[llvm-branch-commits] [llvm] [mlir] [OpenMP][MLIR] Descriptor explicit member map lowering changes (PR #113556)

via llvm-branch-commits llvm-branch-commits at lists.llvm.org
Thu Nov 7 20:24:43 PST 2024


https://github.com/agozillon updated https://github.com/llvm/llvm-project/pull/113556

>From 5b3cfc1554da9e32badbf5525b163698d977087d Mon Sep 17 00:00:00 2001
From: agozillon <Andrew.Gozillon at amd.com>
Date: Fri, 4 Oct 2024 12:59:06 -0500
Subject: [PATCH 1/2] [OpenMP] Allocatable explicit member mapping fortran
 offloading tests

This PR is one in a series of 3 that aim to add support for explicit member mapping of
allocatable components in derived types within OpenMP+Fortran for Flang.

This PR provides all of the runtime tests that are currently upstreamable, unfortunately
some of the other tests would require linking of the fortran runtime for offload which
we currently do not do. But regardless, this is plenty to ensure that the mapping is
working in most cases.
---
 .../basic-target-region-1D-array-section.f90  |   4 +-
 .../basic-target-region-3D-array-section.f90  |   4 +-
 .../fortran/basic-target-region-3D-array.f90  |   4 +-
 .../offloading/fortran/constant-arr-index.f90 |   6 +-
 .../declare-target-vars-in-target-region.f90  |   6 +-
 ...double-target-call-with-declare-target.f90 |   7 +-
 .../dtype-array-constant-index-map.f90        |  20 +-
 .../fortran/dtype-member-map-syntax-1.f90     | 109 +++++++++++
 .../fortran/dtype-member-map-syntax-2.f90     | 176 ++++++++++++++++++
 .../offloading/fortran/dump_map_tables.f90    |   6 +-
 .../fortran/local-descriptor-map-regress.f90  |  24 +--
 .../test/offloading/fortran/target-depend.f90 |   3 +-
 .../target-map-all-common-block-members.f90   |   5 +-
 ...map-alloca-dtype-alloca-array-of-dtype.f90 |  36 ++++
 .../target-map-alloca-dtype-alloca-array.f90  |  29 +++
 ...t-map-alloca-dtype-and-alloca-array-v2.f90 |  34 ++++
 ...rget-map-alloca-dtype-and-alloca-array.f90 |  33 ++++
 ...rget-map-alloca-dtype-array-and-scalar.f90 |  39 ++++
 ...target-map-alloca-dtype-array-of-dtype.f90 |  35 ++++
 ...ap-allocatable-array-section-1d-bounds.f90 |   5 +-
 ...ap-allocatable-array-section-3d-bounds.f90 |   5 +-
 .../fortran/target-map-allocatable-dtype.f90  |  31 +++
 .../target-map-allocatable-map-scopes.f90     |   4 +-
 .../fortran/target-map-common-block.f90       |   5 +-
 ...t-map-declare-target-link-common-block.f90 |   6 +-
 .../target-map-derived-type-full-1.f90        |   6 +-
 .../target-map-derived-type-full-2.f90        |   5 +-
 ...arget-map-derived-type-full-implicit-1.f90 |   6 +-
 ...arget-map-derived-type-full-implicit-2.f90 |   6 +-
 ...double-large-nested-dtype-multi-member.f90 |   5 +-
 ...t-map-double-nested-dtype-array-bounds.f90 |   5 +-
 ...ouble-nested-dtype-double-array-bounds.f90 |   5 +-
 ...-map-double-nested-dtype-single-member.f90 |   5 +-
 ...-map-dtype-3d-alloca-array-with-bounds.f90 |  41 ++++
 ...-map-dtype-alloca-and-non-alloca-array.f90 |  36 ++++
 ...type-alloca-array-and-non-alloca-dtype.f90 |  44 +++++
 ...target-map-dtype-alloca-array-of-dtype.f90 |  34 ++++
 ...get-map-dtype-alloca-array-with-bounds.f90 |  29 +++
 .../target-map-dtype-allocatable-array.f90    |  28 +++
 ...map-dtype-allocatable-scalar-and-array.f90 |  34 ++++
 ...pe-arr-bounds-member-enter-exit-update.f90 |   8 +-
 ...map-dtype-arr-bounds-member-enter-exit.f90 |   8 +-
 ...dtype-explicit-individual-array-member.f90 |   5 +-
 ...-multi-explicit-array-3D-member-bounds.f90 |   6 +-
 ...ype-multi-explicit-array-member-bounds.f90 |   6 +-
 ...-map-dtype-multi-explicit-array-member.f90 |   5 +-
 ...target-map-dtype-multi-explicit-member.f90 |   5 +-
 .../target-map-enter-exit-allocatables.f90    |   4 +-
 .../fortran/target-map-enter-exit-array-2.f90 |   4 +-
 .../target-map-enter-exit-array-bounds.f90    |   5 +-
 .../fortran/target-map-enter-exit-array.f90   |   4 +-
 .../fortran/target-map-enter-exit-scalar.f90  |   3 +-
 .../target-map-first-common-block-member.f90  |   5 +-
 ...target-map-individual-dtype-member-map.f90 |   5 +-
 ...et-map-large-nested-dtype-multi-member.f90 |   5 +-
 ...t-map-mix-imp-exp-common-block-members.f90 |   6 +-
 ...lloca-dtypes-with-multi-alloca-members.f90 |  90 +++++++++
 ...alloca-dtypes-with-multi-mixed-members.f90 |  82 ++++++++
 ...ed-alloca-dtype-3d-alloca-array-bounds.f90 |  51 +++++
 ...ested-alloca-dtype-alloca-array-bounds.f90 |  43 +++++
 ...sted-dtype-3d-alloca-array-with-bounds.f90 |  50 +++++
 ...sted-dtype-alloca-and-non-alloca-array.f90 |  47 +++++
 ...type-alloca-array-and-non-alloca-dtype.f90 |  51 +++++
 ...-nested-dtype-alloca-array-with-bounds.f90 |  43 +++++
 .../target-map-nested-dtype-alloca-array.f90  |  37 ++++
 ...target-map-nested-dtype-complex-member.f90 |   5 +-
 ...target-map-nested-dtype-derived-member.f90 |   6 +-
 .../target-map-nested-dtype-multi-member.f90  |   5 +-
 .../target-map-nested-dtype-single-member.f90 |   5 +-
 .../target-map-pointer-scopes-enter-exit.f90  |   4 +-
 ...pointer-target-array-section-3d-bounds.f90 |   5 +-
 .../target-map-pointer-target-scopes.f90      |   4 +-
 ...ap-pointer-to-dtype-allocatable-member.f90 |  45 +++++
 .../target-map-second-common-block-member.f90 |   5 +-
 .../target-map-two-dtype-explicit-member.f90  |   5 +-
 ...type-individual-member-array-1D-bounds.f90 |   6 +-
 ...type-mixed-implicit-explicit-capture-1.f90 |   5 +-
 ...type-mixed-implicit-explicit-capture-2.f90 |   6 +-
 ...two-dtype-multi-member-array-1D-bounds.f90 |   6 +-
 ...-map-two-nested-dtype-member-array-map.f90 |   4 +-
 .../fortran/target-nested-target-data.f90     |   3 +-
 .../fortran/target-region-implicit-array.f90  |   4 +-
 82 files changed, 1433 insertions(+), 193 deletions(-)
 create mode 100644 offload/test/offloading/fortran/dtype-member-map-syntax-1.f90
 create mode 100644 offload/test/offloading/fortran/dtype-member-map-syntax-2.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array-of-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array-v2.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-array-and-scalar.f90
 create mode 100644 offload/test/offloading/fortran/target-map-alloca-dtype-array-of-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-allocatable-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-3d-alloca-array-with-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-alloca-and-non-alloca-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-alloca-array-and-non-alloca-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-alloca-array-of-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-alloca-array-with-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-allocatable-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-dtype-allocatable-scalar-and-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-alloca-members.f90
 create mode 100644 offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-mixed-members.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-alloca-dtype-3d-alloca-array-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-alloca-dtype-alloca-array-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-dtype-3d-alloca-array-with-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-dtype-alloca-and-non-alloca-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-and-non-alloca-dtype.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-with-bounds.f90
 create mode 100644 offload/test/offloading/fortran/target-map-nested-dtype-alloca-array.f90
 create mode 100644 offload/test/offloading/fortran/target-map-pointer-to-dtype-allocatable-member.f90

diff --git a/offload/test/offloading/fortran/basic-target-region-1D-array-section.f90 b/offload/test/offloading/fortran/basic-target-region-1D-array-section.f90
index 15795fbcf7f354..61244a965a1b88 100644
--- a/offload/test/offloading/fortran/basic-target-region-1D-array-section.f90
+++ b/offload/test/offloading/fortran/basic-target-region-1D-array-section.f90
@@ -1,5 +1,5 @@
-! Basic offloading test of arrays with provided lower
-! and upper bounds as specified by OpenMP's sectioning
+! Basic offloading test of arrays with provided lower and upper bounds as
+! specified by OpenMP's sectioning
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/basic-target-region-3D-array-section.f90 b/offload/test/offloading/fortran/basic-target-region-3D-array-section.f90
index ca0364543c20f8..2e1c6a61379c3f 100644
--- a/offload/test/offloading/fortran/basic-target-region-3D-array-section.f90
+++ b/offload/test/offloading/fortran/basic-target-region-3D-array-section.f90
@@ -1,5 +1,5 @@
-! Basic offloading test of a regular array explicitly
-! passed within a target region
+! Basic offloading test of a regular array explicitly passed within a target
+! region
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/basic-target-region-3D-array.f90 b/offload/test/offloading/fortran/basic-target-region-3D-array.f90
index d575810713895e..61fced35980f41 100644
--- a/offload/test/offloading/fortran/basic-target-region-3D-array.f90
+++ b/offload/test/offloading/fortran/basic-target-region-3D-array.f90
@@ -1,5 +1,5 @@
-! Basic offloading test of a regular array explicitly
-! passed within a target region
+! Basic offloading test of a regular array explicitly passed within a target
+! region
 ! REQUIRES: flang
 ! REQUIRES: gpu
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/constant-arr-index.f90 b/offload/test/offloading/fortran/constant-arr-index.f90
index 80440b6836ac8a..f819f1f0d994df 100644
--- a/offload/test/offloading/fortran/constant-arr-index.f90
+++ b/offload/test/offloading/fortran/constant-arr-index.f90
@@ -1,7 +1,5 @@
-! Basic offloading test with a target region
-! that checks constant indexing on device
-! correctly works (regression test for prior
-! bug).
+! Basic offloading test with a target region that checks constant indexing on
+! device correctly works (regression test for prior bug).
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/declare-target-vars-in-target-region.f90 b/offload/test/offloading/fortran/declare-target-vars-in-target-region.f90
index c0bc9304b3030d..5410a0681e8e2a 100644
--- a/offload/test/offloading/fortran/declare-target-vars-in-target-region.f90
+++ b/offload/test/offloading/fortran/declare-target-vars-in-target-region.f90
@@ -1,6 +1,6 @@
-! Offloading test with a target region mapping a declare target
-! Fortran array writing some values to it and checking the host
-! correctly receives the updates made on the device.
+! Offloading test with a target region mapping a declare target Fortran array
+! writing some values to it and checking the host correctly receives the
+! updates made on the device.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/double-target-call-with-declare-target.f90 b/offload/test/offloading/fortran/double-target-call-with-declare-target.f90
index fa909815757f92..f7fef9583214bd 100644
--- a/offload/test/offloading/fortran/double-target-call-with-declare-target.f90
+++ b/offload/test/offloading/fortran/double-target-call-with-declare-target.f90
@@ -1,7 +1,6 @@
-! Offloading test with two target regions mapping the same
-! declare target Fortran array and writing some values to
-! it before checking the host correctly receives the
-! correct updates made on the device.
+! Offloading test with two target regions mapping the same declare target
+! Fortran array and writing some values to it before checking the host
+! correctly receives the correct updates made on the device.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/dtype-array-constant-index-map.f90 b/offload/test/offloading/fortran/dtype-array-constant-index-map.f90
index 7e168b846f858c..e120c95019c602 100644
--- a/offload/test/offloading/fortran/dtype-array-constant-index-map.f90
+++ b/offload/test/offloading/fortran/dtype-array-constant-index-map.f90
@@ -1,17 +1,9 @@
-! Offloading test which maps a specific element of a
-! derived type to the device and then accesses the
-! element alongside an individual element of an array
-! that the derived type contains. In particular, this
-! test helps to check that we can replace the constants
-! within the kernel with instructions and then replace
-! these instructions with the kernel parameters.
-! REQUIRES: flang
-! UNSUPPORTED: nvptx64-nvidia-cuda
-! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
-! UNSUPPORTED: aarch64-unknown-linux-gnu
-! UNSUPPORTED: aarch64-unknown-linux-gnu-LTO
-! UNSUPPORTED: x86_64-unknown-linux-gnu
-! UNSUPPORTED: x86_64-unknown-linux-gnu-LTO
+! Offloading test which maps a specific element of a derived type to the device
+! and then accesses the element alongside an individual element of an array
+! that the derived type contains. In particular, this test helps to check that
+! we can replace the constants within the kernel with instructions and then
+! replace these instructions with the kernel parameters.
+! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
 module test_0
diff --git a/offload/test/offloading/fortran/dtype-member-map-syntax-1.f90 b/offload/test/offloading/fortran/dtype-member-map-syntax-1.f90
new file mode 100644
index 00000000000000..fa4f20149c017e
--- /dev/null
+++ b/offload/test/offloading/fortran/dtype-member-map-syntax-1.f90
@@ -0,0 +1,109 @@
+! This test checks a number of more complex derived type member mapping
+! syntaxes utilising a non-allocatable parent derived type.
+
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type dtype2
+    integer int
+    real float
+    real float_elements(10)
+    end type dtype2
+
+    type dtype1
+    character (LEN=30) characters
+    character (LEN=1) character
+    type(dtype2) number
+    end type dtype1
+
+    type nonallocatabledtype
+    integer elements(20)
+    type(dtype1) num_chars
+    integer value
+    type(dtype2) internal_dtypes(5)
+    end type nonallocatabledtype
+
+    type (nonallocatabledtype) array_dtype(5)
+
+  !$omp target map(tofrom: array_dtype(5))
+      do i = 1, 20
+        array_dtype(5)%elements(i) = 20 + i
+      end do
+
+      array_dtype(5)%num_chars%number%float_elements(5) = 10
+      array_dtype(5)%value = 40
+  !$omp end target
+
+    print *, array_dtype(5)%elements
+    print *, array_dtype(5)%num_chars%number%float_elements(5)
+    print *, array_dtype(5)%value
+
+  !$omp target map(tofrom: array_dtype(4)%elements(3))
+    array_dtype(4)%elements(3) = 74
+  !$omp end target
+
+   print *, array_dtype(4)%elements(3)
+
+  !$omp target map(tofrom: array_dtype(5)%elements(3:5))
+    do i = 3, 5
+       array_dtype(5)%elements(i) = i + 1
+    end do
+  !$omp end target
+
+   do i = 3, 5
+      print *, array_dtype(5)%elements(i)
+   end do
+
+  !$omp target map(tofrom: array_dtype(3:5))
+    do i = 3, 5
+      array_dtype(i)%value = i + 2
+    end do
+  !$omp end target
+
+    do i = 3, 5
+        print *, array_dtype(i)%value
+    end do
+
+  !$omp target map(tofrom: array_dtype(4)%num_chars%number%float_elements(8))
+    array_dtype(4)%num_chars%number%float_elements(8) = 250
+  !$omp end target
+
+  print *, array_dtype(4)%num_chars%number%float_elements(8)
+
+  !$omp target map(tofrom: array_dtype(4)%num_chars%number%float_elements(5:10))
+    do i = 5, 10
+      array_dtype(4)%num_chars%number%float_elements(i) = i + 3
+    end do
+  !$omp end target
+
+  do i = 5, 10
+    print *, array_dtype(4)%num_chars%number%float_elements(i)
+  end do
+
+  !$omp target map(tofrom: array_dtype(4)%internal_dtypes(3)%float_elements(4))
+    array_dtype(4)%internal_dtypes(3)%float_elements(4) = 200
+  !$omp end target
+
+  print *, array_dtype(4)%internal_dtypes(3)%float_elements(4)
+
+end program main
+
+! CHECK: 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
+! CHECK: 10.
+! CHECK: 40
+! CHECK: 74
+! CHECK: 4
+! CHECK: 5
+! CHECK: 6
+! CHECK: 5
+! CHECK: 6
+! CHECK: 7
+! CHECK: 250.
+! CHECK: 8.
+! CHECK: 9.
+! CHECK: 10.
+! CHECK: 11.
+! CHECK: 12.
+! CHECK: 13.
+! CHECK: 200
diff --git a/offload/test/offloading/fortran/dtype-member-map-syntax-2.f90 b/offload/test/offloading/fortran/dtype-member-map-syntax-2.f90
new file mode 100644
index 00000000000000..8d666360066d9b
--- /dev/null
+++ b/offload/test/offloading/fortran/dtype-member-map-syntax-2.f90
@@ -0,0 +1,176 @@
+! This test checks a number of more complex derived type member mapping
+! syntaxes utilising an allocatable parent derived type and mixed
+! allocatable members.
+
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    implicit none
+
+    integer :: i
+    integer :: N1, N2
+
+    type :: vertexes
+        integer :: test
+        integer :: testarray(10)
+        integer(4), allocatable :: vertexx(:)
+        integer(4), allocatable :: vertexy(:)
+    end type vertexes
+
+    type testing_tile_type
+        TYPE(vertexes) :: field
+    end type testing_tile_type
+
+    type :: dtype
+        real(4) :: i
+        type(vertexes), allocatable :: vertexes(:)
+        TYPE(testing_tile_type), DIMENSION(:), allocatable :: test_tile
+        integer(4) :: array_i(10)
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype) :: alloca_dtype
+    type(dtype), DIMENSION(:), allocatable :: alloca_dtype_arr
+
+    allocate(alloca_dtype%vertexes(4))
+    allocate(alloca_dtype%vertexes(1)%vertexx(10))
+    allocate(alloca_dtype%vertexes(1)%vertexy(10))
+    allocate(alloca_dtype%vertexes(2)%vertexx(10))
+    allocate(alloca_dtype%vertexes(2)%vertexy(10))
+    allocate(alloca_dtype%vertexes(3)%vertexx(10))
+    allocate(alloca_dtype%vertexes(3)%vertexy(10))
+    allocate(alloca_dtype%vertexes(4)%vertexx(10))
+    allocate(alloca_dtype%vertexes(4)%vertexy(10))
+    allocate(alloca_dtype%test_tile(4))
+    allocate(alloca_dtype%test_tile(1)%field%vertexx(10))
+    allocate(alloca_dtype%test_tile(1)%field%vertexy(10))
+    allocate(alloca_dtype%test_tile(2)%field%vertexx(10))
+    allocate(alloca_dtype%test_tile(2)%field%vertexy(10))
+    allocate(alloca_dtype%test_tile(3)%field%vertexx(10))
+    allocate(alloca_dtype%test_tile(3)%field%vertexy(10))
+    allocate(alloca_dtype%test_tile(4)%field%vertexx(10))
+    allocate(alloca_dtype%test_tile(4)%field%vertexy(10))
+
+    allocate(alloca_dtype_arr(3))
+
+    N1 = 1
+    N2 = 2
+
+!$omp target map(tofrom: alloca_dtype%vertexes(N1)%test)
+        alloca_dtype%vertexes(N1)%test = 3
+!$omp end target
+
+print *, alloca_dtype%vertexes(N1)%test
+
+!$omp target map(tofrom: alloca_dtype%vertexes(N1)%test, alloca_dtype%vertexes(N2)%test)
+        alloca_dtype%vertexes(N1)%test = 5
+        alloca_dtype%vertexes(N2)%test = 10
+!$omp end target
+
+print *, alloca_dtype%vertexes(N1)%test
+print *, alloca_dtype%vertexes(N2)%test
+
+!$omp target map(tofrom: alloca_dtype%test_tile(N1)%field%vertexx, &
+!$omp                    alloca_dtype%test_tile(N1)%field%vertexy)
+    do i = 1, 10
+        alloca_dtype%test_tile(N1)%field%vertexx(i) = i + 4
+        alloca_dtype%test_tile(N1)%field%vertexy(i) = i + 4
+    end do
+!$omp end target
+
+print *, alloca_dtype%test_tile(N1)%field%vertexx
+print *, alloca_dtype%test_tile(N1)%field%vertexy
+
+!$omp target map(tofrom:  alloca_dtype%test_tile(N1)%field%test, &
+!$omp                     alloca_dtype%test_tile(N2)%field%test, &
+!$omp                     alloca_dtype%test_tile(N1)%field%vertexy, &
+!$omp                     alloca_dtype%test_tile(N2)%field%vertexy)
+    alloca_dtype%test_tile(N2)%field%test = 9999
+    alloca_dtype%test_tile(N2)%field%vertexy(2) = 9998
+    alloca_dtype%test_tile(N1)%field%test = 9997
+    alloca_dtype%test_tile(N1)%field%vertexy(2) = 9996
+!$omp end target
+
+print *, alloca_dtype%test_tile(N1)%field%test
+print *, alloca_dtype%test_tile(N2)%field%test
+print *, alloca_dtype%test_tile(N1)%field%vertexy(2)
+print *, alloca_dtype%test_tile(N2)%field%vertexy(2)
+
+!$omp target map(tofrom:  alloca_dtype%test_tile(N2)%field%vertexy)
+   alloca_dtype%test_tile(N2)%field%vertexy(2) = 2000
+!$omp end target
+
+!$omp target map(tofrom: alloca_dtype%vertexes(N1)%vertexx, &
+!$omp                    alloca_dtype%vertexes(N1)%vertexy, &
+!$omp                    alloca_dtype%vertexes(N2)%vertexx, &
+!$omp                    alloca_dtype%vertexes(N2)%vertexy)
+    do i = 1, 10
+        alloca_dtype%vertexes(N1)%vertexx(i) = i * 2
+        alloca_dtype%vertexes(N1)%vertexy(i) = i * 2
+        alloca_dtype%vertexes(N2)%vertexx(i) = i * 2
+        alloca_dtype%vertexes(N2)%vertexy(i) = i * 2
+    end do
+!$omp end target
+
+print *, alloca_dtype%vertexes(N1)%vertexx
+print *, alloca_dtype%vertexes(N1)%vertexy
+print *, alloca_dtype%vertexes(N2)%vertexx
+print *, alloca_dtype%vertexes(N2)%vertexy
+
+!$omp target map(tofrom: alloca_dtype%vertexes(N1)%vertexx, &
+!$omp                    alloca_dtype%vertexes(N1)%vertexy, &
+!$omp                    alloca_dtype%vertexes(4)%vertexy, &
+!$omp                    alloca_dtype%vertexes(4)%vertexx, &
+!$omp                    alloca_dtype%vertexes(N2)%vertexx, &
+!$omp                    alloca_dtype%vertexes(N2)%vertexy)
+    do i = 1, 10
+        alloca_dtype%vertexes(N1)%vertexx(i) = i * 3
+        alloca_dtype%vertexes(N1)%vertexy(i) = i * 3
+        alloca_dtype%vertexes(4)%vertexx(i) = i * 3
+        alloca_dtype%vertexes(4)%vertexy(i) = i * 3
+        alloca_dtype%vertexes(N2)%vertexx(i) = i * 3
+        alloca_dtype%vertexes(N2)%vertexy(i) = i * 3
+    end do
+!$omp end target
+
+
+    print *, alloca_dtype%vertexes(1)%vertexx
+    print *, alloca_dtype%vertexes(1)%vertexy
+    print *, alloca_dtype%vertexes(4)%vertexx
+    print *, alloca_dtype%vertexes(4)%vertexy
+    print *, alloca_dtype%vertexes(2)%vertexx
+    print *, alloca_dtype%vertexes(2)%vertexy
+
+!$omp target map(tofrom: alloca_dtype_arr(N2)%array_i)
+    do i = 1, 10
+        alloca_dtype_arr(N2)%array_i(i) = i + 2
+    end do
+!$omp end target
+
+print *, alloca_dtype_arr(N2)%array_i
+
+end program main
+
+! CHECK: 3
+! CHECK: 5
+! CHECK: 10
+! CHECK: 5 6 7 8 9 10 11 12 13 14
+! CHECK: 5 6 7 8 9 10 11 12 13 14
+! CHECK: 9997
+! CHECK: 9999
+! CHECK: 9996
+! CHECK: 9998
+! CHECK: 2 4 6 8 10 12 14 16 18 20
+! CHECK: 2 4 6 8 10 12 14 16 18 20
+! CHECK: 2 4 6 8 10 12 14 16 18 20
+! CHECK: 2 4 6 8 10 12 14 16 18 20
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 6 9 12 15 18 21 24 27 30
+! CHECK: 3 4 5 6 7 8 9 10 11 12
diff --git a/offload/test/offloading/fortran/dump_map_tables.f90 b/offload/test/offloading/fortran/dump_map_tables.f90
index e25b05b9e04db5..efde4ee56ca1a0 100644
--- a/offload/test/offloading/fortran/dump_map_tables.f90
+++ b/offload/test/offloading/fortran/dump_map_tables.f90
@@ -1,6 +1,6 @@
-! Offloading test with runtine call to ompx_dump_mapping_tables
-! Fortran array writing some values and printing the variable mapped to device
-! correctly receives the updates made on the device.
+! Offloading test with runtine call to ompx_dump_mapping_tables Fortran array
+! writing some values and printing the variable mapped to device correctly
+! receives the updates made on the device.
 ! REQUIRES: flang
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
 ! UNSUPPORTED: aarch64-unknown-linux-gnu
diff --git a/offload/test/offloading/fortran/local-descriptor-map-regress.f90 b/offload/test/offloading/fortran/local-descriptor-map-regress.f90
index 5f628b8ad8c7af..e6afc4a6fb9b1a 100644
--- a/offload/test/offloading/fortran/local-descriptor-map-regress.f90
+++ b/offload/test/offloading/fortran/local-descriptor-map-regress.f90
@@ -1,17 +1,13 @@
-! Small regression test that checks that we do not cause
-! a runtime map error in cases where we are required to
-! allocate a local variable for the fortran descriptor
-! to store into and then load from it, done so by
-! re-using the temporary local variable across all
-! maps related to the mapped variable and associated
-! local variable to make sure that each map does as
-! it's intended to do with the original data. This
-! prevents blobs of local descriptor data remaining
-! attatched on device long after it's supposed to,
-! which can cause weird map issues later in susbequent
-! function invocations. However, it doesn't avoid a user
-! shooting themselves in the foot by mapping data via enter
-! and then not providing a corresponding exit.
+! Small regression test that checks that we do not cause a runtime map error in
+! cases where we are required to allocate a local variable for the fortran
+! descriptor to store into and then load from it, done so by re-using the
+! temporary local variable across all maps related to the mapped variable and
+! associated local variable to make sure that each map does as it is intended
+! to do with the original data. This prevents blobs of local descriptor data
+! remaining attatched on device long after it's supposed to, which can cause
+! weird map issues later in susbequent function invocations. However, it
+! doesn't avoid a user shooting themselves in the foot by mapping data via
+! enter and then not providing a corresponding exit.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-depend.f90 b/offload/test/offloading/fortran/target-depend.f90
index 1bb320d6bd93c1..dc9351dd8c20af 100644
--- a/offload/test/offloading/fortran/target-depend.f90
+++ b/offload/test/offloading/fortran/target-depend.f90
@@ -1,5 +1,4 @@
-! Offloading test checking the use of the depend clause on
-! the target construct
+! Offloading test checking the use of the depend clause on the target construct
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-all-common-block-members.f90 b/offload/test/offloading/fortran/target-map-all-common-block-members.f90
index f92e5818a76405..4e78ed1accf3ed 100644
--- a/offload/test/offloading/fortran/target-map-all-common-block-members.f90
+++ b/offload/test/offloading/fortran/target-map-all-common-block-members.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of
-! mapping all the members of a common block
-! to a target region
+! Offloading test checking interaction of mapping all the members of a common
+! block to a target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array-of-dtype.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array-of-dtype.f90
new file mode 100644
index 00000000000000..c430726cfbe93a
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array-of-dtype.f90
@@ -0,0 +1,36 @@
+! Offloading test checking interaction of explicit member mapping of an
+! allocatable array of derived types contained within an allocatable
+! derived type
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: nested_dtype
+        real(4) :: i
+        real(4) :: j
+        integer(4) :: array_i(10)
+        integer(4) :: k
+    end type nested_dtype
+
+    type :: dtype
+        real(4) :: i
+        integer(4) :: array_i(10)
+        real(4) :: j
+        type(nested_dtype), allocatable :: array_dtype(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: dtyped
+    allocate(dtyped)
+    allocate(dtyped%array_dtype(10))
+
+!$omp target map(tofrom: dtyped%array_dtype)
+    do i = 1, 10
+        dtyped%array_dtype(i)%k = i
+    end do
+!$omp end target
+
+    print *, dtyped%array_dtype%k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array.f90
new file mode 100644
index 00000000000000..df51128fc5b7d8
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-alloca-array.f90
@@ -0,0 +1,29 @@
+! Offload test that checks an allocatable array within an allocatable derived
+! type can be mapped explicitly using member mapping.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: dtype
+        real(4) :: i
+        integer(4) :: array_i(10)
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: alloca_dtype
+    allocate(alloca_dtype)
+    allocate(alloca_dtype%array_j(10))
+
+!$omp target map(tofrom: alloca_dtype%array_j)
+    do i = 1, 10
+        alloca_dtype%array_j(i) = i
+    end do
+!$omp end target
+
+print *, alloca_dtype%array_j
+
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array-v2.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array-v2.f90
new file mode 100644
index 00000000000000..54a32160c7a61d
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array-v2.f90
@@ -0,0 +1,34 @@
+! Offload test that checks an allocatable derived type can be mapped alongside
+! one of its own allocatable components without disrupting either mapping,
+! different from original as the argument ordering is reversed (similar to C++
+! mapping of a struct and a pointer, in concept at least).
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: dtype
+        real(4) :: i
+        integer, allocatable :: scalar
+        integer(4) :: array_i(10)
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: alloca_dtype
+    allocate(alloca_dtype)
+    allocate(alloca_dtype%array_j(10))
+
+!$omp target map(tofrom: alloca_dtype%array_j, alloca_dtype)
+    do i = 1, 10
+        alloca_dtype%array_j(i) = i
+    end do
+    alloca_dtype%k = 50
+!$omp end target
+
+print *, alloca_dtype%array_j
+print *, alloca_dtype%k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 50
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array.f90
new file mode 100644
index 00000000000000..2b1957a8451c13
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-and-alloca-array.f90
@@ -0,0 +1,33 @@
+! Offload test that checks an allocatable derived type can be mapped alongside
+! one of its own allocatable components without disrupting either mapping
+! (similar to C++ mapping of a struct and a pointer, in concept at least).
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: dtype
+        real(4) :: i
+        integer, allocatable :: scalar
+        integer(4) :: array_i(10)
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: alloca_dtype
+    allocate(alloca_dtype)
+    allocate(alloca_dtype%array_j(10))
+
+!$omp target map(tofrom: alloca_dtype, alloca_dtype%array_j)
+    do i = 1, 10
+        alloca_dtype%array_j(i) = i
+    end do
+    alloca_dtype%k = 50
+!$omp end target
+
+print *, alloca_dtype%array_j
+print *, alloca_dtype%k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 50
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-array-and-scalar.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-array-and-scalar.f90
new file mode 100644
index 00000000000000..30f35be98af605
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-array-and-scalar.f90
@@ -0,0 +1,39 @@
+! Offloading test checking interaction of explicit member mapping of
+! non-allocatable members of an allocatable derived type.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: nested_dtype
+        real(4) :: i
+        real(4) :: j
+        integer(4) :: array_i(10)
+        integer(4) :: k
+    end type nested_dtype
+
+    type :: dtype
+        real(4) :: i
+        integer, allocatable :: scalar 
+        integer(4) :: array_i(10)
+        type(nested_dtype) :: nested_dtype
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: alloca_dtype
+    allocate(alloca_dtype)
+
+!$omp target map(tofrom: alloca_dtype%nested_dtype%array_i, alloca_dtype%k)
+    do i = 1, 10
+        alloca_dtype%nested_dtype%array_i(i) = i
+    end do
+    alloca_dtype%k = 50
+!$omp end target
+
+print *, alloca_dtype%k
+print *, alloca_dtype%nested_dtype%array_i
+end program main
+
+!CHECK: 50
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-alloca-dtype-array-of-dtype.f90 b/offload/test/offloading/fortran/target-map-alloca-dtype-array-of-dtype.f90
new file mode 100644
index 00000000000000..bad8ff55e638ba
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-alloca-dtype-array-of-dtype.f90
@@ -0,0 +1,35 @@
+! Offloading test checking interaction of explicit member mapping of an array
+! of derived types contained within an allocatable derived type
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: nested_dtype
+        real(4) :: i
+        real(4) :: j
+        integer(4) :: array_i(10)
+        integer(4) :: k
+    end type nested_dtype
+
+    type :: dtype
+        real(4) :: i
+        integer(4) :: array_i(10)
+        real(4) :: j
+        type(nested_dtype) :: array_dtype(10)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: dtyped
+    allocate(dtyped)
+
+!$omp target map(tofrom: dtyped%array_dtype)
+    do i = 1, 10
+        dtyped%array_dtype(i)%k = i
+    end do
+!$omp end target
+
+print *, dtyped%array_dtype%k
+
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-allocatable-array-section-1d-bounds.f90 b/offload/test/offloading/fortran/target-map-allocatable-array-section-1d-bounds.f90
index 3e981116ab42f1..7734ca9d0cf0d4 100644
--- a/offload/test/offloading/fortran/target-map-allocatable-array-section-1d-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-allocatable-array-section-1d-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of a
-! two 1-D allocatable arrays with a target region
-! while providing the map upper and lower bounds
+! Offloading test checking interaction of a two 1-D allocatable arrays with a
+! target region while providing the map upper and lower bounds
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-allocatable-array-section-3d-bounds.f90 b/offload/test/offloading/fortran/target-map-allocatable-array-section-3d-bounds.f90
index 31b6daa35ca5e8..7dba860a4a8328 100644
--- a/offload/test/offloading/fortran/target-map-allocatable-array-section-3d-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-allocatable-array-section-3d-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of allocatables
-! with multi-dimensional bounds (3-D in this case) and
-! a target region
+! Offloading test checking interaction of allocatables with multi-dimensional
+! bounds (3-D in this case) and a target region
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-allocatable-dtype.f90 b/offload/test/offloading/fortran/target-map-allocatable-dtype.f90
new file mode 100644
index 00000000000000..13b8890eee9a87
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-allocatable-dtype.f90
@@ -0,0 +1,31 @@
+! Offload test that checks an allocatable derived type can be mapped and at the
+! least non-allocatable components can be accessed.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: dtype
+        real(4) :: i
+        integer, allocatable :: scalar
+        integer(4) :: array_i(10)
+        real(4) :: j
+        integer, allocatable :: array_j(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype), allocatable :: alloca_dtype
+    allocate(alloca_dtype)
+
+!$omp target map(tofrom: alloca_dtype)
+    do i = 1, 10
+        alloca_dtype%array_i(i) = i
+    end do
+    alloca_dtype%k = 50
+!$omp end target
+
+print *, alloca_dtype%k
+print *, alloca_dtype%array_i
+end program main
+
+!CHECK: 50
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-allocatable-map-scopes.f90 b/offload/test/offloading/fortran/target-map-allocatable-map-scopes.f90
index 4a97a9ade91cdc..fe9c70900fefac 100644
--- a/offload/test/offloading/fortran/target-map-allocatable-map-scopes.f90
+++ b/offload/test/offloading/fortran/target-map-allocatable-map-scopes.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of allocatables
-! with target in different scopes
+! Offloading test checking interaction of allocatables with target in different
+! scopes
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-common-block.f90 b/offload/test/offloading/fortran/target-map-common-block.f90
index d2705db5d6fffb..f368762b46d49d 100644
--- a/offload/test/offloading/fortran/target-map-common-block.f90
+++ b/offload/test/offloading/fortran/target-map-common-block.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of
-! mapping a full common block in a target
-! region
+! Offloading test checking interaction of mapping a full common block in a
+! target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-declare-target-link-common-block.f90 b/offload/test/offloading/fortran/target-map-declare-target-link-common-block.f90
index 857e79c5fa6e21..697365709d2197 100644
--- a/offload/test/offloading/fortran/target-map-declare-target-link-common-block.f90
+++ b/offload/test/offloading/fortran/target-map-declare-target-link-common-block.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of
-! mapping a declare target link common
-! block with device_type any to a target
-! region
+! Offloading test checking interaction of mapping a declare target link common
+! block with device_type any to a target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-derived-type-full-1.f90 b/offload/test/offloading/fortran/target-map-derived-type-full-1.f90
index fe3b84398eaf01..9f2aeb4bf9fc07 100644
--- a/offload/test/offloading/fortran/target-map-derived-type-full-1.f90
+++ b/offload/test/offloading/fortran/target-map-derived-type-full-1.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type mapping when mapped
-! to target and assinging one derived type
-! to another
+! Offloading test checking interaction of an explicit derived type mapping when
+! mapped to target and assinging one derived type to another
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-derived-type-full-2.f90 b/offload/test/offloading/fortran/target-map-derived-type-full-2.f90
index 302af23a3c7e8b..f4f63f233cc046 100644
--- a/offload/test/offloading/fortran/target-map-derived-type-full-2.f90
+++ b/offload/test/offloading/fortran/target-map-derived-type-full-2.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type mapping when mapped to
-! target and assigning to individual members
+! Offloading test checking interaction of an explicit derived type mapping when
+! mapped to target and assigning to individual members
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-derived-type-full-implicit-1.f90 b/offload/test/offloading/fortran/target-map-derived-type-full-implicit-1.f90
index f43fa3cc69055f..8632b951d6fbd6 100644
--- a/offload/test/offloading/fortran/target-map-derived-type-full-implicit-1.f90
+++ b/offload/test/offloading/fortran/target-map-derived-type-full-implicit-1.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! implicit derived type mapping when mapped
-! to target and assinging one derived type
-! to another
+! Offloading test checking interaction of an implicit derived type mapping when
+! mapped to target and assinging one derived type to another
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-derived-type-full-implicit-2.f90 b/offload/test/offloading/fortran/target-map-derived-type-full-implicit-2.f90
index f6c8f9812b14c0..9331a48c3eb75e 100644
--- a/offload/test/offloading/fortran/target-map-derived-type-full-implicit-2.f90
+++ b/offload/test/offloading/fortran/target-map-derived-type-full-implicit-2.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type mapping when mapped
-! to target and assinging one derived type
-! to another
+! Offloading test checking interaction of an explicit derived type mapping when
+! mapped to target and assinging one derived type to another
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-double-large-nested-dtype-multi-member.f90 b/offload/test/offloading/fortran/target-map-double-large-nested-dtype-multi-member.f90
index e390481bbaba0a..ef27ad387ea88b 100644
--- a/offload/test/offloading/fortran/target-map-double-large-nested-dtype-multi-member.f90
+++ b/offload/test/offloading/fortran/target-map-double-large-nested-dtype-multi-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit member map from two large nested
-! derived types
+! Offloading test checking interaction of an explicit member map from two large
+! nested derived types
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-double-nested-dtype-array-bounds.f90 b/offload/test/offloading/fortran/target-map-double-nested-dtype-array-bounds.f90
index 7f260ac4cdc1ed..b90fbd05fb2e83 100644
--- a/offload/test/offloading/fortran/target-map-double-nested-dtype-array-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-double-nested-dtype-array-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of two
-! explicit arrau member maps with bounds from
-! two nested derived types
+! Offloading test checking interaction of two explicit array member maps with
+! bounds from two nested derived types
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-double-nested-dtype-double-array-bounds.f90 b/offload/test/offloading/fortran/target-map-double-nested-dtype-double-array-bounds.f90
index bd9bbb1b7eed50..f637dbed470c2f 100644
--- a/offload/test/offloading/fortran/target-map-double-nested-dtype-double-array-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-double-nested-dtype-double-array-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of two
-! explicit array member maps with array bounds
-! from two nested derived types
+! Offloading test checking interaction of two explicit array member maps with
+! array bounds from two nested derived types
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-double-nested-dtype-single-member.f90 b/offload/test/offloading/fortran/target-map-double-nested-dtype-single-member.f90
index 3c3b5e4b91e2fd..cbf1c6877bdefa 100644
--- a/offload/test/offloading/fortran/target-map-double-nested-dtype-single-member.f90
+++ b/offload/test/offloading/fortran/target-map-double-nested-dtype-single-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of two
-! derived types for a single array member each
+! Offloading test checking interaction of an explicit derived type member
+! mapping of two derived types for a single array member each
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-3d-alloca-array-with-bounds.f90 b/offload/test/offloading/fortran/target-map-dtype-3d-alloca-array-with-bounds.f90
new file mode 100644
index 00000000000000..9fa13339aaddb1
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-3d-alloca-array-with-bounds.f90
@@ -0,0 +1,41 @@
+! Offload test that checks an allocatable array can be mapped with a specified
+! 3-D bounds when contained within a derived type and mapped via member mapping.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:,:,:)
+    integer(4) :: k
+    end type top_layer
+
+    type(top_layer) :: one_l
+    integer :: inArray(3,3,3)
+
+    allocate(one_l%array_j(3,3,3))
+
+    do i = 1, 3
+        do j = 1, 3
+          do k = 1, 3
+              inArray(i, j, k) = 42
+              one_l%array_j(i, j, k) = 0
+          end do
+         end do
+      end do
+
+!$omp target map(tofrom: one_l%array_j(1:3, 1:3, 2:2)) map(to: inArray(1:3, 1:3, 1:3))
+    do j = 1, 3
+        do k = 1, 3
+            one_l%array_j(k, j, 2) = inArray(k, j, 2)
+        end do
+      end do
+!$omp end target
+
+    print *, one_l%array_j
+end program main
+
+!CHECK: 0 0 0 0 0 0 0 0 0 42 42 42 42 42 42 42 42 42 0 0 0 0 0 0 0 0 0
diff --git a/offload/test/offloading/fortran/target-map-dtype-alloca-and-non-alloca-array.f90 b/offload/test/offloading/fortran/target-map-dtype-alloca-and-non-alloca-array.f90
new file mode 100644
index 00000000000000..c72f82a806732a
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-alloca-and-non-alloca-array.f90
@@ -0,0 +1,36 @@
+! Offload test that checks an allocatable array can be mapped alongside a
+! non-allocatable array when both are contained within a derived type can be
+! mapped correctly via member mapping and then written to.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: one_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    end type one_layer
+
+    type(one_layer) :: one_l
+
+    allocate(one_l%array_j(10))
+
+    do i = 1, 10
+        one_l%array_i(i) = i
+    end do
+
+!$omp target map(tofrom: one_l%array_i, one_l%array_j)
+    do i = 1, 10
+        one_l%array_j(i) = one_l%array_i(i) + i
+    end do
+!$omp end target
+
+    print *, one_l%array_i
+    print *, one_l%array_j
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 2 4 6 8 10 12 14 16 18 20
diff --git a/offload/test/offloading/fortran/target-map-dtype-alloca-array-and-non-alloca-dtype.f90 b/offload/test/offloading/fortran/target-map-dtype-alloca-array-and-non-alloca-dtype.f90
new file mode 100644
index 00000000000000..172b3c4683f8b9
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-alloca-array-and-non-alloca-dtype.f90
@@ -0,0 +1,44 @@
+! Offload test that checks an allocatable array can be mapped alongside a
+! non-allocatable derived type when both are contained within a derived type
+! can be mapped correctly via member mapping and then written to.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer) :: nest
+    end type top_layer
+
+    type(top_layer) :: one_l
+
+    allocate(one_l%array_j(10))
+    allocate(one_l%scalar)
+
+    do i = 1, 10
+        one_l%nest%array_i(i) = i
+    end do
+
+!$omp target map(tofrom: one_l%nest, one_l%array_j)
+    do i = 1, 10
+        one_l%array_j(i) = one_l%nest%array_i(i) + i
+    end do
+!$omp end target
+
+    print *, one_l%nest%array_i
+    print *, one_l%array_j
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 2 4 6 8 10 12 14 16 18 20
diff --git a/offload/test/offloading/fortran/target-map-dtype-alloca-array-of-dtype.f90 b/offload/test/offloading/fortran/target-map-dtype-alloca-array-of-dtype.f90
new file mode 100644
index 00000000000000..f56a72843e6dfc
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-alloca-array-of-dtype.f90
@@ -0,0 +1,34 @@
+! Offload test that checks it is possible to member map an allocatable array of
+! derived types nested within a non-allocatable derived type.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: nested_dtype
+        real(4) :: i
+        real(4) :: j
+        integer(4) :: array_i(10)
+        integer(4) :: k
+    end type nested_dtype
+
+    type :: dtype
+        real(4) :: i
+        integer(4) :: array_i(10)
+        real(4) :: j
+        type(nested_dtype), allocatable :: array_dtype(:)
+        integer(4) :: k
+    end type dtype
+
+    type(dtype) :: dtyped
+    allocate(dtyped%array_dtype(10))
+
+!$omp target map(tofrom: dtyped%array_dtype)
+    do i = 1, 10
+        dtyped%array_dtype(i)%k = i
+    end do
+!$omp end target
+
+print *, dtyped%array_dtype%k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-dtype-alloca-array-with-bounds.f90 b/offload/test/offloading/fortran/target-map-dtype-alloca-array-with-bounds.f90
new file mode 100644
index 00000000000000..53480d1d569993
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-alloca-array-with-bounds.f90
@@ -0,0 +1,29 @@
+! Offload test that checks an allocatable array can be mapped with a specified
+! 1-D bounds when contained within a derived type and mapped via member mapping
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    end type top_layer
+
+    type(top_layer) :: one_l
+
+    allocate(one_l%array_j(10))
+
+!$omp target map(tofrom: one_l%array_j(2:6))
+    do index = 1, 10
+        one_l%array_j(index) = index
+    end do
+!$omp end target
+
+    print *, one_l%array_j(2:6)
+end program main
+
+!CHECK: 2 3 4 5 6
diff --git a/offload/test/offloading/fortran/target-map-dtype-allocatable-array.f90 b/offload/test/offloading/fortran/target-map-dtype-allocatable-array.f90
new file mode 100644
index 00000000000000..dbd839637f280d
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-allocatable-array.f90
@@ -0,0 +1,28 @@
+! Offload test that checks an allocatable array contained within a derived type
+! can be mapped correctly via member mapping and then written to.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: one_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    end type one_layer
+
+    type(one_layer) :: one_l
+    allocate(one_l%array_j(10))
+
+    !$omp target map(tofrom: one_l%array_j)
+        do i = 1, 10
+            one_l%array_j(i) = i
+        end do
+    !$omp end target
+
+    print *, one_l%array_j
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-dtype-allocatable-scalar-and-array.f90 b/offload/test/offloading/fortran/target-map-dtype-allocatable-scalar-and-array.f90
new file mode 100644
index 00000000000000..352d02e34caccf
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-dtype-allocatable-scalar-and-array.f90
@@ -0,0 +1,34 @@
+! Offload test that checks an allocatable array alongside an allocatable scalar
+! contained within a derived type can be mapped correctly via member mapping
+! and then written to.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: one_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    end type one_layer
+
+    type(one_layer) :: one_l
+
+    allocate(one_l%array_j(10))
+    allocate(one_l%scalar)
+
+    !$omp target map(tofrom: one_l%array_j, one_l%j)
+        do i = 1, 10
+            one_l%array_j(i) = i
+        end do
+        one_l%j = 50
+    !$omp end target
+
+    print *, one_l%j
+    print *, one_l%array_j
+end program main
+
+!CHECK: 50
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit-update.f90 b/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit-update.f90
index e3c28062687c84..74bbdc0a681099 100644
--- a/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit-update.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit-update.f90
@@ -1,8 +1,6 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of
-! an array with bounds when mapped to
-! target using a combination of update,
-! enter and exit directives.
+! Offloading test checking interaction of an explicit derived type member
+! mapping of an array with bounds when mapped to target using a combination of
+! update, enter and exit directives.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit.f90 b/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit.f90
index fd56142c5cbf25..8dcd7b5f39f3ac 100644
--- a/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-arr-bounds-member-enter-exit.f90
@@ -1,8 +1,6 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of
-! an array with bounds when mapped to
-! target using a combination of enter and
-! exit directives.
+! Offloading test checking interaction of an explicit derived type member
+! mapping of an array with bounds when mapped to target using a combination of
+! enter and exit directives.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-explicit-individual-array-member.f90 b/offload/test/offloading/fortran/target-map-dtype-explicit-individual-array-member.f90
index ce8a1b2ba84bae..c4496a3e411a65 100644
--- a/offload/test/offloading/fortran/target-map-dtype-explicit-individual-array-member.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-explicit-individual-array-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of
-! an array when mapped to target
+! Offloading test checking interaction of an explicit derived type member
+! mapping of an array when mapped to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-3D-member-bounds.f90 b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-3D-member-bounds.f90
index c72bb853780952..145818fd4111a4 100644
--- a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-3D-member-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-3D-member-bounds.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of
-! two arrays with explicit bounds when
-! mapped to target
+! Offloading test checking interaction of an explicit derived type member
+! mapping of two arrays with explicit bounds when mapped to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member-bounds.f90 b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member-bounds.f90
index b8defce2abc726..1a4742461f8c27 100644
--- a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member-bounds.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! explicit derived type member mapping of
-! two arrays with explicit bounds when
-! mapped to target
+! Offloading test checking interaction of an explicit derived type member
+! mapping of two arrays with explicit bounds when mapped to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member.f90 b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member.f90
index 5816598b5b50d5..d05b2dafebb1b2 100644
--- a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-array-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! derived type mapping of two explicit array
-! members to target
+! Offloading test checking interaction of an derived type mapping of two 
+! explicit array members to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-member.f90 b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-member.f90
index e869804470165b..1bda6d67f2b271 100644
--- a/offload/test/offloading/fortran/target-map-dtype-multi-explicit-member.f90
+++ b/offload/test/offloading/fortran/target-map-dtype-multi-explicit-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! derived type mapping of two explicit
-! members to target
+! Offloading test checking interaction of an derived type mapping of two 
+! explicit members to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-enter-exit-allocatables.f90 b/offload/test/offloading/fortran/target-map-enter-exit-allocatables.f90
index 60806feb7a1912..497b824b4fc72e 100644
--- a/offload/test/offloading/fortran/target-map-enter-exit-allocatables.f90
+++ b/offload/test/offloading/fortran/target-map-enter-exit-allocatables.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of allocatables
-! with enter, exit and target
+! Offloading test checking interaction of allocatables with enter, exit and
+! target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-enter-exit-array-2.f90 b/offload/test/offloading/fortran/target-map-enter-exit-array-2.f90
index 320b76fcdb467f..b6005e037e8365 100644
--- a/offload/test/offloading/fortran/target-map-enter-exit-array-2.f90
+++ b/offload/test/offloading/fortran/target-map-enter-exit-array-2.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of an
-! enter and exit map of an array of scalars
+! Offloading test checking interaction of an enter and exit map of an array of
+! scalars
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-enter-exit-array-bounds.f90 b/offload/test/offloading/fortran/target-map-enter-exit-array-bounds.f90
index cae1d47013f622..d6264c60ba30ed 100644
--- a/offload/test/offloading/fortran/target-map-enter-exit-array-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-enter-exit-array-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! enter and exit map of an array of scalars
-! with specified bounds
+! Offloading test checking interaction of an enter and exit map of an array of
+! scalars with specified bounds
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-enter-exit-array.f90 b/offload/test/offloading/fortran/target-map-enter-exit-array.f90
index 67946ba66cd961..51de70f092ef6b 100644
--- a/offload/test/offloading/fortran/target-map-enter-exit-array.f90
+++ b/offload/test/offloading/fortran/target-map-enter-exit-array.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of fixed size
-! arrays with enter, exit and target
+! Offloading test checking interaction of fixed size arrays with enter, exit
+! and target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-enter-exit-scalar.f90 b/offload/test/offloading/fortran/target-map-enter-exit-scalar.f90
index 046a906dd524dd..7e252259007fe8 100644
--- a/offload/test/offloading/fortran/target-map-enter-exit-scalar.f90
+++ b/offload/test/offloading/fortran/target-map-enter-exit-scalar.f90
@@ -1,5 +1,4 @@
-! Offloading test checking interaction of an
-! enter and exit map of an scalar
+! Offloading test checking interaction of an enter and exit map of an scalar
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-first-common-block-member.f90 b/offload/test/offloading/fortran/target-map-first-common-block-member.f90
index 57a7380221ad81..4f72f30cd51d7f 100644
--- a/offload/test/offloading/fortran/target-map-first-common-block-member.f90
+++ b/offload/test/offloading/fortran/target-map-first-common-block-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of
-! mapping a member of a common block to a
-! target region
+! Offloading test checking interaction of mapping a member of a common block to
+! a target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-individual-dtype-member-map.f90 b/offload/test/offloading/fortran/target-map-individual-dtype-member-map.f90
index 42a09499691058..081718e0448cde 100644
--- a/offload/test/offloading/fortran/target-map-individual-dtype-member-map.f90
+++ b/offload/test/offloading/fortran/target-map-individual-dtype-member-map.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! single explicit member map from a single
-! derived type.
+! Offloading test checking interaction of an single explicit member map from a
+! single derived type.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-large-nested-dtype-multi-member.f90 b/offload/test/offloading/fortran/target-map-large-nested-dtype-multi-member.f90
index 2e345fd49cfaa5..0b9fcea4422f4a 100644
--- a/offload/test/offloading/fortran/target-map-large-nested-dtype-multi-member.f90
+++ b/offload/test/offloading/fortran/target-map-large-nested-dtype-multi-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit member map a large nested derived
-! type
+! Offloading test checking interaction of an explicit member map a large nested
+! derived type
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-mix-imp-exp-common-block-members.f90 b/offload/test/offloading/fortran/target-map-mix-imp-exp-common-block-members.f90
index a634cd023c57c1..8c2907a87721db 100644
--- a/offload/test/offloading/fortran/target-map-mix-imp-exp-common-block-members.f90
+++ b/offload/test/offloading/fortran/target-map-mix-imp-exp-common-block-members.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of
-! mapping all the members of a common block
-! with a mix of explicit and implicit
-! mapping to a target region
+! Offloading test checking interaction of mapping all the members of a common
+! block with a mix of explicit and implicit mapping to a target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-alloca-members.f90 b/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-alloca-members.f90
new file mode 100644
index 00000000000000..7880069881acc1
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-alloca-members.f90
@@ -0,0 +1,90 @@
+! Offloading test checking interaction of an explicit member map allocatable
+! components of two large nested derived types. NOTE: Unfortunately this test
+! loses a bit of its bite as we do not currently support linking against an
+! offload compiled fortran runtime library which means allocatable scalar
+! assignment isn't going to work in target regions.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer1
+    real(4), allocatable :: i4
+    real(4), allocatable :: j4
+    integer, pointer :: array_ptr(:)
+    real(4), allocatable :: k4
+    end type bottom_layer1
+
+    type :: bottom_layer2
+      integer(4), allocatable :: i3
+      integer(4), allocatable :: j3
+      integer, allocatable :: scalar
+      real, allocatable :: array_j(:)
+      integer(4), allocatable :: k3
+    end type bottom_layer2
+
+    type :: middle_layer
+     real(4) :: array_i2(10)
+     real(4), allocatable :: i2
+     integer, pointer :: scalar_ptr
+     real(4) :: array_j2(10)
+     type(bottom_layer1), allocatable :: nest
+     type(bottom_layer2), allocatable :: nest2
+    end type middle_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer(4), allocatable :: array_i(:)
+    real(4) :: j
+    integer(4) :: k
+    type(middle_layer), allocatable :: nested
+    end type top_layer
+
+    type(top_layer), allocatable :: top_dtype
+    type(top_layer), allocatable :: top_dtype2
+    integer, target :: array_target(10)
+    integer, target :: array_target2(10)
+
+    allocate(top_dtype)
+    allocate(top_dtype2)
+    allocate(top_dtype%nested)
+    allocate(top_dtype2%nested)
+    allocate(top_dtype%nested%nest)
+    allocate(top_dtype2%nested%nest)
+    allocate(top_dtype%nested%nest2)
+    allocate(top_dtype2%nested%nest2)
+    allocate(top_dtype%array_i(10))
+    allocate(top_dtype2%array_i(10))
+
+    top_dtype%nested%nest%array_ptr => array_target
+    allocate(top_dtype%nested%nest2%array_j(10))
+
+    top_dtype2%nested%nest%array_ptr => array_target2
+    allocate(top_dtype2%nested%nest2%array_j(10))
+
+!$omp target map(tofrom: top_dtype%array_i, top_dtype%nested%nest2%array_j, top_dtype%nested%nest%array_ptr) &
+!$omp map(tofrom: top_dtype2%array_i, top_dtype2%nested%nest2%array_j, top_dtype2%nested%nest%array_ptr)
+    do i = 1, 10
+      top_dtype%nested%nest%array_ptr(i) = i
+      top_dtype%nested%nest2%array_j(i) = i
+      top_dtype%array_i(i) = i
+      top_dtype2%nested%nest%array_ptr(i) = i
+      top_dtype2%nested%nest2%array_j(i) = i
+      top_dtype2%array_i(i) = i
+    end do
+!$omp end target
+
+  print *, top_dtype%nested%nest%array_ptr
+  print *, top_dtype%nested%nest2%array_j
+  print *, top_dtype%array_i
+
+  print *, top_dtype2%nested%nest%array_ptr
+  print *, top_dtype2%nested%nest2%array_j
+  print *, top_dtype2%array_i
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
+!CHECK:  1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-mixed-members.f90 b/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-mixed-members.f90
new file mode 100644
index 00000000000000..7f1eda63d9d48b
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-multi-alloca-dtypes-with-multi-mixed-members.f90
@@ -0,0 +1,82 @@
+! Offloading test checking interaction of an explicit member map of mixed
+! allocatable and non-allocatable components of a nested derived types.
+!
+! NOTE: Unfortunately this test loses a bit of its bite as we do not currently
+! support linking against an offload compiled fortran runtime library which
+! means allocatable scalar assignment isn't going to work in target regions.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer1
+    real(4) :: i4
+    real(4), allocatable :: j4
+    real(4) :: k4
+    end type bottom_layer1
+
+    type :: bottom_layer2
+      integer(4) :: i3
+      integer(4) :: j3
+      integer(4), allocatable :: k3
+    end type bottom_layer2
+
+    type :: middle_layer
+     real(4) :: array_i2(10)
+     real(4) :: i2
+     real(4), allocatable :: array_j2(:)
+     type(bottom_layer1) :: nest
+     type(bottom_layer2), allocatable :: nest2
+    end type middle_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(middle_layer) :: nested
+    end type top_layer
+
+    type(top_layer), allocatable :: top_dtype
+
+    allocate(top_dtype)
+    allocate(top_dtype%array_j(10))
+    allocate(top_dtype%nested%nest2)
+    allocate(top_dtype%nested%array_j2(10))
+
+!$omp target map(tofrom: top_dtype%nested%nest%i4, top_dtype%nested%array_j2) &
+!$omp map(tofrom: top_dtype%nested%nest%k4, top_dtype%array_i, top_dtype%nested%nest2%i3) &
+!$omp map(tofrom: top_dtype%nested%i2, top_dtype%nested%nest2%j3, top_dtype%array_j)
+    top_dtype%nested%nest%i4 = 10
+    top_dtype%nested%nest%k4 = 10
+    top_dtype%nested%nest2%i3 = 20
+    top_dtype%nested%nest2%j3 = 40
+
+    top_dtype%nested%i2 = 200
+
+    do i = 1, 10
+        top_dtype%array_j(i) = i
+        top_dtype%array_i(i) = i
+        top_dtype%nested%array_j2(i) = i
+    end do
+!$omp end target
+
+  print *, top_dtype%nested%nest%i4
+  print *, top_dtype%nested%nest%k4
+  print *, top_dtype%nested%nest2%i3
+  print *, top_dtype%nested%nest2%j3
+
+  print *, top_dtype%nested%i2
+  print *, top_dtype%array_i
+  print *, top_dtype%array_j
+  print *, top_dtype%nested%array_j2
+end program main
+
+!CHECK: 10.
+!CHECK: 10.
+!CHECK: 20
+!CHECK: 40
+!CHECK: 200.
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
diff --git a/offload/test/offloading/fortran/target-map-nested-alloca-dtype-3d-alloca-array-bounds.f90 b/offload/test/offloading/fortran/target-map-nested-alloca-dtype-3d-alloca-array-bounds.f90
new file mode 100644
index 00000000000000..a03e8d137ce8a9
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-alloca-dtype-3d-alloca-array-bounds.f90
@@ -0,0 +1,51 @@
+! Offloading test checking interaction of an explicit member map of an
+! allocatable 3d array within a nested allocatable derived type with
+! specified bounds
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:,:,:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer), allocatable :: nest
+    end type top_layer
+
+    type(top_layer), allocatable :: one_l
+    integer :: inArray(3,3,3)
+    allocate(one_l)
+    allocate(one_l%nest)
+    allocate(one_l%nest%array_k(3,3,3))
+
+    do i = 1, 3
+        do j = 1, 3
+          do k = 1, 3
+              inArray(i, j, k) = 42
+              one_l%nest%array_k(i, j, k) = 0
+          end do
+         end do
+      end do
+
+!$omp target map(tofrom: one_l%nest%array_k(1:3, 1:3, 2:2)) map(to: inArray(1:3, 1:3, 1:3))
+    do j = 1, 3
+        do k = 1, 3
+            one_l%nest%array_k(k, j, 2) = inArray(k, j, 2)
+        end do
+      end do
+!$omp end target
+
+    print *, one_l%nest%array_k
+end program main
+
+!CHECK: 0 0 0 0 0 0 0 0 0 42 42 42 42 42 42 42 42 42 0 0 0 0 0 0 0 0 0
diff --git a/offload/test/offloading/fortran/target-map-nested-alloca-dtype-alloca-array-bounds.f90 b/offload/test/offloading/fortran/target-map-nested-alloca-dtype-alloca-array-bounds.f90
new file mode 100644
index 00000000000000..35b487370d1ab4
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-alloca-dtype-alloca-array-bounds.f90
@@ -0,0 +1,43 @@
+! Offloading test checking interaction of an explicit member map of an
+! allocatable array within a nested allocatable derived type with specified
+! bounds
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer), allocatable :: nest
+    end type top_layer
+
+    type(top_layer), allocatable :: one_l
+    allocate(one_l)
+    allocate(one_l%nest)
+    allocate(one_l%nest%array_k(10))
+
+    do index = 1, 10
+        one_l%nest%array_k(index) = 0
+    end do
+
+!$omp target map(tofrom: one_l%nest%array_k(2:6))
+    do index = 2, 6
+        one_l%nest%array_k(index) = index
+    end do
+!$omp end target
+
+    print *, one_l%nest%array_k
+end program main
+
+!CHECK: 0 2 3 4 5 6 0 0 0 0
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-3d-alloca-array-with-bounds.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-3d-alloca-array-with-bounds.f90
new file mode 100644
index 00000000000000..5aa5a094824a80
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-3d-alloca-array-with-bounds.f90
@@ -0,0 +1,50 @@
+! Offload test that checks an allocatable array can be mapped with a specified
+! 3-D bounds when contained within a nested derived type and mapped via member
+! mapping.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:,:,:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer) :: nest
+    end type top_layer
+
+    type(top_layer) :: one_l
+    integer :: inArray(3,3,3)
+
+    allocate(one_l%nest%array_k(3,3,3))
+
+    do i = 1, 3
+        do j = 1, 3
+          do k = 1, 3
+              inArray(i, j, k) = 42
+              one_l%nest%array_k(i, j, k) = 0
+          end do
+         end do
+      end do
+
+!$omp target map(tofrom: one_l%nest%array_k(1:3, 1:3, 2:2)) map(to: inArray(1:3, 1:3, 1:3))
+    do j = 1, 3
+        do k = 1, 3
+            one_l%nest%array_k(k, j, 2) = inArray(k, j, 2)
+        end do
+      end do
+!$omp end target
+
+    print *, one_l%nest%array_k
+end program main
+
+!CHECK: 0 0 0 0 0 0 0 0 0 42 42 42 42 42 42 42 42 42 0 0 0 0 0 0 0 0 0
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-alloca-and-non-alloca-array.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-and-non-alloca-array.f90
new file mode 100644
index 00000000000000..1bbf25e0b0aa1a
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-and-non-alloca-array.f90
@@ -0,0 +1,47 @@
+! Offload test that checks an allocatable array can be mapped via member
+! mapping alongside a non-allocatable array when both are contained within a
+! nested derived type.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer, allocatable :: scalar_i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: one_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer) :: nest
+    end type one_layer
+
+    type(one_layer) :: one_l
+
+    allocate(one_l%nest%array_k(10))
+    allocate(one_l%nest%scalar_i)
+
+    do i = 1, 10
+        one_l%nest%array_i(i) = i
+    end do
+
+    !$omp target map(tofrom: one_l%nest%array_i, one_l%nest%array_k)
+    do i = 1, 10
+        one_l%nest%array_k(i) = one_l%nest%array_i(i) + i
+    end do
+    !$omp end target
+
+    print *, one_l%nest%array_k
+    print *, one_l%nest%array_i
+end program main
+
+!CHECK: 2 4 6 8 10 12 14 16 18 20
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-and-non-alloca-dtype.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-and-non-alloca-dtype.f90
new file mode 100644
index 00000000000000..f278c37e99903e
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-and-non-alloca-dtype.f90
@@ -0,0 +1,51 @@
+! Offload test that checks an allocatable array can be mapped alongside a
+! non-allocatable derived type when both are contained within a nested derived
+! type.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: middle_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    type(bottom_layer) :: nest2
+    integer, allocatable :: array_k(:)
+    integer(4) :: k
+    end type middle_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(middle_layer) :: nest
+    end type top_layer
+
+    type(top_layer) :: one_l
+
+    allocate(one_l%nest%array_k(10))
+
+    do i = 1, 10
+        one_l%nest%nest2%array_i(i) = i
+    end do
+
+    !$omp target map(tofrom: one_l%nest%nest2, one_l%nest%array_k)
+        do i = 1, 10
+            one_l%nest%array_k(i) = one_l%nest%nest2%array_i(i) + i
+        end do
+    !$omp end target
+
+    print *, one_l%nest%nest2%array_i
+    print *, one_l%nest%array_k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
+!CHECK: 2 4 6 8 10 12 14 16 18 20
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-with-bounds.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-with-bounds.f90
new file mode 100644
index 00000000000000..84328f79ef1c90
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array-with-bounds.f90
@@ -0,0 +1,43 @@
+! Offload test that checks an allocatable array can be mapped with a specified
+! 1-D bounds when contained within a nested derived type and mapped via member
+! mapping.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: top_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer) :: nest
+    end type top_layer
+
+    type(top_layer) :: one_l
+
+    allocate(one_l%nest%array_k(10))
+
+    do index = 1, 10
+        one_l%nest%array_k(index) = 0
+    end do
+
+!$omp target map(tofrom: one_l%nest%array_k(2:6))
+    do index = 2, 6
+        one_l%nest%array_k(index) = index
+    end do
+!$omp end target
+
+print *, one_l%nest%array_k
+
+end program main
+
+!CHECK: 0 2 3 4 5 6 0 0 0 0
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array.f90
new file mode 100644
index 00000000000000..7d0b06dd67f020
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-alloca-array.f90
@@ -0,0 +1,37 @@
+! Offload test that checks an allocatable array contained within a nested
+! derived type can be mapped correctly via member mapping and then written to.
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+program main
+    type :: bottom_layer
+    real(4) :: i
+    integer(4) :: array_i(10)
+    integer, allocatable :: array_k(:)
+    integer(4) :: k
+    end type bottom_layer
+
+    type :: one_layer
+    real(4) :: i
+    integer, allocatable :: scalar
+    integer(4) :: array_i(10)
+    real(4) :: j
+    integer, allocatable :: array_j(:)
+    integer(4) :: k
+    type(bottom_layer) :: nest
+    end type one_layer
+
+    type(one_layer) :: one_l
+
+    allocate(one_l%nest%array_k(10))
+
+!$omp target map(tofrom: one_l%nest%array_k)
+    do i = 1, 10
+        one_l%nest%array_k(i) = i
+    end do
+!$omp end target
+
+    print *, one_l%nest%array_k
+end program main
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-complex-member.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-complex-member.f90
index becc88d3157c8b..e1c5fbba538734 100644
--- a/offload/test/offloading/fortran/target-map-nested-dtype-complex-member.f90
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-complex-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! nested derived type member map of a complex
-! number member
+! Offloading test checking interaction of an nested derived type member map of
+! a complex number member
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-derived-member.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-derived-member.f90
index c5f79de22b2963..3d5cb95a78c85b 100644
--- a/offload/test/offloading/fortran/target-map-nested-dtype-derived-member.f90
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-derived-member.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of an
-! nested derived type member map with the
-! inclusion of an entire nested derived
-! type being mapped
+! Offloading test checking interaction of an nested derived type member map
+! with the inclusion of an entire nested derived type being mapped
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-multi-member.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-multi-member.f90
index b3579c39f7f66b..b8153940fbaa2f 100644
--- a/offload/test/offloading/fortran/target-map-nested-dtype-multi-member.f90
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-multi-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! explicit member map from a small nested
-! derived type
+! Offloading test checking interaction of an explicit member map from a small
+! nested derived type
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-nested-dtype-single-member.f90 b/offload/test/offloading/fortran/target-map-nested-dtype-single-member.f90
index 20c245a5877c53..498b461ca25f27 100644
--- a/offload/test/offloading/fortran/target-map-nested-dtype-single-member.f90
+++ b/offload/test/offloading/fortran/target-map-nested-dtype-single-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of an
-! single explicit member map from a nested
-! derived type.
+! Offloading test checking interaction of an single explicit member map from a
+! nested derived type.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-pointer-scopes-enter-exit.f90 b/offload/test/offloading/fortran/target-map-pointer-scopes-enter-exit.f90
index 2dc03641f8a696..f5f284caed64f1 100644
--- a/offload/test/offloading/fortran/target-map-pointer-scopes-enter-exit.f90
+++ b/offload/test/offloading/fortran/target-map-pointer-scopes-enter-exit.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of pointers
-! with target in different scopes
+! Offloading test checking interaction of pointers with target in different 
+! scopes
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-pointer-target-array-section-3d-bounds.f90 b/offload/test/offloading/fortran/target-map-pointer-target-array-section-3d-bounds.f90
index 2b2c60f5d4b50e..6d9c497134ffb5 100644
--- a/offload/test/offloading/fortran/target-map-pointer-target-array-section-3d-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-pointer-target-array-section-3d-bounds.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of pointer
-! and target with target where 3-D bounds have
-! been specified
+! Offloading test checking interaction of pointer and target with target where
+! 3-D bounds have been specified
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-pointer-target-scopes.f90 b/offload/test/offloading/fortran/target-map-pointer-target-scopes.f90
index fcbcf024d9c9df..323f0e055fb546 100644
--- a/offload/test/offloading/fortran/target-map-pointer-target-scopes.f90
+++ b/offload/test/offloading/fortran/target-map-pointer-target-scopes.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of pointer
-! and target with target across multiple scopes
+! Offloading test checking interaction of pointer and target with target across
+! multiple scopes
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-pointer-to-dtype-allocatable-member.f90 b/offload/test/offloading/fortran/target-map-pointer-to-dtype-allocatable-member.f90
new file mode 100644
index 00000000000000..8e1e68528943fd
--- /dev/null
+++ b/offload/test/offloading/fortran/target-map-pointer-to-dtype-allocatable-member.f90
@@ -0,0 +1,45 @@
+! Offloading test checking interaction of implicit captured of a pointer
+! targeting an allocatable member of a derived type, alongside the explicit
+! map of the derived type and allocatable data via target enter and exit
+! directives
+! REQUIRES: flang, amdgpu
+
+! RUN: %libomptarget-compile-fortran-run-and-check-generic
+module dtype
+    type :: my_dtype
+            integer :: s, e
+            integer,dimension(:),allocatable :: values
+    end type
+end module
+
+program offload_types
+    use dtype
+
+    type(my_dtype),target :: my_instance
+    integer,dimension(:),pointer :: values_ptr
+    integer :: i
+
+    allocate(my_instance%values(20))
+    my_instance%s=1
+    my_instance%e=20
+
+    values_ptr => my_instance%values
+
+    !$omp target enter data map(to:my_instance, my_instance%values)
+
+    !$omp target
+      do i = 1,20
+             values_ptr(i) = i
+      end do
+    !$omp end target
+
+    !$omp target exit data map(from:my_instance%values)
+
+    write(*,*) my_instance%values
+
+    !$omp target exit data map(release:my_instance)
+
+    deallocate(my_instance%values)
+end program
+
+!CHECK: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
diff --git a/offload/test/offloading/fortran/target-map-second-common-block-member.f90 b/offload/test/offloading/fortran/target-map-second-common-block-member.f90
index fce876cfb02136..432c6e1f82b9f8 100644
--- a/offload/test/offloading/fortran/target-map-second-common-block-member.f90
+++ b/offload/test/offloading/fortran/target-map-second-common-block-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of
-! mapping a member of a common block to a
-! target region
+! Offloading test checking interaction of mapping a member of a common block to
+! a target region
 ! REQUIRES: flang, amdgcn-amd-amdhsa
 ! UNSUPPORTED: nvptx64-nvidia-cuda
 ! UNSUPPORTED: nvptx64-nvidia-cuda-LTO
diff --git a/offload/test/offloading/fortran/target-map-two-dtype-explicit-member.f90 b/offload/test/offloading/fortran/target-map-two-dtype-explicit-member.f90
index 2d9f3cdfccc91d..0fe153189f1b46 100644
--- a/offload/test/offloading/fortran/target-map-two-dtype-explicit-member.f90
+++ b/offload/test/offloading/fortran/target-map-two-dtype-explicit-member.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of two
-! derived type's with one explicit member
-! each being mapped with bounds to target
+! Offloading test checking interaction of two derived type's with one explicit
+! member each being mapped with bounds to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-two-dtype-individual-member-array-1D-bounds.f90 b/offload/test/offloading/fortran/target-map-two-dtype-individual-member-array-1D-bounds.f90
index 26d030c2f042d6..261443a4e4e964 100644
--- a/offload/test/offloading/fortran/target-map-two-dtype-individual-member-array-1D-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-two-dtype-individual-member-array-1D-bounds.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of two
-! derived type's with a single explicit array
-! member each being mapped with bounds to
-! target
+! Offloading test checking interaction of two derived type's with a single
+! explicit array member each being mapped with bounds to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-1.f90 b/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-1.f90
index 43cdf0497c57de..7d3a3eca340597 100644
--- a/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-1.f90
+++ b/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-1.f90
@@ -1,6 +1,5 @@
-! Offloading test checking interaction of two
-! derived type's with a mix of explicit and
-! implicit member mapping to target
+! Offloading test checking interaction of two derived type's with a mix of
+! explicit and implicit member mapping to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-2.f90 b/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-2.f90
index 336e182f8a16b8..8f0bae32709b86 100644
--- a/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-2.f90
+++ b/offload/test/offloading/fortran/target-map-two-dtype-mixed-implicit-explicit-capture-2.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of two
-! derived type's with a mix of explicit and
-! implicit member mapping of arrays to target
-! one with bounds.
+! Offloading test checking interaction of two derived type's with a mix of
+! explicit and implicit member mapping of arrays to target one with bounds.
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-two-dtype-multi-member-array-1D-bounds.f90 b/offload/test/offloading/fortran/target-map-two-dtype-multi-member-array-1D-bounds.f90
index 308d6a0c361002..4726093c3657ef 100644
--- a/offload/test/offloading/fortran/target-map-two-dtype-multi-member-array-1D-bounds.f90
+++ b/offload/test/offloading/fortran/target-map-two-dtype-multi-member-array-1D-bounds.f90
@@ -1,7 +1,5 @@
-! Offloading test checking interaction of two
-! derived type's with two explicit array
-! members each being mapped with bounds to
-! target
+! Offloading test checking interaction of two derived type's with two explicit
+! array members each being mapped with bounds to target
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-map-two-nested-dtype-member-array-map.f90 b/offload/test/offloading/fortran/target-map-two-nested-dtype-member-array-map.f90
index 1a59a7c95b9c0a..ef105839a7e4a3 100644
--- a/offload/test/offloading/fortran/target-map-two-nested-dtype-member-array-map.f90
+++ b/offload/test/offloading/fortran/target-map-two-nested-dtype-member-array-map.f90
@@ -1,5 +1,5 @@
-! Offloading test checking interaction of an
-! explicit member map utilising array bounds
+! Offloading test checking interaction of an explicit member map utilising 
+! array bounds
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-nested-target-data.f90 b/offload/test/offloading/fortran/target-nested-target-data.f90
index 6041433afbdc6e..1eb52caf278b7d 100644
--- a/offload/test/offloading/fortran/target-nested-target-data.f90
+++ b/offload/test/offloading/fortran/target-nested-target-data.f90
@@ -1,5 +1,4 @@
-! Offloading test for target nested inside
-! a target data region
+! Offloading test for target nested inside a target data region
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic
diff --git a/offload/test/offloading/fortran/target-region-implicit-array.f90 b/offload/test/offloading/fortran/target-region-implicit-array.f90
index 22d0c5ea655fcd..e3852569a5a82a 100644
--- a/offload/test/offloading/fortran/target-region-implicit-array.f90
+++ b/offload/test/offloading/fortran/target-region-implicit-array.f90
@@ -1,5 +1,5 @@
-! Basic offloading test of a regular array explicitly
-! passed within a target region
+! Basic offloading test of a regular array explicitly passed within a target 
+! region
 ! REQUIRES: flang, amdgpu
 
 ! RUN: %libomptarget-compile-fortran-run-and-check-generic

>From d0373c841db06b7ea069e4f1846cb2b81e730b2a Mon Sep 17 00:00:00 2001
From: agozillon <Andrew.Gozillon at amd.com>
Date: Fri, 4 Oct 2024 13:03:22 -0500
Subject: [PATCH 2/2] [OpenMP][MLIR] Descriptor explicit member map lowering
 changes

This is one of 3 PRs in a PR stack that aims to add support for explicit mapping of
allocatable members in derived types.

The primary changes in this PR are the OpenMPToLLVMIRTranslation.cpp changes,
which are small and seek to alter the current member mapping to add an
additional map insertion for pointers. Effectively, if the member is a pointer
(currently indicated by having a varPtrPtr field) we add an additional map for
the pointer and then alter the subsequent mapping of the member (the data)
to utilise the member rather than the parents base pointer. This appears to be
necessary in certain cases when mapping pointer data within record types to
avoid segfaulting on device (due to incorrect data mapping). In general this
record type mapping may be simplifiable in the future.

There are also additions of tests which should help to showcase the affect
of the changes above.
---
 mlir/include/mlir/Dialect/OpenMP/OpenMPOps.td |   2 +-
 mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp  |  58 +++------
 .../OpenMP/OpenMPToLLVMIRTranslation.cpp      |  81 ++++++++-----
 mlir/test/Dialect/OpenMP/ops.mlir             |   4 +-
 ...t-nested-ptr-record-type-mapping-host.mlir |  66 ++++++++++
 ...arget-nested-record-type-mapping-host.mlir |   2 +-
 ...get-record-type-with-ptr-member-host.mlir} | 114 ++++++++++--------
 7 files changed, 197 insertions(+), 130 deletions(-)
 create mode 100644 mlir/test/Target/LLVMIR/omptarget-nested-ptr-record-type-mapping-host.mlir
 rename mlir/test/Target/LLVMIR/{omptarget-fortran-allocatable-types-host.mlir => omptarget-record-type-with-ptr-member-host.mlir} (58%)

diff --git a/mlir/include/mlir/Dialect/OpenMP/OpenMPOps.td b/mlir/include/mlir/Dialect/OpenMP/OpenMPOps.td
index 5fd8184fe0e0f7..9e8aa5d1e786ae 100644
--- a/mlir/include/mlir/Dialect/OpenMP/OpenMPOps.td
+++ b/mlir/include/mlir/Dialect/OpenMP/OpenMPOps.td
@@ -895,7 +895,7 @@ def MapInfoOp : OpenMP_Op<"map.info", [AttrSizedOperandSegments]> {
                        TypeAttr:$var_type,
                        Optional<OpenMP_PointerLikeType>:$var_ptr_ptr,
                        Variadic<OpenMP_PointerLikeType>:$members,
-                       OptionalAttr<AnyIntElementsAttr>:$members_index,
+                       OptionalAttr<IndexListArrayAttr>:$members_index,
                        Variadic<OpenMP_MapBoundsType>:$bounds, /* rank-0 to rank-{n-1} */
                        OptionalAttr<UI64Attr>:$map_type,
                        OptionalAttr<VariableCaptureKindAttr>:$map_capture_type,
diff --git a/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp b/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp
index 4a27a5ed8eb74b..a6abe305263e58 100644
--- a/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp
+++ b/mlir/lib/Dialect/OpenMP/IR/OpenMPDialect.cpp
@@ -1395,16 +1395,15 @@ static void printMapClause(OpAsmPrinter &p, Operation *op,
 }
 
 static ParseResult parseMembersIndex(OpAsmParser &parser,
-                                     DenseIntElementsAttr &membersIdx) {
-  SmallVector<APInt> values;
-  int64_t value;
-  int64_t shape[2] = {0, 0};
-  unsigned shapeTmp = 0;
+                                     ArrayAttr &membersIdx) {
+  SmallVector<Attribute> values, memberIdxs;
+
   auto parseIndices = [&]() -> ParseResult {
+    int64_t value;
     if (parser.parseInteger(value))
       return failure();
-    shapeTmp++;
-    values.push_back(APInt(32, value, /*isSigned=*/true));
+    values.push_back(IntegerAttr::get(parser.getBuilder().getIntegerType(64),
+                                      APInt(64, value, /*isSigned=*/false)));
     return success();
   };
 
@@ -1418,52 +1417,29 @@ static ParseResult parseMembersIndex(OpAsmParser &parser,
     if (failed(parser.parseRSquare()))
       return failure();
 
-    // Only set once, if any indices are not the same size
-    // we error out in the next check as that's unsupported
-    if (shape[1] == 0)
-      shape[1] = shapeTmp;
-
-    // Verify that the recently parsed list is equal to the
-    // first one we parsed, they must be equal lengths to
-    // keep the rectangular shape DenseIntElementsAttr
-    // requires
-    if (shapeTmp != shape[1])
-      return failure();
-
-    shapeTmp = 0;
-    shape[0]++;
+    memberIdxs.push_back(ArrayAttr::get(parser.getContext(), values));
+    values.clear();
   } while (succeeded(parser.parseOptionalComma()));
 
-  if (!values.empty()) {
-    ShapedType valueType =
-        VectorType::get(shape, IntegerType::get(parser.getContext(), 32));
-    membersIdx = DenseIntElementsAttr::get(valueType, values);
-  }
+  if (!memberIdxs.empty())
+    membersIdx = ArrayAttr::get(parser.getContext(), memberIdxs);
 
   return success();
 }
 
 static void printMembersIndex(OpAsmPrinter &p, MapInfoOp op,
-                              DenseIntElementsAttr membersIdx) {
-  llvm::ArrayRef<int64_t> shape = membersIdx.getShapedType().getShape();
-  assert(shape.size() <= 2);
-
+                              ArrayAttr membersIdx) {
   if (!membersIdx)
     return;
 
-  for (int i = 0; i < shape[0]; ++i) {
+  llvm::interleaveComma(membersIdx, p, [&p](Attribute v) {
     p << "[";
-    int rowOffset = i * shape[1];
-    for (int j = 0; j < shape[1]; ++j) {
-      p << membersIdx.getValues<int32_t>()[rowOffset + j];
-      if ((j + 1) < shape[1])
-        p << ",";
-    }
+    auto memberIdx = cast<ArrayAttr>(v);
+    llvm::interleaveComma(memberIdx.getValue(), p, [&p](Attribute v2) {
+      p << cast<IntegerAttr>(v2).getInt();
+    });
     p << "]";
-
-    if ((i + 1) < shape[0])
-      p << ", ";
-  }
+  });
 }
 
 static void printCaptureType(OpAsmPrinter &p, Operation *op,
diff --git a/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp b/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp
index da11ee9960e1f9..64f4b6ba982ff1 100644
--- a/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp
+++ b/mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp
@@ -2878,46 +2878,36 @@ static int getMapDataMemberIdx(MapInfoData &mapData, omp::MapInfoOp memberOp) {
 
 static omp::MapInfoOp getFirstOrLastMappedMemberPtr(omp::MapInfoOp mapInfo,
                                                     bool first) {
-  DenseIntElementsAttr indexAttr = mapInfo.getMembersIndexAttr();
-
+  ArrayAttr indexAttr = mapInfo.getMembersIndexAttr();
   // Only 1 member has been mapped, we can return it.
   if (indexAttr.size() == 1)
-    if (auto mapOp =
-            dyn_cast<omp::MapInfoOp>(mapInfo.getMembers()[0].getDefiningOp()))
-      return mapOp;
+    return cast<omp::MapInfoOp>(mapInfo.getMembers()[0].getDefiningOp());
 
-  llvm::ArrayRef<int64_t> shape = indexAttr.getShapedType().getShape();
-  llvm::SmallVector<size_t> indices(shape[0]);
+  llvm::SmallVector<size_t> indices(indexAttr.size());
   std::iota(indices.begin(), indices.end(), 0);
 
   llvm::sort(indices.begin(), indices.end(),
              [&](const size_t a, const size_t b) {
-               auto indexValues = indexAttr.getValues<int32_t>();
-               for (int i = 0; i < shape[1]; ++i) {
-                 int aIndex = indexValues[a * shape[1] + i];
-                 int bIndex = indexValues[b * shape[1] + i];
+               auto memberIndicesA = cast<ArrayAttr>(indexAttr[a]);
+               auto memberIndicesB = cast<ArrayAttr>(indexAttr[b]);
+               for (const auto it : llvm::zip(memberIndicesA, memberIndicesB)) {
+                 int64_t aIndex = cast<IntegerAttr>(std::get<0>(it)).getInt();
+                 int64_t bIndex = cast<IntegerAttr>(std::get<1>(it)).getInt();
 
                  if (aIndex == bIndex)
                    continue;
 
-                 if (aIndex != -1 && bIndex == -1)
-                   return false;
-
-                 if (aIndex == -1 && bIndex != -1)
-                   return true;
-
-                 // A is earlier in the record type layout than B
                  if (aIndex < bIndex)
                    return first;
 
-                 if (bIndex < aIndex)
+                 if (aIndex > bIndex)
                    return !first;
                }
 
-               // Iterated the entire list and couldn't make a decision, all
-               // elements were likely the same. Return false, since the sort
-               // comparator should return false for equal elements.
-               return false;
+               // Iterated the up until the end of the smallest member and
+               // they were found to be equal up to that point, so select
+               // the member with the lowest index count, so the "parent"
+               return memberIndicesA.size() < memberIndicesB.size();
              });
 
   return llvm::cast<omp::MapInfoOp>(
@@ -3088,17 +3078,8 @@ static llvm::omp::OpenMPOffloadMappingFlags mapParentWithMembers(
       /*isSigned=*/false);
   combinedInfo.Sizes.push_back(size);
 
-  // TODO: This will need to be expanded to include the whole host of logic for
-  // the map flags that Clang currently supports (e.g. it should take the map
-  // flag of the parent map flag, remove the OMP_MAP_TARGET_PARAM and do some
-  // further case specific flag modifications). For the moment, it handles what
-  // we support as expected.
-  llvm::omp::OpenMPOffloadMappingFlags mapFlag =
-      llvm::omp::OpenMPOffloadMappingFlags::OMP_MAP_TO;
-
   llvm::omp::OpenMPOffloadMappingFlags memberOfFlag =
       ompBuilder.getMemberOfFlag(combinedInfo.BasePointers.size() - 1);
-  ompBuilder.setCorrectMemberOfFlag(mapFlag, memberOfFlag);
 
   // This creates the initial MEMBER_OF mapping that consists of
   // the parent/top level container (same as above effectively, except
@@ -3107,6 +3088,12 @@ static llvm::omp::OpenMPOffloadMappingFlags mapParentWithMembers(
   // only relevant if the structure in its totality is being mapped,
   // otherwise the above suffices.
   if (!parentClause.getPartialMap()) {
+    // TODO: This will need to be expanded to include the whole host of logic
+    // for the map flags that Clang currently supports (e.g. it should do some
+    // further case specific flag modifications). For the moment, it handles
+    // what we support as expected.
+    llvm::omp::OpenMPOffloadMappingFlags mapFlag = mapData.Types[mapDataIndex];
+    ompBuilder.setCorrectMemberOfFlag(mapFlag, memberOfFlag);
     combinedInfo.Types.emplace_back(mapFlag);
     combinedInfo.DevicePointers.emplace_back(
         llvm::OpenMPIRBuilder::DeviceInfoTy::None);
@@ -3157,6 +3144,31 @@ static void processMapMembersWithParent(
 
     assert(memberDataIdx >= 0 && "could not find mapped member of structure");
 
+    // If we're currently mapping a pointer to a block of data, we must
+    // initially map the pointer, and then attatch/bind the data with a
+    // subsequent map to the pointer. This segment of code generates the
+    // pointer mapping, which can in certain cases be optimised out as Clang
+    // currently does in its lowering. However, for the moment we do not do so,
+    // in part as we currently have substantially less information on the data
+    // being mapped at this stage.
+    if (checkIfPointerMap(memberClause)) {
+      auto mapFlag = llvm::omp::OpenMPOffloadMappingFlags(
+          memberClause.getMapType().value());
+      mapFlag &= ~llvm::omp::OpenMPOffloadMappingFlags::OMP_MAP_TARGET_PARAM;
+      mapFlag |= llvm::omp::OpenMPOffloadMappingFlags::OMP_MAP_MEMBER_OF;
+      ompBuilder.setCorrectMemberOfFlag(mapFlag, memberOfFlag);
+      combinedInfo.Types.emplace_back(mapFlag);
+      combinedInfo.DevicePointers.emplace_back(
+          llvm::OpenMPIRBuilder::DeviceInfoTy::None);
+      combinedInfo.Names.emplace_back(
+          LLVM::createMappingInformation(memberClause.getLoc(), ompBuilder));
+      combinedInfo.BasePointers.emplace_back(
+          mapData.BasePointers[mapDataIndex]);
+      combinedInfo.Pointers.emplace_back(mapData.BasePointers[memberDataIdx]);
+      combinedInfo.Sizes.emplace_back(builder.getInt64(
+          moduleTranslation.getLLVMModule()->getDataLayout().getPointerSize()));
+    }
+
     // Same MemberOfFlag to indicate its link with parent and other members
     // of.
     auto mapFlag =
@@ -3172,7 +3184,10 @@ static void processMapMembersWithParent(
         mapData.DevicePointers[memberDataIdx]);
     combinedInfo.Names.emplace_back(
         LLVM::createMappingInformation(memberClause.getLoc(), ompBuilder));
-    combinedInfo.BasePointers.emplace_back(mapData.BasePointers[mapDataIndex]);
+    uint64_t basePointerIndex =
+        checkIfPointerMap(memberClause) ? memberDataIdx : mapDataIndex;
+    combinedInfo.BasePointers.emplace_back(
+        mapData.BasePointers[basePointerIndex]);
     combinedInfo.Pointers.emplace_back(mapData.Pointers[memberDataIdx]);
     combinedInfo.Sizes.emplace_back(mapData.Sizes[memberDataIdx]);
   }
diff --git a/mlir/test/Dialect/OpenMP/ops.mlir b/mlir/test/Dialect/OpenMP/ops.mlir
index b606f9eb708cf3..9c9ae3fa6527e6 100644
--- a/mlir/test/Dialect/OpenMP/ops.mlir
+++ b/mlir/test/Dialect/OpenMP/ops.mlir
@@ -2633,8 +2633,8 @@ func.func @omp_map_with_members(%arg0: !llvm.ptr, %arg1: !llvm.ptr, %arg2: !llvm
   // CHECK: %[[MAP4:.*]] = omp.map.info var_ptr(%[[ARG4]] : !llvm.ptr, f32) map_clauses(from) capture(ByRef) -> !llvm.ptr {name = ""}
   %mapv5 = omp.map.info var_ptr(%arg4 : !llvm.ptr, f32) map_clauses(from) capture(ByRef) -> !llvm.ptr {name = ""}
 
-  // CHECK: %[[MAP5:.*]] = omp.map.info var_ptr(%[[ARG5]] : !llvm.ptr, !llvm.struct<(i32, struct<(i32, f32)>)>) map_clauses(from) capture(ByRef) members(%[[MAP3]], %[[MAP4]] : [1,0], [1,1] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "", partial_map = true}
-  %mapv6 = omp.map.info var_ptr(%arg5 : !llvm.ptr, !llvm.struct<(i32, struct<(i32, f32)>)>) map_clauses(from) capture(ByRef) members(%mapv4, %mapv5 : [1,0], [1,1] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "", partial_map = true}
+  // CHECK: %[[MAP5:.*]] = omp.map.info var_ptr(%[[ARG5]] : !llvm.ptr, !llvm.struct<(i32, struct<(i32, f32)>)>) map_clauses(from) capture(ByRef) members(%[[MAP3]], %[[MAP4]] : [1, 0], [1, 1] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "", partial_map = true}
+  %mapv6 = omp.map.info var_ptr(%arg5 : !llvm.ptr, !llvm.struct<(i32, struct<(i32, f32)>)>) map_clauses(from) capture(ByRef) members(%mapv4, %mapv5 : [1, 0], [1, 1] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "", partial_map = true}
 
   // CHECK: omp.target_exit_data map_entries(%[[MAP3]], %[[MAP4]], %[[MAP5]] : !llvm.ptr, !llvm.ptr, !llvm.ptr)
   omp.target_exit_data map_entries(%mapv4, %mapv5, %mapv6 : !llvm.ptr, !llvm.ptr, !llvm.ptr){}
diff --git a/mlir/test/Target/LLVMIR/omptarget-nested-ptr-record-type-mapping-host.mlir b/mlir/test/Target/LLVMIR/omptarget-nested-ptr-record-type-mapping-host.mlir
new file mode 100644
index 00000000000000..5d772a13fe578a
--- /dev/null
+++ b/mlir/test/Target/LLVMIR/omptarget-nested-ptr-record-type-mapping-host.mlir
@@ -0,0 +1,66 @@
+// RUN: mlir-translate -mlir-to-llvmir %s | FileCheck %s
+
+// This test checks the offload sizes, map types and base pointers and pointers
+// provided to the OpenMP kernel argument structure are correct when lowering
+// to LLVM-IR from MLIR when performing explicit member mapping of a record type
+// that includes records with pointer members in various locations of the record
+// types hierarchy.
+
+module attributes {omp.is_target_device = false, omp.target_triples = ["amdgcn-amd-amdhsa"]} {
+  llvm.func @omp_nested_derived_type_alloca_map(%arg0: !llvm.ptr) {
+    %0 = llvm.mlir.constant(4 : index) : i64
+    %1 = llvm.mlir.constant(1 : index) : i64
+    %2 = llvm.mlir.constant(2 : index) : i64
+    %3 = llvm.mlir.constant(0 : index) : i64
+    %4 = llvm.mlir.constant(6 : index) : i64
+    %5 = omp.map.bounds lower_bound(%3 : i64) upper_bound(%0 : i64) extent(%0 : i64) stride(%1 : i64) start_idx(%3 : i64) {stride_in_bytes = true}
+    %6 = llvm.getelementptr %arg0[0, 6] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<(f32, struct<(ptr, i64, i32, i8, i8, i8, i8)>, array<10 x i32>, f32, struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>, i32, struct<(f32, array<10 x i32>, struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>, i32)>)>
+    %7 = llvm.getelementptr %6[0, 2] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<(f32, array<10 x i32>, struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>, i32)>
+    %8 = llvm.getelementptr %7[0, 0] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
+    %9 = omp.map.info var_ptr(%7 : !llvm.ptr, i32) var_ptr_ptr(%8 : !llvm.ptr) map_clauses(tofrom) capture(ByRef) bounds(%5) -> !llvm.ptr {name = ""}
+    %10 = omp.map.info var_ptr(%7 : !llvm.ptr, !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>) map_clauses(tofrom) capture(ByRef) -> !llvm.ptr {name = "one_l%nest%array_k"}
+    %11 = omp.map.info var_ptr(%arg0 : !llvm.ptr, !llvm.struct<(f32, struct<(ptr, i64, i32, i8, i8, i8, i8)>, array<10 x i32>, f32, struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>, i32, struct<(f32, array<10 x i32>, struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>, i32)>)>) map_clauses(tofrom) capture(ByRef) members(%10, %9 : [6,2], [6,2,0] : !llvm.ptr, !llvm.ptr) -> !llvm.ptr {name = "one_l", partial_map = true}
+    omp.target map_entries(%10 -> %arg1, %9 -> %arg2, %11 -> %arg3 : !llvm.ptr, !llvm.ptr, !llvm.ptr) {
+      omp.terminator
+    }
+    llvm.return
+  }
+}
+
+// CHECK: @.offload_sizes{{.*}} = private unnamed_addr constant [4 x i64] [i64 0, i64 48, i64 8, i64 20]
+// CHECK: @.offload_maptypes{{.*}} = private unnamed_addr constant [4 x i64] [i64 32, i64 281474976710659, i64 281474976710659, i64 281474976710675]
+
+// CHECK: define void @omp_nested_derived_type_alloca_map(ptr %[[ARG:.*]]) {
+
+// CHECK: %[[NESTED_DTYPE_MEMBER_GEP:.*]] = getelementptr { float, { ptr, i64, i32, i8, i8, i8, i8 }, [10 x i32], float, { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, i32, { float, [10 x i32], { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, i32 } }, ptr %[[ARG]], i32 0, i32 6
+// CHECK: %[[NESTED_STRUCT_PTR_MEMBER_GEP:.*]] = getelementptr { float, [10 x i32], { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, i32 }, ptr %[[NESTED_DTYPE_MEMBER_GEP]], i32 0, i32 2
+// CHECK: %[[NESTED_STRUCT_PTR_MEMBER_BADDR_GEP:.*]] = getelementptr { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr %[[NESTED_STRUCT_PTR_MEMBER_GEP]], i32 0, i32 0
+// CHECK: %[[NESTED_STRUCT_PTR_MEMBER_BADDR_LOAD:.*]] = load ptr, ptr %[[NESTED_STRUCT_PTR_MEMBER_BADDR_GEP]], align 8
+// CHECK: %[[ARR_OFFSET:.*]] = getelementptr inbounds i32, ptr %[[NESTED_STRUCT_PTR_MEMBER_BADDR_LOAD]], i64 0
+// CHECK: %[[DTYPE_SIZE_SEGMENT_CALC_1:.*]] = getelementptr { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr %[[NESTED_STRUCT_PTR_MEMBER_GEP]], i64 1
+// CHECK: %[[DTYPE_SIZE_SEGMENT_CALC_2:.*]] = ptrtoint ptr %[[DTYPE_SIZE_SEGMENT_CALC_1]] to i64
+// CHECK: %[[DTYPE_SIZE_SEGMENT_CALC_3:.*]] = ptrtoint ptr %[[NESTED_STRUCT_PTR_MEMBER_GEP]] to i64
+// CHECK: %[[DTYPE_SIZE_SEGMENT_CALC_4:.*]] = sub i64 %[[DTYPE_SIZE_SEGMENT_CALC_2]], %[[DTYPE_SIZE_SEGMENT_CALC_3]]
+// CHECK: %[[DTYPE_SIZE_SEGMENT_CALC_5:.*]] = sdiv exact i64 %[[DTYPE_SIZE_SEGMENT_CALC_4]], ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
+
+// CHECK:  %[[BASE_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_baseptrs, i32 0, i32 0
+// CHECK:  store ptr %[[ARG]], ptr %[[BASE_PTRS]], align 8
+// CHECK:  %[[OFFLOAD_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_ptrs, i32 0, i32 0
+// CHECK:  store ptr %[[NESTED_STRUCT_PTR_MEMBER_GEP]], ptr %[[OFFLOAD_PTRS]], align 8
+// CHECK:  %[[OFFLOAD_SIZES:.*]] = getelementptr inbounds [4 x i64], ptr %.offload_sizes, i32 0, i32 0
+// CHECK:  store i64 %[[DTYPE_SIZE_SEGMENT_CALC_5]], ptr %[[OFFLOAD_SIZES]], align 8
+
+// CHECK:  %[[BASE_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_baseptrs, i32 0, i32 1
+// CHECK:  store ptr %[[ARG]], ptr %[[BASE_PTRS]], align 8
+// CHECK:  %[[OFFLOAD_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_ptrs, i32 0, i32 1
+// CHECK:  store ptr %[[NESTED_STRUCT_PTR_MEMBER_GEP]], ptr %[[OFFLOAD_PTRS]], align 8
+
+// CHECK:  %[[BASE_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_baseptrs, i32 0, i32 2
+// CHECK:  store ptr %[[ARG]], ptr %[[BASE_PTRS]], align 8
+// CHECK:  %[[OFFLOAD_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_ptrs, i32 0, i32 2
+// CHECK:  store ptr %[[NESTED_STRUCT_PTR_MEMBER_BADDR_GEP]], ptr %[[OFFLOAD_PTRS]], align 8
+
+// CHECK:  %[[BASE_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_baseptrs, i32 0, i32 3
+// CHECK:  store ptr %[[NESTED_STRUCT_PTR_MEMBER_BADDR_GEP]], ptr %[[BASE_PTRS]], align 8
+// CHECK:  %[[OFFLOAD_PTRS:.*]] = getelementptr inbounds [4 x ptr], ptr %.offload_ptrs, i32 0, i32 3
+// CHECK:  store ptr %[[ARR_OFFSET]], ptr %[[OFFLOAD_PTRS]], align 8
diff --git a/mlir/test/Target/LLVMIR/omptarget-nested-record-type-mapping-host.mlir b/mlir/test/Target/LLVMIR/omptarget-nested-record-type-mapping-host.mlir
index 8c1182c839a257..8837b42f70a447 100644
--- a/mlir/test/Target/LLVMIR/omptarget-nested-record-type-mapping-host.mlir
+++ b/mlir/test/Target/LLVMIR/omptarget-nested-record-type-mapping-host.mlir
@@ -21,7 +21,7 @@ llvm.func @_QQmain() {
     %9 = llvm.getelementptr %4[0, 1] : (!llvm.ptr) -> !llvm.ptr, !llvm.struct<(f32, array<10 x i32>, struct<(f32, i32)>, i32)>
     %10 = omp.map.bounds lower_bound(%2 : i64) upper_bound(%1 : i64) extent(%0 : i64) stride(%2 : i64) start_idx(%2 : i64)
     %11 = omp.map.info var_ptr(%9 : !llvm.ptr, !llvm.array<10 x i32>) map_clauses(tofrom) capture(ByRef) bounds(%10) -> !llvm.ptr
-    %12 = omp.map.info var_ptr(%4 : !llvm.ptr, !llvm.struct<(f32, array<10 x i32>, struct<(f32, i32)>, i32)>) map_clauses(tofrom) capture(ByRef) members(%6, %8, %11 : [3, -1], [2, 1], [1, -1] : !llvm.ptr, !llvm.ptr, !llvm.ptr) -> !llvm.ptr {partial_map = true}
+    %12 = omp.map.info var_ptr(%4 : !llvm.ptr, !llvm.struct<(f32, array<10 x i32>, struct<(f32, i32)>, i32)>) map_clauses(tofrom) capture(ByRef) members(%6, %8, %11 : [3], [2, 1], [1] : !llvm.ptr, !llvm.ptr, !llvm.ptr) -> !llvm.ptr {partial_map = true}
     omp.target map_entries(%6 -> %arg0, %8 -> %arg1, %11 -> %arg2, %12 -> %arg3 : !llvm.ptr, !llvm.ptr, !llvm.ptr, !llvm.ptr) {
       omp.terminator
     }
diff --git a/mlir/test/Target/LLVMIR/omptarget-fortran-allocatable-types-host.mlir b/mlir/test/Target/LLVMIR/omptarget-record-type-with-ptr-member-host.mlir
similarity index 58%
rename from mlir/test/Target/LLVMIR/omptarget-fortran-allocatable-types-host.mlir
rename to mlir/test/Target/LLVMIR/omptarget-record-type-with-ptr-member-host.mlir
index 6fe77bd228ef2f..d5ca33fb38f38e 100644
--- a/mlir/test/Target/LLVMIR/omptarget-fortran-allocatable-types-host.mlir
+++ b/mlir/test/Target/LLVMIR/omptarget-record-type-with-ptr-member-host.mlir
@@ -1,20 +1,20 @@
 // RUN: mlir-translate -mlir-to-llvmir %s | FileCheck %s
 
 // This test checks the offload sizes, map types and base pointers and pointers
-// provided to the OpenMP kernel argument structure are correct when lowering 
-// to LLVM-IR from MLIR when a fortran allocatable descriptor type is provided 
-// alongside the omp.map.info, the test utilises mapping of array sections, 
+// provided to the OpenMP kernel argument structure are correct when lowering
+// to LLVM-IR from MLIR when a structure with a pointer member type is provided
+// alongside the omp.map.info, the test utilises mapping of array sections,
 // full arrays and individual allocated scalars.
 
 module attributes {omp.is_target_device = false, omp.target_triples = ["amdgcn-amd-amdhsa"]} {
-  llvm.func @_QQmain() {
+  llvm.func @main() {
     %0 = llvm.mlir.constant(5 : index) : i64
     %1 = llvm.mlir.constant(2 : index) : i64
     %2 = llvm.mlir.constant(1 : index) : i64
-    %3 = llvm.mlir.addressof @_QFEfull_arr : !llvm.ptr
+    %3 = llvm.mlir.addressof @full_arr : !llvm.ptr
     %4 = llvm.mlir.constant(1 : i64) : i64
-    %5 = llvm.alloca %4 x !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8)> {bindc_name = "scalar"} : (i64) -> !llvm.ptr
-    %6 = llvm.mlir.addressof @_QFEsect_arr : !llvm.ptr
+    %5 = llvm.alloca %4 x !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8)> : (i64) -> !llvm.ptr
+    %6 = llvm.mlir.addressof @sect_arr : !llvm.ptr
     %7 = llvm.mlir.constant(0 : i64) : i64
     %8 = llvm.getelementptr %3[0, 7, %7, 0] : (!llvm.ptr, i64) -> !llvm.ptr, !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
     %9 = llvm.load %8 : !llvm.ptr -> i64
@@ -47,11 +47,11 @@ module attributes {omp.is_target_device = false, omp.target_triples = ["amdgcn-a
     }
     llvm.return
   }
-  llvm.mlir.global internal @_QFEfull_arr() {addr_space = 0 : i32} : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)> {
+  llvm.mlir.global internal @full_arr() {addr_space = 0 : i32} : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)> {
     %0 = llvm.mlir.undef : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
     llvm.return %0 : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
   }
-  llvm.mlir.global internal @_QFEsect_arr() {addr_space = 0 : i32} : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)> {
+  llvm.mlir.global internal @sect_arr() {addr_space = 0 : i32} : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)> {
     %0 = llvm.mlir.undef : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
     llvm.return %0 : !llvm.struct<(ptr, i64, i32, i8, i8, i8, i8, array<1 x array<3 x i64>>)>
   }
@@ -59,11 +59,11 @@ module attributes {omp.is_target_device = false, omp.target_triples = ["amdgcn-a
 
 // CHECK: @[[FULL_ARR_GLOB:.*]] = internal global { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] } undef
 // CHECK: @[[ARR_SECT_GLOB:.*]] = internal global { ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] } undef
-// CHECK: @.offload_sizes = private unnamed_addr constant [9 x i64] [i64 0, i64 48, i64 0, i64 0, i64 48, i64 0, i64 0, i64 24, i64 4]
-// CHECK: @.offload_maptypes = private unnamed_addr constant [9 x i64] [i64 32, i64 281474976710657, i64 281474976710675, i64 32, i64 1125899906842625, i64 1125899906842643, i64 32, i64 1970324836974593, i64 1970324836974611]
-// CHECK: @.offload_mapnames = private constant [9 x ptr] [ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}]
+// CHECK: @.offload_sizes = private unnamed_addr constant [12 x i64] [i64 0, i64 48, i64 8, i64 0, i64 0, i64 48, i64 8, i64 0, i64 0, i64 24, i64 8, i64 4]
+// CHECK: @.offload_maptypes = private unnamed_addr constant [12 x i64] [i64 32, i64 281474976710659, i64 281474976710659, i64 281474976710675, i64 32, i64 1407374883553283, i64 1407374883553283, i64 1407374883553299, i64 32, i64 2533274790395907, i64 2533274790395907, i64 2533274790395923]
+// CHECK: @.offload_mapnames = private constant [12 x ptr] [ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}, ptr @{{.*}}]
 
-// CHECK: define void @_QQmain()
+// CHECK: define void @main()
 // CHECK: %[[SCALAR_ALLOCA:.*]] = alloca { ptr, i64, i32, i8, i8, i8, i8 }, i64 1, align 8
 // CHECK: %[[FULL_ARR_SIZE5:.*]] = load i64, ptr getelementptr ({ ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr @[[FULL_ARR_GLOB]], i32 0, i32 7, i64 0, i32 1), align 4
 // CHECK: %[[FULL_ARR_SIZE4:.*]] = sub i64 %[[FULL_ARR_SIZE5]], 1
@@ -79,67 +79,77 @@ module attributes {omp.is_target_device = false, omp.target_triples = ["amdgcn-a
 // CHECK: %[[ARR_SECT_SIZE2:.*]] = add i64 %[[ARR_SECT_SIZE3]], 1
 // CHECK: %[[ARR_SECT_SIZE1:.*]] = mul i64 1, %[[ARR_SECT_SIZE2]]
 // CHECK: %[[ARR_SECT_SIZE:.*]] = mul i64 %[[ARR_SECT_SIZE1]], 4
-// CHECK: %[[LFULL_ARR:.*]] = load ptr, ptr @_QFEfull_arr, align 8
+// CHECK: %[[LFULL_ARR:.*]] = load ptr, ptr @full_arr, align 8
 // CHECK: %[[FULL_ARR_PTR:.*]] = getelementptr inbounds float, ptr %[[LFULL_ARR]], i64 0
 // CHECK: %[[ARR_SECT_OFFSET1:.*]] = mul i64 %[[ARR_SECT_OFFSET2]], 1
-// CHECK: %[[LARR_SECT:.*]] = load ptr, ptr @_QFEsect_arr, align 8
+// CHECK: %[[LARR_SECT:.*]] = load ptr, ptr @sect_arr, align 8
 // CHECK: %[[ARR_SECT_PTR:.*]] = getelementptr inbounds i32, ptr %[[LARR_SECT]], i64 %[[ARR_SECT_OFFSET1]]
 // CHECK: %[[SCALAR_PTR_LOAD:.*]] = load ptr, ptr %[[SCALAR_BASE]], align 8
-// CHECK: %[[FULL_ARR_DESC_SIZE:.*]] = sdiv exact i64 sub (i64 ptrtoint (ptr getelementptr ({ ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr @_QFEfull_arr, i32 1) to i64), i64 ptrtoint (ptr @_QFEfull_arr to i64)), ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
-// CHECK: %[[ARR_SECT_DESC_SIZE:.*]] = sdiv exact i64 sub (i64 ptrtoint (ptr getelementptr ({ ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr @_QFEsect_arr, i32 1) to i64), i64 ptrtoint (ptr @_QFEsect_arr to i64)), ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
+// CHECK: %[[FULL_ARR_DESC_SIZE:.*]] = sdiv exact i64 sub (i64 ptrtoint (ptr getelementptr ({ ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr @full_arr, i32 1) to i64), i64 ptrtoint (ptr @full_arr to i64)), ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
+// CHECK: %[[ARR_SECT_DESC_SIZE:.*]] = sdiv exact i64 sub (i64 ptrtoint (ptr getelementptr ({ ptr, i64, i32, i8, i8, i8, i8, [1 x [3 x i64]] }, ptr @sect_arr, i32 1) to i64), i64 ptrtoint (ptr @sect_arr to i64)), ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
 // CHECK: %[[SCALAR_DESC_SZ4:.*]] = getelementptr { ptr, i64, i32, i8, i8, i8, i8 }, ptr %[[SCALAR_ALLOCA]], i32 1
 // CHECK: %[[SCALAR_DESC_SZ3:.*]] = ptrtoint ptr %[[SCALAR_DESC_SZ4]] to i64
 // CHECK: %[[SCALAR_DESC_SZ2:.*]] = ptrtoint ptr %[[SCALAR_ALLOCA]] to i64
 // CHECK: %[[SCALAR_DESC_SZ1:.*]] = sub i64 %[[SCALAR_DESC_SZ3]], %[[SCALAR_DESC_SZ2]]
 // CHECK: %[[SCALAR_DESC_SZ:.*]] = sdiv exact i64 %[[SCALAR_DESC_SZ1]], ptrtoint (ptr getelementptr (i8, ptr null, i32 1) to i64)
 
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 0
-// CHECK: store ptr @_QFEfull_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 0
-// CHECK: store ptr @_QFEfull_arr, ptr %[[OFFLOADPTRS]], align 8
-
-// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [9 x i64], ptr %.offload_sizes, i32 0, i32 0
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 0
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 0
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [12 x i64], ptr %.offload_sizes, i32 0, i32 0
 // CHECK: store i64 %[[FULL_ARR_DESC_SIZE]], ptr %[[OFFLOADSIZES]], align 8
-
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 1
-// CHECK: store ptr @_QFEfull_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 1
-// CHECK: store ptr @_QFEfull_arr, ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 2
-// CHECK: store ptr @_QFEfull_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 2
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 1
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 1
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 2
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 2
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 3
+// CHECK: store ptr @full_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 3
 // CHECK: store ptr %[[FULL_ARR_PTR]], ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [9 x i64], ptr %.offload_sizes, i32 0, i32 2
+// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [12 x i64], ptr %.offload_sizes, i32 0, i32 3
 // CHECK: store i64 %[[FULL_ARR_SIZE]], ptr %[[OFFLOADSIZES]], align 8
 
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 3
-// CHECK: store ptr @_QFEsect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 3
-// CHECK: store ptr @_QFEsect_arr, ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [9 x i64], ptr %.offload_sizes, i32 0, i32 3
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 4
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 4
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [12 x i64], ptr %.offload_sizes, i32 0, i32 4
 // CHECK: store i64 %[[ARR_SECT_DESC_SIZE]], ptr %[[OFFLOADSIZES]], align 8
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 4
-// CHECK: store ptr @_QFEsect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 4
-// CHECK: store ptr @_QFEsect_arr, ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 5
-// CHECK: store ptr @_QFEsect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 5
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 5
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 5
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 6
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 6
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 7
+// CHECK: store ptr @sect_arr, ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 7
 // CHECK: store ptr %[[ARR_SECT_PTR]], ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [9 x i64], ptr %.offload_sizes, i32 0, i32 5
+// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [12 x i64], ptr %.offload_sizes, i32 0, i32 7
 // CHECK: store i64 %[[ARR_SECT_SIZE]], ptr %[[OFFLOADSIZES]], align 8
 
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 6
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 8
 // CHECK: store ptr %[[SCALAR_ALLOCA]], ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 6
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 8
 // CHECK: store ptr %[[SCALAR_ALLOCA]], ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [9 x i64], ptr %.offload_sizes, i32 0, i32 6
+// CHECK: %[[OFFLOADSIZES:.*]] = getelementptr inbounds [12 x i64], ptr %.offload_sizes, i32 0, i32 8
 // CHECK: store i64 %[[SCALAR_DESC_SZ]], ptr %[[OFFLOADSIZES]], align 8
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 7
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 9
 // CHECK: store ptr %[[SCALAR_ALLOCA]], ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 7
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 9
 // CHECK: store ptr %[[SCALAR_ALLOCA]], ptr %[[OFFLOADPTRS]], align 8
-// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_baseptrs, i32 0, i32 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 10
 // CHECK: store ptr %[[SCALAR_ALLOCA]], ptr %[[OFFLOADBASEPTRS]], align 8
-// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [9 x ptr], ptr %.offload_ptrs, i32 0, i32 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 10
+// CHECK: store ptr %[[SCALAR_BASE]], ptr %[[OFFLOADPTRS]], align 8
+// CHECK: %[[OFFLOADBASEPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_baseptrs, i32 0, i32 11
+// CHECK: store ptr %[[SCALAR_BASE]], ptr %[[OFFLOADBASEPTRS]], align 8
+// CHECK: %[[OFFLOADPTRS:.*]] = getelementptr inbounds [12 x ptr], ptr %.offload_ptrs, i32 0, i32 11
 // CHECK: store ptr %[[SCALAR_PTR_LOAD]], ptr %[[OFFLOADPTRS]], align 8



More information about the llvm-branch-commits mailing list