similar to: [LLVMdev] LLVM Instruction*->getOperand() not working properly for ICMP

Hi, On 10/01/13 10:56, Alexandru Ionut Diaconescu wrote: > Hello everyone ! > > In my pass I inspect the penultimate instruction from every basic block in > runOnFunction(). I am interested in ICMP instructions only. > > |if(BB->size()>1) > if(last->getPrevNode()) > { > previous = last->getPrevNode(); > ok=1; > } > |

Hi Duncan, Thank you for your response! Now it works. But I have another problem: Do you know how can I iterate through a Value*? For instance, I have : Instruction *I1; // I take only I1 that are ICmp errs()<<"\n "<<*I1->getOperand(0)<<" \n"; // %3 = load i32* %c, align 4 As expected. Now I want to get %c from the previous to use "%c" in

Dear members, Is there a way to turn a strwidth of a string into a number of lines that ist needed to display the string when using par(mar = c(?,4,4,2)) #### x = 1:5 names(x) = c("ZZZZZ","ZZZZZzzzzz","ZZZlllll","TTTTT","Zzhtsddfg" ) par(mfrow = c(1,2)) par(mar = c(8,4,4,2)) barplot(x, las = 3, main = "8 is to much") par(mar =

The IR after LSR is: *** IR Dump After Loop Strength Reduction *** ; Preheader: entry: tail call void @fill_array(i32* getelementptr inbounds ([10 x i32], [10 x i32]* @buffer, i32 0, i32 0)) #2 br label %while.body ; Loop: while.body: ; preds = %while.body, %entry %lsr.iv = phi i32 [ %lsr.iv.next, %while.body ], [ 0, %entry ] %uglygep = getelementptr

Hm, no. I expect byte addresses - everywhere. The compiler should not know that the arch needs word addresses. During lowering LOAD and STORE get explicit conversion operations for the memory address. Even if my arch was byte addressed the code would be false/illegal. Boris > Am 09.06.2020 um 19:36 schrieb Eli Friedman <efriedma at quicinc.com>: > > Blindly guessing here,

Hi. In my backend I get false code after using StrengthLoopReduction. In the generated code the loop index variable is multiplied by 8 (correct, everything is 64 bit aligned) to get an address offset, and the index variable is incremented by 1*8, which is not correct. It should be incremented by 1 only. The factor 8 appears again. I compared the debug output

On Thu, 2012-01-26 at 15:12 -0600, Sebastian Pop wrote: > On Thu, Jan 26, 2012 at 2:49 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > Thanks! Did you compile with any non-default flags other than -mllvm > > -vectorize? > > I used -O3 and -vectorize, no other non-default flags. If I run clang -O3 -mllvm -vectorize -S -emit-llvm -o test.ll test.c then I get no

Hello John, I was following your procedures and I isolated the problem. The problem are represented by the basic blocks with only one element. for (Function::iterator II = F.begin(), EE = F.end(); II != EE; ++II, ++ii) { BasicBlock* BB=II; if (BB->getTerminator()) { Instruction* current = BB->getTerminator(); Instruction* previous;

Hi all, I stumbled over a problem in ilist_node::getPrevNode(). It crashes when invoked for the first element in a list. It's because the Prev pointer of a first list element does not point to the sentinel but is just null. First question: Is this really a bug or am I doing something wrong? Second question: If it is a bug, what should be the correct behaviour? Either change insert() to let

Hello everyone, I have a segmentation fault while running an LLVM pass. I need to use BBterminators array outside the iterating "for" loop for basic blocks. It seems that LLVM does not protect the addresses ( note: TerminatorInst *BasicBlock::getTerminator() ) when iterating through the loop, so I need to keep in BBterminators "Instruction" type elements, not

Hello John, I was following your procedures and I isolated the problem. The problem are represented by the basic blocks with only one elements. for (Function::iterator II = F.begin(), EE = F.end(); II != EE; ++II, ++ii) { BasicBlock* BB=II; if (BB->getTerminator()) { Instruction* current = BB->getTerminator(); Instruction* previous;

Hello, Thank you for your answer. If I want to use then I have error: ‘NodeTy* llvm::ilist_half_node<NodeTy>::getPrev() [with NodeTy = llvm::Instruction]’ is protected error: ‘llvm::ilist_half_node<llvm::Instruction>’ is not an accessible base of ‘llvm::Instruction’ Do you know any other method to access the previous instruction of a terminator instruction? PS: back() is not an

getPrevNode<http://llvm.org/docs/doxygen/html/classllvm_1_1ilist__node.html#a77b897207ef0a1ae95c404695aed9a4b>() Get the previous node, or 0 for the list head. I don't see any method like hasPrevNode. It can be a weird problem because "current->getPrevNode()" is indicating to "current" itself (the problem appears for the BB with only one element)? On Thu, Dec

Hi All, For various obscure reasons I'd like to detect the condition when X86 CALL instruction immediately precedes a function epilogue in the final emitted code, and insert a NOP between them if that happens. My initial attempt at it looked like this: MachineBasicBlock& MBB; MachineBasicBlock::iterator MBBI; <-- points to where the epilogue would be inserted if (MBBI != MBB.begin()

I solved by checking if(BB->size()>1) Thank you all for the help ! Now debugging the next segfault. On Thu, Dec 20, 2012 at 12:59 PM, Alexandru Ionut Diaconescu < alexandruionutdiaconescu at gmail.com> wrote: > getPrevNode<http://llvm.org/docs/doxygen/html/classllvm_1_1ilist__node.html#a77b897207ef0a1ae95c404695aed9a4b>() > Get the previous node, or 0 for the list

On Thu, Jan 26, 2012 at 3:41 PM, Hal Finkel <hfinkel at anl.gov> wrote: > On Thu, 2012-01-26 at 15:36 -0600, Sebastian Pop wrote: >> arm-none-linux-gnueabi > > Indeed, adding -ccc-host-triple arm-none-linux-gnueabi I also get Minor remark: please use -target instead of -ccc-host-triple that is now deprecated. Thanks for looking at this testcase. Sebastian -- Qualcomm

Hello, I am kind of confused with this situation. I am attempting to build a PDC using TLS/SSL with the following version of software. Samba 3.0.23 OpenLDAP 2.3.19 Fedora Core 5 When I startup the Samba server via the "service" command (service smb start) I get the following errors in my logs. Using SSL: Jul 13 09:52:34 prism smbd[23161]: smbldap_search_suffix: Problem during

Hi all, Please i need help, I have a method that takes 2 arguments with type char *: void branchPredict(char *b1, char *b2){ --- -- } i'm supposed to add this method, in an IR basic bloc: to add it into a basic bloc i do: //i: is the basic bloc std::vector<Value*> void_43_params; Constant* tbname = ConstantArray::get(M.getContext(),i->getNameStr() , true); Constant* pbname =

On Thu, 2012-01-26 at 15:36 -0600, Sebastian Pop wrote: > arm-none-linux-gnueabi Indeed, adding -ccc-host-triple arm-none-linux-gnueabi I also get vectorization (even though I don't get vectorization when targeting x86_64). I'll let you know what I find. -Hal -- Hal Finkel Postdoctoral Appointee Leadership Computing Facility Argonne National Laboratory

This is a follow-up to the thread ending with: http://rubyforge.org/pipermail/vpim-talk/2008/000120.html I too am in search of some ruby parser for icalendar which properly handles timezones on the datetimes in the icalendar RFC. As I understand it there are actually three types of times. 1) UTC times with a string form of yyyymmddThhmmssZ note the trailing Z indicates zulu time aka utc. 2)