On Nov 20, 2007 2:59 PM, Christophe Braillon <[EMAIL PROTECTED]> wrote: > Are you sure that you have no warning like "non-atomic read or write"? > It would be really surprising if the hardware or compiler had bugs on > that kind of basic operation...
For msp430-gcc, this would alas be less surprising than it should be - the -mdisable-hwmul option (disable use of hardware multiply) is used in the Makefile because msp430-gcc's use of the multiply instruction is broken :-( (and has been for years) David Gay > > Federico Fiorentin a écrit : > > > Bug In Arithmetic Ops > > > > I'm developing a synchronization algorithm for Tmote Sky in TinyOS2 > > with a Poller and 2 Clients. > > The poller sends periodically a Request Packet, and the clients reply it. > > > > In the Reply packet there's: > > the Request Packet receive timestamp (TS) , > > a time value (TV) always inferior the timestamp > > and them difference (TS - TV). > > > > I found out that the subtraction evaluated by the mote isn't always > > correct, sometimes it diverge of 1,2,4,8.... it seems that the mote > > doesn't calculate correctly when it's stressed. > > > > Any suggestion is welcome. > > > > Thanks,Regards. > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Tinyos-help mailing list > > [email protected] > > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > _______________________________________________ > Tinyos-help mailing list > [email protected] > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
