RE: [Xen-ia64-devel] [PATCH] unify vtlb and vhpt

[Kouya Shimura <kouya@xxxxxxxxxxxxxx>]

tgingold@xxxxxxx writes:
> to be honest I have mixed feelings about your patch.  I like it but I don't
> really understand its purpose.  See my comment.
> 
> I still think it would be nice if the vTLB were TR-mapped.

IMO, it is not worthwhile to map the vTLB on a TR. 
I surely explained one of advantages, but that is a extra feature.

> By merging vTLB and VHPT the user can't anymore set the size of the vTLB.
> This is obvious.  But is your patch different from increasing vTLB size ?
> Did I miss a point ?
> 
> I am not sure it is a good idea to remove vTLB size.  On a real processor
> the TLB structure is fixed and defined.

My explanation was bad. Yes, my patch increases vTLB size equal to
VHPT size.
I assumed that VHPT size is much much bigger than vTLB size. 
What I wanted to say is, a user can get good performance without
user's setting of vTLB size.

In addition, the advantage of my patch is not to take time for ptc.e
emulation rather speedup. I strongly recommend my patch for this advantage.

> This seems to be the same algorithm as the one for Madison.  Cf Matthew
> Chapman pages.

Thank you for pointing out. 
Although the algorithm is not so concerned with unification of vTLB
and VHPT.

Thanks,
Kouya

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel