From: Jeff D. <jd...@ka...> - 2001-08-27 19:04:17
|
cem...@ch... said: > That address at frame #5 is the faulting address, and isn't mapped; so > it's a bit odd that is_write=1. Within segv, the "is_write && !(vma-> > vm_flags & VM_WRITE)" test is setting ok=0. Well, if that's code, then nothing should be attempting to modify it. If !(vma->vm_flags & VM_WRITE) then that's a readonly area of virtual memory and UML is correct to segfault writes to it. Are you sure your SC_FAULT_WRITE is correct? It seems to me that it can't be, because if the fault is the result of trying to execute whatever code is there, that's not an attempt to write. That's a read. Jeff |