Re: make PROT_WRITE imply PROT_READ

[Chase Venters]

On Wed, 28 Jun 2006, Pavel Machek wrote:

> On Wed 2006-06-28 09:43:22, Ulrich Drepper wrote:
> > On 6/27/06, Pavel Machek <pavel@xxxxxx> wrote:
> > > Usability for "normal" C applications is probably not too high... so
> > > why not work around it in glibc, if at all?
> >
> > Because it wouldn't affect all b inaries.  Existing code could still
> > cause the problem.  Also, there are other callers of the syscalls
>
> _There is no problem_.
>
> mmap() behaviour always was platform-specific, and it happens to be
> quite strange on i386. So what.

Hell, IIRC, on ConvexOS 11, the second argument to mmap() is a /pointer/ 
to the length.

> > (direct, other libcs, etc).  The only reliable way to get rid of this
> > problem is to enforce it in the kernel.  Since the kernel cannot make
> > sense of this setting in all situations it is IMO even necessary since
> > you really don't want to have anything as unstable as this code.
>
> Current kernel behaviour is useful for specialized apps. If you do not
> want to see that weirdness in regular c application, work around it in
> glibc.
> 									Pavel

Thanks,
Chase
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/