NB: Please trim the CC list and keep it to concerned maintainers.
On Fri, Apr 12, 2019 at 08:36:48AM +0000, Z.q. Hou wrote:
> From: Hou Zhiqiang <Zhiqiang.Hou@xxxxxxx>
>
> Avoid to issue CFG transactions to link partner when the PCIe
> link is not up.
>
> Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@xxxxxxx>
> ---
> V5:
> - Corrected the subject.
>
> drivers/pci/controller/pcie-mobiveil.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/pci/controller/pcie-mobiveil.c b/drivers/pci/controller/pcie-mobiveil.c
> index 621852078caf..1ee3ea2570c0 100644
> --- a/drivers/pci/controller/pcie-mobiveil.c
> +++ b/drivers/pci/controller/pcie-mobiveil.c
> @@ -283,6 +283,10 @@ static bool mobiveil_pcie_valid_device(struct pci_bus *bus, unsigned int devfn)
> {
> struct mobiveil_pcie *pcie = bus->sysdata;
>
> + /* If there is no link, then there is no device */
> + if (bus->number > pcie->root_bus_nr && !mobiveil_pcie_link_up(pcie))
I think Bjorn mentioned this a million times already, eg:
https://lore.kernel.org/linux-pci/20190411201535.GS256045@xxxxxxxxxx/
this is racy and gives a false sense of robustness. We have code in the
kernel that checks the link but adding more seems silly to me, so I am
inclined to drop this patch.
Lorenzo
> + return false;
> +
> /* Only one device down on each root port */
> if ((bus->number == pcie->root_bus_nr) && (devfn > 0))
> return false;
> --
> 2.17.1
>
