- fix: IQ demodulator (FPGA)
- fix: approximately align reader signal delay to tag response delay (FPGA)
- fix: remove deprecated RSSI calculation to improve decoder speed (iso14443b.c)
- fix: better approximation of signal amplitude to avoid false carrier detection (iso14443b.c)
- fix: remove initial power off in iso14443b raw command (iso14443b.c)
- add: enable tracing for iso14443b raw command (iso14443b.c)
- fix: client crashed when checking CRC for incomplete responses (iso14433b.c)
- speeding up snoop to avoid circular buffer overflow
- added some comments for better documentation
- rename functions (iso14443 -> iso14443b)
- remove unused code in hi_read_rx_xcorr.v
- most significant bit of tag data (which happens to be the sign bit)
had been dropped when snooping (FPGA change)
- avoid trying to decode both tag and reader data when snooping (we don't
have the time to do so).
- added compiler options -fdata-sections and -ffunction-sections (thanks to iceman for the hint)
- removed float operations from common/lfdemod.c to avoid adding float libraries to the ARM os image
- moved the fpga images to the data section to avoid reserving unused space for a separate section
In Miller Decoder: don't wait too long for a stable signal
In Miller Decoder: Don't accept sequences of four or more zeroes as start bit
In EmSendCmd14443aRaw: don't wait for emptying the FPGA delay queue if it isn't filled
- provided a BigBuf_malloc() function to dynamically allocate parts of BigBuf
e.g. for DMA-Buffers, Frame-Buffers, Emulator-Memory
- the whole rest of BigBuf is now available for traces (instead of a small fixed amount)
- send actual traceLen together with trace data
- changed client side to cope with varying traceLen
- changed small buffers to automatic variables instead of parts of BigBuf
- introduced with the big frame and parity support (commit 6a1f2d82): tag responses with len%8 == 0 were dropped - thanks iceman for testing and finding
- after unsuccessful hf 14a reader the field stayed on. Thanks to iceman for proposing the fix.
Support for bigger iso14443 frames (including parity calculation)
unify hw tune and data tune
fix hf epa cnonces
adapt hf iclass list to new trace format
- unified hw tune and Enio's great data tune
- don't use BigBuf (and hardcoded Offset)
- removed special handling of CMD_MEASURED_ANTENNA_TUNING
in UsbCommandReceived()
- buffers were too small to handle 256 byte frames
- parity bits were only handled for up to 32 byte frames
- trace format was inefficient
- removed parity calculation from decoders in iclass.c (parity not used on air anyway)
- hf mf dump: abort on unrecoverable errors. Don't create file dumpdata.bin in this case.
- hf mf chk: re-enabled and fixed option d (dump keys to dumpkeys.bin).
if there are unknown keys, write 0xffffffffffff instead to the file.
- provide a default key dictionary file for hf mf chk (default_keys.dic). Contents taken from
mf_default_keys.lua
- hf mf rdsc (fix): didn't account for 16 block sectors, allowed max sector 63 instead of 39
- hf mf ecfill (add): added (optional) card size parameter and support for non 1K cards
- hf mf dump (add): added (optional) card size parameter and support for non 1K cards
- hf mf dump (fix): Access Condition 011 not handled correctly (tried to access with key A)
- hf mf restore (add): added (optional) card size parameter and support for non 1K cards
- hf mf nested (fix): didn't account for 16 block sectors, allowed max sector 63 instead of 39
- hf mf nested (fix): always dumped 16 keys to dumpkeys.bin instead of correct number
- hf mf chk (fix): always dumped 16 keys to dumpkeys.bin instead of correct number
- hf mf eget (fix): displayed three instead of one block
- hf mf eload (add): load 4K .eml files (but accepts 1K .eml files for backwards compatibility)
- hf mf esave (add): always save the whole emulator memory (4K) instead of 1K only
- hf mf ecfill (add): added (optional) card size parameter and support for non 1K cards
- output of debug messages caused communication failures due to timing issues.
hf mf dbg 4 now required to see these debug messages.
- changed help text for hf mf dbg
- fixed minor bugs in help texts for hf mf sim and hf mf ecset
- display "key A" or "key B" instead of "key=0" or "key=1 in hf mf sim
- tracing was not always enabled when starting hf mf sniff or hf 14a snoop
- ATQA was displayed in wrong byte order in hf mf sniff
- 4 Byte UIDs were displayed as 7 Byte UIDs (padded with 0x000000) in hf mf sniff
- same for logfile names.
- assignment (=) had been used instead of == in comparisons (shouldn't have been relevant though)
In order to be able to distinguish between Mifare Classic and Mifare Plus
in Security Level 1 (SL1, Mifare Classic Compatibility Mode), hf 14a reader
now always tries RATS - even if SAK claims not to support ISO14443-4.
- fixed a circular buffer rollover bug in iso14443a.c
- fixed 7 Byte UID handling in hf mf sniff
- fixed "cannot append" error in hf mf sniff d
- fixed hint on mfkey32 in hf mf sim x
- fixed hf mf sniff sometimes showing rogue data from previous calloc
- improve snooping/sniffing by syncing modulation detector window with reader signal (hi_iso14443a.v)
- code cleanup of hi_iso14443a.v
- enhanced tracing: hf 14a list now shows meaningful timing information. With new option f it also shows the frame delay times (fdt)
- small fix for hf 14b list - it used to run into the trace trailer
- hf 14a sim now obeys iso14443 timing (fdt of 1172 or 1234 resp.)
Note: you need to flash FPGA as well.
More details in http://www.proxmark.org/forum/viewtopic.php?pid=9721#p9721
- PM: used GetCountMifare in MifareNested() for improved timing accuracy and to deliver better quality nonces
- PM: MifareNested now delivers exactly two different nonces to avoid time consuming multiple lfsr_recovery32() on client side
- Client: replaced quicksort by bucketsort in crapto1.c which is faster
- Client: use multithreading (two parallel calls to lfsr_recovery32())
- Client: fixed a small bug in mfnested() (always showed trgkey=0)
- Client: introduced a mutex for PrintAndLog() to avoid interlaced printing
Minor rework of hf mf chk:
- Avoid time consuming off/on cycles. Send a "halt" instead.