similar to: [LLVMdev] Kaleidoscope Build Error

Hi everyone, sorry to bother you guys. I’m moving a linux application that uses LLVM from version 3.0 to 3.2, and I’m getting a relocation error that I didn’t used to get with the older version of LLVM. I believe that I correctly turned on fPIC, but I’m getting this warning anyway: /usr/bin/ld: /scratch/thirdparty/default/lib/libLLVMBitReader.a(BitcodeReader.o): relocation R_X86_64_PC32

Please ignore, I think this is an issue with an older version of binutils. ________________________________ From: Damien D Neff Sent: 1/4/2013 1:25 PM To: llvmdev at cs.uiuc.edu Subject: Help with linking llvm 3.2 into an .so Hi everyone, sorry to bother you guys. I’m moving a linux application that uses LLVM from version 3.0 to 3.2, and I’m getting a relocation error that I didn’t used to get

"Michael Ness" <mike at liquido2.com> writes: > Some additional information about the setup may be of interest: > > I'm using CMake to generate the make files for MSYS. All appears to go well > until the tools are compiled. > > The undefined references seem to be coming from the libraries that were just > compiled: > >

Some additional information about the setup may be of interest: I'm using CMake to generate the make files for MSYS. All appears to go well until the tools are compiled. The undefined references seem to be coming from the libraries that were just compiled: ../../lib/libLLVMBitReader.a(BitcodeReader.cpp.obj):fake:(.rdata$linkonce_ZTVN4llvm8ConstantE+0x10): undefined reference to

Thanks Óscar, I decided to forget 2.5 and move on to 2.6 which built fine with CMake. However, I continue to have problems with llvm-config when using the following command: $ llvm-g++ out.s -o out `llvm-config --ldflags --libs core support system x86` I consistently get undefined reference errors that are from libraries that should've been linked already. The errors change depending

I'm running into a compile problem with commit 131656 of compiler-rt. I'm compiling llvm with: $ CXX='g++ -m32' CC='gcc -m32' CFLAGS=-m32 CXXFLAGS=-m32 LDFLAGS=-m32 ./configure --disable-bindings --{build,host,target}=i686-apple-darwin --enable-targets=x86,x86_64,cbe --enable-optimized $ make $ make install It will eventually error out with:

Dear R-friends, ? I am using the epicalc package and the manual by V. Chongsuvivatwong "Analysis of epidemiological data using R and Epicalc" to get the hang of some basic epidemiological analyses.??? ? After running all the analyses of chapter 7, one is supposed to wrap it up by saving the data writing: ? ? > save(.data, file = "Chapter7.Rdata") ? ...?after writing the

Hi Gregory, I had a similar problem a while back, see: http://lists.cs.uiuc.edu/pipermail/llvmdev/2010-June/032508.html for my solution to this problem. Basically I ended up linking in the *.o files needed to get exactly what I needed (in your case, the file defining CloneModule and whatever else you need), without pulling in too much (causing duplicate symbol issues with opt). Hopefully this

I thought it would, but it seems to be missing the CloneModule symbol (which isn't called from anywhere in the LLVM codebase). On Fri, Jul 15, 2011 at 7:43 AM, John Criswell <criswell at illinois.edu>wrote: > On 7/14/11 7:36 PM, Gregory Malecha wrote: > > Hello, > > I have a question about how to set up a Makefile that generate a .so that I > can load into opt

Hi Duncan, On Wed, Jun 29, 2011 at 11:32 PM, Duncan Sands <baldrick at free.fr> wrote: > Hi Gregory, > > > I've tried the LTO options with dragonegg, but I haven't had any luck > getting it > > to work, > > details? > I think this is a misunderstanding on my part about what the dragonegg plugin is supposed to do. I am looking for something that will

On 7/14/11 7:36 PM, Gregory Malecha wrote: > Hello, > > I have a question about how to set up a Makefile that generate a .so > that I can load into opt without getting duplicate symbols. Here's > what I have right now: Have you tried removing LINK_COMPONENTS below? I suspect that the opt binary already contains everying in the transformUtils library. -- John T. > >

I would recommend Newlib. It's easy to configure and compile using Clang. http://sourceware.org/newlib/ - xi On Jun 23, 2011, at 2:07 AM, Gregory Malecha wrote: > Hello, > > I'm wondering if anyone had any success (even a small amount) compiling any variant of libc to llvm bitcode? > > -- > gregory malecha > _______________________________________________ >

Hi Xi, How did you compile it with clang? I ran into problems: clang -emit-llvm -DPACKAGE_NAME=\"newlib\" -DPACKAGE_TARNAME=\"newlib\" -DPACKAGE_VERSION=\"1.19.0\" -DPACKAGE_STRING=\"newlib\ 1.19.0\" -DPACKAGE_BUGREPORT=\"\" -DPACKAGE_URL=\"\" -I. -DMISSING_SYSCALL_NAMES -fno-builtin -DMISSING_SYSCALL_NAMES -fno-builtin -O2 -c -o

Hi Dmitri, You might want to try replacing the call to JMM->invalidInstructionCache() with a call to TheExecutionEngine->finalizeObject(). If you are getting a non-NULL pointer from getPointerToFunction but it crashes when you try to call it, that is most likely because the memory for the generated code has not been marked as executable. That happens inside finalizeObject, which also

Hi, Per my previous post about a patch to run the test suite under valgrind (btw, can I commit it?), I've tracked a few memory leaks in LLVM. I've fixed 4 trivial ones, but there are a few more not-so-trivial remaining. Take a look at two common reports by valgrind: 240 (144 direct, 96 indirect) bytes in 3 blocks are definitely lost in loss record 17 of 20 at 0x4023614: operator

For my application it is important that I get bit code out the other end and I couldn't get the linux linker to link llvm. Is there a way to do this with other tools? I noticed a TODO in the llvm-link.cpp file that said that this should be done so I figured that that would be the right place to do it. 2011/7/30 Rafael Ávila de Espíndola <rafael.espindola at gmail.com> > On

Hi Gregory, > I've tried the LTO options with dragonegg, but I haven't had any luck getting it > to work, details? Ciao, Duncan. I'm not sure if it is related to my installation of gold or my gcc > configuration. > > I'm really interested in getting the resulting program as bitcode though. Will > LTO do that for me? What I'm trying to do is build a

Hi Rafael, I've tried the LTO options with dragonegg, but I haven't had any luck getting it to work, I'm not sure if it is related to my installation of gold or my gcc configuration. I'm really interested in getting the resulting program as bitcode though. Will LTO do that for me? What I'm trying to do is build a bitcode version of libc that I can combine with a program and

Hi, I am using current SVN and in the last week or so, something causing the following assertion failure has changed. void llvm::PATypeHolder::addRef(): Assertion `Ty && "Type Holder has a null type!"' failed. The corresponding stack trace is: #0 0x000000339ec332f5 in raise () from /lib64/libc.so.6 #1 0x000000339ec34b20 in abort () from /lib64/libc.so.6 #2

Hi Andrew, Am 04.06.2013 02:13, schrieb Kaylor, Andrew: > Hi Dmitri, > > You might want to try replacing the call to JMM->invalidInstructionCache() with a call to TheExecutionEngine->finalizeObject(). If you are getting a non-NULL pointer from getPointerToFunction but it crashes when you try to call it, that is most likely because the memory for the generated code has not been