passing NULL to clock_getres (VDSO): terminated by unexpected signal 11

[Thomas Gleixner]

On Sun, 20 Oct 2019, Andreas Schwab wrote:

> On Okt 20 2019, Thomas Gleixner <tglx at linutronix.de> wrote:
> 
> > clock_getres(NULL) is hardly valid.
> 
> Of course not, it lacks a parameter.

You know exactly what I mean.
 
> > So special casing
> >
> >     clock_getres(clock, NULL);
> >
> > just to make a test case happy is a pointless exercise which does not make
> > any sense at all.
> 
> POSIX requires it to work.

POSIX does not mention anything about the validity of the pointer handed to
clock_getres().

https://pubs.opengroup.org/onlinepubs/9699919799/

 "The clock_getres(), clock_gettime(), and clock_settime() functions shall fail if:

   [EINVAL]
     The clock_id argument does not specify a known clock."

EINVAL is the only documented error code for clock_getres(). 

Again. The VDSO is not a syscall and therefore neither address space
validation of the supplied pointer nor catching an invalid access works
like it does in a syscall.

The only invalid pointer it can catch is NULL, but that's ONE case out of a
gazillion. So what's the value of catching NULL and only NULL?

If you can come up with a solution to handle invalid pointers in the VDSO
gracefully, then I'm surely all ears.

If not then can you please explain why VDSO usage worked without pointer
validation since the time VDSO was invented more than a decade ago and just
now requires full invalid pointer handling?

Thanks,

	tglx