[ARM] Simplify extract of VMOVDRR
authorDavid Green <david.green@arm.com>
Mon, 1 Feb 2021 10:24:57 +0000 (10:24 +0000)
committerDavid Green <david.green@arm.com>
Mon, 1 Feb 2021 10:24:57 +0000 (10:24 +0000)
commit6ab792b68d5779e1668a2bd6ecd0a52c91b56387
treefb6c0ad5c59ab00f1f6d02c6f9a204648ba62ee1
parent4648098f97fa2a7c08c04632c70cf29293528812
[ARM] Simplify extract of VMOVDRR

Under SoftFP calling conventions, we can be left with
extract(bitcast(BUILD_VECTOR(VMOVDRR(a, b), ..))) patterns that can
simplify to a or b, depending on the extract lane.

Differential Revision: https://reviews.llvm.org/D94990
llvm/lib/Target/ARM/ARMISelLowering.cpp
llvm/test/CodeGen/Thumb2/mve-soft-float-abi.ll