diff options
| author | bors <bors@rust-lang.org> | 2023-08-05 01:53:32 +0000 |
|---|---|---|
| committer | bors <bors@rust-lang.org> | 2023-08-05 01:53:32 +0000 |
| commit | 90f0b24ad3e7fc0dc0e419c9da30d74629cd5736 (patch) | |
| tree | d3a6e8648994cb031983ab2c7c3b806e16900bec /compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp | |
| parent | e173a8e6634d787a1529ff6ec5dc63273425ca34 (diff) | |
| parent | 20bd0c377163f384ae66943674ed2e397dd1c73a (diff) | |
| download | rust-90f0b24ad3e7fc0dc0e419c9da30d74629cd5736.tar.gz rust-90f0b24ad3e7fc0dc0e419c9da30d74629cd5736.zip | |
Auto merge of #98333 - SimonSapin:riscv-atomic, r=Amanieu
Re-enable atomic loads and stores for all RISC-V targets This roughly reverts PR https://github.com/rust-lang/rust/pull/66548 Atomic "CAS" are still disabled for targets without the *“A” Standard Extension for Atomic Instructions*. However this extension only adds instructions for operations more complex than simple loads and stores, which are always atomic when aligned. In the [Unprivileged Spec v. 20191213](https://riscv.org/technical/specifications/) section 2.6 *Load and Store Instructions* of chapter 2 *RV32I Base Integer Instruction Set* (emphasis mine): > Even when misaligned loads and stores complete successfully, these accesses might run extremely slowly depending on the implementation (e.g., when implemented via an invisible trap). Further-more, whereas **naturally aligned loads and stores are guaranteed to execute atomically**, misaligned loads and stores might not, and hence require additional synchronization to ensure atomicity. Unfortunately PR https://github.com/rust-lang/rust/pull/66548 did not provide much details on the bug that motivated it, but https://github.com/rust-lang/rust/issues/66240 and https://github.com/rust-lang/rust/issues/85736 appear related and happen with targets that do have the A extension.
Diffstat (limited to 'compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp')
0 files changed, 0 insertions, 0 deletions
