similar to: [LLVMdev] Dominance frontier & Postdominance frontier

Hi, I was trying to use gold plugin to compile and run some of my passes on some large softwares like mysql, apache etc. When I compile using llvm-gcc, everything works fine. But, when I followed the instructions on http://llvm.org/docs/GoldPlugin.html#build to use gold plugin, I got errors like: /home/vadve/ssahoo2/local/bin/ld: error: modules/standard/libstandard.a: no archive symbol table

Swarup Kumar Sahoo wrote: > Hi, > I was trying to use gold plugin to compile and run some of my passes on some large softwares like mysql, apache etc. When I compile using llvm-gcc, everything works fine. But, when I followed the instructions on http://llvm.org/docs/GoldPlugin.html#build to use gold plugin, I got errors like: > > /home/vadve/ssahoo2/local/bin/ld: error:

On 1/22/13 12:07 PM, Sahoo, Swarup Kumar wrote: > Hi John and Silky, > > I can see a copy of 'giri' slicing project branch here http://llvm.org/viewvc/llvm-project/giri/. Though it may be little older, it will work I think. You can look at the code to see how we do the instrumentation. The giri project is supposed to contain both the static slicing code and the dynamic

Yes, the latest version can be downloaded from here https://github.com/liuml07/giri . The LLVM instrumentation code is in TracingNoGiri.cpp<https://github.com/liuml07/giri/blob/master/lib/Giri/TracingNoGiri.cpp> file and runtime code in Tracing.cpp. You may need to delete other unnecessary instrumentation code which you don't need. Thanks, Swarup. ________________________________

Hi, I was trying to build clang (svn head version) using clang-3.1 version. I am building a debug version using LTO/gold plugin. It is crashing with signal 9 when trying to link the bitcode files to build the executable. I am not sure, if this is a bug or just a memory issue. Has anyone done anything similar or noticed similar behavior? Any help would be welcome. Thanks, Swarup.

You will have to insert callinst to tracing functions immediately after the value is produced, not at the entry points. Giri code has many such examples of how to do this. Thanks, Swarup. ________________________________ From: llvmdev-bounces at cs.uiuc.edu [llvmdev-bounces at cs.uiuc.edu] on behalf of Jin Huang [54jin.huang at gmail.com] Sent: Thursday, January 02, 2014 9:47 PM To: llvmdev at

Hi Jin Huang, The updated Giri code is available at https://github.com/liuml07/giri<https://github.com/liuml07/giri.>. This will give you an idea of how to instrument the program to trace different values. You can modify the code to achieve your goals. -Swarup. ________________________________ From: Criswell, John T Sent: Thursday, January 02, 2014 2:12 PM To: Jin Huang; llvmdev at

Oh, OK. I didn't check the code. I think it only contains your flow tracking analysis code, isn't it. Our 'Giri' project was completely separate from it. Should we merge it with this or keep it as a separate project? -Swarup. ________________________________________ From: John Criswell [criswell at illinois.edu] Sent: Tuesday, January 22, 2013 12:12 PM To: Sahoo, Swarup Kumar Cc:

Thank you , James. But this way needs to determine whether the instruction is a LOAD/STORE ,and if I want to preserve the operations on these variables for further dynamic symbolic execution usage(just like KLEE ,but it's a static analysis ,not dynamic),I had to backtrace instructions ,this may be not very efficient if the operation is very complex! -Jin Huang 2014/1/3 Jin Huang

> >> Yes, ar and nm actually take -plugin option and if I pass this option, then I don't get these errors. But, llvm-gcc doesn't automatically pass this option to ar. > >You probably just need to copy the gold plugin to >$PREFIX/lib/bfd-plugins. Let me know if that works. If not I will >check if there is a bug in the plugin search in binutils. There is no problem in

Two suggestions: - Use a 64-bit operating system. - You could disable the LTO optimizations by commenting out a few lines in tools/lto/LTOCodeGenerator.cpp to determine if the problem is a pass or with LTO itself. On Wed, Oct 17, 2012 at 10:38 AM, Sahoo, Swarup Kumar <ssahoo2 at illinois.edu> wrote: > Hi, > > I was trying to build clang (svn head version) using clang-3.1

On 1/13/13 11:06 PM, Criswell, John T wrote: > There is code that does this for older versions of LLVM. I believe it is in the giri project in the LLVM SVN repository. I can look into more details when I get back from vacation. Swarup may also be able to provide information on the giri code. I took a quick look, and the dynamic slicing code doesn't appear to be checked into the giri

Hi, I want to find the dynamic slice of a given code. Is this already implemented in LLVM?. If not which functions are there in the LLVM to do dynamic analysis.

Hi John and Silky, I can see a copy of 'giri' slicing project branch here http://llvm.org/viewvc/llvm-project/giri/. Though it may be little older, it will work I think. You can look at the code to see how we do the instrumentation. Thanks, Swarup. ________________________________________ From: John Criswell [criswell at illinois.edu] Sent: Tuesday, January 22, 2013 10:29 AM To:

---- Original message ---- >Date: Wed, 10 Mar 2010 22:09:22 -0800 >From: Nick Lewycky <nicholas at mxc.ca> >Subject: Re: [LLVMdev] glod plugin use >To: Swarup Kumar Sahoo <ssahoo2 at illinois.edu> >Cc: llvmdev at cs.uiuc.edu, Rafael Espindola <espindola at google.com> > >Swarup Kumar Sahoo wrote: >> Hi, >> I was trying to use gold plugin

Swarup Kumar Sahoo wrote: >>> Yes, ar and nm actually take -plugin option and if I pass this option, then I don't get these errors. But, llvm-gcc doesn't automatically pass this option to ar. >>> >> You probably just need to copy the gold plugin to >> $PREFIX/lib/bfd-plugins. Let me know if that works. If not I will >> check if there is a bug in

Hi, I found the -postdomfrontier pass in *llvm*.org/docs/Passes.html, but 'opt' does not accept it. I could not find the relevant codes in PostDominance.cpp in SVN trunk, but I found some relevant codes here http://opensource.apple.com/source/clang/clang-137/src/lib/Analysis/PostDominators.cpp. I am wondering why this pass is removed and what should I do if I want the post

On Thu, 9 Nov 2006, Ryan M. Lefever wrote: Sorry I never responded to this: > In the literature (see below for a reference), when a dominance frontier > is computed, it is computed from a CFG that contains a dummy entry node > and dummy exit node. Further, those dummy nodes are potential members > of the (post-)dominance frontier for a given basic block. In LLVM, I > could not

In the literature (see below for a reference), when a dominance frontier is computed, it is computed from a CFG that contains a dummy entry node and dummy exit node. Further, those dummy nodes are potential members of the (post-)dominance frontier for a given basic block. In LLVM, I could not figure out a way to determine if the dummy entry node is a member of the post-dominance frontier of

I just finished upgrading our LLVM to 2.4 and I immediately ran into a problem with dominance frontier calculation: llvm/lib/VMCore/PassManager.cpp:714: void llvm::PMDataManager::verifyDomInfo(llvm::Pass&, llvm::Function&): Assertion `0 && "Invalid dominator info"' failed. Strangely enough, the Pass running when the assert triggers is Dominance Frontier