Data is still dummy.
Not even compiled.
modified: executer/kernel/mcctrl.h
modified: executer/kernel/syscall.c
modified: kernel/include/syscall.h
modified: kernel/mem.c
We can create a dummy core file when gpe occurs.
modified: executer/kernel/mcctrl.h
modified: executer/kernel/syscall.c
modified: kernel/include/syscall.h
modified: kernel/mem.c
When a open system call request is delegated, mcexec will check whether
a path name is known one such as "/proc/meminfo" and "/proc/cpuinfo".
During this check, if mcexec causes a SIGSEGV or SIGBUS signal,
mcexec will begin a loop, because mcexec has a signal handler which
returns simply.
The mcexec may send a signal to itself in order that the mcexec
terminates abnormally.
Generally, the signal is delivered to the thread which sent the signal.
In this case, the signal terminates the mcexec immediately.
However, if the signal is delivered to another thread,
the thread which sent the signal can call exit(3)
before the signal terminates the mcexec.
When a signal is forwarded from McKernel,
ioctl(MCEXEC_UP_WAIT_SYSCALL) may return EINTR error.
In this case, though no syscall request has been received,
the mcexec starts to process the syscall requst that is garbage.
implemented:
- Pages can be shared between maps.
- A change made to a map is written to the file, at munmap().
not yet implemented:
- VM operation during page IO execution.
Because page IO is executed with VM's lock.
- Page IO, which does not change a file size with any case.
When munmap() races with truncate(), the file size may be changed
illegally.
It was able to change a read-only memory using a system call that
forwarded to host.
To fix this, when read-only mmap(), it places read-only vm_area in the
mcexec's space.
