Is there anyone who will speak in favor of keeping the current mspgcc behavior of disabling the watchdog during initialization as the default?
Peter On Mon, Mar 14, 2011 at 11:24 AM, Crazy Casta <[email protected]> wrote: > I would like to concur with the suggestion that we leave the wdt alone > (except for pinging) by default. All the ti datasheets talk about the need > to disable the watchdog. It would seem to be more confusing to operate > differently wrt the to docs (harder for those of us who have read and are > familiar with the to docs). > > On Mar 14, 2011 5:12 AM, "JMGross" <[email protected]> wrote: > > > > > > ----- Ursprüngliche Nachricht ----- > Von: Peter Bigot > Gesendet am: 13 Mrz 2011 12:49:47 > > > >> As a minor irritant, mspgcc is the only msp430 toolchain I know that >> immediately disables the ... > Yes, this was the reason why I reqrote the startup code in my projects (to > ensure > proper operation in an ESD-prone environment, the WDT needs to be > active ALL the time, even during startup) > It took me some time to figure out why the WDT was not triggering even if I > didn't > disable it and it should be on after a reset. > It is as annoying as the 'feature' in CCS (TIs own compiler suite) which > does not > initialize uninitialized variable space to 0 (which will speedup the startup > and > therefore usually prevent the WDT from triggering) > > > >> main() will no longer jump to some arbitrary location instead of > returning. >> It will return, an... > Well, especially on the new G devices with only 128 bytes Ram, this > "new way" will waste ~2% of the ram since there needs to be one additional > return address on the stack. On these devices, every byte counts. > On 430X-devices, this will be 4 bytes (but then there is much more ram). > > >> A new option -menable-watchdog will be added which change which libcrt0.a >> file gets linked in. ... > Great. > > Well, since the loop counters in the init code are word registers, there > is only one 3-instruction-loop for each init stage (copy/clear). > And running a complete loop may easily exceed default WDT timeout > on 1x/2x/4x devices (only on 5x/6x the default WDT interval is long enough) > So in my startup code, I ended up with simply triggering the WDT on each > iteration. > With the value preloaded in a register, it only takes 3 cycles/4 bytes > and is not slower than any fancy code with additional > checks/counts/whatever. > >> Comments? > > given :) > > JMGross > > > ------------------------------------------------------------------------------ > Colocation vs. Manag... > Mspgcc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/mspgcc-users > > ------------------------------------------------------------------------------ > Colocation vs. Managed Hosting > A question and answer guide to determining the best fit > for your organization - today and in the future. > http://p.sf.net/sfu/internap-sfd2d > _______________________________________________ > Mspgcc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/mspgcc-users > > ------------------------------------------------------------------------------ Colocation vs. Managed Hosting A question and answer guide to determining the best fit for your organization - today and in the future. http://p.sf.net/sfu/internap-sfd2d _______________________________________________ Mspgcc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mspgcc-users
