[RFC PATCH] lib: Introduce generic __cmpxchg_u64() and use it where needed

[Peter Zijlstra]

On Thu, Nov 01, 2018 at 10:01:46AM -0700, Paul E. McKenney wrote:
> On Thu, Nov 01, 2018 at 05:32:12PM +0100, Peter Zijlstra wrote:
> > On Thu, Nov 01, 2018 at 03:22:15PM +0000, Trond Myklebust wrote:
> > > On Thu, 2018-11-01 at 15:59 +0100, Peter Zijlstra wrote:
> > > > On Thu, Nov 01, 2018 at 01:18:46PM +0000, Mark Rutland wrote:
> > 
> > > > > > My one question (and the reason why I went with cmpxchg() in the
> > > > > > first place) would be about the overflow behaviour for
> > > > > > atomic_fetch_inc() and friends. I believe those functions should
> > > > > > be OK on x86, so that when we overflow the counter, it behaves
> > > > > > like an unsigned value and wraps back around.  Is that the case
> > > > > > for all architectures?
> > > > > > 
> > > > > > i.e. are atomic_t/atomic64_t always guaranteed to behave like
> > > > > > u32/u64 on increment?
> > > > > > 
> > > > > > I could not find any documentation that explicitly stated that
> > > > > > they should.
> > > > > 
> > > > > Peter, Will, I understand that the atomic_t/atomic64_t ops are
> > > > > required to wrap per 2's-complement. IIUC the refcount code relies
> > > > > on this.
> > > > > 
> > > > > Can you confirm?
> > > > 
> > > > There is quite a bit of core code that hard assumes 2s-complement.
> > > > Not only for atomics but for any signed integer type. Also see the
> > > > kernel using -fno-strict-overflow which implies -fwrapv, which
> > > > defines signed overflow to behave like 2s-complement (and rids us of
> > > > that particular UB).
> > > 
> > > Fair enough, but there have also been bugfixes to explicitly fix unsafe
> > > C standards assumptions for signed integers. See, for instance commit
> > > 5a581b367b5d "jiffies: Avoid undefined behavior from signed overflow"
> > > from Paul McKenney.
> > 
> > Yes, I feel Paul has been to too many C/C++ committee meetings and got
> > properly paranoid. Which isn't always a bad thing :-)
> 
> Even the C standard defines 2s complement for atomics.  

Ooh good to know.

> Just not for
> normal arithmetic, where yes, signed overflow is UB.  And yes, I do
> know about -fwrapv, but I would like to avoid at least some copy-pasta
> UB from my kernel code to who knows what user-mode environment.  :-/
> 
> At least where it is reasonably easy to do so.

Fair enough I suppose; I just always make sure to include the same
-fknobs for the userspace thing when I lift code.

> And there is a push to define C++ signed arithmetic as 2s complement,
> but there are still 1s complement systems with C compilers.  Just not
> C++ compilers.  Legacy...

*groan*; how about those ancient hardwares keep using ancient compilers
and we all move on to the 70s :-)

> > But for us using -fno-strict-overflow which actually defines signed
> > overflow, I myself am really not worried. I'm also not sure if KASAN has
> > been taught about this, or if it will still (incorrectly) warn about UB
> > for signed types.
> 
> UBSAN gave me a signed-overflow warning a few days ago.  Which I have
> fixed, even though 2s complement did the right thing.  I am also taking
> advantage of the change to use better naming.

Oh too many *SANs I suppose; and yes, if you can make the code better,
why not.

> > > Anyhow, if the atomic maintainers are willing to stand up and state for
> > > the record that the atomic counters are guaranteed to wrap modulo 2^n
> > > just like unsigned integers, then I'm happy to take Paul's patch.
> > 
> > I myself am certainly relying on it.
> 
> Color me confused.  My 5a581b367b5d is from 2013.  Or is "Paul" instead
> intended to mean Paul Mackerras, who happens to be on CC?

Paul Burton I think, on a part of the thread before we joined :-)