On 9/25/24 1:39 PM, Andrew Jones wrote:
> commit 6df664f87c73 ("Revert "hw/riscv/virt.c: imsics DT: add
> '#msi-cells'"") removed #msi-cells. Now that we have a Linux
> commit to reference add a comment explaining why it was removed
> to avoid it getting added back due to DT validation failures.
>
> Signed-off-by: Andrew Jones <ajones@ventanamicro.com>
> ---
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
> hw/riscv/virt.c | 10 ++++++++++
> 1 file changed, 10 insertions(+)
>
> diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c
> index cef41c150aaf..4fd9c0edf038 100644
> --- a/hw/riscv/virt.c
> +++ b/hw/riscv/virt.c
> @@ -552,6 +552,16 @@ static void create_fdt_one_imsic(RISCVVirtState *s, hwaddr base_addr,
> FDT_IMSIC_INT_CELLS);
> qemu_fdt_setprop(ms->fdt, imsic_name, "interrupt-controller", NULL, 0);
> qemu_fdt_setprop(ms->fdt, imsic_name, "msi-controller", NULL, 0);
> + /*
> + * Per the DT binding, we should also add #msi-cells with a value of zero.
> + * But, Linux which does not include commit db8e81132cf0 ("of/irq: Support
> + * #msi-cells=<0> in of_msi_get_domain") does not properly handle #msi-cells
> + * with a value of zero. For this reason, the property has intentionally
> + * been left out. It's harmless to not add it since the absence of the
> + * property and a value of zero for the property mean the same thing.
> + * However, since the DT binding requires the property, DT validation may
> + * fail.
> + */
> qemu_fdt_setprop(ms->fdt, imsic_name, "interrupts-extended",
> imsic_cells, ms->smp.cpus * sizeof(uint32_t) * 2);
> qemu_fdt_setprop(ms->fdt, imsic_name, "reg", imsic_regs,