From nobody Tue Jul 28 19:29:23 2026 X-Original-To: dev-commits-src-all@mlmmj.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mlmmj.nyi.freebsd.org (Postfix) with ESMTP id 4h8lrz5Ftdz6n9h2 for ; Tue, 28 Jul 2026 19:29:23 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from mxrelay.nyi.freebsd.org (mxrelay.nyi.freebsd.org [IPv6:2610:1c1:1:606c::19:3]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256 client-signature RSA-PSS (4096 bits) client-digest SHA256) (Client CN "mxrelay.nyi.freebsd.org", Issuer "YR1" (not verified)) by mx1.freebsd.org (Postfix) with ESMTPS id 4h8lrz3N8hz43xT for ; Tue, 28 Jul 2026 19:29:23 +0000 (UTC) (envelope-from git@FreeBSD.org) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=freebsd.org; s=dkim; t=1785266963; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=h7UY2xkhkQ8pHWyE2L6IgiT3UbUl5j+E6ZSDzcD26wU=; b=d172zPCvHViWOY+2DsdbcrqjE1URkm/yCcefOf/lnyYd8t89kkGauYYyOObv4EMqBuCE4A RG2E6dDJtYYxyL6GtfSmCnMqRieIsOYVkfEvwC5FCuee47ZSRZRHAp1yHWwEDYKOxbLgam JLnEpNHksDxC3LGaRSNQlLOYMchcqJbfJOxhWSf6RkLUjABDwKYg6lFke9LjcesDmiLaVx j6RH3yKzhb4wVKsbW5Qv/AMCFVFuMAW3DbkhZVWEfuDlArlufvh1B7IYMqpatswR9OSpXV pC+9HrtmuR6bObgNjrI0svyOXeASZ2PaaK4XxSwGSwvz0SHN6UC24quSOzusCA== ARC-Seal: i=1; s=dkim; d=freebsd.org; t=1785266963; a=rsa-sha256; cv=none; b=yZP9fZrnefOBKpX/yYyRuskHqttzCnYgRfgaoifk46s4aspU4/RMj9zmMIyoIY+vXdrsyY q4Ia+5DZxdvG5ZGzGxtKMkYO2N7n5hCUcV6zYBXVYOn5Q1bIOyr+3FS1N2DHbCDenwP7B/ htFTk1dLTcPmC7Bp3FX2Tqu0Ua/0KVCnWnsveOSnfcwiEjb3f0gPIpEqKrEuT5QIx/D6DE GItN5PykM20UGmJnnOYunIHV/wzATnI/9F+RqPJkI/dNRnJeUKylGfhVLlPtIyRlNqpIjj 8z6ew4mjeHYP+nYbcEkmWY+tgdIpgLy98Vhn+NYmRv1o7jY6EWTcVGxA4s9lBw== ARC-Authentication-Results: i=1; mx1.freebsd.org; none ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=freebsd.org; s=dkim; t=1785266963; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=h7UY2xkhkQ8pHWyE2L6IgiT3UbUl5j+E6ZSDzcD26wU=; b=tsCvBJOlfKD7442/dywbpsJNiO6VwcUY0BPfssmlPb93Tfn/tFMniqC8IiLny1CVQQZRV+ Aj2grR5KUtVKlo5/siA+XJ3I0uc/BSp5kNbprh43ENvLJYd0TmObbK5SGXF6FBftJRxS7y HGpZhUeXuR9G0ZIESHUtWYRPDkq+Pa0rV7LR8vSVHJyAFJr7O+sfYc6AcK5ueavYiLbsbx 4o6BDPdf+VFEK1L7CGKL0ZaCIZDy/ywjS0DpXlHV9meyYxt0HLoGV+AjmbZPJTut+1ELRO 5ids5/C0cSQPabJ6/jtOU8BNGqfYdb1P6lXkAXT/A4Fl/4b/AO6vzjuPrGIIig== Received: from gitrepo.freebsd.org (gitrepo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:5]) by mxrelay.nyi.freebsd.org (Postfix) with ESMTP id 4h8lrz1t3vztMx for ; Tue, 28 Jul 2026 19:29:23 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from git (uid 1279) (envelope-from git@FreeBSD.org) id 43239 by gitrepo.freebsd.org (DragonFly Mail Agent v0.13+ on gitrepo.freebsd.org); Tue, 28 Jul 2026 19:29:23 +0000 To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-main@FreeBSD.org Cc: Ali Mashtizadeh From: Warner Losh Subject: git: 93da997ef759 - main - pmc: new pmc log processing framework List-Id: Commit messages for all branches of the src repository List-Archive: https://lists.freebsd.org/archives/dev-commits-src-all List-Help: List-Post: List-Subscribe: List-Unsubscribe: X-BeenThere: dev-commits-src-all@freebsd.org Sender: owner-dev-commits-src-all@FreeBSD.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit X-Git-Committer: imp X-Git-Repository: src X-Git-Refname: refs/heads/main X-Git-Reftype: branch X-Git-Commit: 93da997ef759061670c1575eec92f922a645a5fd Auto-Submitted: auto-generated Date: Tue, 28 Jul 2026 19:29:23 +0000 Message-Id: <6a690313.43239.6961872a@gitrepo.freebsd.org> The branch main has been updated by imp: URL: https://cgit.FreeBSD.org/src/commit/?id=93da997ef759061670c1575eec92f922a645a5fd commit 93da997ef759061670c1575eec92f922a645a5fd Author: Ali Mashtizadeh AuthorDate: 2026-07-28 17:46:22 +0000 Commit: Warner Losh CommitDate: 2026-07-28 19:27:44 +0000 pmc: new pmc log processing framework View is a class for building PMC log processing tools it is designed to work with the new PMC record command that adds a header with additional CPU information. The new framework processes PMC logs about 2.5 times faster and in about half the code as libpmcstat. Sponsored by: Netflix Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D57776 --- usr.sbin/pmc/headers.hh | 82 +++++ usr.sbin/pmc/util.cc | 82 +++++ usr.sbin/pmc/util.hh | 39 ++ usr.sbin/pmc/view.cc | 960 ++++++++++++++++++++++++++++++++++++++++++++++++ usr.sbin/pmc/view.hh | 411 +++++++++++++++++++++ 5 files changed, 1574 insertions(+) diff --git a/usr.sbin/pmc/headers.hh b/usr.sbin/pmc/headers.hh new file mode 100644 index 000000000000..29f76cd7cc2c --- /dev/null +++ b/usr.sbin/pmc/headers.hh @@ -0,0 +1,82 @@ +/*- + * SPDX-License-Identifier: BSD-2-Clause + * + * Copyright (c) 2026, Netflix, Inc. + * + * This software was developed by Ali Mashtizadeh under the sponsorship from + * Netflix, Inc. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + */ + +#ifndef __HEADERS_HH__ +#define __HEADERS_HH__ + +#define PMC_HEADER_MAGIC 0x504D434C +#define PMC_HEADER_VERSION 0x00 + +#define PMC_ARCH_AMD64 0x01 +#define PMC_ARCH_ARM64 0x02 +#define PMC_ARCH_PPC64 0x03 +#define PMC_ARCH_RISCV64 0x04 + +struct pmchdr_header +{ + uint32_t magic; + uint16_t version; + uint16_t arch; +}; + +#define INFOHDR_TYPE_DONE 0x00 +#define INFOHDR_TYPE_SYSINFO 0x01 +#define INFOHDR_TYPE_PMCINFO 0x02 +#define INFOHDR_TYPE_CPUID 0x03 + +struct pmchdr_infohdr +{ + uint8_t type; + uint8_t reserved0; + uint16_t length; + uint32_t _reserved1; +}; + +struct pmchdr_sysinfo +{ + char cpumodel[64]; + char osrelease[64]; + char buildid[64]; +}; + +struct pmchdr_pmcinfo +{ + uint64_t rate; + char pmc[]; +}; + +struct pmchdr_cpuidinfo +{ + uint32_t cpuid[]; +}; + +#endif + diff --git a/usr.sbin/pmc/util.cc b/usr.sbin/pmc/util.cc new file mode 100644 index 000000000000..b6e43b9dd130 --- /dev/null +++ b/usr.sbin/pmc/util.cc @@ -0,0 +1,82 @@ +/*- + * SPDX-License-Identifier: BSD-2-Clause + * + * Copyright (c) 2026, Netflix, Inc. + * + * This software was developed by Ali Mashtizadeh under the sponsorship from + * Netflix, Inc. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + */ + +#include +#include + +#include "util.hh" + +std::string +basename(const std::string &path) +{ + size_t s; + + s = path.rfind("/"); + if (s == std::string::npos) + return (path); + else + return (path.substr(s + 1)); +} + +void +split_and_insert(std::unordered_set *set, const std::string &str) +{ + size_t pos = 0; + + while (pos < str.length()) { + size_t end = str.find(",", pos); + if (end == str.npos) { + set->insert(std::stoi(str.substr(pos))); + break; + } + + set->insert(std::stoi(str.substr(pos, end - pos))); + pos = end + 1; + } +} + +void +split_and_insert(std::unordered_set *set, const std::string &str) +{ + size_t pos = 0; + + while (pos < str.length()) { + size_t end = str.find(",", pos); + if (end == str.npos) { + set->insert(str.substr(pos)); + break; + } + + set->insert(str.substr(pos, end - pos)); + pos = end + 1; + } +} + diff --git a/usr.sbin/pmc/util.hh b/usr.sbin/pmc/util.hh new file mode 100644 index 000000000000..03c350721b2b --- /dev/null +++ b/usr.sbin/pmc/util.hh @@ -0,0 +1,39 @@ +/*- + * SPDX-License-Identifier: BSD-2-Clause + * + * Copyright (c) 2026, Netflix, Inc. + * + * This software was developed by Ali Mashtizadeh under the sponsorship from + * Netflix, Inc. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + */ +#ifndef __UTIL_H__ +#define __UTIL_H__ + +std::string basename(const std::string &path); +void split_and_insert(std::unordered_set *set, const std::string &input); +void split_and_insert(std::unordered_set *set, const std::string &input); + +#endif /* __UTIL_HH__ */ + diff --git a/usr.sbin/pmc/view.cc b/usr.sbin/pmc/view.cc new file mode 100644 index 000000000000..352856939a66 --- /dev/null +++ b/usr.sbin/pmc/view.cc @@ -0,0 +1,960 @@ +/*- + * SPDX-License-Identifier: BSD-2-Clause + * + * Copyright (c) 2026, Netflix, Inc. + * + * This software was developed by Ali Mashtizadeh under the sponsorship from + * Netflix, Inc. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + */ + +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include "util.hh" +#include "view.hh" + +std::string +syminfo::to_string(bool show_line) +{ + int status; + char *demangled; + std::stringstream ss; + + if (name == "") { + ss << "0x" << std::hex << offset; + } else if (name.length() >= 2 && name[0] == '_' && name[1] == 'Z') { + demangled = abi::__cxa_demangle(name.c_str(), nullptr, nullptr, &status); + if (status == 0) { + ss << demangled; + free(demangled); + } else { + ss << name; + } + } else { + ss << name; + } + + if (!show_line) + return ss.str(); + + if (line) { + ss << ":" << std::dec << line; + } else if (funcoff) { + ss << "+0x" << std::hex << funcoff; + } + + return ss.str(); +} + +pmcview::pmcview() : tscfreq(0), pmcid(), pmcinfo(), procs(), tidtopid(), + images(), sysroot(""), filter() +{ + char *root; + + root = getenv("SYSROOT"); + if (root) { + sysroot = root; + } +} + +pmcview::~pmcview() +{ +} + +void +pmcview::setfilter(const pmcfilter &pmcfilter) +{ + filter = pmcfilter; +} + +static int +readlog(int logfd, void *buf, size_t len) +{ + int status; + char *cur = (char *)buf; + size_t left = len; + + while (left != 0) { + status = read(logfd, cur, left); + if (status < 0) { + if (errno == EINTR || errno == EAGAIN) + continue; + else + return status; + } + if (status == 0) + return status; + + cur += status; + left -= status; + } + + return len; +} + +int +pmcview::process_sysinfo(int logfd, const pmchdr_infohdr &infohdr) +{ + int status; + pmchdr_sysinfo sysinfo; + + if (infohdr.length != sizeof(sysinfo)) + errx(EX_IOERR, "PMC log headers have an unexpected size"); + + status = readlog(logfd, &sysinfo, sizeof(sysinfo)); + if (status < 0 || status != infohdr.length) + errx(EX_IOERR, "readlog"); + + cpumodel = sysinfo.cpumodel; + osrelease = sysinfo.osrelease; + buildid = sysinfo.buildid; + + return 0; +} + +int +pmcview::process_pmcinfo(int logfd, const pmchdr_infohdr &infohdr) +{ + int status; + union { + pmchdr_pmcinfo pmcinfo; + __unused char reserve_max_size[256]; + }; + + status = readlog(logfd, &pmcinfo, infohdr.length); + if (status < 0 || status != infohdr.length) + errx(EX_IOERR, "readlog"); + + extpmcinfo.emplace_back(pmcinfo.rate, std::string(pmcinfo.pmc)); + + return 0; +} + +int +pmcview::process_cpuidinfo(int logfd, const pmchdr_infohdr &infohdr) +{ + int status; + pmchdr_cpuidinfo *cpuidinfo; + int offset, len; + uint32_t root, maxleaf, count; + + len = infohdr.length / 4; + cpuidinfo = (pmchdr_cpuidinfo *)new uint32_t[len]; + + status = readlog(logfd, cpuidinfo, infohdr.length); + if (status < 0 || status != infohdr.length) + errx(EX_IOERR, "readlog"); + + offset = 0; + + while (offset < len) { + maxleaf = cpuidinfo->cpuid[offset]; + + /* + * In x86 the roots contains the maximum leaf number present. + */ + root = maxleaf & 0xFFFF0000; + count = maxleaf & 0x0000FFFF; + + for (uint32_t i = 0; i <= count; i++) { + cpuid[root + i] = { cpuidinfo->cpuid[4 * i + offset], + cpuidinfo->cpuid[4 * i + 1 + offset], + cpuidinfo->cpuid[4 * i + 2 + offset], + cpuidinfo->cpuid[4 * i + 3 + offset] }; + } + + offset += 4 * (count + 1); + } + + delete[] cpuidinfo; + + return 0; +} + +void +pmcview::process(int logfd) +{ + int status; + pmchdr_header hdr; + struct pmclog_parse_state *ps; + + // Read header + status = readlog(logfd, &hdr, sizeof(hdr)); + if (status < 0 || status != sizeof(hdr)) + err(EX_IOERR, "readlog"); + if (hdr.magic != PMC_HEADER_MAGIC) + errx(EX_DATAERR, "PMC log magic mismatch!"); + if (hdr.version > PMC_HEADER_VERSION) + errx(EX_DATAERR, "PMC log version newer than supported!"); + + // Save architecture + arch = hdr.arch; + + while (1) { + pmchdr_infohdr infohdr; + + status = readlog(logfd, &infohdr, sizeof(infohdr)); + if (status < 0 || status != sizeof(infohdr)) + errx(EX_IOERR, "readlog"); + + if (infohdr.type == INFOHDR_TYPE_DONE) { + break; + } else if (infohdr.type == INFOHDR_TYPE_SYSINFO) { + process_sysinfo(logfd, infohdr); + } else if (infohdr.type == INFOHDR_TYPE_PMCINFO) { + process_pmcinfo(logfd, infohdr); + } else if (infohdr.type == INFOHDR_TYPE_CPUID) { + process_cpuidinfo(logfd, infohdr); + } + } + + ps = static_cast(pmclog_open(logfd)); + if (ps == NULL) { + errx(EX_OSERR, "ERROR: Cannot allocate pmclog parse state: %s\n", + strerror(errno)); + } + + process(ps); + + pmclog_close(ps); +} + +void +pmcview::process(struct pmclog_ev &p) +{ + switch (p.pl_type) { + case PMCLOG_TYPE_INITIALIZE: + process(p.pl_u.pl_i); + return; + case PMCLOG_TYPE_CLOSELOG: + process(p.pl_u.pl_cl); + return; + case PMCLOG_TYPE_PMCALLOCATE: + process(p.pl_u.pl_a); + return; + case PMCLOG_TYPE_PMCALLOCATEDYN: + process(p.pl_u.pl_ad); + return; + case PMCLOG_TYPE_PROC_CREATE: + process(p.pl_u.pl_pc); + return; + case PMCLOG_TYPE_PROCEXIT: + process(p.pl_u.pl_e); + return; + case PMCLOG_TYPE_PROCEXEC: + process(p.pl_u.pl_x); + return; + case PMCLOG_TYPE_PROCFORK: + process(p.pl_u.pl_f); + return; + case PMCLOG_TYPE_SYSEXIT: + process(p.pl_u.pl_se); + return; + case PMCLOG_TYPE_MAP_IN: + process(p.pl_u.pl_mi); + return; + case PMCLOG_TYPE_MAP_OUT: + process(p.pl_u.pl_mo); + return; + case PMCLOG_TYPE_THR_CREATE: + process(p.pl_u.pl_tc); + return; + case PMCLOG_TYPE_THR_EXIT: + process(p.pl_u.pl_te); + return; + case PMCLOG_TYPE_CALLCHAIN: + process(p.pl_u.pl_cc); + return; + case PMCLOG_TYPE_PMCATTACH: [[fallthrough]]; + case PMCLOG_TYPE_PMCDETACH: + case PMCLOG_TYPE_USERDATA: + case PMCLOG_TYPE_PROCCSW: + case PMCLOG_TYPE_DROPNOTIFY: + return; + }; +} + +void +pmcview::process(struct pmclog_parse_state *p) +{ + struct pmclog_ev ev; + + while (pmclog_read(p, &ev) == 0) { + process(ev); + } +} + +void +pmcview::process(struct pmclog_ev_initialize &p) +{ + tscfreq = p.pl_tsc_freq; +} + +void +pmcview::process(__unused struct pmclog_ev_closelog &p) +{ +} + +void +pmcview::process(struct pmclog_ev_pmcallocate &p) +{ + pmcid[p.pl_pmcid] = p.pl_event; + if (pmcinfo.find(p.pl_event) == pmcinfo.end()) { + pmcinfo.emplace(p.pl_event, p); + } +} + +void +pmcview::process(struct pmclog_ev_pmcallocatedyn &p) +{ + pmcid[p.pl_pmcid] = p.pl_event; + if (pmcinfo.find(p.pl_event) == pmcinfo.end()) { + pmcinfo.emplace(p.pl_event, p); + } +} + +void +pmcview::process(struct pmclog_ev_proccreate &p) +{ + procs[p.pl_pid] = procinfo(p); + + proccreate(p.pl_pid); +} + +void +pmcview::process(struct pmclog_ev_procexit &p) +{ + /* + * XXX: Seems we can recieve samples after the process exits we need + * some way to delay cleanup, then after the delayed cleanup discard + * delayed events. + */ + procs.erase(p.pl_pid); + + procexit(p.pl_pid); +} + +void +pmcview::process(struct pmclog_ev_procexec &p) +{ + procs[p.pl_pid].map.clear(); + procs[p.pl_pid].name = basename(p.pl_pathname); + procs[p.pl_pid].fullpath = p.pl_pathname; + procs[p.pl_pid].baseaddr = p.pl_baseaddr; + procs[p.pl_pid].dynaddr = p.pl_dynaddr; + + image im = loadimage(p.pl_pathname); + + mapimage(p.pl_pid, im, im.vaddr + p.pl_dynaddr); + + /* + * Map the dynamic runtime loader + */ + if (im.isdynamic) { + image rtldim = loadimage(im.loader); + + mapimage(p.pl_pid, rtldim, p.pl_baseaddr); + } + + procexec(p.pl_pid); +} + +void +pmcview::process(struct pmclog_ev_procfork &p) +{ + procs[p.pl_newpid] = procs[p.pl_oldpid]; + + proccreate(p.pl_newpid); +} + +void +pmcview::process(struct pmclog_ev_sysexit &p) +{ + procs.erase(p.pl_pid); +} + +void +pmcview::process(struct pmclog_ev_threadcreate &p) +{ + procs[p.pl_pid].threads[p.pl_tid] = threadinfo(p.pl_tdname); + tidtopid[p.pl_tid] = p.pl_pid; +} + +void +pmcview::process(struct pmclog_ev_threadexit &p) +{ + pid_t pid = tidtopid[p.pl_tid]; + procs[pid].threads.erase(p.pl_tid); + tidtopid.erase(p.pl_tid); +} + +/* + * Load the ELF file to compute the values needed for the memory map and check + * if the dwarf symbols are available. + */ +image +pmcview::loadimage(const std::string &path) +{ + std::string fullpath; + image im; + GElf_Ehdr eh; + Elf *e; + const char *elf; + uint64_t start; + uint64_t end; + size_t shstrndx; + int fd, i; + bool foundexec; + + if (images.find(path) != images.end()) + return images[path]; + + im = image(); + + if (path == "unknown") { + return (im); + } + + fullpath = sysroot + path; + + if (access(fullpath.c_str(), R_OK) != 0) + return (im); + + fd = open(fullpath.c_str(), O_RDONLY, 0); + if (fd < 0) { + warnx("WARNING: Cannot open \"%s\".", + fullpath.c_str()); + return (im); + } + + e = elf_begin(fd, ELF_C_READ, NULL); + if (e == NULL) { + warnx("WARNING: Cannot read \"%s\".", + fullpath.c_str()); + close(fd); + return (im); + } + + if (gelf_getehdr(e, &eh) != &eh) { + warnx("WARNING: Cannot read the ELF header for \"%s\": %s.", + fullpath.c_str(), elf_errmsg(-1)); + goto done; + } + + if (eh.e_type != ET_EXEC && eh.e_type != ET_DYN && eh.e_type != ET_REL) { + warnx("WARNING: ELF file type is unsupported for \"%s\".", + fullpath.c_str());; + goto done; + } + + elf = elf_rawfile(e, NULL); + if (elf == NULL) { + warnx("WARNING: Cannot read the ELF file for \"%s\": %s.", + fullpath.c_str(), elf_errmsg(-1)); + goto done; + } + + if (eh.e_type != ET_REL) { + foundexec = false; + for (i = 0; i < eh.e_phnum; i++) { + GElf_Phdr ph; + + if (gelf_getphdr(e, i, &ph) != &ph) { + warnx("WARNING: Cannot read program header for \"%s\": %s.", + fullpath.c_str(), elf_errmsg(-1)); + goto done; + } + + if (ph.p_type == PT_DYNAMIC) { + im.isdynamic = 1; + continue; + } + + if (ph.p_type == PT_INTERP) { + im.loader = elf + ph.p_offset; + } + + if (ph.p_type == PT_LOAD) { + if ((ph.p_flags & PF_X) != 0 && !foundexec) { + im.vaddr = ph.p_vaddr & ~(ph.p_align - 1); + foundexec = true; + } + } + } + } + + elf_getshdrstrndx(e, &shstrndx); + + start = ~(0x0ULL); + end = 0; + for (i = 0; i < eh.e_shnum; i++) { + GElf_Shdr sh; + Elf_Scn *scn; + char *scname; + + scn = elf_getscn(e, i); + if (scn == NULL) { + warnx("WARNING: Could not retrieve section descriptor for \"%s\".", + fullpath.c_str()); + goto done; + } + + if (gelf_getshdr(scn, &sh) != &sh) { + warnx("WARNING: Could not retrieve section header for \"%s\".", + fullpath.c_str()); + goto done; + } + + if (sh.sh_flags & SHF_EXECINSTR) { + start = std::min(start, sh.sh_addr); + end = std::max(end, sh.sh_addr + sh.sh_size); + } + + // Check if dwarf is embedded + scname = elf_strptr(e, shstrndx, sh.sh_name); + if (scname != NULL && strcmp(scname, ".debug_info") == 0) + im.dwarf = fullpath; + } + + im.start = start; + im.end = end; + im.binary = path; + im.path = fullpath; + im.name = basename(fullpath); + + /* + * If the dwarf symbols aren't embedded check: + * 1. SYSROOT + /path + .debug + * 2. SYSROOT + /usr/lib/debug + /path + .debug + */ + if (im.dwarf == "") { + std::string sympath = fullpath + ".debug"; + if (access(sympath.c_str(), R_OK) == 0) { + im.dwarf = sympath; + } + } + if (im.dwarf == "") { + std::string sympath = sysroot + "/usr/lib/debug" + path + ".debug"; + if (access(sympath.c_str(), R_OK) == 0) { + im.dwarf = sympath; + } + } + + images[path] = im; + +done: + elf_end(e); + close(fd); + + return (im); +} + +void +pmcview::loadsymboltable(image *im, Elf *e, Elf_Scn *scn, GElf_Shdr *sh) +{ + size_t n, nsyms; + char *fname; + GElf_Sym sym; + Elf_Data *data; + syminfo si; + + si = syminfo(); + + if ((data = elf_getdata(scn, nullptr)) == nullptr) + return; + + nsyms = sh->sh_size / sh->sh_entsize; + + for (n = 0; n < nsyms; n++) { + if (gelf_getsym(data, (int) n, &sym) != &sym) + return; + + // XXX: We should load globals as well + if (GELF_ST_TYPE(sym.st_info) != STT_FUNC) + continue; + + if (sym.st_shndx == STN_UNDEF) + continue; + + if ((fname = elf_strptr(e, sh->sh_link, sym.st_name)) == NULL) + continue; + + // XXX: Extra checks to make sure we don't get corrupted + for (int i = 0; fname[i] != 0 && i < 32; i++) { + if (fname[i] < 0) { + printf("EEEK SYMBOL\n"); + printf("%s\n", fname); + assert(false); + } + } + + si.offset = sym.st_value; + si.length = sym.st_size; + si.binary = im->name; + si.name = fname; + + im->symbols[si.offset] = si; + } +} + +void +pmcview::loadsymbols(image *im) +{ + int i; + int fd; + Elf *e; + Elf_Scn *scn; + GElf_Ehdr eh; + GElf_Shdr sh; + + if (access(im->path.c_str(), R_OK) != 0) + return; + + fd = open(im->path.c_str(), O_RDONLY, 0); + if (fd < 0) { + warnx("WARNING: Cannot open \"%s\".", + im->path.c_str()); + return; + } + + e = elf_begin(fd, ELF_C_READ, NULL); + if (e == NULL) { + warnx("WARNING: Cannot read \"%s\".", + im->path.c_str()); + close(fd); + return; + } + + if (gelf_getehdr(e, &eh) != &eh) { + warnx("WARNING: Cannot read the ELF header for \"%s\": %s.", + im->path.c_str(), elf_errmsg(-1)); + goto done; + } + + for (i = 0; i < eh.e_shnum; i++) { + scn = elf_getscn(e, i); + if (scn == NULL) { + warnx("WARNING: Could not retrieve section descriptor for \"%s\".", + im->path.c_str()); + goto done; + } + + if (gelf_getshdr(scn, &sh) != &sh) { + warnx("WARNING: Could not retrieve section header for \"%s\".", + im->path.c_str()); + goto done; + } + + if (sh.sh_type == SHT_SYMTAB || sh.sh_type == SHT_DYNSYM) { + loadsymboltable(im, e, scn, &sh); + } + } + +done: + elf_end(e); + close(fd); +} + +void +pmcview::mapimage(int pid, const image &im, uint64_t start) +{ + uint64_t offset; + vmmap map; + + // Ignore empty images + if (im.start == 0 && im.end == 0) + return; + + /* + * XXX: Need to adjust the address for the PowerPC kernel that is + * dynamic. Wonder if we can fix this elsewhere, because we should need + * a way to deal with this for KASLR?. + */ + + offset = start - im.vaddr; + map.lowpc = im.start + offset; + map.highpc = im.end + offset; + map.image = im.binary; + + procs[pid].map[start] = map; +} + +void +pmcview::process(struct pmclog_ev_map_in &p) +{ + // Kernel map-in events should be mapped to pid 0 + pid_t pid = (p.pl_pid == -1) ? 0 : p.pl_pid; + + image im = loadimage(p.pl_pathname); + *** 646 LINES SKIPPED ***