[PPC] Boot problems after the pci-v6.18-changes

[Christian Zigotzky]

> On 13. October 2025 at 07:03 am, Christian Zigotzky <chzigotzky at xenosoft.de> wrote:
> 
> 
> 
>> On 13 October 2025 at 06:47 am, Christian Zigotzky <chzigotzky at xenosoft.de> wrote:
>> 
>> 
>>>> On 11 October 2025 at 07:36 pm, Manivannan Sadhasivam <mani at kernel.org> wrote:
>>> 
>>> Hi Lukas,
>>> 
>>> Thanks for looping me in. The referenced commit forcefully enables ASPM on all
>>> DT platforms as we decided to bite the bullet finally.
>>> 
>>> Looks like the device (0000:01:00.0) doesn't play nice with ASPM even though it
>>> advertises ASPM capability.
>>> 
>>> Christian, could you please test the below change and see if it fixes the issue?
>>> 
>>> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
>>> index 214ed060ca1b..e006b0560b39 100644
>>> --- a/drivers/pci/quirks.c
>>> +++ b/drivers/pci/quirks.c
>>> @@ -2525,6 +2525,15 @@ static void quirk_disable_aspm_l0s_l1(struct pci_dev *dev)
>>> */
>>> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ASMEDIA, 0x1080, quirk_disable_aspm_l0s_l1);
>>> 
>>> +
>>> +static void quirk_disable_aspm_all(struct pci_dev *dev)
>>> +{
>>> +       pci_info(dev, "Disabling ASPM\n");
>>> +       pci_disable_link_state(dev, PCIE_LINK_STATE_ALL);
>>> +}
>>> +
>>> +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6738, quirk_disable_aspm_all);
>>> +
>>> /*
>>> * Some Pericom PCIe-to-PCI bridges in reverse mode need the PCIe Retrain
>>> * Link bit cleared after starting the link retrain process to allow this
>>> 
>>> 
>>> Going forward, we should be quirking the devices if they behave erratically.
>>> 
>>> - Mani
>>> 
>>> --
>>> மணிவண்ணன் சதாசிவம்
>> 
>> Hello Mani,
>> 
>>> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6738, quirk_disable_aspm_all);
>> 
>> Is this only for my AMD Radeon HD6870?
>> 
>> My AMD Radeon HD5870 is also affected.
>> 
>> And I tested it with my AMD Radeon HD5870.
>> 
>> What would the line be for all AMD graphics cards?
>> 
>> Thanks,
>> Christian
> 
> I see. 0x6738 is for the AMD Radeon HD 6800 series.
> 
> It could be, that your patch works because I tested it with an AMD Radeon HD5870 instead of an AMD Radeon HD6870. Sorry
> 
> This could be the correct line for the HD5870:
> 
>>> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6898, quirk_disable_aspm_all);
> 
> There are some more id numbers for the HD5870.
> 
> Correct:
> 
> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> index 214ed060ca1b..e006b0560b39 100644
> --- a/drivers/pci/quirks.c
> +++ b/drivers/pci/quirks.c
> @@ -2525,6 +2525,15 @@ static void quirk_disable_aspm_l0s_l1(struct pci_dev *dev)
> */
> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ASMEDIA, 0x1080, quirk_disable_aspm_l0s_l1);
> 
> +
> +static void quirk_disable_aspm_all(struct pci_dev *dev)
> +{
> +       pci_info(dev, "Disabling ASPM\n");
> +       pci_disable_link_state(dev, PCIE_LINK_STATE_ALL);
> +}
> +
> +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6738, quirk_disable_aspm_all);
> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6898, quirk_disable_aspm_all);
> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x6899, quirk_disable_aspm_all);
> DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, 0x689E, quirk_disable_aspm_all);
> +
> /*
> * Some Pericom PCIe-to-PCI bridges in reverse mode need the PCIe Retrain
> * Link bit cleared after starting the link retrain process to allow this

Better for testing (All AMD graphics cards):

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 214ed060ca1b..e006b0560b39 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -2525,6 +2525,15 @@ static void quirk_disable_aspm_l0s_l1(struct pci_dev *dev)
*/
DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ASMEDIA, 0x1080, quirk_disable_aspm_l0s_l1);

+
+static void quirk_disable_aspm_all(struct pci_dev *dev)
+{
+       pci_info(dev, "Disabling ASPM\n");
+       pci_disable_link_state(dev, PCIE_LINK_STATE_ALL);
+}
+
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATI, PCI_ANY_ID, quirk_disable_aspm_all);
+
/*
* Some Pericom PCIe-to-PCI bridges in reverse mode need the PCIe Retrain
* Link bit cleared after starting the link retrain process to allow this