Re: [PATCH v2 10/13] KVM: x86: Shrink the usercopy region of the emulation context

[Sean Christopherson <sean.j.christopherson@xxxxxxxxx>]

On Mon, Mar 02, 2020 at 07:40:27PM +0100, Paolo Bonzini wrote:
> On 26/02/20 18:51, Vitaly Kuznetsov wrote:
> >> +
> >> +	/* Here begins the usercopy section. */
> >> +	struct operand src;
> >> +	struct operand src2;
> >> +	struct operand dst;
> > Out of pure curiosity, how certain are we that this is going to be
> > enough for userspaces?
> > 
> 
> And also, where exactly are the user copies done?

Anything that funnels into ctxt->ops->read_std() or ctxt->ops->write_std(),
e.g.

	if (ctxt->src2.type == OP_MEM) {
		rc = segmented_read(ctxt, ctxt->src2.addr.mem,
				    &ctxt->src2.val, ctxt->src2.bytes);
		if (rc != X86EMUL_CONTINUE)
			goto done;
	}


segmented_read() : @data = &ctxt->src2.val
|
|-> read_emulated()
    |
    |-> ctxt->ops->read_emulated() / emulator_read_emulated()
        |
        |-> emulator_read_write()
            |
            |-> emulator_read_write_onepage()
                |
                |-> ops->read_write_emulate() / read_emulate()
                    |
                    |-> kvm_vcpu_read_guest()
                        |
                        ...
                          |-> __kvm_read_guest_page()
                              |
                              |-> __copy_from_user(data, ...)