From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-9.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 07DC5C43441 for ; Tue, 20 Nov 2018 05:06:27 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id C216F206BB for ; Tue, 20 Nov 2018 05:06:26 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org C216F206BB Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=firstfloor.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1731786AbeKTPdg (ORCPT ); Tue, 20 Nov 2018 10:33:36 -0500 Received: from mga12.intel.com ([192.55.52.136]:52748 "EHLO mga12.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1730962AbeKTPdg (ORCPT ); Tue, 20 Nov 2018 10:33:36 -0500 X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by fmsmga106.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 19 Nov 2018 21:06:24 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.56,255,1539673200"; d="scan'208";a="87760804" Received: from tassilo.jf.intel.com (HELO tassilo.localdomain) ([10.7.201.126]) by fmsmga008.fm.intel.com with ESMTP; 19 Nov 2018 21:06:24 -0800 Received: by tassilo.localdomain (Postfix, from userid 1000) id 5833C3018C4; Mon, 19 Nov 2018 21:06:24 -0800 (PST) From: Andi Kleen To: acme@kernel.org Cc: linux-kernel@vger.kernel.org, jolsa@kernel.org, Andi Kleen , milian.wolff@kdab.com, adrian.hunter@intel.com Subject: [PATCH] perf script: Fix LBR skid dump problems in brstackinsn Date: Mon, 19 Nov 2018 21:06:17 -0800 Message-Id: <20181120050617.4119-1-andi@firstfloor.org> X-Mailer: git-send-email 2.17.2 Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Andi Kleen This is a fix for another instance of the skid problem Milian recently found [1] The LBRs don't freeze at the exact same time as the PMI is triggered. The perf script brstackinsn code that dumps LBR assembler assumes that the last branch in the LBR leads to the sample point. But with skid it's possible that the CPU executes one or more branches before the sample, but which do not appear in the LBR. What happens then is either that the sample point is before the last LBR branch. In this case the dumper sees a negative length and ignores it. Or it the sample point is long after the last branch. Then the dumper sees a very long block and dumps it upto its block limit (16k bytes), which is noise in the output. On typical sample session this can happen regularly. This patch tries to detect and handle the situation. On the last block that is dumped by the LBR dumper we always stop on the first branch. If the block length is negative just scan forward to the first branch. Otherwise scan until a branch is found. The PT decoder already has a function that uses the instruction decoder to detect branches, so we can just reuse it here. Then when a terminating branch is found print an indication and stop dumping. This might miss a few instructions, but at least shows no runaway blocks. Cc: milian.wolff@kdab.com Cc: adrian.hunter@intel.com Signed-off-by: Andi Kleen --- tools/perf/builtin-script.c | 18 +++++++++++++++++- tools/perf/util/dump-insn.c | 8 ++++++++ tools/perf/util/dump-insn.h | 2 ++ .../intel-pt-decoder/intel-pt-insn-decoder.c | 8 ++++++++ 4 files changed, 35 insertions(+), 1 deletion(-) diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 4da5e32b9e03..11868bf39e66 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -1049,9 +1049,18 @@ static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample, /* * Print final block upto sample + * + * Due to pipeline delays the LBRs might be missing a branch + * or two, which can result in very large or negative blocks + * between final branch and sample. When this happens just + * continue walking after the last TO until we hit a branch. */ start = br->entries[0].to; end = sample->ip; + if (end < start) { + /* Missing jump. Scan 128 bytes for the next branch */ + end = start + 128; + } len = grab_bb(buffer, start, end, machine, thread, &x.is64bit, &x.cpumode, true); printed += ip__fprintf_sym(start, thread, x.cpumode, x.cpu, &lastsym, attr, fp); if (len <= 0) { @@ -1060,7 +1069,6 @@ static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample, machine, thread, &x.is64bit, &x.cpumode, false); if (len <= 0) goto out; - printed += fprintf(fp, "\t%016" PRIx64 "\t%s\n", sample->ip, dump_insn(&x, sample->ip, buffer, len, NULL)); goto out; @@ -1070,6 +1078,14 @@ static int perf_sample__fprintf_brstackinsn(struct perf_sample *sample, dump_insn(&x, start + off, buffer + off, len - off, &ilen)); if (ilen == 0) break; + if (arch_is_branch(buffer + off, len - off, x.is64bit) && + start + off != sample->ip) { + /* + * Hit a missing branch. Just stop. + */ + printed += fprintf(fp, "\t... not reaching sample ...\n"); + break; + } } out: return printed; diff --git a/tools/perf/util/dump-insn.c b/tools/perf/util/dump-insn.c index 10988d3de7ce..2bd8585db93c 100644 --- a/tools/perf/util/dump-insn.c +++ b/tools/perf/util/dump-insn.c @@ -13,3 +13,11 @@ const char *dump_insn(struct perf_insn *x __maybe_unused, *lenp = 0; return "?"; } + +__weak +int arch_is_branch(const unsigned char *buf __maybe_unused, + size_t len __maybe_unused, + int x86_64 __maybe_unused) +{ + return 0; +} diff --git a/tools/perf/util/dump-insn.h b/tools/perf/util/dump-insn.h index 0e06280a8860..650125061530 100644 --- a/tools/perf/util/dump-insn.h +++ b/tools/perf/util/dump-insn.h @@ -20,4 +20,6 @@ struct perf_insn { const char *dump_insn(struct perf_insn *x, u64 ip, u8 *inbuf, int inlen, int *lenp); +int arch_is_branch(const unsigned char *buf, size_t len, int x86_64); + #endif diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c index 54818828023b..1c0e289f01e6 100644 --- a/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c +++ b/tools/perf/util/intel-pt-decoder/intel-pt-insn-decoder.c @@ -180,6 +180,14 @@ int intel_pt_get_insn(const unsigned char *buf, size_t len, int x86_64, return 0; } +int arch_is_branch(const unsigned char *buf, size_t len, int x86_64) +{ + struct intel_pt_insn in; + if (intel_pt_get_insn(buf, len, x86_64, &in) < 0) + return -1; + return in.branch != INTEL_PT_BR_NO_BRANCH; +} + const char *dump_insn(struct perf_insn *x, uint64_t ip __maybe_unused, u8 *inbuf, int inlen, int *lenp) { -- 2.17.2