paweld/fpc
1
0
Fork 0
mirror of https://gitlab.com/freepascal.org/fpc/source.git synced 2025-04-16 00:42:06 +02:00
Code Issues Packages Projects Releases Wiki Activity

* x86: Fixed inefficiency in the long-range MOV optimisations

Browse Source
This commit is contained in:
J. Gareth "Curious Kit" Moreton 2023-10-29 01:55:11 +01:00 committed by FPK
parent 736f5edf98
commit af1eb33a75
1 changed files with 17 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
compiler/x86/aoptx86.pas
View File

@ -4369,21 +4369,25 @@ unit aoptx86;
if (taicpu(hp2).oper[1]^.typ = top_ref) then
ReplaceRegisterInRef(taicpu(hp2).oper[1]^.ref^, p_TargetReg, p_SourceReg);
{ Don't remove the first instruction if the temporary register is in use }
if not TempRegUsed and
{ ReplaceRegisterInRef won't actually replace the register if it's a different size }
not RegInOp(p_TargetReg, taicpu(hp2).oper[1]^) then
{ ReplaceRegisterInRef won't actually replace the register if it's a different size }
if not RegInOp(p_TargetReg, taicpu(hp2).oper[1]^) then
begin
DebugMsg(SPeepholeOptimization + 'MovMov2Mov 6 done',p);
RemoveCurrentP(p, hp1);
Result:=true;
JumpTracking.Free;
Exit;
end;
{ Don't remove the first instruction if the temporary register is in use }
if not TempRegUsed then
begin
DebugMsg(SPeepholeOptimization + 'MovMov2Mov 6 done',p);
RemoveCurrentP(p, hp1);
Result:=true;
JumpTracking.Free;
Exit;
end;
{ No need to set Result to True here. If there's another instruction later
on that can be optimised, it will be detected when the main Pass 1 loop
reaches what is now hp2 and passes it through OptPass1MOV. [Kit] }
{ No need to set Result to True here. If there's another instruction later
on that can be optimised, it will be detected when the main Pass 1 loop
reaches what is now hp2 and passes it through OptPass1MOV. [Kit] }
hp3 := hp2;
Continue;
end;
end;
end;
top_const: