mprotect

[Jan Althaus]

Hi,
I'm having some trouble using mprotect. A short code snippet to give
you an idea of where this is heading:
char *p;

p = malloc(1024+PAGESIZE-1);
if (!p)
exit(errno);
p = (char *)(((int) p + PAGESIZE-1) & ~(PAGESIZE-1));
memset( (void*)p, 0xc3, 1024 ); /* set everything to "ret" */

/* here we set p to contain some binary code */

if (mprotect(p, 1024, PROT_EXEC|PROT_WRITE|PROT_READ)) {
exit(errno);
}
__asm__ __volatile__ (
"call *%0 \n"
:
: "m"(p)
);
exit(0);

Now the problem is that the call always causes a segfault. I tried
setting the memory to just PROT_EXEC with no luck... so I'm wondering:
Is there something else I need to do in order for Linux to allow me
execution of a certain page? I'm really out of ideas here... :/
When I pass the last asm segment a pointer to a function that I
declared in c++ it works fine. So it has to be the memory protection
not doing what I expect it to...

 

[santosh]

Hi,
I'm having some trouble using mprotect. A short code snippet to give
you an idea of where this is heading:

<snip>

This is a Linux related question try posting to one of the groups under
the comp.os.linux hierarchy. Also try comp.unix.programmer.

 

[Keith Thompson]

I'm having some trouble using mprotect.

mprotect() is not a standard C function; try comp.unix.programmer.
But I'll add a few C-related comments as well.

A short code snippet to give
you an idea of where this is heading:
char *p;

p = malloc(1024+PAGESIZE-1);
if (!p)
exit(errno);

errno values and arguments to exit() are not necessarily related
(beyond the fact that 0 means "no error" for both). In standard C,
the only portable arguments to exit() are 0, EXIT_SUCCESS, and
EXIT_FAILURE. If you're not concerned with portability to all C
implementations (and since you're using mprotect(), presumably you're
not), you can use other values, but errno values aren't necessarily
sensible.

p = (char *)(((int) p + PAGESIZE-1) & ~(PAGESIZE-1));
memset( (void*)p, 0xc3, 1024 ); /* set everything to "ret" */

There's no need to cast the first argument to memset(); char*, or any
other pointer-to-object type, is implicitly converted to void*.

/* here we set p to contain some binary code */

if (mprotect(p, 1024, PROT_EXEC|PROT_WRITE|PROT_READ)) {
exit(errno);
}
__asm__ __volatile__ (
"call *%0 \n"
:
: "m"(p)
);

Obviously this is extremely non-portable; it's probably specific to
whatever compiler you're using (gcc?).

exit(0);

Now the problem is that the call always causes a segfault. I tried
setting the memory to just PROT_EXEC with no luck... so I'm wondering:
Is there something else I need to do in order for Linux to allow me
execution of a certain page? I'm really out of ideas here... :/
When I pass the last asm segment a pointer to a function that I
declared in c++ it works fine. So it has to be the memory protection
not doing what I expect it to...

Declared in C++? If your code is C++ rather than C (which may also
affect implicit conversions to void*), you should have wrongly posted
to comp.lang.c++ before being redirected to comp.unix.programmer.
}

 

[Jan Althaus]

Sorry, last one was a typo. It's C code of course.
Thanks for the feedback! I'll keep it in mind.
Regarding portability: Seeing as I'd only like to get the basic
principle working I'm not really cocerned with that right now.

 

[Kenny McCormack]

Sorry, last one was a typo. It's C code of course.

In the theology of this ng, it is *not* C code.

You may (quite sensibly) think otherwise.

Thanks for the feedback! I'll keep it in mind.
Regarding portability: Seeing as I'd only like to get the basic
principle working I'm not really cocerned with that right now.

That is no more permissible a view here than would be the view that
abortion is permissible when addressing a KofC audience.