From mboxrd@z Thu Jan 1 00:00:00 1970 From: Keir Fraser Subject: Re: xc_get_pfn_list() creates broken core files Date: Thu, 23 Nov 2006 19:05:08 +0000 Message-ID: References: <8A87A9A84C201449A0C56B728ACF491E01F981@liverpoolst.ad.cl.cam.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <8A87A9A84C201449A0C56B728ACF491E01F981@liverpoolst.ad.cl.cam.ac.uk> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Ian Pratt , Keir Fraser , John Levon Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org On 23/11/06 6:53 pm, "Ian Pratt" wrote: >> Pagetables etc. are all in 'p' address space, so there should >> be no need for p2m or m2p in that case I think >> >> In fact pagetables could be canonicalised into 'p' space for >> PV guests too (xc_linux_save has the code for this) then >> there'd be no need to dump the p2m table in the core dump at >> all. Save format and core dump format should probably be one >> and the same! > > Trouble is, the PV guest hasn't done an orderly suspend, and may be > holding machine address references in registers/memory and in pagetables > that haven't been pinned yet. > > The guest's core dump will already contain the p2m table, and its root > location has been registered with xen during boot, so this just needs to > be saved too. It's questionable how useful the ability to m2p-translate those marginally possible bits of state is though. Currently the p2m table is used *only* to construct an m2p table allowing pagetable walks in xc_ptrace_core(). -- Keir