Research ways to handle total memory corruption (i.e. causing a double-fault)

Question

Research ways to handle total memory corruption (i.e. causing a double-fault)

vvaltchev opened this issue 8 years ago · 2 comments

Answer 1 · 2018-05-19T15:12:14.000Z

The stack pointer validation does a great job most of the time but, if for any reason the overflow is so big that messes up stuff like the ESP in the current TSS, regular fault handlers cannot run, the CPU triggers a double-fault, but it cannot run as well (because it triggers a page fault) and therefore we get a triple fault and the system reboots. The point of this (non high-priority task) is to research safe ways to handle the double fault and, at least, being able to display a panic message on the screen.

Answer 2 · 2019-08-25T19:59:19.000Z

I'm closing this issue from here, as I'm not using this tool (Github's issues) anymore for self-driven projects/improvements for Tilck. Still, this tool will remain enabled for bug reporting.