> -----Original Message-----
> From: linux-rdma-owner@xxxxxxxxxxxxxxx <linux-rdma-
> owner@xxxxxxxxxxxxxxx> On Behalf Of Shiraz Saleem
> Sent: Saturday, January 26, 2019 10:59 AM
> To: dledford@xxxxxxxxxx; jgg@xxxxxxxx; linux-rdma@xxxxxxxxxxxxxxx
> Cc: Shiraz, Saleem <shiraz.saleem@xxxxxxxxx>; Steve Wise
> <swise@xxxxxxxxxxx>
> Subject: [PATCH RFC 05/12] RDMA/cxgb4: Use for_each_sg_dma_page
> iterator on umem SGL
>
> From: "Shiraz, Saleem" <shiraz.saleem@xxxxxxxxx>
>
> Use the for_each_sg_dma_page iterator variant to walk the umem
> DMA-mapped SGL and get the page DMA address. This avoids the extra
> loop to iterate pages in the SGE when for_each_sg iterator is used.
>
> Additionally, purge umem->page_shift usage in the driver
> as its only relevant for ODP MRs. Use system page size and
> shift instead.
Hey Shiraz, Doesn't umem->page_shift allow registering huge pages
efficiently? IE is umem->page_shift set for the 2MB shift if the memory in
this umem region is from the 2MB huge page pool?
>
> Cc: Steve Wise <swise@xxxxxxxxxxx>
> Signed-off-by: Shiraz, Saleem <shiraz.saleem@xxxxxxxxx>
> ---
> drivers/infiniband/hw/cxgb4/mem.c | 33 ++++++++++++++-------------------
> 1 file changed, 14 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/infiniband/hw/cxgb4/mem.c
> b/drivers/infiniband/hw/cxgb4/mem.c
> index 96760a3..a9cd6f1 100644
> --- a/drivers/infiniband/hw/cxgb4/mem.c
> +++ b/drivers/infiniband/hw/cxgb4/mem.c
> @@ -502,10 +502,9 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd,
> u64 start, u64 length,
> u64 virt, int acc, struct ib_udata *udata)
> {
> __be64 *pages;
> - int shift, n, len;
> - int i, k, entry;
> + int shift, n, i;
> int err = -ENOMEM;
> - struct scatterlist *sg;
> + struct sg_dma_page_iter sg_iter;
> struct c4iw_dev *rhp;
> struct c4iw_pd *php;
> struct c4iw_mr *mhp;
> @@ -541,7 +540,7 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd,
> u64 start, u64 length,
> if (IS_ERR(mhp->umem))
> goto err_free_skb;
>
> - shift = mhp->umem->page_shift;
> + shift = PAGE_SHIFT;
>
> n = mhp->umem->nmap;
> err = alloc_pbl(mhp, n);
> @@ -556,21 +555,17 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd,
> u64 start, u64 length,
>
> i = n = 0;
>
> - for_each_sg(mhp->umem->sg_head.sgl, sg, mhp->umem->nmap,
> entry) {
> - len = sg_dma_len(sg) >> shift;
> - for (k = 0; k < len; ++k) {
> - pages[i++] = cpu_to_be64(sg_dma_address(sg) +
> - (k << shift));
> - if (i == PAGE_SIZE / sizeof *pages) {
> - err = write_pbl(&mhp->rhp->rdev,
> - pages,
> - mhp->attr.pbl_addr + (n << 3), i,
> - mhp->wr_waitp);
> - if (err)
> - goto pbl_done;
> - n += i;
> - i = 0;
> - }
> + for_each_sg_dma_page(mhp->umem->sg_head.sgl, &sg_iter, mhp-
> >umem->nmap, 0) {
> + pages[i++] =
> cpu_to_be64(sg_page_iter_dma_address(&sg_iter));
> + if (i == PAGE_SIZE / sizeof *pages) {
> + err = write_pbl(&mhp->rhp->rdev,
> + pages,
> + mhp->attr.pbl_addr + (n << 3), i,
> + mhp->wr_waitp);
> + if (err)
> + goto pbl_done;
> + n += i;
> + i = 0;
> }
> }
>
> --
> 1.8.3.1
