rust/tests/mir-opt/dataflow-const-prop/default_boxed_slice.main.GVN.64bit.panic-abort.diff, branch masterhttps://github.com/rust-lang/rust
http://git.dreamy.place/mirrors/rust/atom?h=master2025-01-19T02:44:41+00:00Update tests for std::simd subtree sync2025-01-19T02:44:41+00:00Caleb Zulawskicaleb.zulawski@gmail.com2025-01-19T02:44:41+00:00urn:sha1:52b42d71876d509340c432671c63217427eef165Less unsafe in `dangling`/`without_provenance`2025-01-16T06:17:57+00:00Scott McMurrayscottmcm@users.noreply.github.com2025-01-10T21:00:45+00:00urn:sha1:c18718c9c28cc4c700312775c2b7f87cd79d3063[mir-opt] GVN some more transmute cases2025-01-09T02:46:30+00:00Scott McMurrayscottmcm@users.noreply.github.com2024-11-22T01:59:43+00:00urn:sha1:8dcc676c9283b040d1ef6d66f156ff2212544b10
We already did `Transmute`-then-`PtrToPtr`; this adds the nearly-identical `PtrToPtr`-then-`Transmute`.
It also adds `transmute(Foo(x))` → `transmute(x)`, when `Foo` is a single-field transparent type. That's useful for things like `NonNull { pointer: p }.as_ptr()`.
Found these as I was looking at MCP807-related changes.
be even more precise about "cast" vs "coercion"2024-09-24T21:12:02+00:00Lukas Markeffsky@2024-09-15T17:35:06+00:00urn:sha1:bd31e3ed70b75e5936880b18f6b5d1c8f5ee344fBless *all* the mir-opt tests2024-08-18T23:07:33+00:00Scott McMurrayscottmcm@users.noreply.github.com2024-08-18T21:26:34+00:00urn:sha1:99cb0c6bc399fb94a0ddde7e9b38e9c00d523badBless mir-opt for excluded alloc bytes2024-06-26T22:30:47+00:00Josh Stonejistone@redhat.com2024-06-14T23:05:22+00:00urn:sha1:7c3673ff6f2ff4e8e85344c091c9feb4b5da1290Ban `ArrayToPointer` and `MutToConstPointer` from runtime MIR2024-06-19T17:44:01+00:00Scott McMurrayscottmcm@users.noreply.github.com2024-06-12T08:22:31+00:00urn:sha1:4630d1b23b996793e9188490cc99b1576176233f
Apparently MIR borrowck cares about at least one of these for checking variance.
In runtime MIR, though, there's no need for them as `PtrToPtr` does the same thing.
(Banning them simplifies passes like GVN that no longer need to handle multiple cast possibilities.)
At debuginfo=0, don't inline debuginfo when inlining2024-04-18T16:35:35+00:00Scott McMurrayscottmcm@users.noreply.github.com2024-04-15T01:22:35+00:00urn:sha1:d05545c05d9399a70a796cfd18b60416909e2d6bRemove MIR unsafe check2024-04-03T08:50:12+00:00Matthew Jaspermjjasper1@gmail.com2024-02-27T11:57:52+00:00urn:sha1:a277c901d95c5fcbb3dbd6608731eebb9e1a01ce
This also remove safety information from MIR.
refactor check_{lang,library}_ub: use a single intrinsic, put policy into library2024-03-23T17:45:05+00:00Ralf Jungpost@ralfj.de2024-03-17T09:29:02+00:00urn:sha1:61775304200826b412f18fc87218161fb8a0dc5c