Fixed build with C++20 standard

[vedranmiletic]

Building LLVM with CMAKE_CXX_STANDARD set to 20 fails since the iterator facade is not fully compatible with C++20. To make it compatible, specific operator overloads have to be constrained.

Overload for operator- in ADT iterator is now constrained with concept BaseT::IsRandomAccess.

Patch by Jonathan Wakely.

Fixes #139072.

[llvmbot]

@llvm/pr-subscribers-llvm-adt

Author: Vedran Miletić (vedranmiletic)

Changes

Overload for operator- in ADT iterator is now constrained with concept BaseT::IsRandomAccess.

Patch by Jonathan Wakely.

Fixes #139072.

---

Full diff: https://github.com/llvm/llvm-project/pull/169772.diff

1 Files Affected:

• (modified) llvm/include/llvm/ADT/iterator.h (+5-1)

diff --git a/llvm/include/llvm/ADT/iterator.h b/llvm/include/llvm/ADT/iterator.h
index 6f0c42fe08bec..f55f9d4a3e7f2 100644
--- a/llvm/include/llvm/ADT/iterator.h
+++ b/llvm/include/llvm/ADT/iterator.h
@@ -267,7 +267,11 @@ class iterator_adaptor_base
     return *static_cast<DerivedT *>(this);
   }
   using BaseT::operator-;
-  difference_type operator-(const DerivedT &RHS) const {
+  difference_type operator-(const DerivedT &RHS) const
+#ifdef __cpp_concepts
+    requires(bool(BaseT::IsRandomAccess))
+#endif
+  {
     static_assert(
         BaseT::IsRandomAccess,
         "The '-' operator is only defined for random access iterators.");

[jwakely]

As noted at #139072 (comment) there are other operators which should really be constrained similarly. So this fixes the immediate issue, but there are still latent bugs that might cause similar errors later.

[jwakely]

but there are still latent bugs that might cause similar errors later.

But they could be fixed later, if needed.

[cor3ntin]

Please update the description of the PR with a description of the problem you are trying to fix, I had to find this comment to understand the context. Thanks
#139072 (comment)

[github-actions]

✅ With the latest revision this PR passed the C/C++ code formatter.

[vedranmiletic]

Please update the description of the PR with a description of the problem you are trying to fix, I had to find this comment to understand the context. Thanks #139072 (comment)

Hope it is better now.

[vedranmiletic]

Can someone merge this for me? I don't have commit access.

[llvm-ci]

LLVM Buildbot has detected a new failure on builder ppc64le-mlir-rhel-clang running on ppc64le-mlir-rhel-test while building llvm at step 6 "test-build-check-mlir-build-only-check-mlir".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/129/builds/35481

Here is the relevant piece of the build log for the reference

Step 6 (test-build-check-mlir-build-only-check-mlir) failure: 1200 seconds without output running [b'ninja', b'check-mlir'], attempting to kill
...
PASS: MLIR-Unit :: Dialect/SparseTensor/./MLIRSparseTensorTests/95/210 (3773 of 3784)
PASS: MLIR-Unit :: Dialect/SparseTensor/./MLIRSparseTensorTests/97/210 (3774 of 3784)
PASS: MLIR-Unit :: IR/./MLIRIRTests/0/131 (3775 of 3784)
PASS: MLIR-Unit :: IR/./MLIRIRTests/39/131 (3776 of 3784)
PASS: MLIR-Unit :: IR/./MLIRIRTests/38/131 (3777 of 3784)
PASS: MLIR-Unit :: Interfaces/./MLIRInterfacesTests/11/22 (3778 of 3784)
PASS: MLIR-Unit :: Interfaces/./MLIRInterfacesTests/12/22 (3779 of 3784)
PASS: MLIR-Unit :: Interfaces/./MLIRInterfacesTests/13/22 (3780 of 3784)
PASS: MLIR-Unit :: Pass/./MLIRPassTests/11/14 (3781 of 3784)
PASS: MLIR :: mlir-reduce/dce-test.mlir (3782 of 3784)
command timed out: 1200 seconds without output running [b'ninja', b'check-mlir'], attempting to kill
process killed by signal 9
program finished with exit code -1
elapsedTime=4205.691615

[vedranmiletic]

Thanks!