From: Leon Romanovsky <leon@kernel.org>
To: Lukas Wunner <lukas@wunner.de>
Cc: Vidya Sagar <vidyas@nvidia.com>,
Bjorn Helgaas <helgaas@kernel.org>,
Dongdong Liu <liudongdong3@huawei.com>,
"hch@infradead.org" <hch@infradead.org>,
"logang@deltatee.com" <logang@deltatee.com>,
"linux-pci@vger.kernel.org" <linux-pci@vger.kernel.org>,
"rajur@chelsio.com" <rajur@chelsio.com>,
Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com>
Subject: Re: [PATCH V11 7/8] PCI: Enable 10-Bit Tag support for PCIe Endpoint device
Date: Sun, 25 Jan 2026 19:03:23 +0200 [thread overview]
Message-ID: <20260125170323.GH13967@unreal> (raw)
In-Reply-To: <aXUT7Xg7V6Dusek8@wunner.de>
On Sat, Jan 24, 2026 at 07:48:13PM +0100, Lukas Wunner wrote:
> On Sat, Jan 24, 2026 at 02:35:12PM +0000, Vidya Sagar wrote:
> > On 05/11/21 23:09, Bjorn Helgaas wrote:
> > > On Fri, Nov 05, 2021 at 04:24:24PM +0800, Dongdong Liu wrote:
> > > > Current we use sysfs to enable/disable 10-bit tags for a requester also
> > > > depend on the RP support 10-bit tag completer, so it will be ok.
> > >
> > > Ah, OK. So we can never *enable* 10-bit tags unless the Root Port
> > > supports them.
> > >
> > > I misunderstood the purpose of this file. When the Root Port doesn't
> > > support 10-bit tags, we won't enable them during enumeration. I
> > > though the point was that if we want to do P2PDMA to a peer that
> > > *does* support them, we could use this file to enable them.
> > >
> > > But my understanding was wrong -- the real purpose of the file is to
> > > *disable* 10-bit tags for the case when a P2PDMA peer doesn't support
> > > them.
> [...]
> > Hi Dongdong Liu,
> > Thanks for pushing this patch series.
> > I would like to know your plan to pursue this series in getting them merged.
> > Given that there is no action on this series since 2021, I would like to
> > take it up if you don't have an imminent plan.
>
> Smita (+cc) took another stab at 10-bit tag support in 2023:
>
> https://lore.kernel.org/all/20230828095429.GA17864@wunner.de/
>
> The approach proposed by Dongdong Liu -- to disable 10-bit tag support
> via sysfs -- doesn't seem right. The kernel has all the information
> it needs to automatically determine when to enable or disable 10-bit tags.
> User space has no business overriding that. There is an in-kernel API
> to set up P2PDMA and it can automatically disable or re-enable 10-bit tags
> when it is appropriate. I think the only reason why a sysfs interface
> would be needed is to support setting up P2PDMA behind the kernel's back
> for proprietary technologies such as Nvidia GPU Direct. I hereby
> preemptively NAK patches that add a sysfs interface for this purpose.
There is no need to be so dramatic. NVIDIA GPUDirect works correctly with the
upstream kernel, without requiring any special hacks.
Thanks
next prev parent reply other threads:[~2026-01-25 17:03 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-30 13:53 [PATCH V11 0/8] PCI: Enable 10-Bit tag support for PCIe devices Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 1/8] PCI: Use cached devcap in more places Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 2/8] PCI: Cache Device Capabilities 2 Register Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 3/8] PCI: Add 10-Bit Tag register definitions Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 4/8] PCI/sysfs: Add a tags sysfs file for PCIe Endpoint devices Dongdong Liu
2021-11-01 20:54 ` Bjorn Helgaas
2021-11-02 13:03 ` Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 5/8] PCI/IOV: Add tags sysfs files for VF devices Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 6/8] PCI/P2PDMA: Add a 10-Bit Tag check in P2PDMA Dongdong Liu
2021-10-30 13:53 ` [PATCH V11 7/8] PCI: Enable 10-Bit Tag support for PCIe Endpoint device Dongdong Liu
2021-11-01 22:02 ` Bjorn Helgaas
2021-11-01 22:33 ` Bjorn Helgaas
2021-11-03 10:05 ` Dongdong Liu
2021-11-03 16:02 ` Bjorn Helgaas
2021-11-05 8:24 ` Dongdong Liu
2021-11-05 17:39 ` Bjorn Helgaas
2026-01-24 14:35 ` Vidya Sagar
2026-01-24 18:48 ` Lukas Wunner
2026-01-25 17:03 ` Leon Romanovsky [this message]
2021-10-30 13:53 ` [PATCH V11 8/8] PCI/IOV: Enable 10-Bit Tag support for PCIe VF devices Dongdong Liu
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=20260125170323.GH13967@unreal \
--to=leon@kernel.org \
--cc=Smita.KoralahalliChannabasappa@amd.com \
--cc=hch@infradead.org \
--cc=helgaas@kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=liudongdong3@huawei.com \
--cc=logang@deltatee.com \
--cc=lukas@wunner.de \
--cc=rajur@chelsio.com \
--cc=vidyas@nvidia.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