[6.12.y] Please apply one upstream fix

[6.12.y] Please apply one upstream fix

[Artem Dinaburg]

Hi,

Please consider applying the following upstream commit to 6.12.y. It
applies cleanly to 6.12.112 and is already in 6.18.y and 7.2.y.

97110d42680e692020154f9aa89298c038196055
  PCI: endpoint: test: Synchronously cancel command handler work
  pci_epf_test_epc_deinit() and pci_epf_test_link_down() use the
  asynchronous cancel_delayed_work(), so the command handler can still
  run while DMA channels and BARs are released. The fix waits for it.

Build-tested on 6.12.y (CONFIG_PCI_EPF_TEST=m, CONFIG_WERROR=y); no
runtime testing.

Prepared with LLM assistance.

Thanks,
Artem Dinaburg

Re: [6.12.y] Please apply one upstream fix

[Manivannan Sadhasivam]

On Wed, Oct 14, 2026 at 03:08:15PM -0400, Artem Dinaburg wrote:
> Hi,
> 
> Please consider applying the following upstream commit to 6.12.y. It
> applies cleanly to 6.12.112 and is already in 6.18.y and 7.2.y.
> 

Please explain why you want this commit to be backported to stable kernels.

- Mani

-- 
மணிவண்ணன் சதாசிவம்

Re: [6.12.y] Please apply one upstream fix

[Greg KH]

On Thu, Oct 15, 2026 at 11:13:42AM +0530, Manivannan Sadhasivam wrote:
> On Wed, Oct 14, 2026 at 03:08:15PM -0400, Artem Dinaburg wrote:
> > Hi,
> > 
> > Please consider applying the following upstream commit to 6.12.y. It
> > applies cleanly to 6.12.112 and is already in 6.18.y and 7.2.y.
> > 
> 
> Please explain why you want this commit to be backported to stable kernels.

Yeah, this is odd, it's a test function?  Artem, how are you hitting
this and why is it needed in older kernels?  Can't you just update to a
newer one if you have this hardware and want to test the endpoint
hardware?

thanks,

greg k-h

Re: [6.12.y] Please apply one upstream fix

[Artem Dinaburg]

Hi Mani and Greg,

You're right, sorry. We haven't hit this and don't have the hardware.

I was preparing a 6.6.y backport of 934e9d137d93 ("PCI: endpoint:
pci-epf-test: Fix double free that causes kernel to oops").

I had planned to send this commit ahead of it, but upon further inspection
that is unnecessary because 6.12.y already carries the double-free
fix without it.

Please drop this request. I apologize for my mistake.

Thanks,
Artem Dinaburg

On Thu, Oct 15, 2026 at 5:55 AM Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Thu, Oct 15, 2026 at 11:13:42AM +0530, Manivannan Sadhasivam wrote:
> > On Wed, Oct 14, 2026 at 03:08:15PM -0400, Artem Dinaburg wrote:
> > > Hi,
> > >
> > > Please consider applying the following upstream commit to 6.12.y. It
> > > applies cleanly to 6.12.112 and is already in 6.18.y and 7.2.y.
> > >
> >
> > Please explain why you want this commit to be backported to stable kernels.
>
> Yeah, this is odd, it's a test function?  Artem, how are you hitting
> this and why is it needed in older kernels?  Can't you just update to a
> newer one if you have this hardware and want to test the endpoint
> hardware?
>
> thanks,
>
> greg k-h