On Fri, Jul 31, 2015 at 05:27:13PM +0200, Borislav Petkov wrote: > On Fri, Jul 31, 2015 at 05:08:06PM +0200, Peter Zijlstra wrote: > > But its things like set_memory_XX(), and afaict that's all buggy against > > MTRR modifications. > > I think the idea is to not do any MTRR modifications at some point: > > From Documentation/x86/pat.txt: > > "... Ideally mtrr_add() usage will be phased out in favor of > arch_phys_wc_add() which will be a no-op on PAT enabled systems. The > region over which a arch_phys_wc_add() is made, should already have been > ioremapped with WC attributes or PAT entries, this can be done by using > ioremap_wc() / set_memory_wc()." I need to update this documentation to remove set_memory_wc() there as we've learned with the MTRR --> PAT conversion that set_memory_wc() cannot be used on IO memory, it can only be used for RAM. I am not sure if I would call it being broken that you cannot use set_memory_*() for IO memory that may have been by design. Luis -- 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/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>