[flang] MOD 7x performance regression

[vzakhari]

Polyhedron/ac slowed down by 3.7x after #77167
_FortranAModReal8 itself slowed down by 7.4x

I wonder if it can use the previous approach for not so different a and p.

[llvmbot]

@llvm/issue-subscribers-flang-runtime

Author: Slava Zakharin (vzakhari)

Polyhedron/ac slowed down by 3.7x after https://github.com//pull/77167 `_FortranAModReal8` itself slowed down by 7.4x

I wonder if it can use the previous approach for not so different a and p.

[vzakhari]

Another observation is that in the ac benchmark both values seem to be integers.

[klausler]

Can you give me a link to the exact call site in the source that you believe is passing integer arguments?

[vzakhari]

There is a single call site in file ac.f90:

      REAL FUNCTION GGL(Ds)

!     +     D1,
      DOUBLE PRECISION Ds , d2
!      DATA D1/2147483648.D0/
      DATA d2/2147483647.D0/
      Ds = DMOD(16807.D0*Ds,d2)
!     Generate U(0,1] distributed random numbers:
      GGL = Ds/d2

      CONTINUE
      END

[klausler]

#78745 should improve things (it's faster in my experiments) but may not be fast enough; please let me know.