ACK: [PATCH 3/3] acpi: method: add check for _FPS

[Colin Ian King]

On 06/10/14 09:35, Alex Hung wrote:
> Signed-off-by: Alex Hung <alex.hung at canonical.com>
> ---
>  src/acpi/method/method.c | 90 ++++++++++++++++++++++++++++++++++++++++++++++--
>  1 file changed, 88 insertions(+), 2 deletions(-)
> 
> diff --git a/src/acpi/method/method.c b/src/acpi/method/method.c
> index 3ad9357..65cfa94 100644
> --- a/src/acpi/method/method.c
> +++ b/src/acpi/method/method.c
> @@ -97,7 +97,7 @@
>   * _FDM  9.9.3		N (floppy controller, ignore)
>   * _FIF  11.3.1.1	Y
>   * _FIX  6.2.5		Y
> - * _FPS  11.3.1.2	N
> + * _FPS  11.3.1.2	Y
>   * _FSL  11.3.1.3	Y
>   * _FST  11.3.1.4	Y
>   * _GAI  10.4.5		Y
> @@ -4501,6 +4501,92 @@ static int method_test_FIF(fwts_framework *fw)
>  		"_FIF", NULL, 0, method_test_FIF_return, NULL);
>  }
>  
> +static void method_test_FPS_return(
> +	fwts_framework *fw,
> +	char *name,
> +	ACPI_BUFFER *buf,
> +	ACPI_OBJECT *obj,
> +	void *private)
> +{
> +	uint32_t i;
> +	bool failed = false;
> +
> +	FWTS_UNUSED(private);
> +
> +	if (method_check_type(fw, name, buf, ACPI_TYPE_PACKAGE) != FWTS_OK)
> +		return;
> +
> +	if (obj->Package.Elements[0].Type == ACPI_TYPE_INTEGER) {
> +		if (obj->Package.Elements[0].Integer.Value != 0) {
> +			fwts_failed(fw, LOG_LEVEL_MEDIUM,
> +				"Method_FPSBadRevision",
> +				"%s element 0 is not revision 0.", name);
> +			failed = true;
> +		}
> +	} else {
> +		fwts_failed(fw, LOG_LEVEL_MEDIUM,
> +			"Method_FPSBadReturnType",
> +			"%s element 0 is not an integer.", name);
> +		failed = true;
> +	}
> +
> +	/* Could be one or more sub-packages */
> +	for (i = 1; i < obj->Package.Count; i++) {
> +		ACPI_OBJECT *pkg;
> +		uint32_t j;
> +		bool elements_ok = true;
> +
> +		if (obj->Package.Elements[i].Type != ACPI_TYPE_PACKAGE) {
> +			fwts_failed(fw, LOG_LEVEL_MEDIUM,
> +				"Method_FPSBadReturnType",
> +				"%s element %" PRIu32 " is not a package.",
> +				name, i);
> +			failed = true;
> +			continue;
> +		}
> +
> +		pkg = &obj->Package.Elements[i];
> +		if (pkg->Package.Count != 5) {
> +			fwts_failed(fw, LOG_LEVEL_MEDIUM,
> +				"Method_FPSBadSubPackageElementCount",
> +				"%s sub-package %" PRIu32 " was expected to "
> +				"have 5 elements, got %" PRIu32 " elements instead.",
> +				name, i, pkg->Package.Count);
> +			failed = true;
> +			continue;
> +		}
> +
> +		for (j = 0; j < 5; j++) {
> +			/* TODO - field 0 and 1 can be related to other control method */


I wouldn't worry too much about the TODO, i've not seen any firmware
that implements _FPS

> +			if (pkg->Package.Elements[j].Type != ACPI_TYPE_INTEGER) {
> +				fwts_failed(fw, LOG_LEVEL_MEDIUM,
> +					"Method_FPSBadSubPackageReturnType",
> +					"%s sub-package %" PRIu32
> +					" element %" PRIu32 " is not "
> +					"an integer.",
> +					name, i, j);
> +				elements_ok = false;
> +			}
> +		}
> +
> +		if (!elements_ok) {
> +			failed = true;
> +			continue;
> +		}
> +	}
> +
> +	if (!failed)
> +		method_passed_sane(fw, name, "package");
> +
> +	method_passed_sane(fw, name, "package");
> +}
> +
> +static int method_test_FPS(fwts_framework *fw)
> +{
> +	return method_evaluate_method(fw, METHOD_OPTIONAL,
> +		"_FPS", NULL, 0, method_test_FPS_return, NULL);
> +}
> +
>  static int method_test_FSL(fwts_framework *fw)
>  {
>  	ACPI_OBJECT arg[1];
> @@ -5749,7 +5835,7 @@ static fwts_framework_minor_test method_tests[] = {
>  	/* Section 11.3 Fan Devices */
>  
>  	{ method_test_FIF, "Test _FIF (Fan Information)." },
> -	/* { method_test_FPS, "Test _FPS (Fan Performance States)." }, */
> +	{ method_test_FPS, "Test _FPS (Fan Performance States)." },
>  	{ method_test_FSL, "Test _FSL (Fan Set Level)." },
>  	{ method_test_FST, "Test _FST (Fan Status)." },
>  
> 
Acked-by: Colin Ian King <colin.king at canonical.com>