* Remove reserved_reg functionality.
This wasn't implemented, and if we need it in the future, it seems like
it would be better to extend the concept of global values to cover this.
* Use GlobalValue::reserved_value() for sentinal values.
* Added initial Rust codegen-meta implementation.
* Replace 'Cretonne' in comments.
* Prevent iterator overflow.
* 1.25.0 compatibility changes.
* Implemented debug traits for type variants.
* Added consistent comments.
* Cleaned up a loop via clippy fix.
* Added new license to codegen-meta Cargo.toml
* Edited lane type iterator `next` method.
* Removed functions that are not needed in Rust, and edited desc.
* Debug trait derived for valuetype.
* Added comments for iterator types in the base types submodule.
* Numbering is now handled in the cdsl/types.rs file.
* Moved type number logic into cdsl/types.
* Repeating the lane change cleanup.
* Removed codegen-meta crate from codegen deps.
* Typo fix.
* Addressing a patch note.
* Addressing patch note.
* Lowercase in vector names.
* Fixing a comment bug.
* Added a copy of the license file.
* Formatting changes.
* Cleaned up the vector type numbering.
* 1.25 compatibility.
* Fixed pattern match arms.
The individual crates are published separately from the main repository
on crates.io. To ensure that a copy of the LICENSE file accompanies all
published copies of the code, give each crate its own LICENSE file.
- don't generate "&& true" when generating instruction eq() fn;
- use more Self;
- use subsec_millis instead of subsec_nanos and divide;
- coalesce two ifs;
* Don't implement Eq and Hash for EntityList
* Generate eq and hash methods for InstructionData
* Use the eq() and hash() methods of InstructionData in simple_gvn
* Now diagnosing missing vmctx arguments (fixes#376).
* Added filetest for fix of #376.
* Respect formatting rules in verifier/mod.rs.
* Added parameters for each use of vmctx in test files.
* Added comments on additions on vmctx verifications.
The documentation for these instructions suggests that immediate values
which could lead to trapping should be invalid.
While it seems nice to have these instructions be always non-trapping,
it's also nice to say that the `_imm` forms of instructions are
interchangeable with the corresponding non-`_imm` forms accompanied
by `iconst` instructions.
This requires splitting X86PCRel4 into two separate relocations, to
distinguish the case where the instruction is a call, as Mach-O uses a
different relocation in that case.
This also makes it explicit that only x86-64 relocations are supported
currently.
