Messages in this thread | | | Subject | Re: Lazy FPU restoration / moving kernel_fpu_end() to context switch | From | Rik van Riel <> | Date | Fri, 15 Jun 2018 16:56:58 -0400 |
| |
On Fri, 2018-06-15 at 13:42 -0700, Dave Hansen wrote: > On 06/15/2018 01:33 PM, Jason A. Donenfeld wrote: > > On Fri, Jun 15, 2018 at 8:32 PM Andy Lutomirski <luto@kernel.org> > > wrote: > > > quite in the form you imagined. The idea that we've tossed > > > around is > > > to restore FPU state on return to user mode. Roughly, we'd > > > introduce > > > a new thread flag TIF_FPU_UNLOADED (name TBD). > > > prepare_exit_to_usermode() would notice this flag, copy the > > > fpstate to > > > fpregs, and clear the flag. (Or maybe exit_to_usermode_loop() -- > > > No > > > one has quite thought it through, but I think it should be > > > outside the > > > loop.) We'd update all the FPU accessors to understand the flag. > > > > Yes! This is exactly what I was thinking. Then those calls to > > begin() > > and end() could be placed as close to the actual FPU usage as > > possible. > > Andy, what was the specific concern about PKRU? That we might do: > > kernel_fpu_begin(); <- Saves the first time > something() > kernel_fpu_end(); <- Does not XRSTOR > > copy_from_user(); <- Sees old PKRU, does the wrong thing > > prepare_exit_to_usermode(); <- Does the XRSTOR > // only now does PKRU have the right value > SYSRET/IRET > > ? > > Does that *matter* unless something() modified PKRU? We could just > make > the rule that nobody is supposed to mess with it and that it's not > covered by kernel_fpu_begin/end() semantics. We could even > theoretically enforce that in a debug environment if we watch its > value.
KVM needs to change out guest and host PKRU values when switching between guest and host mode, but since f775b13eedee ("x86,kvm: move qemu/guest FPU switching out to vcpu_run") that no longer happens under kernel_fpu_begin/end so we don't need to care about that :)
-- All Rights Reversed.[unhandled content-type:application/pgp-signature] | |