>>> On 09.06.17 at 22:07, <boris.ostrov...@oracle.com> wrote: > On 06/09/2017 10:50 AM, Jan Beulich wrote: >>>>> On 19.05.17 at 17:50, <boris.ostrov...@oracle.com> wrote: >>> + unsigned int first_dirty; >> On x86 this change is fine at present, albeit not optimal. Its ARM >> equivalent, however, grows struct page_info in the 32-bit case, > > It does? I am looking at include/asm-arm/mm.h and I don't see this.
The union with field name u is 32 bits on 32-bit ARM right now. You change grows it to 64 bits. >> which I don't think is wanted or needed. You really only need >> MAX_ORDER+1 bits here, so I'd suggest making this a bit field > > Just to make sure --- when you say "bit field" you mean masking various > bits in a word, not C language bit fields? No, actually the latter. Jan _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
