Re: [Xen-devel] Uses of &frame_table[xfn] - Keir Fraser - com.xensource.lists.xen-devel

29 messages in com.xensource.lists.xen-develRe: [Xen-devel] Uses of &frame_table[...

From	Sent On	Attachments
Magenheimer, Dan (HP Labs Fort Collins)	18 Dec 2005 19:18
Ian Pratt	18 Dec 2005 20:23
Magenheimer, Dan (HP Labs Fort Collins)	22 Dec 2005 15:28
Ewan Mellor	23 Dec 2005 02:28
Keir Fraser	23 Dec 2005 03:46
Magenheimer, Dan (HP Labs Fort Collins)	23 Dec 2005 05:05
Magenheimer, Dan (HP Labs Fort Collins)	23 Dec 2005 05:27
Keir Fraser	23 Dec 2005 06:37
Ewan Mellor	23 Dec 2005 07:27
Magenheimer, Dan (HP Labs Fort Collins)	23 Dec 2005 12:33
Tian, Kevin	26 Dec 2005 01:12
Magenheimer, Dan (HP Labs Fort Collins)	28 Dec 2005 14:15
Muli Ben-Yehuda	28 Dec 2005 14:23
Magenheimer, Dan (HP Labs Fort Collins)	28 Dec 2005 14:30
Magenheimer, Dan (HP Labs Fort Collins)	28 Dec 2005 14:42
Cihula, Joseph	28 Dec 2005 17:49
Tian, Kevin	28 Dec 2005 17:58
Keir Fraser	29 Dec 2005 05:47
Magenheimer, Dan (HP Labs Fort Collins)	29 Dec 2005 08:12
Keir Fraser	29 Dec 2005 08:15
Magenheimer, Dan (HP Labs Fort Collins)	29 Dec 2005 10:51
Keir Fraser	29 Dec 2005 12:33
Magenheimer, Dan (HP Labs Fort Collins)	29 Dec 2005 13:38
Tian, Kevin	29 Dec 2005 18:15
Ewan Mellor	30 Dec 2005 05:40
Kevin Fox	30 Dec 2005 08:35
Magenheimer, Dan (HP Labs Fort Collins)	30 Dec 2005 11:49
Mark Williamson	02 Jan 2006 11:22
Hollis Blanchard	03 Jan 2006 13:54

Subject:	Re: [Xen-devel] Uses of &frame_table[xfn]
From:	Keir Fraser (Keir...@cl.cam.ac.uk)
Date:	12/23/2005 03:46:50 AM
List:	com.xensource.lists.xen-devel

On 22 Dec 2005, at 23:29, Magenheimer, Dan (HP Labs Fort Collins) wrote:

The balloon driver in xenlinux was also fixed, but I think this is a syntactic but not a semantic fix. In balloon_init, an assumption is made that all pages below xen_start_info->nr_pages are valid RAM. What if all pages of valid RAM are above nr_pages? Then (I think) the balloon driver will eat up all the domain's RAM. (Unlikely, but you get the drift.)

The assumption is that you are initially allocated contiguous RAM from pseudophys page frame 0 up to pseudophys page frame nr_pages-1.

A further assumption is that, if the max_pfn is larger than nr_pages, the physical space between nr_pages and max_pfn is not (yet) populated with RAM. Thus those pages properly belong to the balloon driver.

Are these assumptions not workable on ia64?

-- Keir