similar to: [LLVMdev] difference between replaceAllUsesWith and replaceUsesOfWith?

On Sat, Aug 9, 2014 at 6:06 AM, Tim Northover <t.p.northover at gmail.com> wrote: > Hi Rob, > > On 9 August 2014 02:03, Rob Jansen <jansen at cs.umn.edu> wrote: > > Why is the first for loop not equivalent to the second? > > In the second loop, "*ui" is an llvm::Use object. It's owned by a > User, but isn't a subclass of one. To match the

> I think this is incorrect since the use_iterator is defined as: > > `typedef value_use_iterator<User> use_iterator;` > > and indeed `ui->getUser();` is not a callable function on a User. What am I > missing? Ah, looks like this has changed this year. What I said applies to trunk; you may be right there about 3.4 or earlier (looks plausible from the key commit).

Ahh ha! I believe that `User::replaceUsesOfWith` modifies the use list, which probably invalidated the use_iterator and caused my code to miss some of the uses. The following works great! (I'll spare you the details about what I am actually trying to do here, unless you really want to know more.) Thanks, Rob =========================== while(GV->getNumUses() > 0) { User* u =

Hi, I'm trying to change all uses of a global variable so they'll use a different global variable. For instance, I have: @a = global [100 x [100 x i64]] zeroinitializer, align 16 @b = global [100 x [100 x i64]] zeroinitializer, align 16 And I want to change the use %arrayidx = getelementptr inbounds [100 x [100 x i64]]* @a, i32 0, i64 %3 to %arrayidx = getelementptr inbounds

On Apr 16, 2008, at 2:50 AM, heisenbug wrote: > > And now here is my educated speculation: > There are 2 things that became slower > 1) Use::getUser() > 2) Use::get/set due to tagging. > > The former is seldom called: > > $ find lib -name "*.cpp" | xargs grep "getUser(" | wc -l > 41 The majority of those aren't actually Use::getUser, but

On Apr 16, 2:13 am, Dan Gohman <goh... at apple.com> wrote: > Hi Gabor, > > Can you provide performance data for this? I'd > like to know what affect these changes have on > compile time. Hi Dan, Unfortunately, no. I can feed you with some speculation, though, see below. The reason why I cannot do measurements (at the moment) is that - I have no experience with

Hello, Calling `replaceUsesOfWith` with a value in a different addrspace does not keep the addrspace of a GEP consistent. Is this known? Is this a bug or expected behaviour? Minimal counterexample link <https://gist.github.com/bollu/152ba5e1c20c03c7fc6d8c7b23ba828f> Reproduced here: #include <iostream> #include "llvm/ADT/APFloat.h" #include

Use *last = nullptr; for (Use &U : Src->uses()) last = &U; last->getUser()->dump(); Or any better solution? 2018-04-20 19:19 GMT+02:00 Dávid Bolvanský <david.bolvansky at gmail.com>: > Is: > > > 2018-04-20 18:07 GMT+02:00 Dávid Bolvanský <david.bolvansky at gmail.com>: > >> Hello, >> >> Code: https://godbolt.org/g/EG4Wi6

Hi all, I can get access to the data stored in globals as follows: GlobalVariable* pGvarAct=_set_a_valid_global_; const_ptr_indicesVec=_the_indexes_ llvm::Constant* pConst; pConst=ConstantExpr::getGetElementPtr(pGvarAct, const_ptr_indicesVec); and e.g. use this to repalece an argumrnt of an instruction like: I->setOperand(someArgumentIndex,pConst); However, getGetElementPtr returns a

Hello, today I download poolalloc from " https://github.com/llvm-mirror/poolalloc". and I compiled it with LLVM3.3. Then when I excute "make",I get the error: AddressTakenAnalysis.cpp:18:30:fatal error:llvm/IR/CallSite.h: No such file or directory. ​I correct the path of callsite.h to "llvm/Support/Callsite.h". there is another error: AddressTakenAnalysis.cpp:In

getInitializer returns the complete initializer of the global. My globals are complex nested structs, from which I want to extract e.g. one double datafield. Example: >From a struct Stuct having double,array(3xint),float fields, I could extract the second int value using the index sequence 0,1,1 with getGetElementPtr. This gives me a constantPointer, but I would need to get a constInt...

Right now I am trying to capture the function name and the number of arguments , so this following is the pass I wrote . ------------------------------------------------------------- struct pass06a : public ModulePass { virtual bool runOnModule(Module &M) { std::vector<const Type*> pList; pList.push_back( PointerType::get(Type::SByteTy) ); pList.push_back(

There were fixes in 3.17 that were supposed to help this, but apparently they didn't help enough. See https://bugs.freedesktop.org/show_bug.cgi?id=72180 -- basically some sort of card setup failure on our part is causing our ctxsw to die, but nvidia's appears to be more resilient to the screwups. On Fri, Dec 5, 2014 at 1:16 PM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >

Hi all, I have a pass that duplicates all instructions and keeps a "reference map" of duplicated values, which is later used to update all the references, thus creating a whole copy of a previous function. How I was doing it is calling "replaceUsesOfWith" in all the replicated instructions. This used to work on 2.9 and older, in svn it no longer does in the case of PHI nodes,

We have finished a beta version of the feature. A webrev for it can be found here: http://cr.illumos.org/~webrev/sensille/fits-send/ It adds a command ''zfs fits-send''. The resulting streams can currently only be received on btrfs, but more receivers will follow. It would be great if anyone interested could give it some testing and/or review. If there are no objections,

Hello, I am trying to reset the operands of instructions. What I am doing is, I am finding all Uses of a specific operand in an Instruction and resetting it with a new value using "setOperands(int, Value*)". I am doing as: for (auto vmitr=vm.begin(), vsitr=vs.begin(); vmitr!=vm.end() && vsitr!=vs.end(); vmitr++, vsitr++){ // I have two *Value ( operands) for ( auto myitr

On 10/21/11 11:15 AM, Carlos Sánchez de La Lama wrote: > Hi all, > > I am trying to replace all uses of a value1 *inside of a given function* > to use value2. My strategy was to iterate all the basic blocks, again > iterating all the instructions in each basic block, and using > replaceUsesOfWith(value1, value2) for every instruction. > > This used to work all right, but

Hi, I was tracking down a performance regression and noticed that LoopStrengthReduce hoists loop invariants (e.g., the initial formulae of indvars) in the reverse order of how they appear in the loop. This reverse order creates troubles for the StraightLineStrengthReduce pass I recently add. While I understand ultimately SLSR should be able to sort independent candidates in an optimal order,

https://bugzilla.samba.org/show_bug.cgi?id=7194 Summary: Getting --inplace and --sparse to work together Product: rsync Version: 3.0.7 Platform: All OS/Version: All Status: NEW Severity: enhancement Priority: P3 Component: core AssignedTo: wayned at samba.org ReportedBy: jansen at

Hi LLVMdev members I found something strange in ConstantExpr::replaceUsesOfWithOnConstant() function. in lib/VMCore/Constants.cpp file 2118 if (getOpcode() == Instruction::GetElementPtr) { 2119 SmallVector<Constant*, 8> Indices; 2120 Constant *Pointer = getOperand(0); 2121 Indices.reserve(getNumOperands()-1); 2122 if (Pointer == From) Pointer = To; 2123 2124 for