OffTarget

Observed issues › catalog/cce-reduction-inscan-ice

Appears to behave incorrectlyCompiler crash · Fortran · OpenMP · Fixed in 21.0.3

ftn: internal compiler error on reduction(inscan) with a scan directive

Legality: independently confirmed. 2 independent readings find the program legal and the result required. How this is checked

Affected variants

CompilerMI300A gfx942
unified memory (XNACK 1)
MI300A gfx942
copy (XNACK 0)
MI250X gfx90a
copy (XNACK 0)
MI300X gfx942
copy (XNACK 0)
MI350X gfx950
copy (XNACK 0)
A100 sm_80
separate memory
A100 sm_80
managed memory
AFAR 24.3.0fails another wayfails another wayfails another wayfails another wayfails another wayn/an/a
CCE 19.0.0behaves incorrectlybehaves incorrectlybehaves incorrectlyunknown
not run
unknown
not run
n/an/a
CCE 20.0.2behaves incorrectlybehaves incorrectlybehaves incorrectlyunknown
not run
unknown
not run
n/an/a
CCE 21.0.3correctcorrectcorrectunknown
not run
unknown
not run
n/an/a
CCE 21.0.4correctcorrectcorrectunknown
not run
unknown
not run
n/an/a
NVHPC 25.11n/an/an/an/an/afails another wayfails another way
NVHPC 26.9n/an/an/an/an/afails another wayfails another way
GCC 16.2.0correctcorrectcorrectcorrectcorrectcorrectn/a

"behaves incorrectly" / "not supported": this issue, confirmed in up to four environments (another build directory, another node, a repeat). "fails another way": a different failure, not counted as this issue. "unknown": not run there, or the runs disagree.

Messages

ResultCompiler and GPUMessage
behaves incorrectlyCCE 19.0.0, MI300A unified
CCE 19.0.0, MI300A copy
CCE 19.0.0, MI250X
show
Warning:
 Headers and libraries from cray-libsci/25.03.0 will be ignored because depends on PE_MPICH which is unavailable..
   Error message      ::  analyze_object: unknown object type
   Error detected     ::  File '/workspace/crayftn/pdgcs/v_alias_new.c', line 1357
   Initiated from     ::  Line 858 (v_main.c)
   Optimizer built    ::  2025-02-07 (production)
   File               ::  cce-reduction-inscan-ice.f90
   Function           ::  gen_prog
behaves incorrectlyCCE 20.0.2, MI300A unified
CCE 20.0.2, MI300A copy
CCE 20.0.2, MI250X
show
Warning:
 Headers and libraries from cray-libsci/25.09.0 will be ignored because depends on PE_MPICH which is unavailable..
   Error message      ::  analyze_object: unknown object type
   Error detected     ::  File '/workspace/crayftn/pdgcs/v_alias_new.c', line 1358
   Initiated from     ::  Line 858 (v_main.c)
   Optimizer built    ::  2025-11-17 (production)
   File               ::  cce-reduction-inscan-ice.f90
   Function           ::  gen_prog
fails another wayAFAR 24.3.0, MI300A unified
AFAR 24.3.0, MI300A copy
AFAR 24.3.0, MI250X
AFAR 24.3.0, MI300X
AFAR 24.3.0, MI350X
show
loc("cce-reduction-inscan-ice.f90":20:11): error: not yet implemented: Unhandled clause reduction with modifier in omp.wsloop operation
loc("cce-reduction-inscan-ice.f90":20:11): error: LLVM Translation failed for operation: omp.wsloop
loc("cce-reduction-inscan-ice.f90":20:11): error: LLVM Translation failed for operation: omp.parallel
loc("cce-reduction-inscan-ice.f90":18:9): error: LLVM Translation failed for operation: omp.target
error: failed to create the LLVM module
fails another wayNVHPC 25.11, A100 separate
NVHPC 25.11, A100 managed
NVHPC 26.9, A100 separate
NVHPC 26.9, A100 managed
show
NVFORTRAN-S-0034-Syntax error at or near + (cce-reduction-inscan-ice.f90: 20)
NVFORTRAN-W-0287-Unrecognized OpenMP directive - scan (cce-reduction-inscan-ice.f90: 23)
  0 inform,   1 warnings,   1 severes, 0 fatal for gen_prog

Reproduce

CCE 19.0.0, MI300A unified

module load cce/19.0.0 craype-accel-amd-gfx942 rocm/6.4.3
module unload cray-mpich
ftn -fopenmp -O2 repro.f90 -o repro
HSA_XNACK=1 OMP_TARGET_OFFLOAD=mandatory ./repro

CCE 20.0.2, MI300A unified

module load cce/20.0.2 craype-accel-amd-gfx942 rocm/6.4.3
module unload cray-mpich
ftn -fopenmp -O2 repro.f90 -o repro
HSA_XNACK=1 OMP_TARGET_OFFLOAD=mandatory ./repro

GCC 16.2.0, MI300A unified (runs correctly)

export LD_LIBRARY_PATH=$GCC_16_2/lib64:/opt/rocm-7.2.1/lib
$GCC_16_2/bin/gfortran -fopenmp -foffload=amdgcn-amdhsa -foffload-options=amdgcn-amdhsa=-march=gfx942 -O2 repro.f90 -o repro
HSA_XNACK=1 OMP_TARGET_OFFLOAD=mandatory ./repro
Program (61 lines)
! # Draft finding: omp52-clause-reduction-inscan on cce19
!
! - compiler: cce19
! - toolchain: cce19; arch: gfx942; machine: tuolumne; memory mode: unified; opt: -O2
! - program: catalog omp52-clause-reduction-inscan (f90, version 52, sha256 f26bdfc791de283c)
! - expected: None (None; )
! - actual: ice  (ice: INTERNAL COMPILER ERROR:  "analyze_object: unknown object type" (/workspace/crayftn/pdgcs/v_alias_new.c, line
!   1357, version cc4d36e4ff3377d45f0e6e892b5dacd82009a0ca))
! - verdict: compiler_bug by rule R10-compile-ice: an internal compiler error (or the compiler rejecting code it generated) on a
!   legal program: always the compiler's defect
! - reason: 
! - citations: RUL-COMPILE-TIME (rulings.md)
! - legality: legal (legal by construction: offload.engine.gen / c_gen / acc_gen check every program against the catalog rules
!   and the
!   offload scope as it is built (no compiler))
! - witnesses: none
! - withdrawn: none
!
! Generated by compliance.gen (OpenMP 5.2, free-form Fortran). Do not edit.
! id: omp52-clause-reduction-inscan
! primary atom: omp.clause.reduction.inscan
! legal: yes
! observable: yes
! ran_on_host: inside the target region of interest
! evidence: construct
! needs: none
! std: f2008
! derivation (check=pass): reduction(inscan, +: s) with scan inclusive(s) (5.2 §5.6 p.141): the statement before the scan directive
! derivation: is the input phase, the one after it the scan phase, which sees the sum of the updates of every iteration up to and
! derivation: including its own; s starts at 0, so b(i) = a(1) + ... + a(i) = i*(i + 1)/2 = ref(i); the check accepts
program gen_prog
  use omp_lib, only: omp_is_initial_device
  implicit none
  integer, parameter :: n = 64
  integer, allocatable, target :: a(:), b(:)
  integer :: ref(n)
  integer :: i, host_flag
  logical :: ok
  integer :: s
  allocate(a(n), b(n))
  do i = 1, n
    a(i) = i
    b(i) = 0
    ref(i) = i*(i + 1)/2
  end do
  host_flag = 1
  s = 0
  !$omp target map(to: a) map(from: b) map(tofrom: s) map(tofrom: host_flag)
    host_flag = merge(1, 0, omp_is_initial_device())
    !$omp parallel do default(none) shared(a, b) reduction(inscan, +: s)
      do i = 1, n
        s = s + a(i)
        !$omp scan inclusive(s)
        b(i) = s
      end do
    !$omp end parallel do
  !$omp end target
  ok = all(b == ref)
  print '(a,i0,a,a)', 'RESULT ran_on_host=', host_flag, ' check=', merge('pass', 'fail', ok)
  deallocate(a, b)
end program gen_prog

Source in the repository

Something to add? Dispute its legality · Report it fixed in a newer release · Report a mistake on this page

Expected and observed

Observed where it behaves incorrectly

Warning:
 Headers and libraries from cray-libsci/25.03.0 will be ignored because depends on PE_MPICH which is unavailable..
   Error message      ::  analyze_object: unknown object type
   Error detected     ::  File '/workspace/crayftn/pdgcs/v_alias_new.c', line 1357
   Initiated from     ::  Line 858 (v_main.c)
   Optimizer built    ::  2025-02-07 (production)
   File               ::  cce-reduction-inscan-ice.f90
   Function           ::  gen_prog

What the specification requires

The specification text this program relies on and the compiler is expected to support.

OpenMP 5.2 §5.6 p.141

Evidence

Same compiler, one change

  • the same scan on the host (the target directive removed): AFAR 24.3.0: fails another way; CCE 19.0.0: correct; CCE 20.0.2: correct; CCE 21.0.3: correct; CCE 21.0.4: correct
    Control program
    program gen_prog
      use omp_lib, only: omp_is_initial_device
      implicit none
      integer, parameter :: n = 64
      integer, allocatable, target :: a(:), b(:)
      integer :: ref(n)
      integer :: i, host_flag
      logical :: ok
      integer :: s
      allocate(a(n), b(n))
      do i = 1, n
        a(i) = i
        b(i) = 0
        ref(i) = i*(i + 1)/2
      end do
      host_flag = 1
      s = 0
        host_flag = merge(1, 0, omp_is_initial_device())
        !$omp parallel do default(none) shared(a, b) reduction(inscan, +: s)
          do i = 1, n
            s = s + a(i)
            !$omp scan inclusive(s)
            b(i) = s
          end do
        !$omp end parallel do
      ok = all(b == ref)
      print '(a,i0,a,a)', 'RESULT ran_on_host=', host_flag, ' check=', merge('pass', 'fail', ok)
      deallocate(a, b)
    end program gen_prog
    

Other compilers run it correctly

GCC 16.2.0