RDV4 Client vs BootmRom/OS firmware not matching

[bodhisattva]

Hi - guys, I have a RDV4 that has been flashed before and it shows

MCU… AT91SAM7S512 Rev A
Memory… 512 KB ( 80% used )
Target… RDV4

Client… Iceman/master/v4.20728-288-g2a2c06e61 2026-01-02 10:10:55
Bootrom… Iceman/master/v4.20728-288-g2a2c06e61-suspect 2026-01-02 09:57:53 7cdeebf82
OS… Iceman/master/v4.20728-288-g2a2c06e61-suspect 2026-01-02 10:01:07 7cdeebf82

From what I have gathered Client / Bootrom and OS should all match. I tried reflashing it to the version from the ‘Getting started with Proxmark3 Guide ( keeping in mind not to change the makefile and leave it as RDV4 then compiling it as needed) with the hopes of having everything match. However it still shows as the output above.
Is there something I am missing?

[Pilgrimsmaster]

Try this

@Pilgrimsmaster

• cd proxmark3
• git pull
• make clean && make all OR make -j (it will allow for parallel compiling and goes faster. - Thanks @ Iceman) OR make -j4 (for compiling in parallel on all cores - Thanks @ equipter)
• ./pm3-flash-bootrom
• ./pm3-flash-fullimage
• pm3
• hw tune

@Iceman

to clarify:

make -j == use all available cores to parallel compile.

make -j4 == use 4 cores to parallel compile.

If you have a 4 core CPU or more you can specify using -j8, -j6 etc, but 99.9% of the common use case is do not specify the number of cores.

The make all is from earlier Makefile’s when we didn’t rebuild it. Today running make implies make all

To conclude:

use: make -j

for the fastest build time using the maximum number of available cores on your system.

Personally: make clean; make -j

[bodhisattva]

Thank you again for the tips.

I pulled from git and it looked like I had the latest version, ran make file as needed and pushed both the bootrom and full image. Running ./pm3 still produces a mismatch.

MCU… AT91SAM7S512 Rev A
Memory… 512 KB ( 80% used )
Target… RDV4

Client… Iceman/master/v4.20728-288-g2a2c06e61 2026-01-02 14:15:59
Bootrom… Iceman/master/v4.20728-288-g2a2c06e61-suspect 2026-01-02 14:01:34 7cdeebf82
OS… Iceman/master/v4.20728-288-g2a2c06e61-suspect 2026-01-02 14:05:52 7cdeebf8

I have a Blueshark module and was wondering ( perhaps poorly). If I install it and run the BlueShark add-on Build from Proxmarkbuilds.org is that just adding more complexity to this?
I can run hw tune and even do an hf mdes info on my Desfire EV3 K8kb fob…But when running other commands like hw status I am seeing that none of my dictionaries are not loaded..which would suggest that something is up

] Flash memory dictionary loaded
[#] Mifare… 0 keys - dict_mf.bin
[#] T55xx… 0 keys - dict_t55xx.bin
[#] iClass… 0 keys - dict_iclass.bin
[#] UL-C… 0 keys - dict_mfulc.bin
[#] UL-AES… 0 keys - dict_mfulaes.bin

[amal]

@bodhisattva

still produces a mismatch.

Why do you think this? Looks matched to me. At this level of client version it will tell you flat out with a red line if text if there is a mismatch.

[bodhisattva]

Hey Amal - Happy New Years. So I noticed that the code looks identical….but what threw me off is that I was expecting all the lines to be exactly the same.

[amal]

Happy new year!

Yep the version number and date are the parts that matter.

@bodhisattva

v4.20728-288-g2a2c06e61 2026-01-02

That’s the important bit, and they all match

[bodhisattva]

Thanks a bunch, I’ll go ahead and install the BTmodule and flash it with its build!!