> > 2.0.x preserves the read position when you get an EFAULT. Nothing I can find
> > in the holy manual of posix says what happens if you get an EFAULT on a tcp
> > operation, so I think we can do what we like.
> >
>
> In fact, POSIX doesn't distinguish between system calls and library
> functions, so according to POSIX it is perfectly valid to throw
> SIGSEGV instead of return EFAULT. Personally, I think there should be
> a kernel option to turn EFAULT into SIGSEGV, because that's really
> what it is -- a segfault where the access happens to be from kernel
> space.
I *really* like this idea. I'm now doing some library hacking, and I
think that SIGSEGV is just what they deserve. (You can not guarantee
-EFAULT if read is emulated using library :-(, you'll get sigsegv
instead.)
Pavel
-- I'm really pavel@atrey.karlin.mff.cuni.cz. Pavel Look at http://atrey.karlin.mff.cuni.cz/~pavel/ ;-).- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu