[PATCH] [ACM] Comilation fix for 32 bit

[PATCH] [ACM] Comilation fix for 32 bit

[Stefan Berger]

[-- Attachment #1: Type: text/plain, Size: 84 bytes --]

This fixes a compilation problem on 32 bit machines.

Signed-off-by: Stefan Berger


[-- Attachment #2: acm_buffer_type.diff --]
[-- Type: text/x-patch, Size: 1126 bytes --]

Index: root/xen-unstable.hg/xen/include/public/acm_ops.h
===================================================================
--- root.orig/xen-unstable.hg/xen/include/public/acm_ops.h
+++ root/xen-unstable.hg/xen/include/public/acm_ops.h
@@ -50,7 +50,7 @@
 struct acm_setpolicy {
     /* IN */
     uint32_t interface_version;
-    XEN_GUEST_HANDLE_64(void) pushcache;
+    XEN_GUEST_HANDLE(void) pushcache;
     uint32_t pushcache_size;
 };
 
@@ -59,7 +59,7 @@ struct acm_setpolicy {
 struct acm_getpolicy {
     /* IN */
     uint32_t interface_version;
-    XEN_GUEST_HANDLE_64(void) pullcache;
+    XEN_GUEST_HANDLE(void) pullcache;
     uint32_t pullcache_size;
 };
 
@@ -68,7 +68,7 @@ struct acm_getpolicy {
 struct acm_dumpstats {
     /* IN */
     uint32_t interface_version;
-    XEN_GUEST_HANDLE_64(void) pullcache;
+    XEN_GUEST_HANDLE(void) pullcache;
     uint32_t pullcache_size;
 };
 
@@ -84,7 +84,7 @@ struct acm_getssid {
         domaintype_t domainid;
         ssidref_t    ssidref;
     } id;
-    XEN_GUEST_HANDLE_64(void) ssidbuf;
+    XEN_GUEST_HANDLE(void) ssidbuf;
     uint32_t ssidbuf_size;
 };
 

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

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

Re: [PATCH] [ACM] Comilation fix for 32 bit

[Keir Fraser]

On 26/4/07 04:36, "Stefan Berger" <stefanb@us.ibm.com> wrote:

> This fixes a compilation problem on 32 bit machines.
> 
> Signed-off-by: Stefan Berger

What's the issue?

You're better off keeping the GUEST_HANDLE_64() and getting rid of the
compat acm stuff. As far as I can see, all your interfaces are *currently*
32/64-bit invariant. So the compat stuff can safely be removed and then you
don't need to continually keep it sync.

 -- Keir

Re: Re: [PATCH] [ACM] Comilation fix for 32 bit

[Stefan Berger]

[-- Attachment #1.1: Type: text/plain, Size: 1027 bytes --]

xen-devel-bounces@lists.xensource.com wrote on 04/26/2007 02:50:49 AM:

> On 26/4/07 04:36, "Stefan Berger" <stefanb@us.ibm.com> wrote:
> 
> > This fixes a compilation problem on 32 bit machines.
> > 
> > Signed-off-by: Stefan Berger
> 
> What's the issue?

It does not compile on 32bit because of those 4 lines. 64bit is fine.

> 
> You're better off keeping the GUEST_HANDLE_64() and getting rid of the
> compat acm stuff. As far as I can see, all your interfaces are 
*currently*
> 32/64-bit invariant. So the compat stuff can safely be removed and then 
you
> don't need to continually keep it sync.

I was going to ask you about this. I am writing on a patch that puts all 
the ACM hypercall structures into a union. I am wondering what the compat 
stuff is for (xen/common/compat) and what the criterion is for needing 
this compat code. 

   Stefan 

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

[-- Attachment #1.2: Type: text/html, Size: 1428 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: Re: [PATCH] [ACM] Comilation fix for 32 bit

[Keir Fraser]

[-- Attachment #1.1: Type: text/plain, Size: 1084 bytes --]

On 26/4/07 11:49, "Stefan Berger" <stefanb@us.ibm.com> wrote:

>> > You're better off keeping the GUEST_HANDLE_64() and getting rid of the
>> > compat acm stuff. As far as I can see, all your interfaces are *currently*
>> > 32/64-bit invariant. So the compat stuff can safely be removed and then you
>> > don't need to continually keep it sync.
> 
> I was going to ask you about this. I am writing on a patch that puts all the
> ACM hypercall structures into a union. I am wondering what the compat stuff is
> for (xen/common/compat) and what the criterion is for needing this compat
> code. 
> 
If you have structures that differ in layout in 32-bit mode vs. 64-bit mode
then you need a compat shim when running 32-bit dom0 on 64-bit Xen (which we
support doing nowadays). Currently you have no such issues that I¹m aware
of. You should be able to delete the compat stuff that is there currently
with no problems.

Is the build failure in Xen itself or in tools? I¹m sure it¹s easily fixed
without needing to throw away the GUEST_HANDLE_64() usage.

 -- Keir


[-- Attachment #1.2: Type: text/html, Size: 1647 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: Re: [PATCH] [ACM] Comilation fix for 32 bit

[Stefan Berger]

[-- Attachment #1.1: Type: text/plain, Size: 1380 bytes --]

xen-devel-bounces@lists.xensource.com wrote on 04/26/2007 06:55:06 AM:

> On 26/4/07 11:49, "Stefan Berger" <stefanb@us.ibm.com> wrote:

> > You're better off keeping the GUEST_HANDLE_64() and getting rid of the
> > compat acm stuff. As far as I can see, all your interfaces are 
*currently*
> > 32/64-bit invariant. So the compat stuff can safely be removed and 
then you
> > don't need to continually keep it sync. 
> 
> I was going to ask you about this. I am writing on a patch that puts
> all the ACM hypercall structures into a union. I am wondering what 
> the compat stuff is for (xen/common/compat) and what the criterion 
> is for needing this compat code. 

> If you have structures that differ in layout in 32-bit mode vs. 64-
> bit mode then you need a compat shim when running 32-bit dom0 on 64-
> bit Xen (which we support doing nowadays). Currently you have no 
> such issues that I?m aware of. You should be able to delete the 
> compat stuff that is there currently with no problems.

Ok.
> 
> Is the build failure in Xen itself or in tools? I?m sure it?s easily
> fixed without needing to throw away the GUEST_HANDLE_64() usage.

It does not build in Xen.

   Stefan

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

[-- Attachment #1.2: Type: text/html, Size: 1859 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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