Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[libc++][test] Avoid using allocator<const T> #73545

Merged
merged 1 commit into from
Nov 28, 2023

Conversation

StephanTLavavej
Copy link
Member

Found while running libc++'s test suite with MSVC's STL.

MSVC's STL rejects allocator<const T>. This may or may not be justified by the current Standardese (it was bogus in the C++03 era), but it's how we reject usage like vector<const T>.

A bunch of mdspan tests are failing for us because some centralized machinery is using allocator<const T>. Testing that mdspan and its associated types work properly with const T is good and necessary, but directly allocating const T is what's a problem for MSVC's STL. I'd like to ask for a very targeted change here that preserves all of the test coverage but changes how ElementPool interacts with allocator.

This intentionally leaves ElementPool::get_ptr() returning T* (pointer-to-possibly-const), so there's no externally visible difference.

This intentionally leaves `ElementPool::get_ptr()` returning `T*`, so there's no externally visible difference.
@StephanTLavavej StephanTLavavej requested a review from a team as a code owner November 27, 2023 17:02
@llvmbot llvmbot added the libc++ libc++ C++ Standard Library. Not GNU libstdc++. Not libc++abi. label Nov 27, 2023
@llvmbot
Copy link
Member

llvmbot commented Nov 27, 2023

@llvm/pr-subscribers-libcxx

Author: Stephan T. Lavavej (StephanTLavavej)

Changes

Found while running libc++'s test suite with MSVC's STL.

MSVC's STL rejects allocator&lt;const T&gt;. This may or may not be justified by the current Standardese (it was bogus in the C++03 era), but it's how we reject usage like vector&lt;const T&gt;.

A bunch of mdspan tests are failing for us because some centralized machinery is using allocator&lt;const T&gt;. Testing that mdspan and its associated types work properly with const T is good and necessary, but directly allocating const T is what's a problem for MSVC's STL. I'd like to ask for a very targeted change here that preserves all of the test coverage but changes how ElementPool interacts with allocator.

This intentionally leaves ElementPool::get_ptr() returning T* (pointer-to-possibly-const), so there's no externally visible difference.


Full diff: https://github.com/llvm/llvm-project/pull/73545.diff

1 Files Affected:

  • (modified) libcxx/test/std/containers/views/mdspan/MinimalElementType.h (+5-4)
diff --git a/libcxx/test/std/containers/views/mdspan/MinimalElementType.h b/libcxx/test/std/containers/views/mdspan/MinimalElementType.h
index b1fbd6ed944d179..fe7f0e1f2383790 100644
--- a/libcxx/test/std/containers/views/mdspan/MinimalElementType.h
+++ b/libcxx/test/std/containers/views/mdspan/MinimalElementType.h
@@ -9,7 +9,8 @@
 #ifndef TEST_STD_CONTAINERS_VIEWS_MDSPAN_MINIMAL_ELEMENT_TYPE_H
 #define TEST_STD_CONTAINERS_VIEWS_MDSPAN_MINIMAL_ELEMENT_TYPE_H
 
-#include<memory>
+#include <memory>
+#include <type_traits>
 
 // Idiosyncratic element type for mdspan
 // Make sure we don't assume copyable, default constructible, movable etc.
@@ -25,7 +26,7 @@ struct MinimalElementType {
 template<class T, size_t N>
 struct ElementPool {
   constexpr ElementPool() {
-    ptr_ = std::allocator<T>().allocate(N);
+    ptr_ = std::allocator<std::remove_const_t<T>>().allocate(N);
     for (int i = 0; i != N; ++i)
       std::construct_at(ptr_ + i, 42);
   }
@@ -35,11 +36,11 @@ struct ElementPool {
   constexpr ~ElementPool() {
     for (int i = 0; i != N; ++i)
       std::destroy_at(ptr_ + i);
-    std::allocator<T>().deallocate(ptr_, N);
+    std::allocator<std::remove_const_t<T>>().deallocate(ptr_, N);
   }
 
 private:
-  T* ptr_;
+  std::remove_const_t<T>* ptr_;
 };
 
 #endif // TEST_STD_CONTAINERS_VIEWS_MDSPAN_MINIMAL_ELEMENT_TYPE_H

Copy link
Contributor

@philnik777 philnik777 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM assuming the CI is green.

@StephanTLavavej
Copy link
Member Author

CI is green, modulo unrelated timeouts and MemorySanitizer failures, I'll go ahead and merge this. Thanks!

@StephanTLavavej StephanTLavavej merged commit 7cbf959 into llvm:main Nov 28, 2023
34 of 39 checks passed
@StephanTLavavej StephanTLavavej deleted the stl-8-allocator-const-t branch November 28, 2023 01:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
libc++ libc++ C++ Standard Library. Not GNU libstdc++. Not libc++abi.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants