Re: [PATCH] locking/atomics: don't alias ____ptr

[Mark Rutland]

On Wed, Jun 28, 2017 at 01:21:43PM +0200, Thomas Gleixner wrote:
> On Wed, 28 Jun 2017, Dmitry Vyukov wrote:
> > On Wed, Jun 28, 2017 at 1:10 PM, Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:
> > >> In my case I ended up with something like:
> > >>
> > >> __typeof__(foo) __ptr = __ptr;
> > >>
> > >> which compiler decided to turn into 0.
> > >>
> > >> Thank you, macros.
> > >>
> > >> We can add more underscores, but the problem can happen again. Should
> > >> we prefix current function/macro name to all local vars?..
> > >
> > > Actually we can void that ___ptr dance completely.
> > >
> > > Thanks,
> > >
> > >         tglx
> > >
> > > 8<--------------------
> > >
> > > --- a/include/asm-generic/atomic-instrumented.h
> > > +++ b/include/asm-generic/atomic-instrumented.h
> > > @@ -359,37 +359,32 @@ static __always_inline bool atomic64_add
> > >
> > >  #define cmpxchg(ptr, old, new)                         \
> > >  ({                                                     \
> > > -       __typeof__(ptr) ___ptr = (ptr);                 \
> > > -       kasan_check_write(___ptr, sizeof(*___ptr));     \
> > > +       kasan_check_write((ptr), sizeof(*(ptr)));       \
> > >         arch_cmpxchg((ptr), (old), (new));              \
> > >  })
> > >
> > >  #define sync_cmpxchg(ptr, old, new)                    \
> > >  ({                                                     \
> > > -       __typeof__(ptr) ___ptr = (ptr);                 \
> > > -       kasan_check_write(___ptr, sizeof(*___ptr));     \
> > > -       arch_sync_cmpxchg(___ptr, (old), (new));        \
> > > +       kasan_check_write((ptr), sizeof(*(ptr)));       \
> > > +       arch_sync_cmpxchg((ptr), (old), (new));         \
> > >  })
> > >
> > >  #define cmpxchg_local(ptr, old, new)                   \
> > >  ({                                                     \
> > > -       __typeof__(ptr) ____ptr = (ptr);                \
> > > -       kasan_check_write(____ptr, sizeof(*____ptr));   \
> > > -       arch_cmpxchg_local(____ptr, (old), (new));      \
> > > +       kasan_check_write((ptr), sizeof(*(ptr)));       \
> > > +       arch_cmpxchg_local((ptr), (old), (new));        \
> > 
> > 
> > /\/\/\/\/\/\/\/\/\/\/\/\
> > 
> > These are macros.
> > If ptr is foo(), then we will call foo() twice.
> 
> Sigh, is that actually used?

For better or worse, we can't rule it out.

We'd risk even more subtle bugs in future trying to rely on that not
being the case. :/

> That's all insane. The whole crap gets worse because:
> 
>       cmpxchg() can be used on u8, u16, u32 ....

Yup, that's the whole reason for the macro insanity in the fist place.

Anoother option is something like:

static inline unsigned long
cmpxchg_size(unsigned long *ptr, unsigned long old, unsigned long new, int size)
{
	kasan_check_write(ptr, size);

	switch (size) {
	case 1:
		return arch_cmpxchg((u8 *)ptr, (u8)old, (u8)new);
	case 2:
		return arch_cmpxchg((u16 *)ptr, (u16)old, (u16)new);
	case 4:
		return arch_cmpxchg((u32 *)ptr, (u32)old, (u32)new);
	case 8:
		return arch_cmpxchg((u64 *)ptr, (u64)old, (u64)new);
	}

	BUILD_BUG();
}

#define cmpxchg(ptr, old, new)	\
	cmpxchg_size(ptr, old, new, sizeof(*ptr))

Thanks,
Mark.