On 07/24/13 18:28, Or Gerlitz wrote:
On 23/07/2013 17:47, Bart Van Assche wrote:
Still regarding page sizes: shouldn't ib_alloc_fast_reg_page_list()
and ib_alloc_fast_reg_mr() multiply the SG list length by PAGE_SIZE /
SIZE_4K to compensate for page size differences on architectures where
virtual memory pages are larger than 4KB ?
My understanding of the API is that it assumes that the HW is capable to
change the page shift (size) associated with certain fast_reg MR for
each mapping.
I think you have a point here. Although I'm not sure this is what you
had in mind, using PAGE_SIZE / PAGE_SHIFT instead of SIZE_4K / SHIFT_4K
should work better on architectures with a page size larger than 4KB. On
mlx4 the largest supported page list length is 511
(MLX4_MAX_FAST_REG_PAGES). So the former option should allow to support
larger sg lists on such architectures (up to length 511) compared to the
latter (up to length 511/(64/4) = 31 for 64KB pages).
Bart.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html