HomePhabricator

[AMDGPU] Fix DPP combiner

Description

[AMDGPU] Fix DPP combiner

Differential revision: https://reviews.llvm.org/D55444

dpp move with uses and old reg initializer should be in the same BB.
bound_ctrl:0 is only considered when bank_mask and row_mask are fully enabled (0xF). Otherwise the old register value is checked for identity.
Added add, subrev, and, or instructions to the old folding function.
Kill flag is cleared for the src0 (DPP register) as it may be copied into more than one user.

The pass is still disabled by default.

Details

Committed
vpykhtinFeb 8 2019, 3:59 AM
Differential Revision
D55444: AMDGPU: Fix DPP combiner
Parents
rL353512: Merging r351387, r351765, and r353374:
Branches
Unknown
Tags
Unknown