1.15.3

• Don't use append the section name if it is known when disassembling elfs
  • This special cases the sections .text, .data, .rodata and .bss.
  • Avoids the redundant filename_.text/ naming scheme

1.15.2

• Fix hardcoded shift value in alignment directive

1.15.1

• Emit string alignment directives even when the section isn't aligned to a multiple of 8.
  • Some projects can have rodata sections aligned to just a multiple of 4, and not emitting the directive in those cases can break their builds

1.15.0

• Change the string guesser to work with multiple levels instead of plainly enabled/disabled and the aggressive toggle.
  • The new option is used via the API GlobalConfig.RODATA_STRING_GUESSER_LEVEL or via the CLI --rodata-string-guesser level.
  • The old GlobalConfig.STRING_GUESSER and GlobalConfig.AGGRESSIVE_STRING_GUESSER options are now deprecated, same as the CLI options --string-guesser and --aggressive-string-guesser.
    • The old GlobalConfig.STRING_GUESSER = True is equivalent to the new GlobalConfig.RODATA_STRING_GUESSER_LEVEL = 1
    • The old GlobalConfig.AGGRESSIVE_STRING_GUESSER = True is equivalent to the new GlobalConfig.RODATA_STRING_GUESSER_LEVEL = 4
  • Meaning of the new levels:
    • level 0: Completely disable the guessing feature.
    • level 1: The most conservative guessing level. Imposes the following restrictions:
      • Do not try to guess if the user provided a type for the symbol.
      • Do no try to guess if type information for the symbol can be inferred by other means.
      • A string symbol must be referenced only once.
      • Strings must not be empty.
    • level 2: A string no longer needs to be referenced only once to be considered a possible string. This can happen because of a deduplication optimization.
    • level 3: Empty strings are allowed.
    • level 4: Symbols with autodetected type information but no user type information can still be guessed as strings.
  • The level defaults to 1.
• Implement string guesser for the data section.
  • Controlled by the API GlobalConfig.DATA_STRING_GUESSER_LEVEL or via the CLI --data-string-guesser level.
  • Decodes strings with the ASCII encoding by default.
  • The meaning of each level are the same as the rodata string guesser.
  • The level defaults to 2.
• Add experimental Pascal string guesser.
  • Works for both rodata and data sections.
  • Follows the same level logic as the C string guesser.
  • It is disabled by default.
• Start emitting .size directives by default.
• Emit jlabel instead of dlabel for jumptable labels by default
• Emit dlabel instead of dlabel for data, rodata and bss symbols by default

1.14.3

• Failcheck for non aligned doubles
• (Hopefully) Fix same-vram overlays using symbols from other overlays
• elfObjDisasm: Can now disassemble sections with arbitrary names
• disasmdis: Disable pseudo instructions by default

1.14.2

• Actually add py.typed to pyproject.toml
• Use bytearray as little as possible
• writeBytearrayToFile is now deprecated, use writeBytesToFile instead

1.14.1

• Emit a previous alignment directive for strings.
  • Ensures strings are always word aligned
• Purge .balign directive in favor of .align directive
• Add py.typed file. Whoops

1.14.0

• Try to support better N32 PIC programs.
  • The current issue was spimdisasm was not able to properly generate symbol references for $gp accesses.
  • GOT table now gets its own address from the reginfo instead of the dynamic table.
  • Accesses pointing outside the GOT table are tried to be redirected to sdata, srdata and sbss sections.
• Implement --dyn-syms on readelf-like mode.
• Minor improvements to readelf output format

1.13.3

• Fix not writing to subfolders properly when a csv filesplit entry has a slash on its name.

1.13.2

• Add support for .dummy section in csv file split format
• Add readelf's --section-headers flag to elfObjDisasm