6502/millfork
1
0
Fork 0
mirror of https://github.com/KarolS/millfork.git synced 2024-11-05 15:06:24 +00:00
Code Issues Projects Releases Wiki Activity
acaaa5bdfe
millfork/CHANGELOG.md
Karol Stasiak acaaa5bdfe Implicit ranges in for loops are dumb, remove
2018-12-22 01:30:54 +01:00

4.5 KiB
Raw Blame History

Change log

Current version

  • Almost complete support for the Zilog Z80, Intel 8080 and Sharp LR35902 microprocessors.

  • A very incomplete support for NEC PC-88, ZX Spectrum, CP/M and Amstrad CPC.

  • Unified the syntax of commandline switches.

  • Added aliases.

  • Added enumeration types.

  • Added preprocessor.

  • Added for loops over enum types and in-place lists

  • Added align keyword for choosing data and code alignment.

  • Added original line numbers in assembler output.

  • Added sizeof operator.

  • Added preliminary support for volatile keyword.

  • Added multiplication of a 16-bit number by an unsigned 8-bit number.

  • Added more warnings.

  • Automatic selection of text encoding based on target platform.

  • Text literals can be now used as expressions of type pointer.

  • Extra z at the name of the encoding means that the string is zero-terminated.

  • Potentially breaking change! No longer allowed to define things with names that are keywords or builtins.

  • Potentially breaking change! Curly braces in text literals are now used for escape sequences.

  • Potentially breaking change! Changed the c64_basic module.

  • Potentially breaking change! scr now refers to the default screencodes as defined for the platform. Code that uses both a custom platform definition and the scr encoding needs attention (either change scr to petscr or add screen_encoding=petscr in the platform definition file).

  • Potentially breaking change! Platform definitions now need appropriate feature definitions. Code that uses a custom platform definitions will cause extra warnings until fixed.

  • Software variable stack for 6502.

  • Other optimizations and fixes for stack variables.

  • Various code deduplication optimizations.

  • Fixed emitting constant decimal expressions.

  • Fixed decimal subtraction.

  • Fixed signed comparison.

  • Fixed for...downto loops.

  • Fixed invalid optimization of loads before BEQ and similar

  • Fixed miscompiled nonet calls.

  • Parser performance improvement.

  • Standard libraries improvements.

  • Other improvements.

0.3.0

  • Finally faster than C.

  • Licenced the standard library more permissively.

  • Preliminary Atari 2600, BBC Micro and LUnix support.

  • Added array initialization syntax with for.

  • Added multiple new text codecs.

  • Added character literals.

  • Added 24-bit farword type.

  • Special array layouts, e.g. @word.

  • Fixed invalid offsets for branching instructions.

  • Fixed incorrectly overlapping local variables.

  • Fixed broken downto loops.

  • Fixed broken comparisons between variables of different sizes.

  • Fixed several other bugs.

  • Tons of optimizer improvements.

  • Other improvements.

0.2.2

  • Allowed adding constant words to variable bytes without the zeropage pseudoregister.

  • -fzp-register is now enabled by default, as the documentation has already been saying.

  • Allowed more kinds of constants within variable and array initializers.

  • Fixed several bugs.

  • Other improvements.

0.2

  • Breaking change! Renamed inline to macro.

  • Breaking change! Added support for memory segments. Changed the platform definition file syntax.

  • Added preliminary support for 65CE02, HuC6280 and 65816 processors.

  • Added support for Famicom/NES and C64 with SuperCPU.

  • Added new -O1 optimization preset; old -O1 became -O2, old -O2 became -O3 and so on.

  • Added command line options for controlling the size-speed trade-offs.

  • Added support for parameters for macros written in Millfork.

  • Enabled calling macros with index expression parameters.

  • Enabled calling macros from assembly.

  • Added optimizer hints: inline, noinline, register.

  • Added command line flags --size, --fast, --blast-processing.

  • Removed command line flag --detailed-flow. Detailed flow analysis was slow, broken, hard to maintain, and didn't even help that much.

  • Added *'=, nonet, hi and lo operators.

  • Added support for zeropage pseudoregisters, allowing for some operators work with more types of operands.

  • Added return dispatch statements.

  • Added break and continue statements.

  • Allowed the else if combination (i.e. the else branch doesn't have to be in braces if it's just a single if-else statement).

  • Added octal and quaternary literals.

  • Fixed several allocation bugs.

  • Fixed several optimization bugs.

  • Fixed several C64 and C16 library bugs.

  • Fixed several other bugs.

  • Other improvements.

0.1

  • Initial numbered version.