[PATCH][1/6] add a hypercall number for virtual device in unmodified guest

[PATCH][1/6] add a hypercall number for virtual device in unmodified guest

[Xiaofeng Ling]

add a hypercall number for virtual device in unmodified guest
There are 6 patches for para-driver support in vmx guest. This is the 
first one.

Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>


diff -r 287d36b46fa3 xen/arch/x86/x86_32/entry.S
--- a/xen/arch/x86/x86_32/entry.S	Tue Aug 30 20:36:49 2005
+++ b/xen/arch/x86/x86_32/entry.S	Fri Sep  2 22:46:13 2005
@@ -812,6 +812,7 @@
          .long do_ni_hypercall       /* 25 */
          .long do_mmuext_op
          .long do_acm_op             /* 27 */
+        .long do_virtual_device_op       /* virutal device op for VMX */
          .rept NR_hypercalls-((.-hypercall_table)/4)
          .long do_ni_hypercall
          .endr

RE: [PATCH][1/6] add a hypercall number for virtual device in unmodified guest

[Ian Pratt]

> Subject: [PATCH][1/6] add a hypercall number for virtual 
> device in unmodified guest
> 
> add a hypercall number for virtual device in unmodified guest 
> There are 6 patches for para-driver support in vmx guest. 
> This is the first one.

Xiaofeng,

Please can you write a few paragraphs describing the xen-side changes. I
thought we were going to have a separate hypercall table for hvm guests,
but it looks like you've not gone down this route. Please can you
describe your approach to simplify the review.

Thanks,
Ian

> Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
> Signed-off-by: Arun Sharma <arun.sharma@intel.com>
> 
> 
> diff -r 287d36b46fa3 xen/arch/x86/x86_32/entry.S
> --- a/xen/arch/x86/x86_32/entry.S	Tue Aug 30 20:36:49 2005
> +++ b/xen/arch/x86/x86_32/entry.S	Fri Sep  2 22:46:13 2005
> @@ -812,6 +812,7 @@
>           .long do_ni_hypercall       /* 25 */
>           .long do_mmuext_op
>           .long do_acm_op             /* 27 */
> +        .long do_virtual_device_op       /* virutal device 
> op for VMX */
>           .rept NR_hypercalls-((.-hypercall_table)/4)
>           .long do_ni_hypercall
>           .endr
> 
> 

RE: [PATCH][1/6] add a hypercall number for virtual device in unmodified guest

[Ling, Xiaofeng]

Ian Pratt <mailto:m+Ian.Pratt@cl.cam.ac.uk> wrote:
>> Subject: [PATCH][1/6] add a hypercall number for virtual device in
>> unmodified guest 
> Xiaofeng,
> 
> Please can you write a few paragraphs describing the xen-side
> changes. I thought we were going to have a separate hypercall table
> for hvm guests, but it looks like you've not gone down this route.
> Please can you describe your approach to simplify the review.   
> 
ok. The changes for hypervisor part include these 6 patches.
Currently, the hypervisor change has reduced a lot compare to previous 
patch.
The changes mainly for 3 parts.
1. make vmx guest can use hypercall.(patch 1,2)
2. deliver event by a irq for vmx guest.( patch 3,5)
3. setup process for 2 kinds of share page: (patch 4, 6)
      1) hypercall parameter sharepage, 
      2) grant table share page


patch1: Just reserver one hypercall number for para-driver. Even use a
seperate call table, we need to reserver one number.
patch2: hypercall entry for vmcall, in this patch, I use a permit bitmap
to check and then call the hypercall_table. To use a seperate call table
is also ok. I'll send the patch later.
patch3 add a call backirq member 
patch4 two virtual device operation
patch5 grant table share page setup
patch6 inject callback irq

For xen-linux patch.
ctrl_if.c   just resolve the setup_irq problem
gnttab.c  for setup share page
blkfront  use macro for virt_to_mfn and page_to_phys
xenbus  use macro for xenstore share page. 
a hypercall wrap.   copy paramters to the share page before the hypercall.

I've removed the support of phys_to_machine_map in guest.
for blkfront, the translation will be done in grant table for shadow_translate mode