Helge Skrivervik

Thank you @ghaerr. Agreed - and as the mount commands show, there is ample space on the fs. Also I agree it's likely a separate issue, it would be interesting...

Finally some (real) progress on this issue: - The problem is a ‘memory leak’. 4 bytes at a fixed address in the kernel text segment are being overwritten at some...

Hi @ghaerr, I just noticed the omission in my report, please ignore the '2d0' vs 'd0' segment difference. I forgot to correct that - they’re from two different runs, different...

Thank you @ghaerr - > > With regards to the origin of the memory corruption: > > elks17# hd d0:16238#4. 006f38: c7 ef 00 f0 eb cc 56 57 55...

Thank you @tkchia and @ghaerr - that's a lot of useful info and suggestions. I'll attack them one by one. - I did take out IDE_PROBE, but as expected it's...

Thank you @tkchia, I did rush to get this one teated, AND IT SEEMS TO WORK. Admittedly I need to sit down tomorrow and understand exactly ’the magic’, but it...

Hi @tkchia and @ghaerr - while there is still some tests I'd like to do - time permitting, I'm quite confident that the bug has been found and fixed. I've...

> It's indeed fascinating that a 4 week hunt for what initially seemed like something else, suddenly was fixed by changing one single letter in the bootblock code. > >...

OK, here's the file - and a partial disasm skipping the strings and the partition table. [SX-MBR.bin.gz](https://github.com/jbruchon/elks/files/9331781/SX-MBR.bin.gz) Looking at this it's easy to want more, such as the ability to...

Thank you @tkchia - actually I had the same thought when a saw the string towards the end. Looking forward to your verdict :-) -M BTW, continuing on the same...