Reading fuses over icsp error

[vdudouyt]

Reading fuses over icsp error

[vdudouyt]

Hi,

That looks to be fixed by #47d4e33 (thanks to Lubomir). Could you give a try again?
Looking forward to hear from you.

Valentin

[z-op]

minipro -i -p "atmega16a" -c config -r atmega16a.fuses
Found Minipro TL866A v3.63.2
IO error: expected 8 bytes but 5 bytes transferred

same with atmega32
over zif socket fuses reads fine.

[danilorcesar]

Hi,

I have the following error message:

minipro -p W25Q16BV -r teste.bin
Found Minipro TL866A v3.72.2
IO error: expected 8 bytes but 5 bytes transferred

Can you help me?
Regards,

[elosha]

Could be caused by potentially buggy programmer information in our devices.h. If so, it would cause a difference between the number of fuse bytes we expect (correct) and what we instruct the programmer to deliver (incorrect, but also intransparent).

Evidence: Similar issue existed in the original Windows tool until 2015-06-12, changelog of version 6.16:

Fix：some ATMEGA8 Fuse bits Program

In this case, one would need to extract updated device info from a recent Windows tool. (Where did devices.h originally came from?)

Similar: #44

[DavidGriffith]

@elosha I just now emailed @vdudouyt asking where devices.h came from.

[chenz]

The device database must have been reverse engineered from InfoIC.dll from the windows software. I have also been looking into it. One cannot read the tables straight from the dll though, because parts of them are populated by code at runtime. I think it might be easiest to write a small windows tool that loads the dll and dumps the tables. I might try to do that next.

More info here: http://www.nullsecurity.org/article/minipro_reverse_engineering_the_infoic_dll (see the comments - they dumped the populated tables from a memory snapshot taken in a debugger).