debuggers.hg
changeset 20854:db8a882f5515
Make sure the minimum shadow allocation is never zero.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
author | Keir Fraser <keir.fraser@citrix.com> |
---|---|
date | Thu Jan 14 14:11:25 2010 +0000 (2010-01-14) |
parents | 4b8843ecd553 |
children | 88d2273c3942 |
files | xen/arch/x86/mm/shadow/common.c |
line diff
1.1 --- a/xen/arch/x86/mm/shadow/common.c Thu Jan 14 14:10:40 2010 +0000 1.2 +++ b/xen/arch/x86/mm/shadow/common.c Thu Jan 14 14:11:25 2010 +0000 1.3 @@ -1244,10 +1244,11 @@ int shadow_cmpxchg_guest_entry(struct vc 1.4 * instruction, we must be able to map a large number (about thirty) VAs 1.5 * at the same time, which means that to guarantee progress, we must 1.6 * allow for more than ninety allocated pages per vcpu. We round that 1.7 - * up to 128 pages, or half a megabyte per vcpu. */ 1.8 + * up to 128 pages, or half a megabyte per vcpu, and add 1 more vcpu's 1.9 + * worth to make sure we never return zero. */ 1.10 static unsigned int shadow_min_acceptable_pages(struct domain *d) 1.11 { 1.12 - u32 vcpu_count = 0; 1.13 + u32 vcpu_count = 1; 1.14 struct vcpu *v; 1.15 1.16 for_each_vcpu(d, v)