ee3caa0d62
path. The basic problem is that we cannot set the single stepping flag directly, because we don't leave the kernel via an interrupt return. So, we need another way to set the single stepping flag. The way we do this is by enabling the lower-privilege transfer trap, which gets raised when we drop the privilege level. However, since we're still running in kernel space (sec), we're not yet done. We clear the lower- privilege transfer trap, enable the taken-branch trap and continue exiting the kernel until we branch into user space. Given the current code, there's a total of two traps this way before we can raise SIGTRAP. |
||
---|---|---|
.. | ||
acpica | ||
compile | ||
conf | ||
disasm | ||
ia32 | ||
ia64 | ||
include | ||
isa | ||
pci |