Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754769AbZJ0NE7 (ORCPT ); Tue, 27 Oct 2009 09:04:59 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754530AbZJ0NE6 (ORCPT ); Tue, 27 Oct 2009 09:04:58 -0400 Received: from hera.kernel.org ([140.211.167.34]:49172 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754495AbZJ0NE5 (ORCPT ); Tue, 27 Oct 2009 09:04:57 -0400 Date: Tue, 27 Oct 2009 13:04:06 GMT From: tip-bot for Arnaldo Carvalho de Melo Cc: linux-kernel@vger.kernel.org, paulus@samba.org, acme@redhat.com, hpa@zytor.com, mingo@redhat.com, efault@gmx.de, peterz@infradead.org, fweisbec@gmail.com, tglx@linutronix.de, mingo@elte.hu Reply-To: mingo@redhat.com, hpa@zytor.com, acme@redhat.com, paulus@samba.org, linux-kernel@vger.kernel.org, fweisbec@gmail.com, peterz@infradead.org, efault@gmx.de, tglx@linutronix.de, mingo@elte.hu In-Reply-To: <1256592199-9608-3-git-send-email-acme@redhat.com> References: <1256592199-9608-3-git-send-email-acme@redhat.com> To: linux-tip-commits@vger.kernel.org Subject: [tip:perf/core] perf top: Support userspace symbols too Message-ID: Git-Commit-ID: 5b2bb75a0d4b08cd16bc35ecd674f957fc3b0eb7 X-Mailer: tip-git-log-daemon MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 10000 Lines: 296 Commit-ID: 5b2bb75a0d4b08cd16bc35ecd674f957fc3b0eb7 Gitweb: http://git.kernel.org/tip/5b2bb75a0d4b08cd16bc35ecd674f957fc3b0eb7 Author: Arnaldo Carvalho de Melo AuthorDate: Mon, 26 Oct 2009 19:23:19 -0200 Committer: Ingo Molnar CommitDate: Tue, 27 Oct 2009 13:51:54 +0100 perf top: Support userspace symbols too Example: Compiling the kernel with 'make -k 22 allyesconfig' [root@emilia linux-2.6-tip]# perf top -r 90 ------------------------------------------------------------------------------ PerfTop: 3669 irqs/sec kernel:59.9% [1000Hz cycles], (all, 8 CPUs) ------------------------------------------------------------------------------ samples pcnt function DSO _______ _____ ________________________________ ________________ 3062.00 6.5% clear_page_c [kernel] 2233.00 4.8% _int_malloc /lib64/libc-2.5.so 2100.00 4.5% yylex /home/acme/git/build/allyesconfig/scripts/genksyms/genksyms 2029.00 4.3% memset /lib64/libc-2.5.so 1224.00 2.6% page_fault [kernel] 1075.00 2.3% __GI_strlen /lib64/libc-2.5.so 863.00 1.8% sub_preempt_count [kernel] 822.00 1.8% __GI_memcpy /lib64/libc-2.5.so 810.00 1.7% __GI_vfprintf /lib64/libc-2.5.so 786.00 1.7% _int_free /lib64/libc-2.5.so 775.00 1.7% __GI_strcmp /lib64/libc-2.5.so 748.00 1.6% _spin_lock [kernel] 699.00 1.5% main /home/acme/git/build/allyesconfig/scripts/basic/fixdep 659.00 1.4% add_preempt_count [kernel] 649.00 1.4% yyparse /home/acme/git/build/allyesconfig/scripts/genksyms/genksyms 645.00 1.4% preempt_trace [kernel] 635.00 1.4% __GI___libc_free /lib64/libc-2.5.so 597.00 1.3% trace_preempt_on [kernel] 551.00 1.2% __GI___libc_malloc /lib64/libc-2.5.so 516.00 1.1% _spin_lock_irqsave [kernel] 481.00 1.0% copy_user_generic_string [kernel] 479.00 1.0% unmap_vmas [kernel] 429.00 0.9% _IO_file_xsputn_internal /lib64/libc-2.5.so 425.00 0.9% __GI_strncpy /lib64/libc-2.5.so 416.00 0.9% get_page_from_freelist [kernel] 414.00 0.9% malloc_consolidate /lib64/libc-2.5.so 406.00 0.9% get_parent_ip [kernel] 362.00 0.8% __rmqueue [kernel] 347.00 0.7% in_lock_functions [kernel] 316.00 0.7% __d_lookup [kernel] [root@emilia linux-2.6-tip]# More polishing is needed to print just DSO basename when not --verbose, etc. Supporting a 'comm' column requires some more reworking of 'perf top' internals as we will need to use something like the hist entries 'perf report' uses and will be done in another patch. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Mike Galbraith LKML-Reference: <1256592199-9608-3-git-send-email-acme@redhat.com> Signed-off-by: Ingo Molnar --- tools/perf/builtin-top.c | 143 ++++++++++++++++++++++++++++++++------------- 1 files changed, 101 insertions(+), 42 deletions(-) diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 4a9fe22..a02fc41 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -318,7 +318,7 @@ static void show_details(struct sym_entry *syme) } /* - * Symbols will be added here in record_ip and will get out + * Symbols will be added here in event__process_sample and will get out * after decayed. */ static LIST_HEAD(active_symbols); @@ -459,18 +459,18 @@ static void print_sym_table(void) } if (nr_counters == 1) - printf(" samples pcnt"); + printf(" samples pcnt"); else - printf(" weight samples pcnt"); + printf(" weight samples pcnt"); if (verbose) printf(" RIP "); - printf(" kernel function\n"); - printf(" %s _______ _____", + printf(" function DSO\n"); + printf(" %s _______ _____", nr_counters == 1 ? " " : "______"); if (verbose) - printf(" ________________"); - printf(" _______________\n\n"); + printf(" ________________"); + printf(" ________________________________ ________________\n\n"); for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) { struct symbol *sym; @@ -486,16 +486,15 @@ static void print_sym_table(void) sum_ksamples)); if (nr_counters == 1 || !display_weighted) - printf("%20.2f - ", syme->weight); + printf("%20.2f ", syme->weight); else - printf("%9.1f %10ld - ", syme->weight, syme->snap_count); + printf("%9.1f %10ld ", syme->weight, syme->snap_count); percent_color_fprintf(stdout, "%4.1f%%", pcnt); if (verbose) - printf(" - %016llx", sym->start); - printf(" : %s", sym->name); - if (syme->map->dso->name[0] == '[') - printf(" \t%s", syme->map->dso->name); + printf(" %016llx", sym->start); + printf(" %-32s", sym->name); + printf(" %s", syme->map->dso->short_name); printf("\n"); } } @@ -818,41 +817,97 @@ static int parse_symbols(void) return 0; } -/* - * Binary search in the histogram table and record the hit: - */ -static void record_ip(u64 ip, int counter) +static void event__process_sample(const event_t *self, int counter) { + u64 ip = self->ip.ip; struct map *map; - struct symbol *sym = kernel_maps__find_symbol(ip, &map); - - if (sym != NULL) { - struct sym_entry *syme = dso__sym_priv(map->dso, sym); - - if (!syme->skip) { - syme->count[counter]++; - record_precise_ip(syme, counter, ip); - pthread_mutex_lock(&active_symbols_lock); - if (list_empty(&syme->node) || !syme->node.next) - __list_insert_active_sym(syme); - pthread_mutex_unlock(&active_symbols_lock); + struct sym_entry *syme; + struct symbol *sym; + + switch (self->header.misc & PERF_RECORD_MISC_CPUMODE_MASK) { + case PERF_RECORD_MISC_USER: { + struct thread *thread = threads__findnew(self->ip.pid); + + if (thread == NULL) return; + + map = thread__find_map(thread, ip); + if (map != NULL) { + ip = map->map_ip(map, ip); + sym = map->dso->find_symbol(map->dso, ip); + if (sym == NULL) + return; + userspace_samples++; + break; } } + /* + * If this is outside of all known maps, + * and is a negative address, try to look it + * up in the kernel dso, as it might be a + * vsyscall or vdso (which executes in user-mode). + */ + if ((long long)ip >= 0) + return; + /* Fall thru */ + case PERF_RECORD_MISC_KERNEL: + sym = kernel_maps__find_symbol(ip, &map); + if (sym == NULL) + return; + break; + default: + return; + } + + syme = dso__sym_priv(map->dso, sym); - samples--; + if (!syme->skip) { + syme->count[counter]++; + record_precise_ip(syme, counter, ip); + pthread_mutex_lock(&active_symbols_lock); + if (list_empty(&syme->node) || !syme->node.next) + __list_insert_active_sym(syme); + pthread_mutex_unlock(&active_symbols_lock); + ++samples; + return; + } } -static void process_event(u64 ip, int counter, int user) +static void event__process_mmap(event_t *self) { - samples++; + struct thread *thread = threads__findnew(self->mmap.pid); + + if (thread != NULL) { + struct map *map = map__new(&self->mmap, NULL, 0, + sizeof(struct sym_entry), + symbol_filter); + if (map != NULL) + thread__insert_map(thread, map); + } +} - if (user) { - userspace_samples++; - return; +static void event__process_comm(event_t *self) +{ + struct thread *thread = threads__findnew(self->comm.pid); + + if (thread != NULL) + thread__set_comm(thread, self->comm.comm); +} + +static int event__process(event_t *event) +{ + switch (event->header.type) { + case PERF_RECORD_COMM: + event__process_comm(event); + break; + case PERF_RECORD_MMAP: + event__process_mmap(event); + break; + default: + break; } - record_ip(ip, counter); + return 0; } struct mmap_data { @@ -925,13 +980,11 @@ static void mmap_read_counter(struct mmap_data *md) event = &event_copy; } + if (event->header.type == PERF_RECORD_SAMPLE) + event__process_sample(event, md->counter); + else + event__process(event); old += size; - - if (event->header.type == PERF_RECORD_SAMPLE) { - int user = - (event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK) == PERF_RECORD_MISC_USER; - process_event(event->ip.ip, md->counter, user); - } } md->prev = old; @@ -973,6 +1026,7 @@ static void start_counter(int i, int counter) } attr->inherit = (cpu < 0) && inherit; + attr->mmap = 1; try_again: fd[i][counter] = sys_perf_event_open(attr, target_pid, cpu, group_fd, 0); @@ -1031,6 +1085,11 @@ static int __cmd_top(void) int i, counter; int ret; + if (target_pid != -1) + event__synthesize_thread(target_pid, event__process); + else + event__synthesize_threads(event__process); + for (i = 0; i < nr_cpus; i++) { group_fd = -1; for (counter = 0; counter < nr_counters; counter++) -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/