similar to: [LLVMdev] What's Going Wrong with GCC 4.6.3 + dragonegg 3.1

Arnamoy Bhattacharyya wrote: > Hi all; > > Problem: Running a bitcode file produced by llvm-gcc (gcc 4.6.3 + > dragonegg 3.1 + llvm 3.1) with lli (Ubuntu 12.04) > > /llvm-gcc:/ > > llvm-gcc -c hello.c -emit-llvm -o hello.bc > lli hello.bc > > output: > lli: hello.bc:1:1: error: expected top-level entity > ELF@4( > > /clang:/ > > clang -c hello.c

Hello; I wrote this simple loop pass to collect the number of instructions in each loop of the program. The code is as follows- #define DEBUG_TYPE "loopinst" #include "llvm/Pass.h" #include "llvm/Analysis/LoopPass.h" #include "llvm/Support/raw_ostream.h" #include "llvm/ADT/Statistic.h" #include "llvm/Instructions.h" #include

Hi, > Problem solved. I was building llvm in a separate llvm-build directory. I > built it within the llvm-src directory (which kept all the llvm .so and my > pass' .so in the llvm-src/Release+Asserts/lib directory) to solve the > problem. I do not fully understand what you mean, there should be no difference on building out of source AFAIK. > Can anyone tell me what's

Problem solved. I was building llvm in a separate llvm-build directory. I built it within the llvm-src directory (which kept all the llvm .so and my pass' .so in the llvm-src/Release+Asserts/lib directory) to solve the problem. Can anyone tell me what's the difference between writing a pass as a "struct" (as in the tutorial) and as a "class" (as most developers do)?

Hello all; So here is my goal: *** If there is a Call instruction to myFunction(int * val), I need to identify all the instructions that uses val in the IR and replace the uses with a newly created pointer. I will also be changing the argument of myFunction to the new pointer. int * val = malloc/calloc; ... myFunction(val); .... *val = 45; becomes== int * val = malloc/calloc; int * val1 =

Hi; I was playing with the -lda pass of LLVM on the following program- #include <stdio.h> void main() { int a[10]; int i; for(i = 0; i < 4; i ++) { a[i] = a[i-1]+1; } } I run the following commands - clang a.c -emit-llvm -S opt -analyze -stats -lda a.s The output is - Printing analysis 'Loop Dependence Analysis': Loop at depth 1, header block: %for.cond Load/store

Thanks for the reply. For the source code: int main() { int a = 0; int *p; p = &a; call_arnamoy(p); int *p1; p1 = p; return 1; } The bit code: %retval = alloca i32, align 4 %a = alloca i32, align 4 %p = alloca i32*, align 8 %p1 = alloca i32*, align 8 store i32 0, i32* %retval store i32 0, i32* %a, align 4 store i32* %a, i32** %p, align 8 %0 = load i32** %p, align 8 %call = call i32

Hello; I wanted to insert two new terminator instructions in LLVM IR. I modified the necessary files (as mentioned in the tutorial in LLVM site and some more to meet the new structure of LLVM, (e.g no Lexer.l now, we have LLLexer.cpp. Again in the tutorial, files like LLVMBitCodes.h, include/llvm-c/core.h that need to be modified are not mentioned) But I am getting the error while building

Hello; I was planning to add a new instruction to the LLVM IR (and later to MIPS backend) for TLS(Thread level speculation) support. For this I tried to follow the steps described in http://llvm.org/docs/ExtendingLLVM dot html#instruction. But I could not find any llvm/lib/AsmParser/Lexer.l and llvm/lib/AsmParser/llvmAsmParser.y file in both the svn repository and the source code downloaded

Hi; I was wondering how can I get the name of the bitcode file on which I am running the pass. I am making my pass as a loadable module. And if I run my pass like this - opt -S -loops -load /home/arnie/llvm-clang/llvm/Release+Asserts/lib/XYZ.so -xyz <b.bc> b1.bc I want to get b.bc in the pass. Thank you; -- Arnamoy Bhattacharyya Athabasca Hall 143 Department of Computing Science -

I think I understood the problem you're trying to solve. I didn't, however, understand the problems you are having (maybe you're not running mem2reg?). I do have a little piece of code that I think does (more or less) what you want. Take a look at it, try understanding it, and see if it helps you out. You might have to change it a bit since you're replacing uses after an

Hi all; I have been trying to use the llvm::ConstantArray::get(llvm::LLVMContext&, llvm::StringRef, bool) function but seems it has been deprecated. ProfileDependence.cpp:68:73: error: no matching function for call to ‘llvm::ConstantArray::get(llvm::LLVMContext&, llvm::StringRef, bool)’ ProfileDependence.cpp:68:73: note: candidate is:

Hi; I wrote an LLVM transformation pass and I put my code inside the llvm-src/lib/Transforms/my-pass directory. I want to know what is the correct method for compiling it and building my plugin for the pass. After reading the http://llvm.org/docs/WritingAnLLVMPass.html page, I felt that I have to keep my pass inside the above said directory. But when I run make from inside my-pass directory

Hello; Thanks to Eli for the pointer to the ConstantDataArray::getString() fucntion. Now I am trying to declare a global char array with the content "hi" and call a function "print" (which takes a char pointer and return an insteger. I am doing the following in the code - Function Creation: PointerType* array = PointerType::get(IntegerType::getInt8Ty(getGlobalContext())

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

> From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu] On Behalf Of Arnamoy Bhattacharyya > Subject: [LLVMdev] Is Loop Dependence Analysis Printing Correct Stats? > for(i = 0; i < 4; i ++) > { > a[i] = a[i-1]+1; > } > Am I doing anything wrong or the code seems buggy? Your code is buggy. The first time through the loop, you're referencing

Hello all; I have been writing this LLVM pass that extracts loops in a function and then duplicates the function so the loop is executed twice. The pass executes fine but the produced bit code does not behave correctly as it should (both the original and duplicated function in which the loop is extracted is called, but from the output it seems only one of them is executed.) Here is the source:

Hi Arnamoy, > Any idea, why the output looks like - > Hello World > Hello World > val is 58 > > While the it should print 4 times? Both your loops are sharing the counter variable (%i1) without resetting it to 0 in between calls. I think that accounts for the behaviour. Cheers. Tim.

Hello; I saw a discussion went i the llvm mailing list about work on supporting OpenMP in Clang (llvm-gcc already supports that).( http://lists.cs.uiuc.edu/pipermail/llvmdev/2012-January/046640.html) I just want to know, anybody has made any progress in it? It would be really helpful for me. Thanks a lot; -- Arnamoy Bhattacharyya Athabasca Hall 143 Department of Computing Science -

Hi; Both -fplugin-arg-dragonegg-emit-ir and -flto works for creating llvm bitcode that can be run with lli. Thanks to Nick for the suggestion. Now, trying to run an OpenMP hello world program. llvm-gcc -fopenmp p_hello.c -flto -S -o p_hello.ll lli p_hello.ll output: LLVM ERROR: Program used external function 'GOMP_parallel_start' which could not be resolved! I am guessing that I have