On 04/26/2012 12:07 AM, Minchan Kim wrote: > > Quick patch - totally untested. > > We can implement new TLB flush function > "local_flush_tlb_kernel_range" If architecture is very smart, it > could flush only tlb entries related to vaddr. If architecture is > smart, it could flush only tlb entries related to a CPU. If > architecture is _NOT_ smart, it could flush all entries of all CPUs. > > Now there are few architectures have "local_flush_tlb_kernel_range". > MIPS, sh, unicore32, arm, score and x86 by this patch. So I think > it's good candidate other arch should implement. Until that, we can > add stub for other architectures which calls only [global/local] TLB > flush. We can expect maintainer could respond then they can > implement best efficient method. If the maintainer doesn't have any > interest, zsmalloc could be very slow in that arch and users will > blame that architecture. > > Any thoughts? I had this same idea a while back. It is encouraging to know that someone else independently thought of this solution too :) Makes me think it is a good solution. Let me build and test on x86, make sure there are no unforseen consequences. Thanks again for your work here! Seth -- 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>