Messages in this thread |  | | Date | Fri, 5 Feb 2021 13:46:55 +0100 | From | Peter Zijlstra <> | Subject | Re: [PATCH 1/2] x86/hw_breakpoint: Prevent data breakpoints on __per_cpu_offset |
| |
On Fri, Feb 05, 2021 at 12:45:54PM +0100, Thomas Gleixner wrote: > On Thu, Feb 04 2021 at 16:11, Andy Lutomirski wrote: > > On Thu, Feb 4, 2021 at 6:26 AM Lai Jiangshan <jiangshanlai@gmail.com> wrote: > >> When FSGSBASE is enabled, paranoid_entry() fetches the per-CPU > >> GSBASE value via __per_cpu_offset or pcpu_unit_offsets. > >> > >> When data breakpoint is set on __per_cpu_offset[cpu] (read-write > >> operation), the specific cpu will be stuck in the infinite #DB loop. > >> RCU will try to send NMI to the specific cpu, but it is not working > >> either since NMI also relies on paranoid_entry(). > > > > Should we consider having a .percpu..noinstr section and having > > objtool enforce this? > > I think so.
I'll put it on the TODO list somewhere ...
|  |