From 0890e97c20333c439a9bda6a94dd16ed5f508429 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 1 Sep 2016 18:53:58 -0300 Subject: perf machine: Remove machine->symbol_filter and friends Including machines__set_symbol_filter(), not used anymore. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-7o1qgmrpvzuis4a9f0t8mnri@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index 551ff6f640be..d594052c8184 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -477,7 +477,7 @@ static int intel_pt_walk_next_insn(struct intel_pt_insn *intel_pt_insn, start_ip = *ip; /* Load maps to ensure dso->is_64_bit has been updated */ - map__load(al.map, machine->symbol_filter); + map__load(al.map, NULL); x86_64 = al.map->dso->is_64_bit; @@ -1294,7 +1294,7 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip) if (!map) return 0; - if (map__load(map, machine->symbol_filter)) + if (map__load(map, NULL)) return 0; start = dso__first_symbol(map->dso, MAP__FUNCTION); -- cgit From be39db9f2932f0ce4e116c71ba5ae2b542e536a7 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 1 Sep 2016 19:25:52 -0300 Subject: perf symbols: Remove symbol_filter_t machinery We're not using it anymore, few users were, but we really could do without it, simplify lots of functions by removing it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-1zng8wdznn00iiz08bb7q3vn@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index d594052c8184..b9cc353cace2 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -477,7 +477,7 @@ static int intel_pt_walk_next_insn(struct intel_pt_insn *intel_pt_insn, start_ip = *ip; /* Load maps to ensure dso->is_64_bit has been updated */ - map__load(al.map, NULL); + map__load(al.map); x86_64 = al.map->dso->is_64_bit; @@ -1294,7 +1294,7 @@ static u64 intel_pt_switch_ip(struct intel_pt *pt, u64 *ptss_ip) if (!map) return 0; - if (map__load(map, NULL)) + if (map__load(map)) return 0; start = dso__first_symbol(map->dso, MAP__FUNCTION); -- cgit From 810c398bc09b2f2dfde52a7d2483a710612c5fb8 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:41 +0300 Subject: perf intel-pt: Fix snapshot overlap detection decoder errors Fix occasional decoder errors decoding trace data collected in snapshot mode. Snapshot mode can take successive snapshots of trace which might overlap. The decoder checks whether there is an overlap but only looks at the current and previous buffer. However buffers that do not contain synchronization (i.e. PSB) packets cannot be decoded or used for overlap checking. That means the decoder actually needs to check overlaps between the current buffer and the previous buffer that contained usable data. Make that change. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Cc: stable@vger.kernel.org # v4.3+ Link: http://lkml.kernel.org/r/1474641528-18776-10-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index b9cc353cace2..b744ea812a2e 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -241,7 +241,7 @@ static int intel_pt_get_trace(struct intel_pt_buffer *b, void *data) } queue = &ptq->pt->queues.queue_array[ptq->queue_nr]; - +next: buffer = auxtrace_buffer__next(queue, buffer); if (!buffer) { if (old_buffer) @@ -264,9 +264,6 @@ static int intel_pt_get_trace(struct intel_pt_buffer *b, void *data) intel_pt_do_fix_overlap(ptq->pt, old_buffer, buffer)) return -ENOMEM; - if (old_buffer) - auxtrace_buffer__drop_data(old_buffer); - if (buffer->use_data) { b->len = buffer->use_size; b->buf = buffer->use_data; @@ -276,6 +273,16 @@ static int intel_pt_get_trace(struct intel_pt_buffer *b, void *data) } b->ref_timestamp = buffer->reference; + /* + * If in snapshot mode and the buffer has no usable data, get next + * buffer and again check overlap against old_buffer. + */ + if (ptq->pt->snapshot_mode && !b->len) + goto next; + + if (old_buffer) + auxtrace_buffer__drop_data(old_buffer); + if (!old_buffer || ptq->pt->sampling_mode || (ptq->pt->snapshot_mode && !buffer->consecutive)) { b->consecutive = false; -- cgit From fa8025c37454501a2df4a90ae84ff01f4aff8ba8 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:42 +0300 Subject: perf intel-pt: Add support for recording the max non-turbo ratio Previously the maximum non-turbo ratio was calculated from TSC assuming a 100 MHz multiplier which is correct for current hardware supporting Intel PT. However more recent kernels also now export the value, so use that in preference to the calculated value. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Link: http://lkml.kernel.org/r/1474641528-18776-11-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index b744ea812a2e..77fbf02c8e41 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -2023,6 +2023,7 @@ static const char * const intel_pt_info_fmts[] = { [INTEL_PT_TSC_CTC_N] = " TSC:CTC numerator %"PRIu64"\n", [INTEL_PT_TSC_CTC_D] = " TSC:CTC denominator %"PRIu64"\n", [INTEL_PT_CYC_BIT] = " CYC bit %#"PRIx64"\n", + [INTEL_PT_MAX_NONTURBO_RATIO] = " Max non-turbo ratio %"PRIu64"\n", }; static void intel_pt_print_info(u64 *arr, int start, int finish) @@ -2087,6 +2088,15 @@ int intel_pt_process_auxtrace_info(union perf_event *event, INTEL_PT_CYC_BIT); } + if (auxtrace_info->header.size >= sizeof(struct auxtrace_info_event) + + (sizeof(u64) * INTEL_PT_MAX_NONTURBO_RATIO)) { + pt->max_non_turbo_ratio = + auxtrace_info->priv[INTEL_PT_MAX_NONTURBO_RATIO]; + intel_pt_print_info(&auxtrace_info->priv[0], + INTEL_PT_MAX_NONTURBO_RATIO, + INTEL_PT_MAX_NONTURBO_RATIO); + } + pt->timeless_decoding = intel_pt_timeless_decoding(pt); pt->have_tsc = intel_pt_have_tsc(pt); pt->sampling_mode = false; @@ -2156,7 +2166,9 @@ int intel_pt_process_auxtrace_info(union perf_event *event, if (pt->tc.time_mult) { u64 tsc_freq = intel_pt_ns_to_ticks(pt, 1000000000); - pt->max_non_turbo_ratio = (tsc_freq + 50000000) / 100000000; + if (!pt->max_non_turbo_ratio) + pt->max_non_turbo_ratio = + (tsc_freq + 50000000) / 100000000; intel_pt_log("TSC frequency %"PRIu64"\n", tsc_freq); intel_pt_log("Maximum non-turbo ratio %u\n", pt->max_non_turbo_ratio); -- cgit From 4d34e10a9f9a38c611cac0deda8f91b064282747 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:43 +0300 Subject: perf intel-pt: Fix missing error codes processing auxtrace_info Fix 2 places where the err variable was not being set. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Link: http://lkml.kernel.org/r/1474641528-18776-12-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index 77fbf02c8e41..96519e801e53 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -2138,11 +2138,13 @@ int intel_pt_process_auxtrace_info(union perf_event *event, pt->switch_evsel = intel_pt_find_sched_switch(session->evlist); if (!pt->switch_evsel) { pr_err("%s: missing sched_switch event\n", __func__); + err = -EINVAL; goto err_delete_thread; } } else if (pt->have_sched_switch == 2 && !intel_pt_find_switch(session->evlist)) { pr_err("%s: missing context_switch attribute flag\n", __func__); + err = -EINVAL; goto err_delete_thread; } -- cgit From 40b746a06332799786ba557fe84184428bef62fb Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:44 +0300 Subject: perf intel-pt: Add a helper function for processing AUXTRACE_INFO Add a helper function 'intel_pt_has()' to make it easier to determine which members the AUXTRACE_INFO event contains. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Link: http://lkml.kernel.org/r/1474641528-18776-13-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index 96519e801e53..f16b00f55a19 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -2037,6 +2037,12 @@ static void intel_pt_print_info(u64 *arr, int start, int finish) fprintf(stdout, intel_pt_info_fmts[i], arr[i]); } +static bool intel_pt_has(struct auxtrace_info_event *auxtrace_info, int pos) +{ + return auxtrace_info->header.size >= + sizeof(struct auxtrace_info_event) + (sizeof(u64) * (pos + 1)); +} + int intel_pt_process_auxtrace_info(union perf_event *event, struct perf_session *session) { @@ -2077,8 +2083,7 @@ int intel_pt_process_auxtrace_info(union perf_event *event, intel_pt_print_info(&auxtrace_info->priv[0], INTEL_PT_PMU_TYPE, INTEL_PT_PER_CPU_MMAPS); - if (auxtrace_info->header.size >= sizeof(struct auxtrace_info_event) + - (sizeof(u64) * INTEL_PT_CYC_BIT)) { + if (intel_pt_has(auxtrace_info, INTEL_PT_CYC_BIT)) { pt->mtc_bit = auxtrace_info->priv[INTEL_PT_MTC_BIT]; pt->mtc_freq_bits = auxtrace_info->priv[INTEL_PT_MTC_FREQ_BITS]; pt->tsc_ctc_ratio_n = auxtrace_info->priv[INTEL_PT_TSC_CTC_N]; @@ -2088,8 +2093,7 @@ int intel_pt_process_auxtrace_info(union perf_event *event, INTEL_PT_CYC_BIT); } - if (auxtrace_info->header.size >= sizeof(struct auxtrace_info_event) + - (sizeof(u64) * INTEL_PT_MAX_NONTURBO_RATIO)) { + if (intel_pt_has(auxtrace_info, INTEL_PT_MAX_NONTURBO_RATIO)) { pt->max_non_turbo_ratio = auxtrace_info->priv[INTEL_PT_MAX_NONTURBO_RATIO]; intel_pt_print_info(&auxtrace_info->priv[0], -- cgit From 2b9e32c47fd3edb0373067de7a151775b0e005c2 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:46 +0300 Subject: perf intel-pt: Read address filter from AUXTRACE_INFO event Read the address filter from the AUXTRACE_INFO event in preparation for using it to assist in decoding. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Link: http://lkml.kernel.org/r/1474641528-18776-15-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 51 insertions(+) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index f16b00f55a19..c9fec19a7914 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -103,6 +103,8 @@ struct intel_pt { unsigned max_non_turbo_ratio; unsigned long num_events; + + char *filter; }; enum switch_state { @@ -1774,6 +1776,7 @@ static void intel_pt_free(struct perf_session *session) intel_pt_free_events(session); session->auxtrace = NULL; thread__put(pt->unknown_thread); + zfree(&pt->filter); free(pt); } @@ -2024,6 +2027,7 @@ static const char * const intel_pt_info_fmts[] = { [INTEL_PT_TSC_CTC_D] = " TSC:CTC denominator %"PRIu64"\n", [INTEL_PT_CYC_BIT] = " CYC bit %#"PRIx64"\n", [INTEL_PT_MAX_NONTURBO_RATIO] = " Max non-turbo ratio %"PRIu64"\n", + [INTEL_PT_FILTER_STR_LEN] = " Filter string len. %"PRIu64"\n", }; static void intel_pt_print_info(u64 *arr, int start, int finish) @@ -2037,6 +2041,14 @@ static void intel_pt_print_info(u64 *arr, int start, int finish) fprintf(stdout, intel_pt_info_fmts[i], arr[i]); } +static void intel_pt_print_info_str(const char *name, const char *str) +{ + if (!dump_trace) + return; + + fprintf(stdout, " %-20s%s\n", name, str ? str : ""); +} + static bool intel_pt_has(struct auxtrace_info_event *auxtrace_info, int pos) { return auxtrace_info->header.size >= @@ -2049,6 +2061,8 @@ int intel_pt_process_auxtrace_info(union perf_event *event, struct auxtrace_info_event *auxtrace_info = &event->auxtrace_info; size_t min_sz = sizeof(u64) * INTEL_PT_PER_CPU_MMAPS; struct intel_pt *pt; + void *info_end; + u64 *info; int err; if (auxtrace_info->header.size < sizeof(struct auxtrace_info_event) + @@ -2101,6 +2115,42 @@ int intel_pt_process_auxtrace_info(union perf_event *event, INTEL_PT_MAX_NONTURBO_RATIO); } + info = &auxtrace_info->priv[INTEL_PT_FILTER_STR_LEN] + 1; + info_end = (void *)info + auxtrace_info->header.size; + + if (intel_pt_has(auxtrace_info, INTEL_PT_FILTER_STR_LEN)) { + size_t len; + + len = auxtrace_info->priv[INTEL_PT_FILTER_STR_LEN]; + intel_pt_print_info(&auxtrace_info->priv[0], + INTEL_PT_FILTER_STR_LEN, + INTEL_PT_FILTER_STR_LEN); + if (len) { + const char *filter = (const char *)info; + + len = roundup(len + 1, 8); + info += len >> 3; + if ((void *)info > info_end) { + pr_err("%s: bad filter string length\n", __func__); + err = -EINVAL; + goto err_free_queues; + } + pt->filter = memdup(filter, len); + if (!pt->filter) { + err = -ENOMEM; + goto err_free_queues; + } + if (session->header.needs_swap) + mem_bswap_64(pt->filter, len); + if (pt->filter[len - 1]) { + pr_err("%s: filter string not null terminated\n", __func__); + err = -EINVAL; + goto err_free_queues; + } + } + intel_pt_print_info_str("Filter string", pt->filter); + } + pt->timeless_decoding = intel_pt_timeless_decoding(pt); pt->have_tsc = intel_pt_have_tsc(pt); pt->sampling_mode = false; @@ -2218,6 +2268,7 @@ err_free_queues: auxtrace_queues__free(&pt->queues); session->auxtrace = NULL; err_free: + zfree(&pt->filter); free(pt); return err; } -- cgit From 2acee108f58045d07475516852d4282ab73904dc Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 23 Sep 2016 17:38:48 +0300 Subject: perf intel-pt: Fix decoding when there are address filters Due to errata SKL014 "Intel PT TIP.PGD May Not Have Target IP Payload", the Intel PT decoder needs to match address filters against TIP.PGD packets. Parse the address filters and implement the decoder's 'pgd_ip()' callback to match the IP against the filter regions. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Mathieu Poirier Link: http://lkml.kernel.org/r/1474641528-18776-17-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/intel-pt.c | 82 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 82 insertions(+) (limited to 'tools/perf/util/intel-pt.c') diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index c9fec19a7914..dc041d4368c8 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -105,6 +105,7 @@ struct intel_pt { unsigned long num_events; char *filter; + struct addr_filters filts; }; enum switch_state { @@ -550,6 +551,76 @@ out_no_cache: return 0; } +static bool intel_pt_match_pgd_ip(struct intel_pt *pt, uint64_t ip, + uint64_t offset, const char *filename) +{ + struct addr_filter *filt; + bool have_filter = false; + bool hit_tracestop = false; + bool hit_filter = false; + + list_for_each_entry(filt, &pt->filts.head, list) { + if (filt->start) + have_filter = true; + + if ((filename && !filt->filename) || + (!filename && filt->filename) || + (filename && strcmp(filename, filt->filename))) + continue; + + if (!(offset >= filt->addr && offset < filt->addr + filt->size)) + continue; + + intel_pt_log("TIP.PGD ip %#"PRIx64" offset %#"PRIx64" in %s hit filter: %s offset %#"PRIx64" size %#"PRIx64"\n", + ip, offset, filename ? filename : "[kernel]", + filt->start ? "filter" : "stop", + filt->addr, filt->size); + + if (filt->start) + hit_filter = true; + else + hit_tracestop = true; + } + + if (!hit_tracestop && !hit_filter) + intel_pt_log("TIP.PGD ip %#"PRIx64" offset %#"PRIx64" in %s is not in a filter region\n", + ip, offset, filename ? filename : "[kernel]"); + + return hit_tracestop || (have_filter && !hit_filter); +} + +static int __intel_pt_pgd_ip(uint64_t ip, void *data) +{ + struct intel_pt_queue *ptq = data; + struct thread *thread; + struct addr_location al; + u8 cpumode; + u64 offset; + + if (ip >= ptq->pt->kernel_start) + return intel_pt_match_pgd_ip(ptq->pt, ip, ip, NULL); + + cpumode = PERF_RECORD_MISC_USER; + + thread = ptq->thread; + if (!thread) + return -EINVAL; + + thread__find_addr_map(thread, cpumode, MAP__FUNCTION, ip, &al); + if (!al.map || !al.map->dso) + return -EINVAL; + + offset = al.map->map_ip(al.map, ip); + + return intel_pt_match_pgd_ip(ptq->pt, ip, offset, + al.map->dso->long_name); +} + +static bool intel_pt_pgd_ip(uint64_t ip, void *data) +{ + return __intel_pt_pgd_ip(ip, data) > 0; +} + static bool intel_pt_get_config(struct intel_pt *pt, struct perf_event_attr *attr, u64 *config) { @@ -726,6 +797,9 @@ static struct intel_pt_queue *intel_pt_alloc_queue(struct intel_pt *pt, params.tsc_ctc_ratio_n = pt->tsc_ctc_ratio_n; params.tsc_ctc_ratio_d = pt->tsc_ctc_ratio_d; + if (pt->filts.cnt > 0) + params.pgd_ip = intel_pt_pgd_ip; + if (pt->synth_opts.instructions) { if (pt->synth_opts.period) { switch (pt->synth_opts.period_type) { @@ -1776,6 +1850,7 @@ static void intel_pt_free(struct perf_session *session) intel_pt_free_events(session); session->auxtrace = NULL; thread__put(pt->unknown_thread); + addr_filters__exit(&pt->filts); zfree(&pt->filter); free(pt); } @@ -2073,6 +2148,8 @@ int intel_pt_process_auxtrace_info(union perf_event *event, if (!pt) return -ENOMEM; + addr_filters__init(&pt->filts); + perf_config(intel_pt_perf_config, pt); err = auxtrace_queues__init(&pt->queues); @@ -2147,6 +2224,10 @@ int intel_pt_process_auxtrace_info(union perf_event *event, err = -EINVAL; goto err_free_queues; } + err = addr_filters__parse_bare_filter(&pt->filts, + filter); + if (err) + goto err_free_queues; } intel_pt_print_info_str("Filter string", pt->filter); } @@ -2268,6 +2349,7 @@ err_free_queues: auxtrace_queues__free(&pt->queues); session->auxtrace = NULL; err_free: + addr_filters__exit(&pt->filts); zfree(&pt->filter); free(pt); return err; -- cgit