WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Xen

xen-devel

[Top] [All Lists]

[Xen-devel] confusion about monitor table in shadow mode

To: "'Tim Deegan'" <Tim.Deegan@xxxxxxxxxx>
Subject: [Xen-devel] confusion about monitor table in shadow mode
From: "Li Yaqiong" <liyaq04@xxxxxxxxxx>
Date: 2009年1月15日 11:32:14 +0800
Cc: xen-devel@xxxxxxxxxxxxxxxxxxx
Delivery-date: 2009年1月14日 19:33:10 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
Importance: Normal
In-reply-to: <20090112124730.GE22263@xxxxxxxxxxxxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
Hi,Tim
 I have confusion when understanding monitor table in shadow
mode. In arch/x86/mm/shadow/mutli.c, the definition of
sh_make_monitor_table is following:
 #if SHADOW_PAGING_LEVELS == GUEST_PAGING_LEVELS
 mfn_t
 sh_make_monitor_table(struct vcpu *v)
 {
 .......
 }
The macro SHADOW_PAGING_LEVELS is defined in arch/x86/mm/shadow/types.h:
 #if GUEST_PAGING_LEVELS == 4
 #define SHADOW_PAGING_LEVELS 4
 #else
 #define SHADOW_PAGING_LEVELS 3
 #endif
If a 32-bit hvm is created, its GUEST_PAGING_LEVELS should be 2. In this
case, SHADOW_PAGING_LEVELS will be 3 and sh_make_monitor_table will not
be compiled. Is it correct? But, when a hvm domain is initialized, Xen
will call sh_update_paging_modes to create a monitor table for it:
In sh_update_paging_modes:
 #ifdef __x86_64__
 if ( hvm_long_mode_enabled(v) )
 {
 // long mode guest...
 v->arch.paging.mode =
 &SHADOW_INTERNAL_NAME(sh_paging_mode, 4);
 }
 else
 #endif
 if ( hvm_pae_enabled(v) )
 {
 // 32-bit PAE mode guest...
 v->arch.paging.mode =
 &SHADOW_INTERNAL_NAME(sh_paging_mode, 3);
 }
 else
 {
 // 32-bit 2 level guest...
 v->arch.paging.mode =
 &SHADOW_INTERNAL_NAME(sh_paging_mode, 2);
 }
 }
 if ( pagetable_is_null(v->arch.monitor_table) )
 {
 mfn_t mmfn =
v->arch.paging.mode->shadow.make_monitor_table(v);
 v->arch.monitor_table = pagetable_from_mfn(mmfn);
 make_cr3(v, mfn_x(mmfn));
 hvm_update_host_cr3(v);
 }
This function will finally be called by hvm_domain_initialise. But
sh_make_monitor_table is NULL-poinitor. What will happened in this case?
Need your help!
-Techie
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
<Prev in Thread] Current Thread [Next in Thread>
Previous by Date: Re: [Xen-devel] [PATCH] Support cross-bitness guest when core-dumping , Isaku Yamahata
Next by Date: Re: [Xen-devel] [PATCH] Support cross-bitness guest when core-dumping , Bruce Rogers
Previous by Thread: [Xen-devel] Move some of the PCI device manage/control into pciback? , Cui, Dexuan
Next by Thread: [Xen-devel] Re: confusion about monitor table in shadow mode , Tim Deegan
Indexes: [Date] [Thread] [Top] [All Lists]

Copyright ©, Citrix Systems Inc. All rights reserved. Legal and Privacy
Citrix This site is hosted by Citrix

AltStyle によって変換されたページ (->オリジナル) /