Xen devel - Sep 2007 - [PATCH][VTPM] re-send:Fix vTPM function in HVM domain

Hi,

In Windows on HVM domain, GetPubkey function and Seal function become an error.
Attached is a patch that fixes the error.

In this patch, I use tpm-emulator0.5 code that fixed this bug.

Signed-off-by: Kouichi YASAKI <yasaki.kouichi@jp.fujitsu.com>

Thanks
  Kouichi YASAKI


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

Hi,

In HVM domain, MA_Transmit function in tcgbios sometimes become an error
(TCG_NO_RESPONSE). The cause of the error is not to make connection of
QEMU and TPMD instance within a timeout of MA_Transmit function.

Before the MA_Transmit function was called, the attached patch corrected so
that connection of QEMU and TPMD might be completed.

Signed-off-by: Kouichi YASAKI <yasaki.kouichi@jp.fujitsu.com>

Thanks
  Kouichi YASAKI



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

xen-devel-bounces@lists.xensource.com wrote on 12/19/2007 07:53:11 PM:
> Hi,
> 
> In HVM domain, MA_Transmit function in tcgbios sometimes become an error
> (TCG_NO_RESPONSE). The cause of the error is not to make connection of
> QEMU and TPMD instance within a timeout of MA_Transmit function.
> 
> Before the MA_Transmit function was called, the attached patch corrected 
so
> that connection of QEMU and TPMD might be completed.
> 
> Signed-off-by: Kouichi YASAKI <yasaki.kouichi@jp.fujitsu.com>
> 
> Thanks
>   Kouichi YASAKI
> 
> diff -r d9ab9eb2bfee tools/ioemu/hw/tpm_tis.c
> --- a/tools/ioemu/hw/tpm_tis.c   Sat Dec 15 18:29:27 2007 +0000
> +++ b/tools/ioemu/hw/tpm_tis.c   Mon Dec 17 19:46:42 2007 +0900
> @@ -904,6 +904,10 @@ void tpm_tis_init(SetIRQFunc *set_irq, v
>      memset(s->buffer.buf,0,sizeof(s->buffer.buf));
> 
>      register_savevm("tpm-tis", 0, 1, tpm_save, tpm_load, s);
> +
> +    while(!IS_COMM_WITH_VTPM(s)){
> +       open_vtpm_channel(s);
> +    }
>  }
I''ll have a look at this. The problem probably stems from the vTPM
manager
starting the vTPM up too late while qemu is already up and running, so 
it''s a timing problem between the two processes. I don''t think
it should
try to connect endlessly in a busy loop. At least there should be a 
counter that tries this for maybe 5 times followed by a [u]sleep() in the 
loop.

    int ctr = 0;
    while(!IS_COMM_WITH_VTPM(s) && ctr < 5){
       open_vtpm_channel(s);
       ctr ++;
       sleep(1);
    }


   Stefan
> 
> 
> 
/****************************************************************************/
> _______________________________________________
> 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

Hi Stefan-san,

Thank you for correcting my patch.
I also think that it should not try to connect endlessly in a busy loop.

attached file is the corrected patch.

Thanks
   Kouichi YASAKI
> 
> xen-devel-bounces@lists.xensource.com wrote on 12/19/2007 07:53:11 PM:
> 
>  > Hi,
>  >
>  > In HVM domain, MA_Transmit function in tcgbios sometimes become an
error
>  > (TCG_NO_RESPONSE). The cause of the error is not to make connection
of
>  > QEMU and TPMD instance within a timeout of MA_Transmit function.
>  >
>  > Before the MA_Transmit function was called, the attached patch 
> corrected so
>  > that connection of QEMU and TPMD might be completed.
>  >
>  > Signed-off-by: Kouichi YASAKI <yasaki.kouichi@jp.fujitsu.com>
>  >
>  > Thanks
>  >   Kouichi YASAKI
>  >
>  > diff -r d9ab9eb2bfee tools/ioemu/hw/tpm_tis.c
>  > --- a/tools/ioemu/hw/tpm_tis.c   Sat Dec 15 18:29:27 2007 +0000
>  > +++ b/tools/ioemu/hw/tpm_tis.c   Mon Dec 17 19:46:42 2007 +0900
>  > @@ -904,6 +904,10 @@ void tpm_tis_init(SetIRQFunc *set_irq, v
>  >      memset(s->buffer.buf,0,sizeof(s->buffer.buf));
>  >  
>  >      register_savevm("tpm-tis", 0, 1, tpm_save, tpm_load,
s);
>  > +
>  > +    while(!IS_COMM_WITH_VTPM(s)){
>  > +       open_vtpm_channel(s);
>  > +    }
>  >  }
> 
> I''ll have a look at this. The problem probably stems from the vTPM
> manager starting the vTPM up too late while qemu is already up and 
> running, so it''s a timing problem between the two processes. I
don''t
> think it should try to connect endlessly in a busy loop. At least there 
> should be a counter that tries this for maybe 5 times followed by a 
> [u]sleep() in the loop.
> 
>     int ctr = 0;
>     while(!IS_COMM_WITH_VTPM(s) && ctr < 5){
>       open_vtpm_channel(s);
>        ctr ++;
>        sleep(1);
>    }
> 
> 
>    Stefan
> 
>  >  
>  >  
>  > 
>
/****************************************************************************/
>  > _______________________________________________
>  > 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