Re: [Xen-devel] [PATCH] xenpaging:add a new array to speed up page-in in xenpaging

[Hongkaixing <hongkaixing@xxxxxxxxxx>]

> -----邮件原件-----
> 发件人: Ian Jackson [mailto:Ian.Jackson@xxxxxxxxxxxxx]
> 发送时间: 2012年1月6日 2:27
> 收件人: hongkaixing@xxxxxxxxxx
> 抄送: Olaf Hering; xen-devel@xxxxxxxxxxxxxxxxxxx
> 主题: Re: [Xen-devel] [PATCH] xenpaging:add a new array to speed up
page-in
> in xenpaging
> 
> hongkaixing@xxxxxxxxxx writes ("[Xen-devel] [PATCH] xenpaging:add a new
> array to speed up page-in in xenpaging"):
> > xenpaging:add a new array to speed up page-in in xenpaging
> >
> > This patch adds a new array named page_out_index to reserve the
> > victim's index.  When page in a page,it has to go through a for loop
> > from 0 to num_pages to find the right page to read,and it costs much
> > time in this loop.After adding the page_out_index array,it just
> > reads the arrry to get the right page,and saves much time.
> 
> Thanks for this submission.  Olaf may well have some comments, but I
> have some too:
> 
> > @@ -660,6 +672,7 @@
> >              break;
> >          if ( i % 100 == 0 )
> >              DPRINTF("%d pages evicted\n", i);
> > +        page_out_index[victims[i].gfn].index=i;
> 
> Surely this should be better done much closer to the actual point
> where we page out ?

After we improve the page-in speed, search index in the for loop becomes the
bottleneck, extremely when the page number is big.

> 
> And you should reset the index entry when the page is read back in, I
> think ?

En, our idea is let it go lazy way. I will try my best to make it perfect.

> 
> Ian.


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