linux-kernel - Re: [PATCH v1 1/2] PCI: dw-rockchip: Add remove callback for resource cleanup

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <ushhqxwggsmji2vfhk2onb7yzme2ajziruvynlkvfbhaxxk3ht@5ov7qn6i3i52>
Date: Fri, 31 Oct 2025 20:38:53 +0530
From: Manivannan Sadhasivam <mani@...nel.org>
To: Nicolas Frattaroli <nicolas.frattaroli@...labora.com>
Cc: Anand Moon <linux.amoon@...il.com>, 
	Lorenzo Pieralisi <lpieralisi@...nel.org>, Krzysztof Wilczyński <kwilczynski@...nel.org>, 
	Rob Herring <robh@...nel.org>, Bjorn Helgaas <bhelgaas@...gle.com>, 
	Heiko Stuebner <heiko@...ech.de>, Niklas Cassel <cassel@...nel.org>, 
	Shawn Lin <shawn.lin@...k-chips.com>, Hans Zhang <18255117159@....com>, 
	"open list:PCI NATIVE HOST BRIDGE AND ENDPOINT DRIVERS" <linux-pci@...r.kernel.org>, 
	"moderated list:ARM/Rockchip SoC support" <linux-arm-kernel@...ts.infradead.org>, 
	"open list:ARM/Rockchip SoC support" <linux-rockchip@...ts.infradead.org>, open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v1 1/2] PCI: dw-rockchip: Add remove callback for
 resource cleanup

On Fri, Oct 31, 2025 at 01:29:25PM +0100, Nicolas Frattaroli wrote:
> On Friday, 31 October 2025 09:36:05 Central European Standard Time Manivannan Sadhasivam wrote:
> > On Mon, Oct 27, 2025 at 08:25:29PM +0530, Anand Moon wrote:
> > > Introduce a .remove() callback to the Rockchip DesignWare PCIe
> > > controller driver to ensure proper resource deinitialization during
> > > device removal. This includes disabling clocks and deinitializing the
> > > PCIe PHY.
> > > 
> > 
> > How can you remove a driver that is only built-in? You are just sending some
> > pointless patches that were not tested and does not make sense at all.
> 
> The better question would be: why does Kconfig make PCIE_ROCKCHIP_DW
> a bool rather than a tristate? I see other PCIE_DW drivers using
> tristate, so this doesn't seem like a technical limitation with the
> IP.
> 

The limitation is due to irqchip maintainers not allowing (or recommending) to
remove a controller driver which implements an irqchip. So if any controller
driver that does so, we make them built-in. Recently, I suggested some
controller drivers to be tristate, but without the remove() callback. This will
allow the controller drivers to be built as a module, but not get removed
during runtime.

The reason why an irqchip controller should not be removed during runtime is
that the concerns around disposing the IRQs consumed by the client drivers.
Current irqchip framework doesn't guarantee that all IRQs would be disposed when
the controller is removed.

But irrespective of the above explanation, my statement was correct that this
patch is pointless.

- Mani

-- 
மணிவண்ணன் சதாசிவம்