From: Abhijit Gangurde <abhijit.gangurde@amd.com>
To: <jgg@ziepe.ca>, <leon@kernel.org>
Cc: <allen.hubbe@amd.com>, <a.k@amd.com>, <nikhil.agarwal@amd.com>,
<linux-rdma@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
"Abhijit Gangurde" <abhijit.gangurde@amd.com>
Subject: [for-next 1/3] RDMA/ionic: size the MR PTE reservation to max_sg
Date: Tue, 27 Oct 2026 14:53:51 +0530 [thread overview]
Message-ID: <20261027092353.1961552-1-abhijit.gangurde@amd.com> (raw)
From: Amit Kumar <a.k@amd.com>
CREATE_MR with map_count zero makes firmware reserve its default number
of PTEs. A fast-registration memory region (MR) whose max_sg is larger
than that default fails at REG_MR.
Pass max_sg as map_count. The virtual address (VA) and length stay zero
because no buffer is mapped yet.
Fixes: e8521822c733 ("RDMA/ionic: Register device ops for control path")
Signed-off-by: Amit Kumar <a.k@amd.com>
Signed-off-by: Abhijit Gangurde <abhijit.gangurde@amd.com>
---
.../infiniband/hw/ionic/ionic_controlpath.c | 18 +++++++++++-------
1 file changed, 11 insertions(+), 7 deletions(-)
diff --git a/drivers/infiniband/hw/ionic/ionic_controlpath.c b/drivers/infiniband/hw/ionic/ionic_controlpath.c
index 18ad967812f5..00f9086f3bab 100644
--- a/drivers/infiniband/hw/ionic/ionic_controlpath.c
+++ b/drivers/infiniband/hw/ionic/ionic_controlpath.c
@@ -1121,21 +1121,25 @@ struct ib_mr *ionic_alloc_mr(struct ib_pd *ibpd, enum ib_mr_type type,
mr->flags = IONIC_MRF_PHYS_MR;
- rc = ionic_pgtbl_init(dev, &mr->buf, mr->umem, 0, max_sg, PAGE_SIZE);
+ mr->buf.tbl_pages = max_sg;
+ rc = ionic_create_mr_cmd(dev, pd, mr, 0, 0);
+ if (rc)
+ goto err_cmd;
+
+ rc = ionic_pgtbl_init(dev, &mr->buf, NULL, 0, max_sg, 0);
if (rc)
goto err_pgtbl;
+ /* NIC reservation was already posted. tbl_pages means "mapped". */
mr->buf.tbl_pages = 0;
- rc = ionic_create_mr_cmd(dev, pd, mr, 0, 0);
- if (rc)
- goto err_cmd;
-
return &mr->ibmr;
-err_cmd:
- ionic_pgtbl_unbuf(dev, &mr->buf);
err_pgtbl:
+ /* Keep the id if the NIC still holds the MR. */
+ if (ionic_destroy_mr_cmd(dev, mr->mrid))
+ goto err_mrid;
+err_cmd:
ionic_put_mrid(dev, mr->mrid);
err_mrid:
kfree(mr);
--
2.43.0
next reply other threads:[~2026-10-27 9:24 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-27 9:23 Abhijit Gangurde [this message]
2026-10-27 9:23 ` [for-next 2/3] RDMA/ionic: set the SQ WQE color bit Abhijit Gangurde
2026-10-27 9:23 ` [for-next 3/3] RDMA/ionic: key recv completions by producer index Abhijit Gangurde
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=20261027092353.1961552-1-abhijit.gangurde@amd.com \
--to=abhijit.gangurde@amd.com \
--cc=a.k@amd.com \
--cc=allen.hubbe@amd.com \
--cc=jgg@ziepe.ca \
--cc=leon@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=nikhil.agarwal@amd.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®