Re: [PATCHv5 25/37] x86/vdso: Switch image on setns()/clone()

2019-08-01 Thread Andy Lutomirski
On Wed, Jul 31, 2019 at 11:09 PM wrote: > > On July 31, 2019 10:34:26 PM PDT, Andy Lutomirski wrote: > >On Mon, Jul 29, 2019 at 2:58 PM Dmitry Safonov wrote: > >> > >> As it has been discussed on timens RFC, adding a new conditional > >branch > >> `if (inside_time_ns)` on VDSO for all processes

Re: [PATCHv5 25/37] x86/vdso: Switch image on setns()/clone()

2019-07-31 Thread hpa
On July 31, 2019 10:34:26 PM PDT, Andy Lutomirski wrote: >On Mon, Jul 29, 2019 at 2:58 PM Dmitry Safonov wrote: >> >> As it has been discussed on timens RFC, adding a new conditional >branch >> `if (inside_time_ns)` on VDSO for all processes is undesirable. >> It will add a penalty for everybody

Re: [PATCHv5 25/37] x86/vdso: Switch image on setns()/clone()

2019-07-31 Thread Andy Lutomirski
On Mon, Jul 29, 2019 at 2:58 PM Dmitry Safonov wrote: > > As it has been discussed on timens RFC, adding a new conditional branch > `if (inside_time_ns)` on VDSO for all processes is undesirable. > It will add a penalty for everybody as branch predictor may mispredict > the jump. Also there are in

[PATCHv5 25/37] x86/vdso: Switch image on setns()/clone()

2019-07-29 Thread Dmitry Safonov
As it has been discussed on timens RFC, adding a new conditional branch `if (inside_time_ns)` on VDSO for all processes is undesirable. It will add a penalty for everybody as branch predictor may mispredict the jump. Also there are instruction cache lines wasted on cmp/jmp. Those effects of introd