[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH V2 2/2] x86/vm_event: fix race between __context_switch() and vm_event_resume()

>>> On 03.05.17 at 11:10, <rcojocaru@xxxxxxxxxxxxxxx> wrote:
> The introspection agent can reply to a vm_event faster than
> vmx_vmexit_handler() can complete in some cases, where it is then
> not safe for vm_event_set_registers() to modify v->arch.user_regs.
> In the test scenario, we were stepping over an INT3 breakpoint by
> setting RIP += 1. The quick reply tended to complete before the VCPU
> triggering the introspection event had properly paused and been
> descheduled. If the reply occurs before __context_switch() happens,
> __context_switch() clobbers the reply by overwriting
> v->arch.user_regs from the stack. If the reply occurs after
> __context_switch(), we don't pass through __context_switch() when
> transitioning to idle.

This last sentence still looks to be wrong (and even self-contradictory).
The reply can't occur after __context_switch() if we don't make it there
in the first place. How about "If we don't pass through
__context_switch() (due to switching to the idle vCPU), reply data
wouldn't be picked up when switching back straight to the original

> This patch ensures that vm_event_resume() code only sets per-VCPU
> data to be used for the actual setting of registers later in
> hvm_do_resume() (similar to the model used to control setting of CRs
> and MSRs).
> The patch additionally removes the sync_vcpu_execstate(v) call from
> vm_event_resume(), which is no longer necessary, which removes the
> associated broadcast TLB flush (read: performance improvement).
> Signed-off-by: Razvan Cojocaru <rcojocaru@xxxxxxxxxxxxxxx>
> Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>

Code changes themselves
Reviewed-by: Jan Beulich <jbeulich@xxxxxxxx>


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.