Hi, On Mon, Dec 19, 2011 at 02:24:53PM -0500, KOSAKI Motohiro wrote: > (12/19/11 1:38 PM), Naoya Horiguchi wrote: > > This flag shows that a given pages is a subpage of transparent hugepage. > > It does not care about whether it is a head page or a tail page, because > > it's clear from pfn of the target page which you should know when you read > > /proc/kpageflags. > > > > Signed-off-by: Naoya Horiguchi<n-horiguchi@xxxxxxxxxxxxx> > > NAK. > > The detail of transparent hugepage are hidden by design. We hope it > keep 'transparent'. > Until any explain why we should expose KPF_THP, we don't agree it. The reason why I want to know physical address of thp is testing. I'm working on memory error recovery and writing test code to confirm that memory recovery really works when an error occurs on thps. There I need to locate thps on the physical memory. IMO, transparency in thp means that we need no manual setup to use it (as a contrast with hugetlbfs,) so it seems to me that exporting pageflag of thp does not break the design of thp. Anyway, I should have written the purpose in the patch description. Thanks for the comment. Naoya -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>