Skip to contentSkip to conversation

Optimize SGL registration

PATCH Superseded Leon Romanovsky 2 participants 8 messages Oct 7, 2019

Subsystems (5)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

cover Optimize SGL registration

From: Leon Romanovsky <[email protected]>

Changelog v0->v1: https://lore.kernel.org/linux-rdma/[email protected]
* Reorganized patches to have IB/core changes separated from mlx5
* Moved SGL check before rdma_rw_force_mr
* Added and rephrased original comment.

-----------------------------------------------------------------------------
Hi,

This series from Yamin implements long standing "TODO" existed in rw.c.

Thanks

Yamin Friedman (3):
net/mlx5: Expose optimal performance scatter entries capability RDMA/rw: Support threshold for registration vs scattering to local
pages
RDMA/mlx5: Add capability for max sge to get optimized performance

 drivers/infiniband/core/rw.c      | 14 ++++++++------
 drivers/infiniband/hw/mlx5/main.c |  2 ++
 include/linux/mlx5/mlx5_ifc.h     |  2 +-
 include/rdma/ib_verbs.h           |  2 ++
 4 files changed, 13 insertions(+), 7 deletions(-)

--
2.20.1

#1 lore raw

PATCH 1/3 net/mlx5: Expose optimal performance scatter entries capability

From: Yamin Friedman <[email protected]>

Expose maximum scatter entries per RDMA READ for optimal performance.

---
 include/linux/mlx5/mlx5_ifc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/mlx5/mlx5_ifc.h +1 -1
index 138c50d5a353..c0bfb1d90dd2 100644
@@ -1153,7 +1153,7 @@ struct mlx5_ifc_cmd_hca_cap_bits {
 	u8         log_max_srq[0x5];
 	u8         reserved_at_b0[0x10];
 
-	u8         reserved_at_c0[0x8];
+	u8         max_sgl_for_optimized_performance[0x8];
 	u8         log_max_cq_sz[0x8];
 	u8         reserved_at_d0[0xb];
 	u8         log_max_cq[0x5];

--
2.20.1

#2 lore raw

PATCH 2/3 RDMA/rw: Support threshold for registration vs scattering to local pages

In reply to Leon Romanovsky

From: Yamin Friedman <[email protected]>

If there are more scatter entries than the recommended limit provided by the ib device, UMR registration is used. This will provide optimal performance when performing large RDMA READs over devices that advertise the threshold capability.

With ConnectX-5 running NVMeoF RDMA with FIO single QP 128KB writes:
Without use of cap: 70Gb/sec
With use of cap: 84Gb/sec

---
 drivers/infiniband/core/rw.c | 14 ++++++++------
 include/rdma/ib_verbs.h      |  2 ++
 2 files changed, 10 insertions(+), 6 deletions(-)
drivers/infiniband/core/rw.c +8 -6
index 5337393d4dfe..8739bd28232b 100644
@@ -20,14 +20,16 @@ module_param_named(force_mr, rdma_rw_force_mr, bool, 0);
 MODULE_PARM_DESC(force_mr, "Force usage of MRs for RDMA READ/WRITE operations");
 
 /*
- * Check if the device might use memory registration.  This is currently only
- * true for iWarp devices. In the future we can hopefully fine tune this based
- * on HCA driver input.
+ * Check if the device might use memory registration. This is currently
+ * true for iWarp devices and devices that have optimized SGL registration
+ * logic.
  */
 static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num)
 {
 	if (rdma_protocol_iwarp(dev, port_num))
 		return true;
+	if (dev->attrs.max_sgl_rd)
+		return true;
 	if (unlikely(rdma_rw_force_mr))
 		return true;
 	return false;
@@ -37,15 +39,15 @@ static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num)
  * Check if the device will use memory registration for this RW operation.
  * We currently always use memory registrations for iWarp RDMA READs, and
  * have a debug option to force usage of MRs.
- *
- * XXX: In the future we can hopefully fine tune this based on HCA driver
- * input.
  */
 static inline bool rdma_rw_io_needs_mr(struct ib_device *dev, u8 port_num,
 		enum dma_data_direction dir, int dma_nents)
 {
 	if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE)
 		return true;
+	if (dev->attrs.max_sgl_rd && dir == DMA_FROM_DEVICE &&
+	    dma_nents > dev->attrs.max_sgl_rd)
+		return true;
 	if (unlikely(rdma_rw_force_mr))
 		return true;
 	return false;
include/rdma/ib_verbs.h +2 -0
index 4f671378dbfc..60fd98a9b7e8 100644
@@ -445,6 +445,8 @@ struct ib_device_attr {
 	struct ib_tm_caps	tm_caps;
 	struct ib_cq_caps       cq_caps;
 	u64			max_dm_size;
+	/* Max entries for sgl for optimized performance per READ */
+	u32			max_sgl_rd;
 };
 
 enum ib_mtu {

--
2.20.1

#3 lore raw

Re: [PATCH rdma-next v1 2/3] RDMA/rw: Support threshold for registration vs scattering to local pages

Sorry for nitpicking again, but..

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 42

Quoted 7 lines · Leon Romanovsky
drivers/infiniband/core/rw.c +0 -3
@@ -37,15 +39,15 @@ static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num)
  * Check if the device will use memory registration for this RW operation.
  * We currently always use memory registrations for iWarp RDMA READs, and
  * have a debug option to force usage of MRs.
- *
- * XXX: In the future we can hopefully fine tune this based on HCA driver
- * input.

The above comment needs an updated a la:

* Check if the device will use memory registration for this RW operation.
* For RDMA READs we must use MRs on iWarp and can optionaly use them as an
* optimaztion otherwise. Additionally we have a debug option to force usage
* of MRs to help testing this code path.

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 50

Quoted 5 lines
drivers/infiniband/core/rw.c +3 -0
 	if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE)
 		return true;
+	if (dev->attrs.max_sgl_rd && dir == DMA_FROM_DEVICE &&
+	    dma_nents > dev->attrs.max_sgl_rd)
+		return true;

This can be simplified to:

	if (dir == DMA_FROM_DEVICE &&
	    (rdma_protocol_iwarp(dev, port_num) ||
	     (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)))
		return true;
#4 lore raw

Re: [PATCH rdma-next v1 2/3] RDMA/rw: Support threshold for registration vs scattering to local pages

Quoted 27 lines · Christoph Hellwig

Sorry for nitpicking again, but..

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 42

Quoted 7 lines · Leon Romanovsky
drivers/infiniband/core/rw.c +0 -3
@@ -37,15 +39,15 @@ static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num)
  * Check if the device will use memory registration for this RW operation.
  * We currently always use memory registrations for iWarp RDMA READs, and
  * have a debug option to force usage of MRs.
- *
- * XXX: In the future we can hopefully fine tune this based on HCA driver
- * input.

The above comment needs an updated a la:

* Check if the device will use memory registration for this RW operation.
* For RDMA READs we must use MRs on iWarp and can optionaly use them as an
* optimaztion otherwise. Additionally we have a debug option to force usage
* of MRs to help testing this code path.

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 50

Quoted 5 lines
drivers/infiniband/core/rw.c +3 -0
 	if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE)
 		return true;
+	if (dev->attrs.max_sgl_rd && dir == DMA_FROM_DEVICE &&
+	    dma_nents > dev->attrs.max_sgl_rd)
+		return true;

This can be simplified to:

Quoted 4 lines
	if (dir == DMA_FROM_DEVICE &&
	    (rdma_protocol_iwarp(dev, port_num) ||
	     (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)))
		return true;

I don't think that it simplifies and wanted to make separate checks to be separated. For example, rdma_protocol_iwarp() has nothing to do with attrs.max_sgl_rd.

I'll fix comment.

Thanks

#5 lore raw

Re: [PATCH rdma-next v1 2/3] RDMA/rw: Support threshold for registration vs scattering to local pages

Quoted 8 lines · Leon Romanovsky
Quoted 8 lines

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 50

Quoted 5 lines
drivers/infiniband/core/rw.c +3 -0
 	if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE)
 		return true;
+	if (dev->attrs.max_sgl_rd && dir == DMA_FROM_DEVICE &&
+	    dma_nents > dev->attrs.max_sgl_rd)
+		return true;

This can be simplified to:

Quoted 8 lines
Quoted 4 lines
	if (dir == DMA_FROM_DEVICE &&
	    (rdma_protocol_iwarp(dev, port_num) ||
	     (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)))
		return true;

I don't think that it simplifies and wanted to make separate checks to be separated. For example, rdma_protocol_iwarp() has nothing to do with attrs.max_sgl_rd.

The important bit is to have the DMA_FROM_DEVICE check only once, as we only do the registration for reads with either parameter. So if you want it more verbose the wya would be:

 	if (dir == DMA_FROM_DEVICE) {
		if (rdma_protocol_iwarp(dev, port_num))
			return true;
		if (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)
			return true;
	}
#6 lore raw

Re: [PATCH rdma-next v1 2/3] RDMA/rw: Support threshold for registration vs scattering to local pages

Quoted 22 lines · Christoph Hellwig
Quoted 8 lines · Leon Romanovsky
Quoted 8 lines

drivers/infiniband/core/rw.c static inline bool rdma_rw_can_use_mr(struct ib_device *dev, u8 port_num) · line 50

Quoted 5 lines
drivers/infiniband/core/rw.c +3 -0
 	if (rdma_protocol_iwarp(dev, port_num) && dir == DMA_FROM_DEVICE)
 		return true;
+	if (dev->attrs.max_sgl_rd && dir == DMA_FROM_DEVICE &&
+	    dma_nents > dev->attrs.max_sgl_rd)
+		return true;

This can be simplified to:

Quoted 8 lines
Quoted 4 lines
	if (dir == DMA_FROM_DEVICE &&
	    (rdma_protocol_iwarp(dev, port_num) ||
	     (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)))
		return true;

I don't think that it simplifies and wanted to make separate checks to be separated. For example, rdma_protocol_iwarp() has nothing to do with attrs.max_sgl_rd.

The important bit is to have the DMA_FROM_DEVICE check only once, as we only do the registration for reads with either parameter. So if you want it more verbose the wya would be:

Quoted 6 lines
 	if (dir == DMA_FROM_DEVICE) {
		if (rdma_protocol_iwarp(dev, port_num))
			return true;
		if (dev->attrs.max_sgl_rd && dma_nents > dev->attrs.max_sgl_rd)
			return true;
	}

I'm doing it now, Thank you for taking time to explain.

#7 lore raw

PATCH 3/3 RDMA/mlx5: Add capability for max sge to get optimized performance

In reply to Leon Romanovsky

From: Yamin Friedman <[email protected]>

Allows the IB device to provide a value of maximum scatter gather entries per RDMA READ.

In certain cases it may be preferable for a device to perform UMR memory registration rather than have many scatter entries in a single RDMA READ. This provides a significant performance increase in devices capable of using different memory registration schemes based on the number of scatter gather entries. This general capability allows each device vendor to fine tune when it is better to use memory registration.

---
 drivers/infiniband/hw/mlx5/main.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/infiniband/hw/mlx5/main.c +2 -0
index fa23c8e7043b..39d54e285ae9 100644
@@ -1012,6 +1012,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size);
 	props->max_pi_fast_reg_page_list_len =
 		props->max_fast_reg_page_list_len / 2;
+	props->max_sgl_rd =
+		MLX5_CAP_GEN(mdev, max_sgl_for_optimized_performance);
 	get_atomic_caps_qp(dev, props);
 	props->masked_atomic_cap   = IB_ATOMIC_NONE;
 	props->max_mcast_grp	   = 1 << MLX5_CAP_GEN(mdev, log_max_mcg);

--
2.20.1