search for: jhom

...unction `nsCacheEntryDescriptor::Close()': : undefined reference to `nsCacheService::CloseDescriptor(nsCacheEntryDescriptor*)' 2) On a whim, I wanted to try using llvm-gcc4. However, when I invoke it, I get a relocation error. prompt% llvm-gcc --version llvm-gcc: relocation error: /home/jhom/LLVM/1.8/llvm-gcc4-1.8-x86-linux/lib/libstdc++.so.6: undefined symbol: _ZNSt13basic_istreamIwSt11char_traitsIwEE6ignoreEi Any suggestions are greatly appreciated! -- Jerry Hom

...e()': > : undefined reference to `nsCacheService::CloseDescriptor(nsCacheEntryDescriptor*)' > > > 2) On a whim, I wanted to try using llvm-gcc4. However, when I invoke > it, I get a relocation error. > > prompt% llvm-gcc --version > llvm-gcc: relocation error: /home/jhom/LLVM/1.8/llvm-gcc4-1.8-x86-linux/lib/libstdc++.so.6: undefined symbol: _ZNSt13basic_istreamIwSt11char_traitsIwEE6ignoreEi > Did you compile llvm-gcc4, or did you download a precompiled version from somewhere? Most likely, the problem is that llvm-gcc4 was built against a different version...

Hi, All I''m just new to this list, I''m started to dealing with OSS Xen to XCP migration earlier this week, seem the domUs installed via virt-install are function well, but domUs copied via dom0 are failed during bootup process, finally I realize if I could fix this on my OSS Xen will resolve all. Can I have any suggestion from you please ? *Details* *Source XenHost* OS:

Hello. I'm new to LLVM, so I apologize if this is a basic question. I've used llvm-gcc to produce an LLVM assembly file (-S) as well as the bytecode file. I then used 'analyze -print-cfg' on the bytecode and realized that some optimizations have been applied against the corresponding assembly file. If I understand the operations, the emitted assembly file has no optimizations,

Well, I'm off to a bad start in this group. I found the answer to my question just 10 minutes after I posted, and it was a basic question. To disable optimizations in the assembler and linker phases, you pass the -disable-opt flag in the command line like this: prompt% llvm-gcc -Wa,-disable-opt -Wl,-disable-opt <source> -- Jerry Hom

I'm trying to insert a call to fprintf(stderr, ...). I've looked at the emitted assembly from llvm-gcc, and it consists of a LoadInst (of stderr) and CallInst. It looks like this: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %tmp.1 = load %struct._IO_FILE** %stderr ; ty=%struct._IO_FILE* %tmp.0 = call int (%struct._IO_FILE*, sbyte*, ...)* %fprintf(%struct._IO_FILE* %tmp.1, sbyte*

Hello. I'm wondering if LLVM can generate a call graph across multiple files? Does 'analyze' accept multiple bytecode files? Can I concatenate multiple bytecode files into a single file for input? Thanks. -- Jerry Hom

I'm running into an error message about PHINode which I don't understand too well. Hopefully I can explain this clearly. At a high level, I'm just trying to add a fprintf() statement at the end of functions to show me a timestamp. However, I also add an if/then such that I can turn the printing on/off. When using llvm-gcc, I've disabled linker and assembler optimizations in

This question is similar to Ryan Lefever's post on May 1, 2006 about printf declaration. In my case, I want to insert fprintf(stderr, ...) calls. I'm new to LLVM, and I don't know what's the recipe for putting together the arguments. Can someone give me basic instructions or point me in the direction on what to do? I can't find any more documentation on this. Thanks!