Revert "Work around CS9363 for nameof of pointer-typed fields" - #130308
Conversation
This reverts commit 22c452b. The upstream roslyn fix is now in a compiler used by the repo.
|
Tagging subscribers to this area: @steveisok, @dotnet/area-system-reflection |
There was a problem hiding this comment.
Pull request overview
This PR updates System.Runtime reflection tests to use nameof(...) again for pointer-typed and function-pointer-typed fields, removing the prior string-literal workaround in the test data now that the compiler issue is no longer needed.
Changes:
- Replace hard-coded field name strings (
"s_intPtr","intPtr","s_fcnPtr", etc.) withnameof(...)for pointer and function pointer fields inGetValue/SetValuetest data.
Copilot Code ReviewHolistic AssessmentMotivation: Justified — the PR reverts a temporary workaround (string literals instead of Approach: Correct — this is a clean revert restoring the idiomatic Summary: ✅ LGTM. All 8 replacements correctly restore Detailed FindingsDetailed Findings✅ Correctness — All
|
This reverts commit 22c452b. The upstream roslyn fix is now in a compiler used by the repo.
This reverts commit 22c452b. The upstream roslyn fix is now in a compiler used by the repo.