Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Wu Zongyong <wuzongyong@linux.alibaba.com>
To: Bjorn Helgaas <helgaas@kernel.org>
Cc: Wu Zongyong <wuzongyong@linux.alibaba.com>,
	lukas@wunner.de, sdonthineni@nvidia.com, bhelgaas@google.com,
	linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org,
	wllenyj@linux.alibaba.com, wutu.xq2@linux.alibaba.com,
	gerry@linux.alibaba.com, pjaroszynski@nvidia.com
Subject: Re: [PATCH] PCI: Mark NVIDIA T4 GPUs to avoid bus reset
Date: Fri, 8 Sep 2023 10:50:48 +0800	[thread overview]
Message-ID: <ZPqMCDWvITlOLHgJ@wuzongyong-alibaba> (raw)
In-Reply-To: <20230809230518.GA12130@bhelgaas>

On Wed, Aug 09, 2023 at 06:05:18PM -0500, Bjorn Helgaas wrote:
> On Mon, Apr 10, 2023 at 08:34:11PM +0800, Wu Zongyong wrote:
> > NVIDIA T4 GPUs do not work with SBR. This problem is found when the T4
> > card is direct attached to a Root Port only. So avoid bus reset by
> > marking T4 GPUs PCI_DEV_FLAGS_NO_BUS_RESET.
> > 
> > Fixes: 4c207e7121fa ("PCI: Mark some NVIDIA GPUs to avoid bus reset")
> > Signed-off-by: Wu Zongyong <wuzongyong@linux.alibaba.com>
> 
> Applied to pci/virtualization for v6.6, thanks!

I talk about the issue with NVIDIA, and they think the issue is probably related
the pci link instead of the T4 GPU card.

I will try to describe the issue I met in detail.

The T4 card which is direct attached to a Root Port and I rebind it to
vfio-pci driver. Then I try to use to call some vfio-related api and the
ioctl VFIO_GROUP_GET_DEVICE_FD failed.

The stack is (base on kernel v5.10):
    vfio_group_fops_unl_ioctl
         vfio_group_get_device_fd
            vfio_pci_open
                vfio_pci_enable // return value is -19
                    pci_try_reset_function
                        __pci_reset_function_locked

After the __pci_reset_function_locked(), the dmesg shows:
   [12207494.508467] pcieport 0000:3f:00.0: pciehp: Slot(5-1): Link Down
   [12207494.508535] vfio-pci 0000:40:00.0: No device request channel registered, blocked until released by user
   [12207494.518426] pci 0000:40:00.0: Removing from iommu group 84
   [12207495.532365] pcieport 0000:3f:00.0: pciehp: Slot(5-1): Card present
   [12207495.532367] pcieport 0000:3f:00.0: pciehp: Slot(5-1): Link Up

NVIDIA people thinks this root port is not going through this reset logic and getting the
link down/hot plug interrupts[1].

Can you revert the patch I sent and maybe we should dig it deeply.

> 
> > ---
> >  drivers/pci/quirks.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> > index 44cab813bf95..be86b93b9e38 100644
> > --- a/drivers/pci/quirks.c
> > +++ b/drivers/pci/quirks.c
> > @@ -3618,7 +3618,7 @@ static void quirk_no_bus_reset(struct pci_dev *dev)
> >   */
> >  static void quirk_nvidia_no_bus_reset(struct pci_dev *dev)
> >  {
> > -	if ((dev->device & 0xffc0) == 0x2340)
> > +	if ((dev->device & 0xffc0) == 0x2340 || dev->device == 0x1eb8)
> >  		quirk_no_bus_reset(dev);
> >  }
> >  DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID,
> > -- 
> > 2.34.3
> > 

  reply	other threads:[~2023-09-08  2:50 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-03-29 11:58 [RFC PATCH] PCI: avoid SBR for NVIDIA T4 Wu Zongyong
2023-03-29 17:05 ` Bjorn Helgaas
2023-03-30  2:10   ` Wu Zongyong
2023-03-30 15:49     ` Bjorn Helgaas
2023-03-31  2:11       ` Wu Zongyong
2023-04-03  4:02         ` Wu Zongyong
2023-03-30  2:41 ` Lukas Wunner
2023-04-10 12:34 ` [PATCH] PCI: Mark NVIDIA T4 GPUs to avoid bus reset Wu Zongyong
2023-04-26  8:13   ` Wu Zongyong
2023-08-09 23:05   ` Bjorn Helgaas
2023-09-08  2:50     ` Wu Zongyong [this message]
2023-09-08  3:40       ` Alex Williamson
2023-09-08 20:11         ` Bjorn Helgaas

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=ZPqMCDWvITlOLHgJ@wuzongyong-alibaba \
    --to=wuzongyong@linux.alibaba.com \
    --cc=bhelgaas@google.com \
    --cc=gerry@linux.alibaba.com \
    --cc=helgaas@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=lukas@wunner.de \
    --cc=pjaroszynski@nvidia.com \
    --cc=sdonthineni@nvidia.com \
    --cc=wllenyj@linux.alibaba.com \
    --cc=wutu.xq2@linux.alibaba.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