allocations are allowed to grow (but not shrink) - #159729
Conversation
|
r? @Darksonn rustbot has assigned @Darksonn. Use Why was this reviewer chosen?The reviewer was selected based on:
|
|
@rfcbot merge opsem |
|
@RalfJung has proposed to merge this. The next step is review by the rest of the tagged team members: No concerns currently listed. Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up! See this document for info about what commands tagged team members can give me. |
|
🔔 This is now entering its final comment period, as per the review above. 🔔 |
d8c3c0c to
122a5c6
Compare
|
📋 This PR cannot be approved because it currently has the following label: |
|
The final comment period, with a disposition to merge, as per the review above, is now complete. As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed. |
|
@bors r+ rollup |
allocations are allowed to grow (but not shrink) This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g. `malloc`, `alloca`, or the Rust global allocator operations) never change their size, so we have to exclude that case. This resolves a part of rust-lang/unsafe-code-guidelines#430. It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612181257) and [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612191402) for some details about their use-cases. Cc @rust-lang/opsem @rust-lang/lang
allocations are allowed to grow (but not shrink) This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g. `malloc`, `alloca`, or the Rust global allocator operations) never change their size, so we have to exclude that case. This resolves a part of rust-lang/unsafe-code-guidelines#430. It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612181257) and [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612191402) for some details about their use-cases. Cc @rust-lang/opsem @rust-lang/lang
Rollup of 6 pull requests Successful merges: - #159844 (Subtree cg_gcc sync (2026-07-24)) - #156527 (Move `std::io` tests to `alloctests` & add prelude) - #159525 (Stabilize passing 128-bit integers via vector registers with `asm!` on x86) - #160342 (Specialize `advance_by` method of `Fuse`) - #106643 (Allow only implementing `Read::read_buf`) - #159729 (allocations are allowed to grow (but not shrink))
allocations are allowed to grow (but not shrink) This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g. `malloc`, `alloca`, or the Rust global allocator operations) never change their size, so we have to exclude that case. This resolves a part of rust-lang/unsafe-code-guidelines#430. It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612181257) and [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612191402) for some details about their use-cases. Cc @rust-lang/opsem @rust-lang/lang
Rollup of 6 pull requests Successful merges: - #156527 (Move `std::io` tests to `alloctests` & add prelude) - #159525 (Stabilize passing 128-bit integers via vector registers with `asm!` on x86) - #160342 (Specialize `advance_by` method of `Fuse`) - #106643 (Allow only implementing `Read::read_buf`) - #159729 (allocations are allowed to grow (but not shrink)) - #159881 (fix: Do not stop `visible_parent_map` breadth-first search reaching children of `#[doc(hidden)]` modules)
allocations are allowed to grow (but not shrink) This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g. `malloc`, `alloca`, or the Rust global allocator operations) never change their size, so we have to exclude that case. This resolves a part of rust-lang/unsafe-code-guidelines#430. It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612181257) and [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612191402) for some details about their use-cases. Cc @rust-lang/opsem @rust-lang/lang
…uwer Rollup of 12 pull requests Successful merges: - #159906 (Semantic check of `mut` restrictions) - #156527 (Move `std::io` tests to `alloctests` & add prelude) - #159525 (Stabilize passing 128-bit integers via vector registers with `asm!` on x86) - #160342 (Specialize `advance_by` method of `Fuse`) - #160358 (borrowck: Simplify deps to build compiler 30s faster) - #160375 (miri subtree update) - #106643 (Allow only implementing `Read::read_buf`) - #159499 (tests: prefer max-llvm-major-version over open LLVM ranges) - #159729 (allocations are allowed to grow (but not shrink)) - #159881 (fix: Do not stop `visible_parent_map` breadth-first search reaching children of `#[doc(hidden)]` modules) - #160189 (Move codegen_stmt_debuginfo to debuginfo.rs) - #160356 (Add more tests for `must_implement_one_of`)
Rollup merge of #159729 - RalfJung:alloc-grow, r=Darksonn allocations are allowed to grow (but not shrink) This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g. `malloc`, `alloca`, or the Rust global allocator operations) never change their size, so we have to exclude that case. This resolves a part of rust-lang/unsafe-code-guidelines#430. It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612181257) and [here](https://rust-lang.zulipchat.com/#narrow/channel/136281-t-opsem/topic/Shrinking.20allocations.20in-place/near/612191402) for some details about their use-cases. Cc @rust-lang/opsem @rust-lang/lang
…uwer Rollup of 12 pull requests Successful merges: - rust-lang/rust#159906 (Semantic check of `mut` restrictions) - rust-lang/rust#156527 (Move `std::io` tests to `alloctests` & add prelude) - rust-lang/rust#159525 (Stabilize passing 128-bit integers via vector registers with `asm!` on x86) - rust-lang/rust#160342 (Specialize `advance_by` method of `Fuse`) - rust-lang/rust#160358 (borrowck: Simplify deps to build compiler 30s faster) - rust-lang/rust#160375 (miri subtree update) - rust-lang/rust#106643 (Allow only implementing `Read::read_buf`) - rust-lang/rust#159499 (tests: prefer max-llvm-major-version over open LLVM ranges) - rust-lang/rust#159729 (allocations are allowed to grow (but not shrink)) - rust-lang/rust#159881 (fix: Do not stop `visible_parent_map` breadth-first search reaching children of `#[doc(hidden)]` modules) - rust-lang/rust#160189 (Move codegen_stmt_debuginfo to debuginfo.rs) - rust-lang/rust#160356 (Add more tests for `must_implement_one_of`)
This got permitted on the LLVM side about a year ago (llvm/llvm-project#141338), but didn't require any code changes since LLVM's optimizations were already compatible with growing allocations. That said, LLVM assumes that allocations created via operations it recognizes (e.g.
malloc,alloca, or the Rust global allocator operations) never change their size, so we have to exclude that case.This resolves a part of rust-lang/unsafe-code-guidelines#430.
It may seem strange that allocations can grow but not shrink, but we did have multiple users show up in #t-opsem with exactly that request. Cc @nhusung @foonathan; see here and here for some details about their use-cases.
Cc @rust-lang/opsem @rust-lang/lang