Short: ASM-Pro v1.22 (update) Author: amigo@binary.si (a/b) Uploader: amigo binary si (a/b) Type: dev/asm Version: 1.22 Replaces: dev/asm/ASMPro1.21.lha Architecture: m68k-amigaos >= 2.0.4 ====================================================================== Update for the ASM-Pro v1.18 package, includes latest executable. Complete source code is available in a separate package. [ ADDITIONS/MODIFICATIONS ] 1. ASM: Added QREPT directive (paired with ENDR, just like REPT). It generates identical copies of the repeat block resulting in shorter assembly time. It evaluates the block (at most) once and then copies it as many times as needed. Consequently, you can declare labels and constants within, there will be no double symbol errors. 2. DISASM: FPU k-factor is now displayed as a signed 7-bit integer instead of unsigned. This is more in line with Motorola documentation. 3. Additional SE (special edition) build of Asm-Pro without gfx assets (logotips), resulting in lower chipmem usage by about 16KB. [ BUGFIXES ] 01. UI: Duplicate menu Commands->Output. 02. EDIT: Pasting a larger block freezes the editor until ESC is pressed. 03. ASM: Using directives WO/WX/WL to write a file with a large number of relocs could overflow the stack. Added stack protection. 04. ASM: Incorrect debug line info when a macro contains a directive, instruction, or macro call with constructed name (e.g. using macro arguments). 05. ASM: REM and TEXT blocks are missing debug line info, causing the source level debugger to misbehave when manually setting a breakpoint. 06. EDIT: Incorrect free memory amount displayed if more than 99999KB is available. 07. EDIT: Long filenames (more than 30 characters) of source files overrun several internal buffers. 08. EDIT: Old source's current directory filename buffer is padded instead of new source's buffer during source switching, which could lead to path corruption. 09: UI: Tag lists incorrectly terminated in some cases (-1 instead of 0/TAG_DONE). 10: ASM: K-factor in fmove.p is not parsed properly (in both data register and numerical form), and it could also result in entirely incorrect opcode. 11: EDIT: Syntax coloring in 1-bitplane mode is not disabled due to a bad screen depth check. 12: UI: Screen depth exceeding maximum (4) causes endless loop. Default screen depth will be used instead in such cases. ======================================================================