From: Yishai Hadas <yishaih@nvidia.com>
To: <bmerry@gmail.com>, <netdev@vger.kernel.org>
Subject: Re: Help with PCIe THP on ConnectX-7
Date: Thu, 23 Apr 2026 17:55:53 +0300 [thread overview]
Message-ID: <58a85178-c00f-4705-aa20-61e668bb9216@nvidia.com> (raw)
In-Reply-To: <aejnzWR_szvAOxkK@brucemerry.org.za>
On 22/04/2026 18:22, Bruce Merry wrote:
> Hello
>
> I'm hoping someone from NVIDIA Networking can help with this; I wasn't
> sure of the best way to get in contact with the engineers working on
> mlx5 kernel code so thought I'd try here. I'm trying to write some
> (userspace) code using ibv_reg_mr_ex to register a memory region using
> TLP Processing Hints (THP) to improve performance on an Epyc Turin
> (Zen 5) system. However, in the kernel mlx5_st_create is bailing out
> because of this check:
>
> if (!MLX5_CAP_GEN(dev, mkey_pcie_tph))
> return NULL;
>
> As far as I can tell, that is checking a capability bit returned by
> the firmware.
Right
>
> Before I spend a lot more time debugging, can you tell me whether
> ConnectX-7 supports this feature at all (I'm on the latest 28.48.1000
> firmware)? I see it mentioned in release notes for ConnectX-8
> firmware, but not in ConnectX-7 release notes.
>
It's supported only on CX8 for now.
Yishai
> If ConnectX-7 does support it, do you have any tips on what might
> cause that capability bit to be false and how to determine what the
> cause is? For example, could that happen if the motherboard BIOS
> doesn't support TPH?
>
> Thanks
> Bruce
next prev parent reply other threads:[~2026-04-23 14:56 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-04-22 15:22 Help with PCIe THP on ConnectX-7 Bruce Merry
2026-04-23 14:55 ` Yishai Hadas [this message]
2026-04-23 19:17 ` Bruce Merry
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=58a85178-c00f-4705-aa20-61e668bb9216@nvidia.com \
--to=yishaih@nvidia.com \
--cc=bmerry@gmail.com \
--cc=netdev@vger.kernel.org \
/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