I patched troglobit sysklogd 2.4.0 to avoid the possibility of advancing past the end of a null terminated string, by always testing the value at the pointer before advancing the pointer.
Patch is here:
https://github.com/troglobit/sysklogd/pull/52/files
I do not know of a real-world condition that could trigger this buffer overread. A truncated kernel message buffer would trigger it, and I don't think that's a realistic possibility. But processes running as root should bend over backwards to avoid overread.
