I've cc'd the nesc god too.

On Thu, Oct 4, 2012 at 5:27 AM, Flemming Nyboe <[email protected]>wrote:

> Hello,
>
> After upgrading to msp430-gcc 4.6.3 (from cygwinports), I get warnings like
>
> RootC.nc:141:7: warning: pointer targets in passing argument 1 of
> ‘sprintf’ differ in signedness [-Wpointer-sign]
> /usr/lib/gcc/msp430/4.6.3/../../../../msp430/include/stdio.h:51:41:
> note: expected ‘char *’ but argument is of type ‘uint8_t *’
>
> The argument to sprintf() in this case is in fact char*, and looking
> in app.c, I noticed that the reason is NesC 1.3.4 sometimes, but not
> always, represents char as uint8_t.
>

That's interesting.   And seems weird.

David?


>
> I one application, when I declare
> 'char txt[2][128];', it becomes 'char SNC__txt[2][128];' in app.c
> In another, when I declare
> 'char txt[2][128];', it becomes 'uint8_t SNC__txt[2][128];' in app.c
>

that is strange.   I think it should be consistent.

I'd say the proper conversion would be char to char.


>
> This raises two questions
> 1) What's up with the occasional char/uint8_t conversion in NesC
>

David?


> 2) Since I did not get these warnings on my previous dated msp430-gcc
> 3.2.3, were they just disabled?
>

yes.   3.2.3 had lots of stuff disabled.   It worked and was good for its
time.

The newer toolchain is complaining correctly.


>
> Regards
> Flemming Nyboe
>
> _______________________________________________
> Tinyos-help mailing list
> [email protected]
> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>



-- 
Eric B. Decker
Senior (over 50 :-) Researcher
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to