PerformVMOVRRDCombine ommits adding a offset
of 4 to the PointerInfo, when converting a
f64 = load[M]
to
{i32, i32} = {load[M], load[M + 4]}
Which would allow the machine scheduller
to break dependencies with the second load.
- pr42638
Differential D64870
[ARM][DAGCOMBINE][FIX] PerformVMOVRRDCombine Authored by dnsampaio on Jul 17 2019, 7:56 AM.
Details PerformVMOVRRDCombine ommits adding a offset Which would allow the machine scheduller
Diff Detail
|