Stlink: [feature] Calculate checksums for flash operations

Do you mean a checksum of the memory on the device? And compare that to a file on your computer?

I was thinking of a binary file, on my use case I was having problems with a binary compiled on my computer and one downloaded, the source code was the same, but the binaries did not match, I was using a newer tool chain, so the new binary was corrupt

on st-link tool when you load a binary, ti calculates a checksum, so I used that to figure out when I had the correct toolchain

https://i.ibb.co/5TS7twF/figure.png

Ah okay. These tools don't do that for you as it stands. On Windows I can recommend HxD hex editor to checksum a binary file (or simply compare two binary files quickly). Are you on Windows?

@vitormhenrique: Unfortunately I can't read the content of the image you linked as the resolution is too low for that. Can you paste a larger screenshot or a relevant detail of it?

no idea why that image was so small....

https://i.ibb.co/mq8vvpP/pic.png

I see - so we are actually looking for a feature request here.
The flashloader does verify the integrity of the flashed content behind the scenes, so I assume that this functionality can be extended. It wants to be looked at certainly.

If no special algorithm limitations are given, I think it’s quite easy to integrate some checksum algorithm(md5, sha256, etc) into the file loading process. Maybe I can give it a try when I am free.

stlink utility (the one in the capture) defines the checksum in its documentation (UM0892) as a simple arithmetic sum, truncated to 32bit ... (sigh)

Memory checksum
Calculates the checksum value of a specified memory zone defined by the
address and the size fields in the Memory display section of the main window.
The checksum is calculated based on an arithmetic sum algorithm, byte per
byte. The result is truncated to 32-bit word.
The checksum value is displayed in the log window

While this is quite crude and not reaaally reliable, this is cheap to implement and maintaining "compatibility" with st-utility would be nice.

@grevaillot Thanks for your information, I will add support for this on my PR.

Should we have some documentation for this in our tutorial?

The algorithm used by ST’s windows program may need a documentation at least for developers I think.

Ok, can you add a section with a few lines to our tutorial then (at the end of the reviewed part)? That would be nice. Don't bother too much about formatting there though, as I am still working on this document anyway...

I am still wondering, if the two methods should be merged all together.
If so, I may need to adjust the two functions for a better format. The current two commits would cause a somewhat too verbose output.