Am 21.03.2018 um 22:31 schrieb Erlo Haugen: >> >> Today, I looked into the mcs51 --model-large --stack-auto Dhrystone code >> size situation in the SDCC 3.6.0 - SDCC 3.7.0 timeframe. Code size data >> for Dhrystone can be seen in the attached graph, the black arrows mark >> the releases. >> >> As can be seen, about 63% of the code size increase from 3.6.0 to 3.7.0 >> came from a single commit. >> >> That commit enabled support for a standard-compliant bool for mcs51. >> Previously, the mcs51 used __bit for bool (which meant that arrays of >> bool were not supported, nor pointers to bool not bool in struct/union). >> >> I hope we can partially undo that regression by optimizing the handling >> of bool in the mcs51 backend a bit. >> >> Philipp >> > Or maybe allow for choosing between the standard compliant way and the > __bit way by means if a flag or compiler directive. > I do realize that linking modules with different implementation of bool > is bound to be problematic or even fail, but I'm not sure if that will > be a problem in real life. > > BR > Erlo
That could already be done by the user via a preprocessor define. And we want the optimizations for bool anyway, so those using the standard-compliant bool get better code as well (other backends that had the standard-compliant bool for longer than mcs51 already have some bool-related optimizations and generate better code). Philipp ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Sdcc-user mailing list Sdcc-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sdcc-user
