Bug 23077 – codegen cannot generage XMM load/store for optimized operation that uses byte/short/...

The following code produces an Illegal instruction (core dumped) on the compiler: float bug(float x) { short i = *cast(short*)&x; ++i; return *cast(float*)&i; } The codegen try to generate optimized calls with XMM opcodes but fail to do it. This might happen to other types too.

Blocking `intel-intrinsics` buildkite job.

@ljmf00 updated dlang/dmd pull request #14058 "fix(backend): define XMM load/store codegen for short type variants" fixing this issue: - fix(backend): define XMM load/store codegen for short type variants The codegen try to generate optimized calls with XMM opcodes but fail to do it. This patch adds logic to generate XMM load/store opcodes for TYshort,TYushort,TYchar,TYwchar,TYdchar. It also adds tests for previous similar ICEs. Fix issue 23077. Reference: https://issues.dlang.org/show_bug.cgi?id=23077 Reference: https://github.com/AuburnSounds/intel-intrinsics/commit/047e9f01ae4092600f99abab1a34e32db26b3a4f Signed-off-by: Luís Ferreira <[email protected]> https://github.com/dlang/dmd/pull/14058

*** This issue has been marked as a duplicate of issue 23084 ***

@WalterBright created dlang/dmd pull request #14081 "fix Issue 23077 - codegen cannot generage XMM load/store for optimize…" fixing this issue: - fix Issue 23077 - codegen cannot generage XMM load/store for optimized operation that uses byte/short https://github.com/dlang/dmd/pull/14081

*** Issue 23084 has been marked as a duplicate of this issue. ***

dlang/dmd pull request #14081 "fix Issue 23077 - codegen cannot generage XMM load/store for optimize…" was merged into stable: - 843c0d5cb44812121c227376e0948c5fa97606bd by Walter Bright: fix Issue 23077 - codegen cannot generage XMM load/store for optimized operation that uses byte/short https://github.com/dlang/dmd/pull/14081

*** Issue 23085 has been marked as a duplicate of this issue. ***