[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-devel] [PATCH for-4.10] xen/x86: p2m-pod: Prevent infinite loop when shattering 1GB pages
The PoD subsystem only have pool of 4KB and 2MB pages. When it comes accross a 1GB mapping, it will be splitted in 2MB one using p2m_set_entry and request the caller to retry (see ept_get_entry for instance). p2m_set_entry may fail to shatter if it is not possible to allocate memory for the new page table. However, the error is not progated resulting to the callers to retry infinitely the PoD. Prevent the infinite loop by return false when it is not possible to shatter the 1GB mapping. Signed-off-by: Julien Grall <julien.grall@xxxxxxxxxx> --- This is a potential candidate to backport and for Xen 4.10. Without it, there a potential for infinite loop if the memory is exhausted. --- xen/arch/x86/mm/p2m-pod.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/xen/arch/x86/mm/p2m-pod.c b/xen/arch/x86/mm/p2m-pod.c index 0a811ccf28..69269a0bd1 100644 --- a/xen/arch/x86/mm/p2m-pod.c +++ b/xen/arch/x86/mm/p2m-pod.c @@ -1103,6 +1103,8 @@ p2m_pod_demand_populate(struct p2m_domain *p2m, gfn_t gfn, */ if ( order == PAGE_ORDER_1G ) { + int rc; + pod_unlock(p2m); /* * Note that we are supposed to call p2m_set_entry() 512 times to @@ -1113,9 +1115,9 @@ p2m_pod_demand_populate(struct p2m_domain *p2m, gfn_t gfn, * NOTE: In a fine-grained p2m locking scenario this operation * may need to promote its locking from gfn->1g superpage */ - p2m_set_entry(p2m, gfn_aligned, INVALID_MFN, PAGE_ORDER_2M, - p2m_populate_on_demand, p2m->default_access); - return true; + rc = p2m_set_entry(p2m, gfn_aligned, INVALID_MFN, PAGE_ORDER_2M, + p2m_populate_on_demand, p2m->default_access); + return !rc; } /* Only reclaim if we're in actual need of more cache. */ -- 2.11.0 _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxx https://lists.xen.org/xen-devel
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |