Displaying 2 results from an estimated 2 matches for "r258".
Did you mean:
258
2011 Jan 04
4
[LLVMdev] Bug in MachineInstr::isIdenticalTo
...; <i16> [#uses=1]
store i16 %conv67, i16 addrspace(1)* %arrayidx64
ret void
Notice how except for the sizes of the pointer, the sequence is the same. This translates into the following for my backend at the MI level.
BB#9: derived from LLVM BB %switch.case55
Live Ins: %R258 %R260 %R259
Predecessors according to CFG: BB#8
%R257<def> = CUSTOM_ADD_i32 %R260<kill>, %R258<kill>
%R258<def> = VEXTRACT_v4i32 %R259<kill>, 1
GLOBALTRUNCSTORE_i32 %R258<kill>, %R257<...
2011 Jan 04
0
[LLVMdev] Bug in MachineInstr::isIdenticalTo
...uses=1]
> store i16 %conv67, i16 addrspace(1)* %arrayidx64
> ret void
>
> Notice how except for the sizes of the pointer, the sequence is the same. This translates into the following for my backend at the MI level.
> BB#9: derived from LLVM BB %switch.case55
> Live Ins: %R258 %R260 %R259
> Predecessors according to CFG: BB#8
> %R257<def> = CUSTOM_ADD_i32 %R260<kill>, %R258<kill>
> %R258<def> = VEXTRACT_v4i32 %R259<kill>, 1
> GLOBALTRUNCSTORE_i32 %R258<...