ivampiresp/Spdk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

nvmf/rdma: Separate filling wr->sg_list from filling req->iov for DIF case

Browse Source 
This patch separates filling wr->sg_list from filling req->iov
in nvmf_rdma_fill_buffers_with_md_interleave() and create an new helper function
nvmf_rdma_fill_wr_sgl_with_md_interleave() to fill wr->sg_list by adding iovcnt to
struct spdk_nvmf_rdma_request.

The subsequent patches will merge nvmf_rdma_fill_buffers() into
spdk_nvmf_request_get_buffers().

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I03206895e37cf385fb8bd7498f2f4a24797c7ce1
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/469204
Community-CI: Broadcom SPDK FC-NVMe CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This commit is contained in:
Shuhei Matsumoto 2019-09-25 09:42:08 +09:00 committed by Jim Harris
parent 5593b61f93
commit b48a97d454
1 changed files with 16 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

41
lib/nvmf/rdma.c
View File

@ -1562,11 +1562,12 @@ static bool
nvmf_rdma_fill_wr_sge_with_md_interleave(struct spdk_nvmf_rdma_device *device,
struct spdk_nvmf_request *req,
struct ibv_send_wr *wr,
int iovpos,
uint32_t *_remaining_data_block,
uint32_t *_offset,
const struct spdk_dif_ctx *dif_ctx)
{
struct iovec *iov = &req->iov[req->iovcnt];
struct iovec *iov = &req->iov[iovpos];
struct ibv_sge *sg_ele;
uint32_t lkey = 0;
uint32_t sge_len;
@ -1616,39 +1617,33 @@ nvmf_rdma_fill_wr_sge_with_md_interleave(struct spdk_nvmf_rdma_device *device,
* and points to part of buffer
*/
static int
nvmf_rdma_fill_buffers_with_md_interleave(struct spdk_nvmf_rdma_transport *rtransport,
struct spdk_nvmf_rdma_poll_group *rgroup,
nvmf_rdma_fill_wr_sgl_with_md_interleave(struct spdk_nvmf_rdma_poll_group *rgroup,
struct spdk_nvmf_rdma_device *device,
struct spdk_nvmf_request *req,
struct spdk_nvmf_rdma_request *rdma_req,
struct ibv_send_wr *wr,
uint32_t length,
const struct spdk_dif_ctx *dif_ctx)
{
struct spdk_nvmf_request *req = &rdma_req->req;
uint32_t remaining_length = length;
uint32_t remaining_data_block = dif_ctx->block_size - dif_ctx->md_size;
uint32_t offset = 0;
struct iovec *iov;
wr->num_sge = 0;
while (remaining_length && wr->num_sge < SPDK_NVMF_MAX_SGL_ENTRIES) {
iov = &req->iov[req->iovcnt];
iov->iov_base = (void *)((uintptr_t)(req->buffers[req->iovcnt] + NVMF_DATA_BUFFER_MASK)
& ~NVMF_DATA_BUFFER_MASK);
iov->iov_len = spdk_min(remaining_length, rtransport->transport.opts.io_unit_size);
while (spdk_unlikely(!nvmf_rdma_fill_wr_sge_with_md_interleave(device, req, wr,
&remaining_data_block, &offset, dif_ctx))) {
if (nvmf_rdma_replace_buffer(rgroup, &req->buffers[req->iovcnt]) == -ENOMEM) {
rdma_req->iovpos, &remaining_data_block, &offset, dif_ctx))) {
if (nvmf_rdma_replace_buffer(rgroup, &req->buffers[rdma_req->iovpos]) == -ENOMEM) {
return -ENOMEM;
}
req->iov[req->iovcnt].iov_base = (void *)((uintptr_t)(req->buffers[req->iovcnt] +
NVMF_DATA_BUFFER_MASK) &
~NVMF_DATA_BUFFER_MASK);
req->iov[rdma_req->iovpos].iov_base = (void *)((uintptr_t)(req->buffers[rdma_req->iovpos] +
NVMF_DATA_BUFFER_MASK) &
~NVMF_DATA_BUFFER_MASK);
}
remaining_length -= iov->iov_len;
req->iovcnt++;
remaining_length -= req->iov[rdma_req->iovpos].iov_len;
rdma_req->iovpos++;
}
if (remaining_length) {
@ -1746,16 +1741,12 @@ spdk_nvmf_rdma_request_fill_iovs(struct spdk_nvmf_rdma_transport *rtransport,
req->iovcnt = 0;
rdma_req->iovpos = 0;
nvmf_rdma_fill_buffers(rtransport, req, length);
if (spdk_unlikely(rdma_req->dif_insert_or_strip)) {
rc = nvmf_rdma_fill_buffers_with_md_interleave(rtransport,
rgroup,
device,
req,
wr,
length,
&rdma_req->dif_ctx);
rc = nvmf_rdma_fill_wr_sgl_with_md_interleave(rgroup, device, rdma_req,
wr, length, &rdma_req->dif_ctx);
} else {
nvmf_rdma_fill_buffers(rtransport, req, length);
rc = nvmf_rdma_fill_wr_sgl(rgroup, device, rdma_req, wr, length);
}
if (rc != 0) {