Content originally posted in LPCWare by ABR15 on Tue May 10 10:05:27 MST 2016
Quote: vtw.433e
Some question:
Is there a pattern to the corruption?
Have you verified that what is written is what you meant to write?
Over what time does the corruption occur?
Thanks for answer!
Is there a pattern to the corruption? No, the devices that I've verified had an irregular data on this sector of flash.
Have you verified that what is written is what you meant to write? Yes, but is different. This area should has only ASCII characteres (32 to 122 decimal), and now has other values.
Over what time does the corruption occur? I could not reproduce the problem here, in fact it is customer's devices. I think that there is an event there that is changing the values.
There is only one way to write the flash in the firmware. In the main() there is a test like that:
if(flag)
{
write_flash();
}
And there is no pointer, no incorrect use to set this flag, only with the programming interface connected.