treat host effect params as erased in codegen
This fixes the changes brought to codegen tests when effect params are added to libcore, by not attempting to monomorphize functions that get the host param by being `const fn`.
r? `@oli-obk`
This fixes the changes brought to codegen tests when effect params are
added to libcore, by not attempting to monomorphize functions that get
the host param by being `const fn`.
Rollup of 6 pull requests
Successful merges:
- #115736 (Remove `verbose_generic_activity_with_arg`)
- #115771 (cleanup leftovers of const_err lint)
- #115798 (add helper method for finding the one non-1-ZST field)
- #115812 (Merge settings.css into rustdoc.css)
- #115815 (fix: return early when has tainted in mir pass)
- #115816 (Disabled socketpair for Vita)
r? `@ghost`
`@rustbot` modify labels: rollup
Remove `verbose_generic_activity_with_arg`
This removes `verbose_generic_activity_with_arg` and changes users to `generic_activity_with_arg`. This keeps the output of `-Z time` readable while these repeated events are still available with the self profiling mechanism.
fix: return early when has tainted in mir-lint
Fixes#115203
`a[..]` is of indeterminate size, it had been reported error during borrow check, therefore we skip the mir lint process.
coverage: Simplify the `coverageinfo` query
The `coverageinfo` query walks through a `mir::Body`'s statements to find the total number of coverage counter IDs and coverage expression IDs that have been used, as this information is needed by coverage codegen.
This PR makes 3 nice simplifications to that query:
- Extract a common iterator over coverage statements, shared by both coverage-related queries
- Simplify the query's visitor from two passes to just one pass
- Explicitly track the highest seen IDs in the visitor, and only convert to a count right at the end
I also updated some related comments. Some had been invalidated by these changes, while others had already been invalidated by previous coverage changes.
Don't report any errors in `lower_intrinsics`.
Intrinsics should have been type checked earlier.
This is part of moving all mir-opt diagnostics early enough so that they are reliably emitted even in check builds: https://github.com/rust-lang/rust/issues/49292#issuecomment-1692212095
Both of the coverage queries can now use this one helper function to iterate
over all of the `mir::Coverage` payloads in the statements of a `mir::Body`.
Represent MIR composite debuginfo as projections instead of aggregates
Composite debuginfo for MIR is currently represented as
```
debug name => Type { projection1 => place1, projection2 => place2 };
```
ie. a single `VarDebugInfo` object with that name, and its value a `VarDebugInfoContents::Composite`.
This PR proposes to reverse the representation to be
```
debug name.projection1 => place1;
debug name.projection2 => place2;
```
ie. multiple `VarDebugInfo` objects with each their projection.
This simplifies the handling of composite debuginfo by the compiler by avoiding weird nesting.
Based on https://github.com/rust-lang/rust/pull/115139
Use relative positions inside a SourceFile.
This allows to remove the normalization of start positions for hashing, and simplify allocation of global address space.
cc `@Zoxc`
Rollup of 5 pull requests
Successful merges:
- #115353 (Emit error instead of ICE when optimized MIR is missing)
- #115488 (Take `&mut Results` in `ResultsVisitor`)
- #115492 (Allow `large_assignments` for Box/Arc/Rc initialization)
- #115519 (Don't ICE on associated type projection without feature gate in new solver)
- #115534 (Expose more information with DefId in smir)
r? `@ghost`
`@rustbot` modify labels: rollup
Fix inlining with -Zalways-encode-mir
Only inline functions that are considered eligible for inlining
by the reachability pass.
This constraint was previously indirectly enforced by only exporting MIR
of eligible functions, but that approach doesn't work with
-Zalways-encode-mir enabled.
Only inline functions that are considered eligible for inlining
by the reachability pass.
This constraint was previously indirectly enforced by only exporting MIR
of eligible functions, but that approach doesn't work with
-Zalways-encode-mir enabled.
Don't do intra-pass validation on MIR shims
Fixes#114375
In the test that was committed, we end up generating the drop shim for `struct Foo` that looks like:
```
fn std::ptr::drop_in_place(_1: *mut Foo) -> () {
let mut _0: ();
bb0: {
goto -> bb5;
}
bb1: {
return;
}
bb2 (cleanup): {
resume;
}
bb3: {
goto -> bb1;
}
bb4 (cleanup): {
drop(((*_1).0: foo::WrapperWithDrop<()>)) -> [return: bb2, unwind terminate];
}
bb5: {
drop(((*_1).0: foo::WrapperWithDrop<()>)) -> [return: bb3, unwind: bb2];
}
}
```
In `bb4` and `bb5`, we assert that `(*_1).0` has type `WrapperWithDrop<()>`. However, In a user-facing param env, the type is actually `WrapperWithDrop<Tait>`. These types are not equal in a user-facing param-env (and can't be made equal even if we use `DefiningAnchor::Bubble`, since it's a non-local TAIT).
coverage: Give the instrumentor its own counter type, separate from MIR
Within the MIR representation of coverage data, `CoverageKind` is an important part of `StatementKind::Coverage`, but the `InstrumentCoverage` pass also uses it heavily as an internal data structure. This means that any change to `CoverageKind` also needs to update all of the internal parts of `InstrumentCoverage` that manipulate it directly, making the MIR representation difficult to modify.
---
This change fixes that by giving the instrumentor its own `BcbCounter` type for internal use, which is then converted to a `CoverageKind` when injecting coverage information into MIR.
The main change is mostly mechanical, because the initial `BcbCounter` is drop-in compatible with `CoverageKind`, minus the unnecessary `CoverageKind::Unreachable` variant.
I've then removed the `function_source_hash` field from `BcbCounter::Counter`, as a small example of how the two types can now usefully differ from each other. Every counter in a MIR-level function should have the same source hash, so we can supply the hash during the conversion to `CoverageKind::Counter` instead.
---
*Background:* BCB stands for “basic coverage block”, which is a node in the simplified control-flow graph used by coverage instrumentation. The instrumentor pass uses the function's actual MIR control-flow graph to build a simplified BCB graph, then assigns coverage counters and counter expressions to various nodes/edges in that simplified graph, and then finally injects corresponding coverage information into the underlying MIR.
Add MIR validation for unwind out from nounwind functions + fixes to make validation pass
`@Nilstrieb` This is the MIR validation you asked in https://github.com/rust-lang/rust/pull/112403#discussion_r1222739722.
Two passes need to be fixed to get the validation to pass:
* `RemoveNoopLandingPads` currently unconditionally introduce a resume block (even there is none to begin with!), changed to not do that
* Generator state transform introduces a `assert` which may unwind, and its drop elaboration also introduces many new `UnwindAction`s, so in this case run the AbortUnwindingCalls after the transformation.
I believe this PR should also fixRust-for-Linux/linux#1016, cc `@ojeda`
r? `@Nilstrieb`
This shows one small benefit of separating `BcbCounter` from `CoverageKind`.
The function source hash will be the same for all counters within a function,
so instead of passing it through `CoverageCounters` and storing it in every
counter, we can just supply it during the final conversion to `CoverageKind`.
Otherwise the file name generated for generator_drop will become
core.ptr-drop_in_place.[generator@<FILEPATH>_<NUMBERS>].generator_drop.0.mir
instead of main-{closure#0}.generator_drop.0.mir which breaks a mir-opt
test.
Normalize before checking if local is freeze in `deduced_param_attrs`
Not normalizing the local type eagerly results in possibly exponential amounts of normalization happening downstream in `is_freeze_raw`.
Fixes#113372
Storing coverage counter information in `CoverageCounters` has a few advantages
over storing it directly inside BCB graph nodes:
- The graph doesn't need to be mutable when making the counters, making it
easier to see that the graph itself is not modified during this step.
- All of the counter data is clearly visible in one place.
- It becomes possible to use a representation that doesn't correspond 1:1 to
graph nodes, e.g. storing all the edge counters in a single hashmap instead of
several.
feat: `riscv-interrupt-{m,s}` calling conventions
Similar to prior support added for the mips430, avr, and x86 targets this change implements the rough equivalent of clang's [`__attribute__((interrupt))`][clang-attr] for riscv targets, enabling e.g.
```rust
static mut CNT: usize = 0;
pub extern "riscv-interrupt-m" fn isr_m() {
unsafe {
CNT += 1;
}
}
```
to produce highly effective assembly like:
```asm
pub extern "riscv-interrupt-m" fn isr_m() {
420003a0: 1141 addi sp,sp,-16
unsafe {
CNT += 1;
420003a2: c62a sw a0,12(sp)
420003a4: c42e sw a1,8(sp)
420003a6: 3fc80537 lui a0,0x3fc80
420003aa: 63c52583 lw a1,1596(a0) # 3fc8063c <_ZN12esp_riscv_rt3CNT17hcec3e3a214887d53E.0>
420003ae: 0585 addi a1,a1,1
420003b0: 62b52e23 sw a1,1596(a0)
}
}
420003b4: 4532 lw a0,12(sp)
420003b6: 45a2 lw a1,8(sp)
420003b8: 0141 addi sp,sp,16
420003ba: 30200073 mret
```
(disassembly via `riscv64-unknown-elf-objdump -C -S --disassemble ./esp32c3-hal/target/riscv32imc-unknown-none-elf/release/examples/gpio_interrupt`)
This outcome is superior to hand-coded interrupt routines which, lacking visibility into any non-assembly body of the interrupt handler, have to be very conservative and save the [entire CPU state to the stack frame][full-frame-save]. By instead asking LLVM to only save the registers that it uses, we defer the decision to the tool with the best context: it can more accurately account for the cost of spills if it knows that every additional register used is already at the cost of an implicit spill.
At the LLVM level, this is apparently [implemented by] marking every register as "[callee-save]," matching the semantics of an interrupt handler nicely (it has to leave the CPU state just as it found it after its `{m|s}ret`).
This approach is not suitable for every interrupt handler, as it makes no attempt to e.g. save the state in a user-accessible stack frame. For a full discussion of those challenges and tradeoffs, please refer to [the interrupt calling conventions RFC][rfc].
Inside rustc, this implementation differs from prior art because LLVM does not expose the "all-saved" function flavor as a calling convention directly, instead preferring to use an attribute that allows for differentiating between "machine-mode" and "superivsor-mode" interrupts.
Finally, some effort has been made to guide those who may not yet be aware of the differences between machine-mode and supervisor-mode interrupts as to why no `riscv-interrupt` calling convention is exposed through rustc, and similarly for why `riscv-interrupt-u` makes no appearance (as it would complicate future LLVM upgrades).
[clang-attr]: https://clang.llvm.org/docs/AttributeReference.html#interrupt-risc-v
[full-frame-save]: 9281af2ecf/src/lib.rs (L440-L469)
[implemented by]: b7fb2a3fec/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp (L61-L67)
[callee-save]: 973f1fe7a8/llvm/lib/Target/RISCV/RISCVCallingConv.td (L30-L37)
[rfc]: https://github.com/rust-lang/rfcs/pull/3246
Similar to prior support added for the mips430, avr, and x86 targets
this change implements the rough equivalent of clang's
[`__attribute__((interrupt))`][clang-attr] for riscv targets, enabling
e.g.
```rust
static mut CNT: usize = 0;
pub extern "riscv-interrupt-m" fn isr_m() {
unsafe {
CNT += 1;
}
}
```
to produce highly effective assembly like:
```asm
pub extern "riscv-interrupt-m" fn isr_m() {
420003a0: 1141 addi sp,sp,-16
unsafe {
CNT += 1;
420003a2: c62a sw a0,12(sp)
420003a4: c42e sw a1,8(sp)
420003a6: 3fc80537 lui a0,0x3fc80
420003aa: 63c52583 lw a1,1596(a0) # 3fc8063c <_ZN12esp_riscv_rt3CNT17hcec3e3a214887d53E.0>
420003ae: 0585 addi a1,a1,1
420003b0: 62b52e23 sw a1,1596(a0)
}
}
420003b4: 4532 lw a0,12(sp)
420003b6: 45a2 lw a1,8(sp)
420003b8: 0141 addi sp,sp,16
420003ba: 30200073 mret
```
(disassembly via `riscv64-unknown-elf-objdump -C -S --disassemble ./esp32c3-hal/target/riscv32imc-unknown-none-elf/release/examples/gpio_interrupt`)
This outcome is superior to hand-coded interrupt routines which, lacking
visibility into any non-assembly body of the interrupt handler, have to
be very conservative and save the [entire CPU state to the stack
frame][full-frame-save]. By instead asking LLVM to only save the
registers that it uses, we defer the decision to the tool with the best
context: it can more accurately account for the cost of spills if it
knows that every additional register used is already at the cost of an
implicit spill.
At the LLVM level, this is apparently [implemented by] marking every
register as "[callee-save]," matching the semantics of an interrupt
handler nicely (it has to leave the CPU state just as it found it after
its `{m|s}ret`).
This approach is not suitable for every interrupt handler, as it makes
no attempt to e.g. save the state in a user-accessible stack frame. For
a full discussion of those challenges and tradeoffs, please refer to
[the interrupt calling conventions RFC][rfc].
Inside rustc, this implementation differs from prior art because LLVM
does not expose the "all-saved" function flavor as a calling convention
directly, instead preferring to use an attribute that allows for
differentiating between "machine-mode" and "superivsor-mode" interrupts.
Finally, some effort has been made to guide those who may not yet be
aware of the differences between machine-mode and supervisor-mode
interrupts as to why no `riscv-interrupt` calling convention is exposed
through rustc, and similarly for why `riscv-interrupt-u` makes no
appearance (as it would complicate future LLVM upgrades).
[clang-attr]: https://clang.llvm.org/docs/AttributeReference.html#interrupt-risc-v
[full-frame-save]: 9281af2ecf/src/lib.rs (L440-L469)
[implemented by]: b7fb2a3fec/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp (L61-L67)
[callee-save]: 973f1fe7a8/llvm/lib/Target/RISCV/RISCVCallingConv.td (L30-L37)
[rfc]: https://github.com/rust-lang/rfcs/pull/3246
Make module inner and function run_analysis_to_runtime_passes in
rustc_mir_transform public to allow re-implementing the query from the
rust compiler interface.
Make `unconditional_recursion` warning detect recursive drops
Closes#55388
Also closes#50049 unless we want to keep it for the second example which this PR does not solve, but I think it is better to track that work in #57965.
r? `@oli-obk` since you are the mentor for #55388
Unresolved questions:
- [x] There are two false positives that must be fixed before merging (see diff). I suspect the best way to solve them is to perform analysis after drop elaboration instead of before, as now, but I have not explored that any further yet. Could that be an option? **Answer:** Yes, that solved the problem.
`@rustbot` label +T-compiler +C-enhancement +A-lint
Steal MIR for CTFE when possible.
Some bodies, like constants, have CTFE MIR but no optimized MIR.
In that case, have `mir_for_ctfe` steal the MIR instead of cloning it.
Add documentation to has_deref
Documentation of `has_deref` needed some polish to be more clear about where it should be used and what's it's purpose.
cc https://github.com/rust-lang/rust/issues/114401
r? `@RalfJung`
Do not run ConstProp on mir_for_ctfe.
This pass does not seem to be useful any more. The const-prop lints are now run by `tcx.mir_drops_elaborated_and_const_checked`, and the const-prop opt should never emit any diagnostic.
Forbid old-style `simd_shuffleN` intrinsics
Don't merge before https://github.com/rust-lang/packed_simd/pull/350 has made its way to crates.io
We used to support specifying the lane length of simd_shuffle ops by attaching the lane length to the name of the intrinsic (like `simd_shuffle16`). After this PR, you cannot do that anymore, and need to instead either rely on inference of the `idx` argument type or specify it as `simd_shuffle::<_, [u32; 16], _>`.
r? `@workingjubilee`
Operand types are now tracked explicitly, so there is no need to reserve ID 0
for the special always-zero counter.
As part of the renumbering, this change fixes an off-by-one error in the way
counters were counted by the `coverageinfo` query. As a result, functions
should now have exactly the number of counters they actually need, instead of
always having an extra counter that is never used.
Operand types are now tracked explicitly, so there is no need for expression
IDs to avoid counter IDs by descending from `u32::MAX`. Instead they can just
count up from 0, and can be used directly as indices when necessary.
Because the three kinds of operand are now distinguished explicitly, we no
longer need fiddly code to disambiguate counter IDs and expression IDs based on
the total number of counters/expressions in a function.
This does increase the size of operands from 4 bytes to 8 bytes, but that
shouldn't be a big deal since they are mostly stored inside boxed structures,
and the current coverage code is not particularly size-optimized anyway.
Substitute types before checking inlining compatibility.
Addresses https://github.com/rust-lang/rust/issues/112332 and https://github.com/rust-lang/rust/issues/113781
I don't have a minimal test, but I this seems to remove the ICE locally.
This whole pre-inlining validation mirrors the "real" MIR validation pass to verify that inlined MIR will still pass validation.
The debuginfo loop is added because MIR validation check projections in debuginfo.
Likewise, MIR validation only checks `is_subtype`, so there is no reason for a stronger check.
The types were not being substituted in `check_equal`, so we were not bailing out of inlining if the substituted MIR callee body would not pass validation.
It makes it sound like the `ExprKind` and `Rvalue` are supposed to represent all pointer related
casts, when in reality their just used to share a some enum variants. Make it clear there these
are only coercion to make it clear why only some pointer related "casts" are in the enum.
Rewrite `UnDerefer`
Currently, `UnDerefer` is used by drop elaboration to undo the effects of the `Derefer` pass. However, it just recreates the original places with derefs in the middle of the projection. Because `ProjectionElem::Deref` is intended to be removed completely in the future, this will not work forever.
This PR introduces a `deref_chain` method that returns the places behind `DerefTemp` locals in a place and rewrites the move path code to use this. In the process, `UnDerefer` was merged into `MovePathLookup`. Now that move paths use the same places as in the MIR, the other uses of `UnDerefer` no longer require it.
See #98145
cc `@ouz-a`
r? `@oli-obk`
mir opt + codegen: handle subtyping
fixes#107205
the same issue was caused in multiple places:
- mir opts: both copy and destination propagation
- codegen: assigning operands to locals (which also propagates values)
I changed codegen to always update the type in the operands used for locals which should guard against any new occurrences of this bug going forward. I don't know how to make mir optimizations more resilient here. Hopefully the added tests will be enough to detect any trivially wrong optimizations going forward.
Use `CoverageKind::as_operand_id` instead of manually reimplementing it
These two pieces of code are functionally equivalent to the `CoverageKind::as_operand_id` method that already exists, and is already used elsewhere in this file.
This slightly reduces the amount of code that manually pattern-matches on `CoverageKind`.
Migrate `TyCtxt::predicates_of` and `ParamEnv::caller_bounds` to `Clause`
The last big change in the series.
I will follow-up with additional filed issues once this PR lands:
- [ ] Investigate making `TypeFoldable<TyCtxt<'tcx>> for ty::Clause<'tcx>` implementation less weird: 2efe091705/compiler/rustc_middle/src/ty/structural_impls.rs (L672)
- [ ] Clean up the elaborator since it should only be emitting child clauses, not predicates
- [ ] Rename identifiers like `pred` and `predicates` to `clause` if they're actually clauses around the codebase
- [ ] Validate that all of the `ToPredicate` impls are acutally still needed, or prune them if they're not
r? `@ghost` until the other branch lands
Add retag in MIR transform: `Adt` for `Unique` may contain a reference
Following #112662 , `may_contain_reference` in `rustc_mir_transform::add_retag` underapproximates too much the types that require retagging.
r? ``@RalfJung``
Add a fully fledged `Clause` type, rename old `Clause` to `ClauseKind`
Does two basic things before I put up a more delicate set of PRs (along the lines of #112714, but hopefully much cleaner) that migrate existing usages of `ty::Predicate` to `ty::Clause` (`predicates_of`/`item_bounds`/`ParamEnv::caller_bounds`).
1. Rename `Clause` to `ClauseKind`, so it's parallel with `PredicateKind`.
2. Add a new `Clause` type which is parallel to `Predicate`.
* This type exposes `Clause::kind(self) -> Binder<'tcx, ClauseKind<'tcx>>` which is parallel to `Predicate::kind` 😸
The new `Clause` type essentially acts as a newtype wrapper around `Predicate` that asserts that it is specifically a `PredicateKind::Clause`. Turns out from experimentation[^1] that this is not negative performance-wise, which is wonderful, since this a much simpler design than something that requires encoding the discriminant into the alignment bits of a predicate kind, or something else like that...
r? ``@lcnr`` or ``@oli-obk``
[^1]: https://github.com/rust-lang/rust/pull/112714#issuecomment-1595653910
Switch the BB CFG cache from postorder to RPO
The `BasicBlocks` CFG cache is interesting:
- it stores a postorder, but `traversal::postorder` doesn't use it
- `traversal::reverse_postorder` does traverse the postorder cache backwards
- we do more RPO traversals than postorder traversals (around 20x on the perf.rlo benchmarks IIRC) but it's not cached
- a couple places here and there were manually reversing the non-cached postorder traversal
This PR switches the order of the cache, and makes a bit more use of it. This is a tiny win locally, but it's also for consistency and aesthetics.
r? `@ghost`
Launch a non-unwinding panic for misaligned pointer deref
This panic already never unwinds, but that's only because it always hits the unwind guard that's created by our `UnwindAction::Terminate`. Hitting the unwind guard generates a huge double-panic backtrace. Now we generate a normal-looking panic message when this check is hit.
r? `@thomcc`
Prevent `.eh_frame` from being emitted for `-C panic=abort`
Since `CheckAlignment` pass is after the `AbortUnwindingCalls` pass, the `UnwindAction::Terminate` inserted in it has no chance to be converted to `UnwindAction::Unreachable` anymore, causing us to emit landing pads that are not necessary. Although these landing pads can themselves be eliminated by LLVM, `.eh_frame` sections are still generated. This causes trouble for Rust-for-Linux project recently.
This PR changes it to generate `UnwindAction::Terminate` when we opt for `-Cpanic=unwind`, and `UnwindAction::Unreachable` for `-Cpanic=abort`.
`@ojeda`
Add MVP suggestion for `unsafe_op_in_unsafe_fn`
Rebase of https://github.com/rust-lang/rust/pull/99827
cc tracking issue https://github.com/rust-lang/rust/issues/71668
No real changes since the original PR, just migrated the new suggestion to use fluent messages and added a couple more testcases, AFAICT from the discussion there were no outstanding changes requested.
Write to stdout if `-` is given as output file
With this PR, if `-o -` or `--emit KIND=-` is provided, output will be written to stdout instead. Binary output (those of type `obj`, `llvm-bc`, `link` and `metadata`) being written this way will result in an error unless stdout is not a tty. Multiple output types going to stdout will trigger an error too, as they will all be mixded together.
This implements https://github.com/rust-lang/compiler-team/issues/431
The idea behind the changes is to introduce an `OutFileName` enum that represents the output - be it a real path or stdout - and to use this enum along the code paths that handle different output types.
Take MIR dataflow analyses by mutable reference
The main motivation here is any analysis requiring dynamically sized scratch memory to work. One concrete example would be pointer target tracking, where tracking the results of a dereference can result in multiple possible targets. This leads to processing multi-level dereferences requiring the ability to handle a changing number of potential targets per step. A (simplified) function for this would be `fn apply_deref(potential_targets: &mut Vec<Target>)` which would use the scratch space contained in the analysis to send arguments and receive the results.
The alternative to this would be to wrap everything in a `RefCell`, which is what `MaybeRequiresStorage` currently does. This comes with a small perf cost and loses the compiler's guarantee that we don't try to take multiple borrows at the same time.
For the implementation:
* `AnalysisResults` is an unfortunate requirement to avoid an unconstrained type parameter error.
* `CloneAnalysis` could just be `Clone` instead, but that would result in more work than is required to have multiple cursors over the same result set.
* `ResultsVisitor` now takes the results type on in each function as there's no other way to have access to the analysis without cloning it. This could use an associated type rather than a type parameter, but the current approach makes it easier to not care about the type when it's not necessary.
* `MaybeRequiresStorage` now no longer uses a `RefCell`, but the graphviz formatter now does. It could be removed, but that would require even more changes and doesn't really seem necessary.
If `-o -` or `--emit KIND=-` is provided, output will be written
to stdout instead. Binary output (`obj`, `llvm-bc`, `link` and
`metadata`) being written this way will result in an error unless
stdout is not a tty. Multiple output types going to stdout will
trigger an error too, as they will all be mixded together.
Only check inlining counter after recursing.
This PR aims to reduce the strength of https://github.com/rust-lang/rust/pull/105119 even more.
In the current implementation, we check the inline count before recursing. This means that we never actually reach inlining depth 3.
This PR checks the counter after recursion, to give a chance to inline at depth >= 3.
r? `@scottmcm`
cc `@JakobDegen`
Use translatable diagnostics in `rustc_const_eval`
This PR:
* adds a `no_span` parameter to `note` / `help` attributes when using `Subdiagnostic` to allow adding notes/helps without using a span
* has minor tweaks and changes to error messages
Enable ConstGoto and SeparateConstSwitch passes by default
These 2 passes implement a limited form of jump-threading.
Filing this PR to see if enabling them would be lighter than https://github.com/rust-lang/rust/pull/107009.
Enable ScalarReplacementOfAggregates in optimized builds
Like MatchBranchSimplification, this pass is known to produce significant runtime improvements in Cranelift artifacts, and I believe based on the perf runs here that the primary effect of this pass is to empower MatchBranchSimplification. ScalarReplacementOfAggregates on its own has little effect on anything, but when this was rebased up to include https://github.com/rust-lang/rust/pull/112001 we started seeing significant and majority-positive results.
Based on the fact that we see most of the regressions in debug builds (https://github.com/rust-lang/rust/pull/112002#issuecomment-1566270144) and some rather significant ones in cycles and wall time, I'm only enabling this in optimized builds at the moment.
