similar to: [LLVMdev] Alias analysis information using Profile Pass.

Hi all, Can anybody tell me how to generate a mov instruction in llvm. I found out there is a MemMove Inst but I dont think that is what I want. I want to bypass a load instruction. So i need to generate a mov instruction which will help me remove the redundant load. Thanks -- Abhinav Karhu MS Computer Science Georgia Institute of Technology

Hi Eli, I want to do redundant load elimination. A simple example for that would be load (a1) , r1; --- --- --- load (a2) , r2; Now if a1 and a2 are alias of each other and the value stored in a1 and a2 remains the same then I can say that the second load is redundant. I can replace the second load with these instructions mov r0 r1; mov r0 r2; and remove the load instruction. This is the thing

On Fri, Nov 14, 2008 at 2:13 PM, Karhu, Abhinav R <akarhu3 at gatech.edu> wrote: > > Hi all, > Can anybody tell me how to generate a mov instruction in llvm. I found out there is a MemMove Inst but I dont think that is what I want. > I want to bypass a load instruction. So i need to generate a mov instruction which will help me remove the redundant load. At the LLVM IL level,

r2->replaceAllUsesWith(r1); r2->eraseFromParent(); LLVM makes it easy! ;-) --Owen On Nov 15, 2008, at 11:02 PM, Karhu, Abhinav R wrote: > Hi owen, > Can you please elaborate as to what should I do? Find all the > instruction which have r2 in it and replace all of them with r1 and > then remove the load instruction? > Thanks > Abhinav > ----- Original Message

Hi owen, Can you please elaborate as to what should I do? Find all the instruction which have r2 in it and replace all of them with r1 and then remove the load instruction? Thanks Abhinav ----- Original Message ----- From: "Owen Anderson" <resistor at mac.com> To: "LLVM Developers Mailing List" <llvmdev at cs.uiuc.edu> Sent: Friday, November 14, 2008 11:22:32 PM

Hey all, I compiled the code using both the -c and -S options. I can get the human readable IR but I still cannot run it using lli. The output of the IR looks like this. Can some of the information tell me about why the program is not running. ; ModuleID = 'Hel.c' target datalayout = "e-p:32:32" target endian = little target pointersize = 32 target triple =

I have compiled the program using the following option. llvm-gcc --emit-llvm Hello.c -o Hel.bc I have used llvm in the past but this sort of problem is new for me. I dont know how to go about it. Thanks for the reply. -- Abhinav Karhu MS Computer Science Georgia Institute of Technology

Hey all, I am getting this weird error which suddenly popped up today. Whenever I am trying to execute a piece of code using lli I get this following error. Error loading program 'sample.bc': Invalid bytecode signature: 464C457F (Vers=0, Pos=4) I dont know what to do..Please reply to this question. Please... -- Abhinav Karhu MS Computer Science Georgia Institute of Technology

Hi all, I am facing new problem with lli now. I am having a sample code in which I am doing malloc for two variables. If I compile the code with normal gcc the program runs without any warning of any sorts. If I compile the program to convert it into a bytecode file and then run it through lli it segfaults and the program aborts. This is the stack trace after execution. lli((anonymous

Hey All, I am running into a problem with running the program using LLVM JIT. I get the following error. lli: error loading program 'Hel.bc': Bitcode stream should be a multiple of 4 bytes in length. Hel.bc is the bytecode for my simple HelloWorld program. Please give me some insight into this. Thanks -- Abhinav Karhu MS Computer Science Georgia Institute of Technology

Hey all, I want to run the profile.pl file in the utils folder but I am having trouble running the script because the profile_rt.so file is not generated properly. So in order to generate that I am running make in the /runtime/libprofile. when I do that I get the following message. llvm[0]: Linking Release Lodable Module profile_rt.so llvm[0]: Compiling BasicBlockTracing.c for Release build

Hello all, I am new to the llvm infrastructure so if this question is already resolved please redirect me to that link. I am writing a pass for flow sensitive and context sensitive alias analysis.for that i require the previous and next instruction of all the instructions.Is there an intrinsic that can help me getting the instruction numbers directly because i read at the llvm.org website making

I think you can refer to lib/VMCore/AsmWriter.cpp for these things. E.g., You can use 'if(I.isTerminator())' if a instruction 'I' is terminator. You can use 'if (isa<BranchInst>(I))' if a instruction 'I' is 'br' instruction. and so on... Thx, Seung Jae Lee ---- Original message ---- >Date: Mon, 2 Jul 2007 17:15:00 -0400 (EDT) >From: abhi232

Thanks a lot for the help guys.. I will try to get the information on these lines... Thank you once again... Will bug you guys after some time now. :) > I think you can refer to lib/VMCore/AsmWriter.cpp for these things. > > E.g., > You can use 'if(I.isTerminator())' if a instruction 'I' is terminator. > You can use 'if (isa<BranchInst>(I))' if a

> On Mon, 2 Jul 2007 abhi232 at cc.gatech.edu wrote: >> Hello all, >> I am new to the llvm infrastructure so if this question is already >> resolved please redirect me to that link. >> >> I am writing a pass for flow sensitive and context sensitive alias >> analysis.for that i require the previous and next instruction of all the >> instructions.Is there

On Fri, 3 Oct 2008 18:03:17 -0400 (EDT) "Kolakkar, Pranay B" <pranay.kolakkar at gatech.edu> wrote: > Hi Paalanen, > > I am a graduate student for Computer Science at Georgia tech and I came > across your page on freedesktop.org/~pq/rules-ng > > It was interesting that you had listed all the registers and addresses of > the nvidia gpu. Please let me know as

Dear UseRs: I intend to write a package to handle the basic operations a revenue management analyst has to deal with on a regular basis (e.g., demand untruncation, capacity allocation, pricing decisions, etc.). Following the directions posted by Peter Rossi (Making R Packages under Windows: A Tutorial, January 2006) I was able to build an interim package that currently consists of one simple

I'm running samba 3.0.13 on a debian box, sharing out NFS filesystems from a solaris box. I know, bad idea.. file locking, etc... non-issues at the moment. I've discovered that if I make a share from a local filesystem, the PC's mounting the share see their quota (hard limit, not soft which is another matter) as their capacity and their appropriate free space... this is good.

On 7/23/07, abhi232 at cc.gatech.edu <abhi232 at cc.gatech.edu> wrote: > Hi Bill, > Thanks a lot for your response.But my problem still remains.The thing is i > am having a data type std::vector<Value*> as i am checking for the > variables in Store Instructions and Malloc Instructions.For store the case > is straightforward as discussed earlier.I want the same Value*

I'm loading a program from CD-ROM .. the splash screen comes up, but then a dialog box comes up with "Failed to initialize ..." btw, I'm running wine under the 2.2.16 linux kernel and the partition I'm running this program from is a win2k ntfs partition (the program is not limited to win2k however.. just what i have) Here are some of the errors (randomly selected) that