Après une année fort calme, voici une nouvelle version de maintenance : 1.6.2
  • Fixed : Bad generated code in 16-bit multiply by 4 routine/inline code. Also used for 16-bit SHL 2. Processors <= PIC16 only.
  • Added : New syntax for pseudo-set: when the assigned variable is a SFR, a non-boolean field constant value may now be defined inside the pseudo-set: OSCCON := [IOSCF = 0b110, SCS];
  • Fixed : In case of "unterminated conditional block" error, the source position was not available.
  • Fixed : Some recent versions of MPLAB files syntax changes made some CONFIG bits unavailable for some processors.
  • Fixed : During interrupt context save PRODH/PRODL was not saved as a pair if a multiply is used in an ISR, when only one is used as a result; PIC18 processors.
  • Changed : The latest SynEdit package has been integrated, with Unicode support, but files are still saved in ANSI format. This integration fixes some editor's bugs and opens future features.
  • Fixed : Use of ROMABLE parameter with POINTERS SMALL (or with processors with only one GPR page) caused internal error. Noticeable with the LCD unit with LCD_WRITESTRING enabled.
  • Fixed : 16-bits integer divide routine generates an assembler error.
  • Fixed : Some PIC12 memory schemes was badly interpreted and caused bad banking optimizations (eg: PIC12F635).
  • Added : PIC12F635 extended 14 bits core instructions support, enabling better code generation.
  • Fixed : Bad generated code for IN when the set is 16-bit or 32-bit wide; all processors.
  • Fixed : STR() applied to 32-bit integers produced internal error when {$POINTERS SMALL} was used; all processors.
http://www.pmpcomp.fr/plugins/opendo...de=one&pg=1041