Re: [RFC PATCH] kvm-unit-tests : Basic architecture of VMX nested test case

[Gleb Natapov <gleb@redhat.com>]

On Fri, Jul 19, 2013 at 08:42:20AM +0200, Paolo Bonzini wrote:
> Il 18/07/2013 21:57, Gleb Natapov ha scritto:
> > On Thu, Jul 18, 2013 at 02:08:51PM +0200, Paolo Bonzini wrote:
> >> Il 18/07/2013 13:06, Gleb Natapov ha scritto:
> >>> On Thu, Jul 18, 2013 at 12:47:46PM +0200, Paolo Bonzini wrote:
> >>>>>> and for a testsuite I'd prefer the latter---which means I'd still favor
> >>>>>> setjmp/longjmp.
> >>>>>>
> >>>>>> Now, here is the long explanation.
> >>>>>>
> >>>>>> I must admit that the code looks nice.  There are some nits I'd like to
> >>>>>> see done differently (such as putting vmx_return at the beginning of the
> >>>>>> while (1), and the vmresume asm at the end), but it is indeed nice.
> >>>>>
> >>>>> Why do you prefer setjmp/longjmp then?
> >>>>
> >>>> Because it is still deceiving, and I dislike the deceit more than I like
> >>>> the linear code flow.
> >>>>
> >>> What is deceiving about it? Of course for someone who has no idea how
> >>> vmx works the code will not be obvious, but why should we care. For
> >>> someone who knows what is deceiving about returning into the same
> >>> function guest was launched from by using VMX mechanism
> >>
> >> The way the code is written is deceiving.  If I see
> >>
> >>   asm("vmlaunch; seta %0")
> >>   while (ret)
> >>
> >> I expect HOST_RIP to point at the seta or somewhere near, not at a
> >> completely different label somewhere else.
> >>
> > Why would you expect that assuming you know what vmlaunch is?
> 
> Because this is written in C, and I know trying to fool the compiler is
> a losing game.  So my reaction is "okay, HOST_RIP must be set so that
> code will not jump around".  If I see
> 
>    asm("vmlaunch")
>    exit(-1)
> 
> the reaction is the opposite: "hmm, anything that jumps around would
> have a hard time with the compiler, there must be some assembly
> trampoline somewhere; let's check what HOST_RIP is".
> 
We do try to fool compiler often even without vmx: code patching. This is
why asm goto was invented, no? So, like you said in previous emails,
asm goto is appropriate way here to tell compiler what is going on.

> >>>> instead of longjmp()?
> >>
> >> Look again at the setjmp/longjmp version.  longjmp is not used to handle
> >> vmexit.  It is used to jump back from the vmexit handler to main, which
> >> is exactly what setjmp/longjmp is meant for.
> >>
> > That's because simple return will not work in that version, this is
> > artifact of how vmexit was done.
> 
> I think it can be made to work without setjmp/longjmp, but the code
> would be ugly.
> 
> >>>> the compiler, and you rely on the compiler not changing %rsp between the
> >>>> vmlaunch and the vmx_return label.  Minor nit, you cannot anymore print
> >>> HOST_RSP should be loaded on each guest entry.
> >>
> >> Right, but my point is: without a big asm blob, you don't know the right
> >> value to load.  It depends on the generated code.  And the big asm blob
> >> limits a lot the "code looks nice" value of this approach.
> >>
> > I said it number of times already, this is not about "code looking nice",
> > "code looks like KVM" or use less assembler as possible", this is about
> > linear data flow. It is not fun chasing code execution path. Yes, you
> > can argue that vmlaunch/vmresume inherently non linear, but there is a
> > difference between skipping one instruction and remain in the same
> > function and on the same stack, or jump completely to a different
> > context.
> 
> I don't see anything bad in jumping completely to a different context.
> The guest and host are sort of like two coroutines, they hardly have any
> connection with the code that called vmlaunch.
You can see it as two coroutines, or you can see it as linear logic:
  do host stuff
  enter guest
  do guest stuff
  exit guest
  continue doing host stuff

Both can be implemented, I prefer linear one. I would prefer linear one
to coroutine in any code design, no connection to vmx. Sometimes
coroutine are better than alternative (and in those cases alternative is
never a linear logic), but this is not the case.


> 
> > The actually differences in asm instruction between both
> > version will not be bigger then a couple of lines (if we will not take
> > setjmp/longjmp implementation into account :)),
> 
> I was waiting for this parenthetical remark to appear. ;)
> 
I've put a smile there :) I realize this argument is not completely fair,
but for the sake of argument everything goes!

--
			Gleb.