* Different: nr_pages vs. max_pfn
@ 2008-11-19 9:45 Y. D.
2008-11-19 9:54 ` Keir Fraser
0 siblings, 1 reply; 4+ messages in thread
From: Y. D. @ 2008-11-19 9:45 UTC (permalink / raw)
To: xen-devel
Hi,
In start_info, there is a field nr_pages definig how much memory is allocated to the domain, which is obvious.
While in shared_info, there is a field max_pfn. I don't know why max_pfn is not equal to nr_pages.
Say for 32MB memory, nr_pages = 8192, while max_pfn = 10240 (may vary?), of which 2048 pfn's have invalid mfn.
Can anybody explain that why max_pfn is larger? Is it static or dynamic?
Furture, does pfn to mfn frame mapping ever change? Hopefully not.
Thanks,
Shawn
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Different: nr_pages vs. max_pfn
2008-11-19 9:45 Different: nr_pages vs. max_pfn Y. D.
@ 2008-11-19 9:54 ` Keir Fraser
2008-11-19 10:23 ` Jayaraman, Bhaskar
2008-11-19 12:22 ` Y. D.
0 siblings, 2 replies; 4+ messages in thread
From: Keir Fraser @ 2008-11-19 9:54 UTC (permalink / raw)
To: Y. D., xen-devel
The shared_info value is set up by the guest itself during boot. It's larger
because Linux adds extra space at the end of its p2m map for backend
drivers.
-- Keir
On 19/11/08 09:45, "Y. D." <duyuyang@gmail.com> wrote:
> Hi,
>
> In start_info, there is a field nr_pages definig how much memory is allocated
> to the domain, which is obvious.
> While in shared_info, there is a field max_pfn. I don't know why max_pfn is
> not equal to nr_pages.
> Say for 32MB memory, nr_pages = 8192, while max_pfn = 10240 (may vary?), of
> which 2048 pfn's have invalid mfn.
> Can anybody explain that why max_pfn is larger? Is it static or dynamic?
> Furture, does pfn to mfn frame mapping ever change? Hopefully not.
>
> Thanks,
> Shawn
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: Different: nr_pages vs. max_pfn
2008-11-19 9:54 ` Keir Fraser
@ 2008-11-19 10:23 ` Jayaraman, Bhaskar
2008-11-19 12:22 ` Y. D.
1 sibling, 0 replies; 4+ messages in thread
From: Jayaraman, Bhaskar @ 2008-11-19 10:23 UTC (permalink / raw)
To: Keir Fraser, Y. D., xen-devel
Keir, does Dom0 Linux do this for each Mini-OS instance that is spawned? If so then shouldn't the memory usage value in xm list for each PV guest take into account this additional usage of memory? i.e. 32MB + 8MB?
Bhaskar.
-----Original Message-----
From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Keir Fraser
Sent: Wednesday, November 19, 2008 3:24 PM
To: Y. D.; xen-devel
Subject: Re: [Xen-devel] Different: nr_pages vs. max_pfn
The shared_info value is set up by the guest itself during boot. It's larger
because Linux adds extra space at the end of its p2m map for backend
drivers.
-- Keir
On 19/11/08 09:45, "Y. D." <duyuyang@gmail.com> wrote:
> Hi,
>
> In start_info, there is a field nr_pages definig how much memory is allocated
> to the domain, which is obvious.
> While in shared_info, there is a field max_pfn. I don't know why max_pfn is
> not equal to nr_pages.
> Say for 32MB memory, nr_pages = 8192, while max_pfn = 10240 (may vary?), of
> which 2048 pfn's have invalid mfn.
> Can anybody explain that why max_pfn is larger? Is it static or dynamic?
> Furture, does pfn to mfn frame mapping ever change? Hopefully not.
>
> Thanks,
> Shawn
>
>
> _______________________________________________
> 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
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: RE: Different: nr_pages vs. max_pfn
2008-11-19 9:54 ` Keir Fraser
2008-11-19 10:23 ` Jayaraman, Bhaskar
@ 2008-11-19 12:22 ` Y. D.
1 sibling, 0 replies; 4+ messages in thread
From: Y. D. @ 2008-11-19 12:22 UTC (permalink / raw)
To: Jayaraman, Bhaskar, Keir Fraser, xen-devel
>The shared_info value is set up by the guest itself during boot. It's larger
>because Linux adds extra space at the end of its p2m map for backend
>drivers.
I tested for various memories, i.e. 64M and 128M, the extra memory is 8M unchanged.
Also, I tested the extra 8M memory is dispersed in the p2m map, not at the end totally.
Can we say the p2m mapping table stays put during dom's lifecycle?
Thanks, keir.
Shawn
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2008-11-19 12:22 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-11-19 9:45 Different: nr_pages vs. max_pfn Y. D.
2008-11-19 9:54 ` Keir Fraser
2008-11-19 10:23 ` Jayaraman, Bhaskar
2008-11-19 12:22 ` Y. D.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.