similar to: [LLVMdev] How to supply a non-default constructor to RegisterPass ?

On Sep 30, 2013, at 10:11 PM, Dirk Schroetter <dschroetter at googlemail.com> wrote: > Folks, > > I did read the documentation but I am at a loss still. > > I do have the following: > > Pass *createFooPass (char *foo1) { > return new FooPass (foo1); > } > > static RegisterPass<FooPass> X("foo-pass", "foo", false, true); >

Originally I was planning on allowing all kinds of mail settings inside namespace {}, including imapc_* settings. But that's a bit difficult to implement (although I think it will happen some day). So for now I was thinking: imapc foo { host = imap.foo.com master_user = foomaster password = foopass } imapc bar { host = imap.bar.com ... } namespace { prefix = foo/ location =

Does anyone know what the guidelines are supposed to be for properly initializing a pass? Looking around, we seem to have three styles of pass registration in use. INITIALIZE_PASS(...) INITIALIZE_PASS_BEGIN(...) INITIALIZE_PASS_DEPENDENCY(...) ... INITIALIZE_PASS_END(...) static RegisterPass<FooPass> X(...); (This is the one encouraged in the docs, but seemingly the least widely used

When migrating my project to 2.9, I've encountered a strange segfault where if a ModulePass's getAnalysisUsage adds LoopInfo and DominatorTree, then llvm::PMTopLevelManager::findAnalysisUsage will segfault. What's odd is that if I rearrange this (add required for DominatorTree before LoopInfo), it does not segfault. I realize that LoopInfo requires and preserves DominatorTree, but this

Folks, I am using a two-stagesimple build like this one: make -G Ninja -DCLANG_ENABLE_BOOTSTRAP=ON ../src ninja stage2-check-all Stage 1 gets built, and when looking at rules.ninja in the stage2-bins I see: rule CXX_COMPILER_AnalysisTests depfile = $DEP_FILE deps = gcc command = <stage1 build dir>./bin/clang++ $DEFINES $INCLUDES $FLAGS -MMD -MT $out -MF $DEP_FILE -o $out -c $in

Dear all, I have the following data, which has \\n in place of \n. I introduced \n's in the csv file so that I could use it in barchart in lattice. When I did that and read it into R using read.csv, it read it as \\n. My question is how do I introduce "\n" in the middle of a long string of quoted text so that lattice can make multiline strips. Hitting Enter which is supposed to

Thanks for the response. I do have assertions enabled, and none of them are getting hit. I did do a search of the mailing list for the past year (approximately) before writing my email, and what I found was that you should be allowed to use LoopInfo and other analysis function passes from a module pass, with the only difference being that getAnalysis is passed the function. The example code I

Hi Michael, > When migrating my project to 2.9, I've encountered a strange segfault > where if a ModulePass's getAnalysisUsage adds LoopInfo and > DominatorTree, then llvm::PMTopLevelManager::findAnalysisUsage will > segfault. I suggest you build LLVM with assertions enabled - then you should get a helpful error message rather than a segfault. I think you are not allowed to

Hi Michael, hi Duncan, yesterday I stumbled over something that might be related. At least I could also just be doing some initialization wrong or something in this direction... In my case, I hit a segfault in PassInfo::isAnalysisGroup() after PassManager.add(myModulePass) is called. My setup seems fairly simple, the attached code should reproduce the error. Compile with g++ test.cpp

Your constructor is not calling initializeTestMPPass(), and you're using RegisterPass which I think was deprecated in favor of INITIALIZE_PASS. You can look at, for example, lib/Transforms/Scalar/IndVarSimplify.cpp for examples of how to initialize, e.g. having "INITIALIZE_PASS_DEPENDENCY(LoopInfo)" sandwiched between BEGIN and END. Note that you'll want a forward declaration of

On 05/31/2013 04:21 PM, Timo Sirainen wrote: > Originally I was planning on allowing all kinds of mail settings inside namespace {}, including imapc_* settings. But that's a bit difficult to implement (although I think it will happen some day). So for now I was thinking: > > imapc foo { > host = imap.foo.com > master_user = foomaster > password = foopass > } >

Hello, I dont fully understand how the domain name replacement in mail_location takes place. If I connect to the server using my username as in the text file i can not access my Maildirs as %d gets not replaced. If i user username at domain.com the %d is correctly replaced. But, how would i configure my mail_location, to take the domain part from the server name i connect to? So i could set

Hello, I''m having a trouble when I try to connect RoR and Oracle. I believe that it has something to do with database.yml, this is how my database.yml looks: development: adapter: oci database: 127.0.0.1:1521/XE username: system password: y1984nip0359 host: 127.0.0.1 And this is the Error that i get: -> http://127.0.0.1:3000/blog OCIError in BlogController#index

Let me rephrase to make sure I understand. Once we have this settled, I'll update the comments in the file to summarize and possibly update the docs. On 08/21/2015 10:27 AM, Owen Anderson wrote: > The macro versions should be preferred, as they reduce static initializaters. Makes sense. But the tradeoff is that they need to be baked into the list of known passes so that someone knows

On Thursday 25 February 2010 19:24:30 Trevor Harmon wrote: > Hi, > > I'm confused about the is_analysis parameter of the RegisterPass > constructor (defined in PassSupport.h). The only explanation I can > find is that is_analysis should be set to true if the pass is "an > analysis pass, for example dominator tree pass". Can someone please > clarify what is meant

On Fri, Feb 26, 2010 at 05:22:49PM -0800, Trevor Harmon wrote: > On Feb 25, 2010, at 5:45 PM, David A. Greene wrote: > > > The RegisterPass template takes care of notifying the PassManager > > that the pass exists. One of the advantages of PassManager is > > that it decouples pass execution from the rest of the infrastructure > > so the class need not be visible to

On Feb 26, 2010, at 7:42 PM, Tom Prince wrote: > You cannot refer to anything in anonymous namespace by name from > another file, however, if you have a pointer, say, then you can > still use the > pointer. Yes, that would make sense, except I don't see any pointers being passed. The relevant line in lib/Transforms/Hello.cpp is: static RegisterPass<Hello>

Hey folks, As part of our ongoing quest to eliminate static constructors from the LLVM libraries (in order to improve startup times for applications that link against them), I'm going to be eliminating the RegisterPass class. They will be replaced with an initialization-function based system that I will detail in a future email. But, for now, what you need to know is that these classes are

HI, Developers: I linked libLTO.so in $LLVMROOT/Debug/lib however, my program crashes in assertion at 149line [Pass.cpp] void RegisterPass(const PassInfo &PI) { bool Inserted = PassInfoMap.insert(std::make_pair(PI.getTypeInfo(),&PI)).second; assert(Inserted && "Pass registered multiple times!"); } i confirm that there is only one static variable

On Feb 25, 2010, at 5:45 PM, David A. Greene wrote: > The RegisterPass template takes care of notifying the PassManager > that the pass exists. One of the advantages of PassManager is > that it decouples pass execution from the rest of the infrastructure > so the class need not be visible to anything outside the > implementation > of the pass itself. Even though the