Re: [PATCH v2 3/3] x86/fpu/xstate: Invalidate fpregs when __fpu_restore_sig() fails

[Sebastian Andrzej Siewior]

On 2019-12-18 12:53:59 [-0800], Yu-cheng Yu wrote:
> I could have explained this better, sorry!  I will explain the first
> case below; other cases are similar.
> 
> In copy_user_to_fpregs_zeroing(), we have:
> 
>     if (user_xsave()) {
>         ...
>         if (unlikely(init_bv))
>             copy_kernel_to_xregs(&init_fpstate.xsave, init_bv);
>         return copy_user_to_xregs(buf, xbv);
>         ...
>     }
> 
> The copy_user_to_xregs() may fail, and when that happens, before going to
> the slow path, there is fpregs_unlock() and context switches may happen.

The context switch may only happen after fpregs_unlock().

> However, at this point, fpu_fpregs_owner_ctx has not been changed; it could
> still be another task's FPU.

TIF_NEED_FPU_LOAD is set for the task in __fpu__restore_sig() and its
context (__fpu_invalidate_fpregs_state()) has been invalidated. So the
FPU register may contain another task's content and
fpu_fpregs_owner_ctx points to another context.

>                               For this to happen and to be detected, the user
> stack page needs to be non-present, fpu_fpregs_owner_ctx need to be another task,
> and that other task needs to be able to detect its registers are modified.
> The last factor is not easy to reproduce, and a CET control-protection fault
> helps.

So far everything is legal. However. If there is a context switch before
fpregs_lock() then this is bad before we don't account for that.
So that:

diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c
--- a/arch/x86/kernel/fpu/signal.c
+++ b/arch/x86/kernel/fpu/signal.c
@@ -352,6 +352,7 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size)
 			fpregs_unlock();
 			return 0;
 		}
+		fpregs_deactivate(fpu);
 		fpregs_unlock();
 	}
 
@@ -403,6 +404,8 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size)
 	}
 	if (!ret)
 		fpregs_mark_activate();
+	else
+		fpregs_deactivate(fpu);
 	fpregs_unlock();
 
 err_out:


Should be enough.

> > Can you tell me which glibc test? I would like to reproduce this.
> > 
> > > The introduction of supervisor xstates and CET, while not contributing to
> > > the problem, makes it more detectable.  After init_fpstate and the Shadow
> > > Stack pointer have been restored to xregs, the XRSTOR from user stack
> > > fails and fpu_fpregs_owner_ctx is not updated.  The task currently owning
> > > fpregs then uses the corrupted Shadow Stack pointer and triggers a control-
> > > protection fault.
> > 
> > So I don't need new HW with supervisor and CET? A plain KVM box with
> > SSE2 and so should be enough?
> 
> What I do is, clone the whole glibc source, and run mutiple copies of
> "make check".  In about 40 minutes or so, there are unexplained seg faults,
> or a few control-protection faults (if you enable CET).  Please let me
> know if more clarification is needed.

Okay. Can you please try the above and if not, I try that glibc thing myself.

> Thanks,
> Yu-cheng

Sebastian