[PATCH] acpi: gtdt: Fix the wrong checking of GTDT block

[Alex Hung]

On 2016-03-23 04:45 PM, Vikas C Sajjan wrote:
> when "fwts gtdt" is executed, fwts throws below error even when the
> GTDT table looks fine in the "fwts acpidump".
>
> gtdt: GTDT Generic Timer Description Table test
> 	  -------------------------------
> Test 1 of 1: GTDT Generic Timer Description Table test.
> FAILED [HIGH] GTDTShortBlock: Test 1, GTDT block is too short
> 	  -------------------------------
>
> This patch removes the error checking,
> since (ptr + 20) is always less than end_ptr even for the valid
> GTDT block.
>
> Signed-off-by: Vikas C Sajjan <vikas.cha.sajjan at hpe.com>
> Signed-off-by: Sunil V L <sunil.vl at hpe.com>
> ---
>   src/acpi/gtdt/gtdt.c | 7 -------
>   1 file changed, 7 deletions(-)
>
> diff --git a/src/acpi/gtdt/gtdt.c b/src/acpi/gtdt/gtdt.c
> index 421f17f..05ccea1 100644
> --- a/src/acpi/gtdt/gtdt.c
> +++ b/src/acpi/gtdt/gtdt.c
> @@ -75,13 +75,6 @@ static int gtdt_test1(fwts_framework *fw)
>   		case 0x00:
>   			/* GT Block Structure */
>   			block = (fwts_acpi_table_gtdt_block *)ptr;
> -			if (ptr + 20 < end_ptr) {
> -				passed = false;
> -				fwts_failed(fw, LOG_LEVEL_HIGH,
> -					"GTDTShortBlock",
> -					"GTDT block is too short");
> -				goto done;
> -			}
I think the original intention is to have a greater sign such as "ptr + 
20 > end_ptr", instead of the less sign.

I do think this patch is fine but let's wait for comments from the 
original author, Colin King.

>   			if (block->length < 20) {
>   				passed = false;
>   				fwts_failed(fw, LOG_LEVEL_HIGH,
>


-- 
Cheers,
Alex Hung