Xen devel - Apr 2006 - [PATCH 5/5] changes of common files for xen/ia64 dom0 vp model: netback work around

Work around for vnif support of xen/ia64 with dom0 vp model.
This is a temporal work around until figuring out how to place a new page
after grant table page transfer.

-- 
yamahata


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 28 Apr 2006, at 03:56, Isaku Yamahata wrote:
> Work around for vnif support of xen/ia64 with dom0 vp model.
> This is a temporal work around until figuring out how to place a new 
> page
> after grant table page transfer.
Does netback work *at all* for you with vp model? Looks like this patch 
may make it build but not actually work, which makes it seem a bit 
pointless.

Anyway, two possible solutions are apparent to me:
  1. netback gathers ''transmitted'' skbuffs and then, in one
batch, does
a populate_physmap() call to get new pages for all of them, the 
kfree_skb()s each one back to system pool.
  2. change semantics of grant transfers for vp guests so that the 
operation automatically gets you a fresh page at the same 
pseudo-physical address.

Option (2) is maybe the better one? Then this patch would make more 
sense (except that the code would be made run-time dependent on 
autotranslate feature flag).

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On Fri, Apr 28, 2006 at 02:53:20PM +0100, Keir Fraser
wrote:
> 
> On 28 Apr 2006, at 03:56, Isaku Yamahata wrote:
> 
> >Work around for vnif support of xen/ia64 with dom0 vp model.
> >This is a temporal work around until figuring out how to place a new 
> >page
> >after grant table page transfer.
> 
> Does netback work *at all* for you with vp model? Looks like this patch 
> may make it build but not actually work, which makes it seem a bit 
> pointless.
> 
> Anyway, two possible solutions are apparent to me:
>  1. netback gathers ''transmitted'' skbuffs and then, in
one batch, does
> a populate_physmap() call to get new pages for all of them, the 
> kfree_skb()s each one back to system pool.
>  2. change semantics of grant transfers for vp guests so that the 
> operation automatically gets you a fresh page at the same 
> pseudo-physical address.
> 
> Option (2) is maybe the better one? Then this patch would make more 
> sense (except that the code would be made run-time dependent on 
> autotranslate feature flag).
xen/ia64 with vp model adopted (2).
steal_page_for_grant_transfer() does it.
I attached the patch for netback.c. I tested only on xen/ia64 very roughly.
However I think steal_page_for_grant_transfer() in xen/arch/x86/mm.c
(or gnttab_transfer()) also needs modification. Right?

-- 
yamahata


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 29 Apr 2006, at 05:27, Isaku Yamahata wrote:
> xen/ia64 with vp model adopted (2).
> steal_page_for_grant_transfer() does it.
> I attached the patch for netback.c. I tested only on xen/ia64 very 
> roughly.
> However I think steal_page_for_grant_transfer() in xen/arch/x86/mm.c
> (or gnttab_transfer()) also needs modification. Right?
You need to take a bit more care with this patch -- you''ve broken 
non-autotranslate mode (look at the first hunk of the patch).

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On Sun, Apr 30, 2006 at 09:42:12AM +0100, Keir Fraser
wrote:
> 
> On 29 Apr 2006, at 05:27, Isaku Yamahata wrote:
> 
> >xen/ia64 with vp model adopted (2).
> >steal_page_for_grant_transfer() does it.
> >I attached the patch for netback.c. I tested only on xen/ia64 very 
> >roughly.
> >However I think steal_page_for_grant_transfer() in xen/arch/x86/mm.c
> >(or gnttab_transfer()) also needs modification. Right?
> 
> You need to take a bit more care with this patch -- you''ve broken 
> non-autotranslate mode (look at the first hunk of the patch).
Oh sorry. I fixed it and tested on xen/x86_32 and xen/ia64
with ping flooding.

Is there anyone who are working on/using netback with xen/x86 autotranslate
mode set?
This patch may break netback with autotranlsate mode set on xen/x86.
If necessary I can give it a try, however it would be desirable
that someone else who are familiar with shadow mode does it...

-- 
yamahata


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 1 May 2006, at 04:07, Isaku Yamahata wrote:
> Oh sorry. I fixed it and tested on xen/x86_32 and xen/ia64
> with ping flooding.
>
> Is there anyone who are working on/using netback with xen/x86 
> autotranslate
> mode set?
> This patch may break netback with autotranlsate mode set on xen/x86.
> If necessary I can give it a try, however it would be desirable
> that someone else who are familiar with shadow mode does it...
No, netback has never worked with auto-translate on x86. You patch will 
at least make it easier to get it working in future.

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel