Xen devel - Sep 2005 - Build breakage in vtpm (missing dependency on libgpm)

Howdy,

The recent vtpm checkin has broken the build for me since it introduces 
a dependency on libgpm.

I''d prefer that we only build vtpm if libgpm is present.  In the very 
least, we should add a check for it if we always want to build it.

Thoughts?

Regards,

Anthony Liguori

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

>Howdy,
>
>The recent vtpm checkin has broken the build for me since it introduces 
>a dependency on libgpm.
Ah - didn''t spot that since I had it installed on all my machines. 
>I''d prefer that we only build vtpm if libgpm is present.  In the
very
>least, we should add a check for it if we always want to build it.
>
Certainly - can you knock up a patch?

[note - the vtpm stuff needs more testing, particularly with save/restore; 
if someone is running w/out netgrants it''d be good if they could
confirm
it doesn''t break save/restore for them] 

cheers,

S.

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

xen-devel-bounces@lists.xensource.com wrote on 08/31/2005 11:51:08 PM:
> 
> >Howdy,
> >
> >The recent vtpm checkin has broken the build for me since it introduces
> >a dependency on libgpm.
> 
> Ah - didn''t spot that since I had it installed on all my machines.
> 
> >I''d prefer that we only build vtpm if libgpm is present.  In
the very
> >least, we should add a check for it if we always want to build it.
> >
> 
> Certainly - can you knock up a patch?
> 
> [note - the vtpm stuff needs more testing, particularly with 
save/restore; 
> if someone is running w/out netgrants it''d be good if they could
confirm
> it doesn''t break save/restore for them] 
There''s one problem that I know of. There cannot be any outstanding 
request in the backend, but all repsonses have to be collected. I have 
tried to use a time in the frontend''s .suspend function, but that did
not
work. The reason I believe is the __cli() call in the reboot.c before 
xenbus_suspend() which prevents functions like wait_event(), 
interruptible_sleep_on_timeout() and related ones to work properly. Can 
this function be moved to a  later point?

  Stefan
> 
> cheers,
> 
> S.
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

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

Steven Hand wrote:
>>I''d prefer that we only build vtpm if libgpm is present.  In
the very
>>least, we should add a check for it if we always want to build it.
>>
>>    
>>
>
>Certainly - can you knock up a patch?
>  
>
Looks like Keir beat me to it (I swear he''s a machine :-)).
>[note - the vtpm stuff needs more testing, particularly with save/restore; 
>if someone is running w/out netgrants it''d be good if they could
confirm
>it doesn''t break save/restore for them] 
>  
>
Dan Smith has been working on a regression suite for xm.  It might make 
sense to look into adding some TPM tests to it.

Regards,

Anthony Liguori
>cheers,
>
>S.
>
>  
>

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