Re: [PATCH v9 3/3] PCI: Remove device links to PHY

From: Bjorn Helgaas

Date: Sat Sep 12 2026 - 11:30:20 EST


On Sat, Sep 12, 2026 at 06:23:11PM +0300, Vladimir Oltean wrote:
> On Sat, Sep 12, 2026 at 09:02:07AM -0500, Bjorn Helgaas wrote:
> > On Sat, Sep 12, 2026 at 02:39:44PM +0300, Vladimir Oltean wrote:
> > > Hi Bjorn,
> > >
> > > On Fri, Sep 11, 2026 at 05:12:09PM -0500, Bjorn Helgaas wrote:
> > > > [+cc Felix, author of device link leak fixes]
> > > >
> > > > On Mon, Aug 10, 2026 at 03:43:15PM +0300, Vladimir Oltean wrote:
> > > > > This is practically a full revert of commit
> > > > > 7a4db656a635 ("PCI: dra7xx: Create functional dependency between PCIe and PHY")
> > > > > and a partial revert of the device link pieces from commits
> > > > > dfb80534692d ("PCI: cadence: Add generic PHY support to host and EP drivers")
> > > > > 49229238ab47 ("PCI: keystone: Cleanup PHY handling")
> > > >
> > > > Can we split this into three commits (dra7xx, cadence, keystone)?
> > > > This combined commit is applied on the pci/controller/misc branch, and
> > > > then we have conflicts between that branch and
> > > > pci/controller/dwc-dra7xx.
> > > >
> > > > I think if we split this into three separate commits, we can apply the
> > > > dra7xx one on pci/controller/dwc-dra7xx and drop Felix's device link
> > > > leak fixes below so we don't clutter the git history with conflict
> > > > resolutions or irrelevant commits:
> > > >
> > > > https://git.kernel.org/cgit/linux/kernel/git/pci/pci.git/commit/?id=2b9a036b568c ("PCI: dra7xx: Fix device link leak when devm_phy_get() fails")
> > > >
> > > > https://git.kernel.org/cgit/linux/kernel/git/pci/pci.git/commit/?id=bdfdf654b4b2 ("PCI: dra7xx: Fix device links leak when dra7xx_pcie_enable_phy() fails")
> > >
> > > Isn't this patch already merged in pci/next as
> > > https://git.kernel.org/pub/scm/linux/kernel/git/pci/pci.git/commit/?id=9a9d18547ddc97a73d5d90cff4d5aef90bd8c15d
> > > ?
> >
> > Yes, it is, but until Linus pulls it, the PCI tree is a draft subject
> > to revision. IMO the git history will be much cleaner. The end
> > result will be the same but we won't have conflict resolutions and
> > fixes to code that was removed.
>
> OK, and what is your request from me? Submit 3 patches as a replacement
> to this one against which tree?

Simplest would be three patches based on v7.3-rc1.