Commits


Accidently hard-coded the size of the line buffer in display_buffer(). Replaced with sizeof(line).


Fill the line buffer in display_buffer() beforce actually using it. This should prevent garbage at the end of the line. This bug only showed up at x86 FreeBSD and PPC32 Mac OS X. PPC64 Mac OS X and x86 Mac OS X do not displayed garbage.


Forgot to add -s to the usage() info.


Changed README according to the latest command line switch changes.


Added new -s option to display __TEXT,__cstring section


Added a new option -d to display the __data, __DATA section


Do not optimize for size, optimize for speed.


Added newline at EOF to silence gcc.


Reworked the complete structure of dermob: - Nearly all functions for different parts of the header return a pointer to a struct. - Complete new set of display functions, which uses the new structs - Various functions in dermob.c are way more simplyfied. Tested on x86 and PPC. Code should be stable, but watch out for bugs.


New header file _only_ for function prototypes.


Added a short note about portability


Removed spaghetti code from a wierd code session yesterday :)


Added newlines at end of some files to silence gcc *grml*


Changed NXLittleEndian to our own constant LE.


New header file with most, but not all(!), structs and defines for the mach-o header. Most of the content derives from /usr/include on Mac OS X, so the copyright belongs to Apple Computer, Inc. All rights reserved, thus the license is APPL!


Removed all Mac OS X dependend code. Replaced NXGetLocalArchInfo(3) with two sysctl querys for HW_MACHINE and HW_BYTEORDER. Added includes for a new header file with most, but not all(!), structs and defines for the header.


Added Sparc and HPPA in display_cpu to support older binaries (eg NeXtstep)


Replace NXSwapInt32 with own function function to be more plattform indipendent.


- Added content to the README - Changed my email addresses - infiniteVOID rocks


Commented a debug line


Extend the line array in display_buffer() and increment the counter after storing the characters in line to avoid garbage at the end of the string.


- Added new option -x: Displays a hexdump of the complete binary. Simple, yet powerfull - Sorted all options in usage() alphabetical


Fixed an offset related bug. The __text section starts at VM offset, which is relative to the beginning of the mach-o header. If we parse an universal binary, we have to skip the fat header to calculate the correct offset of the __text section.


- Fixed a typo - Fixed a bug (line array was decremented)


Added ints for __data section New Makefile version