On 10/10/2015 10:33 PM, Or Gerlitz wrote:
On Sat, Oct 10, 2015 at 3:27 AM, Sagi Grimberg <sagig@xxxxxxxxxxxx> wrote:
The block layer reliably enforces no SG lists with gaps
to a 4k virtual boundary so we can completely remove the
entire bounce buffering logic.
Hi Sagi,
Hey Or,
(Very) nice cleanup -- so what's the actual deal here? is that as long
as we plant a slave alloc callback into out scsi host template which
further invokes a
blk_queue_virt_boundary(sdev->request_queue, ~MASK_4K) call, we're
100% safe/sure what all SGs we get meet the alignment criteria?
Correct, the nvme driver has the same alignment constraints for its PRPs
and uses the queue virt_boundary to have the block layer enforce the
SG alignment. Recently a patch set that resolves some issues with front
merges, integrity payload and SG_IO was included by Jens (IIRC you were
CC'd on this patch set).
See commits:
4634845 block: Copy a user iovec if it includes gaps
87a816d block: Refuse adding appending a gapped integrity page to a bio
7f39add block: Refuse request/bio merges with gaps in the integrity
5e7c427 block: Check for gaps on front and back merges
So now the block layer reliably enforces the alignment constraints and
we can stop doing bounce buffering in iser.
Since this is such a critical aspect in our design, I think it would
be good to copy on this commit Martin or Jens.
I can copy Martin and Jens although I don't know how much they care
about iser.
Also change the patch
title to have IB/iser: which complies with the way we prefix iser
commits.
I will.
Thanks!
Sagi.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
