On Mon, 18 Aug 2025 15:18:14 +0200 Benoit Panizzon <[email protected]> wrote: [...] > I suspect it could be a problem affecting only that specific CPU indeed > hitting an invalid opcode. [...]
I don't think that's what's happening. The kernel intentionally has an invalid instruction on each assertion failure path, because this is smaller than a subroutine call and saves more CPU state. Ben. -- Ben Hutchings Time is nature's way of making sure that everything doesn't happen at once.
