From: Yamin Friedman <yaminf@xxxxxxxxxxxx>
Added the interface in the infiniband driver that applies the rdma_dim
adaptive moderation. There is now a special function for allocating an
ib_cq that uses rdma_dim.
Performance improvement (ConnectX-5 100GbE, x86) running FIO benchmark over
NVMf between two equal end-hosts with 56 cores across a Mellanox switch
using null_blk device:
READS without DIM:
blk size | BW | IOPS | 99th percentile latency | 99.99th latency
512B | 3.8GiB/s | 7.7M | 1401 usec | 2442 usec
4k | 7.0GiB/s | 1.8M | 4817 usec | 6587 usec
64k | 10.7GiB/s| 175k | 9896 usec | 10028 usec
IO WRITES without DIM:
blk size | BW | IOPS | 99th percentile latency | 99.99th latency
512B | 3.6GiB/s | 7.5M | 1434 usec | 2474 usec
4k | 6.3GiB/s | 1.6M | 938 usec | 1221 usec
64k | 10.7GiB/s| 175k | 8979 usec | 12780 usec
IO READS with DIM:
blk size | BW | IOPS | 99th percentile latency | 99.99th latency
512B | 4GiB/s | 8.2M | 816 usec | 889 usec
4k | 10.1GiB/s| 2.65M| 3359 usec | 5080 usec
64k | 10.7GiB/s| 175k | 9896 usec | 10028 usec
IO WRITES with DIM:
blk size | BW | IOPS | 99th percentile latency | 99.99th latency
512B | 3.9GiB/s | 8.1M | 799 usec | 922 usec
4k | 9.6GiB/s | 2.5M | 717 usec | 1004 usec
64k | 10.7GiB/s| 176k | 8586 usec | 12256 usec
The rdma_dim algorithm was designed to measure the effectiveness of
moderation on the flow in a general way and thus should be appropriate
for all RDMA storage protocols.
rdma_dim is configured to be the default option based on performance
improvement seen after extensive tests.
Signed-off-by: Yamin Friedman <yaminf@xxxxxxxxxxxx>
Reviewed-by: Max Gurtovoy <maxg@xxxxxxxxxxxx>
Reviewed-by: Sagi Grimberg <sagi@xxxxxxxxxxx>
Signed-off-by: Saeed Mahameed <saeedm@xxxxxxxxxxxx>
Signed-off-by: Leon Romanovsky <leonro@xxxxxxxxxxxx>
---
drivers/infiniband/core/cq.c | 45 ++++++++++++++++++++++++++++++++++++
include/rdma/ib_verbs.h | 4 ++++
2 files changed, 49 insertions(+)
diff --git a/drivers/infiniband/core/cq.c b/drivers/infiniband/core/cq.c
index 00d70f166209..ffd6e24109d5 100644
--- a/drivers/infiniband/core/cq.c
+++ b/drivers/infiniband/core/cq.c
@@ -18,6 +18,40 @@
#define IB_POLL_FLAGS \
(IB_CQ_NEXT_COMP | IB_CQ_REPORT_MISSED_EVENTS)
+static void ib_cq_rdma_dim_work(struct work_struct *w)
+{
+ struct dim *dim = container_of(w, struct dim, work);
+ struct ib_cq *cq = dim->priv;
+
+ u16 usec = rdma_dim_prof[dim->profile_ix].usec;
+ u16 comps = rdma_dim_prof[dim->profile_ix].comps;
+
+ dim->state = DIM_START_MEASURE;
+
+ cq->device->ops.modify_cq(cq, comps, usec);
+}
+
+static void rdma_dim_init(struct ib_cq *cq)
+{
+ struct dim *dim;
+
+ if (!cq->device->ops.modify_cq || !cq->device->use_cq_dim ||
+ cq->poll_ctx == IB_POLL_DIRECT)
+ return;
+
+ dim = kzalloc(sizeof(struct dim), GFP_KERNEL);
+ if (!dim)
+ return;
+
+ dim->state = DIM_START_MEASURE;
+ dim->tune_state = DIM_GOING_RIGHT;
+ dim->profile_ix = RDMA_DIM_START_PROFILE;
+ dim->priv = cq;
+ cq->dim = dim;
+
+ INIT_WORK(&dim->work, ib_cq_rdma_dim_work);
+}
+
static int __ib_process_cq(struct ib_cq *cq, int budget, struct ib_wc *wcs,
int batch)
{
@@ -78,6 +112,7 @@ static void ib_cq_completion_direct(struct ib_cq *cq, void *private)
static int ib_poll_handler(struct irq_poll *iop, int budget)
{
struct ib_cq *cq = container_of(iop, struct ib_cq, iop);
+ struct dim *dim = cq->dim;
int completed;
completed = __ib_process_cq(cq, budget, cq->wc, IB_POLL_BATCH);
@@ -87,6 +122,9 @@ static int ib_poll_handler(struct irq_poll *iop, int budget)
irq_poll_sched(&cq->iop);
}
+ if (dim)
+ rdma_dim(dim, completed);
+
return completed;
}
@@ -105,6 +143,8 @@ static void ib_cq_poll_work(struct work_struct *work)
if (completed >= IB_POLL_BUDGET_WORKQUEUE ||
ib_req_notify_cq(cq, IB_POLL_FLAGS) > 0)
queue_work(cq->comp_wq, &cq->work);
+ else if (cq->dim)
+ rdma_dim(cq->dim, completed);
}
static void ib_cq_completion_workqueue(struct ib_cq *cq, void *private)
@@ -161,6 +201,8 @@ struct ib_cq *__ib_alloc_cq_user(struct ib_device *dev, void *private,
rdma_restrack_kadd(&cq->res);
+ rdma_dim_init(cq);
+
switch (cq->poll_ctx) {
case IB_POLL_DIRECT:
cq->comp_handler = ib_cq_completion_direct;
@@ -223,6 +265,9 @@ void ib_free_cq_user(struct ib_cq *cq, struct ib_udata *udata)
rdma_restrack_del(&cq->res);
cq->device->ops.destroy_cq(cq, udata);
+ if (cq->dim)
+ cancel_work_sync(&cq->dim->work);
+ kfree(cq->dim);
kfree(cq->wc);
kfree(cq);
}
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 50806bef9f20..30eb68f36109 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -61,6 +61,7 @@
#include <linux/cgroup_rdma.h>
#include <linux/irqflags.h>
#include <linux/preempt.h>
+#include <linux/dim.h>
#include <uapi/rdma/ib_user_verbs.h>
#include <rdma/rdma_counter.h>
#include <rdma/restrack.h>
@@ -1509,6 +1510,7 @@ struct ib_cq {
struct work_struct work;
};
struct workqueue_struct *comp_wq;
+ struct dim *dim;
/*
* Implementation details of the RDMA core, don't use in drivers:
*/
@@ -2576,6 +2578,8 @@ struct ib_device {
u16 is_switch:1;
/* Indicates kernel verbs support, should not be used in drivers */
u16 kverbs_provider:1;
+ /* CQ adaptive moderation (RDMA DIM) */
+ u16 use_cq_dim:1;
u8 node_type;
u8 phys_port_cnt;
struct ib_device_attr attrs;
--
2.20.1
