Re: [linux-usb-devel] 2.6.23-rc2-mm2

[Gabriel C <nix.or.die@googlemail.com>]

David Brownell wrote:
> On Friday 10 August 2007, Gabriel C wrote:
>> Getting that with gcc 4.2.1 :
>>
>> drivers/usb/host/ohci-dbg.c: In function 'show_registers':
>> drivers/usb/host/ohci-dbg.c:620: warning: the address of 'next' will always evaluate as 'true'
>> drivers/usb/host/ohci-dbg.c:639: warning: the address of 'next' will always evaluate as 'true'
> 
> Seems like a pretty annoying warning to have added ... the relevant
> test *expects* to have two constant branches, where the compiler
> optimizes one out of existence.
> 
> Does this patch get rid of it?
> 

Yes it does.


> ---
>  drivers/usb/host/ohci-dbg.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> --- g26.orig/drivers/usb/host/ohci-dbg.c	2007-08-11 13:51:42.000000000 -0700
> +++ g26/drivers/usb/host/ohci-dbg.c	2007-08-11 13:52:07.000000000 -0700
> @@ -74,7 +74,7 @@ urb_print (struct urb * urb, char * str,
>  
>  #define ohci_dbg_sw(ohci, next, size, format, arg...) \
>  	do { \
> -	if (next) { \
> +	if (next != NULL) { \
>  		unsigned s_len; \
>  		s_len = scnprintf (*next, *size, format, ## arg ); \
>  		*size -= s_len; *next += s_len; \
>