Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-28 Thread Zhenzhong Duan
On 2019/1/28 16:36, Thomas Gleixner wrote: On Mon, 28 Jan 2019, Zhenzhong Duan wrote: On 2019/1/26 2:03, Thomas Gleixner wrote: Bah, nonsense. Brain was clearly still out for lunch and I confused IBPB and STIBP for a moment. cond_ibpb() is the thing issues in switch_mm() and that is not leaving

Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-28 Thread Thomas Gleixner
On Mon, 28 Jan 2019, Zhenzhong Duan wrote: > On 2019/1/26 2:03, Thomas Gleixner wrote: > > Bah, nonsense. Brain was clearly still out for lunch and I confused IBPB > > and STIBP for a moment. cond_ibpb() is the thing issues in switch_mm() and > > that is not leaving a stale MSR around because we on

Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-28 Thread Zhenzhong Duan
On 2019/1/26 2:03, Thomas Gleixner wrote: On Fri, 25 Jan 2019, Thomas Gleixner wrote: On Wed, 23 Jan 2019, Thomas Gleixner wrote: On Fri, 18 Jan 2019, Zhenzhong Duan wrote: When a task is set for updating TIF_SPEC_IB throuth SECCOMP by others and it's scheduled in the first time, a stale TIF

Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-25 Thread Thomas Gleixner
On Fri, 25 Jan 2019, Thomas Gleixner wrote: > On Wed, 23 Jan 2019, Thomas Gleixner wrote: > > > On Fri, 18 Jan 2019, Zhenzhong Duan wrote: > > > > > When a task is set for updating TIF_SPEC_IB throuth SECCOMP by others > > > and it's scheduled in the first time, a stale TIF_SPEC_IB value is > > >

Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-25 Thread Thomas Gleixner
On Wed, 23 Jan 2019, Thomas Gleixner wrote: > On Fri, 18 Jan 2019, Zhenzhong Duan wrote: > > > When a task is set for updating TIF_SPEC_IB throuth SECCOMP by others > > and it's scheduled in the first time, a stale TIF_SPEC_IB value is > > picked in cond_ibpb(). This is due to TIF_SPEC_IB is upda

Re: [PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-23 Thread Thomas Gleixner
On Fri, 18 Jan 2019, Zhenzhong Duan wrote: > When a task is set for updating TIF_SPEC_IB throuth SECCOMP by others > and it's scheduled in the first time, a stale TIF_SPEC_IB value is > picked in cond_ibpb(). This is due to TIF_SPEC_IB is updated later at > __switch_to_xtra(). > > Add an extra ca

[PATCH] x86/speculation: Update TIF_SPEC_IB before ibpb barrier

2019-01-18 Thread Zhenzhong Duan
When a task is set for updating TIF_SPEC_IB throuth SECCOMP by others and it's scheduled in the first time, a stale TIF_SPEC_IB value is picked in cond_ibpb(). This is due to TIF_SPEC_IB is updated later at __switch_to_xtra(). Add an extra call to speculation_ctrl_update_tif() to update it before