On Fri, 19 May 2017, Richard Biener wrote:

> On Fri, 19 May 2017, Marek Polacek wrote:
> 
> > On Fri, May 19, 2017 at 09:58:45AM +0200, Richard Biener wrote:
> > > On Fri, 19 May 2017, Marek Polacek wrote:
> > > 
> > > > extract_muldiv folds 
> > > > 
> > > >   (n * 10000 * z) * 50
> > > > 
> > > > to
> > > > 
> > > >   (n * 500000) * z
> > > > 
> > > > which is a wrong transformation to do, because it may introduce an 
> > > > overflow.
> > > > This resulted in a ubsan false positive.  So we should just disable this
> > > > folding altogether.  Does the approach I took make sense?

I think it's possible to keep this folding, note that it's valid to transform to

    (n * 1 * z) * 500000

(i.e. accumulate multiplications on the outermost factor)

> > > > 
> > > > Bootstrapped/regtested on x86_64-linux, ok for trunk?
> > > 
> > > Didn't dig very far to identify extract_muldiv, but I guess it's either
> > > of the following recursions that trigger?
> > > 
> > >       /* If we can extract our operation from the LHS, do so and return a
> > >          new operation.  Likewise for the RHS from a MULT_EXPR.  
> > > Otherwise,
> > >          do something only if the second operand is a constant.  */
> > >       if (same_p
> > >           && (t1 = extract_muldiv (op0, c, code, wide_type,
> > >                                    strict_overflow_p)) != 0)
> > >         return fold_build2 (tcode, ctype, fold_convert (ctype, t1),
> > >                             fold_convert (ctype, op1));
> > >       else if (tcode == MULT_EXPR && code == MULT_EXPR
> > >                && (t1 = extract_muldiv (op1, c, code, wide_type,
> > >                                         strict_overflow_p)) != 0)
> > >         return fold_build2 (tcode, ctype, fold_convert (ctype, op0),
> > >                             fold_convert (ctype, t1));
> > 
> > Exactly.  extract_muldiv first gets (n * 10000 * z) * 50 so it tries
> > to fold 50 with (subexpressions) of (n * 10000 * z).  So it then tries
> > (n * 10000) * 50, and then n * 50 and then 10000 * 50 which finally
> > works out, so it uses 50000 and removes the outermost multiplication.

so would it be possible to adjust things here to remove the innermost
multiplication instead?

Alexander

Reply via email to