Re: [Xen-devel] [PATCH] [RFC] x86/domctl: Fix getpageframeinfo* handling.

2015-05-18 Thread Jan Beulich
>>> On 18.05.15 at 15:24, wrote: > On 18/05/15 12:43, Jan Beulich wrote: > On 18.05.15 at 12:59, wrote: >>> +if ( unlikely(num > 1024) || >>> + unlikely(num != domctl->u.getpageframeinfo3.num) ) >>> +{ >>> +ret = -E2BIG; >>> +break; >>> +

Re: [Xen-devel] [PATCH] [RFC] x86/domctl: Fix getpageframeinfo* handling.

2015-05-18 Thread Andrew Cooper
On 18/05/15 12:43, Jan Beulich wrote: On 18.05.15 at 12:59, wrote: >> In tree, there is one single caller of XEN_DOMCTL_getpageframeinfo3 >> (xc_get_pfn_type_batch()), and no callers of the older variants. >> >> getpageframeinfo3 and getpageframeinfo2 are compatible if the parameter >> conten

[Xen-devel] [PATCH] [RFC] x86/domctl: Fix getpageframeinfo* handling.

2015-05-18 Thread Andrew Cooper
In tree, there is one single caller of XEN_DOMCTL_getpageframeinfo3 (xc_get_pfn_type_batch()), and no callers of the older variants. getpageframeinfo3 and getpageframeinfo2 are compatible if the parameter contents are considered to be unsigned long; a compat guest calling getpageframeinfo3 falls t

Re: [Xen-devel] [PATCH] [RFC] x86/domctl: Fix getpageframeinfo* handling.

2015-05-18 Thread Jan Beulich
>>> On 18.05.15 at 12:59, wrote: > In tree, there is one single caller of XEN_DOMCTL_getpageframeinfo3 > (xc_get_pfn_type_batch()), and no callers of the older variants. > > getpageframeinfo3 and getpageframeinfo2 are compatible if the parameter > contents are considered to be unsigned long; a co