https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=292206
--- Comment #2 from Jarmo Jaakkola <[email protected]> --- I did not fully investigate what exactly causes the system to reboot, only the conditions for triggering it. My wild guess is that the kernel does not properly validate the hw.uart.console value and then tries to use 0x0 as memory mapped I/O address and -4 as register shift. If my hunch is correct, I'd suppose that in addition to my suggested fix that prevents loader from generating the invalid value, the kernel should also try to validate the values it receives in hw.uart.console. -- You are receiving this mail because: You are the assignee for the bug.
