[U-Boot] [PATCH 05/22] board/linkstation/ide.c: Fix GCC 4.6 build warnings

[Guennadi Liakhovetski]

Hi Wolfgang

On Sat, 5 Nov 2011, Wolfgang Denk wrote:

> Fix:
> ide.c: In function 'ide_preinit':
> ide.c:69:21: warning: array subscript is above array bounds
> [-Warray-bounds]
> ide.c:69:21: warning: array subscript is above array bounds
> [-Warray-bounds]
> ide.c:70:17: warning: array subscript is above array bounds
> [-Warray-bounds]
> 
> Signed-off-by: Wolfgang Denk <wd at denx.de>
> Cc: Guennadi Liakhovetski <g.liakhovetski at gmx.de>
> ---
>  board/linkstation/ide.c |    2 ++
>  1 files changed, 2 insertions(+), 0 deletions(-)
> 
> diff --git a/board/linkstation/ide.c b/board/linkstation/ide.c
> index 568fdf5..14e3fd0 100644
> --- a/board/linkstation/ide.c
> +++ b/board/linkstation/ide.c
> @@ -64,12 +64,14 @@ int ide_preinit (void)
>  		ide_bus_offset[0] = pci_hose_bus_to_phys(&hose,
>  							 ide_bus_offset[0] & 0xfffffffe,
>  							 PCI_REGION_IO);
> +#if CONFIG_SYS_IDE_MAXBUS > 1

I don't have gcc 4.6 installed here unfortunately, but wondering, wouldn't 
this

	if (CONFIG_SYS_IDE_MAXBUS > 1) {
		ide_bus_offset[1]...
	}

also make the compiler happy? Maybe not, because, probably, the above "if" 
would be optimised away after indicies are verifies. But if it does fix 
the problem too - it would be preferable IMHO. Of course, one could also 
use a perversion like

	if (CONFIG_SYS_IDE_MAXBUS > 1) {
		ide_bus_offset[CONFIG_SYS_IDE_MAXBUS > 1]...
	}

But that'd be even uglier, than the preprocessor "#if":)

Thanks
Guennadi

>  		pci_read_config_dword(devbusfn, PCI_BASE_ADDRESS_2,
>  				      (u32 *) &ide_bus_offset[1]);
>  		ide_bus_offset[1] &= 0xfffffffe;
>  		ide_bus_offset[1] = pci_hose_bus_to_phys(&hose,
>  							 ide_bus_offset[1] & 0xfffffffe,
>  							 PCI_REGION_IO);
> +#endif
>  	}
>  
>  	if (pci_find_device (PCI_VENDOR_ID_ITE, PCI_DEVICE_ID_ITE_8212, 0) != -1) {
> -- 
> 1.7.6.4
> 

---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/