* .h include only the strict minimum for their own parsing
* this forces all files to include explicitment their needs and not count on far streched dependencies
* this helps Makefile to rebuild only the minimum
* according to this rule, most standalone .h are now gone
* big app.h is gone
* remove seldom __cplusplus, if c++ happens, everything will have to be done properly anyway
* all unrequired include were removed
* split common/ into common/ (client+arm) and common_arm/ (os+bootloader)
* bring zlib to common/
* bring stuff not really/not yet used in common back to armsrc/ or client/
* bring liblua into client/
* bring uart into client/
* move some portions of code around (dbprint, protocols,...)
* rename unused files into *_disabled.[ch] to make it explicit
* rename soft Uarts between 14a, 14b and iclass, so a standalone could use several without clash
* remove PrintAndLogDevice
* move deprecated-hid-flasher from client to tools
* Makefiles
* treat deps in armsrc/ as in client/
* client: stop on warning (-Werror), same as for armsrc/
Tested on:
* all standalone modes
* Linux
* use macros for versions
* activate UNDERSTANDS_VERSION in bootrom
* fix missing break; bug in bootrom
* force flash_load to reject images > 256k if bootloader not up to date
* move logic from flasher to flash
- Keep button pressed when powering to force entering bootloader
- Once in bootloader more, you can release the button
- When in bootloader mode (no matter how you entered that mode)
- One button press to quit bootloader mode and boot main image
CHG: flasher - removed unneeded parameter to function
CHG: flasher - increased limit to fpga-files that can be loaded
FIX: main client, wrong windows define fixed.
CHG: device side - recieve usb command does not always get 544 bytes. usb packages can be incomplete.
CHG: usb_cdc.c also got piwi's changes.
CHG: uart_posix.c removed a debug value
CHG: uart_win32.c clean up.
.. ntag simulation stuff from @marshmellows branch "ntag/sim"
.. hf mf mifare fixes from @pwpivi.
.. hw status command
.. speedtest function from @pwpivi
.. Viking Functionalities, (not a proper DEMOD, but a start)
.. GetCountUS better precision from @pwpivi
.. bin2hex, hex2bin from @holiman
...
starting with getting the T55x7 CONFIGURATION_BLOCK for different clone situations. Ripped from Adam Lauries RFidler, nothing working or finished..
...
Started working with the T55x7 read command with password actually performs a write block... See Issue #136https://github.com/Proxmark/proxmark3/issues/136 Not solved yet.
...
Started add SHA256.. not working yet..
MERGED: @piwi changes
MERGED: @marshmellows changes.
I'm not even gonna try write up all that stuff..
ADD: changed some commands inside the "Hf 14a sim" on deviceside.
ADD: @mobeius "two nonce" version for mfkey32. It is also inside the "hf 14a sim" with the "x" parameter.
REM: removed old comments in bootrom.c
CHG: mifare ultralight & desfire commands inside arm has been cleaned up. Next step is to refactor it into armsrc/mifaredesfire.c
512k chip support isn't complete anyway and the incorrect clock would've
ended up destroying Flash anyway. The 256k waitstates were set to 0,
which is outside the spec. Set them to 1, which is allowed.
- Clean up bootloader asm
- Remove fromflash.c - it's not worth doing in C, do it in ASM
- Clean up linker script
- Force use of symbol inside bootphase2 (otherwise linker
garbage-collects it)
- Link bootloader with gcc instead of ld
I have kept whatever copyright notices exist. Please add your own
copyright notice if you have made any nontrivial changes or additions to
the code. There are several files without any attribution, currently.
+ Add common area at end of RAM to communicate between main os and bootrom
+ Lower stack end to make room for common area
+ Implement CMD_DEVICE_INFO in both OS and bootrom to give information about the current state and supported features
+ Allow hands-free firmware update: When CMD_START_FLASH is received over USB in OS mode, the device will reset and enter the bootrom
Pressing the button in hands-free update mode takes precedence: releasing the button will immediately abort firmware update and
perform a reset. Do not press the button.
+ Require each flash sequence to be preceded by a CMD_START_FLASH to set up the boundaries for the following flash sequence
Not compatible with linux flasher before SVN revision 200
Currently no compatible flasher for Windows. WINDOWS USERS: DO NOT UPDATE YOUR BOOTROM YET
+ Protect bootrom flash area unless magic unlock sequence is given in CMD_START_FLASH
Use perl to create the version information (thereby re-creating the perl dependency and adding an svn dependency) but fall back in case of missing perl or svn
Retire rbt2c.pl, instead use objcopy to directly convert the .bit file into an .o that can be linked with the flash image
Rename armsrc/fpga.c to armsrc/fpgaloader.c (since there is now a new fpga.o, created from fpga.bit)
Remove fpgaimg.c from subversion, add fpga.bit
Instead of creating fpgaimage.elf and osimage.elf separately, now create a joined fullimage.elf
first (obsoleting ldscript-full), then extract only the fpga and os sections with objcopy
(This creates unspecific warnings about an empty segment, need to investigate)
Implement a rudimentary .bit parser in the firmware, use that to locate the bitstream in the new
fpgaimage (which is just a plain copy of the fpga.bit file) and send it to the FPGA
The code will check the format that's in flash and fall back to the legacy format
Remove the now unnecessary files (merge-srec.pl, ldscript-ram-jtag)
Note that this drops the dependency on perl for bootrom build, so end-users who don't touch the FPGA bitstream will not need perl anymore
+ Needs GNU make for Windows, a new release of the Windows development environment will follow
+ Is based on the Windows Makefile, so will build everything in Thumb mode, doesn't have flash commands (yet)
- Cleaned up Makefile.linux in armsrc directory
- Added Linux toolchain build script in tools directory
- Made LCD support optional (disabled by default) in armsrc
- Small formatting changes
It is identical to the popular 20081211, with the doob addition (20090301), a
linux client, and two additional commands for LF analysis. Let me know if
you find issues here!
