All of lore.kernel.org
 help / color / mirror / Atom feed
From: Weihang Li <liweihang@huawei.com>
To: <dledford@redhat.com>, <jgg@nvidia.com>
Cc: <leon@kernel.org>, <linux-rdma@vger.kernel.org>,
	<linuxarm@huawei.com>, Yixing Liu <liuyixing1@huawei.com>,
	Weihang Li <liweihang@huawei.com>
Subject: [PATCH for-next] RDMA/hns: Add window selection field of congestion control
Date: Tue, 22 Jun 2021 20:16:03 +0800	[thread overview]
Message-ID: <1624364163-44185-1-git-send-email-liweihang@huawei.com> (raw)

From: Yixing Liu <liuyixing1@huawei.com>

The window selection field is necessary for congestion control of HIP09, it
is got from firmware and then filled into QPC. Some algorithms need it to
decide whether to limit the number of windows.

Fixes: f91696f2f053 ("RDMA/hns: Support congestion control type selection according to the FW")
Signed-off-by: Yixing Liu <liuyixing1@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 12 ++++++++++++
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h |  2 ++
 2 files changed, 14 insertions(+)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 069c7cd..e057ede 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -4585,6 +4585,11 @@ enum {
 	DIP_VALID,
 };
 
+enum {
+	WND_LIMIT,
+	WND_UNLIMIT,
+};
+
 static int check_cong_type(struct ib_qp *ibqp,
 			   struct hns_roce_congestion_algorithm *cong_alg)
 {
@@ -4596,21 +4601,25 @@ static int check_cong_type(struct ib_qp *ibqp,
 		cong_alg->alg_sel = CONG_DCQCN;
 		cong_alg->alg_sub_sel = UNSUPPORT_CONG_LEVEL;
 		cong_alg->dip_vld = DIP_INVALID;
+		cong_alg->wnd_mode_sel = WND_LIMIT;
 		break;
 	case CONG_TYPE_LDCP:
 		cong_alg->alg_sel = CONG_WINDOW;
 		cong_alg->alg_sub_sel = CONG_LDCP;
 		cong_alg->dip_vld = DIP_INVALID;
+		cong_alg->wnd_mode_sel = WND_UNLIMIT;
 		break;
 	case CONG_TYPE_HC3:
 		cong_alg->alg_sel = CONG_WINDOW;
 		cong_alg->alg_sub_sel = CONG_HC3;
 		cong_alg->dip_vld = DIP_INVALID;
+		cong_alg->wnd_mode_sel = WND_LIMIT;
 		break;
 	case CONG_TYPE_DIP:
 		cong_alg->alg_sel = CONG_DCQCN;
 		cong_alg->alg_sub_sel = UNSUPPORT_CONG_LEVEL;
 		cong_alg->dip_vld = DIP_VALID;
+		cong_alg->wnd_mode_sel = WND_LIMIT;
 		break;
 	default:
 		ibdev_err(&hr_dev->ib_dev,
@@ -4651,6 +4660,9 @@ static int fill_cong_field(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 	hr_reg_clear(&qpc_mask->ext, QPCEX_CONG_ALG_SUB_SEL);
 	hr_reg_write(&context->ext, QPCEX_DIP_CTX_IDX_VLD, cong_field.dip_vld);
 	hr_reg_clear(&qpc_mask->ext, QPCEX_DIP_CTX_IDX_VLD);
+	hr_reg_write(&context->ext, QPCEX_SQ_RQ_NOT_FORBID_EN,
+		     cong_field.wnd_mode_sel);
+	hr_reg_clear(&qpc_mask->ext, QPCEX_SQ_RQ_NOT_FORBID_EN);
 
 	/* if dip is disabled, there is no need to set dip idx */
 	if (cong_field.dip_vld == 0)
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
index 04b4ad4..82a2fe4 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
@@ -700,6 +700,7 @@ struct hns_roce_v2_qp_context {
 #define QPCEX_CONG_ALG_SUB_SEL QPCEX_FIELD_LOC(1, 1)
 #define QPCEX_DIP_CTX_IDX_VLD QPCEX_FIELD_LOC(2, 2)
 #define QPCEX_DIP_CTX_IDX QPCEX_FIELD_LOC(22, 3)
+#define QPCEX_SQ_RQ_NOT_FORBID_EN QPCEX_FIELD_LOC(23, 23)
 #define QPCEX_STASH QPCEX_FIELD_LOC(82, 82)
 
 #define	V2_QP_RWE_S 1 /* rdma write enable */
@@ -1337,6 +1338,7 @@ struct hns_roce_congestion_algorithm {
 	u8 alg_sel;
 	u8 alg_sub_sel;
 	u8 dip_vld;
+	u8 wnd_mode_sel;
 };
 
 #define V2_QUERY_PF_CAPS_D_CEQ_DEPTH_S 0
-- 
2.7.4


             reply	other threads:[~2021-06-22 12:16 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-22 12:16 Weihang Li [this message]
2021-06-24 18:17 ` [PATCH for-next] RDMA/hns: Add window selection field of congestion control Jason Gunthorpe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1624364163-44185-1-git-send-email-liweihang@huawei.com \
    --to=liweihang@huawei.com \
    --cc=dledford@redhat.com \
    --cc=jgg@nvidia.com \
    --cc=leon@kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=liuyixing1@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.