Re: mspgcc Red Hat release
David Brown <[email protected]>
| Newsgroups | gmane.comp.hardware.texas-instruments.msp430.gcc.user |
|---|---|
| Message-ID | <[email protected]> |
Malloc is discouraged in almost all embedded programming - it should never be used by other library functions unless it is absolutely obvious and necessary (such as for an implementation of C++ "new"). On 08/07/13 04:55, Eric Decker wrote: > There definitely needs to be a way to not have malloc get used > automatically. > > Malloc will cause all kinds of problems in the tinyos community. We > discourage the use of malloc and that kind of dynamic memory allocation to > minimize wild pointers in very small embedded systems. > > > On Sun, Jul 7, 2013 at 7:49 PM, Mark Rages <[email protected]> wrote: > >> On Sun, Jul 7, 2013 at 7:12 PM, Brendan Conoboy <[email protected]> wrote: >> >>> >>> There is now msp430 support in newlib: >>> >>> http://www.cygwin.com/ml/newlib/2013/msg00362.html >>> >>> Since we did a clean re-implementation of the MSP port we can't comment >>> on the pre-existing mspgcc library solutions: We really don't know. Our >>> plan is to support and enhance standard upstream infrastructure. For >>> libc this means newlib. >>> >>> >> How does newlib's footprint compare with msp430-libc? >> >> Last time I used newlib (on ARM) it seemed to want to do dynamic memory >> allocation for a simple printf() (!). I can't imagine using it on a device >> with 256 bytes RAM... or was I just using newlib wrong? >> >> Regards, >> Mark >> markrages@gmail >> ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev