similar to: [LLVMdev] Post-dominator data.

Is our post-dom pass solid? I thought it's fair expensive so we avoid using it. Evan Sent from my iPhone On May 29, 2009, at 6:14 PM, mgill404 at ucla.edu wrote: > This is surely a simple question, but it appears that > PostDominatorTree does not provide an interface to get at post > dominator data. Also, it is not a concrete subclass of DominatorTree, > as is presented in

Hi folks, For the past month I’ve been working on improving the DominatorTree and PostDominatorTree in LLVM. The RFC that explains the motivations and plans can be found here: http://lists.llvm.org/pipermail/llvm-dev/2017-June/114045.html . Here’s a short summary of what changed upstream since posting it: - We switched from the Simple Lengauer-Tarjan algorithm for computing dominators

Many published analyses which build on post-dominance assume a canonical single-dominator-tree form induced by unifying all exits (and often adding a virtual edge from START to END). In contrast, it seems that the current LLVM post-dominator analysis only operates in a mode in which it generates a forest of post-dominator trees, with one rooted at each exit node. The problem this can cause is

On Aug 24, 2009, at 4:58 PM, Jonathan Ragan-Kelley wrote: > Many published analyses which build on post-dominance assume a > canonical single-dominator-tree form induced by unifying all exits > (and often adding a virtual edge from START to END). In contrast, it > seems that the current LLVM post-dominator analysis only operates in a > mode in which it generates a forest of

Hello, I am working on a project that requires me to perform some analysis on a bitcode file outside the existing LLVM analysis frameworks (opt, etc.).  In what I am doing, I need to be able to access the dominator tree for a given function.  Is there a way I can instantiate a DominatorTree object outside the existing analysis frameworks? I have tried several things all along the lines of:

On 7/29/11 9:43 AM, david.dewey at comcast.net wrote: > > Hello, > > I am working on a project that requires me to perform some analysis on > a bitcode file outside the existing LLVM analysis frameworks (opt, > etc.). In what I am doing, I need to be able to access the dominator > tree for a given function. Is there a way I can instantiate a > DominatorTree object

John, Thanks for the quick reply.  Using a PassManager object should work fine.  As you said, my only objection is to exec'ing opt.  I had actually tried instantiating a PassManager object before, but I was definitely not doing it right.  I will take a look at how clang and SAFECode use it, and see if I can get it working. Thanks, David ----- Original Message ----- From:

On 7/29/11 10:00 AM, david.dewey at comcast.net wrote: > > John, > > Thanks for the quick reply. Using a PassManager object should work > fine. As you said, my only objection is to exec'ing opt. > > I had actually tried instantiating a PassManager object before, but I > was definitely not doing it right. I will take a look at how clang > and SAFECode use it,

On 4/12/13 3:19 PM, Arsen wrote: > Thank you John. > Actually the opt tool (from LLVM 3.2 version) can generate the needed graphs > (with pass "-domfrontier"). > But I just want to surely know is there some pass or builder which can be > integrated somehow so it will be possible directly to generate CDG? Yes and no. There's isn't a control dependence pass in LLVM

On Mon, Jul 19, 2010 at 7:49 AM, John Criswell <criswell at uiuc.edu> wrote: > Chayan Sarkar wrote: >> Hi, >> >> Can anyone tell me how to traverse Dominator tree in pre-order manner? >> > > In previous versions of LLVM, I believe you could get some sort of Node > object from DominatorTree and then use a method to get the children of > the Node There

Chayan Sarkar wrote: > Hi, > > Can anyone tell me how to traverse Dominator tree in pre-order manner? > In previous versions of LLVM, I believe you could get some sort of Node object from DominatorTree and then use a method to get the children of the Node (and then use a method of the Node to get the BasicBlock held in that Node). However, I don't see such a method in the

I'm seeing some very unexpected behavior in a ModulePass I've created. To support both the legacy and new pass managers, I've abstracted away the code to get the passes it depends on: bool MyPass::runOnModule(Module &M) override { auto DominatorGetter = [this] (Function &F) -> DominatorTree & { return

Hi! I wonder if llvm::APFloat::toString() can be const since it should not modify the APFloat. -Jochen

On 03/07/2010 10:33 PM, Jochen Wilhelmy wrote: > Hi! Hi Jochen, > I have seen that findNearestCommonDominator has been added > to class PostDominatorTree, maybe on my request. > > Now there is the following problem: > in class DominatorTreeBase there is an assert in findNearestCommonDominator > that asserts if the tree is not a PostDominator tree: > > assert

There might be a misuse of DominatorTree::splitBasicBlock in CodeExtractor.cpp, line 145. Header is splited into two (OldPred->NewBB). Line 145 updates the dominator tree by calling DT->splitBasicBlock(NewBB). I think it should be DT->splitBasicBlock(OldPred). When I tried to extract a code region whose header has 2 successors, my pass crashed. It was because header (or OldPred) is the

Hello, I was wondering whether or not some behaviour that I am seeing is expected behaviour and that it has been designed like this, or not. A dominator relation is given by "if A dominates B", then all paths to B go through A. For example, take the CFG below (which is a directed graph (couldn’t make the arrow heads but ok.): A / \ B C \ / D | E We can construct

Hi, Can anyone tell me how to traverse Dominator tree in pre-order manner? Regards, Chayan

Thanks for your reply, John. I am using opt to run my pass. Should opt by default link in the library? Would you please give me some idea on fixing it? Thanks! opt -load ../../../Debug+Asserts/lib/LLVMHello.dylib -hello < hello.bc > /dev/null My pass is just: #include "llvm/IR/Function.h" #include "llvm/Pass.h" #include "llvm/Support/raw_ostream.h" #include

On 03/07/2010 11:09 AM, Jochen Wilhelmy wrote: > Hi! Hi Jochen, > I'd like to find the point where the control flow joins after a branch. > > e.g. if I have the following function > > if (a< b) > foo1(); > else > foo2(); > bar(); > > Then the control flow splits at the basic block containing a< b with > a branch as terminator

Hi all, I am trying to traverse a dominator tree and have encountered a weird runtime exception: Here's my simple code: virtual bool runOnFunction(Function &F) { DominatorTree& DT = getAnalysis<DominatorTree>(); * DomTreeNode* rootNode = DT.getRootNode();* return false. } Here's the documentation page: