Re: [PATCH] Add ability to drop pages through testdev

[Gleb Natapov <gleb@redhat.com>]

On Wed, Dec 29, 2010 at 11:58:31AM +0200, Avi Kivity wrote:
> On 12/29/2010 11:55 AM, Gleb Natapov wrote:
> >On Wed, Dec 29, 2010 at 11:50:48AM +0200, Avi Kivity wrote:
> >>  On 12/22/2010 05:18 PM, Gleb Natapov wrote:
> >>  >Signed-off-by: Gleb Natapov<gleb@redhat.com>
> >>  >diff --git a/hw/testdev.c b/hw/testdev.c
> >>  >index d1abf59..29df385 100644
> >>  >--- a/hw/testdev.c
> >>  >+++ b/hw/testdev.c
> >>  >@@ -1,3 +1,4 @@
> >>  >+#include<sys/mman.h>
> >>  >   #include "hw.h"
> >>  >   #include "qdev.h"
> >>  >   #include "isa.h"
> >>  >@@ -46,6 +47,16 @@ static uint32_t test_device_ioport_read(void *opaque, uint32_t addr)
> >>  >       return test_device_ioport_data;
> >>  >   }
> >>  >
> >>  >+static void test_device_flush_page(void *opaque, uint32_t addr, uint32_t data)
> >>  >+{
> >>  >+    target_phys_addr_t len = 4096;
> >>  >+    void *a = cpu_physical_memory_map(data&   ~0xffful,&len, 0);
> >>  >+
> >>  >+    mprotect(a, 4096, PROT_NONE);
> >>  >+    mprotect(a, 4096, PROT_READ|PROT_WRITE);
> >>  >+    cpu_physical_memory_unmap(a, len, 0, 0);
> >>  >+}
> >>  >+
> >>
> >>  Icky.  This is much better done through the api tests.  Applied it
> >>  so as not to spoil all the effort.
> >>
> >How would you do that there?
> 
> Set up all the state using the KVM_SET_REGS family, pointing to the
> instruction you want to test, and KVM_RUN that.  You can even queue
> exceptions and interrupts for complicated cases.
> 
I mean how do you drop a page from shadow/ept tables? Wouldn't you have
to do the same trick there?

--
			Gleb.