));/* * Do not count this as a failure, because we
@@ -328,16 +331,19 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unusedcontinue ; }else { pair =machine__find_kernel_symbol_by_name (&args.kallsyms, - sym->name,NULL ); +symbol__name (sym), +NULL );if (pair) { -if (UM (pair->start) == mem_start) +if (UM (symbol__start (pair)) == mem_start) goto next_pair;pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , - mem_start, sym->name, pair->name); + mem_start,symbol__name (sym), +symbol__name (pair)); }else {pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , - mem_start, sym->name, first_pair->name); + mem_start,symbol__name (sym), +symbol__name (first_pair)); }continue ;
@@ -348,14 +354,14 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused* such as __indirect_thunk_end. */ continue ; - }else if (is_ignored_symbol (sym->name,symbol__type (sym))) { + }else if (is_ignored_symbol (symbol__name (sym),symbol__type (sym))) {/* * Ignore hidden symbols, see scripts/kallsyms.c for the details */ continue ; }else {pr_debug ("ERR : %#" PRIx64": %s not on kallsyms\n" , - mem_start, sym->name); + mem_start,symbol__name (sym)); } err =-1 ;
tools/perf/ui/browsers/annotate.c +6 -5
@@ -196,7 +196,7 @@ static void annotate_browser__draw_current_jump(struct ui_browser *browser)int diff =0 ;/* PLT symbols contain external offsets */ -if (strstr (sym->name,"@plt" )) +if (strstr (symbol__name (sym),"@plt" ))return ;if (!disasm_line__is_valid_local_jump (cursor, sym))
@@ -554,7 +554,8 @@ static void ui_browser__init_asm_mode(struct ui_browser *browser)static int sym_title (struct symbol *sym,struct map *map,char *title,size_t sz,int percent_type) { -return snprintf (title, sz,"%s %s [Percent: %s] %s" , sym->name, +return snprintf (title, sz,"%s %s [Percent: %s] %s" , +symbol__name (sym),dso__long_name (map__dso (map)),percent_type_str (percent_type), annotate_opts.code_with_type ?"[Type]" :"" );
@@ -602,7 +603,7 @@ static bool annotate_browser__callq(struct annotate_browser *browser,if (!symbol__hists (dl->ops.target.sym,evlist__nr_entries (evsel->evlist))) {annotation__unlock (notes);ui__warning ("Not enough memory for annotating '%s' symbol!\n" , - dl->ops.target.sym->name); +symbol__name (dl->ops.target.sym));return true; }
@@ -876,7 +877,7 @@ static void annotate_browser__symbol_annotate_error(struct annotate_browser *brodso__set_annotate_warned (dso);symbol__strerror_disassemble (ms, err, msg,sizeof (msg)); -ui__error ("Couldn't annotate %s:\n%s" , sym->name, msg); +ui__error ("Couldn't annotate %s:\n%s" ,symbol__name (sym), msg); } static int annotate_browser__run(struct annotate_browser *browser,
@@ -1255,7 +1256,7 @@ int __hist_entry__tui_annotate(struct hist_entry *he, struct map_symbol *ms,*/ if (al_addr !=NO_ADDR ) {struct annotation_line *al =annotated_source__get_line (notes->src, - al_addr - sym->start); + al_addr -symbol__start (sym)); browser.selection = al; }
tools/perf/ui/browsers/hists.c +5 -5
@@ -2548,7 +2548,7 @@ add_annotate_opt(struct popup_action *act, char **optstr,if (ms->sym ==NULL ||symbol__annotation (ms->sym)->src ==NULL )return 0 ; -if (asprintf (optstr,"Annotate %s" , ms->sym->name) <0 ) +if (asprintf (optstr,"Annotate %s" ,symbol__name (ms->sym)) <0 )return 0 ;map_symbol__copy (&act->ms, ms);
@@ -2748,7 +2748,7 @@ do_run_script(struct hist_browser *browser,if (act->ms.thread) len +=strlen (thread__comm_str (act->ms.thread));else if (act->ms.sym) - len +=strlen (act->ms.sym->name); + len +=strlen (symbol__name (act->ms.sym)); script_opt =malloc (len);if (!script_opt)return -1 ;
@@ -2759,7 +2759,7 @@ do_run_script(struct hist_browser *browser, thread__comm_str(act->ms.thread)); }else if (act->ms.sym) { n =scnprintf (script_opt, len," -S %s " , - act->ms.sym->name); +symbol__name (act->ms.sym)); }if (act->time) {
@@ -2804,7 +2804,7 @@ add_script_opt_2(struct popup_action *act, char **optstr,return 0 ; }else if (sym) {if (asprintf (optstr,"Run scripts for samples of symbol [%s]%s" , - sym->name, tstr) <0 ) +symbol__name (sym), tstr) <0 )return 0 ; }else {if (asprintf (optstr,"Run scripts for all samples%s" , tstr) <0 )
@@ -3208,7 +3208,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *hui_browser__warning (&browser->b, delay_secs *2 ,"No samples for the \"%s\" symbol.\n\n" "Probably appeared just in a callchain" , - browser->selection->sym->name); +symbol__name (browser->selection->sym));continue ; }
tools/perf/ui/browsers/map.c +5 -4
@@ -31,12 +31,13 @@ static void map_browser__write(struct ui_browser *browser, void *nd, int row)ui_browser__set_percent_color (browser,0 , current_entry);ui_browser__printf (browser,"%*" PRIx64" %*" PRIx64" %c " , - mb->addrlen, sym->start, mb->addrlen, sym->end, + mb->addrlen,symbol__start (sym), mb->addrlen, +symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' ); width = browser->width - ((mb->addrlen *2 ) +4 );if (width >0 ) -ui_browser__write_nstring (browser, sym->name, width); +ui_browser__write_nstring (browser,symbol__name (sym), width); }/* FIXME uber-kludgy, see comment on cmd_report... */
@@ -122,8 +123,8 @@ int map__browse(struct map *map)for (nd =rb_first (mb.b.entries); nd; nd =rb_next (nd)) { struct symbol *pos = rb_entry(nd, struct symbol, rb_node); -if (maxaddr < pos->end) - maxaddr = pos->end; +if (maxaddr <symbol__end (pos)) + maxaddr =symbol__end (pos);if (verbose >0 ) {u32 *idx =symbol__browser_index (pos); *idx = mb.b.nr_entries;
tools/perf/ui/gtk/annotate.c +4 -3
@@ -67,7 +67,7 @@ static int perf_gtk__get_percent(char *buf, size_t size, struct symbol *sym,static int perf_gtk__get_offset (char *buf,size_t size,struct map_symbol *ms,struct disasm_line *dl) { -u64 start =map__rip_2objdump (ms->map, ms->sym->start); +u64 start =map__rip_2objdump (ms->map,symbol__start (ms->sym));strcpy (buf,"" );
@@ -192,7 +192,8 @@ static int symbol__gtk_annotate(struct map_symbol *ms, struct evsel *evsel,dso__set_annotate_warned (dso);symbol__strerror_disassemble (ms, err, msg,sizeof (msg)); -ui__error ("Couldn't annotate %s: %s\n" , sym->name, msg); +ui__error ("Couldn't annotate %s: %s\n" ,symbol__name (sym), + msg);return -1 ; }
@@ -234,7 +235,7 @@ static int symbol__gtk_annotate(struct map_symbol *ms, struct evsel *evsel, } scrolled_window =gtk_scrolled_window_new (); - tab_label =gtk_label_new (sym->name); + tab_label =gtk_label_new (symbol__name (sym)); gtk_scrolled_window_set_policy(GTK_SCROLLED_WINDOW(scrolled_window), GTK_POLICY_AUTOMATIC,
tools/perf/ui/stdio/hist.c +1 -2
@@ -100,13 +100,12 @@ static struct callchain_list rem_hits;static void init_rem_hits (void ) { - rem_sq_bracket =malloc (sizeof (*rem_sq_bracket) +6 ); + rem_sq_bracket =symbol__new (0 ,0 ,0 ,0 ,"[...]" );if (!rem_sq_bracket) {fprintf (stderr,"Not enough memory to display remaining hits\n" );return ; } -strcpy (rem_sq_bracket->name,"[...]" ); rem_hits.ms.sym = rem_sq_bracket; }
tools/perf/util/annotate-arch/annotate-loongarch.c +2 -2
@@ -94,8 +94,8 @@ static int loongarch_jump__parse(const struct arch *arch, struct ins_operands *o ops->target.addr =strtoull (ops->raw,NULL ,16 ); target.addr =map__objdump_2mem (map, ops->target.addr); - start =map__unmap_ip (map, sym->start); - end =map__unmap_ip (map, sym->end); + start =map__unmap_ip (map,symbol__start (sym)); + end =map__unmap_ip (map,symbol__end (sym)); ops->target.outside = target.addr < start || target.addr >= end;
tools/perf/util/annotate-arch/annotate-x86.c +7 -7
@@ -239,7 +239,7 @@ static void update_insn_state_x86(struct type_state *state,/* Try to resolve the call target name */ if (func) - call_name = func->name; + call_name =symbol__name (func);else call_name = dl->ops.target.name;
@@ -254,7 +254,7 @@ static void update_insn_state_x86(struct type_state *state,/* Invalidate caller-saved registers after call */ call_addr =map__rip_2objdump (dloc->ms->map, - dloc->ms->sym->start + dl->al.offset); +symbol__start (dloc->ms->sym) + dl->al.offset);for (unsigned i =0 ; i <ARRAY_SIZE (state->regs); i++) {struct type_state_reg *reg = &state->regs[i];
@@ -286,7 +286,7 @@ static void update_insn_state_x86(struct type_state *state,int offset;const char *var_name =NULL ;struct map_symbol *ms = dloc->ms; -u64 ip = ms->sym->start + dl->al.offset; +u64 ip =symbol__start (ms->sym) + dl->al.offset;if (!has_reg_type (state, dst->reg1))return ;
@@ -487,7 +487,7 @@ static void update_insn_state_x86(struct type_state *state,return ;if (dloc->fb_cfa) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 pc =map__rip_2objdump (dloc->ms->map, ip);if (die_get_cfa (dloc->di->dbg, pc, &fbreg, &fboff) <0 )
@@ -504,7 +504,7 @@ static void update_insn_state_x86(struct type_state *state,if (dso__kernel (map__dso (dloc->ms->map)) && src->segment ==INSN_SEG_X86_GS && src->imm) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 var_addr;int offset;
@@ -653,7 +653,7 @@ static void update_insn_state_x86(struct type_state *state,/* Or check if it's a global variable */ else if (sreg == DWARF_REG_PC) {struct map_symbol *ms = dloc->ms; -u64 ip = ms->sym->start + dl->al.offset; +u64 ip =symbol__start (ms->sym) + dl->al.offset;u64 addr;int offset;
@@ -678,7 +678,7 @@ static void update_insn_state_x86(struct type_state *state,/* And check percpu access with base register */ else if (has_reg_type(state, sreg) && state->regs[sreg].kind == TSR_KIND_PERCPU_BASE) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 var_addr = src->offset;int offset;
tools/perf/util/annotate-data.c +8 -6
@@ -851,9 +851,10 @@ bool get_global_var_info(struct data_loc_info *dloc, u64 addr, sym =thread__find_symbol_fb (dloc->thread, dloc->cpumode, mem_addr, &al);if (sym) { - *var_name = sym->name; + *var_name =symbol__name (sym);/* Calculate type offset from the start of variable */ - *var_offset = mem_addr -map__unmap_ip (al.map, sym->start); + *var_offset = mem_addr -map__unmap_ip (al.map, +symbol__start (sym)); }else { *var_name =NULL ; }
@@ -1503,7 +1504,7 @@ static enum type_match_result find_data_type_insn(struct data_loc_info *dloc, bb->begin->al.offset, bb->end->al.offset);list_for_each_entry_from (dl, ¬es->src->source, al.node) { -u64 this_ip = sym->start + dl->al.offset; +u64 this_ip =symbol__start (sym) + dl->al.offset;u64 addr =map__rip_2objdump (dloc->ms->map, this_ip);/* Skip comment or debug info lines */
@@ -1582,8 +1583,8 @@ static enum type_match_result find_data_type_block(struct data_loc_info *dloc, }pr_debug_dtp ("cannot find a basic block from %" PRIx64" to %" PRIx64"\n" , - src_ip - dloc->ms->sym->start, - dst_ip - dloc->ms->sym->start); + src_ip -symbol__start (dloc->ms->sym), + dst_ip -symbol__start (dloc->ms->sym));continue ; }prepend_basic_blocks (&this_blocks, &basic_blocks);
@@ -1662,7 +1663,8 @@ static int find_data_type_die(struct data_loc_info *dloc, Dwarf_Die *type_die)pr_debug_dtp ("-----------------------------------------------------------\n" );pr_debug_dtp ("find data type for %s%#x(%s) at %s+%#" PRIx64"\n" , offset_sign, offset, buf, - dloc->ms->sym->name, dloc->ip - dloc->ms->sym->start); +symbol__name (dloc->ms->sym), + dloc->ip -symbol__start (dloc->ms->sym));/* * IP is a relative instruction address from the start of the map, as
tools/perf/util/annotate.c +42 -31
216 lines not on this page · raw message
tools/perf/util/auxtrace.c +7 -7
@@ -2707,7 +2707,7 @@ static bool dso_sym_match(struct symbol *sym, const char *name, int *cnt, int idx) {/* Same name, and global or the n'th found or any */ -return !arch__compare_symbol_names (name, sym->name) && +return !arch__compare_symbol_names (name,symbol__name (sym)) && ((!idx &&symbol__binding (sym) ==STB_GLOBAL ) || (idx >0 && ++*cnt == idx) || idx <0 );
@@ -2725,14 +2725,14 @@ static void print_duplicate_syms(struct dso *dso, const char *sym_name)while (sym) {if (dso_sym_match (sym, sym_name, &cnt,-1 )) {pr_err ("#%d\t0x%" PRIx64"\t%c\t%s\n" , - ++cnt, sym->start, + ++cnt,symbol__start (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , - sym->name); +symbol__name (sym)); near = true; }else if (near) { near = false; -pr_err ("\t\twhich is near\t\t%s\n" , sym->name); +pr_err ("\t\twhich is near\t\t%s\n" ,symbol__name (sym)); } sym =dso__next_symbol (sym); }
@@ -2755,7 +2755,7 @@ static int find_dso_sym(struct dso *dso, const char *sym_name, u64 *start,while (sym) {if (*start) {if (!*size) - *size = sym->start - *start; + *size =symbol__start (sym) - *start;if (idx >0 ) {if (*size)return 0 ;
@@ -2764,8 +2764,8 @@ static int find_dso_sym(struct dso *dso, const char *sym_name, u64 *start,return -EINVAL ; } }else if (dso_sym_match (sym, sym_name, &cnt, idx)) { - *start = sym->start; - *size = sym->end - sym->start; + *start =symbol__start (sym); + *size =symbol__end (sym) -symbol__start (sym); } sym =dso__next_symbol (sym); }
tools/perf/util/block-info.c +4 -3
@@ -84,7 +84,7 @@ int64_t __block_info__cmp(struct hist_entry *left, struct hist_entry *right)return 1 ; } - cmp =strcmp (bi_l->sym->name, bi_r->sym->name); + cmp =strcmp (symbol__name (bi_l->sym),symbol__name (bi_r->sym));if (cmp)return cmp;
@@ -305,10 +305,11 @@ static int block_range_entry(struct perf_hpp_fmt *fmt, struct perf_hpp *hpp, symbol_conf.addr2line_disable_warn = true; - start_line =map__srcline (he->ms.map, bi->sym->start + bi->start, + start_line =map__srcline (he->ms.map, +symbol__start (bi->sym) + bi->start, he->ms.sym); - end_line =map__srcline (he->ms.map, bi->sym->start + bi->end, + end_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->end, he->ms.sym); if (start_line != SRCLINE_UNKNOWN &&
tools/perf/util/bpf_lock_contention.c +12 -8
@@ -143,7 +143,8 @@ static void init_numa_data(struct lock_contention *con)"contig_page_data" , &kmap);if (sym) { - skel->rodata->contig_page_data_addr =map__unmap_ip (kmap, sym->start); + skel->rodata->contig_page_data_addr =map__unmap_ip (kmap, +symbol__start (sym));map__put (kmap);return ; }
@@ -159,7 +160,7 @@ static void init_numa_data(struct lock_contention *con)if (sym ==NULL )return ; - skel->rodata->node_data_addr =map__unmap_ip (kmap, sym->start); + skel->rodata->node_data_addr =map__unmap_ip (kmap,symbol__start (sym));map__put (kmap);/* get the number of online nodes using the last node number + 1 */
@@ -266,7 +267,8 @@ int lock_contention_prepare(struct lock_contention *con)continue ; } - addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap, sym->start); + addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap, +symbol__start (sym)); con->filters->addrs = addrs; } naddrs = con->filters->nr_addrs ?: has_mmap_lock;
@@ -288,7 +290,8 @@ int lock_contention_prepare(struct lock_contention *con)continue ; } - con->delays[i].addr =map__unmap_ip (kmap, sym->start); + con->delays[i].addr =map__unmap_ip (kmap, +symbol__start (sym)); } skel->rodata->lock_delay =1 ;bpf_map__set_max_entries (skel->maps.lock_delays, con->nr_delays);
@@ -608,7 +611,7 @@ static const char *lock_contention_get_name(struct lock_contention *con,/* global locks with symbols */ sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup, &kmap);if (sym) -return sym->name; +return symbol__name (sym);/* try semi-global locks collected separately */ if (!bpf_map_lookup_elem (lock_fd, &key->lock_addr_or_cgroup, &flags)) {
@@ -652,12 +655,13 @@ static const char *lock_contention_get_name(struct lock_contention *con,if (sym) {unsigned long offset; - offset =map__map_ip (kmap, addr) - sym->start; + offset =map__map_ip (kmap, addr) -symbol__start (sym);if (offset ==0 ) -return sym->name; +return symbol__name (sym); -snprintf (name_buf,sizeof (name_buf),"%s+%#lx" , sym->name, offset); +snprintf (name_buf,sizeof (name_buf),"%s+%#lx" , +symbol__name (sym), offset); }else {snprintf (name_buf,sizeof (name_buf),"%#lx" , (unsigned long )addr); }
tools/perf/util/c2c-function.c +6 -6
@@ -966,7 +966,7 @@ c2c_function_hists__level1_entry(struct symbol *sym,* true "function view" rather than a per-code-address view. */ u64 sym_start = (sym && detail_he->ms.map) ? -map__unmap_ip (detail_he->ms.map, sym->start) : detail_he->ip; +map__unmap_ip (detail_he->ms.map,symbol__start (sym)) : detail_he->ip; mi =mem_info__new ();if (!mi)
@@ -1058,8 +1058,8 @@ c2c_function_hists__level2_entry(struct c2c_hist_entry *level1_c2c,/* Key by the function symbol start so all code addresses collapse. */ level2_c2c =c2c_child_entry__alloc (level1_he, detail_he,1 , (sym && detail_he->ms.map) ? -map__unmap_ip (detail_he->ms.map, sym->start) : -hist_entry__iaddr (detail_he)); +map__unmap_ip (detail_he->ms.map,symbol__start (sym)) : +hist_entry__iaddr (detail_he));if (!level2_c2c)return NULL ;
@@ -1181,8 +1181,8 @@ static void c2c_he__resort_by_stores(struct hist_entry *parent_he)if (a != b) cmp = a < b ?-1 :1 ;else if (he->ms.sym && iter->ms.sym) - cmp =strcmp (he->ms.sym->name, - iter->ms.sym->name); + cmp =strcmp (symbol__name (he->ms.sym), +symbol__name (iter->ms.sym));else cmp = (iter->ms.sym ?1 :0 ) - (he->ms.sym ?1 :0 );
@@ -1407,7 +1407,7 @@ static int c2c_function__ident_len(struct hist_entry *he)* Match symbol_view_entry(): sort_sym adds the cpumode prefix and, in * verbose mode, the address and DSO origin before the symbol name. */ - size =strlen (he->ms.sym->name) +64 ; + size =strlen (symbol__name (he->ms.sym)) +64 ; symbuf =malloc (size);if (!symbuf)return size -1 ;
tools/perf/util/callchain.c +9 -6
@@ -802,13 +802,15 @@ static enum match_result match_chain(struct callchain_cursor_node *node,* on the symbol start address. */ if (symbol__inlined (cnode->ms.sym) ||symbol__inlined (node->ms.sym)) { - match =match_chain_strings (cnode->ms.sym->name, - node->ms.sym->name); + match =match_chain_strings (symbol__name (cnode->ms.sym), +symbol__name (node->ms.sym));if (match !=MATCH_ERROR )break ; }else { - match =match_chain_dso_addresses (cnode->ms.map, cnode->ms.sym->start, - node->ms.map, node->ms.sym->start); + match =match_chain_dso_addresses (cnode->ms.map, +symbol__start (cnode->ms.sym), + node->ms.map, +symbol__start (node->ms.sym));break ; } }
@@ -1249,11 +1251,12 @@ char *callchain_list__sym_name(struct callchain_list *cl,if (show_srcline && cl->srcline) printed =scnprintf (bf, bfsize,"%s %s%s" , - cl->ms.sym->name, cl->srcline, +symbol__name (cl->ms.sym), + cl->srcline, inlined);else printed =scnprintf (bf, bfsize,"%s%s" , - cl->ms.sym->name, inlined); +symbol__name (cl->ms.sym), inlined); }else printed =scnprintf (bf, bfsize,"%#" PRIx64, cl->ip);
tools/perf/util/capstone.c +10 -10
@@ -333,7 +333,7 @@ static struct symbol *find_symbol_from_addr(struct map *map, u64 addr,return NULL ; } - *sym_offset = al_addr - sym->start; + *sym_offset = al_addr -symbol__start (sym);return sym; }
@@ -368,10 +368,10 @@ static void print_capstone_detail(struct cs_insn *insn, char *buf, size_t len,if (sym_offset ==0 ) {scnprintf (buf, len,"\t# %" PRIx64" <%s>" , - orig_addr, sym->name); + orig_addr,symbol__name (sym)); }else {scnprintf (buf, len,"\t# %" PRIx64" <%s+%#" PRIx64">" , - orig_addr, sym->name, sym_offset); + orig_addr,symbol__name (sym), sym_offset); }map__put (found_map);break ;
@@ -454,10 +454,10 @@ static void format_capstone_insn_arm64(struct cs_insn *insn, char *buf, len = len - (last_imm - buf);if (sym_offset ==0 ) {scnprintf (last_imm, len,"%" PRIx64" <%s>" , - addr, sym->name); + addr,symbol__name (sym)); }else {scnprintf (last_imm, len,"%" PRIx64" <%s+%#" PRIx64">" , - addr, sym->name, sym_offset); + addr,symbol__name (sym), sym_offset); }map__put (found_map); }
@@ -498,7 +498,7 @@ int symbol__disassemble_capstone(const char *filename, struct symbol *sym,struct annotation *notes =symbol__annotation (sym);struct map *map = args->ms->map;struct dso *dso =map__dso (map); -u64 start =map__rip_2objdump (map, sym->start); +u64 start =map__rip_2objdump (map,symbol__start (sym));u64 offset;int i, count, free_count;bool is_64bit = false;
@@ -526,7 +526,7 @@ int symbol__disassemble_capstone(const char *filename, struct symbol *sym,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
@@ -631,8 +631,8 @@ int symbol__disassemble_capstone_powerpc(const char *filename __maybe_unused,struct map *map = args->ms->map;struct dso *dso =map__dso (map);struct nscookie nsc; -u64 start =map__rip_2objdump (map, sym->start); -u64 end =map__rip_2objdump (map, sym->end); +u64 start =map__rip_2objdump (map,symbol__start (sym)); +u64 end =map__rip_2objdump (map,symbol__end (sym));u64 len = end - start;u64 offset;int i, fd, count;
@@ -703,7 +703,7 @@ int symbol__disassemble_capstone_powerpc(const char *filename __maybe_unused,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
tools/perf/util/data-convert-json.c +2 -2
@@ -138,11 +138,11 @@ static void output_sample_callchain_entry(const struct perf_tool *tool,output_json_format (out, false,4 ,"{" );output_json_key_format (out, false,5 ,"ip" ,"\"0x%" PRIx64"\"" , ip); -if (al && al->sym && al->sym->namelen) { +if (al && al->sym &&symbol__namelen (al->sym)) {struct dso *dso = al->map ?map__dso (al->map) :NULL ;fputc (',' , out); -output_json_key_string (out, false,5 ,"symbol" , al->sym->name); +output_json_key_string (out, false,5 ,"symbol" ,symbol__name (al->sym));if (dso) {const char *dso_name =dso__short_name (dso);
tools/perf/util/db-export.c +1 -1
@@ -198,7 +198,7 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al,if (err)return err; *sym_db_id = *db_id; - *offset = al->addr - al->sym->start; + *offset = al->addr -symbol__start (al->sym); } }
tools/perf/util/debug.c +2 -1
@@ -342,7 +342,8 @@ void __dump_stack(FILE *file, void **stackdump, size_t stackdump_size)if (thread &&thread__find_map (thread,PERF_RECORD_MISC_USER , addr, &al)) { al.sym =map__find_symbol (al.map, al.addr);if (al.sym) { -fprintf (file," #%zd %p in %s " , i, stackdump[i], al.sym->name); +fprintf (file," #%zd %p in %s " , i, stackdump[i], +symbol__name (al.sym)); printed = true; } }
tools/perf/util/disasm.c +25 -20
117 lines not on this page · raw message
tools/perf/util/dlfilter.c +6 -6
@@ -47,13 +47,13 @@ static void al_to_d_al(struct addr_location *al, struct perf_dlfilter_al *d_al) d_al->buildid =NULL ; }if (sym) { - d_al->sym = sym->name; - d_al->sym_start = sym->start; - d_al->sym_end = sym->end; -if (al->addr < sym->end) - d_al->symoff = al->addr - sym->start; + d_al->sym =symbol__name (sym); + d_al->sym_start =symbol__start (sym); + d_al->sym_end =symbol__end (sym); +if (al->addr <symbol__end (sym)) + d_al->symoff = al->addr -symbol__start (sym);else if (al->map) - d_al->symoff = al->addr -map__start (al->map) - sym->start; + d_al->symoff = al->addr -map__start (al->map) -symbol__start (sym);else d_al->symoff =0 ; d_al->sym_binding =symbol__binding (sym);
tools/perf/util/dso.c +2 -2
@@ -2027,8 +2027,8 @@ const u8 *dso__read_symbol(struct dso *dso, const char *symfs_filename, const struct map *map, const struct symbol *sym, u8 **out_buf, u64 *out_buf_len, bool *is_64bit) { - u64 start = map__rip_2objdump(map, sym->start); -u64 end =map__rip_2objdump (map, sym->end); + u64 start = map__rip_2objdump(map, symbol__start(sym)); +u64 end =map__rip_2objdump (map,symbol__end (sym));size_t len = end - start; *out_buf =NULL ;
tools/perf/util/event.c +2 -2
@@ -884,11 +884,11 @@ int machine__resolve(struct machine *machine, struct addr_location *al,if (al->sym) { ret =strlist__has_entry (symbol_conf.sym_list, - al->sym->name); +symbol__name (al->sym)); }if (!ret && al->sym) {snprintf (al_addr_str, sz,"0x%" PRIx64, -map__unmap_ip (al->map, al->sym->start)); +map__unmap_ip (al->map,symbol__start (al->sym))); ret =strlist__has_entry (symbol_conf.sym_list, al_addr_str); }
tools/perf/util/evsel_fprintf.c +1 -1
@@ -200,7 +200,7 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,/* Add srccode here too? */ if (bt_stop_list && sym && -strlist__has_entry (bt_stop_list, sym->name)) { +strlist__has_entry (bt_stop_list,symbol__name (sym))) {break ; }
tools/perf/util/hist.c +9 -8
@@ -97,7 +97,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h)* +3 accounts for ' y ' symtab origin info */ if (h->ms.sym) { - symlen = h->ms.sym->namelen +4 ; + symlen =symbol__namelen (h->ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL , symlen);
@@ -120,11 +120,12 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (h->parent) -hists__new_col_len (hists,HISTC_PARENT , h->parent->namelen); +hists__new_col_len (hists,HISTC_PARENT , +symbol__namelen (h->parent));if (h->branch_info) {if (h->branch_info->from.ms.sym) { - symlen = (int )h->branch_info->from.ms.sym->namelen +4 ; + symlen = (int )symbol__namelen (h->branch_info->from.ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL_FROM , symlen);
@@ -139,7 +140,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (h->branch_info->to.ms.sym) { - symlen = (int )h->branch_info->to.ms.sym->namelen +4 ; + symlen = (int )symbol__namelen (h->branch_info->to.ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL_TO , symlen);
@@ -163,7 +164,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h)if (h->mem_info) {if (mem_info__daddr (h->mem_info)->ms.sym) { - symlen = (int )mem_info__daddr (h->mem_info)->ms.sym->namelen +4 + symlen = (int )symbol__namelen (mem_info__daddr (h->mem_info)->ms.sym) +4 + unresolved_col_width +2 ;hists__new_col_len (hists,HISTC_MEM_DADDR_SYMBOL , symlen);
@@ -178,7 +179,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (mem_info__iaddr (h->mem_info)->ms.sym) { - symlen = (int )mem_info__iaddr (h->mem_info)->ms.sym->namelen +4 + symlen = (int )symbol__namelen (mem_info__iaddr (h->mem_info)->ms.sym) +4 + unresolved_col_width +2 ;hists__new_col_len (hists,HISTC_MEM_IADDR_SYMBOL , symlen);
@@ -725,7 +726,7 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,*/ if (hists__has (hists, sym) && he->ms.map != entry->ms.map) {if (he->ms.sym) { -u64 addr = he->ms.sym->start; +u64 addr =symbol__start (he->ms.sym); he->ms.sym =map__find_symbol (entry->ms.map, addr); }
@@ -2298,7 +2299,7 @@ static bool hists__filter_entry_by_symbol(struct hists *hists, struct hist_entry *he) {if (hists->symbol_filter_str !=NULL && - (!he->ms.sym ||strstr (he->ms.sym->name, + (!he->ms.sym ||strstr (symbol__name (he->ms.sym), hists->symbol_filter_str) ==NULL )) { he->filtered |= (1 <<HIST_FILTER__SYMBOL );return true;
tools/perf/util/intel-pt.c +4 -4
@@ -3003,8 +3003,8 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip)for (sym = start; sym; sym =dso__next_symbol (sym)) {if (symbol__binding (sym) ==STB_GLOBAL && - !strcmp (sym->name,"__switch_to" )) { - ip =map__unmap_ip (map, sym->start); + !strcmp (symbol__name (sym),"__switch_to" )) { + ip =map__unmap_ip (map,symbol__start (sym));if (ip >=map__start (map) && ip <map__end (map)) { switch_ip = ip;break ;
@@ -3021,8 +3021,8 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip) ptss ="__perf_event_task_sched_out" ;for (sym = start; sym; sym =dso__next_symbol (sym)) { -if (!strcmp (sym->name, ptss)) { - ip =map__unmap_ip (map, sym->start); +if (!strcmp (symbol__name (sym), ptss)) { + ip =map__unmap_ip (map,symbol__start (sym));if (ip >=map__start (map) && ip <map__end (map)) { *ptss_ip = ip;break ;
tools/perf/util/libbfd.c +3 -2
@@ -552,7 +552,7 @@ int symbol__disassemble_bpf_libbfd(struct symbol *sym __maybe_unused,struct annotation *notes =symbol__annotation (sym);struct bpf_prog_linfo *prog_linfo =NULL ;struct bpf_prog_info_node *info_node; -int len = sym->end - sym->start; +int len =symbol__end (sym) -symbol__start (sym);disassembler_ftype disassemble;struct map *map = args->ms->map;struct perf_bpil *info_linear;
@@ -572,7 +572,8 @@ int symbol__disassemble_bpf_libbfd(struct symbol *sym __maybe_unused,return SYMBOL_ANNOTATE_ERRNO__BPF_INVALID_FILE ;pr_debug ("%s: handling sym %s addr %" PRIx64" len %" PRIx64"\n" , __func__, - sym->name, sym->start, sym->end - sym->start); +symbol__name (sym),symbol__start (sym), +symbol__end (sym) -symbol__start (sym));memset (tpath,0 ,sizeof (tpath));perf_exe (tpath,sizeof (tpath));
tools/perf/util/llvm.c +3 -3
@@ -132,7 +132,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,struct annotation *notes =symbol__annotation (sym);struct map *map = args->ms->map;struct dso *dso =map__dso (map); -u64 start =map__rip_2objdump (map, sym->start); +u64 start =map__rip_2objdump (map,symbol__start (sym));/* Malloc-ed buffer containing instructions read from disk. */ u8 *code_buf =NULL ;/* Pointer to code to be disassembled. */
@@ -196,7 +196,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
@@ -317,7 +317,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,return ret;#else // HAVE_LIBLLVM_SUPPORT pr_debug ("The LLVM disassembler isn't linked in for %s in %s\n" , - sym->name, filename); +symbol__name (sym), filename);return -1 ;#endif }
tools/perf/util/lock-contention.c +1 -1
@@ -135,7 +135,7 @@ bool match_callstack_filter(struct machine *machine, u64 *callstack, int max_stacontinue ;list_for_each_entry (filter, &callstack_filters, list) { -if (strstr (sym->name, filter->name)) +if (strstr (symbol__name (sym), filter->name))return true; } }
tools/perf/util/machine.c +21 -13
@@ -1110,8 +1110,8 @@ static u64 find_entry_trampoline(struct dso *dso)if (symbol__binding (sym) !=STB_GLOBAL )continue ;for (i =0 ; i <ARRAY_SIZE (syms); i++) { -if (!strcmp (sym->name, syms[i])) -return sym->start; +if (!strcmp (symbol__name (sym), syms[i])) +return symbol__start (sym); } }
@@ -2056,7 +2056,7 @@ int machine__process_event(struct machine *machine, union perf_event *event,static bool symbol__match_regex (struct symbol *sym,regex_t *regex) { -return regexec (regex, sym->name,0 ,NULL ,0 ) ==0 ; +return regexec (regex,symbol__name (sym),0 ,NULL ,0 ) ==0 ; } static void ip__resolve_ams(struct thread *thread,
@@ -3283,8 +3283,8 @@ char *machine__resolve_kernel_addr(void *vmachine, unsigned long long *addrp, chreturn NULL ; *modp =__map__is_kmodule (map) ? (char *)dso__short_name (map__dso (map)) :NULL ; - *addrp =map__unmap_ip (map, sym->start); -return sym->name; + *addrp =map__unmap_ip (map,symbol__start (sym)); +return (char *)symbol__name (sym); }struct machine__for_each_dso_cb_args {
@@ -3330,27 +3330,35 @@ bool machine__is_lock_function(struct machine *machine, u64 addr)return false; } - machine->sched.text_start =map__unmap_ip (kmap, sym->start); + machine->sched.text_start =map__unmap_ip (kmap, +symbol__start (sym));/* should not fail from here */ sym =machine__find_kernel_symbol_by_name (machine,"__sched_text_end" , &kmap); - machine->sched.text_end =map__unmap_ip (kmap, sym->start); + machine->sched.text_end =map__unmap_ip (kmap, +symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__lock_text_start" , &kmap); - machine->lock.text_start =map__unmap_ip (kmap, sym->start); + machine->lock.text_start =map__unmap_ip (kmap, +symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__lock_text_end" , &kmap); - machine->lock.text_end =map__unmap_ip (kmap, sym->start); + machine->lock.text_end =map__unmap_ip (kmap, +symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__traceiter_contention_begin" , &kmap);if (sym) { - machine->traceiter.text_start =map__unmap_ip (kmap, sym->start); - machine->traceiter.text_end =map__unmap_ip (kmap, sym->end); + machine->traceiter.text_start =map__unmap_ip (kmap, +symbol__start (sym)); + machine->traceiter.text_end =map__unmap_ip (kmap, +symbol__end (sym)); } sym =machine__find_kernel_symbol_by_name (machine,"trace_contention_begin" , &kmap);if (sym) { - machine->trace.text_start =map__unmap_ip (kmap, sym->start); - machine->trace.text_end =map__unmap_ip (kmap, sym->end); + machine->trace.text_start =map__unmap_ip (kmap, +symbol__start (sym)); + machine->trace.text_end =map__unmap_ip (kmap, +symbol__end (sym)); } }
tools/perf/util/map.c +3 -3
@@ -321,7 +321,7 @@ void map__fixup_start(struct map *map)if (nd !=NULL ) { struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_start (map, sym->start); +map__set_start (map,symbol__start (sym)); } }
@@ -333,7 +333,7 @@ void map__fixup_end(struct map *map)if (nd !=NULL ) { struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_end (map, sym->end); +map__set_end (map,symbol__end (sym)); } }
@@ -606,7 +606,7 @@ u64 map__objdump_2rip(const struct map *map, u64 ip)bool map__contains_symbol (const struct map *map,const struct symbol *sym) { -u64 ip =map__unmap_ip (map, sym->start); +u64 ip =map__unmap_ip (map,symbol__start (sym));return ip >=map__start (map) && ip <map__end (map); }
tools/perf/util/map.h +1 -1
@@ -164,7 +164,7 @@ struct thread;#define __map__for_each_symbol_by_name (map, sym_name, pos, idx) \ for (pos = map__find_symbol_by_name_idx(map, sym_name, &idx); \ pos && \ - !symbol__match_symbol_name(pos->name, sym_name, \ + !symbol__match_symbol_name(symbol__name(pos), sym_name, \ SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); \ pos = dso__next_symbol_by_name(map__dso(map), &idx))
tools/perf/util/probe-event.c +21 -19
@@ -149,7 +149,7 @@ static int kernel_get_symbol_address_by_name(const char *name, u64 *addr, sym =machine__find_kernel_symbol_by_name (host_machine, name, &map);if (!sym)return -ENOENT ; - *addr =map__unmap_ip (map, sym->start) - + *addr =map__unmap_ip (map,symbol__start (sym)) - ((reloc) ?0 :map__reloc (map)) - ((reladdr) ?map__start (map) :0 ); }
@@ -415,13 +415,13 @@ static int find_alternative_probe_point(struct debuginfo *dinfo,/* Find the address of given function */ map__for_each_symbol_by_name (map, pp->function, sym, idx) {if (uprobes) { - address = sym->start; + address =symbol__start (sym);if (symbol__type (sym) ==STT_GNU_IFUNC )pr_warning ("Warning: The probe function (%s) is a GNU indirect function.\n" "Consider identifying the final function used at run time and set the probe directly on that.\n" , pp->function); }else - address =map__unmap_ip (map, sym->start) -map__reloc (map); + address =map__unmap_ip (map,symbol__start (sym)) -map__reloc (map);break ; }if (!address) {
@@ -721,17 +721,17 @@ static int post_process_probe_trace_point(struct probe_trace_point *tp,return (symbol_conf.ignore_vmlinux_buildid) ?0 : -ENOENT ; } -if (strcmp (sym->name, tp->symbol)) { +if (strcmp (symbol__name (sym), tp->symbol)) {/* If we have no realname, use symbol for it */ if (!tp->realname) tp->realname = tp->symbol;else free (tp->symbol); - tp->symbol =strdup (sym->name); + tp->symbol =strdup (symbol__name (sym));if (!tp->symbol)return -ENOMEM ; } - tp->offset = addr - sym->start; + tp->offset = addr -symbol__start (sym); tp->address -= offs;return 0 ;
@@ -2309,8 +2309,8 @@ static int find_perf_probe_point_from_map(struct probe_trace_point *tp, goto out; pp->retprobe = tp->retprobe; - pp->offset = addr -map__unmap_ip (map, sym->start); - pp->function =strdup (sym->name); + pp->offset = addr -map__unmap_ip (map,symbol__start (sym)); + pp->function =strdup (symbol__name (sym)); ret = pp->function ?0 : -ENOMEM ; out:
@@ -3065,7 +3065,7 @@ static int find_probe_functions(struct map *map, char *name, cut_version = false;map__for_each_symbol (map, sym, tmp) { - norm =arch__normalize_symbol_name (sym->name); + norm =arch__normalize_symbol_name (symbol__name (sym));if (!norm)continue ;
@@ -3194,9 +3194,10 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev,/* There can be duplicated symbols in the map */ for (i =0 ; i < j; i++) -if (sym->start == syms[i]->start) { +if (symbol__start (sym) ==symbol__start (syms[i])) {pr_debug ("Found duplicated symbol %s @ %" PRIx64"\n" , - sym->name, sym->start); +symbol__name (sym), +symbol__start (sym));break ; }if (i != j)
@@ -3210,28 +3211,29 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, } ret++; -if (pp->offset > sym->end - sym->start) { +if (pp->offset >symbol__end (sym) -symbol__start (sym)) {pr_warning ("Offset %ld is bigger than the size of %s\n" , - pp->offset, sym->name); + pp->offset,symbol__name (sym)); ret = -ENOENT ; goto err_out; }/* Add one probe point */ - tp->address =map__unmap_ip (map, sym->start) + pp->offset; + tp->address =map__unmap_ip (map,symbol__start (sym)) + pp->offset;/* Check the kprobe (not in module) is within .text */ if (!pev->uprobes && !pev->target && -kprobe_warn_out_range (sym->name, tp->address)) { +kprobe_warn_out_range (symbol__name (sym), tp->address)) { tp->symbol =NULL ;/* Skip it */ skipped++; }else if (reloc_sym) { tp->symbol =strdup_or_goto (reloc_sym->name, nomem_out); tp->offset = tp->address - reloc_sym->addr; }else { - tp->symbol =strdup_or_goto (sym->name, nomem_out); + tp->symbol =strdup_or_goto (symbol__name (sym), + nomem_out); tp->offset = pp->offset; } - tp->realname =strdup_or_goto (sym->name, nomem_out); + tp->realname =strdup_or_goto (symbol__name (sym), nomem_out); tp->retprobe = pp->retprobe;if (pev->target) {
@@ -3813,8 +3815,8 @@ int show_available_funcs(const char *target, struct nsinfo *nsi,for (size_t i =0 ; i <dso__symbol_names_len (dso); i++) {struct symbol *pos =dso__symbol_names (dso)[i]; -if (strfilter__compare (_filter, pos->name)) -printf ("%s\n" , pos->name); +if (strfilter__compare (_filter,symbol__name (pos))) +printf ("%s\n" ,symbol__name (pos)); } end:map__put (map);
tools/perf/util/python.c +4 -4
@@ -759,7 +759,7 @@ static PyObject *pyrf_sample_event__get_symbol(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyUnicode_FromString (pevent->al.sym->name); +return PyUnicode_FromString (symbol__name (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_sym_start(struct pyrf_event *pevent,
@@ -768,7 +768,7 @@ static PyObject *pyrf_sample_event__get_sym_start(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyLong_FromUnsignedLongLong (pevent->al.sym->start); +return PyLong_FromUnsignedLongLong (symbol__start (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_sym_end(struct pyrf_event *pevent,
@@ -777,7 +777,7 @@ static PyObject *pyrf_sample_event__get_sym_end(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyLong_FromUnsignedLongLong (pevent->al.sym->end); +return PyLong_FromUnsignedLongLong (symbol__end (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_raw_buf(struct pyrf_event *pevent,
@@ -890,7 +890,7 @@ static PyObject *pyrf_callchain_node__get_symbol(struct pyrf_callchain_node *pno void *closure __maybe_unused) {if (pnode->sym) -return PyUnicode_FromString (pnode->sym->name); +return PyUnicode_FromString (symbol__name (pnode->sym));return PyUnicode_FromString ("[unknown]" ); }
tools/perf/util/scripting-engines/trace-event-perl.c +4 -4
@@ -301,11 +301,11 @@ static SV *perl_process_callchain(struct perf_sample *sample,hv_undef (elem); goto exit; } -if (!hv_stores (sym,"start" ,newSVuv (node->ms.sym->start)) || - !hv_stores (sym,"end" ,newSVuv (node->ms.sym->end)) || +if (!hv_stores (sym,"start" ,newSVuv (symbol__start (node->ms.sym))) || + !hv_stores (sym,"end" ,newSVuv (symbol__end (node->ms.sym))) || !hv_stores (sym,"binding" ,newSVuv (symbol__binding (node->ms.sym))) || - !hv_stores (sym,"name" ,newSVpvn (node->ms.sym->name, - node->ms.sym->namelen)) || + !hv_stores (sym,"name" ,newSVpvn (symbol__name (node->ms.sym), +symbol__namelen (node->ms.sym))) || !hv_stores (elem,"sym" ,newRV_noinc ((SV *)sym))) {hv_undef (sym);hv_undef (elem);
tools/perf/util/scripting-engines/trace-event-python.c +14 -14
@@ -382,10 +382,10 @@ static unsigned long get_offset(struct symbol *sym, struct addr_location *al) {unsigned long offset; -if (al->addr < sym->end) - offset = al->addr - sym->start; +if (al->addr <symbol__end (sym)) + offset = al->addr -symbol__start (sym);else - offset = al->addr -map__start (al->map) - sym->start; + offset = al->addr -map__start (al->map) -symbol__start (sym);return offset; }
@@ -433,14 +433,14 @@ static PyObject *python_process_callchain(struct perf_sample *sample,if (!pysym)Py_FatalError ("couldn't create Python dictionary" );pydict_set_item_string_decref (pysym,"start" , -PyLong_FromUnsignedLongLong (node->ms.sym->start)); +PyLong_FromUnsignedLongLong (symbol__start (node->ms.sym)));pydict_set_item_string_decref (pysym,"end" , -PyLong_FromUnsignedLongLong (node->ms.sym->end)); +PyLong_FromUnsignedLongLong (symbol__end (node->ms.sym)));pydict_set_item_string_decref (pysym,"binding" ,_PyLong_FromLong (symbol__binding (node->ms.sym)));pydict_set_item_string_decref (pysym,"name" , -_PyUnicode_FromStringAndSize (node->ms.sym->name, - node->ms.sym->namelen)); +_PyUnicode_FromStringAndSize (symbol__name (node->ms.sym), +symbol__namelen (node->ms.sym)));pydict_set_item_string_decref (pyelem,"sym" , pysym);if (node->ms.map) {
@@ -547,15 +547,15 @@ static int get_symoff(struct symbol *sym, struct addr_location *al, {unsigned long offset; -if (!sym || !sym->name[0 ]) +if (!sym || !symbol__name (sym)[0 ])return scnprintf (bf, size,"%s" ,"[unknown]" );if (!print_off) -return scnprintf (bf, size,"%s" , sym->name); +return scnprintf (bf, size,"%s" ,symbol__name (sym)); offset =get_offset (sym, al); -return scnprintf (bf, size,"%s+0x%x" , sym->name, offset); +return scnprintf (bf, size,"%s+0x%x" ,symbol__name (sym), offset); }static int get_br_mspred (struct branch_flags *flags,char *bf,int size)
@@ -908,7 +908,7 @@ static void set_sym_in_dict(PyObject *dict, struct addr_location *al, }if (al->sym) {pydict_set_item_string_decref (dict, sym_field, -_PyUnicode_FromString (al->sym->name)); +_PyUnicode_FromString (symbol__name (al->sym)));pydict_set_item_string_decref (dict, symoff_field,PyLong_FromUnsignedLong (get_offset (al->sym, al))); }
@@ -1379,10 +1379,10 @@ static int python_export_symbol(struct db_export *dbe, struct symbol *sym,tuple_set_d64 (t,0 , *sym_db_id);tuple_set_d64 (t,1 ,dso__db_id (dso)); -tuple_set_d64 (t,2 , sym->start); -tuple_set_d64 (t,3 , sym->end); +tuple_set_d64 (t,2 ,symbol__start (sym)); +tuple_set_d64 (t,3 ,symbol__end (sym));tuple_set_s32 (t,4 ,symbol__binding (sym)); -tuple_set_string (t,5 , sym->name); +tuple_set_string (t,5 ,symbol__name (sym));call_object (tables->symbol_handler, t,"symbol_table" );
tools/perf/util/sort.c +23 -19
117 lines not on this page · raw message
tools/perf/util/srcline.c +7 -6
@@ -116,7 +116,7 @@ struct symbol *new_inline_sym(struct dso *dso, funcname = demangled; } -if (base_sym &&strcmp (funcname, base_sym->name) ==0 ) { +if (base_sym &&strcmp (funcname,symbol__name (base_sym)) ==0 ) {/* reuse the real, existing symbol */ inline_sym = base_sym;/* ensure that we don't alias an inlined symbol, which could
@@ -125,8 +125,9 @@ struct symbol *new_inline_sym(struct dso *dso,assert (!symbol__inlined (base_sym)); }else {/* create a fake symbol for the inline frame */ - inline_sym =symbol__new (base_sym ? base_sym->start :0 , - base_sym ? (base_sym->end - base_sym->start) :0 , + inline_sym =symbol__new (base_sym ?symbol__start (base_sym) :0 , + base_sym ? + (symbol__end (base_sym) -symbol__start (base_sym)) :0 , base_sym ?symbol__binding (base_sym) :0 , base_sym ?symbol__type (base_sym) :0 , funcname);
@@ -307,11 +308,11 @@ char *__get_srcline(struct dso *dso, u64 addr, struct symbol *sym, out:if (!show_addr)return (show_sym && sym) ? -strndup (sym->name, sym->namelen) :SRCLINE_UNKNOWN ; +strndup (symbol__name (sym),symbol__namelen (sym)) :SRCLINE_UNKNOWN ;if (sym) { -if (asprintf (&srcline,"%s+%" PRIu64, show_sym ? sym->name :"" , - ip - sym->start) <0 ) +if (asprintf (&srcline,"%s+%" PRIu64, show_sym ?symbol__name (sym) :"" , + ip -symbol__start (sym)) <0 )return SRCLINE_UNKNOWN ; }else if (asprintf (&srcline,"%s[%" PRIx64"]" ,dso__short_name (dso), addr) <0 )return SRCLINE_UNKNOWN ;
tools/perf/util/symbol-elf.c +6 -5
@@ -350,11 +350,11 @@ static bool get_ifunc_name(Elf *elf, struct dso *dso, GElf_Ehdr *ehdr, sym =dso__find_symbol_nocache (dso, addr);/* Expecting the address to be an IFUNC or IFUNC alias */ -if (!sym || sym->start != addr || +if (!sym ||symbol__start (sym) != addr || (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym)))return false; -snprintf (buf, buf_sz,"%s@plt" , sym->name); +snprintf (buf, buf_sz,"%s@plt" ,symbol__name (sym));return true; }
@@ -629,8 +629,8 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss)* a symbol for .plt header. */ f =dso__find_symbol_nocache (dso, shdr_plt.sh_offset); -if (f && f->start < shdr_plt.sh_offset && f->end > shdr_plt.sh_offset) - f->end = shdr_plt.sh_offset; +if (f &&symbol__start (f) < shdr_plt.sh_offset &&symbol__end (f) > shdr_plt.sh_offset) +symbol__set_end (f, shdr_plt.sh_offset);if (!get_plt_sizes (dso, &ehdr, &shdr_plt, &plt_header_size, &plt_entry_size))return 0 ;
@@ -652,7 +652,8 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss)if (!get_plt_sizes (dso, &ehdr, &plt_sec_shdr, &plt_header_size, &plt_entry_size))return 0 ;/* Extend .plt symbol to entire .plt */ - plt_sym->end = plt_sym->start + shdr_plt.sh_size; +symbol__set_end (plt_sym, +symbol__start (plt_sym) + shdr_plt.sh_size);/* Use .plt.sec offset */ plt_offset = plt_sec_shdr.sh_offset; lazy_plt = false;
tools/perf/util/symbol.c +59 -48
259 lines not on this page · raw message
tools/perf/util/symbol.h +41 -1
54 lines not on this page · raw message
tools/perf/util/symbol_fprintf.c +7 -7
@@ -10,10 +10,10 @@size_t symbol__fprintf (struct symbol *sym,FILE *fp) {return fprintf (fp," %" PRIx64"-%" PRIx64" %c %s\n" , - sym->start, sym->end, +symbol__start (sym),symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , - sym->name); +symbol__name (sym)); } size_t __symbol__fprintf_symname_offs(const struct symbol *sym,
@@ -25,12 +25,12 @@ size_t __symbol__fprintf_symname_offs(const struct symbol *sym,size_t length;if (sym) { - length =fprintf (fp,"%s" , sym->name); + length =fprintf (fp,"%s" ,symbol__name (sym));if (al && print_offsets) { -if (al->addr < sym->end) - offset = al->addr - sym->start; +if (al->addr <symbol__end (sym)) + offset = al->addr -symbol__start (sym);else - offset = al->addr -map__start (al->map) - sym->start; + offset = al->addr -map__start (al->map) -symbol__start (sym); length +=fprintf (fp,"+0x%lx" , offset); }return length;
@@ -67,7 +67,7 @@ size_t dso__fprintf_symbols_by_name(struct dso *dso,for (size_t i =0 ; i <dso__symbol_names_len (dso); i++) {struct symbol *pos =dso__symbol_names (dso)[i]; - ret +=fprintf (fp,"%s\n" , pos->name); + ret +=fprintf (fp,"%s\n" ,symbol__name (pos)); }return ret; }
tools/perf/util/thread-stack.c +4 -4
@@ -1056,7 +1056,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,struct symbol *tsym = to_al->sym;struct call_path *cp; -if (sym &&is_x86_retpoline (sym->name)) { +if (sym &&is_x86_retpoline (symbol__name (sym))) {/* * This is a x86 retpoline fn. It pollutes the call graph by * showing up everywhere there is an indirect branch, but does
@@ -1069,7 +1069,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,*/ ts->cnt -=1 ; sym = ts->stack[ts->cnt -2 ].cp->sym; -if (sym && sym == tsym && to_al->addr != tsym->start) { +if (sym && sym == tsym && to_al->addr !=symbol__start (tsym)) {/* * Target is back to the middle of the symbol we came * from so assume it is an indirect jmp and forget it
@@ -1170,7 +1170,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm,*/ if (!err && rstate ==X86_RETPOLINE_POSSIBLE && to_al->sym && from_al->sym == to_al->sym && - to_al->addr != to_al->sym->start) + to_al->addr !=symbol__start (to_al->sym)) ts->rstate =X86_RETPOLINE_DETECTED ; }else if (sample->flags &PERF_IP_FLAG_RETURN ) {
@@ -1207,7 +1207,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm, err =thread_stack__trace_end (ts, sample, ref); }else if (sample->flags &PERF_IP_FLAG_BRANCH && from_al->sym != to_al->sym && to_al->sym && - to_al->addr == to_al->sym->start) { + to_al->addr ==symbol__start (to_al->sym)) {struct call_path_root *cpr = ts->crp->cpr;struct call_path *cp;
tools/perf/util/unwind-libdw.c +1 -1
@@ -172,7 +172,7 @@ static int entry(u64 ip, struct unwind_info *ui) e->ms.sym = al.sym;pr_debug ("unwind: %s:ip = 0x%" PRIx64" (0x%" PRIx64")\n" , - al.sym ? al.sym->name :"''" , + al.sym ?symbol__name (al.sym) :"''" , ip, al.map ?map__map_ip (al.map, ip) : (u64 )0 );addr_location__exit (&al);
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 2/7 perf symbol: Remove symbol_conf.priv_size and negative-offset allocations
In reply to Ian RogersPreviously struct symbol allocations could be preceded by symbol_conf.priv_size bytes (used by annotation, the TUI map browser index, and db-export symbol IDs). This coupled symbol allocation size to global configuration, allocated unused space ahead of every symbol in the system, and complicated symbol lifetime management.
Without adding extra seldom-used fields to struct symbol:
- Move struct annotation storage to an on-demand hashmap in util/annotate.c keyed by struct symbol pointer, guarded by annotations_lock and tracked via a new SYMBOL_FLAG_ANNOTATED flag in sym->flags so unannotated symbols pay zero lookup or teardown overhead. If allocation fails in symbol__annotation(), return a static empty_annotation fallback so callers that do not check for NULL remain safe while mutating callers check for &empty_annotation.
- Remove symbol__browser_index() in ui/browsers/map.c and compute the TUI map browser index on demand in map_browser__search() by walking the map browser's symbol entries when the user searches with '/'. Previously, '/' search in the TUI map browser was gated behind 'verbose > 0' (-v) in builtin-c2c.c, builtin-report.c, builtin-top.c, and ui/browsers/map.c solely to avoid adding sizeof(u32) to symbol_conf.priv_size for every struct symbol in normal runs. With on- demand index computation there is no per-symbol memory overhead, so remove the verbose > 0 restriction and allow '/' search in the map browser unconditionally.
- Store exported symbol database IDs in an on-demand sym_db_ids hashmap in struct db_export instead of symbol__priv().
- Remove symbol_conf.priv_size and symbol__priv().
--- tools/perf/builtin-c2c.c | 14 -- tools/perf/builtin-report.c | 14 -- tools/perf/tests/builtin-test.c | 1 - tools/perf/ui/browsers/map.c | 35 +++-- tools/perf/util/annotate.c | 128 +++++++++++++++++- tools/perf/util/annotate.h | 6 +- tools/perf/util/db-export.c | 49 ++++++- tools/perf/util/db-export.h | 6 + .../scripting-engines/trace-event-python.c | 10 +- tools/perf/util/symbol.c | 34 ++--- tools/perf/util/symbol.h | 16 ++- tools/perf/util/symbol_conf.h | 1 - 12 files changed, 219 insertions(+), 95 deletions(-)
tools/perf/builtin-c2c.c +0 -14
@@ -3404,20 +3404,6 @@ static int perf_c2c__report(int argc, const char **argv)if (ret <0 ) goto out_mem2node; -/* -* For searching by name on the "Browse map details". -* providing it only in verbose mode not to bloat too -* much struct symbol. -*/ -if (verbose >0 ) { -/* -* XXX: Need to provide a less kludgy way to ask for -* more space per symbol, the u32 is for the index on -* the ui browser. -* See symbol__browser_index. -*/ - symbol_conf.priv_size +=sizeof (u32); - }annotation_config__init (); }
tools/perf/builtin-report.c +0 -14
@@ -1876,20 +1876,6 @@ int cmd_report(int argc, const char **argv) ret =symbol__annotation_init ();if (ret <0 ) goto error; -/* -* For searching by name on the "Browse map details". -* providing it only in verbose mode not to bloat too -* much struct symbol. -*/ -if (verbose >0 ) { -/* -* XXX: Need to provide a less kludgy way to ask for -* more space per symbol, the u32 is for the index on -* the ui browser. -* See symbol__browser_index. -*/ - symbol_conf.priv_size +=sizeof (u32); - }annotation_config__init (); }
tools/perf/tests/builtin-test.c +0 -1
@@ -1827,7 +1827,6 @@ int cmd_test(int argc, const char **argv)if (dont_fork) sequential = true; - symbol_conf.priv_size =sizeof (int ); symbol_conf.try_vmlinux_path = true;
tools/perf/ui/browsers/map.c +17 -18
@@ -40,12 +40,6 @@ static void map_browser__write(struct ui_browser *browser, void *nd, int row)ui_browser__write_nstring (browser,symbol__name (sym), width); } -/* FIXME uber-kludgy, see comment on cmd_report... */ -static u32 *symbol__browser_index (struct symbol *browser) -{ -return ((void *)browser) -sizeof (struct rb_node ) -sizeof (u32); -} -static int map_browser__search (struct map_browser *browser) {char target[512 ];
@@ -63,10 +57,21 @@ static int map_browser__search(struct map_browser *browser) sym =map__find_symbol_by_name (browser->map, target);if (sym !=NULL ) { -u32 *idx =symbol__browser_index (sym); - - browser->b.top = &sym->rb_node; - browser->b.index = browser->b.top_idx = *idx; +struct rb_node *nd; +u32 idx =0 ; + +/* +* Walk the map browser's symbol entries to find the matching +* symbol node and its display row index, then position the +* browser cursor at that entry. +*/ +for (nd =rb_first (browser->b.entries); nd; nd =rb_next (nd), ++idx) { +if (&sym->rb_node == nd) { + browser->b.top = nd; + browser->b.index = browser->b.top_idx = idx; +break ; + } + } }else ui_helpline__fpush ("%s not found!" , target);
@@ -78,8 +83,7 @@ static int map_browser__run(struct map_browser *browser)int key;if (ui_browser__show (&browser->b,dso__long_name (map__dso (browser->map)), -"Press ESC to exit, %s / to search" , - verbose >0 ?"" :"restart with -v to use" ) <0 ) +"Press ESC to exit, / to search" ) <0 )return -1 ;while (1 ) {
@@ -87,8 +91,7 @@ static int map_browser__run(struct map_browser *browser)switch (key) {case '/' : -if (verbose >0 ) -map_browser__search (browser); +map_browser__search (browser);/* fall thru */ default :ui_browser__warn_unhandled_hotkey (&browser->b, key,0 ,NULL );
@@ -125,10 +128,6 @@ int map__browse(struct map *map)if (maxaddr <symbol__end (pos)) maxaddr =symbol__end (pos); -if (verbose >0 ) { -u32 *idx =symbol__browser_index (pos); - *idx = mb.b.nr_entries; - } ++mb.b.nr_entries; }
tools/perf/util/annotate.c +126 -2
160 lines not on this page · raw message
tools/perf/util/annotate.h +2 -4
@@ -449,10 +449,8 @@ static inline struct sym_hist *annotation__histogram(struct annotation *notes,struct sym_hist_entry *annotated_source__hist_entry (struct annotated_source *src,const struct evsel *evsel,u64 offset); -static inline struct annotation *symbol__annotation (struct symbol *sym) -{ -return (void *)sym - symbol_conf.priv_size; -} +struct annotation *symbol__annotation (struct symbol *sym); +void symbol__annotation_delete (struct symbol *sym);int addr_map_symbol__inc_samples (struct addr_map_symbol *ams,struct perf_sample *sample);
tools/perf/util/db-export.c +43 -6
84 lines not on this page · raw message
tools/perf/util/db-export.h +6 -0
@@ -63,6 +63,11 @@ struct db_export { u64 th_in_id, u64 comm_in_id, int flags);struct call_return_processor *crp;struct call_path_root *cpr; +/** +* @sym_db_ids: Hashmap mapping struct symbol pointers to their +* exported 64-bit database IDs. +*/ +struct hashmap *sym_db_ids;u64 evsel_last_db_id;u64 machine_last_db_id;u64 thread_last_db_id;
@@ -90,6 +95,7 @@ int db_export__comm_thread(struct db_export *dbe, struct comm *comm, struct thread *thread);int db_export__dso (struct db_export *dbe,struct dso *dso,struct machine *machine); +u64 db_export__symbol_db_id (struct db_export *dbe,struct symbol *sym);int db_export__symbol (struct db_export *dbe,struct symbol *sym,struct dso *dso); int db_export__branch_type(struct db_export *dbe, u32 branch_type,
tools/perf/util/scripting-engines/trace-event-python.c +3 -7
@@ -1372,12 +1372,12 @@ static int python_export_symbol(struct db_export *dbe, struct symbol *sym, struct dso *dso) { struct tables *tables = container_of(dbe, struct tables, dbe); -u64 *sym_db_id =symbol__priv (sym); +u64 sym_db_id =db_export__symbol_db_id (dbe, sym);PyObject *t; t =tuple_new (6 ); -tuple_set_d64 (t,0 , *sym_db_id); +tuple_set_d64 (t,0 , sym_db_id);tuple_set_d64 (t,1 ,dso__db_id (dso));tuple_set_d64 (t,2 ,symbol__start (sym));tuple_set_d64 (t,3 ,symbol__end (sym));
@@ -1487,7 +1487,7 @@ static int python_export_call_path(struct db_export *dbe, struct call_path *cp)u64 parent_db_id, sym_db_id; parent_db_id = cp->parent ? cp->parent->db_id :0 ; - sym_db_id = cp->sym ? *(u64 *)symbol__priv (cp->sym) :0 ; + sym_db_id =db_export__symbol_db_id (dbe, cp->sym); t =tuple_new (4 );
@@ -1993,10 +1993,6 @@ static void set_table_handlers(struct tables *tables) } tables->db_export_mode = true; -/* -* Reserve per symbol space for symbol->db_id via symbol__priv() -*/ - symbol_conf.priv_size =sizeof (u64);SET_TABLE_HANDLER (evsel);SET_TABLE_HANDLER (machine);
tools/perf/util/symbol.c +13 -21
@@ -319,19 +319,11 @@ void symbols__fixup_end(struct rb_root_cached *symbols, bool is_kallsyms)struct symbol *symbol__new (u64 start,u64 len,u8 binding,u8 type,const char *name) {size_t namelen =strlen (name) +1 ; -struct symbol *sym =calloc (1 , (symbol_conf.priv_size + -sizeof (*sym) + namelen)); +struct symbol *sym =calloc (1 ,sizeof (*sym) + namelen); +if (sym ==NULL )return NULL ; -if (symbol_conf.priv_size) { -if (symbol_conf.init_annotation) { -struct annotation *notes = (void *)sym; -annotation__init (notes); - } - sym = ((void *)sym) + symbol_conf.priv_size; - } - sym->start = start; sym->end = len ? start + len : start; atomic_init(&sym->flags, (type << SYMBOL_FLAG_TYPE_SHIFT) |
@@ -347,14 +339,9 @@ struct symbol *symbol__new(u64 start, u64 len, u8 binding, u8 type, const char *void symbol__delete (struct symbol *sym) { -if (symbol_conf.priv_size) { -if (symbol_conf.init_annotation) { -struct annotation *notes =symbol__annotation (sym); - -annotation__exit (notes); - } - } -free (((void *)sym) - symbol_conf.priv_size); +if (symbol__annotated (sym)) +symbol__annotation_delete (sym); +free (sym); }void symbol__set_ignore (struct symbol *sym,bool ignore)
@@ -389,6 +376,14 @@ void symbol__set_ifunc_alias(struct symbol *sym, bool ifunc_alias)atomic_fetch_and (&sym->flags, ~SYMBOL_FLAG_IFUNC_ALIAS ); } +void symbol__set_annotated (struct symbol *sym,bool annotated) +{ +if (annotated) +atomic_fetch_or (&sym->flags,SYMBOL_FLAG_ANNOTATED ); +else +atomic_fetch_and (&sym->flags, ~SYMBOL_FLAG_ANNOTATED ); +} +static void symbol__set_idle (struct symbol *sym,bool idle) {uint16_t old_flags =atomic_load_explicit (&sym->flags, memory_order_relaxed);
@@ -2504,7 +2499,6 @@ int symbol__annotation_init(void)return -1 ; } - symbol_conf.priv_size +=sizeof (struct annotation ); symbol_conf.init_annotation = true;return 0 ; }
@@ -2547,8 +2541,6 @@ int symbol__init(struct perf_env *env)if (symbol_conf.initialized)return 0 ; - symbol_conf.priv_size =PERF_ALIGN (symbol_conf.priv_size,sizeof (u64)); -symbol__elf_init ();if (symbol_conf.try_vmlinux_path &&vmlinux_path__init (env) <0 )
tools/perf/util/symbol.h +9 -7
@@ -88,9 +88,9 @@ enum symbol_idle_kind {#define SYMBOL_FLAG_INLINED (1 << 11)#define SYMBOL_FLAG_ANNOTATE2 (1 << 12)#define SYMBOL_FLAG_IFUNC_ALIAS (1 << 13) +#define SYMBOL_FLAG_ANNOTATED (1 << 14)/** -* A symtab entry. When allocated this may be preceded by an annotation (see -* symbol__annotation) and/or a browser_index (see symbol__browser_index). +* A symtab entry. */ struct symbol {struct rb_node rb_node;
@@ -185,12 +185,19 @@ static inline bool symbol__ifunc_alias(const struct symbol *sym) SYMBOL_FLAG_IFUNC_ALIAS) != 0; } +static inline bool symbol__annotated (const struct symbol *sym) +{ +return (atomic_load_explicit (&sym->flags, memory_order_relaxed) & +SYMBOL_FLAG_ANNOTATED ) !=0 ; +} +bool symbol__is_idle (struct symbol *sym,const struct dso *dso,struct perf_env *env);void symbol__set_ignore (struct symbol *sym,bool ignore);void symbol__set_annotate2 (struct symbol *sym,bool annotate2);void symbol__set_inlined (struct symbol *sym,bool inlined);void symbol__set_ifunc_alias (struct symbol *sym,bool ifunc_alias); +void symbol__set_annotated (struct symbol *sym,bool annotated);/* symbols__for_each_entry - iterate over symbols (rb_root) *
@@ -224,11 +231,6 @@ static inline int __symbol__join_symfs(char *bf, size_t size, const char *path)extern int vmlinux_path__nr_entries;extern char **vmlinux_path; -static inline void *symbol__priv (struct symbol *sym) -{ -return ((void *)sym) - symbol_conf.priv_size; -} -struct ref_reloc_sym {const char *name;u64 addr;
tools/perf/util/symbol_conf.h +0 -1
@@ -42,7 +42,6 @@ struct symbol_conf {/* Was hybrid_merge given on the command line rather than in a config file? */ bool hybrid_merge_set;bool nanosecs; -unsigned short priv_size;bool try_vmlinux_path, init_annotation, force,
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 3/7 perf symbol: Switch backing storage from rbtree to struct symbols array
In reply to Ian RogersReplace struct rb_node rb_node in struct symbol (which cost 24 bytes per symbol on 64-bit) with a sorted array in struct symbols, matching the design of struct dsos and struct maps:
- Introduce struct symbols containing a rw_semaphore lock, primary address-sorted symbols array, secondary name-sorted symbols_by_name array, counts, capacity, and sorted state flags.
- Lazily sort symbols by [start, end) using qsort (symbols__sort_locked) and use binary search in symbols__find().
- Update symbols__fixup_duplicate(), symbols__fixup_end(), maps__split_kallsyms_for_kcore(), and maps__split_kallsyms() to operate on the sorted array and compact entries in place.
- Convert symbols__for_each_entry(), dso__for_each_symbol(), and map__for_each_symbol() to callback-based iteration under symbols->lock, matching dsos__for_each_dso() and maps__for_each_map().
- Remove the unsynchronized single-entry dso->last_find_result cache and dso__reset_find_symbol_cache() since binary search over a contiguous array is cache-friendly.
- Remove the direct rb_erase_cached() workaround in builtin-annotate.c's add_sample().
- Update the TUI map browser (ui/browsers/map.c) to snapshot the sorted symbol array into an entries array and use ui_browser__argv_refresh and ui_browser__argv_seek instead of ui_browser__rb_tree_refresh.
--- tools/perf/arch/powerpc/util/sym-handling.c | 34 +- tools/perf/builtin-annotate.c | 11 - tools/perf/builtin-kmem.c | 51 ++- tools/perf/tests/symbols.c | 61 ++- tools/perf/tests/vmlinux-kallsyms.c | 189 ++++---- tools/perf/ui/browsers/map.c | 62 ++- tools/perf/util/auxtrace.c | 109 +++-- tools/perf/util/dso.c | 50 +- tools/perf/util/dso.h | 59 +-- tools/perf/util/intel-pt.c | 74 +-- tools/perf/util/machine.c | 28 +- tools/perf/util/map.c | 25 +- tools/perf/util/map.h | 8 +- tools/perf/util/probe-event.c | 82 ++-- tools/perf/util/symbol.c | 478 ++++++++++++-------- tools/perf/util/symbol.h | 70 ++- 16 files changed, 801 insertions(+), 590 deletions(-)
tools/perf/arch/powerpc/util/sym-handling.c +24 -10
@@ -115,13 +115,27 @@ void arch__fix_tev_from_maps(struct perf_probe_event *pev, }#ifdef HAVE_LIBELF_SUPPORT +struct post_process_probe_trace_events_args { +struct perf_probe_event *pev; +struct probe_trace_event *tev; +struct map *map; +}; + +static int post_process_probe_trace_events_cb (struct symbol *sym,void *data) +{ +struct post_process_probe_trace_events_args *args = data; + +if (map__unmap_ip (args->map,symbol__start (sym)) == args->tev->point.address) { +arch__fix_tev_from_maps (args->pev, args->tev, args->map, sym); +return 1 ; + } +return 0 ; +} +void arch__post_process_probe_trace_events (struct perf_probe_event *pev,int ntevs) { -struct probe_trace_event *tev;struct map *map; -struct symbol *sym =NULL ; -struct rb_node *tmp;int i =0 ; map =get_target_map (pev->target, pev->nsi, pev->uprobes);
@@ -129,13 +143,13 @@ void arch__post_process_probe_trace_events(struct perf_probe_event *pev,return ;for (i =0 ; i < ntevs; i++) { - tev = &pev->tevs[i]; -map__for_each_symbol (map, sym, tmp) { -if (map__unmap_ip (map,symbol__start (sym)) == tev->point.address) { -arch__fix_tev_from_maps (pev, tev, map, sym); -break ; - } - } +struct post_process_probe_trace_events_args args = { + .pev = pev, + .tev = &pev->tevs[i], + .map = map, + }; + +map__for_each_symbol (map, post_process_probe_trace_events_cb, &args); } }#endif /* HAVE_LIBELF_SUPPORT */
tools/perf/builtin-annotate.c +0 -11
@@ -254,17 +254,6 @@ static int add_sample(struct perf_sample *sample, (al->sym == NULL || strcmp(ann->sym_hist_filter, symbol__name(al->sym)) != 0)) {/* We're only interested in a symbol named sym_hist_filter */ -/* -* FIXME: why isn't this done in the symbol_filter when loading -* the DSO? -*/ -if (al->sym !=NULL ) { -struct dso *dso =map__dso (al->map); - -rb_erase_cached (&al->sym->rb_node,dso__symbols (dso)); -symbol__delete (al->sym); -dso__reset_find_symbol_cache (dso); - }return 0 ; }
tools/perf/builtin-kmem.c +29 -22
72 lines not on this page · raw message
tools/perf/tests/symbols.c +36 -25
75 lines not on this page · raw message
tools/perf/tests/vmlinux-kallsyms.c +95 -94
225 lines not on this page · raw message
tools/perf/ui/browsers/map.c +41 -21
92 lines not on this page · raw message
tools/perf/util/auxtrace.c +72 -37
133 lines not on this page · raw message
tools/perf/util/dso.c +26 -24
85 lines not on this page · raw message
tools/perf/util/dso.h +10 -49
128 lines not on this page · raw message
tools/perf/util/intel-pt.c +45 -29
94 lines not on this page · raw message
tools/perf/util/machine.c +17 -11
@@ -1095,29 +1095,35 @@ int machine__create_extra_kernel_map(struct machine *machine,return err; } -static u64 find_entry_trampoline (struct dso *dso) +static int find_entry_trampoline_cb (struct symbol *sym,void *data) {/* Duplicates are removed so lookup all aliases */ -const char *syms[] = { +static const char *const syms[] = {"_entry_trampoline" ,"__entry_trampoline_start" ,"entry_SYSCALL_64_trampoline" , }; -struct symbol *sym =dso__first_symbol (dso); -unsigned int i; +u64 *addr = data; -for (; sym; sym =dso__next_symbol (sym)) { -if (symbol__binding (sym) !=STB_GLOBAL ) -continue ; -for (i =0 ; i <ARRAY_SIZE (syms); i++) { -if (!strcmp (symbol__name (sym), syms[i])) -return symbol__start (sym); +if (symbol__binding (sym) !=STB_GLOBAL ) +return 0 ; +for (unsigned int i =0 ; i <ARRAY_SIZE (syms); i++) { +if (!strcmp (symbol__name (sym), syms[i])) { + *addr =symbol__start (sym); +return 1 ; } } -return 0 ; } +static u64 find_entry_trampoline (struct dso *dso) +{ +u64 addr =0 ; + +dso__for_each_symbol (dso, find_entry_trampoline_cb, &addr); +return addr; +} +/* * These values can be used for kernels that do not have symbols for the entry * trampolines in kallsyms.
tools/perf/util/map.c +10 -15
@@ -314,27 +314,22 @@ void map__put(struct map *map)void map__fixup_start (struct map *map) { -struct dso *dso =map__dso (map); -struct rb_root_cached *symbols =dso__symbols (dso); -struct rb_node *nd =rb_first_cached (symbols); - -if (nd !=NULL ) { - struct symbol *sym = rb_entry(nd, struct symbol, rb_node); +struct symbols *symbols =dso__symbols (map__dso (map)); -map__set_start (map,symbol__start (sym)); - } +symbols__sort_read_lock (symbols); +if (symbols->cnt >0 ) +map__set_start (map,symbol__start (symbols->symbols[0 ])); +up_read (&symbols->lock); }void map__fixup_end (struct map *map) { -struct dso *dso =map__dso (map); -struct rb_root_cached *symbols =dso__symbols (dso); -struct rb_node *nd =rb_last (&symbols->rb_root); +struct symbols *symbols =dso__symbols (map__dso (map)); -if (nd !=NULL ) { - struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_end (map,symbol__end (sym)); - } +symbols__sort_read_lock (symbols); +if (symbols->cnt >0 ) +map__set_end (map,symbol__end (symbols->symbols[symbols->cnt -1 ])); +up_read (&symbols->lock); }#define DSO__DELETED "(deleted)"
tools/perf/util/map.h +4 -4
@@ -146,12 +146,12 @@ struct thread;/* map__for_each_symbol - iterate over the symbols in the given map * * @map: the 'struct map *' in which symbols are iterated -* @pos: the 'struct symbol *' to use as a loop cursor -* @n: the 'struct rb_node *' to use as a temporary storage +* @cb: callback function invoked for each symbol +* @data: opaque data pointer passed to @cb * Note: caller must ensure map->dso is not NULL (map is loaded). */ -#define map__for_each_symbol (map, pos, n) \ - dso__for_each_symbol(map__dso(map), pos, n) +#define map__for_each_symbol (map, cb, data) \ + dso__for_each_symbol(map__dso(map), cb, data)/* map__for_each_symbol_with_name - iterate over the symbols in the given map * that have the given name
tools/perf/util/probe-event.c +48 -34
97 lines not on this page · raw message
tools/perf/util/symbol.c +294 -184
690 lines not on this page · raw message
tools/perf/util/symbol.h +50 -20
115 lines not on this page · raw message
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 4/7 perf symbol: Add reference counting and DECLARE_RC_STRUCT(symbol)
In reply to Ian RogersWrap struct symbol with DECLARE_RC_STRUCT(symbol) and add a refcount_t refcnt field along with symbol__get(), symbol__put(), and symbol__zput() helpers:
- Return a counted symbol reference acquired under symbols->lock from symbol lookup functions (symbols__find(), dso__find_symbol(), map__find_symbol(), dso__find_symbol_by_name(), map__find_symbol_by_name(), dso__first_symbol(), dso__last_symbol()) and iteration helpers (symbols__first_by_idx(), symbols__next_by_idx(), symbols__for_each_entry()).
- Update util/annotate.c and util/db-export.c to hash and compare the underlying RC_STRUCT(symbol) pointer via RC_CHK_ACCESS(sym) and RC_CHK_EQUAL() so REFCNT_CHECKING wrapper pointers for the same symbol resolve to the same hashmap entry, and hold a symbol reference in struct db_export's sym_db_ids hashmap.
- Update struct addr_location, struct map_symbol, struct call_path, struct block_info, struct disasm_line, struct annotated_op_loc, struct c2c_sym_Key, inline_list, srcline, thread_stack, and hist/sort/unwind consumers to acquire and release symbol references and use RC_CHK_EQUAL() for pointer identity comparisons.
--- tools/perf/builtin-kallsyms.c | 2 + tools/perf/builtin-lock.c | 17 ++- tools/perf/builtin-script.c | 4 +- tools/perf/builtin-top.c | 2 +- tools/perf/builtin-trace.c | 1 + tools/perf/tests/dlfilter-test.c | 8 +- tools/perf/tests/hists_cumulate.c | 4 +- tools/perf/tests/hists_filter.c | 10 +- tools/perf/tests/hists_link.c | 16 ++- tools/perf/tests/hists_output.c | 7 +- tools/perf/tests/kallsyms-split.c | 5 +- tools/perf/tests/pe-file-parsing.c | 1 + tools/perf/tests/vmlinux-kallsyms.c | 12 +- tools/perf/ui/browsers/hists.c | 7 +- tools/perf/ui/browsers/map.c | 7 +- tools/perf/ui/stdio/hist.c | 3 +- tools/perf/util/addr_location.c | 4 + .../perf/util/annotate-arch/annotate-alpha.c | 2 +- .../util/annotate-arch/annotate-loongarch.c | 4 +- tools/perf/util/annotate-arch/annotate-s390.c | 2 +- tools/perf/util/annotate.c | 22 ++- tools/perf/util/block-info.c | 6 +- tools/perf/util/bpf_lock_contention.c | 27 +++- tools/perf/util/c2c-function.c | 13 +- tools/perf/util/call-path.c | 15 +- tools/perf/util/callchain.c | 4 +- tools/perf/util/capstone.c | 1 + tools/perf/util/db-export.c | 35 ++++- tools/perf/util/disasm.c | 12 +- tools/perf/util/event.c | 8 +- tools/perf/util/hist.c | 13 +- tools/perf/util/libdw.c | 5 +- tools/perf/util/lock-contention.c | 8 +- tools/perf/util/machine.c | 48 +++++-- tools/perf/util/map.h | 10 +- tools/perf/util/map_symbol.c | 4 +- tools/perf/util/maps.c | 3 +- tools/perf/util/probe-event.c | 17 ++- tools/perf/util/python.c | 12 +- tools/perf/util/sort.c | 2 +- tools/perf/util/srcline.c | 6 +- tools/perf/util/symbol-elf.c | 6 +- tools/perf/util/symbol.c | 129 +++++++++++------- tools/perf/util/symbol.h | 60 +++++--- tools/perf/util/thread-stack.c | 12 +- tools/perf/util/threads.c | 17 ++- tools/perf/util/unwind-libdw.c | 2 +- tools/perf/util/unwind-libunwind.c | 2 +- 48 files changed, 427 insertions(+), 190 deletions(-)
tools/perf/builtin-kallsyms.c +2 -0
@@ -53,6 +53,8 @@ static int __cmd_kallsyms(int argc, const char **argv)map__unmap_ip (map,symbol__start (symbol)),map__unmap_ip (map,symbol__end (symbol)), symbol__start(symbol), symbol__end(symbol)); +symbol__put (symbol); +map__put (map); } out:machine__delete (machine);
tools/perf/builtin-lock.c +14 -3
@@ -1003,12 +1003,16 @@ static int report_lock_contention_begin_event(struct perf_sample *sample) (filters.nr_addrs + 1) * sizeof(*addrs));if (addrs ==NULL ) {pr_warning ("memory allocation failure\n" ); +symbol__put (sym); +map__put (kmap);return -ENOMEM ; } addrs[filters.nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); filters.addrs = addrs; +symbol__put (sym); +map__put (kmap); } }
@@ -1019,9 +1023,12 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)switch (aggr_mode) {case LOCK_AGGR_ADDR : - sym =machine__find_kernel_symbol (machine, key, &kmap); -if (sym) - name =symbol__name (sym); + sym =machine__find_kernel_symbol (machine, key,NULL ); +if (sym) { +scnprintf (buf,sizeof (buf),"%s" ,symbol__name (sym)); + name = buf; +symbol__put (sym); + }break ;case LOCK_AGGR_CALLER : name = buf;
@@ -1668,6 +1675,8 @@ static void print_lock_stat_stdio(struct lock_contention *con, struct lock_stat ip = st->callstack[i]; sym =machine__find_kernel_symbol (con->machine, ip, &kmap);get_symbol_name_offset (kmap, sym, ip, buf,sizeof (buf)); +symbol__put (sym); +map__put (kmap);fprintf (lock_output,"\t\t\t%#lx %s\n" , (unsigned long )ip, buf); } }
@@ -1721,6 +1730,8 @@ static void print_lock_stat_csv(struct lock_contention *con, struct lock_stat *s ip = st->callstack[i]; sym =machine__find_kernel_symbol (con->machine, ip, &kmap);get_symbol_name_offset (kmap, sym, ip, buf,sizeof (buf)); +symbol__put (sym); +map__put (kmap);fprintf (lock_output,"%s %#lx %s" , i ?":" : sep, (unsigned long ) ip, buf); }fprintf (lock_output,"\n" );
tools/perf/builtin-script.c +3 -1
@@ -1426,7 +1426,8 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,if (PRINT_FIELD (SRCLINE )) printed +=map__fprintf_srcline (al.map, al.addr,"\t" , fp); printed +=fprintf (fp,"\n" ); - *lastsym = al.sym; +symbol__put (*lastsym); + *lastsym =symbol__get (al.sym); ret = printed; out:
@@ -1593,6 +1594,7 @@ static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample,print_srccode (thread, x.cpumode, start + off); } out: +symbol__put (lastsym);return printed; }
tools/perf/builtin-top.c +1 -1
@@ -211,7 +211,7 @@ static void perf_top__record_precise_ip(struct perf_top *top,if (sym ==NULL || (use_browser ==0 && (top->sym_filter_entry ==NULL || - top->sym_filter_entry->ms.sym != sym))) + !RC_CHK_EQUAL (top->sym_filter_entry->ms.sym, sym))))return ; notes =symbol__annotation (sym);
tools/perf/builtin-trace.c +1 -0
@@ -778,6 +778,7 @@ size_t syscall_arg__scnprintf_ksym(char *bf, size_t size, struct syscall_arg *arelse printed =scnprintf (bf, size,"%s+0x%" PRIx64,symbol__name (sym), offset); +symbol__put (sym);map__put (map);return printed; }
tools/perf/tests/dlfilter-test.c +6 -2
@@ -273,12 +273,16 @@ static int get_ip_addr(struct test_data *td)return -1 ; sym =map__find_symbol_by_name (map,"foo" ); -if (sym) +if (sym) { td->foo =symbol__start (sym); +symbol__put (sym); + } sym =map__find_symbol_by_name (map,"bar" ); -if (sym) +if (sym) { td->bar =symbol__start (sym); +symbol__put (sym); + }map__put (map);
tools/perf/tests/hists_cumulate.c +3 -1
@@ -114,7 +114,8 @@ static int add_hist_entries(struct hists *hists, struct machine *machine) fake_samples[i].thread =thread__get (al.thread);map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); }addr_location__exit (&al);
@@ -155,6 +156,7 @@ static void put_fake_samples(void)size_t i;for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { +symbol__zput (fake_samples[i].sym);map__zput (fake_samples[i].map);thread__zput (fake_samples[i].thread); }
tools/perf/tests/hists_filter.c +7 -3
@@ -93,7 +93,8 @@ static int add_hist_entries(struct evlist *evlist, fake_samples[i].thread =thread__get (al.thread);map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); } }addr_location__exit (&al);
@@ -109,8 +110,11 @@ static void put_fake_samples(void) {size_t i; -for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) -map__put (fake_samples[i].map); +for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { +symbol__zput (fake_samples[i].sym); +map__zput (fake_samples[i].map); +thread__zput (fake_samples[i].thread); + } }static int test__hists_filter (struct test_suite *test__maybe_unused ,int subtest__maybe_unused )
tools/perf/tests/hists_link.c +11 -5
@@ -99,7 +99,8 @@ static int add_hist_entries(struct evlist *evlist, struct machine *machine) fake_common_samples[k].thread =thread__get (al.thread);map__put (fake_common_samples[k].map); fake_common_samples[k].map =map__get (al.map); - fake_common_samples[k].sym = al.sym; +symbol__put (fake_common_samples[k].sym); + fake_common_samples[k].sym =symbol__get (al.sym); }for (k =0 ; k <ARRAY_SIZE (fake_samples[i]); k++) {
@@ -119,7 +120,8 @@ static int add_hist_entries(struct evlist *evlist, struct machine *machine) fake_samples[i][k].thread =thread__get (al.thread);map__put (fake_samples[i][k].map); fake_samples[i][k].map =map__get (al.map); - fake_samples[i][k].sym = al.sym; +symbol__put (fake_samples[i][k].sym); + fake_samples[i][k].sym =symbol__get (al.sym); } i++; }
@@ -136,11 +138,15 @@ static void put_fake_samples(void) {size_t i, j; -for (i =0 ; i <ARRAY_SIZE (fake_common_samples); i++) +for (i =0 ; i <ARRAY_SIZE (fake_common_samples); i++) {map__put (fake_common_samples[i].map); +symbol__put (fake_common_samples[i].sym); + }for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { -for (j =0 ; j <ARRAY_SIZE (fake_samples[0 ]); j++) +for (j =0 ; j <ARRAY_SIZE (fake_samples[0 ]); j++) {map__put (fake_samples[i][j].map); +symbol__put (fake_samples[i][j].sym); + } } }
@@ -150,7 +156,7 @@ static int find_sample(struct sample *samples, size_t nr_samples,while (nr_samples--) {if (RC_CHK_EQUAL (samples->thread, t) &&RC_CHK_EQUAL (samples->map, m) && - samples->sym == s) +RC_CHK_EQUAL (samples->sym, s))return 1 ; samples++; }
tools/perf/tests/hists_output.c +4 -3
@@ -79,7 +79,8 @@ static int add_hist_entries(struct hists *hists, struct machine *machine) fake_samples[i].thread = al.thread;map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); }addr_location__exit (&al);
@@ -120,8 +121,8 @@ static void put_fake_samples(void)size_t i;for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { -map__put (fake_samples[i].map); - fake_samples[i].map =NULL ; +symbol__zput (fake_samples[i].sym); +map__zput (fake_samples[i].map); } }
tools/perf/tests/kallsyms-split.c +4 -1
@@ -99,6 +99,7 @@ static int test__kallsyms_split(struct test_suite *test __maybe_unused, {struct machine m = {0 };struct map *map =NULL ; +struct symbol *sym;int ret =TEST_FAIL ;pr_debug ("try to create fake root directory\n" );
@@ -139,10 +140,12 @@ static int test__kallsyms_split(struct test_suite *test __maybe_unused, goto out; } -if (machine__find_kernel_symbol_by_name (&m,"main_symbol3" , &map) ==NULL ) { + sym =machine__find_kernel_symbol_by_name (&m,"main_symbol3" , &map); +if (sym ==NULL ) {pr_debug ("FAIL: failed to find a symbol\n" ); goto out; } +symbol__put (sym);if (!RC_CHK_EQUAL (map,machine__kernel_map (&m))) {pr_debug ("FAIL: the symbol is not in the kernel map\n" );
tools/perf/tests/pe-file-parsing.c +1 -0
@@ -64,6 +64,7 @@ static int run_dir(const char *d)dso__sort_by_name (dso); sym =dso__find_symbol_by_name (dso,"main" , &idx);TEST_ASSERT_VAL ("Failed to find main" , sym); +symbol__put (sym);dso__delete (dso);return TEST_OK ;
tools/perf/tests/vmlinux-kallsyms.c +10 -2
@@ -208,7 +208,7 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data) mem_end =map__unmap_ip (args->vmlinux_map,symbol__end (sym)); first_pair =machine__find_kernel_symbol (&args->kallsyms, mem_start,NULL ); - pair = first_pair; + pair =symbol__get (first_pair);if (pair &&UM (symbol__start (pair)) == mem_start) { next_pair:
@@ -237,8 +237,11 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data)* possible to get proper function end from * kallsyms. */ +symbol__put (pair); +symbol__put (first_pair);return 0 ; }else { +symbol__put (pair); pair =machine__find_kernel_symbol_by_name (&args->kallsyms,symbol__name (sym),NULL );
@@ -249,15 +252,20 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data)pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , mem_start,symbol__name (sym),symbol__name (pair)); +symbol__put (pair); }else {pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , mem_start,symbol__name (sym),symbol__name (first_pair)); } +symbol__put (first_pair);return 0 ; } - }else if (mem_start ==map__end (args->kallsyms.vmlinux_map)) { + } +symbol__put (pair); +symbol__put (first_pair); +if (mem_start ==map__end (args->kallsyms.vmlinux_map)) {/* * Ignore aliases to _etext, i.e. to the end of the kernel text area, * such as __indirect_thunk_end.
tools/perf/ui/browsers/hists.c +4 -3
@@ -2522,11 +2522,11 @@ static struct symbol *symbol__new_unresolved(u64 addr, struct map *map)if (sym) { src =symbol__hists (sym,1 );if (!src) { -symbol__delete (sym); +symbol__put (sym);return NULL ; } -dso__insert_symbol (map__dso (map), sym); +dso__insert_symbol (map__dso (map),symbol__get (sym)); }return sym;
@@ -2812,7 +2812,7 @@ add_script_opt_2(struct popup_action *act, char **optstr, } act->ms.thread =thread__get (thread); - act->ms.sym = sym; + act->ms.sym =symbol__get (sym); act->fn = do_run_script;return 1 ; }
@@ -3201,6 +3201,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *hmemset (&hotkey_act,0 ,sizeof (hotkey_act));map_symbol__copy (&hotkey_act.ms, &source_ms); +symbol__zput (source_ms.sym);do_annotate (browser, &hotkey_act);map_symbol__exit (&hotkey_act.ms); }else {
tools/perf/ui/browsers/map.c +5 -2
@@ -65,12 +65,13 @@ static int map_browser__search(struct map_browser *browser)* cursor at that entry. */ for (u32 idx =0 ; idx < browser->b.nr_entries; ++idx) { -if (entries[idx] == sym) { +if (RC_CHK_EQUAL (entries[idx], sym)) { browser->b.top = &entries[idx]; browser->b.index = browser->b.top_idx = idx;break ; } } +symbol__put (sym); }else ui_helpline__fpush ("%s not found!" , target);
@@ -136,7 +137,7 @@ int map__browse(struct map *map)for (unsigned int idx =0 ; idx < nr_entries; idx++) {struct symbol *pos = symbols->symbols[idx]; - entries[idx] = pos; + entries[idx] =symbol__get (pos);if (maxaddr <symbol__end (pos)) maxaddr =symbol__end (pos); }
@@ -151,6 +152,8 @@ int map__browse(struct map *map) mb.addrlen =snprintf (tmp,sizeof (tmp),"%" PRIx64, maxaddr); ret =map_browser__run (&mb); +for (unsigned int idx =0 ; idx < nr_entries; idx++) +symbol__put (entries[idx]);free (entries);return ret; }
tools/perf/ui/stdio/hist.c +2 -1
@@ -908,7 +908,8 @@ size_t hists__fprintf(struct hists *hists, bool show_header, int max_rows,free (line); out: -zfree (&rem_sq_bracket); +symbol__zput (rem_sq_bracket); + rem_hits.ms.sym =NULL ;return ret; }
tools/perf/util/addr_location.c +4 -0
@@ -2,6 +2,7 @@#include "addr_location.h" #include "map.h" #include "maps.h" +#include "symbol.h" #include "thread.h" void addr_location__init (struct addr_location *al)
@@ -29,13 +30,16 @@ void addr_location__exit(struct addr_location *al) {map__zput (al->map);thread__zput (al->thread); +symbol__zput (al->sym); }void addr_location__copy (struct addr_location *dst,struct addr_location *src) {thread__put (dst->thread);map__put (dst->map); +symbol__put (dst->sym); *dst = *src; dst->thread =thread__get (src->thread); dst->map =map__get (src->map); + dst->sym =symbol__get (src->sym); }
tools/perf/util/annotate-arch/annotate-alpha.c +1 -1
@@ -89,7 +89,7 @@ static int alpha_call__parse(const struct arch *arch, struct ins_operands *ops,if (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
tools/perf/util/annotate-arch/annotate-loongarch.c +2 -2
@@ -57,7 +57,7 @@ static int loongarch_call__parse(const struct arch *arch, struct ins_operands *oif (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
@@ -101,7 +101,7 @@ static int loongarch_jump__parse(const struct arch *arch, struct ins_operands *oif (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);if (!ops->target.outside) { ops->target.offset = target.addr - start;
tools/perf/util/annotate-arch/annotate-s390.c +1 -1
@@ -52,7 +52,7 @@ static int s390_call__parse(const struct arch *arch, struct ins_operands *ops,if (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
tools/perf/util/annotate.c +15 -7
@@ -194,15 +194,22 @@ static bool sym_ptr_equal(long key1, long key2, void *ctx __maybe_unused)struct annotation *symbol__annotation (struct symbol *sym) {struct annotation *notes =NULL ; -long val =0 ; +long key, val =0 ;if (!sym)return NULL ; +/* +* Key on the underlying RC_STRUCT(symbol) pointer so all +* REFCNT_CHECKING wrapper pointers for the same symbol map to the +* same annotation entry. +*/ + key = (long )RC_CHK_ACCESS (sym); +/* Fast path: look up existing annotation under read lock if flagged. */ if (symbol__annotated (sym)) {down_read (annotations_lock ()); -if (annotations &&hashmap__find (annotations, (long )sym, &val)) +if (annotations &&hashmap__find (annotations, key, &val)) notes = (struct annotation *)val;up_read (annotations_lock ());if (notes)
@@ -217,7 +224,7 @@ struct annotation *symbol__annotation(struct symbol *sym)up_write (annotations_lock ());return &empty_annotation; } - }else if (hashmap__find (annotations, (long )sym, &val)) { + }else if (hashmap__find (annotations, key, &val)) { notes = (struct annotation *)val;up_write (annotations_lock ());return notes;
@@ -229,7 +236,7 @@ struct annotation *symbol__annotation(struct symbol *sym)return &empty_annotation; }annotation__init (notes); -if (hashmap__add (annotations, (long )sym, notes) <0 ) { +if (hashmap__add (annotations, key, notes) <0 ) {free (notes);if (hashmap__size (annotations) ==0 ) {hashmap__free (annotations);
@@ -246,14 +253,15 @@ struct annotation *symbol__annotation(struct symbol *sym)void symbol__annotation_delete (struct symbol *sym) {struct annotation *notes =NULL ; -long val =0 ; +long key, val =0 ;/* Unannotated symbols avoid taking annotations_lock on teardown. */ if (!sym || !symbol__annotated (sym))return ; + key = (long )RC_CHK_ACCESS (sym);down_write (annotations_lock ()); -if (annotations &&hashmap__delete (annotations, (long )sym,NULL , &val)) { +if (annotations &&hashmap__delete (annotations, key,NULL , &val)) { notes = (struct annotation *)val;if (hashmap__size (annotations) ==0 ) {hashmap__free (annotations);
@@ -567,7 +575,7 @@ int addr_map_symbol__account_cycles(struct addr_map_symbol *ams,* it starts on the function start. */ if (start && - (start->ms.sym == ams->ms.sym || + (RC_CHK_EQUAL (start->ms.sym, ams->ms.sym) || (ams->ms.sym && start->addr ==symbol__start (ams->ms.sym) +map__start (ams->ms.map)))) saddr = start->al_addr;
tools/perf/util/block-info.c +4 -2
@@ -64,8 +64,10 @@ static struct block_info *block_info__new(unsigned int br_cntr_nr)void block_info__delete (struct block_info *bi) { -if (bi) +if (bi) { +symbol__zput (bi->sym);free (bi->br_cntr); + }free (bi); }
@@ -105,7 +107,7 @@ static void init_block_info(struct block_info *bi, struct symbol *sym, u64 total_cycles, unsigned int br_cntr_nr, u64 *br_cntr, struct evsel *evsel) { - bi->sym = sym; + bi->sym = symbol__get(sym); bi->start = ch->start; bi->end = offset; bi->cycles = ch->cycles;
tools/perf/util/bpf_lock_contention.c +20 -7
@@ -145,6 +145,7 @@ static void init_numa_data(struct lock_contention *con)if (sym) { skel->rodata->contig_page_data_addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym);map__put (kmap);return ; }
@@ -161,6 +162,7 @@ static void init_numa_data(struct lock_contention *con)return ; skel->rodata->node_data_addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym);map__put (kmap);/* get the number of online nodes using the last node number + 1 */
@@ -264,12 +266,16 @@ int lock_contention_prepare(struct lock_contention *con) (con->filters->nr_addrs + 1) * sizeof(*addrs));if (addrs ==NULL ) {pr_warning ("memory allocation failure\n" ); +symbol__put (sym); +map__put (kmap);continue ; } addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); con->filters->addrs = addrs; +symbol__put (sym); +map__put (kmap); } naddrs = con->filters->nr_addrs ?: has_mmap_lock; skel->rodata->has_addr =1 ;
@@ -292,6 +298,8 @@ int lock_contention_prepare(struct lock_contention *con) con->delays[i].addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym); +map__put (kmap); } skel->rodata->lock_delay =1 ;bpf_map__set_max_entries (skel->maps.lock_delays, con->nr_delays);
@@ -609,9 +617,12 @@ static const char *lock_contention_get_name(struct lock_contention *con,return "siglock" ;/* global locks with symbols */ - sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup, &kmap); -if (sym) -return symbol__name (sym); + sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup,NULL ); +if (sym) { +snprintf (name_buf,sizeof (name_buf),"%s" ,symbol__name (sym)); +symbol__put (sym); +return name_buf; + }/* try semi-global locks collected separately */ if (!bpf_map_lookup_elem (lock_fd, &key->lock_addr_or_cgroup, &flags)) {
@@ -658,13 +669,15 @@ static const char *lock_contention_get_name(struct lock_contention *con, offset =map__map_ip (kmap, addr) -symbol__start (sym);if (offset ==0 ) -return symbol__name (sym); - -snprintf (name_buf,sizeof (name_buf),"%s+%#lx" , -symbol__name (sym), offset); +snprintf (name_buf,sizeof (name_buf),"%s" ,symbol__name (sym)); +else +snprintf (name_buf,sizeof (name_buf),"%s+%#lx" , +symbol__name (sym), offset); +symbol__put (sym); }else {snprintf (name_buf,sizeof (name_buf),"%#lx" , (unsigned long )addr); } +map__put (kmap);return name_buf; }
tools/perf/util/c2c-function.c +8 -5
@@ -976,13 +976,13 @@ c2c_function_hists__level1_entry(struct symbol *sym,/* mem_info__put() will map_symbol__exit() these, so take refs. */ mem_info__iaddr (mi)->ms.thread =thread__get (detail_he->ms.thread);mem_info__iaddr (mi)->ms.map =map__get (detail_he->ms.map); -mem_info__iaddr (mi)->ms.sym = sym; +mem_info__iaddr (mi)->ms.sym =symbol__get (sym);mem_info__daddr (mi)->addr =0 ;addr_location__init (&al); al.thread =thread__get (synthetic_thread); al.map =map__get (detail_he->ms.map); - al.sym = sym; + al.sym =symbol__get (sym); al.addr = sym_start; al.level = detail_he->level; al.cpumode = detail_he->cpumode;
@@ -1064,9 +1064,12 @@ c2c_function_hists__level2_entry(struct c2c_hist_entry *level1_c2c,return NULL ;/* Key this level by the looked-up symbol, not detail_he's. */ - level2_c2c->he.ms.sym = sym; -if (level2_c2c->he.mem_info) -mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym = sym; +symbol__put (level2_c2c->he.ms.sym); + level2_c2c->he.ms.sym =symbol__get (sym); +if (level2_c2c->he.mem_info) { +symbol__put (mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym); +mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym =symbol__get (sym); + }c2c_child_entry__insert (level1_he, &level2_c2c->he, p, parent, leftmost);
tools/perf/util/call-path.c +12 -3
@@ -10,12 +10,13 @@#include <stdlib.h> #include "call-path.h" +#include "symbol.h" static void call_path__init (struct call_path *cp,struct call_path *parent,struct symbol *sym,u64 ip,bool in_kernel) { cp->parent = parent; - cp->sym = sym; + cp->sym =symbol__get (sym); cp->ip = sym ?0 : ip; cp->db_id =0 ; cp->in_kernel = in_kernel;
@@ -38,8 +39,15 @@ struct call_path_root *call_path_root__new(void)void call_path_root__free (struct call_path_root *cpr) {struct call_path_block *pos, *n; +size_t i =0 ; +if (!cpr) +return ; + +symbol__zput (cpr->call_path.sym);list_for_each_entry_safe (pos, n, &cpr->blocks, node) { +for (size_t j =0 ; j <CALL_PATH_BLOCK_SIZE && i < cpr->next; j++, i++) +symbol__zput (pos->cp[j].sym);list_del_init (&pos->node);free (pos); }
@@ -94,10 +102,11 @@ struct call_path *call_path__findnew(struct call_path_root *cpr, node_parent = *p; cp = rb_entry(node_parent, struct call_path, rb_node); -if (cp->sym == sym && cp->ip == ip) +if (RC_CHK_EQUAL (cp->sym, sym) && cp->ip == ip)return cp; -if (sym < cp->sym || (sym == cp->sym && ip < cp->ip)) +if ((sym ?RC_CHK_ACCESS (sym) :NULL ) < (cp->sym ?RC_CHK_ACCESS (cp->sym) :NULL ) || + (RC_CHK_EQUAL (sym, cp->sym) && ip < cp->ip)) p = &(*p)->rb_left;else p = &(*p)->rb_right;
tools/perf/util/callchain.c +3 -1
@@ -1094,6 +1094,7 @@ merge_chain_branch(struct callchain_cursor *cursor,struct map_symbol ms = { .thread =thread__get (list->ms.thread), .map =map__get (list->ms.map), + .sym =symbol__get (list->ms.sym), };callchain_cursor_append (cursor, list->ip, &ms, false,NULL ,0 ,0 ,0 , list->srcline);list_del_init (&list->list);
@@ -1204,7 +1205,8 @@ int fill_callchain_info(struct addr_location *al, struct callchain_cursor_node *map__put (al->map); al->map =map__get (node->ms.map); - al->sym = node->ms.sym; +symbol__put (al->sym); + al->sym =symbol__get (node->ms.sym); al->srcline = node->srcline; al->addr = node->ip;
tools/perf/util/capstone.c +1 -0
@@ -373,6 +373,7 @@ static void print_capstone_detail(struct cs_insn *insn, char *buf, size_t len,scnprintf (buf, len,"\t# %" PRIx64" <%s+%#" PRIx64">" , orig_addr,symbol__name (sym), sym_offset); } +symbol__put (sym);map__put (found_map);break ; }
tools/perf/util/db-export.c +28 -7
@@ -22,14 +22,24 @@#include "db-export.h" #include "hashmap.h" +/* +* Hash and compare the underlying RC_STRUCT(symbol) pointer so all +* REFCNT_CHECKING wrapper pointers for the same symbol resolve to the same +* exported database ID. +*/ static size_t sym_ptr_hash (long key,void *ctx__maybe_unused ) { -return hash_bits ((size_t )key,16 ); +const struct symbol *sym = (const struct symbol *)key; + +return hash_bits ((size_t )RC_CHK_ACCESS (sym),16 ); }static bool sym_ptr_equal (long key1,long key2,void *ctx__maybe_unused ) { -return key1 == key2; +const struct symbol *sym1 = (const struct symbol *)key1; +const struct symbol *sym2 = (const struct symbol *)key2; + +return RC_CHK_EQUAL (sym1, sym2); }int db_export__init (struct db_export *dbe)
@@ -40,8 +50,15 @@ int db_export__init(struct db_export *dbe)void db_export__exit (struct db_export *dbe) { -hashmap__free (dbe->sym_db_ids); - dbe->sym_db_ids =NULL ; +if (dbe->sym_db_ids) { +struct hashmap_entry *cur; +size_t bkt; + +hashmap__for_each_entry (dbe->sym_db_ids, cur, bkt) +symbol__put ((struct symbol *)cur->pkey); +hashmap__free (dbe->sym_db_ids); + dbe->sym_db_ids =NULL ; + }call_return_processor__free (dbe->crp); dbe->crp =NULL ; }
@@ -187,6 +204,7 @@ int db_export__symbol(struct db_export *dbe, struct symbol *sym, struct dso *dso) {long db_id; +struct symbol *sym_key;if (db_export__symbol_db_id (dbe, sym))return 0 ;
@@ -202,8 +220,11 @@ int db_export__symbol(struct db_export *dbe, struct symbol *sym, } db_id = ++dbe->symbol_last_db_id; -if (hashmap__set (dbe->sym_db_ids, (long )sym, db_id,NULL ,NULL )) + sym_key =symbol__get (sym); +if (hashmap__set (dbe->sym_db_ids, (long )sym_key, db_id,NULL ,NULL )) { +symbol__put (sym_key);return -ENOMEM ; + }if (dbe->export_symbol)return dbe->export_symbol (dbe, sym, dso);
@@ -227,7 +248,7 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al,if (!al->sym) { al->sym =symbol__new (al->addr,0 ,0 ,0 ,"unknown" );if (al->sym) -dso__insert_symbol (dso, al->sym); +dso__insert_symbol (dso,symbol__get (al->sym)); }if (al->sym) {
@@ -287,7 +308,7 @@ static struct call_path *call_path_from_sample(struct db_export *dbe,* db_ids_from_al() to perform the export. */ addr_location__init (&al); - al.sym = node->ms.sym; + al.sym =symbol__get (node->ms.sym); al.map =map__get (node->ms.map); al.addr = node->ip; al.thread =thread__get (thread);
tools/perf/util/disasm.c +8 -4
@@ -228,6 +228,7 @@ static void ins_ops__delete(struct ins_operands *ops)zfree (&ops->source.name);zfree (&ops->target.raw);zfree (&ops->target.name); +symbol__zput (ops->target.sym); } int ins__raw_scnprintf(const struct ins *ins, char *bf, size_t size,
@@ -290,7 +291,7 @@ static int call__parse(const struct arch *arch, struct ins_operands *ops, structif (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
@@ -425,7 +426,7 @@ static int jump__parse(const struct arch *arch, struct ins_operands *ops, struct*/ if (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);if (!ops->target.outside) { ops->target.offset = target.addr - start;
@@ -471,12 +472,14 @@ int jump__scnprintf(const struct ins *ins, char *bf, size_t size, ops->target.offset); } -void jump__delete (struct ins_operands *ops__maybe_unused ) +void jump__delete (struct ins_operands *ops) {/* * The ops->jump.raw_comment and ops->jump.raw_func_start belong to the * raw string, don't free them. */ +if (ops) +symbol__zput (ops->target.sym); }const struct ins_ops jump_ops = {
@@ -570,6 +573,7 @@ static void lock__delete(struct ins_operands *ops)zfree (&ops->locked.ins.name);zfree (&ops->target.raw);zfree (&ops->target.name); +symbol__zput (ops->target.sym); }const struct ins_ops lock_ops = {
@@ -1090,7 +1094,7 @@ static int symbol__parse_objdump_line(struct symbol *sym,if (!maps__find_ams (thread__maps (args->ms->thread), &target) &&symbol__start (target.ms.sym) == target.al_addr) - dl->ops.target.sym = target.ms.sym; + dl->ops.target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target); }
tools/perf/util/event.c +4 -4
@@ -770,7 +770,7 @@ struct map *thread__find_map_fb(struct thread *thread, u8 cpumode, u64 addr,struct symbol *thread__find_symbol (struct thread *thread,u8 cpumode,u64 addr,struct addr_location *al) { - al->sym =NULL ; +symbol__zput (al->sym);if (thread__find_map (thread, cpumode, addr, al)) al->sym =map__find_symbol (al->map, al->addr);return al->sym;
@@ -779,7 +779,7 @@ struct symbol *thread__find_symbol(struct thread *thread, u8 cpumode,struct symbol *thread__find_symbol_fb (struct thread *thread,u8 cpumode,u64 addr,struct addr_location *al) { - al->sym =NULL ; +symbol__zput (al->sym);if (thread__find_map_fb (thread, cpumode, addr, al)) al->sym =map__find_symbol (al->map, al->addr);return al->sym;
@@ -829,7 +829,7 @@ int machine__resolve(struct machine *machine, struct addr_location *al,thread__put (thread); thread =NULL ; - al->sym =NULL ; +symbol__zput (al->sym); al->cpu = sample->cpu; al->socket =-1 ; al->srcline =NULL ;
@@ -937,7 +937,7 @@ void thread__resolve(struct thread *thread, struct addr_location *al,thread__find_map_fb (thread, sample->cpumode, sample->addr, al); al->cpu = sample->cpu; - al->sym =NULL ; +symbol__zput (al->sym);if (al->map) al->sym =map__find_symbol (al->map, al->addr);
tools/perf/util/hist.c +12 -1
@@ -542,6 +542,7 @@ static int hist_entry__init(struct hist_entry *he, he->ms.thread =thread__get (he->ms.thread); he->ms.map =map__get (he->ms.map); + he->ms.sym =symbol__get (he->ms.sym);if (he->branch_info) {/*
@@ -558,8 +559,10 @@ static int hist_entry__init(struct hist_entry *he, he->branch_info->from.ms.thread =thread__get (he->branch_info->from.ms.thread); he->branch_info->from.ms.map =map__get (he->branch_info->from.ms.map); + he->branch_info->from.ms.sym =symbol__get (he->branch_info->from.ms.sym); he->branch_info->to.ms.thread =thread__get (he->branch_info->to.ms.thread); he->branch_info->to.ms.map =map__get (he->branch_info->to.ms.map); + he->branch_info->to.ms.sym =symbol__get (he->branch_info->to.ms.sym); }if (he->mem_info) {
@@ -592,6 +595,7 @@ static int hist_entry__init(struct hist_entry *he,INIT_LIST_HEAD (&he->pairs.node); he->thread =thread__get (he->thread); + he->parent =symbol__get (he->parent); he->hroot_in =RB_ROOT_CACHED ; he->hroot_out =RB_ROOT_CACHED ;
@@ -727,6 +731,8 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,if (hists__has (hists, sym) && he->ms.map != entry->ms.map) {if (he->ms.sym) {u64 addr =symbol__start (he->ms.sym); + +symbol__put (he->ms.sym); he->ms.sym =map__find_symbol (entry->ms.map, addr); }
@@ -1028,7 +1034,8 @@ iter_next_branch_entry(struct hist_entry_iter *iter, struct addr_location *al) al->thread =thread__get (bi[i].to.ms.thread);map__put (al->map); al->map =map__get (bi[i].to.ms.map); - al->sym = bi[i].to.ms.sym; +symbol__put (al->sym); + al->sym =symbol__get (bi[i].to.ms.sym); al->addr = bi[i].to.addr;return 1 ; }
@@ -1342,9 +1349,11 @@ int hist_entry_iter__add(struct hist_entry_iter *iter, struct addr_location *al,if (al) alm =map__get (al->map); + iter->parent =NULL ; err =sample__resolve_callchain (iter->sample,get_tls_callchain_cursor (), &iter->parent, al, max_stack_depth);if (err) { +symbol__zput (iter->parent);map__put (alm);return err; }
@@ -1380,6 +1389,7 @@ int hist_entry_iter__add(struct hist_entry_iter *iter, struct addr_location *al,if (!err) err = err2; +symbol__zput (iter->parent);map__put (alm);return err;
@@ -1466,6 +1476,7 @@ void hist_entry__delete(struct hist_entry *he) }thread__zput (he->thread); +symbol__zput (he->parent);map_symbol__exit (&he->ms);if (he->branch_info) {
tools/perf/util/libdw.c +2 -3
@@ -98,7 +98,7 @@ static int libdw_a2l_cb(Dwarf_Die *die, void *_args)* while DWARF keeps the pre-clone linkage name, which * left the outer frame spuriously tagged "(inlined)". */ - inline_sym = args->sym; + inline_sym =symbol__get (args->sym); }else {/* * Prefer DW_AT_linkage_name so C++ inline frames keep
@@ -156,8 +156,7 @@ static int libdw_a2l_cb(Dwarf_Die *die, void *_args)return 0 ; abort_delete_sym: -if (symbol__inlined (inline_sym)) -symbol__delete (inline_sym); +symbol__put (inline_sym); abort_enomem: args->err = -ENOMEM ;return DWARF_CB_ABORT ;
tools/perf/util/lock-contention.c +5 -3
@@ -101,7 +101,6 @@ struct lock_stat *lock_stat_findnew(u64 addr, const char *name, int flags)bool match_callstack_filter (struct machine *machine,u64 *callstack,int max_stack_depth) { -struct map *kmap;struct symbol *sym;u64 ip;uint16_t e_machine =perf_env__e_machine (machine->env,/*e_flags=*/ NULL );
@@ -130,14 +129,17 @@ bool match_callstack_filter(struct machine *machine, u64 *callstack, int max_stabreak ; ip = callstack[i]; - sym =machine__find_kernel_symbol (machine, ip, &kmap); + sym =machine__find_kernel_symbol (machine, ip,NULL );if (sym ==NULL )continue ;list_for_each_entry (filter, &callstack_filters, list) { -if (strstr (symbol__name (sym), filter->name)) +if (strstr (symbol__name (sym), filter->name)) { +symbol__put (sym);return true; + } } +symbol__put (sym); }return false; }
tools/perf/util/machine.c +33 -15
146 lines not on this page · raw message
tools/perf/util/map.h +6 -4
@@ -160,13 +160,15 @@ struct thread;* @sym_name: the symbol name * @pos: the 'struct symbol *' to use as a loop cursor * @idx: the cursor index in the symbol names array +* Note: caller must call symbol__put(pos) if breaking out of the loop early. */ #define __map__for_each_symbol_by_name (map, sym_name, pos, idx) \ for (pos = map__find_symbol_by_name_idx(map, sym_name, &idx); \ - pos && \ - !symbol__match_symbol_name(symbol__name(pos), sym_name, \ - SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); \ - pos = dso__next_symbol_by_name(map__dso(map), &idx)) + pos && \ + (!symbol__match_symbol_name(symbol__name(pos), sym_name, \ + SYMBOL_TAG_INCLUDE__DEFAULT_ONLY) || \ + (symbol__zput(pos), false)); \ + symbol__put(pos), pos = dso__next_symbol_by_name(map__dso(map), &idx))#define map__for_each_symbol_by_name (map, sym_name, pos, idx) \ __map__for_each_symbol_by_name(map, sym_name, (pos), idx)
tools/perf/util/map_symbol.c +3 -1
@@ -2,12 +2,14 @@#include "map_symbol.h" #include "maps.h" #include "map.h" +#include "symbol.h" #include "thread.h" void map_symbol__exit (struct map_symbol *ms) {thread__zput (ms->thread);map__zput (ms->map); +symbol__zput (ms->sym); }void addr_map_symbol__exit (struct addr_map_symbol *ams)
@@ -19,7 +21,7 @@ void map_symbol__copy(struct map_symbol *dst, struct map_symbol *src) { dst->thread =thread__get (src->thread); dst->map =map__get (src->map); - dst->sym = src->sym; + dst->sym =symbol__get (src->sym); }void addr_map_symbol__copy (struct addr_map_symbol *dst,struct addr_map_symbol *src)
tools/perf/util/maps.c +2 -1
@@ -788,7 +788,7 @@ struct symbol *maps__find_symbol_by_name(struct maps *maps, const char *name, st *mapp =map__get (map);break ; } - sym =NULL ; +symbol__zput (sym); }for (unsigned int i =0 ; i < nr_maps; i++)
@@ -810,6 +810,7 @@ int maps__find_ams(struct maps *maps, struct addr_map_symbol *ams) } ams->al_addr =map__map_ip (ams->ms.map, ams->addr); +symbol__put (ams->ms.sym); ams->ms.sym =map__find_symbol (ams->ms.map, ams->al_addr);return ams->ms.sym ?0 :-1 ;
tools/perf/util/probe-event.c +14 -3
@@ -152,6 +152,8 @@ static int kernel_get_symbol_address_by_name(const char *name, u64 *addr, *addr =map__unmap_ip (map,symbol__start (sym)) - ((reloc) ?0 :map__reloc (map)) - ((reladdr) ?map__start (map) :0 ); +symbol__put (sym); +map__put (map); }return 0 ; }
@@ -422,6 +424,7 @@ static int find_alternative_probe_point(struct debuginfo *dinfo, pp->function); }else address =map__unmap_ip (map,symbol__start (sym)) -map__reloc (map); +symbol__put (sym);break ; }if (!address) {
@@ -728,11 +731,14 @@ static int post_process_probe_trace_point(struct probe_trace_point *tp,else free (tp->symbol); tp->symbol =strdup (symbol__name (sym)); -if (!tp->symbol) +if (!tp->symbol) { +symbol__put (sym);return -ENOMEM ; + } } tp->offset = addr -symbol__start (sym); tp->address -= offs; +symbol__put (sym);return 0 ; }
@@ -2314,6 +2320,7 @@ static int find_perf_probe_point_from_map(struct probe_trace_point *tp, ret = pp->function ?0 : -ENOMEM ; out: +symbol__put (sym);map__put (map);return ret;
@@ -3078,7 +3085,7 @@ static int find_probe_functions_cb(struct symbol *sym, void *data)if (strglobmatch (norm, args->name)) { args->found++;if (args->syms && args->found < probe_conf.max_probes) - args->syms[args->found -1 ] = sym; + args->syms[args->found -1 ] =symbol__get (sym); }free (buf);return 0 ;
@@ -3146,7 +3153,7 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, goto out; } - syms =malloc (sizeof (struct symbol *) * probe_conf.max_probes); + syms =calloc (probe_conf.max_probes,sizeof (struct symbol *));if (!syms) { ret = -ENOMEM ; goto out;
@@ -3291,6 +3298,10 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, } out: +if (syms) { +for (i =0 ; i < probe_conf.max_probes; i++) +symbol__put (syms[i]); + }map__put (map);free (syms);return ret;
tools/perf/util/python.c +8 -4
@@ -814,7 +814,7 @@ static PyObject *pyrf_sample_event__srccode(PyObject *self, PyObject *args)addr_location__init (&al); al.thread =thread__get (pevent->al.thread); al.map =map__get (pevent->al.map); - al.sym = pevent->al.sym; + al.sym =symbol__get (pevent->al.sym); al.addr = pevent->al.addr; }
@@ -877,6 +877,7 @@ struct pyrf_callchain_node {static void pyrf_callchain_node__delete (struct pyrf_callchain_node *pnode) {map__put (pnode->map); +symbol__put (pnode->sym);Py_TYPE (pnode)->tp_free ((PyObject *)pnode); }
@@ -944,8 +945,10 @@ struct pyrf_callchain {static void pyrf_callchain__delete (struct pyrf_callchain *pchain) {if (pchain->frames) { -for (u64 i =0 ; i < pchain->nr_frames; i++) +for (u64 i =0 ; i < pchain->nr_frames; i++) {map__put (pchain->frames[i].map); +symbol__put (pchain->frames[i].sym); + }free (pchain->frames); }Py_TYPE (pchain)->tp_free ((PyObject *)pchain);
@@ -974,7 +977,7 @@ static PyObject *pyrf_callchain__item(PyObject *obj, Py_ssize_t i) pnode->ip = pchain->frames[i].ip; pnode->map =map__get (pchain->frames[i].map); - pnode->sym = pchain->frames[i].sym; + pnode->sym =symbol__get (pchain->frames[i].sym);return (PyObject *)pnode; }
@@ -1486,7 +1489,8 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *ev pchain->frames[i].ip = node->ip; pchain->frames[i].map =map__get (node->ms.map); - pchain->frames[i].sym = node->ms.sym; + pchain->frames[i].sym = +symbol__get (node->ms.sym);callchain_cursor_advance (cursor); } pevent->callchain = (PyObject *)pchain;
tools/perf/util/sort.c +1 -1
@@ -466,7 +466,7 @@ int64_t _sort__sym_cmp(struct symbol *sym_l, struct symbol *sym_r)if (!sym_l || !sym_r)return cmp_null (sym_l, sym_r); -if (sym_l == sym_r) +if (RC_CHK_EQUAL (sym_l, sym_r))return 0 ;if (symbol__inlined (sym_l) ||symbol__inlined (sym_r)) {
tools/perf/util/srcline.c +2 -4
@@ -118,7 +118,7 @@ struct symbol *new_inline_sym(struct dso *dso,if (base_sym &&strcmp (funcname,symbol__name (base_sym)) ==0 ) {/* reuse the real, existing symbol */ - inline_sym = base_sym; + inline_sym =symbol__get (base_sym);/* ensure that we don't alias an inlined symbol, which could * lead to double frees in inline_node__delete */
@@ -468,9 +468,7 @@ void inline_node__clear_frames(struct inline_node *node)list_for_each_entry_safe (ilist, tmp, &node->val, list) {list_del_init (&ilist->list);zfree_srcline (&ilist->srcline); -/* only the inlined symbols are owned by the list */ -if (ilist->symbol &&symbol__inlined (ilist->symbol)) -symbol__delete (ilist->symbol); +symbol__zput (ilist->symbol);free (ilist); } }
tools/perf/util/symbol-elf.c +5 -1
@@ -351,10 +351,13 @@ static bool get_ifunc_name(Elf *elf, struct dso *dso, GElf_Ehdr *ehdr,/* Expecting the address to be an IFUNC or IFUNC alias */ if (!sym ||symbol__start (sym) != addr || - (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym))) + (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym))) { +symbol__put (sym);return false; + }snprintf (buf, buf_sz,"%s@plt" ,symbol__name (sym)); +symbol__put (sym);return true; }
@@ -631,6 +634,7 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss) f =dso__find_symbol_nocache (dso, shdr_plt.sh_offset);if (f &&symbol__start (f) < shdr_plt.sh_offset &&symbol__end (f) > shdr_plt.sh_offset)symbol__set_end (f, shdr_plt.sh_offset); +symbol__put (f);if (!get_plt_sizes (dso, &ehdr, &shdr_plt, &plt_header_size, &plt_entry_size))return 0 ;
tools/perf/util/symbol.c +82 -47
270 lines not on this page · raw message
tools/perf/util/symbol.h +44 -16
145 lines not on this page · raw message
tools/perf/util/thread-stack.c +6 -6
@@ -823,7 +823,7 @@ static int thread_stack__pop_cp(struct thread *thread, struct thread_stack *ts,if (ts->cnt ==1 ) {struct thread_stack_entry *tse = &ts->stack[0 ]; -if (tse->cp->sym == sym) +if (RC_CHK_EQUAL (tse->cp->sym, sym))return thread_stack__call_return (thread, ts, --ts->cnt, timestamp, ref, false); }
@@ -941,7 +941,7 @@ static int thread_stack__no_call_return(struct thread *thread,else parent = root; -if (parent->sym == from_al->sym) { +if (RC_CHK_EQUAL (parent->sym, from_al->sym)) {/* * At the bottom of the stack, assume the missing 'call' was * before the trace started. So, pop the current symbol and push
@@ -1069,7 +1069,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,*/ ts->cnt -=1 ; sym = ts->stack[ts->cnt -2 ].cp->sym; -if (sym && sym == tsym && to_al->addr !=symbol__start (tsym)) { +if (sym &&RC_CHK_EQUAL (sym, tsym) && to_al->addr !=symbol__start (tsym)) {/* * Target is back to the middle of the symbol we came * from so assume it is an indirect jmp and forget it
@@ -1078,7 +1078,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts, ts->cnt -=1 ;return 0 ; } - }else if (sym && sym == tsym) { + }else if (sym &&RC_CHK_EQUAL (sym, tsym)) {/* * Target is back to the symbol we came from so assume it is an * indirect jmp and forget it altogether.
@@ -1169,7 +1169,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm,* may be the start of a x86 retpoline. */ if (!err && rstate ==X86_RETPOLINE_POSSIBLE && to_al->sym && - from_al->sym == to_al->sym && +RC_CHK_EQUAL (from_al->sym, to_al->sym) && to_al->addr !=symbol__start (to_al->sym)) ts->rstate =X86_RETPOLINE_DETECTED ;
@@ -1206,7 +1206,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm, }else if (sample->flags &PERF_IP_FLAG_TRACE_END ) { err =thread_stack__trace_end (ts, sample, ref); }else if (sample->flags &PERF_IP_FLAG_BRANCH && - from_al->sym != to_al->sym && to_al->sym && + !RC_CHK_EQUAL (from_al->sym, to_al->sym) && to_al->sym && to_al->addr ==symbol__start (to_al->sym)) {struct call_path_root *cpr = ts->crp->cpr;struct call_path *cp;
tools/perf/util/threads.c +11 -6
@@ -142,10 +142,13 @@ void threads__remove_all_threads(struct threads *threads)down_write (&table->lock);__threads_table_entry__set_last_match (table,NULL );hashmap__for_each_entry_safe (&table->shard, cur, tmp, bkt) { -struct thread *old_value; +struct thread *old_value =NULL ; -hashmap__delete (&table->shard, cur->key,/*old_key=*/ NULL , &old_value); -thread__put (old_value); +if (hashmap__delete (&table->shard, cur->key, +/*old_key=*/ NULL , &old_value)) { +thread__free_stitch_list (old_value); +thread__put (old_value); + } }up_write (&table->lock); }
@@ -154,14 +157,16 @@ void threads__remove_all_threads(struct threads *threads)void threads__remove (struct threads *threads,struct thread *thread) {struct threads_table_entry *table =threads__table (threads,thread__tid (thread)); -struct thread *old_value; +struct thread *old_value =NULL ;down_write (&table->lock);if (table->last_match &&RC_CHK_EQUAL (table->last_match, thread))__threads_table_entry__set_last_match (table,NULL ); -hashmap__delete (&table->shard,thread__tid (thread),/*old_key=*/ NULL , &old_value); -thread__put (old_value); +if (hashmap__delete (&table->shard,thread__tid (thread),/*old_key=*/ NULL , &old_value)) { +thread__free_stitch_list (old_value); +thread__put (old_value); + }up_write (&table->lock); }
tools/perf/util/unwind-libdw.c +1 -1
@@ -169,7 +169,7 @@ static int entry(u64 ip, struct unwind_info *ui) e->ip = ip; e->ms.thread =thread__get (al.thread); e->ms.map =map__get (al.map); - e->ms.sym = al.sym; + e->ms.sym =symbol__get (al.sym); pr_debug("unwind: %s:ip = 0x%" PRIx64 " (0x%" PRIx64 ")\n", al.sym ? symbol__name(al.sym) : "''",
tools/perf/util/unwind-libunwind.c +1 -1
@@ -632,7 +632,7 @@ static int entry(uint64_t ip, struct thread *thread, unwind_entry_cb_t cb, void e.ms.thread =thread__get (al.thread);pr_debug ("unwind: %s:ip = 0x%" PRIx64" (0x%" PRIx64")\n" , - al.sym ? al.sym->name :"''" , + al.sym ?symbol__name (al.sym) :"''" , ip, al.map ?map__map_ip (al.map, ip) : (u64 )0 );
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 5/7 perf symbol: Add LRU memory shrinking for symbols, DSOs, and machines
In reply to Ian RogersAdd Clock/Second-Chance LRU access tracking (SYMBOL_FLAG_ACCESSED on struct symbol, accessed on struct dso, and shrunk on struct symbols) and memory-shrinking helpers (symbols__shrink, dso__can_shrink_symbols, dso__shrink, dsos__shrink, machine__shrink, machines__shrink, and perf_session__shrink):
- symbols__shrink() scans a symbol table under symbols->lock: symbols with refcnt > 1 are retained, symbols with SYMBOL_FLAG_ACCESSED set have their accessed bit cleared (second chance), and unreferenced symbols with SYMBOL_FLAG_ACCESSED clear are evicted. When symbols are evicted, symbols->shrunk is set under symbols->lock and the backing array is compacted when occupancy drops below 50%.
- dso__shrink() closes any cached file descriptor, frees cached 4KB DSO data pages (__dso_cache__free()) and addr2line/libbfd/libdw state under dso__lock(dso), and shrinks dso->symbols if the DSO supports idempotent on-demand reloading (dso__can_shrink_symbols()). Protect dso_cache lookup and memcpy under dso__lock(dso) so concurrent shrinking cannot free a cache page while it is being read.
- dsos__shrink() sweeps a DSO collection using Clock/Second-Chance LRU:
accessed DSOs have their symbol tables shrunk and their accessed bit cleared; unaccessed DSOs are fully shrunk via dso__shrink() and evicted from struct dsos if unreferenced outside the collection (refcnt == 1) and free of unreloadable state.
- When a DSO's symbol table has been shrunk (dso__symbols_shrunk(dso) is true) and a subsequent symbol lookup misses in map__find_symbol() or map__find_symbol_by_name_idx(), transparently reload the DSO on demand while preferring actively referenced (refcnt > 1) symbols in choose_best_symbol() and deduplicating identical (start, name) pairs in symbols__fixup_duplicate() even when symbol_conf.allow_aliases is enabled.
--- tools/perf/util/dso.c | 182 +++++++++++++++++++++++++++-------- tools/perf/util/dso.h | 46 +++++++++ tools/perf/util/dsos.c | 100 +++++++++++++++++++ tools/perf/util/dsos.h | 12 ++- tools/perf/util/machine.c | 39 ++++++++ tools/perf/util/machine.h | 2 + tools/perf/util/map.c | 26 ++++- tools/perf/util/session.c | 15 +++ tools/perf/util/session.h | 3 + tools/perf/util/symbol-elf.c | 5 +- tools/perf/util/symbol.c | 142 +++++++++++++++++++++++++-- tools/perf/util/symbol.h | 36 +++++-- 12 files changed, 551 insertions(+), 57 deletions(-)
tools/perf/util/dso.c +144 -38
@@ -877,6 +877,7 @@ bool dso__data_get_fd(struct dso *dso, struct machine *machine, int *fd)if (dso__data (dso)->status ==DSO_DATA_STATUS_ERROR )return false; +dso__set_accessed (dso, true);mutex_lock (dso__data_open_lock ());try_to_open_dso (dso, machine);
@@ -955,12 +956,11 @@ static int bpf_size(struct dso *dso)#endif // HAVE_LIBBPF_SUPPORT static void -dso_cache__free (struct dso *dso) +__dso_cache__free (struct dso *dso) {struct rb_root *root = &dso__data (dso)->cache;struct rb_node *next =rb_first (root); -mutex_lock (dso__lock (dso));while (next) {struct dso_cache *cache;
@@ -969,6 +969,13 @@ dso_cache__free(struct dso *dso)rb_erase (&cache->rb_node, root);free (cache); } +} + +static void +dso_cache__free (struct dso *dso) +{ +mutex_lock (dso__lock (dso)); +__dso_cache__free (dso);mutex_unlock (dso__lock (dso)); }
@@ -998,7 +1005,7 @@ static struct dso_cache *__dso_cache__find(struct dso *dso, u64 offset) }static struct dso_cache * -dso_cache__insert (struct dso *dso,struct dso_cache *new) +__dso_cache__insert (struct dso *dso,struct dso_cache *new) {struct rb_root *root = &dso__data (dso)->cache;struct rb_node **p = &root->rb_node;
@@ -1006,7 +1013,6 @@ dso_cache__insert(struct dso *dso, struct dso_cache *new)struct dso_cache *cache;u64 offset = new->offset; -mutex_lock (dso__lock (dso));while (*p !=NULL ) {u64 end;
@@ -1019,16 +1025,13 @@ dso_cache__insert(struct dso *dso, struct dso_cache *new)else if (offset >= end) p = &(*p)->rb_right;else - goto out; +return cache; }rb_link_node (&new->rb_node, parent, p);rb_insert_color (&new->rb_node, root); - cache =NULL ; -out: -mutex_unlock (dso__lock (dso)); -return cache; +return NULL ; } static ssize_t dso_cache__memcpy(struct dso_cache *cache, u64 offset, u8 *data,
@@ -1082,68 +1085,69 @@ static ssize_t file_read(struct dso *dso, struct machine *machine,return ret; } -static struct dso_cache *dso_cache__populate (struct dso *dso, -struct machine *machine, -u64 offset,ssize_t *ret) +static ssize_t dso_cache__populate (struct dso *dso,struct machine *machine, +u64 offset,u8 *data,ssize_t size,bool out) {u64 cache_offset = offset &DSO__DATA_CACHE_MASK ;struct dso_cache *cache;struct dso_cache *old; +ssize_t ret; cache =zalloc (sizeof (*cache) +DSO__DATA_CACHE_SIZE ); -if (!cache) { - *ret = -ENOMEM ; -return NULL ; - } +if (!cache) +return -ENOMEM ;#ifdef HAVE_LIBBPF_SUPPORT if (dso__binary_type (dso) ==DSO_BINARY_TYPE__BPF_PROG_INFO ) - *ret =bpf_read (dso, cache_offset, cache->data); + ret =bpf_read (dso, cache_offset, cache->data); else#endif if (dso__binary_type (dso) ==DSO_BINARY_TYPE__OOL ) - *ret =DSO__DATA_CACHE_SIZE ; + ret =DSO__DATA_CACHE_SIZE ;else - *ret =file_read (dso, machine, cache_offset, cache->data); + ret =file_read (dso, machine, cache_offset, cache->data); -if (*ret <=0 ) { +if (ret <=0 ) {free (cache); -return NULL ; +return ret; } cache->offset = cache_offset; - cache->size = *ret; + cache->size = ret; - old =dso_cache__insert (dso, cache); +mutex_lock (dso__lock (dso)); + old =__dso_cache__insert (dso, cache);if (old) {/* we lose the race */ free (cache); cache = old; } + ret =dso_cache__memcpy (cache, offset, data, size, out); +mutex_unlock (dso__lock (dso)); -return cache; -} - -static struct dso_cache *dso_cache__find (struct dso *dso, -struct machine *machine, -u64 offset, -ssize_t *ret) -{ -struct dso_cache *cache =__dso_cache__find (dso, offset); - -return cache ? cache :dso_cache__populate (dso, machine, offset, ret); +return ret; } +/* +* Look up and copy cached DSO data while holding dso__lock(dso) so a +* concurrent dso__shrink() / __dso_cache__free() cannot free the cache node +* while it is being read. +*/ static ssize_t dso_cache_io (struct dso *dso,struct machine *machine,u64 offset,u8 *data,ssize_t size,bool out) {struct dso_cache *cache; -ssize_t ret =0 ; +ssize_t ret; - cache =dso_cache__find (dso, machine, offset, &ret); -if (!cache) +mutex_lock (dso__lock (dso)); + cache =__dso_cache__find (dso, offset); +if (cache) { + ret =dso_cache__memcpy (cache, offset, data, size, out); +mutex_unlock (dso__lock (dso));return ret; + } +mutex_unlock (dso__lock (dso)); -return dso_cache__memcpy (cache, offset, data, size, out); +return dso_cache__populate (dso, machine, offset, data, size, out); }/*
@@ -1157,6 +1161,7 @@ static ssize_t cached_io(struct dso *dso, struct machine *machine,ssize_t r =0 ;u8 *p = data; +dso__set_accessed (dso, true); do { ssize_t ret;
@@ -1723,6 +1728,7 @@ struct dso *dso__new_id(const char *name, const struct dso_id *id) dso->is_kmod =0 ; dso->needs_swap =DSO_SWAP__UNSET ; dso->comp =COMP_ID__NONE ; +atomic_init (&dso->accessed, false);mutex_init (&dso->lock);refcount_set (&dso->refcnt,1 ); data = &dso->data;
@@ -1776,6 +1782,106 @@ void dso__delete(struct dso *dso)RC_CHK_FREE (dso); } +/** +* dso__can_shrink_symbols - Check if a DSO's symbol table can be shrunk and +* reloaded on demand. +* @dso: DSO to check. +* +* Shrinking evicts unreferenced symbols from @dso->symbols and relies on +* dso__load() being able to reload the missing symbols idempotently if a future +* lookup misses in map__find_symbol(). Not all DSO types support idempotent +* reloading: +* +* - User-space ELF binaries, debuginfo files, build-id cache files, and +* perf-<pid>.map JIT files (dso__kernel(dso) == DSO_SPACE__USER), as well as +* standalone kernel module ELF files (dso__is_kmod(dso)), load their symbols +* directly into @dso->symbols from a backing file on disk and can be safely +* reloaded on demand. +* +* - Main kernel and guest kernel DSOs (vmlinux / kallsyms where +* dso__kernel(dso) != DSO_SPACE__USER && !dso__is_kmod(dso)) split their +* loaded symbols across multiple per-section or per-module maps and DSOs in +* machine->kmaps (via maps__split_kallsyms()). Re-running dso__load() on one +* of the split kernel DSOs would either fail to find a backing file or +* re-split and duplicate kernel maps, so they are excluded. +* +* - KCORE / GUEST_KCORE DSOs replace and remap kernel maps in machine->kmaps +* when loaded from /proc/kcore, so they cannot be reloaded in place. +* +* - BPF_PROG_INFO, BPF_IMAGE, and OOL DSOs have their symbols synthesized in +* memory from PERF_RECORD_KSYMBOL / PERF_RECORD_BPF_EVENT meta-events rather +* than loaded from disk by dso__load(), so evicting their symbols would lose +* them permanently. +* +* Return: True if @dso->symbols can be safely shrunk and reloaded on demand. +*/ +bool dso__can_shrink_symbols (const struct dso *dso) +{ +switch (dso__symtab_type (dso)) { +case DSO_BINARY_TYPE__KALLSYMS : +case DSO_BINARY_TYPE__GUEST_KALLSYMS : +case DSO_BINARY_TYPE__VMLINUX : +case DSO_BINARY_TYPE__GUEST_VMLINUX : +case DSO_BINARY_TYPE__JAVA_JIT : +case DSO_BINARY_TYPE__DEBUGLINK : +case DSO_BINARY_TYPE__BUILD_ID_CACHE : +case DSO_BINARY_TYPE__BUILD_ID_CACHE_DEBUGINFO : +case DSO_BINARY_TYPE__FEDORA_DEBUGINFO : +case DSO_BINARY_TYPE__UBUNTU_DEBUGINFO : +case DSO_BINARY_TYPE__MIXEDUP_UBUNTU_DEBUGINFO : +case DSO_BINARY_TYPE__BUILDID_DEBUGINFO : +case DSO_BINARY_TYPE__SYSTEM_PATH_DSO : +case DSO_BINARY_TYPE__GUEST_KMODULE : +case DSO_BINARY_TYPE__GUEST_KMODULE_COMP : +case DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE : +case DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE_COMP : +case DSO_BINARY_TYPE__OPENEMBEDDED_DEBUGINFO : +case DSO_BINARY_TYPE__GNU_DEBUGDATA : +return dso__kernel (dso) ==DSO_SPACE__USER ||dso__is_kmod (dso); +case DSO_BINARY_TYPE__KCORE : +case DSO_BINARY_TYPE__GUEST_KCORE : +case DSO_BINARY_TYPE__BPF_PROG_INFO : +case DSO_BINARY_TYPE__BPF_IMAGE : +case DSO_BINARY_TYPE__OOL : +case DSO_BINARY_TYPE__NOT_FOUND : +default : +return false; + } +} + +/** +* dso__shrink - Reclaim cached file data, debuginfo, and unreferenced symbols +* from a DSO. +* @dso: DSO to shrink. +* +* Closes any cached file descriptor, frees the 4KB page cache (@dso->data.cache), +* releases cached addr2line / libbfd / libdw handles, and shrinks @dso->symbols +* if supported by dso__can_shrink_symbols(). +* +* Return: Number of symbols evicted from @dso->symbols. +*/ +size_t dso__shrink (struct dso *dso) +{ +size_t removed =0 ; + +if (!dso) +return 0 ; + +dso__data_close (dso); + +mutex_lock (dso__lock (dso)); +__dso_cache__free (dso); +dso__free_a2l (dso); +dso__free_a2l_libbfd (dso); +dso__free_libdw (dso); + +if (dso__can_shrink_symbols (dso)) + removed =symbols__shrink (dso__symbols (dso)); +mutex_unlock (dso__lock (dso)); + +return removed; +} +struct dso *dso__get (struct dso *dso) {struct dso *result;
tools/perf/util/dso.h +46 -0
92 lines not on this page · raw message
tools/perf/util/dsos.c +100 -0
124 lines not on this page · raw message
tools/perf/util/dsos.h +9 -3
@@ -13,15 +13,20 @@ struct dso_id;struct kmod_path ;struct machine ; -/* -* Collection of DSOs as an array for iteration speed, but sorted for O(n) -* lookup. +/** +* struct dsos - Collection of DSOs stored as a sorted array for iteration speed +* and O(log n) lookup. */ struct dsos { +/** @lock: Read/write semaphore synchronizing access to @dsos. */ struct rw_semaphore lock; +/** @dsos: Dynamically allocated array of struct dso pointers. */ struct dso **dsos; +/** @cnt: Number of valid entries in @dsos. */ unsigned int cnt; +/** @allocated: Capacity (in entries) of the @dsos allocation. */ unsigned int allocated; +/** @sorted: True when @dsos is sorted by long name, dso_id, and short name. */ bool sorted; };
@@ -48,5 +53,6 @@ struct dso *dsos__findnew_module_dso(struct dsos *dsos, struct machine *machine,struct dso *dsos__find_kernel_dso (struct dsos *dsos);int dsos__for_each_dso (struct dsos *dsos,int (*cb)(struct dso *dso,void *data),void *data); +size_t dsos__shrink (struct dsos *dsos);#endif /* __PERF_DSOS */
tools/perf/util/machine.c +39 -0
@@ -962,6 +962,45 @@ size_t machines__fprintf_dsos_buildid(struct machines *machines, FILE *fp,return ret; } +/** +* machine__shrink - Reclaim unreferenced symbols, DSO caches, and cold DSOs +* for a machine. +* @machine: Machine to shrink. +* +* Return: Number of DSOs evicted from @machine->dsos. +*/ +size_t machine__shrink (struct machine *machine) +{ +if (!machine) +return 0 ; + +return dsos__shrink (&machine->dsos); +} + +/** +* machines__shrink - Reclaim unreferenced symbols, DSO caches, and cold DSOs +* across the host and all guest machines. +* @machines: Collection of host and guest machines to shrink. +* +* Return: Total number of DSOs evicted across all machines. +*/ +size_t machines__shrink (struct machines *machines) +{ +struct rb_node *nd; +size_t ret; + +if (!machines) +return 0 ; + + ret =machine__shrink (&machines->host); +for (nd =rb_first_cached (&machines->guests); nd; nd =rb_next (nd)) { + struct machine *pos = rb_entry(nd, struct machine, rb_node); + + ret +=machine__shrink (pos); + } +return ret; +} +struct machine_fprintf_cb_args {FILE *fp;size_t printed;
tools/perf/util/machine.h +2 -0
@@ -268,6 +268,8 @@ size_t machine__fprintf_dsos_buildid(struct machine *machine, FILE *fp,size_t machines__fprintf_dsos (struct machines *machines,FILE *fp);size_t machines__fprintf_dsos_buildid (struct machines *machines,FILE *fp,bool (skip )(struct dso *dso,int parm),int parm); +size_t machine__shrink (struct machine *machine); +size_t machines__shrink (struct machines *machines);void machine__destroy_kernel_maps (struct machine *machine);int machine__create_kernel_maps (struct machine *machine);
tools/perf/util/map.c +24 -2
@@ -377,15 +377,30 @@ int map__load(struct map *map)struct symbol *map__find_symbol (struct map *map,u64 addr) { +struct dso *dso; +struct symbol *sym; +if (map__load (map) <0 )return NULL ; -return dso__find_symbol (map__dso (map), addr); + dso =map__dso (map); + sym =dso__find_symbol (dso, addr); +/* +* If the DSO's symbol table was previously shrunk by symbols__shrink() +* and the lookup missed, transparently reload the DSO so evicted +* symbols are restored on demand rather than reported as [unknown]. +*/ +if (!sym &&dso__symbols_shrunk (dso)) { +if (dso__load (dso, map) >=0 ) + sym =dso__find_symbol (dso, addr); + } +return sym; }struct symbol *map__find_symbol_by_name_idx (struct map *map,const char *name,size_t *idx) {struct dso *dso; +struct symbol *sym;if (map__load (map) <0 )return NULL ;
@@ -393,7 +408,14 @@ struct symbol *map__find_symbol_by_name_idx(struct map *map, const char *name, s dso =map__dso (map);dso__sort_by_name (dso); -return dso__find_symbol_by_name (dso, name, idx); + sym =dso__find_symbol_by_name (dso, name, idx); +if (!sym &&dso__symbols_shrunk (dso)) { +if (dso__load (dso, map) >=0 ) { +dso__sort_by_name (dso); + sym =dso__find_symbol_by_name (dso, name, idx); + } + } +return sym; }struct symbol *map__find_symbol_by_name (struct map *map,const char *name)
tools/perf/util/session.c +15 -0
@@ -3933,6 +3933,21 @@ size_t perf_session__fprintf_dsos_buildid(struct perf_session *session, FILE *fpreturn machines__fprintf_dsos_buildid (&session->machines, fp, skip, parm); } +/** +* perf_session__shrink - Reclaim unreferenced symbols, DSO caches, and cold +* DSOs across all machines in the session. +* @session: Session to shrink. +* +* Return: Number of DSOs evicted across the session's machines. +*/ +size_t perf_session__shrink (struct perf_session *session) +{ +if (!session) +return 0 ; + +return machines__shrink (&session->machines); +} +size_t perf_session__fprintf_nr_events (struct perf_session *session,FILE *fp) {size_t ret;
tools/perf/util/session.h +3 -0
@@ -91,7 +91,9 @@ struct perf_session {u64 bytes_compressed;/** @zstd_data: Owner of global compression state, buffers, etc. */ struct zstd_data zstd_data; +/** @decomp_data: Decompression state for the session data file. */ struct decomp_data decomp_data; +/** @active_decomp: Currently active decompression context. */ struct decomp_data *active_decomp; };
@@ -175,6 +177,7 @@ size_t perf_session__fprintf_dsos(struct perf_session *session, FILE *fp);size_t perf_session__fprintf_dsos_buildid (struct perf_session *session,FILE *fp,bool (fn )(struct dso *dso,int parm),int parm); +size_t perf_session__shrink (struct perf_session *session);size_t perf_session__fprintf_nr_events (struct perf_session *session,FILE *fp);
tools/perf/util/symbol-elf.c +3 -2
@@ -1826,9 +1826,10 @@ int dso__load_sym(struct dso *dso, struct map *map, struct symsrc *syms_ss,/* * Modules may already have symbols from kallsyms, but those symbols -* have the wrong values for the dso maps, so remove them. +* have the wrong values for the dso maps, so remove them unless we are +* reloading a previously shrunk module. */ -if (kmodule && syms_ss->symtab) +if (kmodule && syms_ss->symtab && !dso__symbols_shrunk (dso))symbols__delete (dso__symbols (dso));if (!syms_ss->symtab) {
tools/perf/util/symbol.c +136 -6
244 lines not on this page · raw message
tools/perf/util/symbol.h +30 -6
@@ -91,21 +91,32 @@ enum symbol_idle_kind {#define SYMBOL_FLAG_ANNOTATE2 (1 << 12)#define SYMBOL_FLAG_IFUNC_ALIAS (1 << 13)#define SYMBOL_FLAG_ANNOTATED (1 << 14) +#define SYMBOL_FLAG_ACCESSED (1U << 15) +/** -* A symtab entry. +* struct symbol - A symtab entry. */ DECLARE_RC_STRUCT (symbol) { -/** Range of symbol [start, end). */ +/** @start: Start address of symbol range [start, end). */ u64 start; +/** @end: End address (exclusive) of symbol range [start, end). */ u64 end; -/** Reference count. */ +/** @refcnt: Reference count tracking active holders of the symbol. */ refcount_t refcnt; -/** Length of the string name. */ +/** @namelen: Length of the string @name (excluding trailing NUL). */ u16 namelen; +/** +* @flags: Atomic bitfield of SYMBOL_FLAG_* values encoding symbol type, +* binding, idle classification, ignore/inlined/annotate/ifunc flags, +* and Clock LRU access state (SYMBOL_FLAG_ACCESSED). +*/ _Atomicuint16_t flags; -/** Architecture specific. Unused except on PPC where it holds st_other. */ +/** +* @arch_sym: Architecture specific value. Unused except on PPC where it +* holds st_other. +*/ u8 arch_sym; -/** The name of length namelen associated with the symbol. */ +/** @name: NUL-terminated symbol name of length @namelen. */ char name[]; };
@@ -138,6 +149,11 @@ struct symbols {* by name. */ bool sorted_by_name; +/** +* @shrunk: Set under @lock when symbols__shrink() evicts one or more +* symbols so a subsequent lookup miss knows to reload the DSO on demand. +*/ + _Atomicbool shrunk; };void symbol__delete (struct symbol *sym);
@@ -145,6 +161,7 @@ void symbols__init(struct symbols *symbols);void symbols__exit (struct symbols *symbols);void symbols__delete (struct symbols *symbols); void symbols__sort_read_lock(struct symbols *symbols) SHARED_LOCK_FUNCTION(symbols->lock); +size_t symbols__shrink (struct symbols *symbols);static inline refcount_t *symbol__refcnt (struct symbol *sym) {
@@ -253,6 +270,12 @@ static inline bool symbol__annotated(const struct symbol *sym) SYMBOL_FLAG_ANNOTATED) != 0; } +static inline bool symbol__accessed (const struct symbol *sym) +{ +return (atomic_load_explicit (&RC_CHK_ACCESS (sym)->flags, memory_order_relaxed) & +SYMBOL_FLAG_ACCESSED ) !=0 ; +} +bool symbol__is_idle (struct symbol *sym,const struct dso *dso,struct perf_env *env);void symbol__set_ignore (struct symbol *sym,bool ignore);
@@ -260,6 +283,7 @@ void symbol__set_annotate2(struct symbol *sym, bool annotate2);void symbol__set_inlined (struct symbol *sym,bool inlined);void symbol__set_ifunc_alias (struct symbol *sym,bool ifunc_alias);void symbol__set_annotated (struct symbol *sym,bool annotated); +void symbol__set_accessed (struct symbol *sym,bool accessed);/* * symbols__for_each_entry - iterate over symbols calling @cb for each entry
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 6/7 perf session: Periodically shrink symbols and DSOs during event processing
In reply to Ian RogersAdd periodic calls to perf_session__shrink() so that both live profiling and trace-processing commands reclaim unused symbol and DSO memory:
- In perf_session__deliver_event(), trigger perf_session__shrink() every 16,384 events, or after at least 512 events when at least 65,536 new symbols have been inserted across DSOs (tracked via symbols__inserted_cnt()). This bounds memory usage for session-based commands such as 'perf script' and 'perf report' even on short traces that load many large binaries, while giving hot symbols in newly loaded DSOs a window to be marked SYMBOL_FLAG_ACCESSED.
- In perf_top__resort_hists(), call perf_session__shrink() on each display refresh cycle so 'perf top' periodically ages out unused symbols and DSOs.
--- tools/perf/builtin-top.c | 2 ++ tools/perf/util/session.c | 16 ++++++++++++++++ tools/perf/util/session.h | 10 ++++++++++ tools/perf/util/symbol.c | 15 +++++++++++++++ tools/perf/util/symbol.h | 1 + 5 files changed, 44 insertions(+)
tools/perf/builtin-top.c +2 -0
@@ -320,6 +320,8 @@ static void perf_top__resort_hists(struct perf_top *t)evlist__for_each_entry (evlist, pos) {evsel__output_resort (pos,NULL ); } + +perf_session__shrink (t->session); }static void perf_top__print_sym_table (struct perf_top *top)
tools/perf/util/session.c +16 -0
@@ -164,6 +164,7 @@ struct perf_session *__perf_session__new(struct perf_data *data, session->tool = tool; session->decomp_data.zstd_decomp = &session->zstd_data; session->active_decomp = &session->decomp_data; + session->last_symbols_inserted =symbols__inserted_cnt ();INIT_LIST_HEAD (&session->auxtrace_index);perf_env__init (&session->header.env);if (machines__init (&session->machines, session))
@@ -2351,6 +2352,19 @@ static int perf_session__deliver_event(struct perf_session *session,if (dump_trace && sample.aux_sample.size)auxtrace__dump_auxtrace_sample (session, &sample); + +/* +* Periodically reclaim unreferenced symbols, DSO caches, and cold DSOs +* during event processing (e.g., perf script, perf report). Trigger +* either every 16,384 events, or after at least 512 events when more +* than 65,536 symbols have been newly loaded across DSOs. +*/ + session->events_since_shrink++; +if (session->events_since_shrink >=16384 || + (session->events_since_shrink >=512 && +symbols__inserted_cnt () - session->last_symbols_inserted >=65536 )) { +perf_session__shrink (session); + } out:perf_sample__exit (&sample);return ret;
@@ -3945,6 +3959,8 @@ size_t perf_session__shrink(struct perf_session *session)if (!session)return 0 ; + session->events_since_shrink =0 ; + session->last_symbols_inserted =symbols__inserted_cnt ();return machines__shrink (&session->machines); }
tools/perf/util/session.h +10 -0
@@ -95,6 +95,16 @@ struct perf_session {struct decomp_data decomp_data;/** @active_decomp: Currently active decompression context. */ struct decomp_data *active_decomp; +/** +* @events_since_shrink: Number of events delivered via +* perf_session__deliver_event() since the last perf_session__shrink(). +*/ +u64 events_since_shrink; +/** +* @last_symbols_inserted: Snapshot of symbols__inserted_cnt() taken at +* the last perf_session__shrink() pass. +*/ +unsigned long last_symbols_inserted; };struct decomp {
tools/perf/util/symbol.c +15 -0
@@ -623,6 +623,20 @@ static void symbol__set_idle(struct symbol *sym, bool idle) }while (!atomic_compare_exchange_weak (&RC_CHK_ACCESS (sym)->flags, &old_flags, new_flags)); } +static atomic_ulong symbols_inserted_cnt; + +/** +* symbols__inserted_cnt - Return cumulative count of symbols inserted. +* +* Monotonically increasing counter incremented each time a symbol is inserted +* into a struct symbols collection. Used by perf_session__deliver_event() to +* detect burst symbol loading across large DSOs and trigger periodic shrinking. +*/ +unsigned long symbols__inserted_cnt (void ) +{ +return atomic_load_explicit (&symbols_inserted_cnt, memory_order_relaxed); +} +int __symbols__insert (struct symbols *symbols,struct symbol *sym) {if (symbols->cnt == symbols->allocated) {
@@ -644,6 +658,7 @@ int __symbols__insert(struct symbols *symbols, struct symbol *sym) symbols->symbols[symbols->cnt++] = sym; symbols->sorted_by_name = false; +atomic_fetch_add_explicit (&symbols_inserted_cnt,1 , memory_order_relaxed);return 0 ; }
tools/perf/util/symbol.h +1 -0
@@ -162,6 +162,7 @@ void symbols__exit(struct symbols *symbols);void symbols__delete (struct symbols *symbols); void symbols__sort_read_lock(struct symbols *symbols) SHARED_LOCK_FUNCTION(symbols->lock);size_t symbols__shrink (struct symbols *symbols); +unsigned long symbols__inserted_cnt (void );static inline refcount_t *symbol__refcnt (struct symbol *sym) {
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerPATCH 7/7 perf test symbols: Add tests for symbol and DSO LRU shrinking
In reply to Ian RogersAdd unit tests in tools/perf/tests/symbols.c covering:
- symbols__shrink(): Second-Chance / Clock eviction of unaccessed symbols, clearing of SYMBOL_FLAG_ACCESSED on survival, and retention of symbols with external reference counts (refcnt > 1).
- dsos__shrink(): Second-Chance / Clock eviction of cold DSOs from struct dsos while retaining accessed DSOs and externally referenced DSOs.
- dso__shrink() and map__find_symbol(): Shrinking a loaded DSO's symbol table and verifying that a subsequent map__find_symbol() transparently reloads the symbol table on demand.
--- tools/perf/tests/symbols.c | 228 +++++++++++++++++++++++++++++++++++++ 1 file changed, 228 insertions(+)
tools/perf/tests/symbols.c +228 -0
@@ -3,6 +3,8 @@#include <linux/string.h> #include <sys/mman.h> #include <limits.h> +#include <stdlib.h> +#include <string.h> #include "debug.h" #include "dso.h" #include "env.h"
@@ -178,6 +180,221 @@ static int process_subdivided_dso(struct machine *machine, struct dso *dso)return ret <0 ?TEST_FAIL :TEST_OK ; } +static int test_symbols_shrink (void ) +{ +struct symbols symbols; +struct symbol *sym1, *sym2, *sym3, *held; +int ret =TEST_FAIL ; + +symbols__init (&symbols); + + sym1 =symbol__new (0x100 ,0x50 ,STB_GLOBAL ,STT_FUNC ,"sym1" ); + sym2 =symbol__new (0x200 ,0x50 ,STB_GLOBAL ,STT_FUNC ,"sym2" ); + sym3 =symbol__new (0x300 ,0x50 ,STB_GLOBAL ,STT_FUNC ,"sym3" ); +if (!sym1 || !sym2 || !sym3) { +symbol__put (sym1); +symbol__put (sym2); +symbol__put (sym3); + goto out; + } + +/* Hold an active reference on sym1, and mark sym2 accessed. */ + held =symbol__get (sym1); +symbol__set_accessed (sym2, true); + +symbols__insert (&symbols, sym1); +symbols__insert (&symbols, sym2); +symbols__insert (&symbols, sym3); + +/* +* First shrink: sym3 (unreferenced) is evicted; sym2 gets a second +* chance because SYMBOL_FLAG_ACCESSED was set; sym1 has refcnt > 1. +*/ +if (symbols__shrink (&symbols) !=1 || symbols.cnt !=2 ) { +pr_debug ("First symbols__shrink failed (cnt=%u)!\n" , symbols.cnt); +symbol__put (held); + goto out; + } + +/* +* Second shrink: sym2 was not accessed since first shrink, so it is now +* evicted; sym1 still has refcnt > 1. +*/ +if (symbols__shrink (&symbols) !=1 || symbols.cnt !=1 ) { +pr_debug ("Second symbols__shrink failed (cnt=%u)!\n" , symbols.cnt); +symbol__put (held); + goto out; + } + +symbol__set_accessed (held, false); +symbol__put (held); +if (symbols__shrink (&symbols) !=1 || symbols.cnt !=0 ) { +pr_debug ("Third symbols__shrink failed (cnt=%u)!\n" , symbols.cnt); + goto out; + } + + ret =TEST_OK ; +out: +symbols__exit (&symbols); +return ret; +} + +static int test_dsos_shrink (void ) +{ +struct dsos dsos; +struct dso *dso1, *dso2; +int ret =TEST_FAIL ; + +dsos__init (&dsos); + + dso1 =dsos__findnew_id (&dsos,"/tmp/perf_test_dso1" ,NULL ); + dso2 =dsos__findnew_id (&dsos,"/tmp/perf_test_dso2" ,NULL ); +if (!dso1 || !dso2) { +dso__put (dso1); +dso__put (dso2); + goto out; + } + +/* Keep the caller's active reference on dso1; drop caller reference on dso2. */ +dso__put (dso2); + +/* +* First shrink: both DSOs had accessed=true from creation, so neither +* is evicted on the first pass (Clock second chance), and accessed is +* cleared. +*/ +if (dsos__shrink (&dsos) !=0 || dsos.cnt !=2 ) { +pr_debug ("First dsos__shrink failed (cnt=%u)!\n" , dsos.cnt); +dso__put (dso1); + goto out; + } + +/* +* Second shrink: dso2 has refcnt == 1 and accessed == false, so it is +* evicted; dso1 has refcnt == 2, so it is kept. +*/ +if (dsos__shrink (&dsos) !=1 || dsos.cnt !=1 ) { +pr_debug ("Second dsos__shrink failed (cnt=%u)!\n" , dsos.cnt); +dso__put (dso1); + goto out; + } + +dso__put (dso1); +if (dsos__shrink (&dsos) !=1 || dsos.cnt !=0 ) { +pr_debug ("Third dsos__shrink failed (cnt=%u)!\n" , dsos.cnt); + goto out; + } + + ret =TEST_OK ; +out: +dsos__exit (&dsos); +return ret; +} + +struct shrink_reload_find_args { +struct symbol *kept; +u64 first_start; +u64 second_start; +char *second_name; +}; + +static int shrink_reload_find_cb (struct symbol *sym,void *data) +{ +struct shrink_reload_find_args *args = data; + +if (!args->kept) { + args->kept =symbol__get (sym); + args->first_start =symbol__start (sym); +return 0 ; + } +if (symbol__start (sym) != args->first_start) { + args->second_start =symbol__start (sym); + args->second_name =strdup (symbol__name (sym)); +return args->second_name ?1 : -ENOMEM ; + } +return 0 ; +} + +static int test_dso_shrink_and_reload (struct dso *dso,struct map *map) +{ +struct shrink_reload_find_args args = { + .kept =NULL , + .first_start =0 , + .second_start =0 , + .second_name =NULL , + }; +struct symbol *sym, *reloaded; +size_t removed; + +if (!dso__can_shrink_symbols (dso) ||dso__kernel (dso)) +return TEST_OK ; + +if (dso__for_each_symbol (dso, shrink_reload_find_cb, &args) <0 ) { +symbol__put (args.kept); +return TEST_FAIL ; + } +if (!args.second_name) { +symbol__put (args.kept); +return TEST_OK ; + } + +/* Unreferenced symbols have SYMBOL_FLAG_ACCESSED == false and are evicted immediately. */ + removed =dso__shrink (dso); +if (removed ==0 || !dso__symbols_shrunk (dso)) { +pr_debug ("dso__shrink did not evict any symbols!\n" ); +free (args.second_name); +symbol__put (args.kept); +return TEST_FAIL ; + } + +/* Kept symbol is still found without reloading. */ + sym =dso__find_symbol (dso, args.first_start); +if (!RC_CHK_EQUAL (sym, args.kept)) { +pr_debug ("Kept symbol not found after dso__shrink!\n" ); +symbol__put (sym); +free (args.second_name); +symbol__put (args.kept); +return TEST_FAIL ; + } +symbol__put (sym); + +/* Evicted symbol is no longer in dso->symbols before reload. */ + sym =dso__find_symbol (dso, args.second_start); +if (sym !=NULL ) { +pr_debug ("Evicted symbol still present after dso__shrink!\n" ); +symbol__put (sym); +free (args.second_name); +symbol__put (args.kept); +return TEST_FAIL ; + } + +/* Lookup via map__find_symbol transparently reloads the DSO. */ + reloaded =map__find_symbol (map, args.second_start); +if (!reloaded ||strcmp (symbol__name (reloaded), args.second_name) !=0 ) { +pr_debug ("Failed to reload evicted symbol %s!\n" , args.second_name); +symbol__put (reloaded); +free (args.second_name); +symbol__put (args.kept); +return TEST_FAIL ; + } +symbol__put (reloaded); + +/* Kept symbol pointer was preserved across reload deduplication. */ + sym =dso__find_symbol (dso, args.first_start); +if (!RC_CHK_EQUAL (sym, args.kept)) { +pr_debug ("Kept symbol pointer was not preserved across reload!\n" ); +symbol__put (sym); +free (args.second_name); +symbol__put (args.kept); +return TEST_FAIL ; + } +symbol__put (sym); + +free (args.second_name); +symbol__put (args.kept); +return TEST_OK ; +} +static int test_file (struct test_info *ti,char *filename) {struct map *map =NULL ;
@@ -209,6 +426,9 @@ static int test_file(struct test_info *ti, char *filename)/* Module dso is split into many dsos by section */ if (ret ==TEST_OK &&dso__kernel (dso) !=DSO_SPACE__USER ) ret =process_subdivided_dso (ti->machine, dso); + +if (ret ==TEST_OK ) + ret =test_dso_shrink_and_reload (dso, map); out_put:map__put (map);
@@ -221,6 +441,14 @@ static int test__symbols(struct test_suite *test __maybe_unused, int subtest __mstruct test_info ti;int ret; + ret =test_symbols_shrink (); +if (ret !=TEST_OK ) +return ret; + + ret =test_dsos_shrink (); +if (ret !=TEST_OK ) +return ret; + ret =init_test_info (&ti);if (ret !=TEST_OK )return ret;
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
Ian Rogers
reviewerQuoted 31 lines · Ian Rogers
This series modernizes struct symbol memory management and backing storage in
tools/perfto support reference counting (DECLARE_RC_STRUCT(symbol)) and periodic LRU memory shrinking during long or large profiling sessions:1. Field accessors (Patch 1): Introduces inline getters/setters for struct symbol fields across
tools/perfin preparation for wrapping struct symbol withDECLARE_RC_STRUCT(symbol).
2. Remove negative-offset private allocations (Patch 2): Replaces symbol_conf.priv_size andsymbol__priv()negative-offset allocations inutil/annotate.candutil/db-export.cwith hashmaps keyed by symbol pointer (structannotated_symbolprotected by anrw_semaphore, andsym_db_idson structdb_export).
3. Flat sorted array backing storage (Patch 3): Replaces the dual address-sorted and name-sorted red-black trees (rb_node) in struct dso with a struct symbols container backed by dynamically sorted flat arrays protected by structrw_semaphore(util/rwsem.h) with Clang thread-safety annotations. Saves 48 bytes per struct symbol (tworb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.
4. Reference counting andDECLARE_RC_STRUCT(symbol)(Patch 4): Wraps struct symbol withDECLARE_RC_STRUCT(symbol)and addsrefcount_trefcnt along withsymbol__get(),symbol__put(), andsymbol__zput(). Lookup functions return a counted reference acquired undersymbols->lock, and lifetime holders (addr_location,map_symbol,hist_entry,call_path,block_info,disasm_line,thread_stack,db_export, etc.) properly manage symbol references.
5. LRU memory shrinking (Patch 5): Adds an accessed bit to struct dso and implementssymbols__shrink(),dso__shrink_symbols(),dsos__shrink(),machine__shrink(), andmachines__shrink(). Unreferenced symbols (refcnt == 1) in cold DSOs are reclaimed, and shrunken DSOs transparently reload on demand viadso__load()while preserving live symbols and theirRC_STRUCT(symbol)
Quoted 8 lines
identity.
6. Periodic session shrinking (Patch 6): Hooksmachines__shrink()intoperf_sessionevent processing so long-running sessions periodically reclaim cold symbols and DSOs.
7. Unit tests (Patch 7): Adds unit tests intools/perf/tests/symbols.ccovering symbol reference counting,symbols__shrink()retention of active symbols,dsos__shrink()two-pass LRU clock behavior, and on-demand reload preserving live symbol identity.
This needs a rebase to after the changes to the python integration, hence no Sashiko reviews. I'll do that asap.
Thanks,
Ian
Quoted 106 lines
Ian Rogers (7):
perf symbol: Add accessor functions for struct symbol fields perf symbol: Remove symbol_conf.priv_size and negative-offset
allocations
perf symbol: Switch backing storage from rbtree to struct symbols
array
perf symbol: Add reference counting andDECLARE_RC_STRUCT(symbol)perf symbol: Add LRU memory shrinking for symbols, DSOs, and machines perf session: Periodically shrink symbols and DSOs during event
processing
perf test symbols: Add tests for symbol and DSO LRU shrinking.../arch/powerpc/util/skip-callchain-idx.c | 2 +- tools/perf/arch/powerpc/util/sym-handling.c | 40 +- tools/perf/builtin-annotate.c | 15 +- tools/perf/builtin-c2c.c | 14 - tools/perf/builtin-diff.c | 5 +- tools/perf/builtin-kallsyms.c | 10 +- tools/perf/builtin-kmem.c | 66 +- tools/perf/builtin-kwork.c | 4 +- tools/perf/builtin-lock.c | 27 +- tools/perf/builtin-mem.c | 2 +- tools/perf/builtin-report.c | 14 - tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 18 +- tools/perf/builtin-timechart.c | 2 +- tools/perf/builtin-top.c | 29 +- tools/perf/builtin-trace.c | 12 +- tools/perf/tests/builtin-test.c | 1 - tools/perf/tests/dlfilter-test.c | 12 +- tools/perf/tests/dwarf-unwind.c | 2 +- tools/perf/tests/hists_common.c | 4 +- tools/perf/tests/hists_cumulate.c | 8 +- tools/perf/tests/hists_filter.c | 10 +- tools/perf/tests/hists_link.c | 16 +- tools/perf/tests/hists_output.c | 9 +- tools/perf/tests/kallsyms-split.c | 5 +- tools/perf/tests/pe-file-parsing.c | 1 + tools/perf/tests/symbols.c | 289 +++++- tools/perf/tests/vmlinux-kallsyms.c | 191 ++-- tools/perf/ui/browsers/annotate.c | 11 +- tools/perf/ui/browsers/hists.c | 17 +- tools/perf/ui/browsers/map.c | 87 +- tools/perf/ui/gtk/annotate.c | 7 +- tools/perf/ui/stdio/hist.c | 6 +- tools/perf/util/addr_location.c | 4 + .../perf/util/annotate-arch/annotate-alpha.c | 2 +- .../util/annotate-arch/annotate-loongarch.c | 8 +- tools/perf/util/annotate-arch/annotate-s390.c | 2 +- tools/perf/util/annotate-arch/annotate-x86.c | 14 +- tools/perf/util/annotate-data.c | 14 +- tools/perf/util/annotate.c | 211 ++++- tools/perf/util/annotate.h | 6 +- tools/perf/util/auxtrace.c | 111 ++- tools/perf/util/block-info.c | 13 +- tools/perf/util/bpf_lock_contention.c | 39 +- tools/perf/util/c2c-function.c | 25 +- tools/perf/util/call-path.c | 15 +- tools/perf/util/callchain.c | 19 +- tools/perf/util/capstone.c | 21 +- tools/perf/util/data-convert-json.c | 4 +- tools/perf/util/db-export.c | 76 +- tools/perf/util/db-export.h | 6 + tools/perf/util/debug.c | 3 +- tools/perf/util/disasm.c | 57 +- tools/perf/util/dlfilter.c | 12 +- tools/perf/util/dso.c | 236 +++-- tools/perf/util/dso.h | 85 +- tools/perf/util/dsos.c | 100 +++ tools/perf/util/dsos.h | 12 +- tools/perf/util/event.c | 12 +- tools/perf/util/evsel_fprintf.c | 2 +- tools/perf/util/hist.c | 30 +- tools/perf/util/intel-pt.c | 74 +- tools/perf/util/libbfd.c | 5 +- tools/perf/util/libdw.c | 5 +- tools/perf/util/llvm.c | 6 +- tools/perf/util/lock-contention.c | 8 +- tools/perf/util/machine.c | 143 ++- tools/perf/util/machine.h | 2 + tools/perf/util/map.c | 53 +- tools/perf/util/map.h | 18 +- tools/perf/util/map_symbol.c | 4 +- tools/perf/util/maps.c | 3 +- tools/perf/util/probe-event.c | 135 +-- tools/perf/util/python.c | 20 +- .../util/scripting-engines/trace-event-perl.c | 8 +- .../scripting-engines/trace-event-python.c | 38 +- tools/perf/util/session.c | 31 + tools/perf/util/session.h | 13 + tools/perf/util/sort.c | 44 +- tools/perf/util/srcline.c | 19 +- tools/perf/util/symbol-elf.c | 22 +- tools/perf/util/symbol.c | 845 ++++++++++++------ tools/perf/util/symbol.h | 203 ++++- tools/perf/util/symbol_conf.h | 1 - tools/perf/util/symbol_fprintf.c | 14 +- tools/perf/util/thread-stack.c | 18 +- tools/perf/util/threads.c | 17 +- tools/perf/util/unwind-libdw.c | 4 +- tools/perf/util/unwind-libunwind.c | 2 +- 89 files changed, 2644 insertions(+), 1192 deletions(-)--
2.56.0.rc1.315.gc6ed9934b7-goog
Quoted 2 lines · Ian Rogers
annotations. Saves 48 bytes per struct symbol (two
rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.
Hi Ian,
Thanks for posting this and for Cc'ing me. I built it on edd8a9fe2eca0, the same base as the v3 series I sent last week, and compared it with v3 in eager mode (today's behavior) and with --lazy-load-symbols.
All runs use perf script --no-inline --max-stack 127
-F comm,tid,time,ip,sym,dso, with one warm-up and three runs; the median is shown below. Peak anon is the maximum RssAnon from /proc/PID/status, sampled every 10 ms.
1) Production: 120 s cgroup profile of a storage service on an 80-CPU host (perf record -a -g -F 99 -G <cgroup>). 54k samples, 11 DSOs, 5% of lines [unknown].
time max RSS peak anon
v3 eager 4.5 s 386 MiB 314 MiB v3 lazy 3.7 s 151 MiB 80 MiB this series 95.9 s 309 MiB 275 MiB
2) A sample fixture from my v3 cover letter, a worst case with 73% of lines [unknown]:
time max RSS
v3 eager 3.2 s 334 MiB v3 lazy 1.9 s 106 MiB this series 205.3 s 449 MiB
Most of the time goes to reloading. On the fixture, a profile shows it under map__find_symbol() -> dso__load() -> dso__load_sym(). After each shrink, the first miss in a shrunk DSO re-parses its whole symtab, and addresses that resolve to [unknown] miss every time.
The peak is still set by dso__load(), since the whole symtab is materialized before a shrink can run. That is the main blocker for us running perf script at scale: we can't bound how much memory it will use, so we have to contain it with memory.max, and then the profiling job fails with an OOM kill instead of degrading.
For continuous profiling, a degraded profile is still useful. We build our view statistically, aggregating many profiles across hosts and time, so some frames reported as [unknown] in the DSOs that hit the cap only add a little noise, and the cap reports when it was hit. A lost profile is worse: the OOM kill removes all its samples and wastes the CPU and memory already spent. A bounded, reported degradation, like perf record --max-size, is the trade-off I'd recommend.
The bound also matters before a profile starts. perf runs next to production workloads, so we have to plan its memory up front: pick a host with enough headroom and size the job's reservation, without pushing those sensitive workloads into memory reclaim. Without a bound we can't do that.
The output of your series matches current eager loading except for a few frames, and there I think your series is right. For example, a sample inside nft_do_chain [nf_tables] is printed as dca_sysfs_exit [dca]. symbols__fixup_end() extends the last symbol of a module to roundup(end + 4096, 4096), which overlaps the next module when it starts
in the next page. The rb-tree lookup can then return the stretched symbol,
while your bsearch picks the right one. This comes from bacefe0c7b77b; I can send a separate fix that clamps the end to the next symbol's start.
A few questions:
- Which workloads did you have in mind? I'd expect the shrinking to help most with perf top and long perf report sessions, which I haven't measured. I'm happy to run those on the same hosts if you have a setup you'd like me to use.
- Would you consider a finer-grained reload? The index in my patch 5/6 keeps a 24-byte entry per symbol, so a miss after a shrink could materialize one symbol instead of re-parsing the symtab. The pieces could fit together: your container and refcounting, the index for reload, and a byte cap for the peak. If the maintainers prefer your container first, I'm happy to rebase my v4 onto struct symbols.
- The cover letter says patch 3 saves 48 bytes per symbol (two rb_nodes). At edd8a9fe2eca0, struct symbol has one rb_node, and the name index is already a struct symbol ** array on the DSO. Am I missing a second node?
Thanks,
Alireza
Ian Rogers
reviewerQuoted 38 lines · Alireza Haghdoost
Quoted 2 lines · Ian Rogers
annotations. Saves 48 bytes per struct symbol (two
rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.Hi Ian,
Thanks for posting this and for Cc'ing me. I built it on
edd8a9fe2eca0, the same base as the v3 series I sent last week, and compared it with v3 in eager mode (today's behavior) and with --lazy-load-symbols.All runs use perf script --no-inline --max-stack 127
-F comm,tid,time,ip,sym,dso, with one warm-up and three runs; the median is shown below. Peak anon is the maximum RssAnon from/proc/PID/status, sampled every 10 ms.1) Production: 120 s cgroup profile of a storage service on an 80-CPU host (perf record -a -g -F 99 -G <cgroup>). 54k samples, 11 DSOs, 5% of lines [unknown].
time max RSS peak anon
v3 eager 4.5 s 386 MiB 314 MiB v3 lazy 3.7 s 151 MiB 80 MiB this series 95.9 s 309 MiB 275 MiB2) A sample fixture from my v3 cover letter, a worst case with 73% of lines [unknown]:
time max RSS
v3 eager 3.2 s 334 MiB v3 lazy 1.9 s 106 MiB this series 205.3 s 449 MiBMost of the time goes to reloading. On the fixture, a profile shows it under
map__find_symbol()->dso__load()->dso__load_sym(). After each shrink, the first miss in a shrunk DSO re-parses its whole symtab, and addresses that resolve to [unknown] miss every time.
Quoted 5 lines
The peak is still set by
dso__load(), since the whole symtab is materialized before a shrink can run. That is the main blocker for us running perf script at scale: we can't bound how much memory it will use, so we have to contain it with memory.max, and then the profiling job fails with an OOM kill instead of degrading.
Hi Alireza, and thanks for the feedback! I'll see what can be done about dso__load in v2.
Quoted 7 lines
For continuous profiling, a degraded profile is still useful. We build our view statistically, aggregating many profiles across hosts and time, so some frames reported as [unknown] in the DSOs that hit the cap only add a little noise, and the cap reports when it was hit. A lost profile is worse: the OOM kill removes all its samples and wastes the CPU and memory already spent. A bounded, reported degradation, like perf record --max-size, is the trade-off I'd recommend.
I don't disagree. Fwiw, a C++ memory allocation failure inside my company's code base is a runtime abort. In perf, many code paths return -ENOMEM or NULL. I wanted to see how far we could get with a "shrunk" set of data inside perf as even if we fix memory allocation around symbols, other code paths can still fail.
Quoted 5 lines
The bound also matters before a profile starts. perf runs next to production workloads, so we have to plan its memory up front: pick a host with enough headroom and size the job's reservation, without pushing those sensitive workloads into memory reclaim. Without a bound we can't do that.
I'm still not massively sold on having explicit bounds. It reminds me of old and often seemingly arbitrary numbers like the number of rows or columns in a spreadsheet. That said, perf has a similar problem with exhausting file descriptors. As long as the default is "unlimited," I'm okay with some attempts to limit size, but we should implement obvious garbage collection as I hope this series introduces.
Quoted 8 lines
The output of your series matches current eager loading except for a few frames, and there I think your series is right. For example, a sample inside
nft_do_chain[nf_tables] is printed asdca_sysfs_exit[dca].symbols__fixup_end()extends the last symbol of a module toroundup(end + 4096, 4096), which overlaps the next module when it starts
in the next page. The rb-tree lookup can then return the stretched symbol,
while your bsearch picks the right one. This comes frombacefe0c7b77b; I can send a separate fix that clamps the end to the next symbol's start.
That's surprising as no behavioral change was intended. That then leads to the question of which symbol is right and whether a latent bug is being fixed or a new one introduced. Oh what fun :-)
Quoted 6 lines
A few questions:
- Which workloads did you have in mind? I'd expect the shrinking to help most with perf top and long perf report sessions, which I haven't measured. I'm happy to run those on the same hosts if you have a setup you'd like me to use.
So most of my experience with excessive memory usage has come from perf top, which seems to leak memory and consume all your RAM before long. That is expected since no DSO is currently unloaded. I'm hoping the shrinking process can be generic.
Quoted 6 lines
- Would you consider a finer-grained reload? The index in my patch 5/6 keeps a 24-byte entry per symbol, so a miss after a shrink could materialize one symbol instead of re-parsing the symtab. The pieces could fit together: your container and refcounting, the index for reload, and a byte cap for the peak. If the maintainers prefer your container first, I'm happy to rebase my v4 onto struct symbols.
Definitely. I was more focused on the reference counting aspect. Things like the policy around shrinking can be amended as needed, I just wanted to establish a policy so the patches made sense. I think lazier loading, shrinking, .. all make sense for things to be faster and use less RAM.
Quoted 4 lines
- The cover letter says patch 3 saves 48 bytes per symbol (two
rb_nodes). Atedd8a9fe2eca0, struct symbol has onerb_node, and the name index is already a struct symbol ** array on the DSO. Am I missing a second node?
No, I'll fix this mistake.
Thanks,
Ian
Quoted 2 lines
Thanks,
Alireza
Namhyung Kim
maintainerHello,
Quoted 39 lines · Ian Rogers
Quoted 38 lines · Alireza Haghdoost
Quoted 2 lines · Ian Rogers
annotations. Saves 48 bytes per struct symbol (two
rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.Hi Ian,
Thanks for posting this and for Cc'ing me. I built it on
edd8a9fe2eca0, the same base as the v3 series I sent last week, and compared it with v3 in eager mode (today's behavior) and with --lazy-load-symbols.All runs use perf script --no-inline --max-stack 127
-F comm,tid,time,ip,sym,dso, with one warm-up and three runs; the median is shown below. Peak anon is the maximum RssAnon from/proc/PID/status, sampled every 10 ms.1) Production: 120 s cgroup profile of a storage service on an 80-CPU host (perf record -a -g -F 99 -G <cgroup>). 54k samples, 11 DSOs, 5% of lines [unknown].
time max RSS peak anon
v3 eager 4.5 s 386 MiB 314 MiB v3 lazy 3.7 s 151 MiB 80 MiB this series 95.9 s 309 MiB 275 MiB2) A sample fixture from my v3 cover letter, a worst case with 73% of lines [unknown]:
time max RSS
v3 eager 3.2 s 334 MiB v3 lazy 1.9 s 106 MiB this series 205.3 s 449 MiBMost of the time goes to reloading. On the fixture, a profile shows it under
map__find_symbol()->dso__load()->dso__load_sym(). After each shrink, the first miss in a shrunk DSO re-parses its whole symtab, and addresses that resolve to [unknown] miss every time.
Quoted 8 lines
Quoted 5 lines
The peak is still set by
dso__load(), since the whole symtab is materialized before a shrink can run. That is the main blocker for us running perf script at scale: we can't bound how much memory it will use, so we have to contain it with memory.max, and then the profiling job fails with an OOM kill instead of degrading.Hi Alireza, and thanks for the feedback! I'll see what can be done about
dso__loadin v2.
Thanks for the patches and the analysis.
I think many of this memory overhead come from the symbol string. Probably we don't use most of them. Then would it be nice if we can lazy-load the strings? Then it'd have an union of a pointer and a file offset for symbol strings. The find-by-name API is used rarely for user DSOs and the kernel symbols should be fully loaded anyway.
Without the string (and the priv part), now it has a fixed size so it should be saved in an array directly. Probably we can add a limit there and print it with offset when it doesn't have the name.
I'm a bit skeptical about the refcount approach here. It may be hard to determine when it reclaims memory. I guess the lazy-load strings with an array would give similar savings like Alireza's work.
Thanks,
Namhyung
Ian Rogers
reviewerQuoted 60 lines · Namhyung Kim
Hello,
Quoted 39 lines · Ian Rogers
Quoted 38 lines · Alireza Haghdoost
Quoted 2 lines · Ian Rogers
annotations. Saves 48 bytes per struct symbol (two
rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.Hi Ian,
Thanks for posting this and for Cc'ing me. I built it on
edd8a9fe2eca0, the same base as the v3 series I sent last week, and compared it with v3 in eager mode (today's behavior) and with --lazy-load-symbols.All runs use perf script --no-inline --max-stack 127
-F comm,tid,time,ip,sym,dso, with one warm-up and three runs; the median is shown below. Peak anon is the maximum RssAnon from/proc/PID/status, sampled every 10 ms.1) Production: 120 s cgroup profile of a storage service on an 80-CPU host (perf record -a -g -F 99 -G <cgroup>). 54k samples, 11 DSOs, 5% of lines [unknown].
time max RSS peak anon
v3 eager 4.5 s 386 MiB 314 MiB v3 lazy 3.7 s 151 MiB 80 MiB this series 95.9 s 309 MiB 275 MiB2) A sample fixture from my v3 cover letter, a worst case with 73% of lines [unknown]:
time max RSS
v3 eager 3.2 s 334 MiB v3 lazy 1.9 s 106 MiB this series 205.3 s 449 MiBMost of the time goes to reloading. On the fixture, a profile shows it under
map__find_symbol()->dso__load()->dso__load_sym(). After each shrink, the first miss in a shrunk DSO re-parses its whole symtab, and addresses that resolve to [unknown] miss every time.Quoted 8 lines
Quoted 5 lines
The peak is still set by
dso__load(), since the whole symtab is materialized before a shrink can run. That is the main blocker for us running perf script at scale: we can't bound how much memory it will use, so we have to contain it with memory.max, and then the profiling job fails with an OOM kill instead of degrading.Hi Alireza, and thanks for the feedback! I'll see what can be done about
dso__loadin v2.Thanks for the patches and the analysis.
I think many of this memory overhead come from the symbol string. Probably we don't use most of them. Then would it be nice if we can lazy-load the strings? Then it'd have an union of a pointer and a file offset for symbol strings. The find-by-name API is used rarely for user DSOs and the kernel symbols should be fully loaded anyway.
Quoted 7 lines
Without the string (and the priv part), now it has a fixed size so it should be saved in an array directly. Probably we can add a limit there and print it with offset when it doesn't have the name.
I'm a bit skeptical about the refcount approach here. It may be hard to determine when it reclaims memory. I guess the lazy-load strings with an array would give similar savings like Alireza's work.
Thanks Namhyung. Without reference counting we lack a good way to discard symbols; we must either assume they always exist as in the current code or implement reference counting. I don't see a way around it. Alireza's patches introduced an "unknown" state for when a symbol limit was hit, but I think that'd be frustrating in practice. We could use it after trying to shrink memory use, in my opinion.
Fwiw, I never like switching the rbtree to an array. The issue is that the rbtree has references that really should be managed by a reference count. Doing that is a challenge in the current code and a sorted array offers similar performance while simplifying the reference counting.
We load all symbols instead of lazily because we need the end of a symbol, which we can only determine by processing the entire symbol table as the end of one symbol is the start of the next. We generally translate one address in a DSO into a symbol, so loading all symbols is overkill. I think we can do better by loading only symbols within a certain address window rather than loading all symbols in a DSO. We can then expand this window as more symbols are needed. This at least bounds the number of symbols but isn't quite lazy loading.
Alireza also made good points about how symbols not being found or being shrunk leads to thrashing patterns. I think we can fix this by maintaining extra state in the DSO. I'm working to add this into v2.
Thanks,
Ian
Quoted 3 lines
Thanks,
Namhyung
Quoted 3 lines · Ian Rogers
Quoted 3 lines
I think many of this memory overhead come from the symbol string. Probably we don't use most of them. Then would it be nice if we can lazy-load the strings?
Namhyung, your guess is right. I counted live symbol allocations in eager mode on the sample fixture. At the peak, 986,688 symbols take 258 MiB:
names (demangled, avg 210 bytes) 197 MiB 77%
struct symbol (48 bytes each, 45 MiB 18%
24 of them the rb_node)
malloc overhead 15 MiB 6%
That is most of the 335 MiB max RSS. With --lazy-load-symbols, only 2,928 symbols are created (0.5 MiB), with identical output, and max RSS is 108 MiB.
Quoted 2 lines
Quoted 2 lines
Without the string (and the priv part), now it has a fixed size so it should be saved in an array directly.
That is close to what patch 5/6 of my v3 does. Its index is a sorted array of fixed-size entries:
struct sym_idx {
u64 start;
u64 end;
u32 name_off;
u8 binding;
u8 type;
u8 flags;
};That is 24 bytes per symbol. The name is read from the string table and demangled only when a sample lands in the symbol. The difference from your description is that v3 then creates a regular struct symbol for the hit and inserts it in the rb-tree, because the rest of perf holds struct symbol pointers. If the entry itself were the symbol, with the name as a pointer/offset union, that extra step would go away. Ian's patch 1 would help: once every name access goes through symbol__name(), that accessor is the only place that has to resolve an offset.
Quoted 2 lines
Quoted 2 lines
Probably we can add a limit there
and print it with offset when it doesn't have the name.
Agreed, that is better than [unknown]. The symbol boundaries are still known, so the frame can be printed as an offset and resolved offline. I can do the same for the byte cap in v4.
Quoted 5 lines
We load all symbols instead of lazily because we need the end of a symbol, which we can only determine by processing the entire symbol table as the end of one symbol is the start of the next.
Ian, I think most symbols don't need the next one for their end. ELF symbols have st_size, so the end is start + st_size. Only zero-size symbols take the next symbol's start. The whole table still has to be scanned to find the symbol for an address, because .symtab isn't sorted by address, so an address window would scan it again each time it grows. The v3 index scans it once, sorts the entries and fixes up the ends there, but keeps only the 24-byte entries, not the names.
On which kernel symbol is right in the nf_tables example: the sample is at 0xffffffffc0c4b280, and /proc/kallsyms has nft_do_chain at 0xffffffffc0c4b0b0 and the next nf_tables symbol at 0xffffffffc0c4b4e0, so your series is right. The last dca symbol starts at 0xffffffffc0c4aa50, and symbols__fixup_end() extends it to 0xffffffffc0c4c000, over the first page of nf_tables. I've sent the fix separately:
https://lore.kernel.org/all/20260928-haghdoost-perf-symbols-fixup-module-end-v1-1-0a70d1edd401@uber.com/
Thanks
Alireza
Namhyung Kim
maintainerQuoted 8 lines · Alireza Haghdoost
Quoted 3 lines · Ian Rogers
Quoted 3 lines
I think many of this memory overhead come from the symbol string. Probably we don't use most of them. Then would it be nice if we can lazy-load the strings?
Namhyung, your guess is right. I counted live symbol allocations in eager mode on the sample fixture. At the peak, 986,688 symbols take 258 MiB:
Quoted 8 lines
names (demangled, avg 210 bytes) 197 MiB 77%
struct symbol (48 bytes each, 45 MiB 18%
24 of them therb_node)
malloc overhead 15 MiB 6%That is most of the 335 MiB max RSS. With --lazy-load-symbols, only 2,928 symbols are created (0.5 MiB), with identical output, and max RSS is 108 MiB.
Thanks for checking this!
Quoted 16 lines
Quoted 3 lines
Quoted 3 lines
Without the string (and the priv part), now it has a fixed size so it should be saved in an array directly.
That is close to what patch 5/6 of my v3 does. Its index is a sorted array of fixed-size entries:
struct sym_idx { u64 start; u64 end; u32 name_off; u8 binding; u8 type; u8 flags; };
Quoted 9 lines
That is 24 bytes per symbol. The name is read from the string table and demangled only when a sample lands in the symbol. The difference from your description is that v3 then creates a regular struct symbol for the hit and inserts it in the rb-tree, because the rest of perf holds struct symbol pointers. If the entry itself were the symbol, with the name as a pointer/offset union, that extra step would go away. Ian's patch 1 would help: once every name access goes through
symbol__name(), that accessor is the only place that has to resolve an offset.
I'm curious if name_off would work well for PLT symbols which come from the dynamic symbol table. Probably you need to handle them differently.
Quoted 22 lines
Quoted 3 lines
Quoted 3 lines
Probably we can add a limit there
and print it with offset when it doesn't have the name.Agreed, that is better than [unknown]. The symbol boundaries are still known, so the frame can be printed as an offset and resolved offline. I can do the same for the byte cap in v4.
Quoted 5 lines
We load all symbols instead of lazily because we need the end of a symbol, which we can only determine by processing the entire symbol table as the end of one symbol is the start of the next.
Ian, I think most symbols don't need the next one for their end. ELF symbols have
st_size, so the end is start +st_size. Only zero-size symbols take the next symbol's start. The whole table still has to be scanned to find the symbol for an address, because .symtab isn't sorted by address, so an address window would scan it again each time it grows. The v3 index scans it once, sorts the entries and fixes up the ends there, but keeps only the 24-byte entries, not the names.
Quoted 8 lines
On which kernel symbol is right in the
nf_tablesexample: the sample is at0xffffffffc0c4b280, and/proc/kallsymshasnft_do_chainat0xffffffffc0c4b0b0and the nextnf_tablessymbol at0xffffffffc0c4b4e0, so your series is right. The last dca symbol starts at0xffffffffc0c4aa50, andsymbols__fixup_end()extends it to0xffffffffc0c4c000, over the first page ofnf_tables. I've sent the fix separately:
https://lore.kernel.org/all/20260928-haghdoost-perf-symbols-fixup-module-end-v1-1-0a70d1edd401@uber.com/
Thanks for the fix!
Namhyung
Quoted 22 lines
Quoted 12 lines
That is close to what patch 5/6 of my v3 does. Its index is a sorted array of fixed-size entries:
struct sym_idx { u64 start; u64 end; u32 name_off; u8 binding; u8 type; u8 flags; };Quoted 9 lines
That is 24 bytes per symbol. The name is read from the string table and demangled only when a sample lands in the symbol. The difference from your description is that v3 then creates a regular struct symbol for the hit and inserts it in the rb-tree, because the rest of perf holds struct symbol pointers. If the entry itself were the symbol, with the name as a pointer/offset union, that extra step would go away. Ian's patch 1 would help: once every name access goes through
symbol__name(), that accessor is the only place that has to resolve an offset.
Quoted 3 lines
I'm curious if
name_offwould work well for PLT symbols which come from the dynamic symbol table. Probably you need to handle them differently.
Hi Namhyung,
Yes, synthesized PLT symbols need separate handling, but current upstream already has that path and my v3 series keeps it.
For a regular ELF symbol, name_off is its st_name: a byte offset into .strtab for .symtab, or .dynstr for .dynsym. So the same index format works for both tables. v3 uses .dynsym when a DSO has no .symtab.
An @plt symbol is synthesized rather than read directly from either symbol table. Upstream's dso__synthesize_plt_symbols() gets its address from the .plt layout, finds its base name through .rela.plt and .dynsym, and appends "@plt". The existing code also handles .plt.got and .plt.sec this way.
Since name_off only identifies the base string, it cannot represent the added "@plt" suffix by itself. I think we should keep synthesizing these as regular symbols, as upstream does now. v3 only makes the byte cap apply to them and, in lazy mode, clips the index at .plt so a preceding zero-size symbol cannot overlap the PLT entries. With the cap disabled, which is the default, PLT synthesis is unchanged.
There aren't many of them. On the sample fixture, has 381 .rela.plt entries against about 319,000 .symtab symbols, and the other binary has 100 against about 637,000. Therefore, I think keeping the existing upstream path has little memory cost compared with the main symbol table.
Thanks,
Alireza
Namhyung Kim
maintainerQuoted 76 lines · Ian Rogers
Quoted 60 lines · Namhyung Kim
Hello,
Quoted 39 lines · Ian Rogers
Quoted 38 lines · Alireza Haghdoost
Quoted 2 lines · Ian Rogers
annotations. Saves 48 bytes per struct symbol (two
rb_nodes) and provides cache-friendly binary search (bsearch) and callback-based iteration.Hi Ian,
Thanks for posting this and for Cc'ing me. I built it on
edd8a9fe2eca0, the same base as the v3 series I sent last week, and compared it with v3 in eager mode (today's behavior) and with --lazy-load-symbols.All runs use perf script --no-inline --max-stack 127
-F comm,tid,time,ip,sym,dso, with one warm-up and three runs; the median is shown below. Peak anon is the maximum RssAnon from/proc/PID/status, sampled every 10 ms.1) Production: 120 s cgroup profile of a storage service on an 80-CPU host (perf record -a -g -F 99 -G <cgroup>). 54k samples, 11 DSOs, 5% of lines [unknown].
time max RSS peak anon
v3 eager 4.5 s 386 MiB 314 MiB v3 lazy 3.7 s 151 MiB 80 MiB this series 95.9 s 309 MiB 275 MiB2) A sample fixture from my v3 cover letter, a worst case with 73% of lines [unknown]:
time max RSS
v3 eager 3.2 s 334 MiB v3 lazy 1.9 s 106 MiB this series 205.3 s 449 MiBMost of the time goes to reloading. On the fixture, a profile shows it under
map__find_symbol()->dso__load()->dso__load_sym(). After each shrink, the first miss in a shrunk DSO re-parses its whole symtab, and addresses that resolve to [unknown] miss every time.Quoted 8 lines
Quoted 5 lines
The peak is still set by
dso__load(), since the whole symtab is materialized before a shrink can run. That is the main blocker for us running perf script at scale: we can't bound how much memory it will use, so we have to contain it with memory.max, and then the profiling job fails with an OOM kill instead of degrading.Hi Alireza, and thanks for the feedback! I'll see what can be done about
dso__loadin v2.Thanks for the patches and the analysis.
I think many of this memory overhead come from the symbol string. Probably we don't use most of them. Then would it be nice if we can lazy-load the strings? Then it'd have an union of a pointer and a file offset for symbol strings. The find-by-name API is used rarely for user DSOs and the kernel symbols should be fully loaded anyway.
Quoted 7 lines
Without the string (and the priv part), now it has a fixed size so it should be saved in an array directly. Probably we can add a limit there and print it with offset when it doesn't have the name.
I'm a bit skeptical about the refcount approach here. It may be hard to determine when it reclaims memory. I guess the lazy-load strings with an array would give similar savings like Alireza's work.
Thanks Namhyung. Without reference counting we lack a good way to discard symbols; we must either assume they always exist as in the current code or implement reference counting. I don't see a way around it. Alireza's patches introduced an "unknown" state for when a symbol limit was hit, but I think that'd be frustrating in practice. We could use it after trying to shrink memory use, in my opinion.
Quoted 5 lines
Fwiw, I never like switching the rbtree to an array. The issue is that the rbtree has references that really should be managed by a reference count. Doing that is a challenge in the current code and a sorted array offers similar performance while simplifying the reference counting.
Once the symbol becomes small enough, we may not need to shrink memory. Alireze shows the lazy loading works fine and IIUC it maintains symbol index which is 24 bytes. Without rbtree we can make it 32 bytes (also without the name string), then I think it'd be enough.
Thanks,
Namhyung
Quoted 20 lines
We load all symbols instead of lazily because we need the end of a symbol, which we can only determine by processing the entire symbol table as the end of one symbol is the start of the next. We generally translate one address in a DSO into a symbol, so loading all symbols is overkill. I think we can do better by loading only symbols within a certain address window rather than loading all symbols in a DSO. We can then expand this window as more symbols are needed. This at least bounds the number of symbols but isn't quite lazy loading.
Alireza also made good points about how symbols not being found or being shrunk leads to thrashing patterns. I think we can fix this by maintaining extra state in the DSO. I'm working to add this into v2.
Thanks,
IanQuoted 3 lines
Thanks,
Namhyung
Ian Rogers
reviewerv2 cover perf symbol: Reference counting, sorted arrays, and memory shrinking
In reply to Ian RogersWhen profiling large binaries or long-running workloads with many DSOs and threads, symbol and DSO tables frequently dominate perf's memory footprint and startup latency. Currently struct symbol has several structural limitations:
1. Every struct symbol embeds a 24-byte struct rb_node for the dso->symbols address rbtree, plus an optional negative-offset allocation prefix (symbol_conf.priv_size) for TUI browser indices, db-export IDs, or struct annotation state even when only a tiny fraction of symbols are ever annotated or exported.
2. Unlike struct map, struct dso, and struct thread, struct symbol is not reference-counted: once loaded into a struct dso, symbols can never be safely evicted while hist_entry, callchain_cursor_node, or map_symbol references exist, and dso__symbols() has no internal synchronization against concurrent readers/writers.
3. ELF symbol table loading (dso__load_sym_internal()) repeatedly invokes libelf section/program header lookups per symbol, calls elf_sym__name() before filtering out non-function/non-allocated symbols, runs Rust/Java demanglers on plain C symbols without fast-path prefix checks, and resorts dso->symbols with qsort() even when ELF symbol tables are already largely ordered in ascending address runs.
This series modernizes struct symbol to match the rest of tools/perf/util (DECLARE_RC_STRUCT, struct rw_semaphore, sorted pointer arrays) and introduces symbol/DSO memory shrinking and faster ELF symbol loading:
- Patch 1 adds perf bench internals symbols and perf bench internals maps to benchmark ELF symbol loading, bulk symbol insertion/sorting/lookup, incremental symbol appends, ascending map insertion, and incremental map appends.
- Patch 2 adds fast-path prefix checks in dso__demangle_sym() (before calling rust_demangle_demangle(), keeping demangle-rust-v0.c unmodified relative to upstream rustc-demangle) and in java_demangle_sym().
- Patches 3-6 encapsulate struct symbol field access behind accessor functions (annotated with __attribute__((nonnull))) and replace the negative-offset symbol_conf.priv_size allocation prefix with on-demand browser index calculation and per-DSO/per-exporter hashmaps for db-export and struct annotation.
- Patch 7 replaces dso->symbols (rb_root_cached) and dso->symbol_names with a locked struct symbols container backed by sorted pointer arrays (symbols and symbols_by_name) protected by a struct rw_semaphore, removing struct rb_node from struct symbol.
- Patch 8 wraps struct symbol in DECLARE_RC_STRUCT(symbol) with refcount_t (symbol__get() / symbol__put() / symbol__zput()) for sanitizer-checked reference counting across map_symbol, addr_location, hist_entry, callchain, thread_stack, and db-export.
- Patch 9 replaces qsort() in struct symbols with a natural-run merge sort and galloping merge (taking advantage of pre-ordered runs in .symtab/.dynsym and kallsyms) plus an in-place binary insertion fast path for small incremental appends.
- Patch 10 applies the same incremental binary-insertion sort and an O(1) tail-insertion fast path to struct maps for /proc/<pid>/maps synthesis and runtime PERF_RECORD_MMAP2 / JIT map insertions.
- Patch 11 caches ELF program and section headers during dso__load_sym_internal(), defers elf_sym__name() string table lookups until after symbol filtering, and runs symbols__fixup_end() / symbols__fixup_duplicate() once per DSO load.
- Patches 12-17 add down_read_trylock() / down_write_trylock(), protect machines->guests with a rw_semaphore, implement symbols__shrink(), dso__shrink(), dsos__shrink(), and perf_session__shrink() (with transparent on-demand reload in dso__load() that preserves pointer identity for retained symbols), and provide perf_session__malloc() / zalloc() / calloc() / realloc() / strdup() wrappers that automatically shrink unreferenced symbols and DSOs on allocation failure.
- Patch 18 adds unit tests in perf test symbols covering sorted symbol insertion, symbol/DSO shrinking, transparent DSO reload after shrink, and session allocation wrappers.
End-to-end benchmark results across the series:
perf bench internals symbols (perf binary):
Average DSO load + sort: 7.348 msec -> 5.324 msec (-27.5%)
Average DSO memory usage: 1401 KB -> 1308 KB ( -6.6%)
Average bulk insert + lookup: 1453.95 us -> 1161.00 us (-20.1%)
Average append + lookup (200 interior inserts into 10,000 symbols):
rbtree: 42.85 us -> array+qsort: 40429.75 us -> array+binary-insert: 152.70 usperf bench internals symbols -b libLLVM.so.21.1:
Average DSO load + sort: 81.196 msec -> 62.582 msec (-22.9%)
Average DSO memory usage: 9530 KB -> 9164 KB ( -3.8%)perf bench internals symbols -b libLLVM.so.21.1 --no-demangle:
Average DSO load + sort: 32.675 msec -> 19.122 msec (-41.5%)
Average DSO memory usage: 7589 KB -> 7229 KB ( -4.7%)perf bench internals maps:
Average ascending insert + lookup: 63.37 us -> 58.97 us ( -6.9%)
Average append + lookup: 7477.45 us -> 96.67 us (-98.7%, 77.3x faster)Net performance and memory trade-offs:
- Replacing the per-symbol rbtree (struct rb_node) and negative-offset symbol_conf.priv_size allocation prefix with a contiguous sorted pointer array reduces per-symbol heap overhead by 16 to 24+ bytes (saving 3.8% to 6.6% of total DSO symbol table heap memory unconditionally, and allowing unreferenced symbols and DSOs to be shrunk on demand under memory pressure).
- For the dominant symbol workloads—loading ELF binaries (-22.9% to -41.5%) and bulk symbol insertion, sorting, and lookup (-20.1%)—the contiguous array, natural-run merge sort, ELF header caching, and demangler fast paths yield substantial speedups over the rbtree baseline.
- For synthetic single-symbol interior insertions interleaved with immediate lookups into an already-large 10,000-symbol DSO (dso__insert_symbol()), an rbtree rotation takes ~0.21 us/op (42.85 us for 200 ops) whereas shifting the contiguous pointer array tail with memmove() after binary search takes ~0.76 us/op (152.70 us for 200 ops, down from 40429.75 us with naive qsort()). In real workloads, symbols are either loaded in bulk (dso__load()) or inserted into small per-prog BPF/JIT DSOs where the array fits in a few cache lines, making this sub-microsecond difference negligible in practice.
- Overall, the series reduces memory consumption across all workloads while significantly improving—or remaining neutral on—runtime performance.
v2:
- Rebased onto perf-tools-next (dropping "perf symbols: Fix broken ELF_C_READ_MMAP fallback guard", which was merged upstream, and splitting "perf libdw: Optimize fallback logic for missing DWARF" into a separate fixes series).
- Added 'perf bench internals symbols' and 'perf bench internals maps' benchmarks (Patch 1) and included before/after numbers in the optimization patches.
- Added __attribute__((nonnull)) annotations to struct symbol accessor functions (Patch 3).
- Split the symbol_conf.priv_size removal patch into three separate patches for map_browser (Patch 4), db-export (Patch 5), and annotate (Patch 6).
- Added dedicated optimization patches for Rust/Java demangler prefix checks (Patch 2), natural-run merge sort and binary insertion in struct symbols (Patch 9), tail-insertion and incremental binary-insertion sort in struct maps (Patch 10), and ELF program/section header caching and deferred symbol name lookup (Patch 11).
- Replaced periodic heuristic LRU shrinking and windowed ELF symbol loading with on-demand shrinking (symbols__shrink(), dso__shrink(), dsos__shrink(), perf_session__shrink()) triggered on allocation failure via session-scoped allocation wrappers, plus transparent DSO reload preserving retained symbol pointers (Patches 12-18).
v1: https://lore.kernel.org/lkml/[email protected]/
Ian Rogers (18):
perf bench internals: Add symbols and maps benchmarks perf demangle: Add fast-path prefix checks for Rust and Java
demanglers
perf symbol: Add accessor functions for struct symbol fields perf map_browser: Compute symbol browser index on demand perf db-export: Store symbol database IDs in a hashmap perf annotate: Move struct annotation to a hashmap and remove
symbol_conf.priv_size
perf symbol: Switch backing storage from rbtree to struct symbols
array
perf symbol: Add reference counting and DECLARE_RC_STRUCT(symbol) perf symbol: Optimize symbol sorting with natural-run merge sort and
binary insertion
perf maps: Optimize tail insertion and small-append sorting perf symbol-elf: Cache ELF program/section headers and defer symbol
name lookup
perf rwsem: Add down_read_trylock() and down_write_trylock() perf vdso: Constify dso parameter in dso__is_vdso()
perf machine: Protect machines->guests rbtree with a rw_semaphore perf symbol: Add memory shrinking for symbols, DSOs, and machines perf session: Add dso__session() and session-scoped allocation
wrappers
perf symbol: Use session allocation wrappers in symbol, DSO, thread,
and hist paths
perf test symbols: Add tests for symbol shrinking and session alloc
wrappers
tools/perf/Documentation/perf-bench.txt | 6 + .../arch/powerpc/util/skip-callchain-idx.c | 2 +- tools/perf/arch/powerpc/util/sym-handling.c | 40 +- tools/perf/bench/Build | 2 + tools/perf/bench/bench.h | 2 + tools/perf/bench/maps.c | 186 +++ tools/perf/bench/symbols.c | 239 +++ tools/perf/builtin-annotate.c | 22 +- tools/perf/builtin-bench.c | 2 + tools/perf/builtin-c2c.c | 14 - tools/perf/builtin-diff.c | 4 +- tools/perf/builtin-kallsyms.c | 9 +- tools/perf/builtin-kmem.c | 82 +- tools/perf/builtin-kwork.c | 4 +- tools/perf/builtin-lock.c | 25 +- tools/perf/builtin-mem.c | 2 +- tools/perf/builtin-report.c | 14 - tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 18 +- tools/perf/builtin-timechart.c | 2 +- tools/perf/builtin-top.c | 24 +- tools/perf/builtin-trace.c | 23 +- tools/perf/tests/builtin-test.c | 1 - tools/perf/tests/dlfilter-test.c | 12 +- tools/perf/tests/dwarf-unwind.c | 2 +- tools/perf/tests/hists_common.c | 9 +- tools/perf/tests/hists_cumulate.c | 8 +- tools/perf/tests/hists_filter.c | 10 +- tools/perf/tests/hists_link.c | 16 +- tools/perf/tests/hists_output.c | 9 +- tools/perf/tests/kallsyms-split.c | 5 +- tools/perf/tests/maps.c | 58 + tools/perf/tests/pe-file-parsing.c | 1 + tools/perf/tests/symbols.c | 454 +++++- tools/perf/tests/vmlinux-kallsyms.c | 187 ++- tools/perf/ui/browser.c | 3 + tools/perf/ui/browsers/annotate.c | 10 +- tools/perf/ui/browsers/hists.c | 35 +- tools/perf/ui/browsers/map.c | 88 +- tools/perf/ui/gtk/annotate.c | 6 +- tools/perf/ui/stdio/hist.c | 6 +- tools/perf/util/addr2line.c | 8 +- tools/perf/util/addr_location.c | 4 + .../perf/util/annotate-arch/annotate-alpha.c | 5 +- .../perf/util/annotate-arch/annotate-arm64.c | 12 +- .../util/annotate-arch/annotate-loongarch.c | 15 +- tools/perf/util/annotate-arch/annotate-s390.c | 6 +- tools/perf/util/annotate-arch/annotate-x86.c | 14 +- tools/perf/util/annotate-data.c | 12 +- tools/perf/util/annotate.c | 211 ++- tools/perf/util/annotate.h | 6 +- tools/perf/util/auxtrace.c | 111 +- tools/perf/util/block-info.c | 12 +- tools/perf/util/bpf_lock_contention.c | 35 +- tools/perf/util/c2c-function.c | 32 +- tools/perf/util/call-path.c | 15 +- tools/perf/util/callchain.c | 23 +- tools/perf/util/capstone.c | 22 +- tools/perf/util/data-convert-json.c | 4 +- tools/perf/util/db-export.c | 99 +- tools/perf/util/db-export.h | 8 + tools/perf/util/debug.c | 3 +- tools/perf/util/demangle-java.c | 23 +- tools/perf/util/disasm.c | 69 +- tools/perf/util/dlfilter.c | 12 +- tools/perf/util/dso.c | 257 ++- tools/perf/util/dso.h | 94 +- tools/perf/util/dsos.c | 76 + tools/perf/util/dsos.h | 3 + tools/perf/util/event.c | 12 +- tools/perf/util/evsel_fprintf.c | 2 +- tools/perf/util/hist.c | 70 +- tools/perf/util/intel-pt.c | 74 +- tools/perf/util/libbfd.c | 14 +- tools/perf/util/libdw.c | 5 +- tools/perf/util/llvm.c | 8 +- tools/perf/util/lock-contention.c | 8 +- tools/perf/util/machine.c | 189 ++- tools/perf/util/machine.h | 9 + tools/perf/util/map.c | 27 +- tools/perf/util/map.h | 18 +- tools/perf/util/map_symbol.c | 4 +- tools/perf/util/maps.c | 284 +++- tools/perf/util/mutex.h | 3 + tools/perf/util/probe-event.c | 148 +- tools/perf/util/python.c | 36 +- tools/perf/util/rwsem.c | 20 + tools/perf/util/rwsem.h | 2 + tools/perf/util/session.c | 78 + tools/perf/util/session.h | 8 + tools/perf/util/sort.c | 47 +- tools/perf/util/srcline.c | 22 +- tools/perf/util/symbol-elf.c | 311 +++- tools/perf/util/symbol.c | 1438 +++++++++++++---- tools/perf/util/symbol.h | 187 ++- tools/perf/util/symbol_conf.h | 1 - tools/perf/util/symbol_fprintf.c | 37 +- tools/perf/util/thread-stack.c | 18 +- tools/perf/util/threads.c | 17 +- tools/perf/util/unwind-libdw.c | 4 +- tools/perf/util/unwind-libunwind.c | 2 +- tools/perf/util/vdso.c | 2 +- tools/perf/util/vdso.h | 2 +- 103 files changed, 4557 insertions(+), 1389 deletions(-) create mode 100644 tools/perf/bench/maps.c create mode 100644 tools/perf/bench/symbols.c
Signature · 1 line
2.56.0.313.g2479322e68-goog
Ian Rogers
reviewerPATCH v2 1/18 perf bench internals: Add symbols and maps benchmarks
Add two new 'perf bench internals' subcommands to measure symbol and map operations directly:
- 'perf bench internals symbols':
1. Loads symbols from an ELF binary (defaulting to the running perf executable, or specified via -b/--binary) and sorts them by name, reporting both average runtime and heap memory consumption.
2. Inserts synthetic symbols in two ascending runs (simulating .symtab and .dynsym), fixes up symbol ends and duplicates, and queries each symbol by address.
3. Incrementally appends symbols at interior addresses and immediately queries each appended symbol by address.
- 'perf bench internals maps':
1. Inserts non-overlapping maps in ascending address order (simulating /proc/<pid>/maps synthesis) and queries each map by address.
2. Incrementally appends maps at interior addresses (simulating runtime PERF_RECORD_MMAP2 / JIT events) and immediately queries each map by address and by name.
--- tools/perf/Documentation/perf-bench.txt | 6 + tools/perf/bench/Build | 2 + tools/perf/bench/bench.h | 2 + tools/perf/bench/maps.c | 186 +++++++++++++++++++ tools/perf/bench/symbols.c | 236 ++++++++++++++++++++++++ tools/perf/builtin-bench.c | 2 + 6 files changed, 434 insertions(+) create mode 100644 tools/perf/bench/maps.c create mode 100644 tools/perf/bench/symbols.c
tools/perf/Documentation/perf-bench.txt +6 -0
@@ -351,6 +351,12 @@ SUITES FOR 'internals' *synthesize*:: Suite for evaluating perf's event synthesis performance. +*symbols*:: +Suite for evaluating ELF symbol loading, sorting, and lookup performance. + +*maps*:: +Suite for evaluating maps insertion, sorting, and lookup performance. + SEE ALSO -------- linkperf:perf[1]
tools/perf/bench/Build +2 -0
@@ -20,6 +20,8 @@ perf-bench-y += evlist-open-close.o perf-bench-y += breakpoint.o perf-bench-y += pmu-scan.o perf-bench-y += uprobe.o +perf-bench-y += symbols.o +perf-bench-y += maps.o perf-bench-$(CONFIG_X86_64) += mem-memcpy-x86-64-asm.o perf-bench-$(CONFIG_X86_64) += mem-memset-x86-64-asm.o
tools/perf/bench/bench.h +2 -0
@@ -52,6 +52,8 @@ int bench_uprobe_trace_printk(int argc, const char **argv);int bench_uprobe_empty_ret (int argc,const char **argv);int bench_uprobe_trace_printk_ret (int argc,const char **argv);int bench_pmu_scan (int argc,const char **argv); +int bench_symbols (int argc,const char **argv); +int bench_maps (int argc,const char **argv);#define BENCH_FORMAT_DEFAULT_STR "default"#define BENCH_FORMAT_DEFAULT 0
tools/perf/bench/maps.c new file +186 -0
186 lines not on this page · raw message
tools/perf/bench/symbols.c new file +236 -0
236 lines not on this page · raw message
tools/perf/builtin-bench.c +2 -0
@@ -97,6 +97,8 @@ static const struct bench internals_benchmarks[] = { { "inject-build-id", "Benchmark build-id injection", bench_inject_build_id }, { "evlist-open-close", "Benchmark evlist open and close", bench_evlist_open_close }, { "pmu-scan", "Benchmark sysfs PMU info scanning", bench_pmu_scan }, + { "symbols", "Benchmark symbol loading, sorting and lookup", bench_symbols }, + { "maps", "Benchmark maps insertion, sorting and lookup", bench_maps }, { NULL, NULL, NULL } };
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 01/18] perf bench internals: Add symbols and maps benchmarks
tools/perf/bench/symbols.c line 57
Quoted 3 lines · Ian Rogers
tools/perf/bench/symbols.c new file +3
-0+#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) +struct stats mem_stats; +#endif
Does this build on a libc without __GLIBC_PREREQ, such as musl ? How about nesting it:
#if defined(__GLIBC__)
#if __GLIBC_PREREQ(2, 33)
#define HAVE_MALLINFO2
#endif
#endif
and then #ifdef HAVE_MALLINFO2 at the five uses.
Two smaller things: 'perf bench internals symbols' exits with an error when the binary has no defined symbols (a stripped executable), even though the synthetic insert and lookup phases don't need the binary. And symbol__init() has no matching symbol__exit().
Namhyung Kim
maintainerRe: [PATCH v2 01/18] perf bench internals: Add symbols and maps benchmarks
In reply to Ian RogersQuoted 13 lines · Ian Rogers
Add two new 'perf bench internals' subcommands to measure symbol and map operations directly:
- 'perf bench internals symbols':
1. Loads symbols from an ELF binary (defaulting to the running perf executable, or specified via -b/--binary) and sorts them by name, reporting both average runtime and heap memory consumption.
2. Inserts synthetic symbols in two ascending runs (simulating .symtab and .dynsym), fixes up symbol ends and duplicates, and queries each symbol by address.
3. Incrementally appends symbols at interior addresses and immediately queries each appended symbol by address.
Quoted 6 lines
- 'perf bench internals maps':
1. Inserts non-overlapping maps in ascending address order (simulating /proc/<pid>/maps synthesis) and queries each map by address.
2. Incrementally appends maps at interior addresses (simulating runtimePERF_RECORD_MMAP2/ JIT events) and immediately queries each map by address and by name.
Can you please add an example output?
Quoted 3 lines
Signed-off-by Ian Rogers <[email protected]>---
[SNIP]
Quoted 70 lines
tools/perf/bench/symbols.c new file +64
-0@@ -0,0 +1,236 @@ +// SPDX-License-Identifier: GPL-2.0 +/* +* Benchmark symbol loading, sorting, and address/name lookup. +*/ +#include <elf.h> +#include <errno.h> +#include <inttypes.h> +#include <limits.h> +#include <malloc.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/time.h> +#include <linux/compiler.h> +#include <linux/kernel.h> +#include <linux/time64.h> +#include <subcmd/parse-options.h> +#include "bench.h" +#include "../util/debug.h" +#include "../util/dso.h" +#include "../util/map.h" +#include "../util/stat.h" +#include "../util/symbol.h" +#include "../util/util.h" + +static unsigned int iterations =20 ; +static unsigned int nr_symbols =10000 ; +static unsigned int nr_appends =200 ; +static const char *binary_path; +static bool no_demangle; + +static const struct option options[] = { +OPT_UINTEGER ('i' ,"iterations" , &iterations, +"Number of iterations used to compute average (default: 20)" ), +OPT_STRING ('b' ,"binary" , &binary_path,"path" , +"Binary path for ELF symbol load benchmark (default: perf binary)" ), +OPT_UINTEGER ('n' ,"nr-symbols" , &nr_symbols, +"Number of synthetic symbols for bulk benchmark (default: 10000)" ), +OPT_UINTEGER ('a' ,"nr-appends" , &nr_appends, +"Number of incremental symbol appends per iteration (default: 200)" ), +OPT_BOOLEAN (0 ,"no-demangle" , &no_demangle, +"Disable symbol demangling during ELF load" ), +OPT_END () +}; + +static const char *const bench_usage[] = { +"perf bench internals symbols <options>" , +NULL +}; + +static int do_dso_load_bench (const char *path) +{ +struct stats time_stats; +#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) +struct stats mem_stats; +#endif +double time_average, time_stddev; +unsigned int i; + +init_stats (&time_stats); +#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) +init_stats (&mem_stats); +#endif
tools/perf/bench/symbols.c line 74
Quoted 10 lines
tools/perf/bench/symbols.c new file +10
-0+ +for (i =0 ; i < iterations; i++) { +struct timeval start, end, diff; +struct map *map; +struct dso *dso; +u64 runtime_us; +#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) +size_t mem_before =mallinfo2 ().uordblks; +size_t mem_after; +#endif
Can we just use maxrss from getrusage()? It doesn't need to be average, probably we just want to know the maximum memory usage.
Quoted 4 lines
tools/perf/bench/symbols.c new file +4
-0+ +gettimeofday (&start,NULL ); + map =dso__new_map (path); +if (!map) {
tools/perf/bench/symbols.c line 88
Quoted 10 lines
tools/perf/bench/symbols.c new file +10
-0+pr_err ("Failed to create map for %s\n" , path); +return -ENOMEM ; + } +if (map__load (map) <0 ) { +pr_err ("Failed to load symbols from %s\n" , path); +map__put (map); +return -EINVAL ; + } + dso =map__dso (map); +dso__sort_by_name (dso);
Probably better to have a separate test not to use symbol names. As we expect to have lazy symbol loading soon.
Thanks,
Namhyung
Quoted 164 lines
tools/perf/bench/symbols.c new file +148
-0+gettimeofday (&end,NULL ); + +#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) + mem_after =mallinfo2 ().uordblks; +if (mem_after > mem_before) +update_stats (&mem_stats, (mem_after - mem_before) /1024 ); +#endif +map__put (map); + +timersub (&end, &start, &diff); + runtime_us = diff.tv_sec *USEC_PER_SEC + diff.tv_usec; +update_stats (&time_stats, runtime_us); + } + + time_average =avg_stats (&time_stats) /USEC_PER_MSEC ; + time_stddev =stddev_stats (&time_stats) /USEC_PER_MSEC ; +printf (" Average DSO load + sort: %.3f msec (+- %.3f msec)\n" , + time_average, time_stddev); + +#if defined(__GLIBC__) &&__GLIBC_PREREQ (2 ,33 ) +printf (" Average DSO memory usage: %.0f KB (+- %.0f KB)\n" , +avg_stats (&mem_stats),stddev_stats (&mem_stats)); +#endif +return 0 ; +} + +static int do_synthetic_symbols_bench (void ) +{ +struct stats bulk_stats, append_stats; +unsigned int i, j, half = nr_symbols /2 ; + +if (half ==0 ) +return 0 ; + +init_stats (&bulk_stats); +init_stats (&append_stats); + +for (i =0 ; i < iterations; i++) { +struct timeval start, end, diff; +struct dso *dso =dso__new ("[bench_symbols]" ); +u64 runtime_us; +char name[32 ]; + +if (!dso) +return -ENOMEM ; + +/* +* Phase 1: Insert symbols in two ascending runs (simulating +* .symtab + .dynsym), fixup ends and duplicates, and look up +* all symbols by address. +*/ +gettimeofday (&start,NULL ); +for (j =0 ; j < nr_symbols; j++) { +u64 addr = (j < half) ? + (0x100000ULL + (u64 )j *0x40ULL ) : + (0x100020ULL + (u64)(j - half) *0x40ULL ); +struct symbol *sym; + +scnprintf (name,sizeof (name),"bench_sym_%06u" , j); + sym =symbol__new (addr,0x18 ,STB_GLOBAL ,STT_FUNC , name); +if (!sym) { +dso__put (dso); +return -ENOMEM ; + } +symbols__insert (dso__symbols (dso), sym); + } +symbols__fixup_end (dso__symbols (dso), false); +symbols__fixup_duplicate (dso__symbols (dso)); + +for (j =0 ; j < nr_symbols; j++) { +u64 addr = (j < half) ? + (0x100000ULL + (u64 )j *0x40ULL +4 ) : + (0x100020ULL + (u64)(j - half) *0x40ULL +4 ); + + (void )dso__find_symbol_nocache (dso, addr); + } +gettimeofday (&end,NULL ); +timersub (&end, &start, &diff); + runtime_us = diff.tv_sec *USEC_PER_SEC + diff.tv_usec; +update_stats (&bulk_stats, runtime_us); + +/* +* Phase 2: Incrementally append one symbol at a time at an +* interior address and immediately query by address. +*/ +gettimeofday (&start,NULL ); +for (j =0 ; j < nr_appends; j++) { +u64 addr =0x100010ULL + ((u64)(j *197 % half)) *0x40ULL ; +struct symbol *sym; + +scnprintf (name,sizeof (name),"jit_sym_%06u" , j); + sym =symbol__new (addr,0x08 ,STB_GLOBAL ,STT_FUNC , name); +if (!sym) { +dso__put (dso); +return -ENOMEM ; + } +dso__insert_symbol (dso, sym); + (void )dso__find_symbol_nocache (dso, addr +2 ); + } +gettimeofday (&end,NULL ); +timersub (&end, &start, &diff); + runtime_us = diff.tv_sec *USEC_PER_SEC + diff.tv_usec; +update_stats (&append_stats, runtime_us); + +dso__put (dso); + } + +printf (" Average bulk insert + lookup: %.3f usec (+- %.3f usec)\n" , +avg_stats (&bulk_stats),stddev_stats (&bulk_stats)); +printf (" Average append + lookup: %.3f usec (+- %.3f usec)\n" , +avg_stats (&append_stats),stddev_stats (&append_stats)); +return 0 ; +} + +int bench_symbols (int argc,const char **argv) +{ +char exe_buf[PATH_MAX ]; +const char *path; +bool orig_demangle; +int err; + + argc =parse_options (argc, argv, options, bench_usage,0 ); +if (argc) { +usage_with_options (bench_usage, options); +exit (EXIT_FAILURE ); + } + +if (symbol__init (NULL ) <0 ) +return -1 ; + + orig_demangle = symbol_conf.demangle; +if (no_demangle) + symbol_conf.demangle = false; + +if (binary_path) { + path = binary_path; + }else { +perf_exe (exe_buf,sizeof (exe_buf)); + path = exe_buf; + } + + err =do_dso_load_bench (path); +if (!err) + err =do_synthetic_symbols_bench (); + + symbol_conf.demangle = orig_demangle; +return err; +}tools/perf/builtin-bench.c +2
-0@@ -97,6 +97,8 @@ static const struct bench internals_benchmarks[] = { { "inject-build-id", "Benchmark build-id injection", bench_inject_build_id }, { "evlist-open-close", "Benchmark evlist open and close", bench_evlist_open_close }, { "pmu-scan", "Benchmark sysfs PMU info scanning", bench_pmu_scan }, + { "symbols", "Benchmark symbol loading, sorting and lookup", bench_symbols }, + { "maps", "Benchmark maps insertion, sorting and lookup", bench_maps }, { NULL, NULL, NULL } };Signature · 1 line
2.56.0.313.g2479322e68-goog
Ian Rogers
reviewerPATCH v2 2/18 perf demangle: Add fast-path prefix checks for Rust and Java demanglers
In reply to Ian RogersDuring DSO symbol loading in dso__load_sym_internal(), dso__demangle_sym() is called for every ELF symbol. Both rust_demangle_demangle() and java_demangle_sym() previously performed string searches or buffer allocations before checking whether the symbol name could possibly be a mangled Rust or Java symbol, adding unnecessary overhead on plain C symbols.
Add fast-path prefix checks in dso__demangle_sym() before calling rust_demangle_demangle() (using strstarts() for '_ZN', 'ZN', '__ZN', '_R', 'R', or '__R', keeping demangle-rust-v0.c unmodified relative to upstream rustc-demangle) and in java_demangle_sym() (requiring 'L', ';', '(', and ')') to reject non-Rust and non-Java symbol names immediately.
Using 'perf bench internals symbols':
Before (perf binary):
Average DSO load + sort: 7.348 msec (+- 0.069 msec)
After (perf binary, -4.2%):
Average DSO load + sort: 7.041 msec (+- 0.051 msec)Before (libLLVM.so.21.1):
Average DSO load + sort: 81.196 msec (+- 0.780 msec)
After (libLLVM.so.21.1, -3.2%):
Average DSO load + sort: 78.599 msec (+- 0.857 msec)--- tools/perf/util/demangle-java.c | 23 +++++++++++++++++++++-- tools/perf/util/symbol.c | 15 ++++++++++++++- 2 files changed, 35 insertions(+), 3 deletions(-)
tools/perf/util/demangle-java.c +21 -2
@@ -158,14 +158,33 @@ char * java_demangle_sym(const char *str, int flags) { char *buf, *ptr; -const char *p; +const char *l, *semi, *open, *p;size_t len, l1 =0 ;if (!str)return NULL ; +/* +* OpenJDK JIT symbols have the form: +* [prefix]L<class>;<method>(<args>)<return_type> +* where the return type (such as 'I' for int or 'V' for void) is at the +* end after ')'. In __demangle_java_sym(), the parser starts in +* MODE_PREFIX and only reaches MODE_TYPE (required when ')' is reached) +* via 'L' (MODE_CLASS) -> ';' (MODE_FUNC) -> '(' (MODE_TYPE). Verify +* those delimiters appear in order before allocating the output buffer. +*/ + l =strchr (str,'L' ); +if (!l) +return NULL ; + semi =strchr (l +1 ,';' ); +if (!semi) +return NULL ; + open =strchr (semi +1 ,'(' ); +if (!open) +return NULL ; +/* find start of return type */ - p =strrchr (str,')' ); + p =strrchr (open +1 ,')' );if (!p)return NULL ;
tools/perf/util/symbol.c +14 -1
@@ -2738,7 +2738,20 @@ char *dso__demangle_sym(struct dso *dso, int kmodule, const char *elf_name)if (!want_demangle ((dso &&dso__kernel (dso)) || kmodule))return demangled; -rust_demangle_demangle (elf_name, &rust_demangle); +/* +* Rust mangled symbols start with one of the prefixes checked by +* rust_demangle_legacy_demangle() ("_ZN", Windows dbghelp "ZN", or +* macOS "__ZN") or rust_demangle_v0_demangle() ("_R", Windows dbghelp +* "R", or macOS "__R") in demangle-rust-v0.c. Check the prefix here so +* we avoid calling rust_demangle_demangle() (which scans the entire +* symbol name with strstr() and strlen() for ThinLTO ".llvm." suffixes) +* on non-Rust symbols, while keeping demangle-rust-v0.c unmodified +* relative to upstream rustc-demangle. +*/ +if (strstarts (elf_name,"_ZN" ) ||strstarts (elf_name,"ZN" ) || +strstarts (elf_name,"__ZN" ) ||strstarts (elf_name,"_R" ) || +strstarts (elf_name,"R" ) ||strstarts (elf_name,"__R" )) +rust_demangle_demangle (elf_name, &rust_demangle);if (rust_demangle_is_known (&rust_demangle)) {/* A rust mangled name. */ if (rust_demangle.mangled_len ==0 )
Signature · 1 line
2.56.0.313.g2479322e68-goog
Ian Rogers
reviewerPATCH v2 3/18 perf symbol: Add accessor functions for struct symbol fields
In reply to Ian RogersTo track symbol lifetimes across DSO reloads, histogram decay, and on-demand memory shrinking, struct symbol needs to be reference-counted and wrapped with DECLARE_RC_STRUCT(symbol) so REFCNT_CHECKING builds can detect missing or leaked reference counts.
To prepare for wrapping struct symbol in DECLARE_RC_STRUCT(symbol), add inline accessor functions for struct symbol fields (symbol__start, symbol__set_start, symbol__end, symbol__set_end, symbol__name, symbol__namelen, symbol__arch_sym, symbol__set_arch_sym) and convert direct member accesses across tools/perf.
--- .../arch/powerpc/util/skip-callchain-idx.c | 2 +- tools/perf/arch/powerpc/util/sym-handling.c | 8 +- tools/perf/builtin-annotate.c | 4 +- tools/perf/builtin-diff.c | 4 +- tools/perf/builtin-kallsyms.c | 7 +- tools/perf/builtin-kmem.c | 29 ++--- tools/perf/builtin-kwork.c | 4 +- tools/perf/builtin-lock.c | 10 +- tools/perf/builtin-mem.c | 2 +- tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 14 +-- tools/perf/builtin-timechart.c | 2 +- tools/perf/builtin-top.c | 22 ++-- tools/perf/builtin-trace.c | 22 ++-- tools/perf/tests/dlfilter-test.c | 4 +- tools/perf/tests/dwarf-unwind.c | 2 +- tools/perf/tests/hists_common.c | 4 +- tools/perf/tests/hists_cumulate.c | 4 +- tools/perf/tests/hists_output.c | 2 +- tools/perf/tests/symbols.c | 12 +- tools/perf/tests/vmlinux-kallsyms.c | 34 +++--- tools/perf/ui/browsers/annotate.c | 10 +- tools/perf/ui/browsers/hists.c | 10 +- tools/perf/ui/browsers/map.c | 8 +- tools/perf/ui/gtk/annotate.c | 6 +- tools/perf/ui/stdio/hist.c | 3 +- .../perf/util/annotate-arch/annotate-arm64.c | 12 +- .../util/annotate-arch/annotate-loongarch.c | 4 +- tools/perf/util/annotate-arch/annotate-x86.c | 14 +-- tools/perf/util/annotate-data.c | 12 +- tools/perf/util/annotate.c | 69 ++++++------ tools/perf/util/auxtrace.c | 14 +-- tools/perf/util/block-info.c | 6 +- tools/perf/util/bpf_lock_contention.c | 16 +-- tools/perf/util/c2c-function.c | 19 ++-- tools/perf/util/callchain.c | 19 ++-- tools/perf/util/capstone.c | 20 ++-- tools/perf/util/data-convert-json.c | 4 +- tools/perf/util/db-export.c | 2 +- tools/perf/util/debug.c | 3 +- tools/perf/util/disasm.c | 49 +++++---- tools/perf/util/dlfilter.c | 12 +- tools/perf/util/dso.c | 4 +- tools/perf/util/event.c | 4 +- tools/perf/util/evsel_fprintf.c | 2 +- tools/perf/util/hist.c | 16 +-- tools/perf/util/intel-pt.c | 8 +- tools/perf/util/libbfd.c | 4 +- tools/perf/util/llvm.c | 6 +- tools/perf/util/lock-contention.c | 2 +- tools/perf/util/machine.c | 26 ++--- tools/perf/util/map.c | 6 +- tools/perf/util/map.h | 2 +- tools/perf/util/probe-event.c | 41 +++---- tools/perf/util/python.c | 18 +-- tools/perf/util/sort.c | 40 +++---- tools/perf/util/srcline.c | 16 +-- tools/perf/util/symbol-elf.c | 10 +- tools/perf/util/symbol.c | 103 +++++++++--------- tools/perf/util/symbol.h | 43 +++++++- tools/perf/util/symbol_fprintf.c | 14 +-- tools/perf/util/thread-stack.c | 8 +- tools/perf/util/unwind-libdw.c | 2 +- 63 files changed, 478 insertions(+), 407 deletions(-)
tools/perf/arch/powerpc/util/skip-callchain-idx.c +1 -1
@@ -239,7 +239,7 @@ int arch_skip_callchain_idx(struct thread *thread, struct ip_callchain *chain) rc =check_return_addr (dso,map__map_ip (al.map, ip));pr_debug ("[DSO %s, sym %s, ip 0x%" PRIx64"] rc %d\n" , -dso__long_name (dso), al.sym->name, ip, rc); +dso__long_name (dso), al.sym ?symbol__name (al.sym) :"unknown" , ip, rc);if (rc ==0 ) {/*
tools/perf/arch/powerpc/util/sym-handling.c +4 -4
@@ -13,7 +13,7 @@int arch__choose_best_symbol (struct symbol *syma,struct symbol *symb__maybe_unused ) { -char *sym = syma->name; +const char *sym =symbol__name (syma);#if !defined(_CALL_ELF) || _CALL_ELF !=2 /* Skip over any initial dot */
@@ -69,7 +69,7 @@ const char *arch__normalize_symbol_name(const char *name)#ifdef HAVE_LIBELF_SUPPORT void arch__sym_update (struct symbol *s,GElf_Sym *sym) { - s->arch_sym = sym->st_other; +symbol__set_arch_sym (s, sym->st_other); }#endif
@@ -102,7 +102,7 @@ void arch__fix_tev_from_maps(struct perf_probe_event *pev,return ; } - lep_offset =PPC64_LOCAL_ENTRY_OFFSET (sym->arch_sym); + lep_offset =PPC64_LOCAL_ENTRY_OFFSET (symbol__arch_sym (sym));if (map__dso (map)->symtab_type ==DSO_BINARY_TYPE__KALLSYMS ) tev->point.offset +=PPC64LE_LEP_OFFSET ;
@@ -131,7 +131,7 @@ void arch__post_process_probe_trace_events(struct perf_probe_event *pev,for (i =0 ; i < ntevs; i++) { tev = &pev->tevs[i];map__for_each_symbol (map, sym, tmp) { -if (map__unmap_ip (map, sym->start) == tev->point.address) { +if (map__unmap_ip (map,symbol__start (sym)) == tev->point.address) {arch__fix_tev_from_maps (pev, tev, map, sym);break ; }
tools/perf/builtin-annotate.c +2 -2
@@ -252,7 +252,7 @@ static int add_sample(struct perf_sample *sample,if ((!ann->has_br_stack || !has_annotation (ann)) && ann->sym_hist_filter !=NULL && (al->sym ==NULL || -strcmp (ann->sym_hist_filter, al->sym->name) !=0 )) { +strcmp (ann->sym_hist_filter,symbol__name (al->sym)) !=0 )) {/* We're only interested in a symbol named sym_hist_filter */ /* * FIXME: why isn't this done in the symbol_filter when loading
@@ -423,7 +423,7 @@ static void hists__find_annotations(struct hists *hists, goto find_next;if (ann->sym_hist_filter && - (strcmp (he->ms.sym->name, ann->sym_hist_filter) !=0 )) + (strcmp (symbol__name (he->ms.sym), ann->sym_hist_filter) !=0 )) goto find_next;if (ann->min_percent) {
tools/perf/builtin-diff.c +2 -2
@@ -1358,10 +1358,10 @@ static int cycles_printf(struct hist_entry *he, struct hist_entry *pair, bi = block_he->block_info; - start_line =map__srcline (he->ms.map, bi->sym->start + bi->start, + start_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->start, he->ms.sym); - end_line =map__srcline (he->ms.map, bi->sym->start + bi->end, + end_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->end, he->ms.sym); if (start_line != SRCLINE_UNKNOWN &&
tools/perf/builtin-kallsyms.c +4 -3
@@ -48,9 +48,10 @@ static int __cmd_kallsyms(int argc, const char **argv) dso =map__dso (map);printf ("%s: %s %s %#" PRIx64"-%#" PRIx64" (%#" PRIx64"-%#" PRIx64")\n" , - symbol->name,dso__short_name (dso),dso__long_name (dso), -map__unmap_ip (map, symbol->start),map__unmap_ip (map, symbol->end), - symbol->start, symbol->end); +symbol__name (symbol),dso__short_name (dso),dso__long_name (dso), +map__unmap_ip (map,symbol__start (symbol)), +map__unmap_ip (map,symbol__end (symbol)), +symbol__start (symbol),symbol__end (symbol)); } out:machine__delete (machine);
tools/perf/builtin-kmem.c +15 -14
@@ -310,7 +310,7 @@ static struct rb_root page_caller_sorted;struct alloc_func {u64 start;u64 end; -char *name; +const char *name; };static int nr_alloc_funcs;
@@ -368,7 +368,7 @@ static int build_alloc_func_list(void) }map__for_each_symbol (kernel_map, sym, node) { -if (regexec (&alloc_func_regex, sym->name,0 ,NULL ,0 )) +if (regexec (&alloc_func_regex,symbol__name (sym),0 ,NULL ,0 ))continue ; func = realloc(alloc_func_list,
@@ -376,10 +376,10 @@ static int build_alloc_func_list(void)if (func ==NULL )return -ENOMEM ; -pr_debug ("alloc func: %s\n" , sym->name); - func[nr_alloc_funcs].start = sym->start; - func[nr_alloc_funcs].end = sym->end; - func[nr_alloc_funcs].name = sym->name; +pr_debug ("alloc func: %s\n" ,symbol__name (sym)); + func[nr_alloc_funcs].start =symbol__start (sym); + func[nr_alloc_funcs].end =symbol__end (sym); + func[nr_alloc_funcs].name =symbol__name (sym); alloc_func_list = func; nr_alloc_funcs++;
@@ -1048,11 +1048,12 @@ static void __print_slab_result(struct rb_root *root, }else addr = data->ptr; -if (sym !=NULL ) -snprintf (buf,sizeof (buf),"%s+%" PRIx64"" , sym->name, - addr -map__unmap_ip (map, sym->start)); -else +if (sym !=NULL ) { +snprintf (buf,sizeof (buf),"%s+%" PRIx64"" ,symbol__name (sym), + addr -map__unmap_ip (map,symbol__start (sym))); + }else {snprintf (buf,sizeof (buf),"%#" PRIx64"" , addr); + }printf (" %-34s |" , buf); printf(" %9llu/%-5lu | %9llu/%-5lu | %8lu | %9lu | %6.3f%%\n",
@@ -1105,12 +1106,12 @@ static void __print_page_alloc_result(struct perf_session *session, int n_lines)struct symbol *sym;struct map *map;char buf[32 ]; -char *caller = buf; +const char *caller = buf; data = rb_entry(next, struct page_stat, node); sym =machine__find_kernel_symbol (machine, data->callsite, &map);if (sym) - caller = sym->name; + caller =symbol__name (sym);else scnprintf (buf,sizeof (buf),"%" PRIx64, data->callsite);
@@ -1147,12 +1148,12 @@ static void __print_page_caller_result(struct perf_session *session, int n_linesstruct symbol *sym;struct map *map;char buf[32 ]; -char *caller = buf; +const char *caller = buf; data = rb_entry(next, struct page_stat, node); sym =machine__find_kernel_symbol (machine, data->callsite, &map);if (sym) - caller = sym->name; + caller =symbol__name (sym);else scnprintf (buf,sizeof (buf),"%" PRIx64, data->callsite);
tools/perf/builtin-kwork.c +2 -2
@@ -768,8 +768,8 @@ static void timehist_save_callchain(struct perf_kwork *kwork, sym = node->ms.sym;if (sym) { -if (!strcmp (sym->name,"__softirqentry_text_start" ) || - !strcmp (sym->name,"__do_softirq" )) +if (!strcmp (symbol__name (sym),"__softirqentry_text_start" ) || + !strcmp (symbol__name (sym),"__do_softirq" ))symbol__set_ignore (sym, true); }
tools/perf/builtin-lock.c +5 -5
@@ -826,12 +826,12 @@ static int get_symbol_name_offset(struct map *map, struct symbol *sym, u64 ip,return 0 ; } - offset =map__map_ip (map, ip) - sym->start; + offset =map__map_ip (map, ip) -symbol__start (sym);if (offset) -return scnprintf (buf, size,"%s+%#lx" , sym->name, offset); +return scnprintf (buf, size,"%s+%#lx" ,symbol__name (sym), offset);else -return strlcpy (buf, sym->name, size); +return strlcpy (buf,symbol__name (sym), size); }static int lock_contention_caller (struct perf_sample *sample,char *buf,int size)
@@ -1006,7 +1006,7 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)return -ENOMEM ; } - addrs[filters.nr_addrs++] =map__unmap_ip (kmap, sym->start); + addrs[filters.nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); filters.addrs = addrs; } }
@@ -1059,7 +1059,7 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)case LOCK_AGGR_ADDR : sym =machine__find_kernel_symbol (machine, key, &kmap);if (sym) - name = sym->name; + name =symbol__name (sym);break ;case LOCK_AGGR_CALLER : name = buf;
tools/perf/builtin-mem.c +1 -1
@@ -249,7 +249,7 @@ dump_raw_samples(const struct perf_tool *tool, sample->data_src, symbol_conf.field_sep, dso ?dso__long_name (dso) :"???" , - al.sym ? al.sym->name : "???"); + al.sym ? symbol__name(al.sym) : "???"); out_put:addr_location__exit (&al);return 0 ;
tools/perf/builtin-sched.c +3 -3
@@ -2629,9 +2629,9 @@ static void save_task_callchain(struct perf_sched *sched, sym = node->ms.sym;if (sym) { -if (!strcmp (sym->name,"schedule" ) || - !strcmp (sym->name,"__schedule" ) || - !strcmp (sym->name,"preempt_schedule" )) +if (!strcmp (symbol__name (sym),"schedule" ) || + !strcmp (symbol__name (sym),"__schedule" ) || + !strcmp (symbol__name (sym),"preempt_schedule" ))symbol__set_ignore (sym, true); }
tools/perf/builtin-script.c +7 -7
@@ -1408,7 +1408,7 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,addr_location__init (&al);thread__find_map (thread, cpumode, addr, &al); -if ((*lastsym) && al.addr >= (*lastsym)->start && al.addr < (*lastsym)->end) +if ((*lastsym) && al.addr >=symbol__start ((*lastsym)) && al.addr <symbol__end ((*lastsym))) goto out; al.cpu = cpu;
@@ -1419,11 +1419,11 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,if (!al.sym) goto out; -if (al.addr < al.sym->end) - off = al.addr - al.sym->start; +if (al.addr <symbol__end (al.sym)) + off = al.addr -symbol__start (al.sym);else - off = al.addr -map__start (al.map) - al.sym->start; - printed +=fprintf (fp,"\t%s" , al.sym->name); + off = al.addr -map__start (al.map) -symbol__start (al.sym); + printed +=fprintf (fp,"\t%s" ,symbol__name (al.sym));if (off) printed +=fprintf (fp,"%+d" , off); printed +=fprintf (fp,":" );
@@ -1642,7 +1642,7 @@ static const char *resolve_branch_sym(struct perf_sample *sample,if (!addr_al->thread)thread__resolve (thread, addr_al, sample);if (addr_al->sym) - name = addr_al->sym->name; + name =symbol__name (addr_al->sym);else *ip = sample->addr; }else {
@@ -1650,7 +1650,7 @@ static const char *resolve_branch_sym(struct perf_sample *sample, } }else if (sample->flags & (PERF_IP_FLAG_RETURN |PERF_IP_FLAG_TRACE_END )) {if (al->sym) - name = al->sym->name; + name =symbol__name (al->sym);else *ip = sample->ip; }
tools/perf/builtin-timechart.c +1 -1
@@ -563,7 +563,7 @@ static char *cat_backtrace(struct perf_sample *sample,addr_location__init (&tal); tal.filtered =0 ;if (thread__find_symbol (al.thread, cpumode, ip, &tal)) -fprintf (f,"..... %016" PRIx64" %s\n" , ip, tal.sym->name); +fprintf (f,"..... %016" PRIx64" %s\n" , ip,symbol__name (tal.sym));else fprintf (f,"..... %016" PRIx64"\n" , ip);
tools/perf/builtin-top.c +11 -11
@@ -133,8 +133,8 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he)* We can't annotate with just /proc/kallsyms */ if (dso__symtab_type (dso) ==DSO_BINARY_TYPE__KALLSYMS && !dso__is_kcore (dso)) { -pr_err ("Can't annotate %s: No vmlinux file was found in the " -"path\n" , sym->name); +pr_err ("Can't annotate %s: No vmlinux file was found in the path\n" , +symbol__name (sym));sleep (1 );return -1 ; }
@@ -145,7 +145,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he)if (!symbol__hists (sym,evlist__nr_entries (top->evlist))) {annotation__unlock (notes);pr_err ("Not enough memory for annotating '%s' symbol!\n" , - sym->name); +symbol__name (sym));sleep (1 );return err; }
@@ -156,7 +156,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he) }else {char msg[BUFSIZ ];symbol__strerror_disassemble (&he->ms, err, msg,sizeof (msg)); -pr_err ("Couldn't annotate %s: %s\n" , sym->name, msg); +pr_err ("Couldn't annotate %s: %s\n" ,symbol__name (sym), msg); }annotation__unlock (notes);
@@ -186,9 +186,9 @@ static void ui__warn_map_erange(struct map *map, struct symbol *sym, u64 ip)"Not all samples will be on the annotation output.\n\n" "Please report to [email protected]\n" , ip,dso__long_name (dso),dso__symtab_origin (dso), -map__start (map),map__end (map), sym->start, sym->end, +map__start (map),map__end (map),symbol__start (sym),symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' : -symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , sym->name, +symbol__binding (sym) ==STB_LOCAL ?'l' :'w' ,symbol__name (sym), err ?"[unknown]" : uts.machine, err ? "[unknown]" : uts.release, perf_version_string);if (use_browser <=0 )
@@ -232,7 +232,7 @@ static void perf_top__record_precise_ip(struct perf_top *top,ui__warn_map_erange (he->ms.map, sym, ip); else if (err == -ENOMEM) {pr_err ("Not enough memory for annotating '%s' symbol!\n" , - sym->name); +symbol__name (sym));sleep (1 ); }
@@ -263,7 +263,7 @@ static void perf_top__show_details(struct perf_top *top)if (notes->src ==NULL ) goto out_unlock; -printf ("Showing %s for %s\n" ,evsel__name (top->sym_evsel), symbol->name); +printf ("Showing %s for %s\n" ,evsel__name (top->sym_evsel),symbol__name (symbol));printf (" Events Pcnt (>=%d%%)\n" , annotate_opts.min_pcnt); more =hist_entry__annotate_printf (he, top->sym_evsel);
@@ -418,7 +418,7 @@ static void perf_top__prompt_symbol(struct perf_top *top, const char *msg) next =rb_first_cached (&hists->entries);while (next) { n = rb_entry(next, struct hist_entry, rb_node); -if (n->ms.sym && !strcmp (buf, n->ms.sym->name)) { +if (n->ms.sym && !strcmp (buf,symbol__name (n->ms.sym))) { found = n;break ; }
@@ -437,11 +437,11 @@ static void perf_top__prompt_symbol(struct perf_top *top, const char *msg)static void perf_top__print_mapped_keys (struct perf_top *top) { -char *name =NULL ; +const char *name =NULL ;if (top->sym_filter_entry) {struct symbol *sym = top->sym_filter_entry->ms.sym; - name = sym->name; + name =symbol__name (sym); }fprintf (stdout,"\nMapped keys:\n" );
tools/perf/builtin-trace.c +12 -10
@@ -789,15 +789,16 @@ size_t syscall_arg__scnprintf_ksym(char *bf, size_t size, struct syscall_arg *ar arg->val, &map);if (sym) { -u64 start =map__unmap_ip (map, sym->start); +u64 start =map__unmap_ip (map,symbol__start (sym));u64 offset = arg->val - start;size_t printed; -if (offset ==0 ) - printed =scnprintf (bf, size,"%s" , sym->name); -else +if (offset ==0 ) { + printed =scnprintf (bf, size,"%s" ,symbol__name (sym)); + }else { printed =scnprintf (bf, size,"%s+0x%" PRIx64, - sym->name, offset); +symbol__name (sym), offset); + }map__put (map);return printed; }
@@ -3669,13 +3670,14 @@ static void print_location(FILE *f, struct perf_sample *sample,if ((verbose >0 || print_dso) && al->map)fprintf (f,"%s@" ,dso__long_name (map__dso (al->map))); -if ((verbose >0 || print_sym) && al->sym) -fprintf (f,"%s+0x%" PRIx64, al->sym->name, - al->addr - al->sym->start); -else if (al->map) +if ((verbose >0 || print_sym) && al->sym) { +fprintf (f,"%s+0x%" PRIx64,symbol__name (al->sym), + al->addr -symbol__start (al->sym)); + }else if (al->map) {fprintf (f,"0x%" PRIx64, al->addr); -else + }else {fprintf (f,"0x%" PRIx64, sample->addr); + } } static int trace__pgfault(struct trace *trace,
tools/perf/tests/dlfilter-test.c +2 -2
@@ -274,11 +274,11 @@ static int get_ip_addr(struct test_data *td) sym =map__find_symbol_by_name (map,"foo" );if (sym) - td->foo = sym->start; + td->foo =symbol__start (sym); sym =map__find_symbol_by_name (map,"bar" );if (sym) - td->bar = sym->start; + td->bar =symbol__start (sym);map__put (map);
tools/perf/tests/dwarf-unwind.c +1 -1
@@ -56,7 +56,7 @@ int test__dwarf_unwind(struct test_suite *test, int subtest);static int unwind_entry (struct unwind_entry *entry,void *arg) {unsigned long *cnt = (unsigned long *) arg; -char *symbol = entry->ms.sym ? entry->ms.sym->name :NULL ; +const char *symbol = entry->ms.sym ?symbol__name (entry->ms.sym) :NULL ;static const char *funcs[MAX_STACK ] = {"test__arch_unwind_sample" ,"test_dwarf_unwind__thread" ,
tools/perf/tests/hists_common.c +2 -2
@@ -184,7 +184,7 @@ void print_hists_in(struct hists *hists)pr_info ("%2d: entry: %-8s [%-8s] %20s: period = %" PRIu64"\n" , i,thread__comm_str (he->thread),dso__short_name (dso), - he->ms.sym->name, he->stat.period); +symbol__name (he->ms.sym), he->stat.period); } i++;
@@ -213,7 +213,7 @@ void print_hists_out(struct hists *hists)pr_info ("%2d: entry: %8s:%5d [%-8s] %20s: period = %" PRIu64"/%" PRIu64"\n" , i,thread__comm_str (he->thread),thread__tid (he->thread),dso__short_name (dso), - he->ms.sym->name, he->stat.period, +symbol__name (he->ms.sym), he->stat.period, he->stat_acc ? he->stat_acc->period :0 ); }
tools/perf/tests/hists_cumulate.c +2 -2
@@ -164,11 +164,11 @@ typedef int (*test_fn_t)(struct evsel *, struct machine *);#define COMM (he) (thread__comm_str(he->thread))#define DSO (he) (dso__short_name(map__dso(he->ms.map))) -#define SYM (he) (he->ms.sym->name) +#define SYM (he) (symbol__name(he->ms.sym))#define CPU (he) (he->cpu)#define DEPTH (he) (he->callchain->max_depth)#define CDSO (cl) (dso__short_name(map__dso(cl->ms.map))) -#define CSYM (cl) (cl->ms.sym->name) +#define CSYM (cl) (symbol__name(cl->ms.sym))struct result {u64 children;
tools/perf/tests/hists_output.c +1 -1
@@ -129,7 +129,7 @@ typedef int (*test_fn_t)(struct evsel *, struct machine *);#define COMM (he) (thread__comm_str(he->thread))#define DSO (he) (dso__short_name(map__dso(he->ms.map))) -#define SYM (he) (he->ms.sym->name) +#define SYM (he) (symbol__name(he->ms.sym))#define CPU (he) (he->cpu)#define PID (he) (thread__tid(he->thread))
tools/perf/tests/symbols.c +6 -6
@@ -130,14 +130,14 @@ static int test_dso(struct dso *dso)continue ;/* Check for overlapping function symbols */ -if (last_sym && sym->start < last_sym->end) { +if (last_sym &&symbol__start (sym) <symbol__end (last_sym)) {pr_debug ("Overlapping symbols:\n" );symbol__fprintf (last_sym, stderr);symbol__fprintf (sym, stderr); ret =TEST_FAIL ; }/* Check for zero-length function symbol */ -if (sym->start == sym->end) { +if (symbol__start (sym) ==symbol__end (sym)) {pr_debug ("Zero-length symbol:\n" );symbol__fprintf (sym, stderr); ret =TEST_FAIL ;
@@ -250,14 +250,14 @@ static int check_symbol(struct dso *dso, u64 addr, const char *name, u64 end) {struct symbol *sym =dso__find_symbol_nocache (dso, addr); -if (!sym ||strcmp (sym->name, name)) { +if (!sym ||strcmp (symbol__name (sym), name)) {pr_debug ("%#" PRIx64": expected %s, got %s\n" , addr, name, - sym ? sym->name :"no symbol" ); + sym ?symbol__name (sym) :"no symbol" );return TEST_FAIL ; } -if (sym->end != end) { +if (symbol__end (sym) != end) {pr_debug ("%s: expected end %#" PRIx64", got %#" PRIx64"\n" , - name, end, sym->end); + name, end,symbol__end (sym));return TEST_FAIL ; }return TEST_OK ;
tools/perf/tests/vmlinux-kallsyms.c +19 -15
@@ -292,18 +292,19 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused sym = rb_entry(nd, struct symbol, rb_node); -if (sym->start == sym->end) +if (symbol__start (sym) ==symbol__end (sym))continue ; - mem_start =map__unmap_ip (args.vmlinux_map, sym->start); - mem_end =map__unmap_ip (args.vmlinux_map, sym->end); + mem_start =map__unmap_ip (args.vmlinux_map,symbol__start (sym)); + mem_end =map__unmap_ip (args.vmlinux_map,symbol__end (sym)); first_pair =machine__find_kernel_symbol (&args.kallsyms, mem_start,NULL ); pair = first_pair; -if (pair &&UM (pair->start) == mem_start) { +if (pair &&UM (symbol__start (pair)) == mem_start) { next_pair: -if (arch__compare_symbol_names (sym->name, pair->name) ==0 ) { +if (arch__compare_symbol_names (symbol__name (sym), +symbol__name (pair)) ==0 ) {/* * kallsyms don't have the symbol end, so we * set that by using the next symbol start - 1,
@@ -313,11 +314,13 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused* off the real size. More than that and we * _really_ have a problem. */ -s64 skew = mem_end -UM (pair->end); -if (llabs (skew) >= page_size) +s64 skew = mem_end -UM (symbol__end (pair)); + +if (llabs (skew) >= page_size) {pr_debug ("WARN: %#" PRIx64": diff end addr for %s v: %#" PRIx64" k: %#" PRIx64"\n" , - mem_start, sym->name, mem_end, -UM (pair->end)); + mem_start,symbol__name (sym), mem_end, +UM (symbol__end (pair))); + }/* * Do not count this as a failure, because we
@@ -328,16 +331,17 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unusedcontinue ; }else { pair =machine__find_kernel_symbol_by_name (&args.kallsyms, - sym->name,NULL ); +symbol__name (sym),NULL );if (pair) { -if (UM (pair->start) == mem_start) +if (UM (symbol__start (pair)) == mem_start) goto next_pair;pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , - mem_start, sym->name, pair->name); + mem_start,symbol__name (sym),symbol__name (pair)); }else {pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , - mem_start, sym->name, first_pair->name); + mem_start,symbol__name (sym), +symbol__name (first_pair)); }continue ;
@@ -348,14 +352,14 @@ static int test__vmlinux_matches_kallsyms(struct test_suite *test __maybe_unused* such as __indirect_thunk_end. */ continue ; - }else if (is_ignored_symbol (sym->name,symbol__type (sym))) { + }else if (is_ignored_symbol (symbol__name (sym),symbol__type (sym))) {/* * Ignore hidden symbols, see scripts/kallsyms.c for the details */ continue ; }else {pr_debug ("ERR : %#" PRIx64": %s not on kallsyms\n" , - mem_start, sym->name); + mem_start,symbol__name (sym)); } err =-1 ;
tools/perf/ui/browsers/annotate.c +5 -5
@@ -196,7 +196,7 @@ static void annotate_browser__draw_current_jump(struct ui_browser *browser)int diff =0 ;/* PLT symbols contain external offsets */ -if (strstr (sym->name,"@plt" )) +if (strstr (symbol__name (sym),"@plt" ))return ;if (!disasm_line__is_valid_local_jump (cursor, sym))
@@ -554,7 +554,7 @@ static void ui_browser__init_asm_mode(struct ui_browser *browser)static int sym_title (struct symbol *sym,struct map *map,char *title,size_t sz,int percent_type) { -return snprintf (title, sz,"%s %s [Percent: %s] %s" , sym->name, +return snprintf (title, sz,"%s %s [Percent: %s] %s" ,symbol__name (sym),dso__long_name (map__dso (map)),percent_type_str (percent_type), annotate_opts.code_with_type ?"[Type]" :"" );
@@ -602,7 +602,7 @@ static bool annotate_browser__callq(struct annotate_browser *browser,if (!symbol__hists (dl->ops.target.sym,evlist__nr_entries (evsel->evlist))) {annotation__unlock (notes);ui__warning ("Not enough memory for annotating '%s' symbol!\n" , - dl->ops.target.sym->name); +symbol__name (dl->ops.target.sym));return true; }
@@ -876,7 +876,7 @@ static void annotate_browser__symbol_annotate_error(struct annotate_browser *brodso__set_annotate_warned (dso);symbol__strerror_disassemble (ms, err, msg,sizeof (msg)); -ui__error ("Couldn't annotate %s:\n%s" , sym->name, msg); +ui__error ("Couldn't annotate %s:\n%s" ,symbol__name (sym), msg); } static int annotate_browser__run(struct annotate_browser *browser,
@@ -1255,7 +1255,7 @@ int __hist_entry__tui_annotate(struct hist_entry *he, struct map_symbol *ms,*/ if (al_addr !=NO_ADDR ) {struct annotation_line *al =annotated_source__get_line (notes->src, - al_addr - sym->start); + al_addr -symbol__start (sym)); browser.selection = al; }
tools/perf/ui/browsers/hists.c +5 -5
@@ -2548,7 +2548,7 @@ add_annotate_opt(struct popup_action *act, char **optstr,if (ms->sym ==NULL ||symbol__annotation (ms->sym)->src ==NULL )return 0 ; -if (asprintf (optstr,"Annotate %s" , ms->sym->name) <0 ) +if (asprintf (optstr,"Annotate %s" ,symbol__name (ms->sym)) <0 )return 0 ;map_symbol__copy (&act->ms, ms);
@@ -2748,7 +2748,7 @@ do_run_script(struct hist_browser *browser,if (act->ms.thread) len +=strlen (thread__comm_str (act->ms.thread));else if (act->ms.sym) - len +=strlen (act->ms.sym->name); + len +=strlen (symbol__name (act->ms.sym)); script_opt =malloc (len);if (!script_opt)return -1 ;
@@ -2759,7 +2759,7 @@ do_run_script(struct hist_browser *browser, thread__comm_str(act->ms.thread)); }else if (act->ms.sym) { n =scnprintf (script_opt, len," -S %s " , - act->ms.sym->name); +symbol__name (act->ms.sym)); }if (act->time) {
@@ -2804,7 +2804,7 @@ add_script_opt_2(struct popup_action *act, char **optstr,return 0 ; }else if (sym) {if (asprintf (optstr,"Run scripts for samples of symbol [%s]%s" , - sym->name, tstr) <0 ) +symbol__name (sym), tstr) <0 )return 0 ; }else {if (asprintf (optstr,"Run scripts for all samples%s" , tstr) <0 )
@@ -3208,7 +3208,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *hui_browser__warning (&browser->b, delay_secs *2 ,"No samples for the \"%s\" symbol.\n\n" "Probably appeared just in a callchain" , - browser->selection->sym->name); +symbol__name (browser->selection->sym));continue ; }
tools/perf/ui/browsers/map.c +4 -4
@@ -31,12 +31,12 @@ static void map_browser__write(struct ui_browser *browser, void *nd, int row)ui_browser__set_percent_color (browser,0 , current_entry);ui_browser__printf (browser,"%*" PRIx64" %*" PRIx64" %c " , - mb->addrlen, sym->start, mb->addrlen, sym->end, + mb->addrlen,symbol__start (sym), mb->addrlen,symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' ); width = browser->width - ((mb->addrlen *2 ) +4 );if (width >0 ) -ui_browser__write_nstring (browser, sym->name, width); +ui_browser__write_nstring (browser,symbol__name (sym), width); }/* FIXME uber-kludgy, see comment on cmd_report... */
@@ -122,8 +122,8 @@ int map__browse(struct map *map)for (nd =rb_first (mb.b.entries); nd; nd =rb_next (nd)) { struct symbol *pos = rb_entry(nd, struct symbol, rb_node); -if (maxaddr < pos->end) - maxaddr = pos->end; +if (maxaddr <symbol__end (pos)) + maxaddr =symbol__end (pos);if (verbose >0 ) {u32 *idx =symbol__browser_index (pos); *idx = mb.b.nr_entries;
tools/perf/ui/gtk/annotate.c +3 -3
@@ -67,7 +67,7 @@ static int perf_gtk__get_percent(char *buf, size_t size, struct symbol *sym,static int perf_gtk__get_offset (char *buf,size_t size,struct map_symbol *ms,struct disasm_line *dl) { -u64 start =map__rip_2objdump (ms->map, ms->sym->start); +u64 start =map__rip_2objdump (ms->map,symbol__start (ms->sym));strcpy (buf,"" );
@@ -192,7 +192,7 @@ static int symbol__gtk_annotate(struct map_symbol *ms, struct evsel *evsel,dso__set_annotate_warned (dso);symbol__strerror_disassemble (ms, err, msg,sizeof (msg)); -ui__error ("Couldn't annotate %s: %s\n" , sym->name, msg); +ui__error ("Couldn't annotate %s: %s\n" ,symbol__name (sym), msg);return -1 ; }
@@ -234,7 +234,7 @@ static int symbol__gtk_annotate(struct map_symbol *ms, struct evsel *evsel, } scrolled_window =gtk_scrolled_window_new (); - tab_label =gtk_label_new (sym->name); + tab_label =gtk_label_new (symbol__name (sym)); gtk_scrolled_window_set_policy(GTK_SCROLLED_WINDOW(scrolled_window), GTK_POLICY_AUTOMATIC,
tools/perf/ui/stdio/hist.c +1 -2
@@ -100,13 +100,12 @@ static struct callchain_list rem_hits;static void init_rem_hits (void ) { - rem_sq_bracket =malloc (sizeof (*rem_sq_bracket) +6 ); + rem_sq_bracket =symbol__new (0 ,0 ,0 ,0 ,"[...]" );if (!rem_sq_bracket) {fprintf (stderr,"Not enough memory to display remaining hits\n" );return ; } -strcpy (rem_sq_bracket->name,"[...]" ); rem_hits.ms.sym = rem_sq_bracket; }
tools/perf/util/annotate-arch/annotate-arm64.c +6 -6
@@ -477,12 +477,12 @@ static void update_call_insn_state(struct type_state *state,Dwarf_Die type_die;u32 insn_offset = dl->al.offset; - call_name = func ? func->name : dl->ops.target.name; + call_name = func ?symbol__name (func) : dl->ops.target.name;pr_debug_dtp ("call [%x] %s\n" , insn_offset, call_name ?:"<unknown>" );/* Invalidate caller-saved registers after call */ call_addr =map__rip_2objdump (dloc->ms->map, - dloc->ms->sym->start + dl->al.offset); +symbol__start (dloc->ms->sym) + dl->al.offset);for (unsigned int i =0 ; i <ARRAY_SIZE (state->regs); i++) {struct type_state_reg *reg = &state->regs[i];
@@ -640,7 +640,7 @@ static int propagate_load_reg_state(struct type_state *state, tsr->lifetime_end =0 ;if (dloc->fb_cfa) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 pc =map__rip_2objdump (dloc->ms->map, ip);if (die_get_cfa (dloc->di->dbg, pc, &fbreg, &fboff) <0 )
@@ -727,7 +727,7 @@ static int propagate_load_reg_state(struct type_state *state,/* Or check if it's a global variable */ if (src_tsr->kind ==TSR_KIND_GLOBAL_ADDR ) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 addr = src_tsr->imm_value + reg_offset;int offset;u8 kind;
@@ -875,7 +875,7 @@ static void propagate_store_reg_state(struct type_state *state, reg_offset += mem_spacing;if (dloc->fb_cfa) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 pc =map__rip_2objdump (dloc->ms->map, ip);if (die_get_cfa (dloc->di->dbg, pc, &fbreg, &fboff) <0 )
@@ -1144,7 +1144,7 @@ static void update_mrs_insn_state(struct type_state *state,return ; } - ip = dloc->ms->sym->start + dl->al.offset; + ip =symbol__start (dloc->ms->sym) + dl->al.offset; pc =map__rip_2objdump (dloc->ms->map, ip);/*
tools/perf/util/annotate-arch/annotate-loongarch.c +2 -2
@@ -94,8 +94,8 @@ static int loongarch_jump__parse(const struct arch *arch, struct ins_operands *o ops->target.addr =strtoull (ops->raw,NULL ,16 ); target.addr =map__objdump_2mem (map, ops->target.addr); - start =map__unmap_ip (map, sym->start); - end =map__unmap_ip (map, sym->end); + start =map__unmap_ip (map,symbol__start (sym)); + end =map__unmap_ip (map,symbol__end (sym)); ops->target.outside = target.addr < start || target.addr >= end;
tools/perf/util/annotate-arch/annotate-x86.c +7 -7
@@ -230,7 +230,7 @@ static void update_insn_state_x86(struct type_state *state,/* Try to resolve the call target name */ if (func) - call_name = func->name; + call_name =symbol__name (func);else call_name = dl->ops.target.name;
@@ -245,7 +245,7 @@ static void update_insn_state_x86(struct type_state *state,/* Invalidate caller-saved registers after call */ call_addr =map__rip_2objdump (dloc->ms->map, - dloc->ms->sym->start + dl->al.offset); +symbol__start (dloc->ms->sym) + dl->al.offset);for (unsigned i =0 ; i <ARRAY_SIZE (state->regs); i++) {struct type_state_reg *reg = &state->regs[i];
@@ -277,7 +277,7 @@ static void update_insn_state_x86(struct type_state *state,int offset;const char *var_name =NULL ;struct map_symbol *ms = dloc->ms; -u64 ip = ms->sym->start + dl->al.offset; +u64 ip =symbol__start (ms->sym) + dl->al.offset;if (!has_reg_type (state, dst->reg1))return ;
@@ -478,7 +478,7 @@ static void update_insn_state_x86(struct type_state *state,return ;if (dloc->fb_cfa) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 pc =map__rip_2objdump (dloc->ms->map, ip);if (die_get_cfa (dloc->di->dbg, pc, &fbreg, &fboff) <0 )
@@ -495,7 +495,7 @@ static void update_insn_state_x86(struct type_state *state,if (dso__kernel (map__dso (dloc->ms->map)) && src->segment ==INSN_SEG_X86_GS && src->imm) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 var_addr;int offset;
@@ -646,7 +646,7 @@ static void update_insn_state_x86(struct type_state *state,/* Or check if it's a global variable */ else if (sreg == DWARF_REG_PC) {struct map_symbol *ms = dloc->ms; -u64 ip = ms->sym->start + dl->al.offset; +u64 ip =symbol__start (ms->sym) + dl->al.offset;u64 addr;int offset;
@@ -671,7 +671,7 @@ static void update_insn_state_x86(struct type_state *state,/* And check percpu access with base register */ else if (has_reg_type(state, sreg) && state->regs[sreg].kind == TSR_KIND_PERCPU_BASE) { -u64 ip = dloc->ms->sym->start + dl->al.offset; +u64 ip =symbol__start (dloc->ms->sym) + dl->al.offset;u64 var_addr = src->offset;int offset;
tools/perf/util/annotate-data.c +6 -6
@@ -887,9 +887,9 @@ bool get_global_var_info(struct data_loc_info *dloc, u64 addr, sym =thread__find_symbol_fb (dloc->thread, dloc->cpumode, mem_addr, &al);if (sym) { - *var_name = sym->name; + *var_name =symbol__name (sym);/* Calculate type offset from the start of variable */ - *var_offset = mem_addr -map__unmap_ip (al.map, sym->start); + *var_offset = mem_addr -map__unmap_ip (al.map,symbol__start (sym)); }else { *var_name =NULL ; }
@@ -1657,7 +1657,7 @@ static enum type_match_result find_data_type_insn(struct data_loc_info *dloc, bb->begin->al.offset, bb->end->al.offset);list_for_each_entry_from (dl, ¬es->src->source, al.node) { -u64 this_ip = sym->start + dl->al.offset; +u64 this_ip =symbol__start (sym) + dl->al.offset;u64 addr =map__rip_2objdump (dloc->ms->map, this_ip);/* Skip comment or debug info lines */
@@ -1737,8 +1737,8 @@ static enum type_match_result find_data_type_block(struct data_loc_info *dloc, }pr_debug_dtp ("cannot find a basic block from %" PRIx64" to %" PRIx64"\n" , - src_ip - dloc->ms->sym->start, - dst_ip - dloc->ms->sym->start); + src_ip -symbol__start (dloc->ms->sym), + dst_ip -symbol__start (dloc->ms->sym));continue ; }prepend_basic_blocks (&this_blocks, &basic_blocks);
@@ -1817,7 +1817,7 @@ static int find_data_type_die(struct data_loc_info *dloc, Dwarf_Die *type_die)pr_debug_dtp ("-----------------------------------------------------------\n" );pr_debug_dtp ("find data type for %s%#x(%s) at %s+%#" PRIx64"\n" , offset_sign, offset, buf, - dloc->ms->sym->name, dloc->ip - dloc->ms->sym->start); +symbol__name (dloc->ms->sym), dloc->ip -symbol__start (dloc->ms->sym));/* * IP is a relative instruction address from the start of the map, as
tools/perf/util/annotate.c +37 -32
213 lines not on this page · raw message
tools/perf/util/auxtrace.c +7 -7
@@ -2707,7 +2707,7 @@ static bool dso_sym_match(struct symbol *sym, const char *name, int *cnt, int idx) {/* Same name, and global or the n'th found or any */ -return !arch__compare_symbol_names (name, sym->name) && +return !arch__compare_symbol_names (name,symbol__name (sym)) && ((!idx &&symbol__binding (sym) ==STB_GLOBAL ) || (idx >0 && ++*cnt == idx) || idx <0 );
@@ -2725,14 +2725,14 @@ static void print_duplicate_syms(struct dso *dso, const char *sym_name)while (sym) {if (dso_sym_match (sym, sym_name, &cnt,-1 )) {pr_err ("#%d\t0x%" PRIx64"\t%c\t%s\n" , - ++cnt, sym->start, + ++cnt,symbol__start (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , - sym->name); +symbol__name (sym)); near = true; }else if (near) { near = false; -pr_err ("\t\twhich is near\t\t%s\n" , sym->name); +pr_err ("\t\twhich is near\t\t%s\n" ,symbol__name (sym)); } sym =dso__next_symbol (sym); }
@@ -2755,7 +2755,7 @@ static int find_dso_sym(struct dso *dso, const char *sym_name, u64 *start,while (sym) {if (*start) {if (!*size) - *size = sym->start - *start; + *size =symbol__start (sym) - *start;if (idx >0 ) {if (*size)return 0 ;
@@ -2764,8 +2764,8 @@ static int find_dso_sym(struct dso *dso, const char *sym_name, u64 *start,return -EINVAL ; } }else if (dso_sym_match (sym, sym_name, &cnt, idx)) { - *start = sym->start; - *size = sym->end - sym->start; + *start =symbol__start (sym); + *size =symbol__end (sym) -symbol__start (sym); } sym =dso__next_symbol (sym); }
tools/perf/util/block-info.c +3 -3
@@ -84,7 +84,7 @@ int64_t __block_info__cmp(struct hist_entry *left, struct hist_entry *right)return 1 ; } - cmp =strcmp (bi_l->sym->name, bi_r->sym->name); + cmp =strcmp (symbol__name (bi_l->sym),symbol__name (bi_r->sym));if (cmp)return cmp;
@@ -305,10 +305,10 @@ static int block_range_entry(struct perf_hpp_fmt *fmt, struct perf_hpp *hpp, symbol_conf.addr2line_disable_warn = true; - start_line =map__srcline (he->ms.map, bi->sym->start + bi->start, + start_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->start, he->ms.sym); - end_line =map__srcline (he->ms.map, bi->sym->start + bi->end, + end_line =map__srcline (he->ms.map,symbol__start (bi->sym) + bi->end, he->ms.sym); if (start_line != SRCLINE_UNKNOWN &&
tools/perf/util/bpf_lock_contention.c +8 -8
@@ -143,7 +143,7 @@ static void init_numa_data(struct lock_contention *con)"contig_page_data" , &kmap);if (sym) { - skel->rodata->contig_page_data_addr =map__unmap_ip (kmap, sym->start); + skel->rodata->contig_page_data_addr =map__unmap_ip (kmap,symbol__start (sym));map__put (kmap);return ; }
@@ -159,7 +159,7 @@ static void init_numa_data(struct lock_contention *con)if (sym ==NULL )return ; - skel->rodata->node_data_addr =map__unmap_ip (kmap, sym->start); + skel->rodata->node_data_addr =map__unmap_ip (kmap,symbol__start (sym));map__put (kmap);/* get the number of online nodes using the last node number + 1 */
@@ -266,7 +266,7 @@ int lock_contention_prepare(struct lock_contention *con)continue ; } - addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap, sym->start); + addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); con->filters->addrs = addrs; } naddrs = con->filters->nr_addrs ?: has_mmap_lock;
@@ -288,7 +288,7 @@ int lock_contention_prepare(struct lock_contention *con)continue ; } - con->delays[i].addr =map__unmap_ip (kmap, sym->start); + con->delays[i].addr =map__unmap_ip (kmap,symbol__start (sym)); } skel->rodata->lock_delay =1 ;bpf_map__set_max_entries (skel->maps.lock_delays, con->nr_delays);
@@ -608,7 +608,7 @@ static const char *lock_contention_get_name(struct lock_contention *con,/* global locks with symbols */ sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup, &kmap);if (sym) -return sym->name; +return symbol__name (sym);/* try semi-global locks collected separately */ if (!bpf_map_lookup_elem (lock_fd, &key->lock_addr_or_cgroup, &flags)) {
@@ -652,12 +652,12 @@ static const char *lock_contention_get_name(struct lock_contention *con,if (sym) {unsigned long offset; - offset =map__map_ip (kmap, addr) - sym->start; + offset =map__map_ip (kmap, addr) -symbol__start (sym);if (offset ==0 ) -return sym->name; +return symbol__name (sym); -snprintf (name_buf,sizeof (name_buf),"%s+%#lx" , sym->name, offset); +snprintf (name_buf,sizeof (name_buf),"%s+%#lx" ,symbol__name (sym), offset); }else {snprintf (name_buf,sizeof (name_buf),"%#lx" , (unsigned long )addr); }
tools/perf/util/c2c-function.c +10 -9
@@ -966,7 +966,7 @@ c2c_function_hists__level1_entry(struct symbol *sym,* true "function view" rather than a per-code-address view. */ u64 sym_start = (sym && detail_he->ms.map) ? -map__unmap_ip (detail_he->ms.map, sym->start) : detail_he->ip; +map__unmap_ip (detail_he->ms.map,symbol__start (sym)) : detail_he->ip; mi =mem_info__new ();if (!mi)
@@ -1058,8 +1058,8 @@ c2c_function_hists__level2_entry(struct c2c_hist_entry *level1_c2c,/* Key by the function symbol start so all code addresses collapse. */ level2_c2c =c2c_child_entry__alloc (level1_he, detail_he,1 , (sym && detail_he->ms.map) ? -map__unmap_ip (detail_he->ms.map, sym->start) : -hist_entry__iaddr (detail_he)); +map__unmap_ip (detail_he->ms.map,symbol__start (sym)) : +hist_entry__iaddr (detail_he));if (!level2_c2c)return NULL ;
@@ -1178,13 +1178,14 @@ static void c2c_he__resort_by_stores(struct hist_entry *parent_he)/* Stable tie-break: instruction address, name, then cacheline. */ u64 a =hist_entry__iaddr (he), b =hist_entry__iaddr (iter); -if (a != b) +if (a != b) { cmp = a < b ?-1 :1 ; -else if (he->ms.sym && iter->ms.sym) - cmp =strcmp (he->ms.sym->name, - iter->ms.sym->name); -else + }else if (he->ms.sym && iter->ms.sym) { + cmp =strcmp (symbol__name (he->ms.sym), +symbol__name (iter->ms.sym)); + }else { cmp = (iter->ms.sym ?1 :0 ) - (he->ms.sym ?1 :0 ); + }if (!cmp) {struct c2c_hist_entry *he_c2c;
@@ -1407,7 +1408,7 @@ static int c2c_function__ident_len(struct hist_entry *he)* Match symbol_view_entry(): sort_sym adds the cpumode prefix and, in * verbose mode, the address and DSO origin before the symbol name. */ - size =strlen (he->ms.sym->name) +64 ; + size =strlen (symbol__name (he->ms.sym)) +64 ; symbuf =malloc (size);if (!symbuf)return size -1 ;
tools/perf/util/callchain.c +11 -8
@@ -802,13 +802,15 @@ static enum match_result match_chain(struct callchain_cursor_node *node,* on the symbol start address. */ if (symbol__inlined (cnode->ms.sym) ||symbol__inlined (node->ms.sym)) { - match =match_chain_strings (cnode->ms.sym->name, - node->ms.sym->name); + match =match_chain_strings (symbol__name (cnode->ms.sym), +symbol__name (node->ms.sym));if (match !=MATCH_ERROR )break ; }else { - match =match_chain_dso_addresses (cnode->ms.map, cnode->ms.sym->start, - node->ms.map, node->ms.sym->start); + match =match_chain_dso_addresses (cnode->ms.map, +symbol__start (cnode->ms.sym), + node->ms.map, +symbol__start (node->ms.sym));break ; } }
@@ -1247,13 +1249,14 @@ char *callchain_list__sym_name(struct callchain_list *cl,if (cl->ms.sym) {const char *inlined =symbol__inlined (cl->ms.sym) ?" (inlined)" :"" ; -if (show_srcline && cl->srcline) +if (show_srcline && cl->srcline) { printed =scnprintf (bf, bfsize,"%s %s%s" , - cl->ms.sym->name, cl->srcline, +symbol__name (cl->ms.sym), cl->srcline, inlined); -else + }else { printed =scnprintf (bf, bfsize,"%s%s" , - cl->ms.sym->name, inlined); +symbol__name (cl->ms.sym), inlined); + } }else printed =scnprintf (bf, bfsize,"%#" PRIx64, cl->ip);
tools/perf/util/capstone.c +10 -10
@@ -333,7 +333,7 @@ static struct symbol *find_symbol_from_addr(struct map *map, u64 addr,return NULL ; } - *sym_offset = al_addr - sym->start; + *sym_offset = al_addr -symbol__start (sym);return sym; }
@@ -368,10 +368,10 @@ static void print_capstone_detail(struct cs_insn *insn, char *buf, size_t len,if (sym_offset ==0 ) {scnprintf (buf, len,"\t# %" PRIx64" <%s>" , - orig_addr, sym->name); + orig_addr,symbol__name (sym)); }else {scnprintf (buf, len,"\t# %" PRIx64" <%s+%#" PRIx64">" , - orig_addr, sym->name, sym_offset); + orig_addr,symbol__name (sym), sym_offset); }map__put (found_map);break ;
@@ -454,10 +454,10 @@ static void format_capstone_insn_arm64(struct cs_insn *insn, char *buf, len = len - (last_imm - buf);if (sym_offset ==0 ) {scnprintf (last_imm, len,"%" PRIx64" <%s>" , - addr, sym->name); + addr,symbol__name (sym)); }else {scnprintf (last_imm, len,"%" PRIx64" <%s+%#" PRIx64">" , - addr, sym->name, sym_offset); + addr,symbol__name (sym), sym_offset); }map__put (found_map); }
@@ -498,7 +498,7 @@ int symbol__disassemble_capstone(const char *filename, struct symbol *sym,struct annotation *notes =symbol__annotation (sym);struct map *map = args->ms->map;struct dso *dso =map__dso (map); -u64 start =map__rip_2objdump (map, sym->start); +u64 start =map__rip_2objdump (map,symbol__start (sym));u64 offset;int i, count, free_count;bool is_64bit = false;
@@ -526,7 +526,7 @@ int symbol__disassemble_capstone(const char *filename, struct symbol *sym,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
@@ -631,8 +631,8 @@ int symbol__disassemble_capstone_powerpc(const char *filename __maybe_unused,struct map *map = args->ms->map;struct dso *dso =map__dso (map);struct nscookie nsc; -u64 start =map__rip_2objdump (map, sym->start); -u64 end =map__rip_2objdump (map, sym->end); +u64 start =map__rip_2objdump (map,symbol__start (sym)); +u64 end =map__rip_2objdump (map,symbol__end (sym));u64 len = end - start;u64 offset;int i, fd, count;
@@ -703,7 +703,7 @@ int symbol__disassemble_capstone_powerpc(const char *filename __maybe_unused,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
tools/perf/util/data-convert-json.c +2 -2
@@ -138,11 +138,11 @@ static void output_sample_callchain_entry(const struct perf_tool *tool,output_json_format (out, false,4 ,"{" );output_json_key_format (out, false,5 ,"ip" ,"\"0x%" PRIx64"\"" , ip); -if (al && al->sym && al->sym->namelen) { +if (al && al->sym &&symbol__namelen (al->sym)) {struct dso *dso = al->map ?map__dso (al->map) :NULL ;fputc (',' , out); -output_json_key_string (out, false,5 ,"symbol" , al->sym->name); +output_json_key_string (out, false,5 ,"symbol" ,symbol__name (al->sym));if (dso) {const char *dso_name =dso__short_name (dso);
tools/perf/util/db-export.c +1 -1
@@ -198,7 +198,7 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al,if (err)return err; *sym_db_id = *db_id; - *offset = al->addr - al->sym->start; + *offset = al->addr -symbol__start (al->sym); } }
tools/perf/util/debug.c +2 -1
@@ -342,7 +342,8 @@ void __dump_stack(FILE *file, void **stackdump, size_t stackdump_size)if (thread &&thread__find_map (thread,PERF_RECORD_MISC_USER , addr, &al)) { al.sym =map__find_symbol (al.map, al.addr);if (al.sym) { -fprintf (file," #%zd %p in %s " , i, stackdump[i], al.sym->name); +fprintf (file," #%zd %p in %s " , i, stackdump[i], +symbol__name (al.sym)); printed = true; } }
tools/perf/util/disasm.c +27 -22
119 lines not on this page · raw message
tools/perf/util/dlfilter.c +6 -6
@@ -47,13 +47,13 @@ static void al_to_d_al(struct addr_location *al, struct perf_dlfilter_al *d_al) d_al->buildid =NULL ; }if (sym) { - d_al->sym = sym->name; - d_al->sym_start = sym->start; - d_al->sym_end = sym->end; -if (al->addr < sym->end) - d_al->symoff = al->addr - sym->start; + d_al->sym =symbol__name (sym); + d_al->sym_start =symbol__start (sym); + d_al->sym_end =symbol__end (sym); +if (al->addr <symbol__end (sym)) + d_al->symoff = al->addr -symbol__start (sym);else if (al->map) - d_al->symoff = al->addr -map__start (al->map) - sym->start; + d_al->symoff = al->addr -map__start (al->map) -symbol__start (sym);else d_al->symoff =0 ; d_al->sym_binding =symbol__binding (sym);
tools/perf/util/dso.c +2 -2
@@ -2147,8 +2147,8 @@ const u8 *dso__read_symbol(struct dso *dso, const char *symfs_filename, const struct map *map, const struct symbol *sym, u8 **out_buf, u64 *out_buf_len, bool *is_64bit) { - u64 start = map__rip_2objdump(map, sym->start); -u64 end =map__rip_2objdump (map, sym->end); + u64 start = map__rip_2objdump(map, symbol__start(sym)); +u64 end =map__rip_2objdump (map,symbol__end (sym));size_t len = end - start; *out_buf =NULL ;
tools/perf/util/event.c +2 -2
@@ -884,11 +884,11 @@ int machine__resolve(struct machine *machine, struct addr_location *al,if (al->sym) { ret =strlist__has_entry (symbol_conf.sym_list, - al->sym->name); +symbol__name (al->sym)); }if (!ret && al->sym) {snprintf (al_addr_str, sz,"0x%" PRIx64, -map__unmap_ip (al->map, al->sym->start)); +map__unmap_ip (al->map,symbol__start (al->sym))); ret =strlist__has_entry (symbol_conf.sym_list, al_addr_str); }
tools/perf/util/evsel_fprintf.c +1 -1
@@ -200,7 +200,7 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,/* Add srccode here too? */ if (bt_stop_list && sym && -strlist__has_entry (bt_stop_list, sym->name)) { +strlist__has_entry (bt_stop_list,symbol__name (sym))) {break ; }
tools/perf/util/hist.c +8 -8
@@ -97,7 +97,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h)* +3 accounts for ' y ' symtab origin info */ if (h->ms.sym) { - symlen = h->ms.sym->namelen +4 ; + symlen =symbol__namelen (h->ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL , symlen);
@@ -120,11 +120,11 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (h->parent) -hists__new_col_len (hists,HISTC_PARENT , h->parent->namelen); +hists__new_col_len (hists,HISTC_PARENT ,symbol__namelen (h->parent));if (h->branch_info) {if (h->branch_info->from.ms.sym) { - symlen = (int )h->branch_info->from.ms.sym->namelen +4 ; + symlen = (int )symbol__namelen (h->branch_info->from.ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL_FROM , symlen);
@@ -139,7 +139,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (h->branch_info->to.ms.sym) { - symlen = (int )h->branch_info->to.ms.sym->namelen +4 ; + symlen = (int )symbol__namelen (h->branch_info->to.ms.sym) +4 ;if (verbose >0 ) symlen +=BITS_PER_LONG /4 +2 +3 ;hists__new_col_len (hists,HISTC_SYMBOL_TO , symlen);
@@ -163,7 +163,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h)if (h->mem_info) {if (mem_info__daddr (h->mem_info)->ms.sym) { - symlen = (int )mem_info__daddr (h->mem_info)->ms.sym->namelen +4 + symlen = (int )symbol__namelen (mem_info__daddr (h->mem_info)->ms.sym) +4 + unresolved_col_width +2 ;hists__new_col_len (hists,HISTC_MEM_DADDR_SYMBOL , symlen);
@@ -178,7 +178,7 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h) }if (mem_info__iaddr (h->mem_info)->ms.sym) { - symlen = (int )mem_info__iaddr (h->mem_info)->ms.sym->namelen +4 + symlen = (int )symbol__namelen (mem_info__iaddr (h->mem_info)->ms.sym) +4 + unresolved_col_width +2 ;hists__new_col_len (hists,HISTC_MEM_IADDR_SYMBOL , symlen);
@@ -727,7 +727,7 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,*/ if (hists__has (hists, sym) && he->ms.map != entry->ms.map) {if (he->ms.sym) { -u64 addr = he->ms.sym->start; +u64 addr =symbol__start (he->ms.sym); he->ms.sym =map__find_symbol (entry->ms.map, addr); }
@@ -2300,7 +2300,7 @@ static bool hists__filter_entry_by_symbol(struct hists *hists, struct hist_entry *he) {if (hists->symbol_filter_str !=NULL && - (!he->ms.sym ||strstr (he->ms.sym->name, + (!he->ms.sym ||strstr (symbol__name (he->ms.sym), hists->symbol_filter_str) ==NULL )) { he->filtered |= (1 <<HIST_FILTER__SYMBOL );return true;
tools/perf/util/intel-pt.c +4 -4
@@ -3003,8 +3003,8 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip)for (sym = start; sym; sym =dso__next_symbol (sym)) {if (symbol__binding (sym) ==STB_GLOBAL && - !strcmp (sym->name,"__switch_to" )) { - ip =map__unmap_ip (map, sym->start); + !strcmp (symbol__name (sym),"__switch_to" )) { + ip =map__unmap_ip (map,symbol__start (sym));if (ip >=map__start (map) && ip <map__end (map)) { switch_ip = ip;break ;
@@ -3021,8 +3021,8 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip) ptss ="__perf_event_task_sched_out" ;for (sym = start; sym; sym =dso__next_symbol (sym)) { -if (!strcmp (sym->name, ptss)) { - ip =map__unmap_ip (map, sym->start); +if (!strcmp (symbol__name (sym), ptss)) { + ip =map__unmap_ip (map,symbol__start (sym));if (ip >=map__start (map) && ip <map__end (map)) { *ptss_ip = ip;break ;
tools/perf/util/libbfd.c +2 -2
@@ -552,7 +552,7 @@ int symbol__disassemble_bpf_libbfd(struct symbol *sym __maybe_unused,struct annotation *notes =symbol__annotation (sym);struct bpf_prog_linfo *prog_linfo =NULL ;struct bpf_prog_info_node *info_node; -int len = sym->end - sym->start; +int len =symbol__end (sym) -symbol__start (sym);disassembler_ftype disassemble;struct map *map = args->ms->map;struct perf_bpil *info_linear;
@@ -572,7 +572,7 @@ int symbol__disassemble_bpf_libbfd(struct symbol *sym __maybe_unused,return SYMBOL_ANNOTATE_ERRNO__BPF_INVALID_FILE ;pr_debug ("%s: handling sym %s addr %" PRIx64" len %" PRIx64"\n" , __func__, - sym->name, sym->start, sym->end - sym->start); +symbol__name (sym),symbol__start (sym),symbol__end (sym) -symbol__start (sym));memset (tpath,0 ,sizeof (tpath));perf_exe (tpath,sizeof (tpath));
tools/perf/util/llvm.c +3 -3
@@ -138,7 +138,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,struct annotation *notes =symbol__annotation (sym);struct map *map = args->ms->map;struct dso *dso =map__dso (map); -u64 start =map__rip_2objdump (map, sym->start); +u64 start =map__rip_2objdump (map,symbol__start (sym));/* Malloc-ed buffer containing instructions read from disk. */ u8 *code_buf =NULL ;/* Pointer to code to be disassembled. */
@@ -202,7 +202,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,/* add the function address and name */ scnprintf (disasm_buf,sizeof (disasm_buf),"%#" PRIx64" <%s>:" , - start, sym->name); + start,symbol__name (sym)); args->offset =-1 ; args->line = disasm_buf;
@@ -323,7 +323,7 @@ int symbol__disassemble_llvm(const char *filename, struct symbol *sym,return ret;#else // HAVE_LIBLLVM_SUPPORT pr_debug ("The LLVM disassembler isn't linked in for %s in %s\n" , - sym->name, filename); +symbol__name (sym), filename);return -1 ;#endif }
tools/perf/util/lock-contention.c +1 -1
@@ -135,7 +135,7 @@ bool match_callstack_filter(struct machine *machine, u64 *callstack, int max_stacontinue ;list_for_each_entry (filter, &callstack_filters, list) { -if (strstr (sym->name, filter->name)) +if (strstr (symbol__name (sym), filter->name))return true; } }
tools/perf/util/machine.c +13 -13
@@ -1110,8 +1110,8 @@ static u64 find_entry_trampoline(struct dso *dso)if (symbol__binding (sym) !=STB_GLOBAL )continue ;for (i =0 ; i <ARRAY_SIZE (syms); i++) { -if (!strcmp (sym->name, syms[i])) -return sym->start; +if (!strcmp (symbol__name (sym), syms[i])) +return symbol__start (sym); } }
@@ -2056,7 +2056,7 @@ int machine__process_event(struct machine *machine, union perf_event *event,static bool symbol__match_regex (struct symbol *sym,regex_t *regex) { -return regexec (regex, sym->name,0 ,NULL ,0 ) ==0 ; +return regexec (regex,symbol__name (sym),0 ,NULL ,0 ) ==0 ; } static void ip__resolve_ams(struct thread *thread,
@@ -3283,8 +3283,8 @@ char *machine__resolve_kernel_addr(void *vmachine, unsigned long long *addrp, chreturn NULL ; *modp =__map__is_kmodule (map) ? (char *)dso__short_name (map__dso (map)) :NULL ; - *addrp =map__unmap_ip (map, sym->start); -return sym->name; + *addrp =map__unmap_ip (map,symbol__start (sym)); +return (char *)symbol__name (sym); }struct machine__for_each_dso_cb_args {
@@ -3330,27 +3330,27 @@ bool machine__is_lock_function(struct machine *machine, u64 addr)return false; } - machine->sched.text_start =map__unmap_ip (kmap, sym->start); + machine->sched.text_start =map__unmap_ip (kmap,symbol__start (sym));/* should not fail from here */ sym =machine__find_kernel_symbol_by_name (machine,"__sched_text_end" , &kmap); - machine->sched.text_end =map__unmap_ip (kmap, sym->start); + machine->sched.text_end =map__unmap_ip (kmap,symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__lock_text_start" , &kmap); - machine->lock.text_start =map__unmap_ip (kmap, sym->start); + machine->lock.text_start =map__unmap_ip (kmap,symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__lock_text_end" , &kmap); - machine->lock.text_end =map__unmap_ip (kmap, sym->start); + machine->lock.text_end =map__unmap_ip (kmap,symbol__start (sym)); sym =machine__find_kernel_symbol_by_name (machine,"__traceiter_contention_begin" , &kmap);if (sym) { - machine->traceiter.text_start =map__unmap_ip (kmap, sym->start); - machine->traceiter.text_end =map__unmap_ip (kmap, sym->end); + machine->traceiter.text_start =map__unmap_ip (kmap,symbol__start (sym)); + machine->traceiter.text_end =map__unmap_ip (kmap,symbol__end (sym)); } sym =machine__find_kernel_symbol_by_name (machine,"trace_contention_begin" , &kmap);if (sym) { - machine->trace.text_start =map__unmap_ip (kmap, sym->start); - machine->trace.text_end =map__unmap_ip (kmap, sym->end); + machine->trace.text_start =map__unmap_ip (kmap,symbol__start (sym)); + machine->trace.text_end =map__unmap_ip (kmap,symbol__end (sym)); } }
tools/perf/util/map.c +3 -3
@@ -321,7 +321,7 @@ void map__fixup_start(struct map *map)if (nd !=NULL ) { struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_start (map, sym->start); +map__set_start (map,symbol__start (sym)); } }
@@ -333,7 +333,7 @@ void map__fixup_end(struct map *map)if (nd !=NULL ) { struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_end (map, sym->end); +map__set_end (map,symbol__end (sym)); } }
@@ -606,7 +606,7 @@ u64 map__objdump_2rip(const struct map *map, u64 ip)bool map__contains_symbol (const struct map *map,const struct symbol *sym) { -u64 ip =map__unmap_ip (map, sym->start); +u64 ip =map__unmap_ip (map,symbol__start (sym));return ip >=map__start (map) && ip <map__end (map); }
tools/perf/util/map.h +1 -1
@@ -164,7 +164,7 @@ struct thread;#define __map__for_each_symbol_by_name (map, sym_name, pos, idx) \ for (pos = map__find_symbol_by_name_idx(map, sym_name, &idx); \ pos && \ - !symbol__match_symbol_name(pos->name, sym_name, \ + !symbol__match_symbol_name(symbol__name(pos), sym_name, \ SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); \ pos = dso__next_symbol_by_name(map__dso(map), &idx))
tools/perf/util/probe-event.c +21 -20
121 lines not on this page · raw message
tools/perf/util/python.c +9 -9
@@ -812,7 +812,7 @@ static PyObject *pyrf_sample_event__get_symbol(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyUnicode_FromString (pevent->al.sym->name); +return PyUnicode_FromString (symbol__name (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_sym_start(struct pyrf_event *pevent,
@@ -821,7 +821,7 @@ static PyObject *pyrf_sample_event__get_sym_start(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyLong_FromUnsignedLongLong (pevent->al.sym->start); +return PyLong_FromUnsignedLongLong (symbol__start (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_sym_offset(struct pyrf_event *pevent,
@@ -830,7 +830,7 @@ static PyObject *pyrf_sample_event__get_sym_offset(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyLong_FromUnsignedLongLong (pevent->al.addr - pevent->al.sym->start); +return PyLong_FromUnsignedLongLong (pevent->al.addr -symbol__start (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_sym_end(struct pyrf_event *pevent,
@@ -839,7 +839,7 @@ static PyObject *pyrf_sample_event__get_sym_end(struct pyrf_event *pevent,if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) Py_RETURN_NONE; -return PyLong_FromUnsignedLongLong (pevent->al.sym->end); +return PyLong_FromUnsignedLongLong (symbol__end (pevent->al.sym)); } static PyObject *pyrf_sample_event__get_raw_buf(struct pyrf_event *pevent,
@@ -960,7 +960,7 @@ static PyObject *pyrf_callchain_node__get_symbol(struct pyrf_callchain_node *pnoif (pnode->sym_name)return PyUnicode_FromString (pnode->sym_name);if (pnode->sym) -return PyUnicode_FromString (pnode->sym->name); +return PyUnicode_FromString (symbol__name (pnode->sym));return PyUnicode_FromString ("[unknown]" ); }
@@ -991,7 +991,7 @@ static PyObject *pyrf_callchain_node__get_sym_offset(struct pyrf_callchain_node Py_RETURN_NONE; addr = pnode->map ?map__map_ip (pnode->map, pnode->ip) : pnode->ip; -return PyLong_FromUnsignedLongLong (addr - pnode->sym->start); +return PyLong_FromUnsignedLongLong (addr -symbol__start (pnode->sym)); }static PyGetSetDef pyrf_callchain_node__getset[] = {
@@ -1330,7 +1330,7 @@ static PyObject *pyrf_sample_event__get_addr_symbol(struct pyrf_event *pevent,PyObject *ret = Py_None;if (pyrf_sample_event__resolve_addr_al (pevent, &addr_al) ==0 && addr_al.sym) - ret =PyUnicode_FromString (addr_al.sym->name); + ret =PyUnicode_FromString (symbol__name (addr_al.sym));else Py_INCREF (Py_None);
@@ -1345,7 +1345,7 @@ static PyObject *pyrf_sample_event__get_addr_sym_offset(struct pyrf_event *pevenPyObject *ret = Py_None;if (pyrf_sample_event__resolve_addr_al (pevent, &addr_al) ==0 && addr_al.sym) - ret =PyLong_FromUnsignedLongLong (addr_al.addr - addr_al.sym->start); + ret =PyLong_FromUnsignedLongLong (addr_al.addr -symbol__start (addr_al.sym));else Py_INCREF (Py_None);
@@ -4505,7 +4505,7 @@ static PyObject *pyrf_call_path__to_callchain(const struct call_path *cp, frame->ip = pos->ip; frame->sym =NULL ; - frame->sym_name = pos->sym ?strdup (pos->sym->name) :NULL ; + frame->sym_name = pos->sym ?strdup (symbol__name (pos->sym)) :NULL ; frame->map =NULL ;if (thread) {struct addr_location al;
tools/perf/util/sort.c +21 -19
@@ -470,18 +470,19 @@ int64_t _sort__sym_cmp(struct symbol *sym_l, struct symbol *sym_r)return 0 ;if (symbol__inlined (sym_l) ||symbol__inlined (sym_r)) { -int ret =strcmp (sym_l->name, sym_r->name); +int ret =strcmp (symbol__name (sym_l),symbol__name (sym_r));if (ret)return ret; -if ((sym_l->start <= sym_r->end) && (sym_l->end >= sym_r->start)) +if ((symbol__start (sym_l) <=symbol__end (sym_r)) && + (symbol__end (sym_l) >=symbol__start (sym_r)))return 0 ; } -if (sym_l->start != sym_r->start) -return (int64_t )(sym_r->start - sym_l->start); +if (symbol__start (sym_l) !=symbol__start (sym_r)) +return (int64_t )(symbol__start (sym_r) -symbol__start (sym_l)); -return (int64_t )(sym_r->end - sym_l->end); +return (int64_t )(symbol__end (sym_r) -symbol__end (sym_l)); } static int64_t
@@ -511,7 +512,7 @@ sort__sym_sort(struct hist_entry *left, struct hist_entry *right)if (!left->ms.sym || !right->ms.sym)return cmp_null (left->ms.sym, right->ms.sym); -return strcmp (right->ms.sym->name, left->ms.sym->name); +return strcmp (symbol__name (right->ms.sym),symbol__name (left->ms.sym)); } static int _hist_entry__sym_snprintf(struct map_symbol *ms,
@@ -537,13 +538,13 @@ static int _hist_entry__sym_snprintf(struct map_symbol *ms, ret +=repsep_snprintf (bf + ret, size - ret,"[%c] " , level);if (sym && map) {if (symbol__type (sym) ==STT_OBJECT ) { - ret +=repsep_snprintf (bf + ret, size - ret,"%s" , sym->name); + ret +=repsep_snprintf (bf + ret, size - ret,"%s" ,symbol__name (sym)); ret +=repsep_snprintf (bf + ret, size - ret,"+0x%llx" , - ip -map__unmap_ip (map, sym->start)); + ip -map__unmap_ip (map,symbol__start (sym))); }else { ret +=repsep_snprintf (bf + ret, size - ret,"%.*s" , width - ret, - sym->name); +symbol__name (sym));if (symbol__inlined (sym)) ret +=repsep_snprintf (bf + ret, size - ret," (inlined)" );
@@ -570,7 +571,7 @@ static int hist_entry__sym_filter(struct hist_entry *he, int type, const void *aif (type !=HIST_FILTER__SYMBOL )return -1 ; -return sym && (!he->ms.sym || !strstr (he->ms.sym->name, sym)); +return sym && (!he->ms.sym || !strstr (symbol__name (he->ms.sym), sym)); }struct sort_entry sort_sym = {
@@ -616,7 +617,8 @@ hist_entry__symoff_snprintf(struct hist_entry *he, char *bf, size_t size, unsignif (sym ==NULL )return repsep_snprintf (bf, size,"[%c] %-#.*llx" , he->level, width -4 , he->ip); -return repsep_snprintf (bf, size,"[%c] %s+0x%llx" , he->level, sym->name, he->ip - sym->start); +return repsep_snprintf (bf, size,"[%c] %s+0x%llx" , he->level, +symbol__name (sym), he->ip -symbol__start (sym)); }static struct sort_entry sort_sym_offset = {
@@ -1016,14 +1018,14 @@ sort__parent_cmp(struct hist_entry *left, struct hist_entry *right)if (!sym_l || !sym_r)return cmp_null (sym_l, sym_r); -return strcmp (sym_r->name, sym_l->name); +return strcmp (symbol__name (sym_r),symbol__name (sym_l)); }static int hist_entry__parent_snprintf (struct hist_entry *he,char *bf,size_t size,unsigned int width) {return repsep_snprintf (bf, size,"%-*.*s" , width, width, - he->parent ? he->parent->name :"[other]" ); + he->parent ?symbol__name (he->parent) :"[other]" ); }struct sort_entry sort_parent = {
@@ -1426,7 +1428,7 @@ static int hist_entry__sym_from_filter(struct hist_entry *he, int type,return -1 ;return sym && !(he->branch_info && he->branch_info->from.ms.sym && -strstr (he->branch_info->from.ms.sym->name, sym)); +strstr (symbol__name (he->branch_info->from.ms.sym), sym)); } static int hist_entry__sym_to_filter(struct hist_entry *he, int type,
@@ -1438,7 +1440,7 @@ static int hist_entry__sym_to_filter(struct hist_entry *he, int type,return -1 ;return sym && !(he->branch_info && he->branch_info->to.ms.sym && -strstr (he->branch_info->to.ms.sym->name, sym)); +strstr (symbol__name (he->branch_info->to.ms.sym), sym)); }struct sort_entry sort_dso_from = {
@@ -1484,14 +1486,14 @@ static int _hist_entry__addr_snprintf(struct map_symbol *ms, ret +=repsep_snprintf (bf + ret, size - ret,"[%c] " , level);if (sym && map) {if (symbol__type (sym) ==STT_OBJECT ) { - ret +=repsep_snprintf (bf + ret, size - ret,"%s" , sym->name); + ret +=repsep_snprintf (bf + ret, size - ret,"%s" ,symbol__name (sym)); ret +=repsep_snprintf (bf + ret, size - ret,"+0x%llx" , - ip -map__unmap_ip (map, sym->start)); + ip -map__unmap_ip (map,symbol__start (sym))); }else { ret +=repsep_snprintf (bf + ret, size - ret,"%.*s" , width - ret, - sym->name); - offs = ip - sym->start; +symbol__name (sym)); + offs = ip -symbol__start (sym);if (offs) ret +=repsep_snprintf (bf + ret, size - ret,"+0x%llx" , offs); }
tools/perf/util/srcline.c +9 -7
@@ -116,7 +116,7 @@ struct symbol *new_inline_sym(struct dso *dso, funcname = demangled; } -if (base_sym &&strcmp (funcname, base_sym->name) ==0 ) { +if (base_sym &&strcmp (funcname,symbol__name (base_sym)) ==0 ) {/* reuse the real, existing symbol */ inline_sym = base_sym;/* ensure that we don't alias an inlined symbol, which could
@@ -125,8 +125,9 @@ struct symbol *new_inline_sym(struct dso *dso,assert (!symbol__inlined (base_sym)); }else {/* create a fake symbol for the inline frame */ - inline_sym =symbol__new (base_sym ? base_sym->start :0 , - base_sym ? (base_sym->end - base_sym->start) :0 , + inline_sym =symbol__new (base_sym ?symbol__start (base_sym) :0 , + base_sym ? + (symbol__end (base_sym) -symbol__start (base_sym)) :0 , base_sym ?symbol__binding (base_sym) :0 , base_sym ?symbol__type (base_sym) :0 , funcname);
@@ -310,13 +311,14 @@ char *__get_srcline(struct dso *dso, u64 addr, struct symbol *sym, }mutex_unlock (dso__lock (dso)); out: -if (!show_addr) +if (!show_addr) {return (show_sym && sym) ? -strndup (sym->name, sym->namelen) :SRCLINE_UNKNOWN ; +strndup (symbol__name (sym),symbol__namelen (sym)) :SRCLINE_UNKNOWN ; + }if (sym) { -if (asprintf (&srcline,"%s+%" PRIu64, show_sym ? sym->name :"" , - ip - sym->start) <0 ) +if (asprintf (&srcline,"%s+%" PRIu64, show_sym ?symbol__name (sym) :"" , + ip -symbol__start (sym)) <0 )return SRCLINE_UNKNOWN ; }else if (asprintf (&srcline,"%s[%" PRIx64"]" ,dso__short_name (dso), addr) <0 )return SRCLINE_UNKNOWN ;
tools/perf/util/symbol-elf.c +5 -5
@@ -351,11 +351,11 @@ static bool get_ifunc_name(Elf *elf, struct dso *dso, GElf_Ehdr *ehdr, sym =dso__find_symbol_nocache (dso, addr);/* Expecting the address to be an IFUNC or IFUNC alias */ -if (!sym || sym->start != addr || +if (!sym ||symbol__start (sym) != addr || (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym)))return false; -snprintf (buf, buf_sz,"%s@plt" , sym->name); +snprintf (buf, buf_sz,"%s@plt" ,symbol__name (sym));return true; }
@@ -630,8 +630,8 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss)* a symbol for .plt header. */ f =dso__find_symbol_nocache (dso, shdr_plt.sh_offset); -if (f && f->start < shdr_plt.sh_offset && f->end > shdr_plt.sh_offset) - f->end = shdr_plt.sh_offset; +if (f &&symbol__start (f) < shdr_plt.sh_offset &&symbol__end (f) > shdr_plt.sh_offset) +symbol__set_end (f, shdr_plt.sh_offset);if (!get_plt_sizes (dso, &ehdr, &shdr_plt, &plt_header_size, &plt_entry_size))return 0 ;
@@ -653,7 +653,7 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss)if (!get_plt_sizes (dso, &ehdr, &plt_sec_shdr, &plt_header_size, &plt_entry_size))return 0 ;/* Extend .plt symbol to entire .plt */ - plt_sym->end = plt_sym->start + shdr_plt.sh_size; +symbol__set_end (plt_sym,symbol__start (plt_sym) + shdr_plt.sh_size);/* Use .plt.sec offset */ plt_offset = plt_sec_shdr.sh_offset; lazy_plt = false;
tools/perf/util/symbol.c +53 -50
257 lines not on this page · raw message
tools/perf/util/symbol.h +42 -1
55 lines not on this page · raw message
tools/perf/util/symbol_fprintf.c +7 -7
@@ -10,10 +10,10 @@size_t symbol__fprintf (struct symbol *sym,FILE *fp) {return fprintf (fp," %" PRIx64"-%" PRIx64" %c %s\n" , - sym->start, sym->end, +symbol__start (sym),symbol__end (sym),symbol__binding (sym) ==STB_GLOBAL ?'g' :symbol__binding (sym) ==STB_LOCAL ?'l' :'w' , - sym->name); +symbol__name (sym)); } size_t __symbol__fprintf_symname_offs(const struct symbol *sym,
@@ -25,12 +25,12 @@ size_t __symbol__fprintf_symname_offs(const struct symbol *sym,size_t length;if (sym) { - length =fprintf (fp,"%s" , sym->name); + length =fprintf (fp,"%s" ,symbol__name (sym));if (al && print_offsets) { -if (al->addr < sym->end) - offset = al->addr - sym->start; +if (al->addr <symbol__end (sym)) + offset = al->addr -symbol__start (sym);else - offset = al->addr -map__start (al->map) - sym->start; + offset = al->addr -map__start (al->map) -symbol__start (sym); length +=fprintf (fp,"+0x%lx" , offset); }return length;
@@ -67,7 +67,7 @@ size_t dso__fprintf_symbols_by_name(struct dso *dso,for (size_t i =0 ; i <dso__symbol_names_len (dso); i++) {struct symbol *pos =dso__symbol_names (dso)[i]; - ret +=fprintf (fp,"%s\n" , pos->name); + ret +=fprintf (fp,"%s\n" ,symbol__name (pos)); }return ret; }
tools/perf/util/thread-stack.c +4 -4
@@ -1056,7 +1056,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,struct symbol *tsym = to_al->sym;struct call_path *cp; -if (sym &&is_x86_retpoline (sym->name)) { +if (sym &&is_x86_retpoline (symbol__name (sym))) {/* * This is a x86 retpoline fn. It pollutes the call graph by * showing up everywhere there is an indirect branch, but does
@@ -1069,7 +1069,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,*/ ts->cnt -=1 ; sym = ts->stack[ts->cnt -2 ].cp->sym; -if (sym && sym == tsym && to_al->addr != tsym->start) { +if (sym && sym == tsym && to_al->addr !=symbol__start (tsym)) {/* * Target is back to the middle of the symbol we came * from so assume it is an indirect jmp and forget it
@@ -1170,7 +1170,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm,*/ if (!err && rstate ==X86_RETPOLINE_POSSIBLE && to_al->sym && from_al->sym == to_al->sym && - to_al->addr != to_al->sym->start) + to_al->addr !=symbol__start (to_al->sym)) ts->rstate =X86_RETPOLINE_DETECTED ; }else if (sample->flags &PERF_IP_FLAG_RETURN ) {
@@ -1207,7 +1207,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm, err =thread_stack__trace_end (ts, sample, ref); }else if (sample->flags &PERF_IP_FLAG_BRANCH && from_al->sym != to_al->sym && to_al->sym && - to_al->addr == to_al->sym->start) { + to_al->addr ==symbol__start (to_al->sym)) {struct call_path_root *cpr = ts->crp->cpr;struct call_path *cp;
tools/perf/util/unwind-libdw.c +1 -1
@@ -172,7 +172,7 @@ static int entry(u64 ip, struct unwind_info *ui) e->ms.sym = al.sym;pr_debug ("unwind: %s:ip = 0x%" PRIx64" (0x%" PRIx64")\n" , - al.sym ? al.sym->name :"''" , + al.sym ?symbol__name (al.sym) :"''" , ip, al.map ?map__map_ip (al.map, ip) : (u64 )0 );addr_location__exit (&al);
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 03/18] perf symbol: Add accessor functions for struct symbol fields
Quoted 4 lines · Ian Rogers
static void
init_rem_hits(void)
{
-rem_sq_bracket=malloc(sizeof(*rem_sq_bracket) + 6);
+rem_sq_bracket=symbol__new(0, 0, 0, 0, "[...]");
Doesn't this make hists__fprintf() pass an interior pointer to free()?
At this commit symbol__new() still returns calloc() + symbol_conf.priv_size, but hists__fprintf() still frees rem_sq_bracket with zfree(). Once symbol__annotation_init() has run, priv_size is not zero and free() gets a pointer into the middle of the block.
Should we use symbol__delete(), which subtracts priv_size before freeing?
Ian Rogers
reviewerPATCH v2 4/18 perf map_browser: Compute symbol browser index on demand
In reply to Ian RogersTo wrap struct symbol in DECLARE_RC_STRUCT(symbol) and reduce per-symbol memory overhead, negative-offset allocations ahead of struct symbol via symbol_conf.priv_size and symbol__priv() need to be eliminated.
Remove symbol__browser_index() in ui/browsers/map.c and compute the TUI map browser index on demand in map_browser__search() by walking the map browser's symbol entries when the user searches with '/'.
While this changes finding a symbol's display row index in map_browser__search() from O(1) to O(N) in the number of symbols in that single map, map_browser__search() is only invoked interactively when a user opens the TUI "Browse map details" view for a map, presses '/', types a query, and hits ENTER. Even for a large DSO such as vmlinux (~100k symbols), scanning the map's symbols once per interactive search takes well under a millisecond (and becomes a flat pointer array scan once symbols are stored in struct symbols).
Previously, '/' search in the TUI map browser was gated behind 'verbose > 0' (-v) in builtin-c2c.c, builtin-report.c, and ui/browsers/map.c solely to avoid adding sizeof(u32) to symbol_conf.priv_size for every struct symbol across all DSOs in normal runs. With on-demand index computation there is no per-symbol memory overhead, so remove the verbose > 0 restriction and allow '/' search in the map browser unconditionally.
--- tools/perf/builtin-c2c.c | 14 -------------- tools/perf/builtin-report.c | 14 -------------- tools/perf/ui/browsers/map.c | 37 ++++++++++++++++++------------------ 3 files changed, 18 insertions(+), 47 deletions(-)
tools/perf/builtin-c2c.c +0 -14
@@ -3412,20 +3412,6 @@ static int perf_c2c__report(int argc, const char **argv)if (ret <0 ) goto out_mem2node; -/* -* For searching by name on the "Browse map details". -* providing it only in verbose mode not to bloat too -* much struct symbol. -*/ -if (verbose >0 ) { -/* -* XXX: Need to provide a less kludgy way to ask for -* more space per symbol, the u32 is for the index on -* the ui browser. -* See symbol__browser_index. -*/ - symbol_conf.priv_size +=sizeof (u32); - }annotation_config__init (); }
tools/perf/builtin-report.c +0 -14
@@ -1885,20 +1885,6 @@ int cmd_report(int argc, const char **argv) ret =symbol__annotation_init ();if (ret <0 ) goto error; -/* -* For searching by name on the "Browse map details". -* providing it only in verbose mode not to bloat too -* much struct symbol. -*/ -if (verbose >0 ) { -/* -* XXX: Need to provide a less kludgy way to ask for -* more space per symbol, the u32 is for the index on -* the ui browser. -* See symbol__browser_index. -*/ - symbol_conf.priv_size +=sizeof (u32); - }annotation_config__init (); }
tools/perf/ui/browsers/map.c +18 -19
@@ -39,12 +39,6 @@ static void map_browser__write(struct ui_browser *browser, void *nd, int row)ui_browser__write_nstring (browser,symbol__name (sym), width); } -/* FIXME uber-kludgy, see comment on cmd_report... */ -static u32 *symbol__browser_index (struct symbol *browser) -{ -return ((void *)browser) -sizeof (struct rb_node ) -sizeof (u32); -} -static int map_browser__search (struct map_browser *browser) {char target[512 ];
@@ -62,10 +56,21 @@ static int map_browser__search(struct map_browser *browser) sym =map__find_symbol_by_name (browser->map, target);if (sym !=NULL ) { -u32 *idx =symbol__browser_index (sym); - - browser->b.top = &sym->rb_node; - browser->b.index = browser->b.top_idx = *idx; +struct rb_node *nd; +u32 idx =0 ; + +/* +* Walk the map browser's symbol entries to find the matching +* symbol node and its display row index, then position the +* browser cursor at that entry. +*/ +for (nd =rb_first (browser->b.entries); nd; nd =rb_next (nd), ++idx) { +if (&sym->rb_node == nd) { + browser->b.top = nd; + browser->b.index = browser->b.top_idx = idx; +break ; + } + } }else ui_helpline__fpush ("%s not found!" , target);
@@ -77,8 +82,7 @@ static int map_browser__run(struct map_browser *browser)int key;if (ui_browser__show (&browser->b,dso__long_name (map__dso (browser->map)), -"Press ESC to exit, %s / to search" , - verbose >0 ?"" :"restart with -v to use" ) <0 ) +"Press ESC to exit, / to search" ) <0 )return -1 ;while (1 ) {
@@ -86,9 +90,8 @@ static int map_browser__run(struct map_browser *browser)switch (key) {case '/' : -if (verbose >0 ) -map_browser__search (browser); -/* fall thru */ +map_browser__search (browser); +continue ;default :ui_browser__warn_unhandled_hotkey (&browser->b, key,0 ,NULL );continue ;
@@ -124,10 +127,6 @@ int map__browse(struct map *map)if (maxaddr <symbol__end (pos)) maxaddr =symbol__end (pos); -if (verbose >0 ) { -u32 *idx =symbol__browser_index (pos); - *idx = mb.b.nr_entries; - } ++mb.b.nr_entries; }
Signature · 1 line
2.56.0.313.g2479322e68-goog
Ian Rogers
reviewerPATCH v2 5/18 perf db-export: Store symbol database IDs in a hashmap
In reply to Ian RogersCurrently, db-export reserves sizeof(u64) bytes ahead of every allocated struct symbol via symbol_conf.priv_size and accesses the database ID at a negative offset via symbol__priv(). This adds 8 bytes to every loaded symbol across all DSOs even when only a subset of symbols are exported, and negative-offset pointer arithmetic ahead of struct symbol is incompatible with wrapping struct symbol in DECLARE_RC_STRUCT(symbol).
Store exported symbol database IDs in an on-demand sym_db_ids hashmap in struct db_export instead of using symbol__priv().
--- tools/perf/util/db-export.c | 68 ++++++++++++++++++++++++++++++------- tools/perf/util/db-export.h | 8 +++++ 2 files changed, 63 insertions(+), 13 deletions(-)
tools/perf/util/db-export.c +55 -13
@@ -4,21 +4,37 @@* Copyright (c) 2014, Intel Corporation. */ +#include "db-export.h" +#include <errno.h> +#include <stdbool.h> +#include <stddef.h> #include <stdlib.h> +#include <linux/err.h> + +#include "call-path.h" +#include "callchain.h" +#include "comm.h" #include "dso.h" +#include "event.h" #include "evsel.h" +#include "hashmap.h" #include "machine.h" -#include "thread.h" -#include "comm.h" -#include "symbol.h" #include "map.h" -#include "event.h" +#include "symbol.h" #include "thread-stack.h" -#include "callchain.h" -#include "call-path.h" -#include "db-export.h" +#include "thread.h" + +static size_t sym_ptr_hash (long key,void *ctx__maybe_unused ) +{ +return hash_bits ((size_t )key,16 ); +} + +static bool sym_ptr_equal (long key1,long key2,void *ctx__maybe_unused ) +{ +return key1 == key2; +}int db_export__init (struct db_export *dbe) {
@@ -28,6 +44,8 @@ int db_export__init(struct db_export *dbe)void db_export__exit (struct db_export *dbe) { +hashmap__free (dbe->sym_db_ids); + dbe->sym_db_ids =NULL ;call_return_processor__free (dbe->crp); dbe->crp =NULL ; }
@@ -156,15 +174,41 @@ int db_export__dso(struct db_export *dbe, struct dso *dso,return 0 ; } +/* +* Return the database export ID assigned to @sym in @dbe->sym_db_ids, or 0 +* if @sym has not been exported yet. +*/ +u64 db_export__symbol_db_id (struct db_export *dbe,struct symbol *sym) +{ +long db_id =0 ; + +if (sym && dbe->sym_db_ids) +hashmap__find (dbe->sym_db_ids, (long )sym, &db_id); +return (u64 )db_id; +} +int db_export__symbol (struct db_export *dbe,struct symbol *sym,struct dso *dso) { -u64 *sym_db_id =symbol__priv (sym); +long db_id; -if (*sym_db_id) +if (db_export__symbol_db_id (dbe, sym))return 0 ; - *sym_db_id = ++dbe->symbol_last_db_id; +if (!dbe->sym_db_ids) { + dbe->sym_db_ids = +hashmap__new (sym_ptr_hash, sym_ptr_equal,NULL ); +if (IS_ERR (dbe->sym_db_ids)) { +int err =PTR_ERR (dbe->sym_db_ids); + + dbe->sym_db_ids =NULL ; +return err; + } + } + + db_id = ++dbe->symbol_last_db_id; +if (hashmap__set (dbe->sym_db_ids, (long )sym, db_id,NULL ,NULL )) +return -ENOMEM ;if (dbe->export_symbol)return dbe->export_symbol (dbe, sym, dso);
@@ -192,12 +236,10 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al, }if (al->sym) { -u64 *db_id =symbol__priv (al->sym); - err =db_export__symbol (dbe, al->sym, dso);if (err)return err; - *sym_db_id = *db_id; + *sym_db_id =db_export__symbol_db_id (dbe, al->sym); *offset = al->addr -symbol__start (al->sym); } }
tools/perf/util/db-export.h +8 -0
@@ -21,6 +21,8 @@ struct call_return_processor;struct call_path_root ;struct call_path ;struct call_return ; +struct hashmap ; +struct symbol ;struct export_sample {union perf_event *event;
@@ -63,6 +65,11 @@ struct db_export { u64 th_in_id, u64 comm_in_id, int flags);struct call_return_processor *crp;struct call_path_root *cpr; +/** +* @sym_db_ids: Hashmap mapping struct symbol pointers to their +* exported 64-bit database IDs. +*/ +struct hashmap *sym_db_ids;u64 evsel_last_db_id;u64 machine_last_db_id;u64 thread_last_db_id;
@@ -90,6 +97,7 @@ int db_export__comm_thread(struct db_export *dbe, struct comm *comm, struct thread *thread);int db_export__dso (struct db_export *dbe,struct dso *dso,struct machine *machine); +u64 db_export__symbol_db_id (struct db_export *dbe,struct symbol *sym);int db_export__symbol (struct db_export *dbe,struct symbol *sym,struct dso *dso); int db_export__branch_type(struct db_export *dbe, u32 branch_type,
Signature · 1 line
2.56.0.313.g2479322e68-goog
Ian Rogers
reviewerPATCH v2 6/18 perf annotate: Move struct annotation to a hashmap and remove symbol_conf.priv_size
In reply to Ian RogersPreviously, enabling annotation added sizeof(struct annotation) to symbol_conf.priv_size, bloating every loaded struct symbol in every DSO even though only a small fraction of symbols are ever annotated. In addition, allocating private data at a negative offset ahead of struct symbol via symbol__priv() prevents wrapping struct symbol with DECLARE_RC_STRUCT(symbol) for reference-count checking.
Move struct annotation storage to an on-demand hashmap in util/annotate.c keyed by struct symbol pointer, guarded by annotations_lock and tracked via a new SYMBOL_FLAG_ANNOTATED flag in sym->flags so unannotated symbols pay zero lookup or teardown overhead. If allocation fails in symbol__annotation(), return a static empty_annotation fallback so callers that do not check for NULL remain safe while mutating callers check for &empty_annotation.
With annotation, TUI map browser indices, and db-export symbol IDs no longer using negative-offset space ahead of struct symbol, remove symbol_conf.priv_size and symbol__priv().
--- tools/perf/tests/builtin-test.c | 1 - tools/perf/ui/browsers/hists.c | 13 +++- tools/perf/util/annotate.c | 132 +++++++++++++++++++++++++++++++- tools/perf/util/annotate.h | 6 +- tools/perf/util/sort.c | 5 +- tools/perf/util/symbol.c | 34 ++++---- tools/perf/util/symbol.h | 16 ++-- tools/perf/util/symbol_conf.h | 1 - 8 files changed, 167 insertions(+), 41 deletions(-)
tools/perf/tests/builtin-test.c +0 -1
@@ -2050,7 +2050,6 @@ int cmd_test(int argc, const char **argv)if (dont_fork) sequential = true; - symbol_conf.priv_size =sizeof (int ); symbol_conf.try_vmlinux_path = true;
tools/perf/ui/browsers/hists.c +10 -3
@@ -796,7 +796,9 @@ static void hist_browser__show_callchain_entry(struct hist_browser *browser, {int color, width;char folded_sign =callchain_list__folded (chain); -bool show_annotated = browser->show_dso && chain->ms.sym &&symbol__annotation (chain->ms.sym)->src; +bool show_annotated = browser->show_dso && chain->ms.sym && +symbol__annotated (chain->ms.sym) && +symbol__annotation (chain->ms.sym)->src; color =HE_COLORSET_NORMAL ; width = browser->b.width - (offset +2 );
@@ -2486,6 +2488,9 @@ do_annotate(struct hist_browser *browser, struct popup_action *act) perf_env__lookup_objdump(browser->env, &annotate_opts.objdump_path))return 0 ; +if (!symbol__annotated (act->ms.sym)) +return 0 ; + notes =symbol__annotation (act->ms.sym);if (!notes->src)return 0 ;
@@ -2545,7 +2550,8 @@ add_annotate_opt(struct popup_action *act, char **optstr,if (!ms->sym) ms->sym =symbol__new_unresolved (addr, ms->map); -if (ms->sym ==NULL ||symbol__annotation (ms->sym)->src ==NULL ) +if (ms->sym ==NULL || !symbol__annotated (ms->sym) || +symbol__annotation (ms->sym)->src ==NULL )return 0 ;if (asprintf (optstr,"Annotate %s" ,symbol__name (ms->sym)) <0 )
@@ -3204,7 +3210,8 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *hdo_annotate (browser, &hotkey_act);map_symbol__exit (&hotkey_act.ms); }else { -if (symbol__annotation (browser->selection->sym)->src ==NULL ) { +if (!symbol__annotated (browser->selection->sym) || +symbol__annotation (browser->selection->sym)->src ==NULL ) { ui_browser__warning(&browser->b, delay_secs * 2, "No samples for the \"%s\" symbol.\n\n" "Probably appeared just in a callchain",
tools/perf/util/annotate.c +130 -2
176 lines not on this page · raw message
tools/perf/util/annotate.h +2 -4
@@ -449,10 +449,8 @@ static inline struct sym_hist *annotation__histogram(struct annotation *notes,struct sym_hist_entry *annotated_source__hist_entry (struct annotated_source *src,const struct evsel *evsel,u64 offset); -static inline struct annotation *symbol__annotation (struct symbol *sym) -{ -return (void *)sym - symbol_conf.priv_size; -} +struct annotation *symbol__annotation (struct symbol *sym); +void symbol__annotation_delete (struct symbol *sym);int addr_map_symbol__inc_samples (struct addr_map_symbol *ams,struct perf_sample *sample);
tools/perf/util/sort.c +3 -2
@@ -795,14 +795,15 @@ static int hist_entry__sym_ipc_snprintf(struct hist_entry *he, char *bf, {struct symbol *sym = he->ms.sym; -struct annotated_branch *branch; +struct annotated_branch *branch =NULL ;double ipc =0.0 , coverage =0.0 ;char tmp[64 ];if (!sym)return repsep_snprintf (bf, size,"%-*s" , width,"-" ); - branch =symbol__annotation (sym)->branch; +if (symbol__annotated (sym)) + branch =symbol__annotation (sym)->branch;if (branch && branch->hit_cycles) ipc = branch->hit_insn / ((double )branch->hit_cycles);
tools/perf/util/symbol.c +13 -21
@@ -319,19 +319,11 @@ void symbols__fixup_end(struct rb_root_cached *symbols, bool is_kallsyms)struct symbol *symbol__new (u64 start,u64 len,u8 binding,u8 type,const char *name) {size_t namelen =strlen (name) +1 ; -struct symbol *sym =calloc (1 , (symbol_conf.priv_size + -sizeof (*sym) + namelen)); +struct symbol *sym =calloc (1 ,sizeof (*sym) + namelen); +if (sym ==NULL )return NULL ; -if (symbol_conf.priv_size) { -if (symbol_conf.init_annotation) { -struct annotation *notes = (void *)sym; -annotation__init (notes); - } - sym = ((void *)sym) + symbol_conf.priv_size; - } - sym->start = start; sym->end = len ? start + len : start; atomic_init(&sym->flags, (type << SYMBOL_FLAG_TYPE_SHIFT) |
@@ -347,14 +339,9 @@ struct symbol *symbol__new(u64 start, u64 len, u8 binding, u8 type, const char *void symbol__delete (struct symbol *sym) { -if (symbol_conf.priv_size) { -if (symbol_conf.init_annotation) { -struct annotation *notes =symbol__annotation (sym); - -annotation__exit (notes); - } - } -free (((void *)sym) - symbol_conf.priv_size); +if (symbol__annotated (sym)) +symbol__annotation_delete (sym); +free (sym); }void symbol__set_ignore (struct symbol *sym,bool ignore)
@@ -389,6 +376,14 @@ void symbol__set_ifunc_alias(struct symbol *sym, bool ifunc_alias)atomic_fetch_and (&sym->flags, ~SYMBOL_FLAG_IFUNC_ALIAS ); } +void symbol__set_annotated (struct symbol *sym,bool annotated) +{ +if (annotated) +atomic_fetch_or (&sym->flags,SYMBOL_FLAG_ANNOTATED ); +else +atomic_fetch_and (&sym->flags, ~SYMBOL_FLAG_ANNOTATED ); +} +static void symbol__set_idle (struct symbol *sym,bool idle) {uint16_t old_flags =atomic_load_explicit (&sym->flags, memory_order_relaxed);
@@ -2482,7 +2477,6 @@ int symbol__annotation_init(void)return -1 ; } - symbol_conf.priv_size +=sizeof (struct annotation ); symbol_conf.init_annotation = true;return 0 ; }
@@ -2525,8 +2519,6 @@ int symbol__init(struct perf_env *env)if (symbol_conf.initialized)return 0 ; - symbol_conf.priv_size =PERF_ALIGN (symbol_conf.priv_size,sizeof (u64)); -symbol__elf_init ();if (symbol_conf.try_vmlinux_path &&vmlinux_path__init (env) <0 )
tools/perf/util/symbol.h +9 -7
@@ -80,9 +80,9 @@ enum symbol_idle_kind {#define SYMBOL_FLAG_INLINED (1 << 11)#define SYMBOL_FLAG_ANNOTATE2 (1 << 12)#define SYMBOL_FLAG_IFUNC_ALIAS (1 << 13) +#define SYMBOL_FLAG_ANNOTATED (1 << 14)/** -* A symtab entry. When allocated this may be preceded by an annotation (see -* symbol__annotation) and/or a browser_index (see symbol__browser_index). +* A symtab entry. */ struct symbol {struct rb_node rb_node;
@@ -178,12 +178,19 @@ static inline bool symbol__ifunc_alias(const struct symbol *sym) SYMBOL_FLAG_IFUNC_ALIAS) != 0; } +static inline bool symbol__annotated (const struct symbol *sym) +{ +return (atomic_load_explicit (&sym->flags, memory_order_relaxed) & +SYMBOL_FLAG_ANNOTATED ) !=0 ; +} +bool symbol__is_idle (struct symbol *sym,const struct dso *dso,struct perf_env *env);void symbol__set_ignore (struct symbol *sym,bool ignore);void symbol__set_annotate2 (struct symbol *sym,bool annotate2);void symbol__set_inlined (struct symbol *sym,bool inlined);void symbol__set_ifunc_alias (struct symbol *sym,bool ifunc_alias); +void symbol__set_annotated (struct symbol *sym,bool annotated);/* symbols__for_each_entry - iterate over symbols (rb_root) *
@@ -217,11 +224,6 @@ static inline int __symbol__join_symfs(char *bf, size_t size, const char *path)extern int vmlinux_path__nr_entries;extern char **vmlinux_path; -static inline void *symbol__priv (struct symbol *sym) -{ -return ((void *)sym) - symbol_conf.priv_size; -} -struct ref_reloc_sym {const char *name;u64 addr;
tools/perf/util/symbol_conf.h +0 -1
@@ -42,7 +42,6 @@ struct symbol_conf {/* Was hybrid_merge given on the command line rather than in a config file? */ bool hybrid_merge_set;bool nanosecs; -unsigned short priv_size;bool try_vmlinux_path, init_annotation, force,
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 06/18] perf annotate: Move struct annotation to a hashmap and remove symbol_conf.priv_size
tools/perf/util/annotate.c annotated_source__hist_entry(struct annotated_source *src, const struct evsel *e · line 191
Quoted 4 lines · Ian Rogers
tools/perf/util/annotate.c +4
-0+static size_t sym_ptr_hash (long key,void *ctx__maybe_unused ) +{ +return hash_bits ((size_t )key,16 ); +}
hashmap.h hashes the callback result again with hash_bits(h, cap_bits), which seems to leave this function with only 64K possible results. If the map grows beyond that, I think the extra buckets cannot help and the chains just get longer. Would it be better to return the key here and let hashmap.h do the hashing? Patch 05/18 appears to have the same issue for the db-export ids.
This seems worth avoiding for the annotation map in particular. After the first annotation access for a symbol, symbol__annotation() is a read-locked hashmap lookup on the per-sample path in perf top and TUI perf report. Before this it was pointer arithmetic.
Quoted 4 lines
+ notes = zalloc(sizeof(*notes));
[...]
+annotation__init(notes);
+ if (hashmap__add(annotations, (long)sym, notes)< 0) {
One smaller thing: symbol__annotation() is now a get-or-create operation, but some callers only want to inspect existing annotation data. Those calls allocate a hashmap entry and set the annotated flag even when there is no annotation data to use. In patch 15/18 that flag also makes symbols__shrink() retain the symbol.
Could the read-only callers use a non-creating lookup instead?
Quoted 5 lines
+ } else if (
hashmap__find(annotations, (long)sym, &val)) {
+ notes = (struct annotation *)val;
+ up_write(annotations_lock());
+ return notes;
+ }
The write path returns an existing entry without setting the annotated flag. I don't see how the entry and flag can get out of sync with the current locking, so this may only be a defensive concern. If they ever do diverge, though, symbol__delete() would skip removing the entry.
Thanks,
Alireza
Ian Rogers
reviewerPATCH v2 7/18 perf symbol: Switch backing storage from rbtree to struct symbols array
In reply to Ian RogersPreviously, DSO symbols were stored in an unprotected rb_root_cached with a 24-byte struct rb_node embedded inside every struct symbol. That design had three drawbacks:
1. Embedding struct rb_node (3 pointers, 24 bytes on 64-bit) accounted for nearly 40% of struct symbol's fixed header size and scattered symbol lookups across non-contiguous heap nodes.
2. Unlike struct dsos and struct maps, the symbol rbtree had no lock protecting concurrent lookups, sorting, and insertions.
3. Reclaiming unreferenced symbols under memory pressure requires compacting the symbol container in place under a lock.
Replace struct rb_node in struct symbol with a locked, lazily-sorted pointer array in struct symbols, matching the design of struct dsos and struct maps:
- Introduce struct symbols containing a rw_semaphore lock, primary address-sorted symbols array, secondary name-sorted symbols_by_name array, counts, capacity, and sorted state flags.
- Lazily sort symbols by [start, end) using qsort (symbols__sort_locked) and use binary search in symbols__find().
- Update symbols__fixup_duplicate(), symbols__fixup_end(), maps__split_kallsyms_for_kcore(), and maps__split_kallsyms() to operate on the sorted array and compact entries in place.
- Convert symbols__for_each_entry(), dso__for_each_symbol(), and map__for_each_symbol() to callback-based iteration under symbols->lock, matching dsos__for_each_dso() and maps__for_each_map().
- Remove the unsynchronized single-entry dso->last_find_result cache and dso__reset_find_symbol_cache() since binary search over a contiguous array is cache-friendly.
- Remove the direct rb_erase_cached() workaround in builtin-annotate.c's add_sample().
- Update the TUI map browser (ui/browsers/map.c) to snapshot the sorted symbol array into an entries array and use ui_browser__argv_refresh and ui_browser__argv_seek instead of ui_browser__rb_tree_refresh.
Using 'perf bench internals symbols':
Before (perf binary):
Average DSO load + sort: 7.041 msec (+- 0.051 msec)
Average DSO memory usage: 1401 KB (+- 4 KB)
Average bulk insert + lookup: 1453.950 usec (+- 10.895 usec)
After (perf binary, -13.5% load, -14.6% memory, -7.7% bulk):
Average DSO load + sort: 6.089 msec (+- 0.039 msec)
Average DSO memory usage: 1197 KB (+- 3 KB)
Average bulk insert + lookup: 1341.500 usec (+- 21.564 usec) Before (libLLVM.so.21.1):
Average DSO load + sort: 78.599 msec (+- 0.857 msec) Average DSO memory usage: 9530 KB (+- 5 KB)
After (libLLVM.so.21.1, -22.3% load, -8.1% memory):
Average DSO load + sort: 61.074 msec (+- 0.236 msec) Average DSO memory usage: 8755 KB (+- 1 KB)
--- tools/perf/arch/powerpc/util/sym-handling.c | 34 +- tools/perf/builtin-annotate.c | 11 - tools/perf/builtin-kmem.c | 51 +- tools/perf/tests/hists_common.c | 5 +- tools/perf/tests/symbols.c | 62 ++- tools/perf/tests/vmlinux-kallsyms.c | 185 +++---- tools/perf/ui/browser.c | 3 + tools/perf/ui/browsers/hists.c | 3 +- tools/perf/ui/browsers/map.c | 62 ++- tools/perf/util/auxtrace.c | 109 ++-- tools/perf/util/db-export.c | 4 +- tools/perf/util/dso.c | 50 +- tools/perf/util/dso.h | 68 +-- tools/perf/util/intel-pt.c | 74 ++- tools/perf/util/libbfd.c | 3 +- tools/perf/util/machine.c | 30 +- tools/perf/util/map.c | 25 +- tools/perf/util/map.h | 8 +- tools/perf/util/probe-event.c | 98 ++-- tools/perf/util/symbol-elf.c | 12 +- tools/perf/util/symbol.c | 571 +++++++++++++------- tools/perf/util/symbol.h | 72 ++- tools/perf/util/symbol_fprintf.c | 25 +- 23 files changed, 941 insertions(+), 624 deletions(-)
tools/perf/arch/powerpc/util/sym-handling.c +24 -10
@@ -115,13 +115,27 @@ void arch__fix_tev_from_maps(struct perf_probe_event *pev, }#ifdef HAVE_LIBELF_SUPPORT +struct post_process_probe_trace_events_args { +struct perf_probe_event *pev; +struct probe_trace_event *tev; +struct map *map; +}; + +static int post_process_probe_trace_events_cb (struct symbol *sym,void *data) +{ +struct post_process_probe_trace_events_args *args = data; + +if (map__unmap_ip (args->map,symbol__start (sym)) == args->tev->point.address) { +arch__fix_tev_from_maps (args->pev, args->tev, args->map, sym); +return 1 ; + } +return 0 ; +} +void arch__post_process_probe_trace_events (struct perf_probe_event *pev,int ntevs) { -struct probe_trace_event *tev;struct map *map; -struct symbol *sym =NULL ; -struct rb_node *tmp;int i =0 ; map =get_target_map (pev->target, pev->nsi, pev->uprobes);
@@ -129,13 +143,13 @@ void arch__post_process_probe_trace_events(struct perf_probe_event *pev,return ;for (i =0 ; i < ntevs; i++) { - tev = &pev->tevs[i]; -map__for_each_symbol (map, sym, tmp) { -if (map__unmap_ip (map,symbol__start (sym)) == tev->point.address) { -arch__fix_tev_from_maps (pev, tev, map, sym); -break ; - } - } +struct post_process_probe_trace_events_args args = { + .pev = pev, + .tev = &pev->tevs[i], + .map = map, + }; + +map__for_each_symbol (map, post_process_probe_trace_events_cb, &args); } }#endif /* HAVE_LIBELF_SUPPORT */
tools/perf/builtin-annotate.c +0 -11
@@ -254,17 +254,6 @@ static int add_sample(struct perf_sample *sample, (al->sym == NULL || strcmp(ann->sym_hist_filter, symbol__name(al->sym)) != 0)) {/* We're only interested in a symbol named sym_hist_filter */ -/* -* FIXME: why isn't this done in the symbol_filter when loading -* the DSO? -*/ -if (al->sym !=NULL ) { -struct dso *dso =map__dso (al->map); - -rb_erase_cached (&al->sym->rb_node,dso__symbols (dso)); -symbol__delete (al->sym); -dso__reset_find_symbol_cache (dso); - }return 0 ; }
tools/perf/builtin-kmem.c +29 -22
72 lines not on this page · raw message
tools/perf/tests/hists_common.c +4 -1
@@ -146,7 +146,10 @@ struct machine *setup_fake_machine(struct machines *machines) goto out; } -symbols__insert (dso__symbols (dso), sym); +if (symbols__insert (dso__symbols (dso), sym)) { +dso__put (dso); + goto out; + } }dso__put (dso);
tools/perf/tests/symbols.c +37 -25
79 lines not on this page · raw message
tools/perf/tests/vmlinux-kallsyms.c +93 -92
221 lines not on this page · raw message
tools/perf/ui/browser.c +3 -0
@@ -622,6 +622,9 @@ static int ui_browser__color_config(const char *var, const char *value,void ui_browser__argv_seek (struct ui_browser *browser,off_t offset,int whence) { +if (browser->nr_entries ==0 ) +return ; +switch (whence) {case SEEK_SET : browser->top = browser->entries;
tools/perf/ui/browsers/hists.c +2 -1
@@ -2531,7 +2531,8 @@ static struct symbol *symbol__new_unresolved(u64 addr, struct map *map)return NULL ; } -dso__insert_symbol (map__dso (map), sym); +if (dso__insert_symbol (map__dso (map), sym)) +return NULL ; }return sym;
tools/perf/ui/browsers/map.c +41 -21
92 lines not on this page · raw message
tools/perf/util/auxtrace.c +72 -37
133 lines not on this page · raw message
tools/perf/util/db-export.c +2 -2
@@ -231,8 +231,8 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al,if (!al->sym) { al->sym =symbol__new (al->addr,0 ,0 ,0 ,"unknown" ); -if (al->sym) -dso__insert_symbol (dso, al->sym); +if (al->sym &&dso__insert_symbol (dso, al->sym)) + al->sym =NULL ; }if (al->sym) {
tools/perf/util/dso.c +26 -24
85 lines not on this page · raw message
tools/perf/util/dso.h +11 -57
131 lines not on this page · raw message
tools/perf/util/intel-pt.c +45 -29
94 lines not on this page · raw message
tools/perf/util/libbfd.c +2 -1
@@ -453,7 +453,8 @@ int dso__load_bfd_symbols(struct dso *dso, const char *debugfile)if (!symbol) goto out_free; -symbols__insert (dso__symbols (dso), symbol); +if (symbols__insert (dso__symbols (dso), symbol)) + goto out_free; }#ifdef bfd_get_section#undef bfd_asymbol_section
tools/perf/util/machine.c +18 -12
@@ -796,7 +796,7 @@ static int machine__process_ksymbol_register(struct machine *machine, err = -ENOMEM ; goto out; } -dso__insert_symbol (dso, sym); + err =dso__insert_symbol (dso, sym); out:map__put (map);dso__put (dso);
@@ -1095,29 +1095,35 @@ int machine__create_extra_kernel_map(struct machine *machine,return err; } -static u64 find_entry_trampoline (struct dso *dso) +static int find_entry_trampoline_cb (struct symbol *sym,void *data) {/* Duplicates are removed so lookup all aliases */ -const char *syms[] = { +static const char *const syms[] = {"_entry_trampoline" ,"__entry_trampoline_start" ,"entry_SYSCALL_64_trampoline" , }; -struct symbol *sym =dso__first_symbol (dso); -unsigned int i; +u64 *addr = data; -for (; sym; sym =dso__next_symbol (sym)) { -if (symbol__binding (sym) !=STB_GLOBAL ) -continue ; -for (i =0 ; i <ARRAY_SIZE (syms); i++) { -if (!strcmp (symbol__name (sym), syms[i])) -return symbol__start (sym); +if (symbol__binding (sym) !=STB_GLOBAL ) +return 0 ; +for (unsigned int i =0 ; i <ARRAY_SIZE (syms); i++) { +if (!strcmp (symbol__name (sym), syms[i])) { + *addr =symbol__start (sym); +return 1 ; } } -return 0 ; } +static u64 find_entry_trampoline (struct dso *dso) +{ +u64 addr =0 ; + +dso__for_each_symbol (dso, find_entry_trampoline_cb, &addr); +return addr; +} +/* * These values can be used for kernels that do not have symbols for the entry * trampolines in kallsyms.
tools/perf/util/map.c +10 -15
@@ -314,27 +314,22 @@ void map__put(struct map *map)void map__fixup_start (struct map *map) { -struct dso *dso =map__dso (map); -struct rb_root_cached *symbols =dso__symbols (dso); -struct rb_node *nd =rb_first_cached (symbols); - -if (nd !=NULL ) { - struct symbol *sym = rb_entry(nd, struct symbol, rb_node); +struct symbols *symbols =dso__symbols (map__dso (map)); -map__set_start (map,symbol__start (sym)); - } +symbols__sort_read_lock (symbols); +if (symbols->cnt >0 ) +map__set_start (map,symbol__start (symbols->symbols[0 ])); +up_read (&symbols->lock); }void map__fixup_end (struct map *map) { -struct dso *dso =map__dso (map); -struct rb_root_cached *symbols =dso__symbols (dso); -struct rb_node *nd =rb_last (&symbols->rb_root); +struct symbols *symbols =dso__symbols (map__dso (map)); -if (nd !=NULL ) { - struct symbol *sym = rb_entry(nd, struct symbol, rb_node); -map__set_end (map,symbol__end (sym)); - } +symbols__sort_read_lock (symbols); +if (symbols->cnt >0 ) +map__set_end (map,symbol__end (symbols->symbols[symbols->cnt -1 ])); +up_read (&symbols->lock); }#define DSO__DELETED "(deleted)"
tools/perf/util/map.h +4 -4
@@ -146,12 +146,12 @@ struct thread;/* map__for_each_symbol - iterate over the symbols in the given map * * @map: the 'struct map *' in which symbols are iterated -* @pos: the 'struct symbol *' to use as a loop cursor -* @n: the 'struct rb_node *' to use as a temporary storage +* @cb: callback function invoked for each symbol +* @data: opaque data pointer passed to @cb * Note: caller must ensure map->dso is not NULL (map is loaded). */ -#define map__for_each_symbol (map, pos, n) \ - dso__for_each_symbol(map__dso(map), pos, n) +#define map__for_each_symbol (map, cb, data) \ + dso__for_each_symbol(map__dso(map), cb, data)/* map__for_each_symbol_with_name - iterate over the symbols in the given map * that have the given name
tools/perf/util/probe-event.c +58 -40
125 lines not on this page · raw message
tools/perf/util/symbol-elf.c +8 -4
@@ -586,7 +586,8 @@ static int dso__synthesize_plt_got_symbols(struct dso *dso, Elf *elf, sym =symbol__new (shdr.sh_offset + i, shdr.sh_entsize,STB_GLOBAL ,STT_FUNC , buf);if (!sym) goto out; -symbols__insert (dso__symbols (dso), sym); +if (symbols__insert (dso__symbols (dso), sym)) + goto out; } err =0 ; out:
@@ -640,7 +641,8 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss) plt_sym =symbol__new (shdr_plt.sh_offset, plt_header_size,STB_GLOBAL ,STT_FUNC ,".plt" );if (!plt_sym) goto out_elf_end; -symbols__insert (dso__symbols (dso), plt_sym); +if (symbols__insert (dso__symbols (dso), plt_sym)) + goto out_elf_end;/* Only x86 has .plt.got */ if (machine_is_x86(ehdr.e_machine) &&
@@ -762,7 +764,8 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss) goto out_elf_end; plt_offset += plt_entry_size; -symbols__insert (dso__symbols (dso), f); +if (symbols__insert (dso__symbols (dso), f)) + goto out_elf_end; ++nr; }
@@ -1785,7 +1788,8 @@ dso__load_sym_internal(struct dso *dso, struct map *map, struct symsrc *syms_ss,arch__sym_update (f, &sym); -__symbols__insert (dso__symbols (curr_dso), f); +if (symbols__insert (dso__symbols (curr_dso), f)) + goto out_elf_end; nr++; }dso__put (curr_dso);
tools/perf/util/symbol.c +382 -189
788 lines not on this page · raw message
tools/perf/util/symbol.h +52 -20
117 lines not on this page · raw message
tools/perf/util/symbol_fprintf.c +18 -7
@@ -59,15 +59,26 @@ size_t symbol__fprintf_symname(const struct symbol *sym, FILE *fp)return __symbol__fprintf_symname_offs (sym,NULL , false, false, fp); } +struct fprintf_symbols_by_name_args { +FILE *fp; +size_t ret; +}; + +static int fprintf_symbols_by_name_cb (struct symbol *pos,void *data) +{ +struct fprintf_symbols_by_name_args *args = data; + + args->ret +=fprintf (args->fp,"%s\n" ,symbol__name (pos)); +return 0 ; +} +size_t dso__fprintf_symbols_by_name (struct dso *dso,FILE *fp) { -size_t ret =0 ; - -for (size_t i =0 ; i <dso__symbol_names_len (dso); i++) { -struct symbol *pos =dso__symbol_names (dso)[i]; +struct fprintf_symbols_by_name_args args = { + .fp = fp, + }; - ret +=fprintf (fp,"%s\n" ,symbol__name (pos)); - } -return ret; +dso__for_each_symbol_by_name (dso, fprintf_symbols_by_name_cb, &args); +return args.ret; }
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 07/18] perf symbol: Switch backing storage from rbtree to struct symbols array
Quoted 3 lines · Ian Rogers
+ if (
symbols->sorted&&symbols->cnt> 0 &&
+ !symbol__less_or_equal(symbols->symbols[symbols->cnt - 1], sym))
+symbols->sorted= false;
Just want to call out that in here, every out-of-order insert clears sorted, and the next lookup qsort()s the whole array. This is fixed in 9/18 I think but moving the sorted insert into this patch make it more clean.
Some behaviour notes:
- Symbols with the same start are now ordered by end instead of by insertion order. That can change which alias symbols__fixup_duplicate() keeps.
- The order of full ties now depends on qsort() being stable. Patch 09/18 replaces it with a stable sort, but still falls back to qsort() if the merge buffer can't be allocated.
- On libc.so.6 I see 62 differing @plt-name lines compared with the previous patch, for example __GI_memcpy@plt, although both sides still have 55 @plt symbols. My guess is that this comes from get_ifunc_name(): for an unnamed R_X86_64_IRELATIVE relocation it resolves the addend with dso__find_symbol_nocache(), and libc has several IFUNC names or aliases at the same address. The old rbtree and the new array do not necessarily return the same alias there, because same-start symbols are now ordered by end and full ties go through qsort(). The PLT address and count stay the same, but the displayed name changes. Was that intended?
Thanks,
Alireza
Ian Rogers
reviewerPATCH v2 8/18 perf symbol: Add reference counting and DECLARE_RC_STRUCT(symbol)
In reply to Ian RogersPreviously, struct symbol pointers returned by dso__find_symbol() and map__find_symbol() were uncounted borrowed pointers into the DSO's symbol table. Callers stored these raw pointers in struct map_symbol, struct addr_location, struct hist_entry, struct callchain_list, and annotation structures without lifetime tracking, making it unsafe to evict unreferenced symbols from a DSO when reclaiming memory.
Wrap struct symbol with DECLARE_RC_STRUCT(symbol) and add a refcount_t refcnt field along with symbol__get(), symbol__put(), and symbol__zput() helpers:
- Return a counted symbol reference acquired under symbols->lock from symbol lookup functions (symbols__find(), dso__find_symbol(), dso__find_symbol_nocache(), map__find_symbol(), dso__find_symbol_by_name(), dso__next_symbol_by_name(), map__find_symbol_by_name()).
- Update util/annotate.c and util/db-export.c to hash and compare the underlying RC_STRUCT(symbol) pointer via RC_CHK_ACCESS(sym) and RC_CHK_EQUAL() so REFCNT_CHECKING wrapper pointers for the same symbol resolve to the same hashmap entry, and hold a symbol reference in struct db_export's sym_db_ids hashmap.
- Update struct addr_location, struct map_symbol, struct call_path, struct block_info, struct disasm_line, struct annotated_op_loc, struct c2c_sym_Key, inline_list, srcline, thread_stack, and hist/sort/unwind consumers to acquire and release symbol references and use RC_CHK_EQUAL() for pointer identity comparisons.
--- tools/perf/bench/symbols.c | 7 +- tools/perf/builtin-annotate.c | 7 +- tools/perf/builtin-kallsyms.c | 2 + tools/perf/builtin-kmem.c | 12 +- tools/perf/builtin-lock.c | 17 ++- tools/perf/builtin-script.c | 4 +- tools/perf/builtin-top.c | 2 +- tools/perf/builtin-trace.c | 1 + tools/perf/tests/dlfilter-test.c | 8 +- tools/perf/tests/hists_cumulate.c | 4 +- tools/perf/tests/hists_filter.c | 10 +- tools/perf/tests/hists_link.c | 16 ++- tools/perf/tests/hists_output.c | 7 +- tools/perf/tests/kallsyms-split.c | 5 +- tools/perf/tests/pe-file-parsing.c | 1 + tools/perf/tests/symbols.c | 14 ++- tools/perf/tests/vmlinux-kallsyms.c | 10 +- tools/perf/ui/browsers/hists.c | 11 +- tools/perf/ui/browsers/map.c | 7 +- tools/perf/ui/stdio/hist.c | 3 +- tools/perf/util/addr2line.c | 8 +- tools/perf/util/addr_location.c | 4 + .../perf/util/annotate-arch/annotate-alpha.c | 5 +- .../util/annotate-arch/annotate-loongarch.c | 11 +- tools/perf/util/annotate-arch/annotate-s390.c | 6 +- tools/perf/util/annotate.c | 22 ++-- tools/perf/util/block-info.c | 6 +- tools/perf/util/bpf_lock_contention.c | 25 +++- tools/perf/util/c2c-function.c | 13 ++- tools/perf/util/call-path.c | 15 ++- tools/perf/util/callchain.c | 4 +- tools/perf/util/capstone.c | 2 + tools/perf/util/db-export.c | 39 +++++-- tools/perf/util/disasm.c | 22 ++-- tools/perf/util/event.c | 8 +- tools/perf/util/hist.c | 13 ++- tools/perf/util/libbfd.c | 7 +- tools/perf/util/libdw.c | 5 +- tools/perf/util/llvm.c | 2 + tools/perf/util/lock-contention.c | 8 +- tools/perf/util/machine.c | 56 ++++++--- tools/perf/util/map.h | 10 +- tools/perf/util/map_symbol.c | 4 +- tools/perf/util/maps.c | 3 +- tools/perf/util/probe-event.c | 17 ++- tools/perf/util/python.c | 20 +--- tools/perf/util/sort.c | 2 +- tools/perf/util/srcline.c | 6 +- tools/perf/util/symbol-elf.c | 6 +- tools/perf/util/symbol.c | 107 +++++++++++------- tools/perf/util/symbol.h | 60 +++++++--- tools/perf/util/thread-stack.c | 12 +- tools/perf/util/threads.c | 17 ++- tools/perf/util/unwind-libdw.c | 2 +- tools/perf/util/unwind-libunwind.c | 2 +- 55 files changed, 478 insertions(+), 219 deletions(-)
tools/perf/bench/symbols.c +5 -2
@@ -159,8 +159,9 @@ static int do_synthetic_symbols_bench(void)u64 addr = (j < half) ? (0x100000ULL + (u64 )j *0x40ULL +4 ) : (0x100020ULL + (u64)(j - half) *0x40ULL +4 ); +struct symbol *sym =dso__find_symbol_nocache (dso, addr); - (void )dso__find_symbol_nocache (dso, addr); +symbol__put (sym); }gettimeofday (&end,NULL );timersub (&end, &start, &diff);
@@ -183,7 +184,9 @@ static int do_synthetic_symbols_bench(void)return -ENOMEM ; }dso__insert_symbol (dso, sym); - (void )dso__find_symbol_nocache (dso, addr +2 ); + + sym =dso__find_symbol_nocache (dso, addr +2 ); +symbol__put (sym); }gettimeofday (&end,NULL );timersub (&end, &start, &diff);
tools/perf/builtin-annotate.c +6 -1
@@ -134,7 +134,8 @@ static void process_basic_block(struct addr_map_symbol *start, entry =block_range_iter (&iter); entry->coverage++; - entry->sym = sym; +symbol__put (entry->sym); + entry->sym =symbol__get (sym);if (branch) branch->max_coverage =max (branch->max_coverage, entry->coverage);
@@ -174,6 +175,10 @@ static void process_branch_stack(struct branch_stack *bs, struct addr_location * prev = &bi[i].to; } +for (i =0 ; i < (int )bs->nr; i++) { +map_symbol__exit (&bi[i].to.ms); +map_symbol__exit (&bi[i].from.ms); + }free (bi); }
tools/perf/builtin-kallsyms.c +2 -0
@@ -52,6 +52,8 @@ static int __cmd_kallsyms(int argc, const char **argv)map__unmap_ip (map,symbol__start (symbol)),map__unmap_ip (map,symbol__end (symbol)), symbol__start(symbol), symbol__end(symbol)); +symbol__put (symbol); +map__put (map); } out:machine__delete (machine);
tools/perf/builtin-kmem.c +7 -5
@@ -1044,7 +1044,7 @@ static void __print_slab_result(struct rb_root *root, struct alloc_stat *data = rb_entry(next, struct alloc_stat, node);struct symbol *sym =NULL ; -struct map *map; +struct map *map =NULL ;char buf[BUFSIZ ];u64 addr;
@@ -1061,6 +1061,8 @@ static void __print_slab_result(struct rb_root *root, }else {snprintf (buf,sizeof (buf),"%#" PRIx64"" , addr); } +symbol__put (sym); +map__put (map);printf (" %-34s |" , buf); printf(" %9llu/%-5lu | %9llu/%-5lu | %8lu | %9lu | %6.3f%%\n",
@@ -1111,12 +1113,11 @@ static void __print_page_alloc_result(struct perf_session *session, int n_lines)while (next && n_lines--) {struct page_stat *data;struct symbol *sym; -struct map *map;char buf[32 ];const char *caller = buf; data = rb_entry(next, struct page_stat, node); - sym =machine__find_kernel_symbol (machine, data->callsite, &map); + sym =machine__find_kernel_symbol (machine, data->callsite,NULL );if (sym) caller =symbol__name (sym);else
@@ -1127,6 +1128,7 @@ static void __print_page_alloc_result(struct perf_session *session, int n_lines) data->nr_alloc, data->order, migrate_type_str[data->migrate_type], gfp_len, compact_gfp_string(data->gfp_flags), caller); +symbol__put (sym); next =rb_next (next); }
@@ -1153,12 +1155,11 @@ static void __print_page_caller_result(struct perf_session *session, int n_lineswhile (next && n_lines--) {struct page_stat *data;struct symbol *sym; -struct map *map;char buf[32 ];const char *caller = buf; data = rb_entry(next, struct page_stat, node); - sym =machine__find_kernel_symbol (machine, data->callsite, &map); + sym =machine__find_kernel_symbol (machine, data->callsite,NULL );if (sym) caller =symbol__name (sym);else
@@ -1169,6 +1170,7 @@ static void __print_page_caller_result(struct perf_session *session, int n_lines data->nr_alloc, data->order, migrate_type_str[data->migrate_type], gfp_len, compact_gfp_string(data->gfp_flags), caller); +symbol__put (sym); next =rb_next (next); }
tools/perf/builtin-lock.c +14 -3
@@ -1003,11 +1003,15 @@ static int report_lock_contention_begin_event(struct perf_sample *sample) (filters.nr_addrs + 1) * sizeof(*addrs));if (addrs ==NULL ) {pr_warning ("memory allocation failure\n" ); +symbol__put (sym); +map__put (kmap);return -ENOMEM ; } addrs[filters.nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); filters.addrs = addrs; +symbol__put (sym); +map__put (kmap); } }
@@ -1057,9 +1061,12 @@ static int report_lock_contention_begin_event(struct perf_sample *sample)switch (aggr_mode) {case LOCK_AGGR_ADDR : - sym =machine__find_kernel_symbol (machine, key, &kmap); -if (sym) - name =symbol__name (sym); + sym =machine__find_kernel_symbol (machine, key,NULL ); +if (sym) { +scnprintf (buf,sizeof (buf),"%s" ,symbol__name (sym)); + name = buf; +symbol__put (sym); + }break ;case LOCK_AGGR_CALLER : name = buf;
@@ -1671,6 +1678,8 @@ static void print_lock_stat_stdio(struct lock_contention *con, struct lock_stat ip = st->callstack[i]; sym =machine__find_kernel_symbol (con->machine, ip, &kmap);get_symbol_name_offset (kmap, sym, ip, buf,sizeof (buf)); +symbol__put (sym); +map__put (kmap);fprintf (lock_output,"\t\t\t%#lx %s\n" , (unsigned long )ip, buf); } }
@@ -1724,6 +1733,8 @@ static void print_lock_stat_csv(struct lock_contention *con, struct lock_stat *s ip = st->callstack[i]; sym =machine__find_kernel_symbol (con->machine, ip, &kmap);get_symbol_name_offset (kmap, sym, ip, buf,sizeof (buf)); +symbol__put (sym); +map__put (kmap);fprintf (lock_output,"%s %#lx %s" , i ?":" : sep, (unsigned long ) ip, buf); }fprintf (lock_output,"\n" );
tools/perf/builtin-script.c +3 -1
@@ -1430,7 +1430,8 @@ static int ip__fprintf_sym(uint64_t addr, struct thread *thread,if (PRINT_FIELD (SRCLINE )) printed +=map__fprintf_srcline (al.map, al.addr,"\t" , fp); printed +=fprintf (fp,"\n" ); - *lastsym = al.sym; +symbol__put (*lastsym); + *lastsym =symbol__get (al.sym); ret = printed; out:
@@ -1597,6 +1598,7 @@ static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample,print_srccode (thread, x.cpumode, start + off); } out: +symbol__put (lastsym);return printed; }
tools/perf/builtin-top.c +1 -1
@@ -209,7 +209,7 @@ static void perf_top__record_precise_ip(struct perf_top *top,if (sym ==NULL || (use_browser ==0 && (top->sym_filter_entry ==NULL || - top->sym_filter_entry->ms.sym != sym))) + !RC_CHK_EQUAL (top->sym_filter_entry->ms.sym, sym))))return ; notes =symbol__annotation (sym);
tools/perf/builtin-trace.c +1 -0
@@ -799,6 +799,7 @@ size_t syscall_arg__scnprintf_ksym(char *bf, size_t size, struct syscall_arg *ar printed =scnprintf (bf, size,"%s+0x%" PRIx64,symbol__name (sym), offset); } +symbol__put (sym);map__put (map);return printed; }
tools/perf/tests/dlfilter-test.c +6 -2
@@ -273,12 +273,16 @@ static int get_ip_addr(struct test_data *td)return -1 ; sym =map__find_symbol_by_name (map,"foo" ); -if (sym) +if (sym) { td->foo =symbol__start (sym); +symbol__put (sym); + } sym =map__find_symbol_by_name (map,"bar" ); -if (sym) +if (sym) { td->bar =symbol__start (sym); +symbol__put (sym); + }map__put (map);
tools/perf/tests/hists_cumulate.c +3 -1
@@ -114,7 +114,8 @@ static int add_hist_entries(struct hists *hists, struct machine *machine) fake_samples[i].thread =thread__get (al.thread);map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); }addr_location__exit (&al);
@@ -155,6 +156,7 @@ static void put_fake_samples(void)size_t i;for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { +symbol__zput (fake_samples[i].sym);map__zput (fake_samples[i].map);thread__zput (fake_samples[i].thread); }
tools/perf/tests/hists_filter.c +7 -3
@@ -93,7 +93,8 @@ static int add_hist_entries(struct evlist *evlist, fake_samples[i].thread =thread__get (al.thread);map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); } }addr_location__exit (&al);
@@ -109,8 +110,11 @@ static void put_fake_samples(void) {size_t i; -for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) -map__put (fake_samples[i].map); +for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { +symbol__zput (fake_samples[i].sym); +map__zput (fake_samples[i].map); +thread__zput (fake_samples[i].thread); + } }static int test__hists_filter (struct test_suite *test__maybe_unused ,int subtest__maybe_unused )
tools/perf/tests/hists_link.c +11 -5
@@ -99,7 +99,8 @@ static int add_hist_entries(struct evlist *evlist, struct machine *machine) fake_common_samples[k].thread =thread__get (al.thread);map__put (fake_common_samples[k].map); fake_common_samples[k].map =map__get (al.map); - fake_common_samples[k].sym = al.sym; +symbol__put (fake_common_samples[k].sym); + fake_common_samples[k].sym =symbol__get (al.sym); }for (k =0 ; k <ARRAY_SIZE (fake_samples[i]); k++) {
@@ -119,7 +120,8 @@ static int add_hist_entries(struct evlist *evlist, struct machine *machine) fake_samples[i][k].thread =thread__get (al.thread);map__put (fake_samples[i][k].map); fake_samples[i][k].map =map__get (al.map); - fake_samples[i][k].sym = al.sym; +symbol__put (fake_samples[i][k].sym); + fake_samples[i][k].sym =symbol__get (al.sym); } i++; }
@@ -136,11 +138,15 @@ static void put_fake_samples(void) {size_t i, j; -for (i =0 ; i <ARRAY_SIZE (fake_common_samples); i++) +for (i =0 ; i <ARRAY_SIZE (fake_common_samples); i++) {map__put (fake_common_samples[i].map); +symbol__put (fake_common_samples[i].sym); + }for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { -for (j =0 ; j <ARRAY_SIZE (fake_samples[0 ]); j++) +for (j =0 ; j <ARRAY_SIZE (fake_samples[0 ]); j++) {map__put (fake_samples[i][j].map); +symbol__put (fake_samples[i][j].sym); + } } }
@@ -150,7 +156,7 @@ static int find_sample(struct sample *samples, size_t nr_samples,while (nr_samples--) {if (RC_CHK_EQUAL (samples->thread, t) &&RC_CHK_EQUAL (samples->map, m) && - samples->sym == s) +RC_CHK_EQUAL (samples->sym, s))return 1 ; samples++; }
tools/perf/tests/hists_output.c +4 -3
@@ -79,7 +79,8 @@ static int add_hist_entries(struct hists *hists, struct machine *machine) fake_samples[i].thread = al.thread;map__put (fake_samples[i].map); fake_samples[i].map =map__get (al.map); - fake_samples[i].sym = al.sym; +symbol__put (fake_samples[i].sym); + fake_samples[i].sym =symbol__get (al.sym); }addr_location__exit (&al);
@@ -120,8 +121,8 @@ static void put_fake_samples(void)size_t i;for (i =0 ; i <ARRAY_SIZE (fake_samples); i++) { -map__put (fake_samples[i].map); - fake_samples[i].map =NULL ; +symbol__zput (fake_samples[i].sym); +map__zput (fake_samples[i].map); } }
tools/perf/tests/kallsyms-split.c +4 -1
@@ -99,6 +99,7 @@ static int test__kallsyms_split(struct test_suite *test __maybe_unused, {struct machine m = {0 };struct map *map =NULL ; +struct symbol *sym;int ret =TEST_FAIL ;pr_debug ("try to create fake root directory\n" );
@@ -139,10 +140,12 @@ static int test__kallsyms_split(struct test_suite *test __maybe_unused, goto out; } -if (machine__find_kernel_symbol_by_name (&m,"main_symbol3" , &map) ==NULL ) { + sym =machine__find_kernel_symbol_by_name (&m,"main_symbol3" , &map); +if (sym ==NULL ) {pr_debug ("FAIL: failed to find a symbol\n" ); goto out; } +symbol__put (sym);if (!RC_CHK_EQUAL (map,machine__kernel_map (&m))) {pr_debug ("FAIL: the symbol is not in the kernel map\n" );
tools/perf/tests/pe-file-parsing.c +1 -0
@@ -64,6 +64,7 @@ static int run_dir(const char *d)dso__sort_by_name (dso); sym =dso__find_symbol_by_name (dso,"main" , &idx);TEST_ASSERT_VAL ("Failed to find main" , sym); +symbol__put (sym);dso__delete (dso);return TEST_OK ;
tools/perf/tests/symbols.c +8 -6
@@ -260,18 +260,19 @@ static const struct kallsyms_sym kallsyms_syms[] = {static int check_symbol (struct dso *dso,u64 addr,const char *name,u64 end) {struct symbol *sym =dso__find_symbol_nocache (dso, addr); +int ret =TEST_OK ;if (!sym ||strcmp (symbol__name (sym), name)) {pr_debug ("%#" PRIx64": expected %s, got %s\n" , addr, name, sym ?symbol__name (sym) :"no symbol" ); -return TEST_FAIL ; - } -if (symbol__end (sym) != end) { + ret =TEST_FAIL ; + }else if (symbol__end (sym) != end) {pr_debug ("%s: expected end %#" PRIx64", got %#" PRIx64"\n" , name, end,symbol__end (sym)); -return TEST_FAIL ; + ret =TEST_FAIL ; } -return TEST_OK ; +symbol__put (sym); +return ret; } static int test__kallsyms_fixup_end(struct test_suite *test __maybe_unused,
@@ -289,7 +290,8 @@ static int test__kallsyms_fixup_end(struct test_suite *test __maybe_unused,if (!sym) goto out; -symbols__insert (dso__symbols (dso), sym); +if (symbols__insert (dso__symbols (dso), sym)) + goto out; }symbols__fixup_end (dso__symbols (dso), true);
tools/perf/tests/vmlinux-kallsyms.c +8 -2
@@ -208,7 +208,7 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data) mem_end =map__unmap_ip (args->vmlinux_map,symbol__end (sym)); first_pair =machine__find_kernel_symbol (&args->kallsyms, mem_start,NULL ); - pair = first_pair; + pair =symbol__get (first_pair);if (pair &&UM (symbol__start (pair)) == mem_start) { next_pair:
@@ -237,8 +237,11 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data)* possible to get proper function end from * kallsyms. */ +symbol__put (pair); +symbol__put (first_pair);return 0 ; } +symbol__put (pair); pair =machine__find_kernel_symbol_by_name (&args->kallsyms,symbol__name (sym),NULL );if (pair) {
@@ -247,14 +250,17 @@ static int test__vmlinux_matches_kallsyms_sym_cb(struct symbol *sym, void *data)pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , mem_start,symbol__name (sym),symbol__name (pair)); +symbol__put (pair); }else {pr_debug ("WARN: %#" PRIx64": diff name v: %s k: %s\n" , mem_start,symbol__name (sym),symbol__name (first_pair)); } +symbol__put (first_pair);return 0 ; } - +symbol__put (pair); +symbol__put (first_pair);if (mem_start ==map__end (args->kallsyms.vmlinux_map)) {/* * Ignore aliases to _etext, i.e. to the end of the kernel text area,
tools/perf/ui/browsers/hists.c +7 -4
@@ -2527,12 +2527,14 @@ static struct symbol *symbol__new_unresolved(u64 addr, struct map *map)if (sym) { src =symbol__hists (sym,1 );if (!src) { -symbol__delete (sym); +symbol__put (sym);return NULL ; } -if (dso__insert_symbol (map__dso (map), sym)) +if (dso__insert_symbol (map__dso (map),symbol__get (sym))) { +symbol__put (sym);return NULL ; + } }return sym;
@@ -2819,7 +2821,7 @@ add_script_opt_2(struct popup_action *act, char **optstr, } act->ms.thread =thread__get (thread); - act->ms.sym = sym; + act->ms.sym =symbol__get (sym); act->fn = do_run_script;return 1 ; }
@@ -3208,6 +3210,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *hmemset (&hotkey_act,0 ,sizeof (hotkey_act));map_symbol__copy (&hotkey_act.ms, &source_ms); +symbol__zput (source_ms.sym);do_annotate (browser, &hotkey_act);map_symbol__exit (&hotkey_act.ms); }else {
@@ -3410,7 +3413,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *h &options[nr_options], &bi->from.ms, bi->from.al_addr); -if (bi->to.ms.sym != bi->from.ms.sym) +if (!RC_CHK_EQUAL (bi->to.ms.sym, bi->from.ms.sym)) nr_options += add_annotate_opt(&actions[nr_options], &options[nr_options], &bi->to.ms,
tools/perf/ui/browsers/map.c +5 -2
@@ -64,12 +64,13 @@ static int map_browser__search(struct map_browser *browser)* cursor at that entry. */ for (u32 idx =0 ; idx < browser->b.nr_entries; ++idx) { -if (entries[idx] == sym) { +if (RC_CHK_EQUAL (entries[idx], sym)) { browser->b.top = &entries[idx]; browser->b.index = browser->b.top_idx = idx;break ; } } +symbol__put (sym); }else ui_helpline__fpush ("%s not found!" , target);
@@ -135,7 +136,7 @@ int map__browse(struct map *map)for (unsigned int idx =0 ; idx < nr_entries; idx++) {struct symbol *pos = symbols->symbols[idx]; - entries[idx] = pos; + entries[idx] =symbol__get (pos);if (maxaddr <symbol__end (pos)) maxaddr =symbol__end (pos); }
@@ -150,6 +151,8 @@ int map__browse(struct map *map) mb.addrlen =snprintf (tmp,sizeof (tmp),"%" PRIx64, maxaddr); ret =map_browser__run (&mb); +for (unsigned int idx =0 ; idx < nr_entries; idx++) +symbol__put (entries[idx]);free (entries);return ret; }
tools/perf/ui/stdio/hist.c +2 -1
@@ -908,7 +908,8 @@ size_t hists__fprintf(struct hists *hists, bool show_header, int max_rows,free (line); out: -zfree (&rem_sq_bracket); +symbol__zput (rem_sq_bracket); + rem_hits.ms.sym =NULL ;return ret; }
tools/perf/util/addr2line.c +7 -1
@@ -274,8 +274,14 @@ static int inline_list__append_record(struct dso *dso, unsigned int line_nr) {struct symbol *inline_sym =new_inline_sym (dso, sym, function); +char *srcline =srcline_from_fileline (filename, line_nr); -return inline_list__append (inline_sym,srcline_from_fileline (filename, line_nr), node); +if (inline_list__append (inline_sym, srcline, node) !=0 ) { +zfree_srcline (&srcline); +symbol__put (inline_sym); +return -1 ; + } +return 0 ; } int cmd__addr2line(const char *dso_name, u64 addr,
tools/perf/util/addr_location.c +4 -0
@@ -2,6 +2,7 @@#include "addr_location.h" #include "map.h" #include "maps.h" +#include "symbol.h" #include "thread.h" void addr_location__init (struct addr_location *al)
@@ -29,13 +30,16 @@ void addr_location__exit(struct addr_location *al) {map__zput (al->map);thread__zput (al->thread); +symbol__zput (al->sym); }void addr_location__copy (struct addr_location *dst,struct addr_location *src) {thread__put (dst->thread);map__put (dst->map); +symbol__put (dst->sym); *dst = *src; dst->thread =thread__get (src->thread); dst->map =map__get (src->map); + dst->sym =symbol__get (src->sym); }
tools/perf/util/annotate-arch/annotate-alpha.c +3 -2
@@ -88,8 +88,9 @@ static int alpha_call__parse(const struct arch *arch, struct ins_operands *ops,if (maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map, -map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; +map__map_ip (target.ms.map, target.addr)) == ops->target.addr && + !RC_CHK_EQUAL (target.ms.sym, ms->sym)) + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
tools/perf/util/annotate-arch/annotate-loongarch.c +7 -4
@@ -56,8 +56,10 @@ static int loongarch_call__parse(const struct arch *arch, struct ins_operands *o };if (maps__find_ams (thread__maps (ms->thread), &target) ==0 && -map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; +map__rip_2objdump (target.ms.map, +map__map_ip (target.ms.map, target.addr)) == ops->target.addr && + !RC_CHK_EQUAL (target.ms.sym, ms->sym)) + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
@@ -99,9 +101,10 @@ static int loongarch_jump__parse(const struct arch *arch, struct ins_operands *o ops->target.outside = target.addr < start || target.addr >= end; -if (maps__find_ams (thread__maps (ms->thread), &target) ==0 && +if (ops->target.outside && +maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);if (!ops->target.outside) { ops->target.offset = target.addr - start;
tools/perf/util/annotate-arch/annotate-s390.c +4 -2
@@ -51,8 +51,10 @@ static int s390_call__parse(const struct arch *arch, struct ins_operands *ops, };if (maps__find_ams (thread__maps (ms->thread), &target) ==0 && -map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; +map__rip_2objdump (target.ms.map, +map__map_ip (target.ms.map, target.addr)) == ops->target.addr && + !RC_CHK_EQUAL (target.ms.sym, ms->sym)) + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
tools/perf/util/annotate.c +15 -7
@@ -198,15 +198,22 @@ static bool sym_ptr_equal(long key1, long key2, void *ctx __maybe_unused)struct annotation *symbol__annotation (struct symbol *sym) {struct annotation *notes =NULL ; -long val =0 ; +long key, val =0 ;if (!sym)return NULL ; +/* +* Key on the underlying RC_STRUCT(symbol) pointer so all +* REFCNT_CHECKING wrapper pointers for the same symbol map to the +* same annotation entry. +*/ + key = (long )RC_CHK_ACCESS (sym); +/* Fast path: look up existing annotation under read lock if flagged. */ if (symbol__annotated (sym)) {down_read (annotations_lock ()); -if (annotations &&hashmap__find (annotations, (long )sym, &val)) +if (annotations &&hashmap__find (annotations, key, &val)) notes = (struct annotation *)val;up_read (annotations_lock ());if (notes)
@@ -221,7 +228,7 @@ struct annotation *symbol__annotation(struct symbol *sym)up_write (annotations_lock ());return &empty_annotation; } - }else if (hashmap__find (annotations, (long )sym, &val)) { + }else if (hashmap__find (annotations, key, &val)) { notes = (struct annotation *)val;up_write (annotations_lock ());return notes;
@@ -233,7 +240,7 @@ struct annotation *symbol__annotation(struct symbol *sym)return &empty_annotation; }annotation__init (notes); -if (hashmap__add (annotations, (long )sym, notes) <0 ) { +if (hashmap__add (annotations, key, notes) <0 ) {free (notes);if (hashmap__size (annotations) ==0 ) {hashmap__free (annotations);
@@ -250,14 +257,15 @@ struct annotation *symbol__annotation(struct symbol *sym)void symbol__annotation_delete (struct symbol *sym) {struct annotation *notes =NULL ; -long val =0 ; +long key, val =0 ;/* Unannotated symbols avoid taking annotations_lock on teardown. */ if (!sym || !symbol__annotated (sym))return ; + key = (long )RC_CHK_ACCESS (sym);down_write (annotations_lock ()); -if (annotations &&hashmap__delete (annotations, (long )sym,NULL , &val)) { +if (annotations &&hashmap__delete (annotations, key,NULL , &val)) { notes = (struct annotation *)val;if (hashmap__size (annotations) ==0 ) {hashmap__free (annotations);
@@ -571,7 +579,7 @@ int addr_map_symbol__account_cycles(struct addr_map_symbol *ams,* it starts on the function start. */ if (start && - (start->ms.sym == ams->ms.sym || + (RC_CHK_EQUAL (start->ms.sym, ams->ms.sym) || (ams->ms.sym && start->addr ==symbol__start (ams->ms.sym) +map__start (ams->ms.map)))) saddr = start->al_addr;
tools/perf/util/block-info.c +4 -2
@@ -64,8 +64,10 @@ static struct block_info *block_info__new(unsigned int br_cntr_nr)void block_info__delete (struct block_info *bi) { -if (bi) +if (bi) { +symbol__zput (bi->sym);free (bi->br_cntr); + }free (bi); }
@@ -105,7 +107,7 @@ static void init_block_info(struct block_info *bi, struct symbol *sym, u64 total_cycles, unsigned int br_cntr_nr, u64 *br_cntr, struct evsel *evsel) { - bi->sym = sym; + bi->sym = symbol__get(sym); bi->start = ch->start; bi->end = offset; bi->cycles = ch->cycles;
tools/perf/util/bpf_lock_contention.c +19 -6
@@ -144,6 +144,7 @@ static void init_numa_data(struct lock_contention *con) &kmap);if (sym) { skel->rodata->contig_page_data_addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym);map__put (kmap);return ; }
@@ -160,6 +161,7 @@ static void init_numa_data(struct lock_contention *con)return ; skel->rodata->node_data_addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym);map__put (kmap);/* get the number of online nodes using the last node number + 1 */
@@ -263,11 +265,15 @@ int lock_contention_prepare(struct lock_contention *con) (con->filters->nr_addrs + 1) * sizeof(*addrs));if (addrs ==NULL ) {pr_warning ("memory allocation failure\n" ); +symbol__put (sym); +map__put (kmap);continue ; } addrs[con->filters->nr_addrs++] =map__unmap_ip (kmap,symbol__start (sym)); con->filters->addrs = addrs; +symbol__put (sym); +map__put (kmap); } naddrs = con->filters->nr_addrs ?: has_mmap_lock; skel->rodata->has_addr =1 ;
@@ -289,6 +295,8 @@ int lock_contention_prepare(struct lock_contention *con) } con->delays[i].addr =map__unmap_ip (kmap,symbol__start (sym)); +symbol__put (sym); +map__put (kmap); } skel->rodata->lock_delay =1 ;bpf_map__set_max_entries (skel->maps.lock_delays, con->nr_delays);
@@ -606,9 +614,12 @@ static const char *lock_contention_get_name(struct lock_contention *con,return "siglock" ;/* global locks with symbols */ - sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup, &kmap); -if (sym) -return symbol__name (sym); + sym =machine__find_kernel_symbol (machine, key->lock_addr_or_cgroup,NULL ); +if (sym) { +snprintf (name_buf,sizeof (name_buf),"%s" ,symbol__name (sym)); +symbol__put (sym); +return name_buf; + }/* try semi-global locks collected separately */ if (!bpf_map_lookup_elem (lock_fd, &key->lock_addr_or_cgroup, &flags)) {
@@ -655,12 +666,14 @@ static const char *lock_contention_get_name(struct lock_contention *con, offset =map__map_ip (kmap, addr) -symbol__start (sym);if (offset ==0 ) -return symbol__name (sym); - -snprintf (name_buf,sizeof (name_buf),"%s+%#lx" ,symbol__name (sym), offset); +snprintf (name_buf,sizeof (name_buf),"%s" ,symbol__name (sym)); +else +snprintf (name_buf,sizeof (name_buf),"%s+%#lx" ,symbol__name (sym), offset); +symbol__put (sym); }else {snprintf (name_buf,sizeof (name_buf),"%#lx" , (unsigned long )addr); } +map__put (kmap);return name_buf; }
tools/perf/util/c2c-function.c +8 -5
@@ -976,13 +976,13 @@ c2c_function_hists__level1_entry(struct symbol *sym,/* mem_info__put() will map_symbol__exit() these, so take refs. */ mem_info__iaddr (mi)->ms.thread =thread__get (detail_he->ms.thread);mem_info__iaddr (mi)->ms.map =map__get (detail_he->ms.map); -mem_info__iaddr (mi)->ms.sym = sym; +mem_info__iaddr (mi)->ms.sym =symbol__get (sym);mem_info__daddr (mi)->addr =0 ;addr_location__init (&al); al.thread =thread__get (synthetic_thread); al.map =map__get (detail_he->ms.map); - al.sym = sym; + al.sym =symbol__get (sym); al.addr = sym_start; al.level = detail_he->level; al.cpumode = detail_he->cpumode;
@@ -1064,9 +1064,12 @@ c2c_function_hists__level2_entry(struct c2c_hist_entry *level1_c2c,return NULL ;/* Key this level by the looked-up symbol, not detail_he's. */ - level2_c2c->he.ms.sym = sym; -if (level2_c2c->he.mem_info) -mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym = sym; +symbol__put (level2_c2c->he.ms.sym); + level2_c2c->he.ms.sym =symbol__get (sym); +if (level2_c2c->he.mem_info) { +symbol__put (mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym); +mem_info__iaddr (level2_c2c->he.mem_info)->ms.sym =symbol__get (sym); + }c2c_child_entry__insert (level1_he, &level2_c2c->he, p, parent, leftmost);
tools/perf/util/call-path.c +12 -3
@@ -10,12 +10,13 @@#include <stdlib.h> #include "call-path.h" +#include "symbol.h" static void call_path__init (struct call_path *cp,struct call_path *parent,struct symbol *sym,u64 ip,bool in_kernel) { cp->parent = parent; - cp->sym = sym; + cp->sym =symbol__get (sym); cp->ip = sym ?0 : ip; cp->db_id =0 ; cp->in_kernel = in_kernel;
@@ -38,8 +39,15 @@ struct call_path_root *call_path_root__new(void)void call_path_root__free (struct call_path_root *cpr) {struct call_path_block *pos, *n; +size_t i =0 ; +if (!cpr) +return ; + +symbol__zput (cpr->call_path.sym);list_for_each_entry_safe (pos, n, &cpr->blocks, node) { +for (size_t j =0 ; j <CALL_PATH_BLOCK_SIZE && i < cpr->next; j++, i++) +symbol__zput (pos->cp[j].sym);list_del_init (&pos->node);free (pos); }
@@ -94,10 +102,11 @@ struct call_path *call_path__findnew(struct call_path_root *cpr, node_parent = *p; cp = rb_entry(node_parent, struct call_path, rb_node); -if (cp->sym == sym && cp->ip == ip) +if (RC_CHK_EQUAL (cp->sym, sym) && cp->ip == ip)return cp; -if (sym < cp->sym || (sym == cp->sym && ip < cp->ip)) +if ((sym ?RC_CHK_ACCESS (sym) :NULL ) < (cp->sym ?RC_CHK_ACCESS (cp->sym) :NULL ) || + (RC_CHK_EQUAL (sym, cp->sym) && ip < cp->ip)) p = &(*p)->rb_left;else p = &(*p)->rb_right;
tools/perf/util/callchain.c +3 -1
@@ -1094,6 +1094,7 @@ merge_chain_branch(struct callchain_cursor *cursor,struct map_symbol ms = { .thread =thread__get (list->ms.thread), .map =map__get (list->ms.map), + .sym =symbol__get (list->ms.sym), };callchain_cursor_append (cursor, list->ip, &ms, false,NULL ,0 ,0 ,0 , list->srcline);list_del_init (&list->list);
@@ -1204,7 +1205,8 @@ int fill_callchain_info(struct addr_location *al, struct callchain_cursor_node *map__put (al->map); al->map =map__get (node->ms.map); - al->sym = node->ms.sym; +symbol__put (al->sym); + al->sym =symbol__get (node->ms.sym); al->srcline = node->srcline; al->addr = node->ip;
tools/perf/util/capstone.c +2 -0
@@ -373,6 +373,7 @@ static void print_capstone_detail(struct cs_insn *insn, char *buf, size_t len,scnprintf (buf, len,"\t# %" PRIx64" <%s+%#" PRIx64">" , orig_addr,symbol__name (sym), sym_offset); } +symbol__put (sym);map__put (found_map);break ; }
@@ -459,6 +460,7 @@ static void format_capstone_insn_arm64(struct cs_insn *insn, char *buf,scnprintf (last_imm, len,"%" PRIx64" <%s+%#" PRIx64">" , addr,symbol__name (sym), sym_offset); } +symbol__put (sym);map__put (found_map); } }
tools/perf/util/db-export.c +31 -8
@@ -13,6 +13,8 @@#include <linux/err.h> +#include <internal/rc_check.h> +#include "call-path.h" #include "callchain.h" #include "comm.h"
@@ -26,14 +28,24 @@#include "thread-stack.h" #include "thread.h" +/* +* Hash and compare the underlying RC_STRUCT(symbol) pointer so all +* REFCNT_CHECKING wrapper pointers for the same symbol resolve to the same +* exported database ID. +*/ static size_t sym_ptr_hash (long key,void *ctx__maybe_unused ) { -return hash_bits ((size_t )key,16 ); +const struct symbol *sym = (const struct symbol *)key; + +return hash_bits ((size_t )RC_CHK_ACCESS (sym),16 ); }static bool sym_ptr_equal (long key1,long key2,void *ctx__maybe_unused ) { -return key1 == key2; +const struct symbol *sym1 = (const struct symbol *)key1; +const struct symbol *sym2 = (const struct symbol *)key2; + +return RC_CHK_EQUAL (sym1, sym2); }int db_export__init (struct db_export *dbe)
@@ -44,8 +56,15 @@ int db_export__init(struct db_export *dbe)void db_export__exit (struct db_export *dbe) { -hashmap__free (dbe->sym_db_ids); - dbe->sym_db_ids =NULL ; +if (dbe->sym_db_ids) { +struct hashmap_entry *cur; +size_t bkt; + +hashmap__for_each_entry (dbe->sym_db_ids, cur, bkt) +symbol__put ((struct symbol *)cur->pkey); +hashmap__free (dbe->sym_db_ids); + dbe->sym_db_ids =NULL ; + }call_return_processor__free (dbe->crp); dbe->crp =NULL ; }
@@ -191,6 +210,7 @@ int db_export__symbol(struct db_export *dbe, struct symbol *sym, struct dso *dso) {long db_id; +struct symbol *sym_key;if (db_export__symbol_db_id (dbe, sym))return 0 ;
@@ -207,8 +227,11 @@ int db_export__symbol(struct db_export *dbe, struct symbol *sym, } db_id = ++dbe->symbol_last_db_id; -if (hashmap__set (dbe->sym_db_ids, (long )sym, db_id,NULL ,NULL )) + sym_key =symbol__get (sym); +if (hashmap__set (dbe->sym_db_ids, (long )sym_key, db_id,NULL ,NULL )) { +symbol__put (sym_key);return -ENOMEM ; + }if (dbe->export_symbol)return dbe->export_symbol (dbe, sym, dso);
@@ -231,8 +254,8 @@ static int db_ids_from_al(struct db_export *dbe, struct addr_location *al,if (!al->sym) { al->sym =symbol__new (al->addr,0 ,0 ,0 ,"unknown" ); -if (al->sym &&dso__insert_symbol (dso, al->sym)) - al->sym =NULL ; +if (al->sym &&dso__insert_symbol (dso,symbol__get (al->sym))) +symbol__zput (al->sym); }if (al->sym) {
@@ -292,7 +315,7 @@ static struct call_path *call_path_from_sample(struct db_export *dbe,* db_ids_from_al() to perform the export. */ addr_location__init (&al); - al.sym = node->ms.sym; + al.sym =symbol__get (node->ms.sym); al.map =map__get (node->ms.map); al.addr = node->ip; al.thread =thread__get (thread);
tools/perf/util/disasm.c +15 -7
@@ -228,6 +228,7 @@ static void ins_ops__delete(struct ins_operands *ops)zfree (&ops->source.name);zfree (&ops->target.raw);zfree (&ops->target.name); +symbol__zput (ops->target.sym); } int ins__raw_scnprintf(const struct ins *ins, char *bf, size_t size,
@@ -289,8 +290,10 @@ static int call__parse(const struct arch *arch, struct ins_operands *ops, struct };if (maps__find_ams (thread__maps (ms->thread), &target) ==0 && -map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; +map__rip_2objdump (target.ms.map, +map__map_ip (target.ms.map, target.addr)) == ops->target.addr && + !RC_CHK_EQUAL (target.ms.sym, ms->sym)) + ops->target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target);return 0 ;
@@ -424,9 +427,10 @@ static int jump__parse(const struct arch *arch, struct ins_operands *ops, struct* Actual navigation will come next, with further understanding of how * the symbol searching and disassembly should be done. */ -if (maps__find_ams (thread__maps (ms->thread), &target) ==0 && +if (ops->target.outside && +maps__find_ams (thread__maps (ms->thread), &target) ==0 &&map__rip_2objdump (target.ms.map,map__map_ip (target.ms.map, target.addr)) == ops->target.addr) - ops->target.sym = target.ms.sym; + ops->target.sym =symbol__get (target.ms.sym);if (!ops->target.outside) { ops->target.offset = target.addr - start;
@@ -473,12 +477,14 @@ int jump__scnprintf(const struct ins *ins, char *bf, size_t size, ops->target.offset); } -void jump__delete (struct ins_operands *ops__maybe_unused ) +void jump__delete (struct ins_operands *ops) {/* * The ops->jump.raw_comment and ops->jump.raw_func_start belong to the * raw string, don't free them. */ +if (ops) +symbol__zput (ops->target.sym); }const struct ins_ops jump_ops = {
@@ -572,6 +578,7 @@ static void lock__delete(struct ins_operands *ops)zfree (&ops->locked.ins.name);zfree (&ops->target.raw);zfree (&ops->target.name); +symbol__zput (ops->target.sym); }const struct ins_ops lock_ops = {
@@ -1090,8 +1097,9 @@ static int symbol__parse_objdump_line(struct symbol *sym, };if (!maps__find_ams (thread__maps (args->ms->thread), &target) && -symbol__start (target.ms.sym) == target.al_addr) - dl->ops.target.sym = target.ms.sym; +symbol__start (target.ms.sym) == target.al_addr && + !RC_CHK_EQUAL (target.ms.sym, sym)) + dl->ops.target.sym =symbol__get (target.ms.sym);addr_map_symbol__exit (&target); }
tools/perf/util/event.c +4 -4
@@ -770,7 +770,7 @@ struct map *thread__find_map_fb(struct thread *thread, u8 cpumode, u64 addr,struct symbol *thread__find_symbol (struct thread *thread,u8 cpumode,u64 addr,struct addr_location *al) { - al->sym =NULL ; +symbol__zput (al->sym);if (thread__find_map (thread, cpumode, addr, al)) al->sym =map__find_symbol (al->map, al->addr);return al->sym;
@@ -779,7 +779,7 @@ struct symbol *thread__find_symbol(struct thread *thread, u8 cpumode,struct symbol *thread__find_symbol_fb (struct thread *thread,u8 cpumode,u64 addr,struct addr_location *al) { - al->sym =NULL ; +symbol__zput (al->sym);if (thread__find_map_fb (thread, cpumode, addr, al)) al->sym =map__find_symbol (al->map, al->addr);return al->sym;
@@ -829,7 +829,7 @@ int machine__resolve(struct machine *machine, struct addr_location *al,thread__put (thread); thread =NULL ; - al->sym =NULL ; +symbol__zput (al->sym); al->cpu = sample->cpu; al->socket =-1 ; al->srcline =NULL ;
@@ -937,7 +937,7 @@ void thread__resolve(struct thread *thread, struct addr_location *al,thread__find_map_fb (thread, sample->cpumode, sample->addr, al); al->cpu = sample->cpu; - al->sym =NULL ; +symbol__zput (al->sym);if (al->map) al->sym =map__find_symbol (al->map, al->addr);
tools/perf/util/hist.c +12 -1
@@ -543,6 +543,7 @@ static int hist_entry__init(struct hist_entry *he, he->ms.thread =thread__get (he->ms.thread); he->ms.map =map__get (he->ms.map); + he->ms.sym =symbol__get (he->ms.sym);if (he->branch_info) {/*
@@ -559,8 +560,10 @@ static int hist_entry__init(struct hist_entry *he, he->branch_info->from.ms.thread =thread__get (he->branch_info->from.ms.thread); he->branch_info->from.ms.map =map__get (he->branch_info->from.ms.map); + he->branch_info->from.ms.sym =symbol__get (he->branch_info->from.ms.sym); he->branch_info->to.ms.thread =thread__get (he->branch_info->to.ms.thread); he->branch_info->to.ms.map =map__get (he->branch_info->to.ms.map); + he->branch_info->to.ms.sym =symbol__get (he->branch_info->to.ms.sym); }if (he->mem_info) {
@@ -593,6 +596,7 @@ static int hist_entry__init(struct hist_entry *he,INIT_LIST_HEAD (&he->pairs.node); he->thread =thread__get (he->thread); + he->parent =symbol__get (he->parent); he->hroot_in =RB_ROOT_CACHED ; he->hroot_out =RB_ROOT_CACHED ;
@@ -728,6 +732,8 @@ static struct hist_entry *hists__findnew_entry(struct hists *hists,if (hists__has (hists, sym) && he->ms.map != entry->ms.map) {if (he->ms.sym) {u64 addr =symbol__start (he->ms.sym); + +symbol__put (he->ms.sym); he->ms.sym =map__find_symbol (entry->ms.map, addr); }
@@ -1029,7 +1035,8 @@ iter_next_branch_entry(struct hist_entry_iter *iter, struct addr_location *al) al->thread =thread__get (bi[i].to.ms.thread);map__put (al->map); al->map =map__get (bi[i].to.ms.map); - al->sym = bi[i].to.ms.sym; +symbol__put (al->sym); + al->sym =symbol__get (bi[i].to.ms.sym); al->addr = bi[i].to.addr;return 1 ; }
@@ -1343,9 +1350,11 @@ int hist_entry_iter__add(struct hist_entry_iter *iter, struct addr_location *al,if (al) alm =map__get (al->map); + iter->parent =NULL ; err =sample__resolve_callchain (iter->sample,get_tls_callchain_cursor (), &iter->parent, al, max_stack_depth);if (err) { +symbol__zput (iter->parent);map__put (alm);return err; }
@@ -1381,6 +1390,7 @@ int hist_entry_iter__add(struct hist_entry_iter *iter, struct addr_location *al,if (!err) err = err2; +symbol__zput (iter->parent);map__put (alm);return err;
@@ -1467,6 +1477,7 @@ void hist_entry__delete(struct hist_entry *he) }thread__zput (he->thread); +symbol__zput (he->parent);map_symbol__exit (&he->ms);if (he->branch_info) {
tools/perf/util/libbfd.c +6 -1
@@ -246,7 +246,12 @@ static int inline_list__append_dso_a2l(struct dso *dso,if (a2l->filename) srcline =srcline_from_fileline (a2l->filename, a2l->line); -return inline_list__append (inline_sym, srcline, node); +if (inline_list__append (inline_sym, srcline, node) !=0 ) { +zfree_srcline (&srcline); +symbol__put (inline_sym); +return -1 ; + } +return 0 ; } int libbfd__addr2line(const char *dso_name, u64 addr,
tools/perf/util/libdw.c +2 -3
@@ -99,7 +99,7 @@ static int libdw_a2l_cb(Dwarf_Die *die, void *_args)* while DWARF keeps the pre-clone linkage name, which * left the outer frame spuriously tagged "(inlined)". */ - inline_sym = args->sym; + inline_sym =symbol__get (args->sym); }else {/* * Prefer DW_AT_linkage_name so C++ inline frames keep
@@ -157,8 +157,7 @@ static int libdw_a2l_cb(Dwarf_Die *die, void *_args)return 0 ; abort_delete_sym: -if (symbol__inlined (inline_sym)) -symbol__delete (inline_sym); +symbol__put (inline_sym); abort_enomem: args->err = -ENOMEM ;return DWARF_CB_ABORT ;
tools/perf/util/llvm.c +2 -0
@@ -63,6 +63,8 @@ int llvm__addr2line(const char *dso_name __maybe_unused, u64 addr __maybe_unused inline_frames[i].line); }if (inline_list__append (inline_sym, srcline, node) !=0 ) { +zfree_srcline (&srcline); +symbol__put (inline_sym);if (file)zfree (file);inline_node__clear_frames (node);
tools/perf/util/lock-contention.c +5 -3
@@ -101,7 +101,6 @@ struct lock_stat *lock_stat_findnew(u64 addr, const char *name, int flags)bool match_callstack_filter (struct machine *machine,u64 *callstack,int max_stack_depth) { -struct map *kmap;struct symbol *sym;u64 ip;uint16_t e_machine =perf_env__e_machine (machine->env,/*e_flags=*/ NULL );
@@ -130,14 +129,17 @@ bool match_callstack_filter(struct machine *machine, u64 *callstack, int max_stabreak ; ip = callstack[i]; - sym =machine__find_kernel_symbol (machine, ip, &kmap); + sym =machine__find_kernel_symbol (machine, ip,NULL );if (sym ==NULL )continue ;list_for_each_entry (filter, &callstack_filters, list) { -if (strstr (symbol__name (sym), filter->name)) +if (strstr (symbol__name (sym), filter->name)) { +symbol__put (sym);return true; + } } +symbol__put (sym); }return false; }
tools/perf/util/machine.c +40 -16
149 lines not on this page · raw message
tools/perf/util/map.h +6 -4
@@ -160,13 +160,15 @@ struct thread;* @sym_name: the symbol name * @pos: the 'struct symbol *' to use as a loop cursor * @idx: the cursor index in the symbol names array +* Note: caller must call symbol__put(pos) if breaking out of the loop early. */ #define __map__for_each_symbol_by_name (map, sym_name, pos, idx) \ for (pos = map__find_symbol_by_name_idx(map, sym_name, &idx); \ - pos && \ - !symbol__match_symbol_name(symbol__name(pos), sym_name, \ - SYMBOL_TAG_INCLUDE__DEFAULT_ONLY); \ - pos = dso__next_symbol_by_name(map__dso(map), &idx)) + pos && \ + (!symbol__match_symbol_name(symbol__name(pos), sym_name, \ + SYMBOL_TAG_INCLUDE__DEFAULT_ONLY) || \ + (symbol__zput(pos), false)); \ + symbol__put(pos), pos = dso__next_symbol_by_name(map__dso(map), &idx))#define map__for_each_symbol_by_name (map, sym_name, pos, idx) \ __map__for_each_symbol_by_name(map, sym_name, (pos), idx)
tools/perf/util/map_symbol.c +3 -1
@@ -2,12 +2,14 @@#include "map_symbol.h" #include "maps.h" #include "map.h" +#include "symbol.h" #include "thread.h" void map_symbol__exit (struct map_symbol *ms) {thread__zput (ms->thread);map__zput (ms->map); +symbol__zput (ms->sym); }void addr_map_symbol__exit (struct addr_map_symbol *ams)
@@ -19,7 +21,7 @@ void map_symbol__copy(struct map_symbol *dst, struct map_symbol *src) { dst->thread =thread__get (src->thread); dst->map =map__get (src->map); - dst->sym = src->sym; + dst->sym =symbol__get (src->sym); }void addr_map_symbol__copy (struct addr_map_symbol *dst,struct addr_map_symbol *src)
tools/perf/util/maps.c +2 -1
@@ -788,7 +788,7 @@ struct symbol *maps__find_symbol_by_name(struct maps *maps, const char *name, st *mapp =map__get (map);break ; } - sym =NULL ; +symbol__zput (sym); }for (unsigned int i =0 ; i < nr_maps; i++)
@@ -810,6 +810,7 @@ int maps__find_ams(struct maps *maps, struct addr_map_symbol *ams) } ams->al_addr =map__map_ip (ams->ms.map, ams->addr); +symbol__put (ams->ms.sym); ams->ms.sym =map__find_symbol (ams->ms.map, ams->al_addr);return ams->ms.sym ?0 :-1 ;
tools/perf/util/probe-event.c +14 -3
@@ -152,6 +152,8 @@ static int kernel_get_symbol_address_by_name(const char *name, u64 *addr, *addr =map__unmap_ip (map,symbol__start (sym)) - ((reloc) ?0 :map__reloc (map)) - ((reladdr) ?map__start (map) :0 ); +symbol__put (sym); +map__put (map); }return 0 ; }
@@ -422,6 +424,7 @@ static int find_alternative_probe_point(struct debuginfo *dinfo, pp->function); }else address =map__unmap_ip (map,symbol__start (sym)) -map__reloc (map); +symbol__put (sym);break ; }if (!address) {
@@ -728,11 +731,14 @@ static int post_process_probe_trace_point(struct probe_trace_point *tp,else free (tp->symbol); tp->symbol =strdup (symbol__name (sym)); -if (!tp->symbol) +if (!tp->symbol) { +symbol__put (sym);return -ENOMEM ; + } } tp->offset = addr -symbol__start (sym); tp->address -= offs; +symbol__put (sym);return 0 ; }
@@ -2314,6 +2320,7 @@ static int find_perf_probe_point_from_map(struct probe_trace_point *tp, ret = pp->function ?0 : -ENOMEM ; out: +symbol__put (sym);map__put (map);return ret;
@@ -3078,7 +3085,7 @@ static int find_probe_functions_cb(struct symbol *sym, void *data)if (strglobmatch (norm, args->name)) { args->found++;if (args->syms && args->found < probe_conf.max_probes) - args->syms[args->found -1 ] = sym; + args->syms[args->found -1 ] =symbol__get (sym); }free (buf);return 0 ;
@@ -3146,7 +3153,7 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, goto out; } - syms =malloc (sizeof (struct symbol *) * probe_conf.max_probes); + syms =calloc (probe_conf.max_probes,sizeof (struct symbol *));if (!syms) { ret = -ENOMEM ; goto out;
@@ -3290,6 +3297,10 @@ static int find_probe_trace_events_from_map(struct perf_probe_event *pev, } out: +if (syms) { +for (i =0 ; i < probe_conf.max_probes; i++) +symbol__put (syms[i]); + }map__put (map);free (syms);return ret;
tools/perf/util/python.c +6 -14
@@ -882,7 +882,7 @@ static PyObject *pyrf_sample_event__srccode(PyObject *self, PyObject *args)addr_location__init (&al); al.thread =thread__get (pevent->al.thread); al.map =map__get (pevent->al.map); - al.sym = pevent->al.sym; + al.sym =symbol__get (pevent->al.sym); al.addr = pevent->al.addr; }
@@ -938,13 +938,12 @@ struct pyrf_callchain_node {u64 ip;struct map *map;struct symbol *sym; -char *sym_name; };static void pyrf_callchain_node__delete (struct pyrf_callchain_node *pnode) {map__put (pnode->map); -free (pnode->sym_name); +symbol__put (pnode->sym);Py_TYPE (pnode)->tp_free ((PyObject *)pnode); }
@@ -957,8 +956,6 @@ static PyObject *pyrf_callchain_node__get_ip(struct pyrf_callchain_node *pnode,static PyObject *pyrf_callchain_node__get_symbol (struct pyrf_callchain_node *pnode,void *closure__maybe_unused ) { -if (pnode->sym_name) -return PyUnicode_FromString (pnode->sym_name);if (pnode->sym)return PyUnicode_FromString (symbol__name (pnode->sym));return PyUnicode_FromString ("[unknown]" );
@@ -1016,7 +1013,6 @@ struct pyrf_callchain_frame {u64 ip;struct map *map;struct symbol *sym; -char *sym_name; };struct pyrf_callchain {
@@ -1030,7 +1026,7 @@ static void pyrf_callchain__delete(struct pyrf_callchain *pchain)if (pchain->frames) {for (u64 i =0 ; i < pchain->nr_frames; i++) {map__put (pchain->frames[i].map); -free (pchain->frames[i].sym_name); +symbol__put (pchain->frames[i].sym); }free (pchain->frames); }
@@ -1060,8 +1056,7 @@ static PyObject *pyrf_callchain__item(PyObject *obj, Py_ssize_t i) pnode->ip = pchain->frames[i].ip; pnode->map =map__get (pchain->frames[i].map); - pnode->sym = pchain->frames[i].sym; - pnode->sym_name = pchain->frames[i].sym_name ?strdup (pchain->frames[i].sym_name) :NULL ; + pnode->sym =symbol__get (pchain->frames[i].sym);return (PyObject *)pnode; }
@@ -1114,7 +1109,7 @@ static int pyrf_sample_event__resolve_callchain(struct pyrf_event *pevent) node =callchain_cursor_current (cursor); pchain->frames[i].ip = node->ip; pchain->frames[i].map =map__get (node->ms.map); - pchain->frames[i].sym = node->ms.sym; + pchain->frames[i].sym =symbol__get (node->ms.sym);callchain_cursor_advance (cursor); } pevent->callchain = (PyObject *)pchain;
@@ -4504,8 +4499,7 @@ static PyObject *pyrf_call_path__to_callchain(const struct call_path *cp,struct pyrf_callchain_frame *frame = &pchain->frames[--nr_frames]; frame->ip = pos->ip; - frame->sym =NULL ; - frame->sym_name = pos->sym ?strdup (symbol__name (pos->sym)) :NULL ; + frame->sym =symbol__get (pos->sym); frame->map =NULL ;if (thread) {struct addr_location al;
@@ -4515,8 +4509,6 @@ static PyObject *pyrf_call_path__to_callchain(const struct call_path *cp,addr_location__init (&al);thread__find_map (thread, cpumode, pos->ip, &al); frame->map =map__get (al.map); -if (frame->map) - frame->sym = pos->sym;addr_location__exit (&al); } }
tools/perf/util/sort.c +1 -1
@@ -466,7 +466,7 @@ int64_t _sort__sym_cmp(struct symbol *sym_l, struct symbol *sym_r)if (!sym_l || !sym_r)return cmp_null (sym_l, sym_r); -if (sym_l == sym_r) +if (RC_CHK_EQUAL (sym_l, sym_r))return 0 ;if (symbol__inlined (sym_l) ||symbol__inlined (sym_r)) {
tools/perf/util/srcline.c +2 -4
@@ -118,7 +118,7 @@ struct symbol *new_inline_sym(struct dso *dso,if (base_sym &&strcmp (funcname,symbol__name (base_sym)) ==0 ) {/* reuse the real, existing symbol */ - inline_sym = base_sym; + inline_sym =symbol__get (base_sym);/* ensure that we don't alias an inlined symbol, which could * lead to double frees in inline_node__delete */
@@ -478,9 +478,7 @@ void inline_node__clear_frames(struct inline_node *node)list_for_each_entry_safe (ilist, tmp, &node->val, list) {list_del_init (&ilist->list);zfree_srcline (&ilist->srcline); -/* only the inlined symbols are owned by the list */ -if (ilist->symbol &&symbol__inlined (ilist->symbol)) -symbol__delete (ilist->symbol); +symbol__zput (ilist->symbol);free (ilist); } }
tools/perf/util/symbol-elf.c +5 -1
@@ -352,10 +352,13 @@ static bool get_ifunc_name(Elf *elf, struct dso *dso, GElf_Ehdr *ehdr,/* Expecting the address to be an IFUNC or IFUNC alias */ if (!sym ||symbol__start (sym) != addr || - (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym))) + (symbol__type (sym) !=STT_GNU_IFUNC && !symbol__ifunc_alias (sym))) { +symbol__put (sym);return false; + }snprintf (buf, buf_sz,"%s@plt" ,symbol__name (sym)); +symbol__put (sym);return true; }
@@ -633,6 +636,7 @@ int dso__synthesize_plt_symbols(struct dso *dso, struct symsrc *ss) f =dso__find_symbol_nocache (dso, shdr_plt.sh_offset);if (f &&symbol__start (f) < shdr_plt.sh_offset &&symbol__end (f) > shdr_plt.sh_offset)symbol__set_end (f, shdr_plt.sh_offset); +symbol__put (f);if (!get_plt_sizes (dso, &ehdr, &shdr_plt, &plt_header_size, &plt_entry_size))return 0 ;
tools/perf/util/symbol.c +66 -41
241 lines not on this page · raw message
tools/perf/util/symbol.h +44 -16
146 lines not on this page · raw message
tools/perf/util/thread-stack.c +6 -6
@@ -823,7 +823,7 @@ static int thread_stack__pop_cp(struct thread *thread, struct thread_stack *ts,if (ts->cnt ==1 ) {struct thread_stack_entry *tse = &ts->stack[0 ]; -if (tse->cp->sym == sym) +if (RC_CHK_EQUAL (tse->cp->sym, sym))return thread_stack__call_return (thread, ts, --ts->cnt, timestamp, ref, false); }
@@ -941,7 +941,7 @@ static int thread_stack__no_call_return(struct thread *thread,else parent = root; -if (parent->sym == from_al->sym) { +if (RC_CHK_EQUAL (parent->sym, from_al->sym)) {/* * At the bottom of the stack, assume the missing 'call' was * before the trace started. So, pop the current symbol and push
@@ -1069,7 +1069,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts,*/ ts->cnt -=1 ; sym = ts->stack[ts->cnt -2 ].cp->sym; -if (sym && sym == tsym && to_al->addr !=symbol__start (tsym)) { +if (sym &&RC_CHK_EQUAL (sym, tsym) && to_al->addr !=symbol__start (tsym)) {/* * Target is back to the middle of the symbol we came * from so assume it is an indirect jmp and forget it
@@ -1078,7 +1078,7 @@ static int thread_stack__x86_retpoline(struct thread_stack *ts, ts->cnt -=1 ;return 0 ; } - }else if (sym && sym == tsym) { + }else if (sym &&RC_CHK_EQUAL (sym, tsym)) {/* * Target is back to the symbol we came from so assume it is an * indirect jmp and forget it altogether.
@@ -1169,7 +1169,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm,* may be the start of a x86 retpoline. */ if (!err && rstate ==X86_RETPOLINE_POSSIBLE && to_al->sym && - from_al->sym == to_al->sym && +RC_CHK_EQUAL (from_al->sym, to_al->sym) && to_al->addr !=symbol__start (to_al->sym)) ts->rstate =X86_RETPOLINE_DETECTED ;
@@ -1206,7 +1206,7 @@ int thread_stack__process(struct thread *thread, struct comm *comm, }else if (sample->flags &PERF_IP_FLAG_TRACE_END ) { err =thread_stack__trace_end (ts, sample, ref); }else if (sample->flags &PERF_IP_FLAG_BRANCH && - from_al->sym != to_al->sym && to_al->sym && + !RC_CHK_EQUAL (from_al->sym, to_al->sym) && to_al->sym && to_al->addr ==symbol__start (to_al->sym)) {struct call_path_root *cpr = ts->crp->cpr;struct call_path *cp;
tools/perf/util/threads.c +11 -6
@@ -142,10 +142,13 @@ void threads__remove_all_threads(struct threads *threads)down_write (&table->lock);__threads_table_entry__set_last_match (table,NULL );hashmap__for_each_entry_safe (&table->shard, cur, tmp, bkt) { -struct thread *old_value; +struct thread *old_value =NULL ; -hashmap__delete (&table->shard, cur->key,/*old_key=*/ NULL , &old_value); -thread__put (old_value); +if (hashmap__delete (&table->shard, cur->key, +/*old_key=*/ NULL , &old_value)) { +thread__free_stitch_list (old_value); +thread__put (old_value); + } }up_write (&table->lock); }
@@ -154,14 +157,16 @@ void threads__remove_all_threads(struct threads *threads)void threads__remove (struct threads *threads,struct thread *thread) {struct threads_table_entry *table =threads__table (threads,thread__tid (thread)); -struct thread *old_value; +struct thread *old_value =NULL ;down_write (&table->lock);if (table->last_match &&RC_CHK_EQUAL (table->last_match, thread))__threads_table_entry__set_last_match (table,NULL ); -hashmap__delete (&table->shard,thread__tid (thread),/*old_key=*/ NULL , &old_value); -thread__put (old_value); +if (hashmap__delete (&table->shard,thread__tid (thread),/*old_key=*/ NULL , &old_value)) { +thread__free_stitch_list (old_value); +thread__put (old_value); + }up_write (&table->lock); }
tools/perf/util/unwind-libdw.c +1 -1
@@ -169,7 +169,7 @@ static int entry(u64 ip, struct unwind_info *ui) e->ip = ip; e->ms.thread =thread__get (al.thread); e->ms.map =map__get (al.map); - e->ms.sym = al.sym; + e->ms.sym =symbol__get (al.sym); pr_debug("unwind: %s:ip = 0x%" PRIx64 " (0x%" PRIx64 ")\n", al.sym ? symbol__name(al.sym) : "''",
tools/perf/util/unwind-libunwind.c +1 -1
@@ -632,7 +632,7 @@ static int entry(uint64_t ip, struct thread *thread, unwind_entry_cb_t cb, void e.ms.thread =thread__get (al.thread);pr_debug ("unwind: %s:ip = 0x%" PRIx64" (0x%" PRIx64")\n" , - al.sym ? al.sym->name :"''" , + al.sym ?symbol__name (al.sym) :"''" , ip, al.map ?map__map_ip (al.map, ip) : (u64 )0 );
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 08/18] perf symbol: Add reference counting and DECLARE_RC_STRUCT(symbol)
Quoted 5 lines · Ian Rogers
struct
map_symbolms = {
.thread =thread__get(list->ms.thread),
.map =map__get(list->ms.map),
+ .sym =symbol__get(list->ms.sym),
};
This fixes a real bug and worth it's own patch ahead of the conversion series so it can go to the LTS independently imo.
Since commit bffb5b0c0976a ("perf map/maps/thread:
Changes to reference counting") the temporary map_symbol in merge_chain_branch() has had no .sym, so callchains merged during collapse lose their symbol. With this line, the default 'perf report --children' output changes on recordings with inlined
frames: a line like "|--35.41%--0x55566d988a0f" becomes "bench_sched_messaging" / "group (inlined)", with different percentages.
Quoted 3 lines
-
frame->sym= NULL;
-frame->sym_name=pos->sym? strdup(symbol__name(pos->sym)) : NULL;
+frame->sym=symbol__get(pos->sym);
[...]
Quoted 2 lines
- if (
frame->map)
-frame->sym=pos->sym;
Before this patch, frame->sym was only set when thread__find_map() found a map. call_path__init() stores ip = 0 whenever the call path has a symbol, so for those frames no map is found, and sym_offset now returns 0 - symbol__start(sym) instead of None. Returning None from pyrf_callchain_node__get_sym_offset() when there is no map would keep the old behaviour.
Smaller things related to this patch:
- call__parse() no longer records a call to the same function, so Enter on a recursive call in TUI annotate says "The called function was not found".
- builtin-annotate.c now takes a symbol reference for each block_range entry, but the process-global block_ranges tree has no teardown. Patch 15/18 retains symbols with a refcount greater than one, so any symbol still referenced by this tree is permanently ineligible for shrinking. Could block_range get a cleanup path that puts the symbol references and frees the tree?
Thanks,
Alireza
Ian Rogers
reviewerPATCH v2 9/18 perf symbol: Optimize symbol sorting with natural-run merge sort and binary insertion
In reply to Ian RogersELF .symtab and .dynsym sections are typically already mostly ordered by address in ascending runs, and dynamic symbol resolution (dso__insert_symbol()) frequently inserts one or a few symbols into an already-sorted DSO symbol table. Invoking full-array qsort() on every sort after incremental insertions or across pre-ordered ELF runs spends avoidable CPU time in symbol comparisons.
Speed up symbol array sorting and out-of-order symbol insertions:
- Track the sorted prefix length (@sorted_cnt) in struct symbols and replace full-array qsort() with a natural-run merge sort (symbols__sort_range_with_tmp()) and galloping binary-search merge (symbols__merge_runs()), plus an O(N) in-place binary insertion fast path when <= 8 symbols are appended.
- Insert directly into sorted position in dso__insert_symbol() when symbols->sorted is true so dynamic single-symbol insertions maintain the sorted invariant without triggering full re-sorts.
Using 'perf bench internals symbols':
Before:
Average DSO load + sort: 6.242 msec (+- 0.057 msec)
Average bulk insert + lookup: 1404.150 usec (+- 18.752 usec)
Average append + lookup: 40429.750 usec (+- 507.578 usec)
After (-7.0% load, -13.9% bulk, -99.6% append):
Average DSO load + sort: 5.808 msec (+- 0.047 msec)
Average bulk insert + lookup: 1209.250 usec (+- 18.446 usec)
Average append + lookup: 152.900 usec (+- 0.736 usec)--- tools/perf/util/symbol.c | 434 ++++++++++++++++++++++++++++++++++++--- tools/perf/util/symbol.h | 5 + 2 files changed, 405 insertions(+), 34 deletions(-)
tools/perf/util/symbol.c +400 -34
@@ -220,6 +220,7 @@ void symbols__init(struct symbols *symbols) symbols->symbols =NULL ; symbols->symbols_by_name =NULL ; symbols->cnt =0 ; + symbols->sorted_cnt =0 ; symbols->allocated =0 ; symbols->num_by_name =0 ; symbols->sorted = true;
@@ -234,6 +235,7 @@ static void symbols__delete_locked(struct symbols *symbols)zfree (&symbols->symbols);zfree (&symbols->symbols_by_name); symbols->cnt =0 ; + symbols->sorted_cnt =0 ; symbols->allocated =0 ; symbols->num_by_name =0 ; symbols->sorted = true;
@@ -266,27 +268,312 @@ static inline bool symbol__less_or_equal(const struct symbol *a, const struct syreturn symbol__end (a) <=symbol__end (b); } +static unsigned int symbols__upper_bound (struct symbol **arr,unsigned int len, +const struct symbol *sym) +{ +unsigned int low =0 , high = len; + +while (low < high) { +unsigned int mid = low + (high - low) /2 ; + +if (symbol__less_or_equal (arr[mid], sym)) + low = mid +1 ; +else + high = mid; + } +return low; +} + +static unsigned int symbols__lower_bound (struct symbol **arr,unsigned int len, +const struct symbol *sym) +{ +unsigned int low =0 , high = len; + +while (low < high) { +unsigned int mid = low + (high - low) /2 ; + +if (!symbol__less_or_equal (sym, arr[mid])) + low = mid +1 ; +else + high = mid; + } +return low; +} + +static void symbols__insertion_sort_range (struct symbol **arr, +unsigned int start, +unsigned int end) +{ +for (unsigned int i = start +1 ; i < end; i++) { +struct symbol *sym = arr[i]; + +if (!symbol__less_or_equal (arr[i -1 ], sym)) { +unsigned int pos = start + +symbols__upper_bound (&arr[start], i - start, sym); + +memmove (&arr[pos +1 ], &arr[pos], + (i - pos) *sizeof (struct symbol *)); + arr[pos] = sym; + } + } +} + +static void symbols__merge_runs (struct symbol **arr, +unsigned int left, +unsigned int mid, +unsigned int right, +struct symbol **tmp) +{ +unsigned int first_left, last_left, left_len, right_len; +unsigned int i, j, dst; + +if (left >= mid || mid >= right) +return ; +if (symbol__less_or_equal (arr[mid -1 ], arr[mid])) +return ; + + first_left = left +symbols__upper_bound (&arr[left], mid - left, arr[mid]); + right_len = right - mid; + +if (mid - first_left <= right_len) { + left_len = mid - first_left; +memcpy (tmp, &arr[first_left], left_len *sizeof (struct symbol *)); + i =0 ; + j = mid; + dst = first_left; +while (i < left_len && j < right) { +if (symbol__less_or_equal (tmp[i], arr[j])) { + arr[dst++] = tmp[i++]; + }else { + arr[dst++] = arr[j++]; +if (right - j >=16 && + (right - j) >8 * (left_len - i) && + !symbol__less_or_equal (tmp[i], arr[j])) { +unsigned int step = +symbols__lower_bound (&arr[j], right - j, tmp[i]); + +if (step >0 ) { +memmove (&arr[dst], &arr[j], + step *sizeof (struct symbol *)); + dst += step; + j += step; + } + } + } + } +if (i < left_len) +memcpy (&arr[dst], &tmp[i], (left_len - i) *sizeof (struct symbol *)); +return ; + } + + last_left = first_left + +symbols__upper_bound (&arr[first_left], mid - first_left, arr[right -1 ]); + left_len = last_left - first_left; + +memcpy (tmp, &arr[mid], right_len *sizeof (struct symbol *)); +if (last_left < mid) { +memmove (&arr[last_left + right_len], &arr[last_left], + (mid - last_left) *sizeof (struct symbol *)); + } + +if (first_left == last_left) { +memcpy (&arr[first_left], tmp, right_len *sizeof (struct symbol *)); +return ; + } + + i = last_left; + j = right_len; + dst = last_left + right_len; +while (i > first_left && j >0 ) { +if (symbol__less_or_equal (arr[i -1 ], tmp[j -1 ])) { + arr[--dst] = tmp[--j]; + }else { + arr[--dst] = arr[--i]; +if (i - first_left >=16 && + (i - first_left) >8 * j && + !symbol__less_or_equal (arr[i -1 ], tmp[j -1 ])) { +unsigned int pos = first_left + +symbols__upper_bound (&arr[first_left], + i - first_left, + tmp[j -1 ]); +unsigned int step = i - pos; + +if (step >0 ) { + dst -= step; + i = pos; +memmove (&arr[dst], &arr[pos], + step *sizeof (struct symbol *)); + } + } + } + } +if (j >0 ) +memcpy (&arr[first_left], tmp, j *sizeof (struct symbol *)); +} + +struct symbol_run { +unsigned int start; +unsigned int len; +}; + +static void symbols__sort_range_with_tmp (struct symbol **arr,unsigned int n, +struct symbol **tmp) +{ +struct symbol_run runs[64 ]; +unsigned int nr_runs =0 ; +unsigned int i =0 ; + +if (n <=32 ) { +symbols__insertion_sort_range (arr,0 , n); +return ; + } + +while (i < n) { +unsigned int start = i++; + +if (i < n) { +if (symbol__less_or_equal (arr[i -1 ], arr[i])) { +while (i < n &&symbol__less_or_equal (arr[i -1 ], arr[i])) + i++; + }else { +while (i < n && !symbol__less_or_equal (arr[i -1 ], arr[i])) + i++; +for (unsigned int lo = start, hi = i -1 ; lo < hi; lo++, hi--) { +struct symbol *swap = arr[lo]; + + arr[lo] = arr[hi]; + arr[hi] = swap; + } + } + } + +if (i - start <32 && i < n) { + i =min (start +32U , n); +symbols__insertion_sort_range (arr, start, i); + } + + runs[nr_runs++] = (struct symbol_run ){ + .start = start, + .len = i - start, + }; + +while (nr_runs >1 ) { +unsigned int r = nr_runs -2 ; + +if (nr_runs >=3 && + runs[nr_runs -3 ].len <= + runs[nr_runs -2 ].len + runs[nr_runs -1 ].len) { +if (runs[nr_runs -3 ].len < runs[nr_runs -1 ].len) + r = nr_runs -3 ; + }else if (runs[nr_runs -2 ].len > runs[nr_runs -1 ].len && + nr_runs <ARRAY_SIZE (runs)) { +break ; + } + +symbols__merge_runs (arr, runs[r].start, + runs[r +1 ].start, + runs[r +1 ].start + runs[r +1 ].len, + tmp); + runs[r].len += runs[r +1 ].len; +if (r +2 < nr_runs) + runs[r +1 ] = runs[r +2 ]; + nr_runs--; + } + } + +while (nr_runs >1 ) { +unsigned int r = nr_runs -2 ; + +if (nr_runs >=3 && runs[nr_runs -3 ].len < runs[nr_runs -1 ].len) + r = nr_runs -3 ; +symbols__merge_runs (arr, runs[r].start, + runs[r +1 ].start, + runs[r +1 ].start + runs[r +1 ].len, + tmp); + runs[r].len += runs[r +1 ].len; +if (r +2 < nr_runs) + runs[r +1 ] = runs[r +2 ]; + nr_runs--; + } +} +static int symbols__sort_cmp (const void *vlhs,const void *vrhs) { -const struct symbol *lhs = *((const struct symbol **)vlhs); -const struct symbol *rhs = *((const struct symbol **)vrhs); +const struct symbol *l = *((const struct symbol **)vlhs); +const struct symbol *r = *((const struct symbol **)vrhs); -if (symbol__start (lhs) !=symbol__start (rhs)) -return symbol__start (lhs) <symbol__start (rhs) ?-1 :1 ; -if (symbol__end (lhs) !=symbol__end (rhs)) -return symbol__end (lhs) <symbol__end (rhs) ?-1 :1 ; +if (symbol__start (l) !=symbol__start (r)) +return symbol__start (l) <symbol__start (r) ?-1 :1 ; +if (symbol__end (l) !=symbol__end (r)) +return symbol__end (l) <symbol__end (r) ?-1 :1 ;return 0 ; }static void symbols__sort_locked (struct symbols *symbols) { -if (symbols->sorted) +unsigned int m = symbols->sorted_cnt; +unsigned int k; + +if (symbols->sorted) { + symbols->sorted_cnt = symbols->cnt;return ; + } + +if (m > symbols->cnt) + m =0 ; + k = symbols->cnt - m; -if (symbols->cnt >1 ) { -qsort (symbols->symbols, symbols->cnt,sizeof (struct symbol *), - symbols__sort_cmp); +if (k ==0 ) { + symbols->sorted = true; +return ; } + +/* +* When only a few symbols have been appended since the last sort +* (including single-symbol insertions), insert them directly at +* their binary-searched sorted positions in O(N) without allocating. +*/ +if (k <=8 || symbols->cnt <=64 ) { +for (unsigned int i =max (m,1U ); i < symbols->cnt; i++) { +struct symbol *sym = symbols->symbols[i]; + +if (!symbol__less_or_equal (symbols->symbols[i -1 ], sym)) { +unsigned int pos =symbols__upper_bound (symbols->symbols, i, sym); + +memmove (&symbols->symbols[pos +1 ], &symbols->symbols[pos], + (i - pos) *sizeof (struct symbol *)); + symbols->symbols[pos] = sym; + } + } + symbols->sorted_cnt = symbols->cnt; + symbols->sorted = true; +return ; + } + +if (m <=1 ) { +struct symbol **tmp =malloc (((symbols->cnt +1 ) /2 ) *sizeof (struct symbol *)); + +if (tmp) { +symbols__sort_range_with_tmp (symbols->symbols, symbols->cnt, tmp); +free (tmp); + }else { +qsort (symbols->symbols, symbols->cnt,sizeof (struct symbol *), + symbols__sort_cmp); + } + }else { +struct symbol **tmp =malloc (k *sizeof (struct symbol *)); + +if (tmp) { +symbols__sort_range_with_tmp (&symbols->symbols[m], k, tmp); +symbols__merge_runs (symbols->symbols,0 , m, symbols->cnt, tmp); +free (tmp); + }else { +qsort (symbols->symbols, symbols->cnt,sizeof (struct symbol *), + symbols__sort_cmp); + } + } + + symbols->sorted_cnt = symbols->cnt; symbols->sorted = true; }
@@ -347,6 +634,7 @@ void symbols__fixup_duplicate(struct symbols *symbols) } } symbols->cnt = write_idx +1 ; + symbols->sorted_cnt = symbols->cnt;zfree (&symbols->symbols_by_name); symbols->num_by_name =0 ; symbols->sorted_by_name = false;
@@ -388,40 +676,62 @@ void symbols__fixup_end(struct symbols *symbols, bool is_kallsyms)* like in: * ffffffffc1937000 T hdmi_driver_init [snd_hda_codec_hdmi] */ -if (symbol__end (prev) ==symbol__start (prev)) { +if (symbol__end (prev) ==symbol__start (prev) && +symbol__start (prev) !=symbol__start (curr)) {const char *prev_mod;const char *curr_mod;if (!is_kallsyms) {symbol__set_end (prev,symbol__start (curr)); -continue ; - } + }else { + prev_mod =strchr (symbol__name (prev),'[' ); + curr_mod =strchr (symbol__name (curr),'[' ); + +/* Last kernel/module symbol mapped to end of page */ +if (!prev_mod != !curr_mod) { +symbol__set_end (prev, +roundup (symbol__end (prev) +4096 ,4096 )); +/* Last symbol in the previous module */ + }else if (prev_mod &&strcmp (prev_mod, curr_mod)) { +symbol__set_end (prev, +roundup (symbol__end (prev) +4096 ,4096 )); + }else { +symbol__set_end (prev,symbol__start (curr)); + } - prev_mod =strchr (symbol__name (prev),'[' ); - curr_mod =strchr (symbol__name (curr),'[' ); +/* The next module can start within that page */ +if (symbol__end (prev) >symbol__start (curr)) +symbol__set_end (prev,symbol__start (curr)); -/* Last kernel/module symbol mapped to end of page */ -if (!prev_mod != !curr_mod) -symbol__set_end (prev,roundup (symbol__end (prev) +4096 ,4096 )); -/* Last symbol in the previous module */ -else if (prev_mod &&strcmp (prev_mod, curr_mod)) -symbol__set_end (prev,roundup (symbol__end (prev) +4096 ,4096 )); -else -symbol__set_end (prev,symbol__start (curr)); +pr_debug4 ("%s sym:%s end:%#" PRIx64"\n" , + __func__,symbol__name (prev),symbol__end (prev)); + } -/* The next module can start within that page */ -if (symbol__end (prev) >symbol__start (curr)) -symbol__set_end (prev,symbol__start (curr)); +for (unsigned int k = i -1 ; k >0 ; k--) { +struct symbol *same = symbols->symbols[k -1 ]; -pr_debug4 ("%s sym:%s end:%#" PRIx64"\n" , - __func__,symbol__name (prev),symbol__end (prev)); +if (symbol__start (same) !=symbol__start (prev) || +symbol__end (same) !=symbol__start (same)) +break ; +symbol__set_end (same,symbol__end (prev)); + } } }/* Last entry */ -if (symbol__end (curr) ==symbol__start (curr)) +if (symbol__end (curr) ==symbol__start (curr)) {symbol__set_end (curr,roundup (symbol__start (curr),4096 ) +4096 ); +for (unsigned int k = symbols->cnt -1 ; k >0 ; k--) { +struct symbol *same = symbols->symbols[k -1 ]; + +if (symbol__start (same) !=symbol__start (curr) || +symbol__end (same) !=symbol__start (same)) +break ; +symbol__set_end (same,symbol__end (curr)); + } + } +up_write (&symbols->lock); }
@@ -535,9 +845,13 @@ static int __symbols__insert(struct symbols *symbols, struct symbol *sym) symbols->allocated = to_allocate; } -if (symbols->sorted && symbols->cnt >0 && - !symbol__less_or_equal (symbols->symbols[symbols->cnt -1 ], sym)) - symbols->sorted = false; +if (symbols->sorted) { +if (symbols->cnt ==0 || +symbol__less_or_equal (symbols->symbols[symbols->cnt -1 ], sym)) + symbols->sorted_cnt = symbols->cnt +1 ; +else + symbols->sorted = false; + } symbols->symbols[symbols->cnt++] = sym;zfree (&symbols->symbols_by_name);
@@ -546,6 +860,44 @@ static int __symbols__insert(struct symbols *symbols, struct symbol *sym)return 0 ; } +static int __symbols__insert_sorted (struct symbols *symbols,struct symbol *sym) +{ +unsigned int idx; + +if (!symbols->sorted) +return __symbols__insert (symbols, sym); + +if (symbols->cnt == symbols->allocated) { +unsigned int to_allocate = symbols->allocated ? symbols->allocated *2 :32 ; +struct symbol **temp =realloc (symbols->symbols, +sizeof (struct symbol *) * to_allocate); + +if (!temp) { +symbol__put (sym); +return -ENOMEM ; + } + symbols->symbols = temp; + symbols->allocated = to_allocate; + } + +if (symbols->cnt ==0 || +symbol__less_or_equal (symbols->symbols[symbols->cnt -1 ], sym)) { + idx = symbols->cnt; + }else { + idx =symbols__upper_bound (symbols->symbols, symbols->cnt, sym); +memmove (&symbols->symbols[idx +1 ], &symbols->symbols[idx], + (symbols->cnt - idx) *sizeof (struct symbol *)); + } + + symbols->symbols[idx] = sym; + symbols->cnt++; + symbols->sorted_cnt = symbols->cnt; +zfree (&symbols->symbols_by_name); + symbols->num_by_name =0 ; + symbols->sorted_by_name = false; +return 0 ; +} +int symbols__insert (struct symbols *symbols,struct symbol *sym) {int ret;
@@ -712,7 +1064,13 @@ static struct symbol *symbols__find_by_name(struct symbol *symbols[],int dso__insert_symbol (struct dso *dso,struct symbol *sym) { -return symbols__insert (dso__symbols (dso), sym); +struct symbols *symbols =dso__symbols (dso); +int ret; + +down_write (&symbols->lock); + ret =__symbols__insert_sorted (symbols, sym); +up_write (&symbols->lock); +return ret; }void dso__delete_symbol (struct dso *dso,struct symbol *sym)
@@ -720,13 +1078,16 @@ void dso__delete_symbol(struct dso *dso, struct symbol *sym)struct symbols *symbols =dso__symbols (dso);down_write (&symbols->lock); -for (unsigned int i =0 ; i < symbols->cnt; i++) { +for (unsigned int i = symbols->cnt; i-- >0 ; ) {if (RC_CHK_EQUAL (symbols->symbols[i], sym)) {struct symbol *old = symbols->symbols[i];memmove (&symbols->symbols[i], &symbols->symbols[i +1 ], (symbols->cnt - i -1 ) *sizeof (struct symbol *)); symbols->cnt--; +if (i < symbols->sorted_cnt) + symbols->sorted_cnt--; + symbols->sorted = (symbols->sorted_cnt == symbols->cnt);zfree (&symbols->symbols_by_name); symbols->num_by_name =0 ; symbols->sorted_by_name = false;
@@ -1082,6 +1443,7 @@ static int maps__split_kallsyms_for_kcore(struct maps *kmaps, struct dso *dso) old_cnt = root->cnt; root->symbols =NULL ; root->cnt =0 ; + root->sorted_cnt =0 ; root->allocated =0 ; root->sorted = true;zfree (&root->symbols_by_name);
@@ -1159,6 +1521,7 @@ static int maps__split_kallsyms(struct maps *kmaps, struct dso *dso, u64 delta, old_allocated = root->allocated; root->symbols =NULL ; root->cnt =0 ; + root->sorted_cnt =0 ; root->allocated =0 ; root->sorted = true;zfree (&root->symbols_by_name);
@@ -1305,6 +1668,7 @@ static int maps__split_kallsyms(struct maps *kmaps, struct dso *dso, u64 delta,if (root->cnt ==0 ) { root->symbols = old_symbols; root->cnt = write_idx; + root->sorted_cnt = write_idx; root->allocated = old_allocated; root->sorted = true; old_symbols =NULL ;
@@ -1334,12 +1698,14 @@ static int maps__split_kallsyms(struct maps *kmaps, struct dso *dso, u64 delta,if (root->cnt ==0 ) { root->symbols = old_symbols; root->cnt = write_idx; + root->sorted_cnt = (delta ==0 ) ? write_idx :0 ; root->allocated = old_allocated; root->sorted = (delta ==0 ); old_symbols =NULL ; }else {for (unsigned int j =0 ; j < write_idx; j++)__symbols__insert (root, old_symbols[j]); + root->sorted_cnt =0 ; root->sorted = false; }zfree (&root->symbols_by_name);
tools/perf/util/symbol.h +5 -0
@@ -119,6 +119,11 @@ struct symbols {struct symbol **symbols_by_name;/** @cnt: Number of valid entries in @symbols. */ unsigned int cnt; +/** +* @sorted_cnt: Number of leading entries in @symbols that are sorted +* by address range. Equal to @cnt when @sorted is true. +*/ +unsigned int sorted_cnt;/** @allocated: Capacity (in entries) of the @symbols allocation. */ unsigned int allocated;/** @num_by_name: Number of valid entries in @symbols_by_name. */
Signature · 1 line
2.56.0.313.g2479322e68-goog
Re: [PATCH v2 09/18] perf symbol: Optimize symbol sorting with natural-run merge sort and binary insertion
Quoted 8 lines · Ian Rogers
+ for (unsigned int k = i - 1; k > 0; k--) {
+ struct symbol *same = symbols->symbols[k - 1];+ if (
symbol__start(same)!=symbol__start(prev)||
+symbol__end(same)!=symbol__start(same))
+ break;
+symbol__set_end(same, symbol__end(prev));
+ }
This changes which kallsyms alias survives. Every kallsyms symbol starts with zero size. Before, only the last alias at an address got an end, so symbols__fixup_duplicate() always kept that one. With this loop every alias gets the same end, and choose_best_symbol() falls through to its underscore and name-length rules. On x86_64, perf record -e cpu-clock -g of a getuid/getpid loop, before and after:
0.20% [k] __x64_sys_getuid
0.20% [k] __ia32_sys_getuid
2.87% [k] __memset
2.87% [k] memset
This host has 79 kernel text addresses with more than one name. The losing alias is deleted, so looking up __x64_sys_getuid by name fails too. If the change is intended, it needs its own patch and a test with alias groups. Otherwise, keeping the old "last alias wins" result would avoid it.
Separately, __symbols__insert_sorted() shifts the tail with memmove(), so n out-of-order inserts cost O(n^2). In a benchmark of this code, 300k random inserts took 3.4 s, where the rb-tree was O(log n) per insert. db-export with many unique unresolved addresses would hit that.
Thanks,
Alireza
Ian Rogers
reviewerPATCH v2 10/18 perf maps: Optimize tail insertion and small-append sorting
In reply to Ian RogersUser-space map insertions (from /proc/pid/maps synthesis, runtime PERF_RECORD_MMAP2, and jitdump) and BPF ksymbol registrations all flow through maps__fixup_overlap_and_insert() -> first_ending_after(). Currently first_ending_after() has a fast-path check for the first map (maps_by_address[0]), but no fast-path check for the last map (maps_by_address[high]). Because /proc/pid/maps entries are synthesized in ascending address order, every synthesized mapping starts at or after maps_by_address[high]'s end, yet currently executes the full O(log N) binary search loop. Add a tail check in first_ending_after() so non-overlapping ascending insertions complete in O(1).
Additionally, track the sorted prefix lengths (nr_maps_by_address_sorted and nr_maps_by_name_sorted) in struct maps:
- In __maps__sort_by_address() and maps__sort_by_name(), when the unsorted tail is small (k <= 8) or the total map count is small (n <= 64), binary-insert the newly appended maps into the sorted prefix instead of re-running qsort() over the entire array. This avoids O(N^2 log N) re-sorting when maps__find_by_name() is called repeatedly between single-map maps__insert() calls in dso__process_kernel_symbol() and dso__split_kallsyms().
- In maps__by_address_index() and maps__by_name_index(), binary-search the known-sorted prefix before scanning only the unsorted tail.
- Add a unit test in tools/perf/tests/maps.c covering the tail fast-path and incremental binary-insertion sort.
Using 'perf bench internals maps':
Before:
Average ascending insert + lookup: 63.370 usec (+- 0.476 usec)
Average append + lookup: 7477.450 usec (+- 20.829 usec)
After (-6.9% ascending, -98.7% append):
Average ascending insert + lookup: 58.970 usec (+- 0.609 usec)
Average append + lookup: 96.670 usec (+- 1.353 usec)--- tools/perf/tests/maps.c | 58 +++++++++ tools/perf/util/maps.c | 281 +++++++++++++++++++++++++++++++++------- 2 files changed, 289 insertions(+), 50 deletions(-)
tools/perf/tests/maps.c +58 -0
@@ -1,5 +1,6 @@// SPDX-License-Identifier: GPL-2.0 #include <inttypes.h> +#include <internal/rc_check.h> #include <linux/compiler.h> #include <linux/kernel.h> #include "tests.h"
@@ -233,9 +234,66 @@ static int test__maps__fixup_overlap_and_insert(struct test_suite *t __maybe_unureturn TEST_OK ; } +static int test__maps__incremental_sort (struct test_suite *t__maybe_unused , +int subtest__maybe_unused ) +{ +struct map_def ascending_maps[] = { + {"map_c" ,1000 ,2000 }, + {"map_a" ,2000 ,3000 }, + {"map_d" ,4000 ,5000 }, + }; +struct map_def out_of_order = {"map_b" ,3000 ,4000 }; +struct map *map, *found; +struct maps *maps =maps__new (NULL ); + +TEST_ASSERT_VAL ("failed to create maps" , maps); + +/* Exercise tail fast-path in maps__fixup_overlap_and_insert. */ +for (unsigned int i =0 ; i <ARRAY_SIZE (ascending_maps); i++) { + map =dso__new_map (ascending_maps[i].name); +TEST_ASSERT_VAL ("failed to create map" , map); +map__set_start (map, ascending_maps[i].start); +map__set_end (map, ascending_maps[i].end); +TEST_ASSERT_VAL ("failed to insert sorted map" , +maps__fixup_overlap_and_insert (maps, map) ==0 ); +map__put (map); + } + +/* Populate maps_by_name and verify lookup. */ + found =maps__find_by_name (maps,"map_a" ); +TEST_ASSERT_VAL ("failed to find map_a by name" , found &&map__start (found) ==2000 ); +map__zput (found); + +/* Append out-of-order map via maps__insert and verify binary-insertion sort. */ + map =dso__new_map (out_of_order.name); +TEST_ASSERT_VAL ("failed to create map_b" , map); +map__set_start (map, out_of_order.start); +map__set_end (map, out_of_order.end); +TEST_ASSERT_VAL ("failed to insert map_b" ,maps__insert (maps, map) ==0 ); + + found =maps__find (maps,3500 ); +TEST_ASSERT_VAL ("failed to find map_b by address" ,RC_CHK_EQUAL (found, map)); +map__zput (found); + + found =maps__find_by_name (maps,"map_b" ); +TEST_ASSERT_VAL ("failed to find map_b by name" ,RC_CHK_EQUAL (found, map)); +map__zput (found); + +maps__remove (maps, map); +map__zput (map); + + found =maps__find_by_name (maps,"map_d" ); +TEST_ASSERT_VAL ("failed to find map_d after removal" , found &&map__start (found) ==4000 ); +map__zput (found); + +maps__zput (maps); +return TEST_OK ; +} +static struct test_case tests__maps[] = {TEST_CASE ("Test merge_in interface" , maps__merge_in),TEST_CASE ("Test fix up overlap interface" , maps__fixup_overlap_and_insert), +TEST_CASE ("Test incremental sort and tail insert" , maps__incremental_sort), { .name =NULL , } };
tools/perf/util/maps.c +231 -50
475 lines not on this page · raw message
Signature · 1 line
2.56.0.313.g2479322e68-goog