UbiquitousOS
/
OpenCloudOS-Kernel
mirror of https://gitee.com/OpenCloudOS/OpenCloudOS-Kernel.git
11
0
Fork 0
Code Issues Projects Releases Wiki Activity

IB/hfi1: Get rid of divide when setting the tx request header 

Div instructions show costly in profiles when
the tx request header is set. Using right shift
instead of a divide operation reduces the cycles
spent in the function that sets the tx request
header as shown in the profile. Use right shift
operation instead.

Profile before change:
43.24%  009
|
|--23.41%-- user_sdma_send_pkts
|          |
|          |--99.90%-- hfi1_user_sdma_process_requestAfter:

Profile after change:
45.75%  009
|
|--14.81%-- user_sdma_send_pkts
|          |
|          |--99.95%-- hfi1_user_sdma_process_request

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
This commit is contained in:
Sebastian Sanchez 2017-05-04 05:14:16 -07:00 committed by Doug Ledford
parent dd1ed10817
commit ade6f8af52
1 changed files with 17 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
drivers/infiniband/hw/hfi1/user_sdma.c
View File

@ -143,7 +143,9 @@ MODULE_PARM_DESC(sdma_comp_size, "Size of User SDMA completion ring. Default: 12
/* KDETH OM multipliers and switch over point */ /* KDETH OM multipliers and switch over point */
#define KDETH_OM_SMALL 4 #define KDETH_OM_SMALL 4
#define KDETH_OM_SMALL_SHIFT 2
#define KDETH_OM_LARGE 64 #define KDETH_OM_LARGE 64
#define KDETH_OM_LARGE_SHIFT 6
#define KDETH_OM_MAX_SIZE (1 << ((KDETH_OM_LARGE / KDETH_OM_SMALL) + 1)) #define KDETH_OM_MAX_SIZE (1 << ((KDETH_OM_LARGE / KDETH_OM_SMALL) + 1))
/* Tx request flag bits */ /* Tx request flag bits */
@ -228,12 +230,6 @@ struct user_sdma_request {
* size of the TID entry. * size of the TID entry.
*/ */
u32 tidoffset; u32 tidoffset;
/*
* KDETH.OM
* Remember this because the header template always sets it
* to 0.
*/
u8 omfactor;
/* /*
* We copy the iovs for this request (based on * We copy the iovs for this request (based on
* info.iovcnt). These are only the data vectors * info.iovcnt). These are only the data vectors
@ -1323,6 +1319,7 @@ static int set_txreq_header(struct user_sdma_request *req,
{ {
struct hfi1_user_sdma_pkt_q *pq = req->pq; struct hfi1_user_sdma_pkt_q *pq = req->pq;
struct hfi1_pkt_header *hdr = &tx->hdr; struct hfi1_pkt_header *hdr = &tx->hdr;
u8 omfactor; /* KDETH.OM */
u16 pbclen; u16 pbclen;
int ret; int ret;
u32 tidval = 0, lrhlen = get_lrh_len(*hdr, pad_len(datalen)); u32 tidval = 0, lrhlen = get_lrh_len(*hdr, pad_len(datalen));
@ -1400,8 +1397,9 @@ static int set_txreq_header(struct user_sdma_request *req,
} }
tidval = req->tids[req->tididx]; tidval = req->tids[req->tididx];
} }
req->omfactor = EXP_TID_GET(tidval, LEN) * PAGE_SIZE >= omfactor = EXP_TID_GET(tidval, LEN) * PAGE_SIZE >=
KDETH_OM_MAX_SIZE ? KDETH_OM_LARGE : KDETH_OM_SMALL; KDETH_OM_MAX_SIZE ? KDETH_OM_LARGE_SHIFT :
KDETH_OM_SMALL_SHIFT;
/* Set KDETH.TIDCtrl based on value for this TID. */ /* Set KDETH.TIDCtrl based on value for this TID. */
KDETH_SET(hdr->kdeth.ver_tid_offset, TIDCTRL, KDETH_SET(hdr->kdeth.ver_tid_offset, TIDCTRL,
EXP_TID_GET(tidval, CTRL)); EXP_TID_GET(tidval, CTRL));
@ -1416,12 +1414,12 @@ static int set_txreq_header(struct user_sdma_request *req,
* transfer. * transfer.
*/ */
SDMA_DBG(req, "TID offset %ubytes %uunits om%u", SDMA_DBG(req, "TID offset %ubytes %uunits om%u",
req->tidoffset, req->tidoffset / req->omfactor, req->tidoffset, req->tidoffset >> omfactor,
req->omfactor != KDETH_OM_SMALL); omfactor != KDETH_OM_SMALL_SHIFT);
KDETH_SET(hdr->kdeth.ver_tid_offset, OFFSET, KDETH_SET(hdr->kdeth.ver_tid_offset, OFFSET,
req->tidoffset / req->omfactor); req->tidoffset >> omfactor);
KDETH_SET(hdr->kdeth.ver_tid_offset, OM, KDETH_SET(hdr->kdeth.ver_tid_offset, OM,
req->omfactor != KDETH_OM_SMALL); omfactor != KDETH_OM_SMALL_SHIFT);
} }
done: done:
trace_hfi1_sdma_user_header(pq->dd, pq->ctxt, pq->subctxt, trace_hfi1_sdma_user_header(pq->dd, pq->ctxt, pq->subctxt,
@ -1433,6 +1431,7 @@ static int set_txreq_header_ahg(struct user_sdma_request *req,
struct user_sdma_txreq *tx, u32 len) struct user_sdma_txreq *tx, u32 len)
{ {
int diff = 0; int diff = 0;
u8 omfactor; /* KDETH.OM */
struct hfi1_user_sdma_pkt_q *pq = req->pq; struct hfi1_user_sdma_pkt_q *pq = req->pq;
struct hfi1_pkt_header *hdr = &req->hdr; struct hfi1_pkt_header *hdr = &req->hdr;
u16 pbclen = le16_to_cpu(hdr->pbc[0]); u16 pbclen = le16_to_cpu(hdr->pbc[0]);
@ -1484,14 +1483,15 @@ static int set_txreq_header_ahg(struct user_sdma_request *req,
} }
tidval = req->tids[req->tididx]; tidval = req->tids[req->tididx];
} }
req->omfactor = ((EXP_TID_GET(tidval, LEN) * omfactor = ((EXP_TID_GET(tidval, LEN) *
PAGE_SIZE) >= PAGE_SIZE) >=
KDETH_OM_MAX_SIZE) ? KDETH_OM_LARGE : KDETH_OM_MAX_SIZE) ? KDETH_OM_LARGE_SHIFT :
KDETH_OM_SMALL; KDETH_OM_SMALL_SHIFT;
/* KDETH.OM and KDETH.OFFSET (TID) */ /* KDETH.OM and KDETH.OFFSET (TID) */
AHG_HEADER_SET(req->ahg, diff, 7, 0, 16, AHG_HEADER_SET(req->ahg, diff, 7, 0, 16,
((!!(req->omfactor - KDETH_OM_SMALL)) << 15 | ((!!(omfactor - KDETH_OM_SMALL_SHIFT)) << 15 |
((req->tidoffset / req->omfactor) & 0x7fff))); ((req->tidoffset >> omfactor)
& 0x7fff)));
/* KDETH.TIDCtrl, KDETH.TID, KDETH.Intr, KDETH.SH */ /* KDETH.TIDCtrl, KDETH.TID, KDETH.Intr, KDETH.SH */
val = cpu_to_le16(((EXP_TID_GET(tidval, CTRL) & 0x3) << 10) | val = cpu_to_le16(((EXP_TID_GET(tidval, CTRL) & 0x3) << 10) |
(EXP_TID_GET(tidval, IDX) & 0x3ff)); (EXP_TID_GET(tidval, IDX) & 0x3ff));