Re: [PATCH] kcsan: Add option to allow watcher interruptions

[peterz]

On Sat, Jul 25, 2020 at 03:07:50PM -0700, Paul E. McKenney wrote:
> On Sat, Jul 25, 2020 at 10:21:31PM +0200, peterz@xxxxxxxxxxxxx wrote:
> > On Sat, Jul 25, 2020 at 10:10:13PM +0200, peterz@xxxxxxxxxxxxx wrote:
> > > On Sat, Jul 25, 2020 at 12:39:09PM -0700, Paul E. McKenney wrote:
> > 
> > > > This gets me the following for __rcu_read_lock():
> > > > 
> > > > 00000000000000e0 <__rcu_read_lock>:
> > > >       e0:	48 8b 14 25 00 00 00 	mov    0x0,%rdx
> > > >       e7:	00 
> > > >       e8:	8b 82 e0 02 00 00    	mov    0x2e0(%rdx),%eax
> > > >       ee:	83 c0 01             	add    $0x1,%eax
> > > >       f1:	89 82 e0 02 00 00    	mov    %eax,0x2e0(%rdx)
> > > >       f7:	c3                   	retq   
> > > >       f8:	0f 1f 84 00 00 00 00 	nopl   0x0(%rax,%rax,1)
> > > >       ff:	00 
> > > > 
> > > > One might hope for a dec instruction, but this isn't bad.  We do lose
> > > > a few instructions compared to the C-language case due to differences
> > > > in address calculation:
> > > > 
> > > > 00000000000000e0 <__rcu_read_lock>:
> > > >       e0:	48 8b 04 25 00 00 00 	mov    0x0,%rax
> > > >       e7:	00 
> > > >       e8:	83 80 e0 02 00 00 01 	addl   $0x1,0x2e0(%rax)
> > > >       ef:	c3                   	retq   
> > > 
> > > Shees, that's daft... I think this is one of the cases where GCC is
> > > perhaps overly cautious when presented with 'volatile'.
> > > 
> > > It has a history of generating excessively crap code around volatile,
> > > and while it has improved somewhat, this seems to show there's still
> > > room for improvement...
> > > 
> > > I suppose this is the point where we go bug a friendly compiler person.
> 
> Sounds very good!  Do you have someone specific in mind?

Jakub perhaps?, Cc'ed

> > Having had a play with godbolt.org, it seems clang isn't affected by
> > this particular flavour of crazy, but GCC does indeed refuse to fuse the
> > address calculation and the addition.
> 
> So there is hope, then!
> 
> And even GCC's current state is an improvement.  Last I messed with this,
> the ACCESS_ONCE()++ approach generated a load, a register increment,
> and a store.
> 
> Do you still have the godbolt.org URLs?  I would be happy to file
> a bugzilla.

https://godbolt.org/z/rP8rYM