Michael Ilseman
2011-May-16 15:55 UTC
[LLVMdev] InstructionCombining.cpp inconsistency in whether it modifies the CFG?
InstCombine says in its getAnalysisUsage that it preserves the CFG, but for the 4th argument in its INITIALIZE_PASS call, it says false, which I believe corresponds to whether it preserves the CFG. Is this a mistake, or is there deeper meaning here? InstructionCombining.cpp:73-82 char InstCombiner::ID = 0; INITIALIZE_PASS(InstCombiner, "instcombine", "Combine redundant instructions", false, false) void InstCombiner::getAnalysisUsage(AnalysisUsage &AU) const { AU.addPreservedID(LCSSAID); AU.setPreservesCFG(); }
Duncan Sands
2011-May-16 19:58 UTC
[LLVMdev] InstructionCombining.cpp inconsistency in whether it modifies the CFG?
Hi Michael,> InstCombine says in its getAnalysisUsage that it preserves the CFG, > but for the 4th argument in its INITIALIZE_PASS call, it says false, > which I believe corresponds to whether it preserves the CFG.that argument should be set to true if the pass only looks at the CFG, i.e. whatever it computes/does is only a function of the CFG, and doesn't otherwise depend on what particular instructions make up the function. Suppose an analysis pass A has this property, and a transform pass P announces (using setPreservesCFG()) that it does not change the CFG. Then A does not need to be recomputed after P has run. So I think it is correct for InstCombine to not set this flag. Ciao, Duncan. Is this a> mistake, or is there deeper meaning here? > > InstructionCombining.cpp:73-82 > > char InstCombiner::ID = 0; > INITIALIZE_PASS(InstCombiner, "instcombine", > "Combine redundant instructions", false, false) > > void InstCombiner::getAnalysisUsage(AnalysisUsage&AU) const { > AU.addPreservedID(LCSSAID); > AU.setPreservesCFG(); > } > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
Michael Ilseman
2011-May-16 20:09 UTC
[LLVMdev] InstructionCombining.cpp inconsistency in whether it modifies the CFG?
Ah, I see, thanks for the clarification. I interpreted "Pass only looks at the CFG" as didn't modify it, only looked at it, but I see now that it can also mean "Pass looks only at the CFG", as in not the instructions. Thanks! On Mon, May 16, 2011 at 1:58 PM, Duncan Sands <baldrick at free.fr> wrote:> Hi Michael, > > > InstCombine says in its getAnalysisUsage that it preserves the CFG, > > but for the 4th argument in its INITIALIZE_PASS call, it says false, > > which I believe corresponds to whether it preserves the CFG. > > that argument should be set to true if the pass only looks at the CFG, i.e. > whatever it computes/does is only a function of the CFG, and doesn't > otherwise > depend on what particular instructions make up the function. Suppose an > analysis pass A has this property, and a transform pass P announces (using > setPreservesCFG()) that it does not change the CFG. Then A does not need > to be recomputed after P has run. > > So I think it is correct for InstCombine to not set this flag. > > Ciao, Duncan. > > Is this a > > mistake, or is there deeper meaning here? > > > > InstructionCombining.cpp:73-82 > > > > char InstCombiner::ID = 0; > > INITIALIZE_PASS(InstCombiner, "instcombine", > > "Combine redundant instructions", false, false) > > > > void InstCombiner::getAnalysisUsage(AnalysisUsage&AU) const { > > AU.addPreservedID(LCSSAID); > > AU.setPreservesCFG(); > > } > > _______________________________________________ > > LLVM Developers mailing list > > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev > > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20110516/89ac99af/attachment.html>
Reasonably Related Threads
- [LLVMdev] InstructionCombining.cpp inconsistency in whether it modifies the CFG?
- [LLVMdev] Function Pass Manager
- [LLVMdev] Function Pass Manager
- [LLVMdev] How to call some transformation passes (LoopRotate and LoopUnroll) from my own pass
- [LLVMdev] YSU_Student