PM3_CMD_DATA_SIZE went 512 -> 624 without a capabilities bump, so a new
client connects to old firmware and every oversized command dies at the
device's length check with no message.
Append max_cmd_data_size, bump to v9. The client now accepts an older
capabilities struct - it only ever grows by appending, so an older layout
is a prefix - and defaults the frame size for pre-v9 firmware.
SendCommandNG bounds by the device value instead of the compile time one.
Also zero init capabilities_t on the device, it leaked stack bytes.
chunked by PM3_CMD_DATA_SIZE. Identical today, but if the NG size moves the
chunk would be built oversized, truncated on the wire, and still announced
at full length in oldarg[1] - the client would copy past the valid bytes and
advance by the wrong stride. Bound the client's OLD download branch by the
same constant.
The OLD frame size was tied to the NG one, but the bootloader only speaks
OLD - growing PM3_CMD_DATA_SIZE would silently change sizeof(PacketCommandOLD)
and break flashing against every deployed bootrom in both directions.
Pin the OLD structs to their own constant and use it on every OLD path:
reply_old and the OLD receive branch on both sides, the bootrom, and the
flasher's write_block/send_finish_write_cmd, which memcpy into a
PacketCommandOLD using the NG size.
No behaviour change - both constants are 512 and armsrc .text is
byte-identical before and after.
Bug triggered with:
lf read -s 50000
hw ping
ReadLF_realtime() in lfsampling.c never knew how many samples the client wanted — for the realtime path, lf_sample_payload_t.samples was never populated on the client side, and the ARM handler in appmain.c (CMD_LF_ACQ_RAW_ADC/CMD_LF_SNIFF_RAW_ADC) ignored it anyway. So the device just streamed samples indefinitely, only checking for a stop request roughly once per 64-sample USB burst, and even after noticing it, flushed one more trailing partial burst via async_usb_write_stop().
Meanwhile the client (WaitForRawDataTimeout in comms.c) stops listening in raw mode the instant it has collected exactly the number of bytes it asked for (e.g. 50000), then switches back to normal framed-packet parsing. Since the device kept sending more bytes past that point, those extra bytes sat unread in the pipe and got misinterpreted as a PacketResponseNGPreamble/OLD-frame header, corrupting every subsequent USB exchange (hw ping, etc.) — reproducible even with a plain lf read -s 50000, no trigger/COTAG needed.
Moreover, we make sure the client doesn't stop listening, even briefly, during stream reception.
Works when the device is running either osimage or bootloader.
- New memory reading command in osimage and bootloader.
- Extended 'hw readmem' command with length parameter, file writing and hex viewer.
- Introduced '--dumpmem' option to proxmark3 executable to support dumping from bootloader.
Simple interactive examples:
hw readmem -f flashdump
hw readmem -l 1024
CLI example:
./pm3 --dumpmem flashdump.bin
Reading from arbitrary memory ranges can be unlocked using the 'raw' option.
Check if memory allocation fails
Fix memory leak
Initialize struct in declaration
Add/Fix some notes
Remove unlikely() in favor of readability
Remove a hard-coded magic number
Rename UART_TCP_CLIENT_RX_TIMEOUT_MS to UART_NET_CLIENT_RX_TIMEOUT_MS
Add UART_UDP_LOCAL_CLIENT_RX_TIMEOUT_MS for shorter timeout
Check if the target address is local in uart_open()