Core/MIPS/IR/IRPassSimplify.h

#pragma once

#include "Core/MIPS/IR/IRInst.h"

typedef bool (*IRPassFunc)(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool IRApplyPasses(const IRPassFunc *passes, size_t c, const IRWriter &in, IRWriter &out, const IROptions &opts);

// Block optimizer passes of varying usefulness.
bool RemoveLoadStoreLeftRight(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool PropagateConstants(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool PurgeTemps(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool ReduceLoads(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool ThreeOpToTwoOp(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool OptimizeFPMoves(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool ReorderLoadStore(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool MergeLoadStore(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool ApplyMemoryValidation(const IRWriter &in, IRWriter &out, const IROptions &opts);
bool ReduceVec4Flush(const IRWriter &in, IRWriter &out, const IROptions &opts);
It's kind of starting to run 2016-05-07 17:37:19 +02:00			`#pragma once`

			`#include "Core/MIPS/IR/IRInst.h"`

irjit: Add options for compile/optimize steps. This way the backend can set flags for the type of IR it wants. It's seems too complex to combine certain things like lwl/lwr in a pass. 2017-12-31 17:14:34 -08:00			`typedef bool (*IRPassFunc)(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool IRApplyPasses(const IRPassFunc *passes, size_t c, const IRWriter &in, IRWriter &out, const IROptions &opts);`
Use the regcache in a new (incomplete) pass, PropagateConstants. 2016-05-08 01:06:07 +02:00
Add a couple more passes (2-op, optimize f<->v fp moves) 2016-05-13 20:14:03 +02:00			`// Block optimizer passes of varying usefulness.`
irjit: Combine lwl/lwr and swl/swr, like before. Still want to inline the operation, because the backend shouldn't have to redo it every time, and we want the temps cleaned up if possible. 2018-01-07 18:37:06 -08:00			`bool RemoveLoadStoreLeftRight(const IRWriter &in, IRWriter &out, const IROptions &opts);`
irjit: Add options for compile/optimize steps. This way the backend can set flags for the type of IR it wants. It's seems too complex to combine certain things like lwl/lwr in a pass. 2017-12-31 17:14:34 -08:00			`bool PropagateConstants(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool PurgeTemps(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool ReduceLoads(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool ThreeOpToTwoOp(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool OptimizeFPMoves(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool ReorderLoadStore(const IRWriter &in, IRWriter &out, const IROptions &opts);`
			`bool MergeLoadStore(const IRWriter &in, IRWriter &out, const IROptions &opts);`
irjit: Add constructs for validing mem access. Basically to allow slow/fast memory to work with IR, including for alignment checks. 2022-08-21 12:44:35 -07:00			`bool ApplyMemoryValidation(const IRWriter &in, IRWriter &out, const IROptions &opts);`
irjit: Add a pass to keep Vec4s in Vec4s. 2023-08-27 11:47:02 -07:00			`bool ReduceVec4Flush(const IRWriter &in, IRWriter &out, const IROptions &opts);`