[PATCH v6 7/8] PCI: imx6: Disable enabled clocks and regulators after link is down

[Hongxing Zhu]

> -----Original Message-----
> From: Fabio Estevam <festevam at gmail.com>
> Sent: 2022年2月9日 10:02
> To: Hongxing Zhu <hongxing.zhu at nxp.com>
> Cc: Lucas Stach <l.stach at pengutronix.de>; Bjorn Helgaas
> <bhelgaas at google.com>; Mark Brown <broonie at kernel.org>; Lorenzo Pieralisi
> <lorenzo.pieralisi at arm.com>; Jingoo Han <jingoohan1 at gmail.com>;
> linux-pci at vger.kernel.org; dl-linux-imx <linux-imx at nxp.com>; moderated
> list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE
> <linux-arm-kernel at lists.infradead.org>; linux-kernel
> <linux-kernel at vger.kernel.org>; Sascha Hauer <kernel at pengutronix.de>
> Subject: Re: [PATCH v6 7/8] PCI: imx6: Disable enabled clocks and regulators
> after link is down
> 
> Hi Richard,
> 
> On Tue, Feb 8, 2022 at 10:56 PM Hongxing Zhu <hongxing.zhu at nxp.com>
> wrote:
> 
> > Do you mean that we should ignore the return of dw_pcie_host_init()
> > and finish the _probe without an error return?
> 
> Yes, we should not return an error on probe if the link is down.
> 
> Thanks
Hi Fabio:
Regarding my understand, the suspend/resume callbacks would be hooked if the
 probe is complete successfully. There would be extra more than 100ms delay
 in every system resume operations.
I'm afraid that it would bring bad experience for the power management.

Is there a policy that we should obey, and let the probe complete without
an error return?
If yes, I can follow your suggestion, wouldn't return an error on probe when
link is down.

Thanks.
Best Regards
Richard Zhu