Messages in this thread | | | From | Joel Fernandes <> | Date | Tue, 15 Jun 2021 07:31:59 -0400 | Subject | Re: [PATCH 18/19] sched: prctl() core-scheduling interface |
| |
On Mon, Jun 14, 2021 at 7:36 PM Josh Don <joshdon@google.com> wrote: > > On Thu, Apr 22, 2021 at 5:36 AM Peter Zijlstra <peterz@infradead.org> wrote: > > > > From: Chris Hyser <chris.hyser@oracle.com> > > > > This patch provides support for setting and copying core scheduling > > 'task cookies' between threads (PID), processes (TGID), and process > > groups (PGID). > > [snip] > > Internally, we have lots of trusted processes that don't have a > security need for coresched cookies. However, these processes could > still decide to create cookies for themselves, which will degrade > machine capacity and performance for other jobs on the machine. > > Any thoughts on whether it would be desirable to have the ability to > restrict use of SCHED_CORE_CREATE? Perhaps a new SCHED_CORE capability > would be appropriate?
Hi, Maybe a capability may not work because then other users who don't care for the issue you mention will be required to manage/assign the capability as well?
How about you use seccomp to filter the prctl based on the PID, and CREATE command?
-Joel
| |