search for: viewcfg

Hey everyone, I want to call F.viewCFG() from within the Hello.so shared library runOnFunction() function, but I keep getting: opt: symbol lookup error: ../../../Debug/lib/LLVMHello.so: undefined symbol: _ZNK4llvm8Function7viewCFGEv while using: "opt -load ../../../Debug/lib/LLVMHello.so -print /temp/llvm/KellyTest/test.bc&qu...

Hello all, I am getting an undefined reference to Function::viewCFG() I have a function list iterator F and I am trying to get the CFG through F->viewCFG(); other function calls through the F pointer are working. - I have included the libraries: pthread, ffi, dl, m, LLVMAsmParse, LLVMCore, LLVMSupport, LLVMSystem. - I have dot and gv in my path, and I have r...

...nd (DomTreeNode *N) { + return df_end<DomTreeNode *> (N); + } No spaces between the function name and the parenthesis. This occurs a few times. @@ -112,7 +59,7 @@ namespace { CFGOnlyViewer() : FunctionPass(&ID) {} virtual bool runOnFunction(Function &F) { - F.viewCFG(); + F.viewCFGOnly(); return false; } Put that back. viewCFG shows the CFG with full instructions inside it while viewCFGOnly shows just the block without the instruction listing inside. This pass is intended to show the details. + template<> + struct DOTGraphTraits...

Hi, I have some problems with the debug build. I cannot use 'call F->viewCFG()', gdb says the function might be inlined. Did not help with make CXXFLAGS+="-O0 -g3". I also have problems setting breakpoints in gdb for methods. I expected this functionality in a Debug build, pleas give any info /Jonas -------------- next part -------------- An HTML attachme...

...*> to enable the GraphWriter on these data structures. 3. The file DomPrinter.cpp containing the actual passes, printers & Co. It is simple as almost everything already existed. 4. A little bug fix in lib/Analysis/CFGPrinter.cpp. virtual bool runOnFunction(Function &F) { - F.viewCFG(); + F.viewCFGOnly(); return false; "opt -view-cfg-only" showed too much. Any feedback is highly appreciated. See you Tobi

On Wed, 2009-07-15 at 00:20 +0200, Tobias Grosser wrote: > Hi, > > I started to work with llvm and liked the CFG dotty printer a lot. > However there was none for the dominance trees. > > I implemented dotty printing for dominance trees And here is the patch. Tobi -------------- next part -------------- A non-text attachment was scrubbed... Name:

Hi, I tried to view the CFG when debugging a pass by "gdb opt," but the gdb does not have DAG.viewGraph function, which is indicated at http://www.llvm.org/docs/ProgrammersManual.html#ViewGraph. Could anyone tell me how to make DAG* functions available in gdb? Thanks in advance. Lu

Hello, LLVM provides several ways to visual IR structure straight in its core - Function::viewCFG() to render control flow graph, then -view-* options to llc to render various stages of transforming to machine code. However, I wasn't able to find a way to render complete DAG visualization of normal IR - which besides CFG would also show dataflow (and other flows, if any). What people use to...

Here's what the doxygen page has for viewCFGOnly() : /// viewCFGOnly - This function is meant for use from the debugger. It works /// just like viewCFG, but it does not include the contents of basic blocks /// into the nodes, just the label. If you are only interested in the CFG /// this can make the graph smaller. /// void viewCFGOnly() c...

Hi Hisham, Most likely the basic blocks are the headers of two different loops. Try running viewCFG() on the function in question to see if this is the case. Tom ----- Original Message ----- From: "Hisham Chowdhury" <hisham_chow at yahoo.com> To: llvmdev at cs.uiuc.edu Sent: Wednesday, June 16, 2010 7:22:00 PM GMT -05:00 US/Canada Eastern Subject: [LLVMdev] Loopinfo Analysis Hel...

Hello, I have a question regrading the analysis pass that generates loop info from an .ll code. My previous understanding was there will be just one loop header(in the loop info) for a particular loop. But, when i use isLoopHeader() member function from the loop info class I get 'true' return value for two different basic blocks. Note both basic blocks are loop conditional block(break

...VM. 14. The target-independent live-variable analysis is quicker and simpler than before. 15. Vladimir Prus contributed code generator support for targets that have object alignments bigger than their size (e.g. bytes aligned on 32-bit boundaries). 16. Alkis implemented MachineFunction::viewCFG and ::viewCFGOnly methods, which make it really easy to visualize machine code in a debugger (operating like the Function::viewCFG method). Note that this does not work with the SparcV9 target yet. 17. The LLVM llc and lli tools now support dynamically loaded target implementations...

Hello David, Thanks for the reply. Would i not need to pass any arguments for viewCFGOnly() method? According to this http://llvm.org/doxygen/Function_8h-source.html seems like we need to? I am building it on top of the example shown at http://llvm.org/docs/WritingAnLLVMPass.html and just added 'F->viewCFGOnly(); line to it. It doesn't compile, i suspect it has to do wi...

...least a path down the CFG where a virtual register "use" is not preceded by its "def". Are you sure that you don't see anything suspicious? You can see what your passes are doing to the code at any compilation stage with dump() (all "Value"s can be dumped), or viewCFG() in the case of functions. For me, this is like the printf() of LLVM debugging :) > I’m also interested in playing with branching between the original > code or one of a few variations of the extracted RPC based on some > runtime info (such as where global addresses resolve to). Just...

Hi Pablo, Your tool sounds really cool. It goes well beyond what I’m trying to do, which is really just extracting blocks of code, serializing and sending the inputs over to another core, running the code over there, and then sending the outputs back to the original caller (like an automatic RPC). So it sounds like most of the things your tool can do would be overkill for my use case. What I’ve

On Sun, Jul 26, 2009 at 1:48 PM, Ralf Karrenberg<Chareos at gmx.de> wrote: > Hey Daniel, > > thanks for the response. > >> I believe all you need to do is call llvm::llvm_shutdown(). >> > I am not sure that this is what I need. When and how should I call > llvm_shutdown()? At the end of the program, this just arranges for LLVM's "managed statics"

..., 2010 7:23 PM > To: Jonas Paulsson > Cc: 'llvmdev at cs.uiuc.edu' > Subject: Re: [LLVMdev] Debug build > > > On Sep 23, 2010, at 4:07 AM, Jonas Paulsson wrote: > >> Hi, >> >> I have some problems with the debug build. I cannot use 'call F->viewCFG()', gdb says the function might be inlined. Did not help with make CXXFLAGS+="-O0 -g3". >> >> I also have problems setting breakpoints in gdb for methods. > > Which compiler did you use to build llvm ? > - > Devang > >> >> I expected this f...

All the functions in LinkAllPasses.h are passes. I am not sure whether it takes inline functions. I added createPostDomTree() to LinkAllPasses.h but I get the same error. I wonder whether it was even worth it making the functions inline for PostDominators.h. George On Tue, Aug 21, 2012 at 9:05 AM, Tobias Grosser <tobias at grosser.es> wrote: > On 08/21/2012 02:40 PM, Duncan Sands

Hi, On 02/24/2013 06:39 PM, Paul Sokolovsky wrote: > Hello, > > LLVM provides several ways to visual IR structure straight in its core - > Function::viewCFG() to render control flow graph, then -view-* options > to llc to render various stages of transforming to machine code. > However, I wasn't able to find a way to render complete DAG > visualization of normal IR - which besides CFG would also show dataflow > (and other flows, if any)...

This is just a quick note that the Function class now has two new methods for directly viewing the CFG of a function being worked on. This is useful in a debugger, where allows you to just say: call F->viewCFG() and a window will pop up with the function in it, with each basic block broken into a node in the graph. This requires having 'dot' and 'gv' in your path. The two new methods are described here: http://mail.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20031020/008606.html -C...