Here is my go at this with what I have - reverseengineering.stackexchange.com/questions/22529/…

Ok I got it to work but now there is probably an issue on the rr side since the step forward packet is not working.

Actually it's IDA side since I've simply ignored the vCont;s:1 packet because I thought it was secondary to the vCont;c one my command is never send

Aha it looks like somehow IDA datects my "bs" GDB command as single steps and so on the next run it issues "vCont;s:1" - now i've prefixed my commands with "debug:command"

Actually no the abnormality seem to came when the program counter is on a breakpoint

But it's not a huge deal I can live with it