On Fri, Nov 04, 2011 at 03:16:03PM +0000, Ralf Baechle wrote: > On Fri, Oct 28, 2011 at 09:26:23PM +0800, Hillf Danton wrote: > > When we were chasing the TLB crash recently for a while I was suspecting > the default get_user_fast implementation so I ported gup.c from x86 but I > never really finished it, so thanks for ridding me of one item from my > to do list. > > This probably wants some more testing in particular on 32-bit systems with > 64-bit pagetables such as some Alchemy configurations or 32-bit kernels > on Sibyte systems. > > I fixed up a reject in arch/mips/mm/Makefile due to whitespace differences > and queued the patch for 3.3. > > Have you made any benchmarks of the new gup.c? Hitting this one in a non-hugepage build of upstream-sfr: CC arch/mips/mm/gup.o arch/mips/mm/gup.c:70:51: error: redefinition of ‘get_huge_page_tail’ include/linux/mm.h:379:51: note: previous definition of ‘get_huge_page_tail’ was here make[2]: *** [arch/mips/mm/gup.o] Error 1 make[1]: *** [arch/mips/mm] Error 2 make: *** [arch/mips] Error 2 I fixed this one up by removing the local definition of get_huge_page_tail but you may want to re-test. Ralf
