Uh oh!
There was an error while loading. Please reload this page.
Rollup of 9 pull requests - #94332
Closed
Dylan-DPC wants to merge 25 commits into
Closed
Conversation
Like I previously did for `reverse`, this leaves it to LLVM to pick how to vectorize it, since it can know better the chunk size to use, compared to the "32 bytes always" approach we currently have. It does still need logic to type-erase where appropriate, though, as while LLVM is now smart enough to vectorize over slices of things like `[u8; 4]`, it fails to do so over slices of `[u8; 3]`. As a bonus, this also means one no longer gets the spurious `memcpy`(s?) at the end up swapping a slice of `__m256`s: <https://rust.godbolt.org/z/joofr4v8Y>
- Test the combinations of --check-cfg with partial values() and --cfg - Test that we detect unexpected value when none are expected
`scan_escape` currently has a fast path (for when the first char isn't '\\') and a slow path. This commit changes `scan_escape` so it only handles the slow path, i.e. the actual escaping code. The fast path is inlined into the two call sites. This change makes the code faster, because there is no function call overhead on the fast path. (`scan_escape` is a big function and doesn't get inlined.) This change also improves readability, because it removes a bunch of mode checks on the the fast paths.
The change looks big because `rustfmt` rearranges things, but the only real change is the inlining annotation.
…error-span, r=jackh726 better ObligationCause for normalization errors in `can_type_implement_copy` Some logic is needed so we can point to the field when given totally nonsense types like `struct Foo(<u32 as Iterator>::Item);` Fixesrust-lang#93687
…r=cjgillot Remove in band lifetimes As discussed in t-lang backlog bonanza, the `in_band_lifetimes` FCP closed in favor for the feature not being stabilized. This PR removes `#![feature(in_band_lifetimes)]` in its entirety. Let me know if this PR is too hasty, and if we should instead do something intermediate for deprecate the feature first. r? ```@scottmcm``` (or feel free to reassign, just saw your last comment on rust-lang#44524) Closesrust-lang#44524
…rochenkov Improve `--check-cfg` implementation This pull-request is a mix of improvements regarding the `--check-cfg` implementation: - Simpler internal representation (usage of `Option` instead of separate bool) - Add --check-cfg to the unstable book (based on the RFC) - Improved diagnostics: * List possible values when the value is unexpected * Suggest if possible a name or value that is similar - Add more tests (well known names, mix of combinations, ...) r? ``@petrochenkov``
Stop manually SIMDing in `swap_nonoverlapping` Like I previously did for `reverse` (rust-lang#90821), this leaves it to LLVM to pick how to vectorize it, since it can know better the chunk size to use, compared to the "32 bytes always" approach we currently have. A variety of codegen tests are included to confirm that the various cases are still being vectorized. It does still need logic to type-erase in some cases, though, as while LLVM is now smart enough to vectorize over slices of things like `[u8; 4]`, it fails to do so over slices of `[u8; 3]`. As a bonus, this change also means one no longer gets the spurious `memcpy`(s?) at the end up swapping a slice of `__m256`s: <https://rust.godbolt.org/z/joofr4v8Y> <details> <summary>ASM for this example</summary> ## Before (from godbolt) note the `push`/`pop`s and `memcpy` ```x86 swap_m256_slice: push r15 push r14 push r13 push r12 push rbx sub rsp, 32 cmp rsi, rcx jne .LBB0_6 mov r14, rsi shl r14, 5 je .LBB0_6 mov r15, rdx mov rbx, rdi xor eax, eax .LBB0_3: mov rcx, rax vmovaps ymm0, ymmword ptr [rbx + rax] vmovaps ymm1, ymmword ptr [r15 + rax] vmovaps ymmword ptr [rbx + rax], ymm1 vmovaps ymmword ptr [r15 + rax], ymm0 add rax, 32 add rcx, 64 cmp rcx, r14 jbe .LBB0_3 sub r14, rax jbe .LBB0_6 add rbx, rax add r15, rax mov r12, rsp mov r13, qword ptr [rip + memcpy@GOTPCREL] mov rdi, r12 mov rsi, rbx mov rdx, r14 vzeroupper call r13 mov rdi, rbx mov rsi, r15 mov rdx, r14 call r13 mov rdi, r15 mov rsi, r12 mov rdx, r14 call r13 .LBB0_6: add rsp, 32 pop rbx pop r12 pop r13 pop r14 pop r15 vzeroupper ret ``` ## After (from my machine) Note no `rsp` manipulation, sorry for different ASM syntax ```x86 swap_m256_slice: cmpq %r9, %rdx jne .LBB1_6 testq %rdx, %rdx je .LBB1_6 cmpq $1, %rdx jne .LBB1_7 xorl %r10d, %r10d jmp .LBB1_4 .LBB1_7: movq %rdx, %r9 andq $-2, %r9 movl $32, %eax xorl %r10d, %r10d .p2align 4, 0x90 .LBB1_8: vmovaps -32(%rcx,%rax), %ymm0 vmovaps -32(%r8,%rax), %ymm1 vmovaps %ymm1, -32(%rcx,%rax) vmovaps %ymm0, -32(%r8,%rax) vmovaps (%rcx,%rax), %ymm0 vmovaps (%r8,%rax), %ymm1 vmovaps %ymm1, (%rcx,%rax) vmovaps %ymm0, (%r8,%rax) addq $2, %r10 addq $64, %rax cmpq %r10, %r9 jne .LBB1_8 .LBB1_4: testb $1, %dl je .LBB1_6 shlq $5, %r10 vmovaps (%rcx,%r10), %ymm0 vmovaps (%r8,%r10), %ymm1 vmovaps %ymm1, (%rcx,%r10) vmovaps %ymm0, (%r8,%r10) .LBB1_6: vzeroupper retq ``` </details> This does all its copying operations as either the original type or as `MaybeUninit`s, so as far as I know there should be no potential abstract machine issues with reading padding bytes as integers. <details> <summary>Perf is essentially unchanged</summary> Though perhaps with more target features this would help more, if it could pick bigger chunks ## Before ``` running 10 tests test slice::swap_with_slice_4x_usize_30 ... bench: 894 ns/iter (+/- 11) test slice::swap_with_slice_4x_usize_3000 ... bench: 99,476 ns/iter (+/- 2,784) test slice::swap_with_slice_5x_usize_30 ... bench: 1,257 ns/iter (+/- 7) test slice::swap_with_slice_5x_usize_3000 ... bench: 139,922 ns/iter (+/- 959) test slice::swap_with_slice_rgb_30 ... bench: 328 ns/iter (+/- 27) test slice::swap_with_slice_rgb_3000 ... bench: 16,215 ns/iter (+/- 176) test slice::swap_with_slice_u8_30 ... bench: 312 ns/iter (+/- 9) test slice::swap_with_slice_u8_3000 ... bench: 5,401 ns/iter (+/- 123) test slice::swap_with_slice_usize_30 ... bench: 368 ns/iter (+/- 3) test slice::swap_with_slice_usize_3000 ... bench: 28,472 ns/iter (+/- 3,913) ``` ## After ``` running 10 tests test slice::swap_with_slice_4x_usize_30 ... bench: 868 ns/iter (+/- 36) test slice::swap_with_slice_4x_usize_3000 ... bench: 99,642 ns/iter (+/- 1,507) test slice::swap_with_slice_5x_usize_30 ... bench: 1,194 ns/iter (+/- 11) test slice::swap_with_slice_5x_usize_3000 ... bench: 139,761 ns/iter (+/- 5,018) test slice::swap_with_slice_rgb_30 ... bench: 324 ns/iter (+/- 6) test slice::swap_with_slice_rgb_3000 ... bench: 15,962 ns/iter (+/- 287) test slice::swap_with_slice_u8_30 ... bench: 281 ns/iter (+/- 5) test slice::swap_with_slice_u8_3000 ... bench: 5,324 ns/iter (+/- 40) test slice::swap_with_slice_usize_30 ... bench: 275 ns/iter (+/- 5) test slice::swap_with_slice_usize_3000 ... bench: 28,277 ns/iter (+/- 277) ``` </detail>
…ointer, r=michaelwoerister properly handle fat pointers to uninhabitable types Calculate the pointee metadata size by using `tcx.struct_tail_erasing_lifetimes` instead of duplicating the logic in `fat_pointer_kind`. Open to alternatively suggestions on how to fix this. Fixesrust-lang#94149 r? ```@michaelwoerister``` since you touched this code last, I think!
…i-obk
Normalize main return type during mono item collection & codegen
The issue can be observed with `-Zprint-mono-items=lazy` in:
```rust
#![feature(termination_trait_lib)]
fn main() -> impl std::process::Termination { }
```
```
BEFORE: MONO_ITEM fn std::rt::lang_start::<impl std::process::Termination> ```@@``` t.93933fa2-cgu.2[External]
AFTER: MONO_ITEM fn std::rt::lang_start::<()> ```@@``` t.df56e625-cgu.1[External]
```update auto trait lint for `PhantomData` cc rust-lang#93367 (comment)
…unescaping, r=petrochenkov Improve string literal unescaping Some easy wins that affect a few popular crates. r? ``@matklad``
…etrochenkov Avoid emitting full macro body into JSON errors While investigating rust-lang#94322, it was noted that currently the JSON diagnostics for macro backtraces include the full def_site span -- the whole macro body. It seems like this shouldn't be necessary, so this PR adjusts the span to just be the "guessed head", typically the macro name. It doesn't look like we keep enough information to synthesize a nicer span here at this time. Atop rust-lang#92123, this reduces output for the src/test/ui/suggestions/missing-lifetime-specifier.rs test from 660 KB to 156 KB locally.
Dylan-DPC
commented
Feb 24, 2022
MemberAuthor
duplicate by mistake |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
9 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Successful merges:
can_type_implement_copy#93714 (better ObligationCause for normalization errors incan_type_implement_copy)--check-cfgimplementation #94175 (Improve--check-cfgimplementation)swap_nonoverlapping#94212 (Stop manually SIMDing inswap_nonoverlapping)PhantomData#94315 (update auto trait lint forPhantomData)Failed merges:
r? @ghost
@rustbot modify labels: rollup
Create a similar rollup