From: Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> To: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org>, Alexei Starovoitov <ast@kernel.org>, Ingo Molnar <mingo@redhat.com>, James Morris <jmorris@namei.org>, Namhyung Kim <namhyung@kernel.org>, Serge Hallyn <serge@hallyn.com>, Jiri Olsa <jolsa@redhat.com>, Song Liu <songliubraving@fb.com>, Andi Kleen <ak@linux.intel.com>, Stephane Eranian <eranian@google.com>, Igor Lubashev <ilubashe@akamai.com>, Thomas Gleixner <tglx@linutronix.de>, linux-kernel <linux-kernel@vger.kernel.org>, "linux-security-module@vger.kernel.org" <linux-security-module@vger.kernel.org>, "selinux@vger.kernel.org" <selinux@vger.kernel.org>, "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>, "linux-doc@vger.kernel.org" <linux-doc@vger.kernel.org>, linux-man@vger.kernel.org Subject: Re: [PATCH v8 00/12] Introduce CAP_PERFMON to secure system performance monitoring and observability Date: Tue, 7 Apr 2020 11:30:14 -0300 [thread overview] Message-ID: <20200407143014.GD11186@kernel.org> (raw) In-Reply-To: <f96f8f8a-e65c-3f36-dc85-fc3f5191e8c5@linux.intel.com> Em Thu, Apr 02, 2020 at 11:42:05AM +0300, Alexey Budankov escreveu: > This patch set introduces CAP_PERFMON capability designed to secure > system performance monitoring and observability operations so that > CAP_PERFMON would assist CAP_SYS_ADMIN capability in its governing role > for performance monitoring and observability subsystems of the kernel. So, what am I doing wrong? [perf@five ~]$ type perf perf is hashed (/home/perf/bin/perf) [perf@five ~]$ [perf@five ~]$ ls -lahF /home/perf/bin/perf -rwxr-x---. 1 root perf_users 24M Apr 7 10:34 /home/perf/bin/perf* [perf@five ~]$ [perf@five ~]$ getcap /home/perf/bin/perf [perf@five ~]$ perf top --stdio Error: You may not have permission to collect system-wide stats. Consider tweaking /proc/sys/kernel/perf_event_paranoid, which controls use of the performance events system by unprivileged users (without CAP_PERFMON or CAP_SYS_ADMIN). The current value is 2: -1: Allow use of (almost) all events by all users Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK >= 0: Disallow ftrace function tracepoint by users without CAP_PERFMON or CAP_SYS_ADMIN Disallow raw tracepoint access by users without CAP_SYS_PERFMON or CAP_SYS_ADMIN >= 1: Disallow CPU event access by users without CAP_PERFMON or CAP_SYS_ADMIN >= 2: Disallow kernel profiling by users without CAP_PERFMON or CAP_SYS_ADMIN To make this setting permanent, edit /etc/sysctl.conf too, e.g.: kernel.perf_event_paranoid = -1 [perf@five ~]$ Ok, the message says I need to have CAP_PERFMON, lets do it, using an unpatched libcap that doesn't know about it but we can use 38, CAP_PERFMON value instead, and I tested this with a patched libcap as well, same results: As root: [root@five bin]# setcap "38,cap_sys_ptrace,cap_syslog=ep" perf [root@five bin]# Back to the 'perf' user in the 'perf_users' group, ok, so now 'perf record -a' works for system wide sampling of cycles:u, i.e. only userspace samples, but 'perf top' is failing: [perf@five ~]$ type perf perf is hashed (/home/perf/bin/perf) [perf@five ~]$ getcap /home/perf/bin/perf /home/perf/bin/perf = cap_sys_ptrace,cap_syslog,38+ep [perf@five ~]$ groups perf perf_users [perf@five ~]$ id uid=1002(perf) gid=1002(perf) groups=1002(perf),1003(perf_users) context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 [perf@five ~]$ perf top --stdio Error: Failed to mmap with 1 (Operation not permitted) [perf@five ~]$ perf record -a ^C[ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.177 MB perf.data (1552 samples) ] [perf@five ~]$ perf evlist cycles:u [perf@five ~]$ - Arnaldo
WARNING: multiple messages have this Message-ID (diff)
From: Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> To: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: linux-man@vger.kernel.org, Song Liu <songliubraving@fb.com>, Andi Kleen <ak@linux.intel.com>, "linux-doc@vger.kernel.org" <linux-doc@vger.kernel.org>, Peter Zijlstra <peterz@infradead.org>, linux-kernel <linux-kernel@vger.kernel.org>, "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>, Igor Lubashev <ilubashe@akamai.com>, Alexei Starovoitov <ast@kernel.org>, Stephane Eranian <eranian@google.com>, James Morris <jmorris@namei.org>, "selinux@vger.kernel.org" <selinux@vger.kernel.org>, "linux-security-module@vger.kernel.org" <linux-security-module@vger.kernel.org>, Ingo Molnar <mingo@redhat.com>, Namhyung Kim <namhyung@kernel.org>, Thomas Gleixner <tglx@linutronix.de>, Jiri Olsa <jolsa@redhat.com>, Serge Hallyn <serge@hallyn.com> Subject: Re: [Intel-gfx] [PATCH v8 00/12] Introduce CAP_PERFMON to secure system performance monitoring and observability Date: Tue, 7 Apr 2020 11:30:14 -0300 [thread overview] Message-ID: <20200407143014.GD11186@kernel.org> (raw) In-Reply-To: <f96f8f8a-e65c-3f36-dc85-fc3f5191e8c5@linux.intel.com> Em Thu, Apr 02, 2020 at 11:42:05AM +0300, Alexey Budankov escreveu: > This patch set introduces CAP_PERFMON capability designed to secure > system performance monitoring and observability operations so that > CAP_PERFMON would assist CAP_SYS_ADMIN capability in its governing role > for performance monitoring and observability subsystems of the kernel. So, what am I doing wrong? [perf@five ~]$ type perf perf is hashed (/home/perf/bin/perf) [perf@five ~]$ [perf@five ~]$ ls -lahF /home/perf/bin/perf -rwxr-x---. 1 root perf_users 24M Apr 7 10:34 /home/perf/bin/perf* [perf@five ~]$ [perf@five ~]$ getcap /home/perf/bin/perf [perf@five ~]$ perf top --stdio Error: You may not have permission to collect system-wide stats. Consider tweaking /proc/sys/kernel/perf_event_paranoid, which controls use of the performance events system by unprivileged users (without CAP_PERFMON or CAP_SYS_ADMIN). The current value is 2: -1: Allow use of (almost) all events by all users Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK >= 0: Disallow ftrace function tracepoint by users without CAP_PERFMON or CAP_SYS_ADMIN Disallow raw tracepoint access by users without CAP_SYS_PERFMON or CAP_SYS_ADMIN >= 1: Disallow CPU event access by users without CAP_PERFMON or CAP_SYS_ADMIN >= 2: Disallow kernel profiling by users without CAP_PERFMON or CAP_SYS_ADMIN To make this setting permanent, edit /etc/sysctl.conf too, e.g.: kernel.perf_event_paranoid = -1 [perf@five ~]$ Ok, the message says I need to have CAP_PERFMON, lets do it, using an unpatched libcap that doesn't know about it but we can use 38, CAP_PERFMON value instead, and I tested this with a patched libcap as well, same results: As root: [root@five bin]# setcap "38,cap_sys_ptrace,cap_syslog=ep" perf [root@five bin]# Back to the 'perf' user in the 'perf_users' group, ok, so now 'perf record -a' works for system wide sampling of cycles:u, i.e. only userspace samples, but 'perf top' is failing: [perf@five ~]$ type perf perf is hashed (/home/perf/bin/perf) [perf@five ~]$ getcap /home/perf/bin/perf /home/perf/bin/perf = cap_sys_ptrace,cap_syslog,38+ep [perf@five ~]$ groups perf perf_users [perf@five ~]$ id uid=1002(perf) gid=1002(perf) groups=1002(perf),1003(perf_users) context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 [perf@five ~]$ perf top --stdio Error: Failed to mmap with 1 (Operation not permitted) [perf@five ~]$ perf record -a ^C[ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.177 MB perf.data (1552 samples) ] [perf@five ~]$ perf evlist cycles:u [perf@five ~]$ - Arnaldo _______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2020-04-07 14:30 UTC|newest] Thread overview: 118+ messages / expand[flat|nested] mbox.gz Atom feed top 2020-04-02 8:42 [PATCH v8 00/12] Introduce CAP_PERFMON to secure system performance monitoring and observability Alexey Budankov 2020-04-02 8:42 ` [Intel-gfx] " Alexey Budankov 2020-04-02 8:45 ` [PATCH v8 01/12] capabilities: introduce CAP_PERFMON to kernel and user space Alexey Budankov 2020-04-02 8:45 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] capabilities: Introduce " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:46 ` [PATCH v8 02/12] perf/core: open access to the core for CAP_PERFMON privileged process Alexey Budankov 2020-04-02 8:46 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] perf/core: Open " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:47 ` [PATCH v8 03/12] perf/core: open access to probes " Alexey Budankov 2020-04-02 8:47 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:47 ` [PATCH v8 04/12] perf tool: extend Perf tool with CAP_PERFMON capability support Alexey Budankov 2020-04-02 8:47 ` [Intel-gfx] " Alexey Budankov 2020-04-03 11:08 ` Jiri Olsa 2020-04-03 11:08 ` [Intel-gfx] " Jiri Olsa 2020-04-03 13:08 ` Alexey Budankov 2020-04-03 13:08 ` [Intel-gfx] " Alexey Budankov 2020-04-04 2:18 ` Namhyung Kim 2020-04-04 2:18 ` [Intel-gfx] " Namhyung Kim 2020-04-04 8:18 ` Alexey Budankov 2020-04-04 8:18 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] perf tools: Support CAP_PERFMON capability tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:48 ` [PATCH v8 05/12] drm/i915/perf: open access for CAP_PERFMON privileged process Alexey Budankov 2020-04-02 8:48 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] drm/i915/perf: Open " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:48 ` [PATCH v8 06/12] trace/bpf_trace: open " Alexey Budankov 2020-04-02 8:48 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] trace/bpf_trace: Open " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:49 ` [PATCH v8 07/12] powerpc/perf: open " Alexey Budankov 2020-04-02 8:49 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:50 ` [PATCH v8 08/12] parisc/perf: " Alexey Budankov 2020-04-02 8:50 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:51 ` [PATCH v8 09/12] drivers/perf: " Alexey Budankov 2020-04-02 8:51 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] drivers/perf: Open " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:53 ` [PATCH v8 10/12] drivers/oprofile: open " Alexey Budankov 2020-04-02 8:53 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] drivers/oprofile: Open " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:54 ` [PATCH v8 11/12] doc/admin-guide: update perf-security.rst with CAP_PERFMON information Alexey Budankov 2020-04-02 8:54 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] doc/admin-guide: Update " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 8:54 ` [PATCH v8 12/12] doc/admin-guide: update kernel.rst " Alexey Budankov 2020-04-02 8:54 ` [Intel-gfx] " Alexey Budankov 2020-04-05 14:10 ` Arnaldo Carvalho de Melo 2020-04-05 14:10 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-05 14:41 ` Alexey Budankov 2020-04-05 14:41 ` [Intel-gfx] " Alexey Budankov 2020-04-05 14:54 ` Alexey Budankov 2020-04-05 14:54 ` [Intel-gfx] " Alexey Budankov 2020-04-05 15:05 ` Arnaldo Carvalho de Melo 2020-04-05 15:05 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-05 15:51 ` Alexey Budankov 2020-04-05 15:51 ` [Intel-gfx] " Alexey Budankov 2020-04-22 12:17 ` [tip: perf/core] " tip-bot2 for Alexey Budankov 2020-04-22 12:17 ` [Intel-gfx] " tip-bot2 for Alexey Budankov 2020-04-02 9:37 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for Introduce CAP_PERFMON to secure system performance monitoring and observability (rev5) Patchwork 2020-04-02 10:02 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork 2020-04-03 11:24 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork 2020-04-05 16:00 ` [Intel-gfx] ✗ Fi.CI.BUILD: failure for Introduce CAP_PERFMON to secure system performance monitoring and observability (rev6) Patchwork 2020-04-07 14:30 ` Arnaldo Carvalho de Melo [this message] 2020-04-07 14:30 ` [Intel-gfx] [PATCH v8 00/12] Introduce CAP_PERFMON to secure system performance monitoring and observability Arnaldo Carvalho de Melo 2020-04-07 14:35 ` Arnaldo Carvalho de Melo 2020-04-07 14:35 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-07 14:54 ` Alexey Budankov 2020-04-07 14:54 ` [Intel-gfx] " Alexey Budankov 2020-04-07 16:36 ` Arnaldo Carvalho de Melo 2020-04-07 16:36 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-07 16:40 ` Arnaldo Carvalho de Melo 2020-04-07 16:40 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-07 17:17 ` Alexey Budankov 2020-04-07 17:17 ` [Intel-gfx] " Alexey Budankov 2020-04-07 16:52 ` Alexey Budankov 2020-04-07 16:52 ` [Intel-gfx] " Alexey Budankov 2020-04-07 17:02 ` Arnaldo Carvalho de Melo 2020-04-07 17:02 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-07 17:32 ` Alexey Budankov 2020-04-07 17:32 ` [Intel-gfx] " Alexey Budankov 2020-04-07 16:56 ` Arnaldo Carvalho de Melo 2020-04-07 16:56 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-07 17:23 ` Arnaldo Carvalho de Melo 2020-04-07 17:23 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-04-22 19:18 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for Introduce CAP_PERFMON to secure system performance monitoring and observability (rev18) Patchwork 2020-04-22 19:43 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork 2020-07-10 13:31 ` [PATCH v8 00/12] Introduce CAP_PERFMON to secure system performance monitoring and observability Ravi Bangoria 2020-07-10 13:31 ` [Intel-gfx] " Ravi Bangoria 2020-07-10 14:30 ` Alexey Budankov 2020-07-10 14:30 ` [Intel-gfx] " Alexey Budankov 2020-07-10 17:09 ` Arnaldo Carvalho de Melo 2020-07-10 17:09 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-07-13 9:48 ` Alexey Budankov 2020-07-13 9:48 ` [Intel-gfx] " Alexey Budankov 2020-07-13 12:17 ` Arnaldo Carvalho de Melo 2020-07-13 12:17 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-07-13 12:37 ` Alexey Budankov 2020-07-13 12:37 ` [Intel-gfx] " Alexey Budankov 2020-07-13 18:51 ` Arnaldo Carvalho de Melo 2020-07-13 18:51 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-07-14 10:59 ` Peter Zijlstra 2020-07-14 10:59 ` [Intel-gfx] " Peter Zijlstra 2020-07-14 15:27 ` Arnaldo Carvalho de Melo 2020-07-14 15:27 ` [Intel-gfx] " Arnaldo Carvalho de Melo 2020-07-21 13:06 ` Alexey Budankov 2020-07-21 13:06 ` [Intel-gfx] " Alexey Budankov 2020-07-22 11:30 ` Arnaldo Carvalho de Melo 2020-07-22 11:30 ` [Intel-gfx] " Arnaldo Carvalho de Melo
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20200407143014.GD11186@kernel.org \ --to=arnaldo.melo@gmail.com \ --cc=ak@linux.intel.com \ --cc=alexey.budankov@linux.intel.com \ --cc=ast@kernel.org \ --cc=eranian@google.com \ --cc=ilubashe@akamai.com \ --cc=intel-gfx@lists.freedesktop.org \ --cc=jmorris@namei.org \ --cc=jolsa@redhat.com \ --cc=linux-doc@vger.kernel.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-man@vger.kernel.org \ --cc=linux-security-module@vger.kernel.org \ --cc=mingo@redhat.com \ --cc=namhyung@kernel.org \ --cc=peterz@infradead.org \ --cc=selinux@vger.kernel.org \ --cc=serge@hallyn.com \ --cc=songliubraving@fb.com \ --cc=tglx@linutronix.de \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.