similar to: [LLVMdev] Controlling the order of a FunctionPass

Trevor Harmon wrote: > Hi, > > I would like my FunctionPasses to be invoked in reverse call graph > order (callees before callers). However, "Writing an LLVM Pass" notes > that "FunctionPass's do not require that they are executed in a > particular order." Is there any way at all to specify an ordering? > > If this is not possible, I'm

On Jul 22, 2010, at 2:05 PM, John Criswell wrote: > If you write your pass as a ModulePass, then you can iterate over the > functions in any order that you want. I had considered that, but my FunctionPass depends on other passes processing the functions first: void MyPass::getAnalysisUsage(AnalysisUsage &AU) const { AU.addRequired<UnifyFunctionExitNodes>();

Trevor Harmon wrote: > On Jul 22, 2010, at 2:05 PM, John Criswell wrote: > > >> If you write your pass as a ModulePass, then you can iterate over the >> functions in any order that you want. >> > > I had considered that, but my FunctionPass depends on other passes > processing the functions first: > Two things to consider: 1) The PassManager

Hey LLVMDev, I'm working on a ModulePass that uses ScalarEvolution along with several other analyses. After some debugging, it looks to me like ScalarEvolutionWrapperPass does not handle memory correctly for this case. Here's my current understanding of the problem. ScalarEvolutionWrapperPass maintains a unique_ptr to a ScalarEvolution. Calling getSE() dereferences this pointer.

John Criswell wrote: > Trevor Harmon wrote: >> On Jul 22, 2010, at 2:05 PM, John Criswell wrote: >> >> >>> If you write your pass as a ModulePass, then you can iterate over the >>> functions in any order that you want. >>> >> >> I had considered that, but my FunctionPass depends on other passes >> processing the functions first:

Hi, I'm writing a function pass which is dynamically loaded by opt and I need some analysis and passes to be run before my pass: virtual void getAnalysisUsage(AnalysisUsage &AU) const { AU.addRequired<LoopInfo>(); AU.addPreserved<LoopInfo>(); AU.addRequiredID(LoopSimplifyID); AU.addPreservedID(LoopSimplifyID);

I have a ModulePass, which we'll call MP, that generates a dependency graph for an entire program. I want MP to require the UnifyFunctionExitNodes pass, which is a FunctionPass. Since its not possible to make a ModulePass depend on a FunctionPass, is my only choice to make MP a FunctionPass in which the runOnFunction() routine does nothing, and the doFinalization routine does all the

Hi again, I come back to this issue with an example. It's a pass which does nothing but throw the 'Unable to schedule' error. namespace { struct MyPass : public FunctionPass { static char ID; // Pass identification, replacement for typeid MyPass() : FunctionPass(ID) { initializeMyPassPass(*PassRegistry::getPassRegistry()); } virtual void

Thank you for your quick reply. Actually I am using a std::map to map Function* to LoopInfo*, but that does not help in this case. Each time I call getAnalysis<llvm::LoopInfo>(*F), it returns the same instance of llvm::LoopInfo, so the std::map is just mapping every function into the same instance. It seems only the analysis result for the last function is valid, because all the result for

Thanks John, I modify my code to like this: bool XXX::ModulePass(Module &M){ .... LoopInfo &li = getAnalysis<LoopInfo>(fi); .... } Here fi is a Function* pointing to main(). Now when I run the pass, another error shows up: AnalysisType& llvm::Pass::getAnalysisID(const llvm::PassInfo*, llvm::Function&) [with AnalysisType = llvm::LoopInfo]: Assertion `ResultPass

On 3/9/12 4:28 PM, Fan Long wrote: > Thank you for your quick reply. > > Actually I am using a std::map to map Function* to LoopInfo*, but that > does not help in this case. Each time I call > getAnalysis<llvm::LoopInfo>(*F), it returns the same instance of > llvm::LoopInfo, so the std::map is just mapping every function into > the same instance. It seems only the

Hi all, I tried to have a LoopInfo object in a function pass, add addRequired<LoopInfo> in getAnalysisUsage, and then use getAnalysis<LoopInfo> in runOnFunction(). It worked OK. Now I want to have a module pass to traverse the functions, and similarly I want to have to loop information of the functions. When I did the above in runOnModule, and run the pass, the following error popped

ether zhhb wrote: > hi John, > > sorry for reply so late. > > On Tue, Apr 13, 2010 at 10:38 PM, John Criswell <criswell at uiuc.edu > <mailto:criswell at uiuc.edu>> wrote: > > Devang Patel wrote: > > On Mon, Apr 12, 2010 at 6:41 PM, ether zhhb > <etherzhhb at gmail.com <mailto:etherzhhb at gmail.com>> wrote: > >

Hello, I am trying to write a new ModulePass using LoopInfo analysis result, but it seems I misunderstand some concept about PassManager. Basically I want to keep LoopInfo analysis result alive. Here is an example showing the problem I encountered, assuming I already addRequired<llvm::LoopInfo>() in getAnalysisUsage: void foo(llvm::Function *F1, llvm::Function *F2) { llvm::LoopInfo

Hello all; So my goal is to insert some (self-written) function calls in the LLVM IR. I can achieve it with a ModulePass and using the getOrInsertFunction() call. For a ModulePass I follow the steps- 1. Compile the code (in which call instructions are to be inserted) to LLVM IR 2. Compile the file (which contains the *called* external function ) to LLVM IR 3. Link them together and run the

I think this assertion failure may be caused by the getAnalysisUsage(). Mine is like the following: class myclass : public ModulePass{ ... virtual void getAnalysisUsage(AnalysisUsage &AU) const { AU.addRequired<LoopInfo>(); } ... } Is it enough? Thanks! Best, --Wenbin ----- Original Message ----- From: Wenbin Zhang To: John Criswell Cc: llvmdev at

On 3/3/11 3:09 PM, Wenbin Zhang wrote: > Hi all, > I tried to have a LoopInfo object in a function pass, add > addRequired<LoopInfo> in getAnalysisUsage, and then use > getAnalysis<LoopInfo> in runOnFunction(). It worked OK. > Now I want to have a module pass to traverse the functions, and > similarly I want to have to loop information of the functions. When I >

On 5/6/15 11:15 AM, Kuperstein, Michael M wrote: > > I’ve always thought that the only guarantee is that > doFinalization(Module &M) runs after runOnFunction() was executed for > all functions in M, and there’s no guarantee it runs *immediately* after. > > That is, a PM may run a bunch of function passes over each function, > and only then call doFinazliation() for each

On 3/9/12 4:10 PM, Fan Long wrote: > Hello, > I am trying to write a new ModulePass using LoopInfo analysis result, but it seems I misunderstand some concept about PassManager. Basically I want to keep LoopInfo analysis result alive. Here is an example showing the problem I encountered, assuming I already addRequired<llvm::LoopInfo>() in getAnalysisUsage: > > void

LLVMers, I have a traversal plan which needs to visit certain types of Functions. For each instruction in such functions, I will need to know: 1. is it located inside a loop 2. if yes, what level of loop nest it is currently in. So on the highest level, it should be a FunctionPass. However, in order to identify loops and loop-nest levels, it should also be a LoopPass. Is there a reasonably