Re: [PATCH v14 5/9] clocksource/drivers/arm_arch_timer: Simplify ACPI support code.

[Mark Rutland]

On Thu, Sep 29, 2016 at 02:17:13AM +0800, fu.wei@xxxxxxxxxx wrote:
> From: Fu Wei <fu.wei@xxxxxxxxxx>
> 
> The patch update arm_arch_timer driver to use the function
> provided by the new GTDT driver of ACPI.
> By this way, arm_arch_timer.c can be simplified, and separate
> all the ACPI GTDT knowledge from this timer driver.
> 
> Signed-off-by: Fu Wei <fu.wei@xxxxxxxxxx>
> Signed-off-by: Hanjun Guo <hanjun.guo@xxxxxxxxxx>

This generally looks fine, but:

> +	arch_timer_ppi[PHYS_SECURE_PPI] = acpi_gtdt_map_ppi(PHYS_SECURE_PPI);

As mentioned on the prior patch, I think we shouldn't bother parsing the
secure interrupt, given the problem with the GSIV, and the fact that we
shouldn't need it in Linux.

> +	arch_timer_ppi[PHYS_NONSECURE_PPI] = acpi_gtdt_map_ppi(PHYS_NONSECURE_PPI);
> +	arch_timer_ppi[VIRT_PPI] = acpi_gtdt_map_ppi(VIRT_PPI);
> +	arch_timer_ppi[HYP_PPI] = acpi_gtdt_map_ppi(HYP_PPI);
> +	/* Always-on capability */
> +	arch_timer_c3stop = acpi_gtdt_c3stop();

... I think we should check the flag on the relevant interrupt, though
that's worth clarifying.

>  
> -	/* Always-on capability */
> -	arch_timer_c3stop = !(gtdt->non_secure_el1_flags & ACPI_GTDT_ALWAYS_ON);
> +	if (timer_count < 0)
> +		pr_err("Failed to get platform timer info.\n");

Why don't we log this in the code that would try to initialise the MMIO
timer? We can still fail after this.

Thanks,
Mark.