Re: [PATCH] PCI: loongson: Add quirk for OHCI device rev 0x02

[Mingcong Bai <jeffbaichina@gmail.com>]

Hi Huacai,

在 2025/1/21 19:42, Huacai Chen 写道:
> The OHCI controller (rev 0x02) under LS7A PCI host has a hardware flaw.
> MMIO register with offset 0x60/0x64 is treated as legacy PS2-compatible
> keyboard/mouse interface, which confuse the OHCI controller. Since OHCI
> only use a 4KB BAR resource indeed, the LS7A OHCI controller's 32KB BAR
> is wrapped around (the second 4KB BAR space is the same as the first 4KB
> internally). So we can add an 4KB offset (0x1000) to the BAR resource as
> a workaround.
> 
> Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
> ---
>   drivers/pci/controller/pci-loongson.c | 8 ++++++++
>   1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/pci/controller/pci-loongson.c b/drivers/pci/controller/pci-loongson.c
> index bc630ab8a283..977f7b15f3a7 100644
> --- a/drivers/pci/controller/pci-loongson.c
> +++ b/drivers/pci/controller/pci-loongson.c
> @@ -32,6 +32,7 @@
>   #define DEV_LS7A_CONF	0x7a10
>   #define DEV_LS7A_GNET	0x7a13
>   #define DEV_LS7A_EHCI	0x7a14
> +#define DEV_LS7A_OHCI	0x7a24
>   #define DEV_LS7A_DC2	0x7a36
>   #define DEV_LS7A_HDMI	0x7a37
>   
> @@ -144,6 +145,13 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_LOONGSON,
>   DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_LOONGSON,
>   			DEV_LS7A_PCIE_PORT6, loongson_mrrs_quirk);
>   
> +static void loongson_ohci_quirk(struct pci_dev *dev)
> +{
> +	if (dev->revision == 0x2)
> +		dev->resource[0].start += 0x1000;
> +}
> +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_LOONGSON, DEV_LS7A_OHCI, loongson_ohci_quirk);
> +
>   static void loongson_pci_pin_quirk(struct pci_dev *pdev)
>   {
>   	pdev->pin = 1 + (PCI_FUNC(pdev->devfn) & 3);

I have tested this patch and can confirm that it fixes USB full-speed 
and low-speed devices (such as keyboards and mice) on ASUS XC-LS3A6M 
boards. The full-/low-speed devices only work consistently with this 
specific patch applied, or at least one of the two full-/low-speed 
devices would not work.

P.S., With more than one SuperSpeed devices plugged in, the issue 
reproduced more reliably, where at least one of the two full-/low-speed 
devices would not work. With three SuperSpeed devices plugged in, 
filling all five downstream facing ports at the back of the board along 
with the two full-/low-speed devices, the issue reproduced consistently.

Similar issues have also been reported by mini PC (NUC) users, but I do 
not own such a device.

Tested-by: Mingcong Bai <jeffbai@aosc.io>

Best Regards,
Mingcong Bai