On 03/22/2013 02:47 AM, Kirill A. Shutemov wrote: > Dave Hansen wrote: >> On 03/14/2013 10:50 AM, Kirill A. Shutemov wrote: >>> +#define RADIX_TREE_PRELOAD_NR 512 /* For THP's benefit */ >> >> This eventually boils down to making the radix_tree_preload array >> larger. Do we really want to do this unconditionally if it's only for >> THP's benefit? > > It will be useful not only for THP. Batching can be useful to solve > scalability issues. Still, it seems like something that little machines with no THP support probably don't want to pay the cost for. Perhaps you could enable it for THP||NR_CPUS>$FOO. >> For those of us too lazy to go compile a kernel and figure this out in >> practice, how much bigger does this make the nodes[] array? > > We have three possible RADIX_TREE_MAP_SHIFT: > > #ifdef __KERNEL__ > #define RADIX_TREE_MAP_SHIFT (CONFIG_BASE_SMALL ? 4 : 6) > #else > #define RADIX_TREE_MAP_SHIFT 3 /* For more stressful testing */ > #endif > > On 64-bit system: > For RADIX_TREE_MAP_SHIFT=3, old array size is 43, new is 107. > For RADIX_TREE_MAP_SHIFT=4, old array size is 31, new is 63. > For RADIX_TREE_MAP_SHIFT=6, old array size is 21, new is 30. > > On 32-bit system: > For RADIX_TREE_MAP_SHIFT=3, old array size is 21, new is 84. > For RADIX_TREE_MAP_SHIFT=4, old array size is 15, new is 46. > For RADIX_TREE_MAP_SHIFT=6, old array size is 11, new is 19. > > On most machines we will have RADIX_TREE_MAP_SHIFT=6. Could you stick that in your patch description? The total cost is "array size" * sizeof(void*) * NR_CPUS, right? -- Dave Hansen, Intel OTC Scalability Team -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
