On Tue, Oct 19, 2010 at 6:32 AM, Scott Wood <scottw...@freescale.com> wrote: > If mem= is used on the kernel command line to create reserved regions > for userspace to map using /dev/mem, let it be mapped cacheable as long > as it is within the memory region described in the device tree. > > Signed-off-by: Scott Wood <scottw...@freescale.com> > --- > This isn't just a performance issue, but it could also be a correctness > issue, if the reserved portion of RAM is mapped cacheable by e.g. a KVM > guest. This patch does not address cases where such regions could show up > as something other than a standard memory node -- such as shared regions > in an AMP configuration. Ideally there would be some means for a platform > to register cacheable regions, without having to completely replace the > entire phys_mem_access_prot function.
Agreed. Such requirement might be special case in server/desktop world, but much more common in embedded market when migrating to multi-core in which shared memory is commonly used for inter-core communication. It will be best to have a common way to achieve this. Not only cache-ability but also coherent property might need to be tunable. - Leo _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
