On Fri, Feb 03, 2023 at 01:49:07PM -0500, David Thompson wrote: > This commit adds ACPI handling for host-gpio[7] to trigger > the power-button event. This has something to discuss... ... > /* Pad Electrical Controls. */ > -#define MLXBF_GPIO_PAD_CONTROL_FIRST_WORD 0x0700 > -#define MLXBF_GPIO_PAD_CONTROL_1_FIRST_WORD 0x0708 > -#define MLXBF_GPIO_PAD_CONTROL_2_FIRST_WORD 0x0710 > -#define MLXBF_GPIO_PAD_CONTROL_3_FIRST_WORD 0x0718 > +#define MLXBF_GPIO_PAD_CONTROL_FIRST_WORD 0x0700 > +#define MLXBF_GPIO_PAD_CONTROL_1_FIRST_WORD 0x0708 > +#define MLXBF_GPIO_PAD_CONTROL_2_FIRST_WORD 0x0710 > +#define MLXBF_GPIO_PAD_CONTROL_3_FIRST_WORD 0x0718 Unrelated change. > +#define MLXBF_GPIO_PIN_DIR_I 0x1040 > +#define MLXBF_GPIO_PIN_DIR_O 0x1048 > +#define MLXBF_GPIO_PIN_STATE 0x1000 > +#define MLXBF_GPIO_SCRATCHPAD 0x20 Ditto. You already have it defined, why to change? > -#define MLXBF_GPIO_PIN_DIR_I 0x1040 > -#define MLXBF_GPIO_PIN_DIR_O 0x1048 > -#define MLXBF_GPIO_PIN_STATE 0x1000 > -#define MLXBF_GPIO_SCRATCHPAD 0x20 ... > + gs->hwirq = irq; > + gc->to_irq = mlxbf_gpio_to_irq; This sounds incorrect. Seems like for _any_ interrupt you will give the same result. Moreover, you should not use to_irq(). Try to model proper IRQ chip. ... > platform_set_drvdata(pdev, gs); > + acpi_gpiochip_request_interrupts(gc); This is done by GPIO library, no? ... > dev_info(&pdev->dev, "registered Mellanox BlueField GPIO"); > + > return 0; Stray change. -- With Best Regards, Andy Shevchenko