AMDGPU/GlobalISel: Mark G_ATOMICRMW_{NAND|FSUB} as lower
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Sat, 25 Jul 2020 14:17:49 +0000 (10:17 -0400)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Mon, 27 Jul 2020 22:47:40 +0000 (18:47 -0400)
commitce944af33c1e011b69665b6892eb3ea142afcdf0
tree35cbf4c13b00bb517e73aa8ef17a54c7dca9f287
parent8b81d0633f1ebaf1392e6509b34694de05c94de7
AMDGPU/GlobalISel: Mark G_ATOMICRMW_{NAND|FSUB} as lower

These aren't implemented and we're still relying on the AtomicExpand
pass, but mark these as lower to eliminate a few of the few remaining
no rules defined cases.
llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp