Re: [PATCH] ACPI: resource: Do IRQ override on MECHREV GM7RG7M

[Rafael J. Wysocki (Intel)]

On Sat, Oct 31, 2026 at 3:41 PM Andrei Galiant <zakat1984@xxxxxxxxx> wrote:
>
> The MECHREV Jiaolong17 Series GM7RG7M (TongFang GM7RG7M, Ryzen 7
> 6800H) describes the PS/2 keyboard interrupt in its DSDT as
> edge-triggered, active-low and shared:
>
>     IRQ (Edge, ActiveLow, Shared) {1}
>
> The keyboard controller actually asserts IRQ 1 active-high.  On AMD
> Zen systems acpi_dev_irq_override() skips the ACPI IRQ override for
> GSI 1, so the IO-APIC redirection entry is programmed with the DSDT's
> active-low polarity and the interrupt is never delivered: the embedded
> controller still scans the matrix and places scancodes in port 0x60,
> but atkbd_receive_byte() is never called and the internal keyboard is
> dead under Linux, while it works in the firmware setup, in GRUB and in
> Windows.
>
> Add the board to irq1_edge_low_force_override[] so that the standard
> edge/high override is applied, matching the already-listed sibling
> GM7XG0M (Jiaolong17KS).
>
> Fixes: 9946e39fe8d0 ("ACPI: resource: skip IRQ override on AMD Zen platforms")
> Cc: stable@xxxxxxxxxxxxxxx
> Tested-by: Andrei Galiant <zakat1984@xxxxxxxxx>
> Signed-off-by: Andrei Galiant <zakat1984@xxxxxxxxx>
> ---
>  drivers/acpi/resource.c | 6 ++++++
>  1 file changed, 6 insertions(+)
>
> diff --git a/drivers/acpi/resource.c b/drivers/acpi/resource.c
> index 56df459..0a944ba 100644
> --- a/drivers/acpi/resource.c
> +++ b/drivers/acpi/resource.c
> @@ -553,6 +553,12 @@ static const struct dmi_system_id irq1_edge_low_force_override[] = {
>                         DMI_MATCH(DMI_BOARD_NAME, "GM7XG0M"),
>                 },
>         },
> +       {
> +               /* MECHREVO Jiaolong17 Series GM7RG7M */
> +               .matches = {
> +                       DMI_MATCH(DMI_BOARD_NAME, "GM7RG7M"),
> +               },
> +       },
>         {
>                 /* XMG APEX 17 (M23) */
>                 .matches = {
> --

Applied as 7.5 material, thanks!