[llvm] [GlobalIsel] Push cast through build vector (PR #104634)

[Dhruv Chawla via llvm-commits]

Thorsten =?utf-8?q?Schütt?= <schuett at gmail.com>,
Thorsten =?utf-8?q?Schütt?= <schuett at gmail.com>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/104634 at github.com>


dc03-work wrote:

> Could I get some eyes on the assembler diffs? Limiting to truncate seems to be not to bad.

I still see the same major regressions (eg. in `llvm/test/CodeGen/AArch64/arm64-subvector-extend.ll`), I don't think anything has really improved.

https://github.com/llvm/llvm-project/pull/104634