| Message ID | 4d95341d-d1f4-fc48-f173-a6fedfc70d33@linux.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Documentation/admin-guide: introduce perf-security.rst file and extend perf_event_paranoid documentation | expand |
On Wed, 21 Nov 2018 12:14:14 +0300 Alexey Budankov <alexey.budankov@linux.intel.com> wrote: > +For the purpose of performing security checks Linux implementation splits > +processes into two categories [6]_ : a) privileged processes (whose effective > +user ID is 0, referred to as superuser or root), and b) unprivileged processes > +(whose effective UID is nonzero). Is that really what's going on here? If I understand things correctly, it's looking for CAP_SYS_PTRACE rather than a specific UID; am I missing something here? (Also, you would want "*the* Linux implementation" in the first sentence above). One other thing: > +(whose effective UID is nonzero). Privileged processes bypass all kernel > +security permission checks so perf_events performance monitoring is fully > +available to privileged processes without *access*, *scope* and *resource* > +restrictions. Could I ask for a slight toning down of the markup here? There's a lot of *emphasis* here that isn't really needed and tends to get in the way. Thanks, jon
Hello Jon, On 25.11.2018 22:47, Jonathan Corbet wrote: > On Wed, 21 Nov 2018 12:14:14 +0300 > Alexey Budankov <alexey.budankov@linux.intel.com> wrote: > >> +For the purpose of performing security checks Linux implementation splits >> +processes into two categories [6]_ : a) privileged processes (whose effective >> +user ID is 0, referred to as superuser or root), and b) unprivileged processes >> +(whose effective UID is nonzero). > > Is that really what's going on here? If I understand things correctly, > it's looking for CAP_SYS_PTRACE rather than a specific UID; am I missing > something here? You are right regarding CAP_SYS_PTRACE but this capability is not the only one which is used by perf_events for security checks, so the capabilities clarification is kept aside of these patches, because patches initial intention is to clarify security specifics of sysctl_perf_even_paranoid settings. I agree that the document can be extended with details clarifying capabilities used by perf_events for security checks. > > (Also, you would want "*the* Linux implementation" in the first sentence > above). Accepted. > > One other thing: > >> +(whose effective UID is nonzero). Privileged processes bypass all kernel >> +security permission checks so perf_events performance monitoring is fully >> +available to privileged processes without *access*, *scope* and *resource* >> +restrictions. > > Could I ask for a slight toning down of the markup here? There's a lot of > *emphasis* here that isn't really needed and tends to get in the way. Accepted. Thanks, Alexey > > Thanks, > > jon >
On Mon, 26 Nov 2018 11:57:21 +0300 Alexey Budankov <alexey.budankov@linux.intel.com> wrote: > >> +For the purpose of performing security checks Linux implementation splits > >> +processes into two categories [6]_ : a) privileged processes (whose effective > >> +user ID is 0, referred to as superuser or root), and b) unprivileged processes > >> +(whose effective UID is nonzero). > > > > Is that really what's going on here? If I understand things correctly, > > it's looking for CAP_SYS_PTRACE rather than a specific UID; am I missing > > something here? > > You are right regarding CAP_SYS_PTRACE but this capability is not the only > one which is used by perf_events for security checks, so the capabilities > clarification is kept aside of these patches, because patches initial intention > is to clarify security specifics of sysctl_perf_even_paranoid settings. > > I agree that the document can be extended with details clarifying capabilities > used by perf_events for security checks. I don't really like the idea of adding a document that we know doesn't really describe how the security decision is made. Even a one-line parenthetical saying that things are more complicated and giving a pointer to a place to look for details would help, IMO. Either way, I can merge this, but I'd like to have an ack from the perf folks first. Thanks, jon
On 26.11.2018 23:28, Jonathan Corbet wrote: > On Mon, 26 Nov 2018 11:57:21 +0300 > Alexey Budankov <alexey.budankov@linux.intel.com> wrote: > >>>> +For the purpose of performing security checks Linux implementation splits >>>> +processes into two categories [6]_ : a) privileged processes (whose effective >>>> +user ID is 0, referred to as superuser or root), and b) unprivileged processes >>>> +(whose effective UID is nonzero). >>> >>> Is that really what's going on here? If I understand things correctly, >>> it's looking for CAP_SYS_PTRACE rather than a specific UID; am I missing >>> something here? >> >> You are right regarding CAP_SYS_PTRACE but this capability is not the only >> one which is used by perf_events for security checks, so the capabilities >> clarification is kept aside of these patches, because patches initial intention >> is to clarify security specifics of sysctl_perf_even_paranoid settings. >> >> I agree that the document can be extended with details clarifying capabilities >> used by perf_events for security checks. > > I don't really like the idea of adding a document that we know doesn't > really describe how the security decision is made. Even a one-line > parenthetical saying that things are more complicated and giving a pointer > to a place to look for details would help, IMO. Well, makes sense. It is possible to have something similar to this: ... For the purpose of performing security checks the Linux implementation splits processes into two categories [6]_ : a) privileged processes (whose effective user ID is 0, referred to as superuser or root), and b) unprivileged processes (whose effective UID is nonzero). Privileged processes bypass all kernel security permission checks so perf_events performance monitoring is fully available to privileged processes without access, scope and resource restrictions. Unprivileged processes are subject to full security permission check based on the process's credentials [5]_ (usually: effective UID, effective GID, and supplementary group list). Linux divides the privileges traditionally associated with superuser into distinct units, known as capabilities [6]_, which can be independently enabled and disabled on per-thread basis for processes and files of unprivileged users. Unprivileged processes with enabled CAP_SYS_ADMIN capability are treated as privileged processes with respect to perf_events performance monitoring and bypass *scope* permissions checks in the kernel. Unprivileged processes using perf_events system call API is also subject for PTRACE_MODE_READ_REALCREDS ptrace access mode check [7]_ , whose outcome determines whether monitoring is permitted. So unprivileged process provided with CAP_SYS_PTRACE capability are effectively permitted to pass the check. Other capabilities being granted to unprivileged processes can effectively enable capturing of additional data required for later performance analysis of monitored processes or system. For example CAP_SYSLOG capability permits reading kernel space memory addresses from /proc/kallsyms file. ... Thanks, Alexey > > Either way, I can merge this, but I'd like to have an ack from the perf > folks first. > > Thanks, > > jon >
diff --git a/Documentation/admin-guide/perf-security.rst b/Documentation/admin-guide/perf-security.rst new file mode 100644 index 000000000000..55ccd9394809 --- /dev/null +++ b/Documentation/admin-guide/perf-security.rst @@ -0,0 +1,77 @@ +.. _perf_security: + +Perf Events and tool security +============================= + +Overview +-------- + +Usage of Performance Counters for Linux (perf_events) [1]_ , [2]_ , [3]_ can +impose a considerable risk of leaking sensitive data accessed by monitored +processes. The data leakage is possible both in scenarios of direct usage of +perf_events system call API [2]_ and over data files generated by Perf tool user +mode utility (Perf) [3]_ , [4]_ . The risk depends on the nature of data that +perf_events performance monitoring units (PMU) [2]_ collect and expose for +performance analysis. Having that said perf_events/Perf performance monitoring +is the subject for security access control management [5]_ . + +perf_events/Perf access control +------------------------------- + +For the purpose of performing security checks Linux implementation splits +processes into two categories [6]_ : a) privileged processes (whose effective +user ID is 0, referred to as superuser or root), and b) unprivileged processes +(whose effective UID is nonzero). Privileged processes bypass all kernel +security permission checks so perf_events performance monitoring is fully +available to privileged processes without *access*, *scope* and *resource* +restrictions. Unprivileged processes are subject to full security permission +check based on the process's credentials [5]_ (usually: effective UID, effective +GID, and supplementary group list). + +perf_events/Perf unprivileged users +-------------------------- + +perf_events/Perf *scope* and *access* control for unprivileged processes is +governed by perf_event_paranoid [2]_ setting: + +-1: + Impose no *scope* and *access* restrictions on using perf_events performance + monitoring. Per-user per-cpu perf_event_mlock_kb [2]_ locking limit is + ignored when allocating memory buffers for storing performance data. + This is the least secure mode since allowed monitored *scope* is + maximized and no perf_events specific limits are imposed on *resources* + allocated for performance monitoring. + +>=0: + *scope* includes per-process and system wide performance monitoring + but excludes raw tracepoints and ftrace function tracepoints monitoring. + CPU and system events happened when executing either in user or + in kernel space can be monitored and captured for later analysis. + Per-user per-cpu perf_event_mlock_kb locking limit is imposed but + ignored for unprivileged processes with CAP_IPC_LOCK [6]_ capability. + +>=1: + *scope* includes per-process performance monitoring only and excludes + system wide performance monitoring. CPU and system events happened when + executing either in user or in kernel space can be monitored and + captured for later analysis. Per-user per-cpu perf_event_mlock_kb + locking limit is imposed but ignored for unprivileged processes with + CAP_IPC_LOCK capability. + +>=2: + *scope* includes per-process performance monitoring only. CPU and system + events happened when executing in user space only can be monitored and + captured for later analysis. Per-user per-cpu perf_event_mlock_kb + locking limit is imposed but ignored for unprivileged processes with + CAP_IPC_LOCK capability. + +Bibliography +------------ + +.. [1] `<https://lwn.net/Articles/337493/>`_ +.. [2] `<http://man7.org/linux/man-pages/man2/perf_event_open.2.html>`_ +.. [3] `<http://web.eece.maine.edu/~vweaver/projects/perf_events/>`_ +.. [4] `<https://perf.wiki.kernel.org/index.php/Main_Page>`_ +.. [5] `<https://www.kernel.org/doc/html/latest/security/credentials.html>`_ +.. [6] `<http://man7.org/linux/man-pages/man7/capabilities.7.html>`_ +
Implement initial version of perf-security.rst documentation file initially covering security concerns related to perf_events/Perf performance monitoring in multiuser environments. Suggested-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Alexey Budankov <alexey.budankov@linux.intel.com> --- Changes in v2: - replaced PCL referencing by perf_events - skipped >=3 setting documentation at the moment --- Documentation/admin-guide/perf-security.rst | 77 +++++++++++++++++++++ 1 file changed, 77 insertions(+) create mode 100644 Documentation/admin-guide/perf-security.rst