search for: juanruiz

Dear LLVM, If I have a DSGraph dsg, I want to duplicate a DSgraph which is exactly the same with dsg. How should I do that? Should I use DSGraph dsg2 = dsg; or use some other methods? Thanks, xiaodong

I just recompiled the opt executable, but when trying to load a dynamic module I get the following error: % opt -load $LLVM_DIR/lib/Debug/libdsgraphshow.so --help Error opening '/usr/dcs/projects/cs426/juanruiz/llvm/lib/Debug/libdsgraphshow.so': ld.so.1: opt: fatal: relocation error: file /usr/dcs/projects/cs426/juanruiz/llvm/lib/Debug/libdsgraphshow.so: symbol _ZN7DSGraphC1ER8FunctionPS_: referenced symbol not found -load request ignored. $LLVM_DIR points to the right directory, libdsgraphshow.so...

does anyone knows how to get the Olden BM bytecode files? We're low on disk space and I don't want to build the whole test branch (even if i can, which I doubt, it failed last time i tried).

...I have found to get around this is to delete all *.d files when header files are moved or deleted: % cd LLVM_OBJ_DIR % /bin/rm `find . -name '*.d'` If someone knows a better way around this,I'd love to hear it too. --Vikram -----Original Message----- From: Juan Nicolas Ruiz [mailto:juanruiz at students.uiuc.edu] Sent: Friday, November 01, 2002 12:01 PM To: Vikram S. Adve Subject: RE: CVS Broken? > I did a full update of the CVS tree on the CSIL machines this morning. > This has fixed several problems: but we still need to use "gmake -k", don't we? If I use &q...

I've been using DSGraph->dump() to get the DSGraph of some functions, but the report make little sense to me. I'm specially looking for the "N" flag mentioned by prof. vikram yesterday. How can I read the results (so that I can manually verify that my passes are making a good job) or there is some other DSGraph function better suited for human consumption? tia nicolas

% df -k /usr/dcs/projects Filesystem kbytes used avail capacity Mounted on csil-server1:/usr/dcs/csil-projects 17332442 16991416 167702 100% /usr/dcs/projects % date Wed Oct 30 21:21:25 CST 2002 if anybody has good ideas on how to reduce disk usage, better speak up now....

shouldn't the malloc() be something like malloc(sizeof(int)*size)? You're allocating 5 bytes, which is probably not the size of an int in llvm (it's definitely not the size of an int in sparc), and later you're trying to access c[1], which starts at byte 4 but is out of bounds. On Fri, 1 Nov 2002, Xiaodong Li wrote: > Dear LLVM: > > I wrote a small testcase and I can

How can I run llvm-gcc in such a way that does not perform optimizations like constant propagation and dead code elimination? In particular, I'm trying to analyze cases where the code will look like: int *op; int *ip=(int*)malloc(sizeof(int)); op=ip; /* store/read into *ip */ free(op); but llvm-gcc transforms the above code to int *ip=(int*)malloc(sizeof(int)); /* store/read

Is there a way get the register on the LHS of a llvm instruction? (in case there is one) for example, given a "free %reg773", I want to find the matching "%something = malloc %", where %something is not necessarily %reg773. One way I found was to inmediately follow the use-def chain up from the free to the malloc. But instead, I want to put %reg773 in a set (possibly with many

It is valid that a function returns without using an Instruction::Ret as the last instruction? If so, can we assume that a BasicBlock with no successors is an exit BB?

How do I recognize that the callee function for a Instruction::Call is external? I have: CallInst *CI = cast<CallInst>(Inst); Function *G = CI->getCalledFunction(); nicolas

I've been trying to merge DSGraphs from different Functions, but i'm not going anywhere. For a Function F and DSGraph G, I tried this (assumming no SCC in the callgraph): vector<DSCallSite> CSV = G->getAuxFunctionCalls(); for (vector<DSCallSite>::iterator I = CSV.begin(), E = CSV.end(); I != E; I++) { Value *V = I->getCallInst().getOperand(0); Function &SF

llvm is giving me some assertion errors when trying to replace an instruction, and I'm not quite sure of what to do. I have the following snippet of code: switch((*I)->getOpcode()) { case Instruction::Malloc: { AllocaInst *AI; AI = new AllocaInst(cast<AllocationInst>(*I)->getAllocatedType(), cast<AllocationInst>(*I)->getArraySize());

lli executed the bytecode corresponding to test_3.0_ml.ll without a failure, even though Function() is accessing an invalid memory address. The original code is in test_3.0.c, and the malloc() in Create() has been replaced by alloca() in test_3.0_ml.ll. I expected lli to segfault or similar when testing my code. Are my assumptions erroneous? -------------- next part --------------

I've been trying to obtain a DSGraph for a complete module, but haven't succeeded. I tried - getting the DSGraph for each function, and copying the nodes to a Module graph. But since there are no nodehandles, apparently there is no merging of nodes. - modifying BUDataStructure Pass so that it will not execute removeDeadNodes() or removeTriviallyDeadNodes(). - couldn't get

Thanks for the tip, but what do you mean by "DSGraphs may be different shapes? On Sun, 1 Dec 2002, David Crowe wrote: > The way we are doing it, per Vikram's recommendation, is to provide a > mapping between the callers nodes and the callees nodes at the time of the > call. Given that the only way they can communicate is through the > parameters, globals, and return node,

can we assume that for(Function::aiterator I = F.abegin(); E = F.aend(); I!=E; ++I) { .... } or Argument *FA = &F.afront(); do { FA = FA->getNext(); .... } while (FA != &F.aback()); would iterate over function "F" formal parameters in order? All the tests I've run so far indicate so, but unless it is guaranteed, I don't see how can I match