similar to: [LLVMdev] C Frontend updated

The problem with llvmgcc invoking as instead of llvm-as has been fixed. 3 llvm-specific executables needed to be copied into a default location where gcc can find them if it does not find them as configured (gcc was configured for the paths on our research machines and copied over). Let us know if you have additional problems. --Vikram > -----Original Message----- > From: llvmdev-admin

Dear Prof. Adve, Now I can use llvmgcc to compile a .c file into .bc file. But I still have trouble simply run the .bc code. Below is the sequence I got when I tried. I really don't know what's going on here. Please let me know how can I fix it. Thanks, xiaodong xli3|csil-suna48|~/cs426|[13]% llvmgcc scalarize.c -o scalarize xli3|csil-suna48|~/cs426|[14]% scalarize Cannot load value of

Thanks, Chris, Below is the output of 'llvmgcc he.c -v' xli3|csil-suna27|~/cs426|[35]% llvmgcc he.c -v Reading specs from /usr/dcs/projects/cs426/Software/gcc_install/bin/../lib/gcc-lib/llvm/3.1/specs Configured with: /home/vadve/lattner/cvs/gcc/configure --srcdir=/home/vadve/lattner/cvs/gcc --prefix=/home/vadve/lattner/cvs/gcc_install_sparc --target=llvm --enable-languages=c

Here are a few bits of news about the C frontend: 1. I have ported it to compile and run under linux. On our research machines, it is available here: /home/vadve/lattner/cvs/gcc_install_x86 Note that this C compiler will still generate "64-bit" LLVM files, thus the code will probably not work right if run through the C backend, then compiled on x86. For middle-end

> ok, I finished recompiling llvm, and when trying to use llvmgcc I get > the following error message: > as -o /var/tmp//ccIuGwwQ.o /var/tmp//ccZvtWKU.s > as: error opening '/var/tmp//ccIuGwwQ.o': file exists! > Use -f command line argument to force output Okay, this is the problem. For some reason, it is trying to use the 'as' in your path, which is the LLVM

Dear llvm, I just tried to compile a simple file hello.c. But each time I used llvmgcc hello.c, it gave me the following error: xli3|csil-suna33|~/cs426|[13]% llvmgcc hello.c as: error opening '/var/tmp//ccapglpE.o': file exists! Use -f command line argument to force output I really got confused. Could you please let me know what's wrong? thanks, Jerry

Dear LLVM: I wrote a small testcase and I can compiler and run it with gcc, But with llvm, I got weird error like the following. I really don't know what's going on. Could you explain? Thanks a lot -Jerry Error Message: xli3|csil-suna38|~/mp2|[25]% llvmgcc testcase3.c testcase3.c: In function `init': testcase3.c:5: warning: cast to pointer from integer of different size

shouldn't the malloc() be something like malloc(sizeof(int)*size)? You're allocating 5 bytes, which is probably not the size of an int in llvm (it's definitely not the size of an int in sparc), and later you're trying to access c[1], which starts at byte 4 but is out of bounds. On Fri, 1 Nov 2002, Xiaodong Li wrote: > Dear LLVM: > > I wrote a small testcase and I can

> > OK, I had an llvmgcc.html (etc.) in my checkout, which I thought was > checked out from the repository (certainly it is none of my doing). > But I see it is not in there, so I just didn't understand how this > works. > >> For the website, can you check again? There was a permissions issues >> on >> something that might have been the problem. I fixed

Since we only have a Solaris version of the front-end (llvmgcc), we need to compile .c -> .ll files on a solaris machines in order to run the tests in Programs/. Doing this manually for all the C source files is really difficult. I added a flag called RUN_GCC_ONLY in test/Programs/Makefile.programs that will cause gmake to run only llvmgcc. Run the following in test/Programs to build .ll

I just had a chance to read some of follow-up comments on Reid's initial document. I agree with Chris's discussion below of what is needed for users to get IPO/lifelong opt'n via LLVM without extensive changes to Makefiles, and about what .o files should contain. This is in perfect agreement with what I just said about how users should view LLVM. --Vikram

On Sat, 10 Jul 2004, Ramu Ramamurthy wrote: > LLVM is an excellent piece of work, Thanks! > LLVM is one of the few project I know that has used the mysterious GCC > front-end. I think it will be a great contribution to the community if > you could writeup and document how to interface to the GCC front-end > data-structures (like LLVM has done). Specifically: > > a)

Perhaps I am missing some link. Need a bit clarification. For the C language, I want to access the LLVM code immediately generated by llvmgcc(cfrontend/bin/gcc) before it undergoes any further transformation or optimization. 1) Are there any libraries that enable me to parse C code and create the Module instance. 2) If answer to 1) is no, then is there some other way to create Module

yue wrote: > hi, > about [LLVMdev] another question > > thanks > > yueqiang One other thing you might want to try is to put your object tree in a directory that is *not* inside of your source tree. Currently, we don't support using separate object trees that are subdirectories of the source tree. In other words: Will work: ========== SRC_ROOT=/home/yue/llvm

Hi, In the quest for better test cases for my MP, I thought of trying the Stepanov Abstraction Penalty benchmark. But apparently llvm-g++ is not ready for such terrible things. Let me know if you want me to send the source code (it is widely available). gaeke|csil-suna37|~/cs/426/MP1/step|[1177]% /usr/dcs/projects/cs426/Software/gcc_install/bin/g++ stepanov.cpp -o stepanov In file included from

Just for the archive or in case of anybody is interested in: The problem is the missing stddef.h which resides more than one time on the OS. I mistakenly thought, that there is an typo in the code, because one of this files is in /us/include/linux so I changed #include <stddef.h> to #include <linux/stddef.h> But this was wrong and it is the wrong file too. The right file to

Adam, I recant my previous message. I built FreeType with GCC not LLVMGCC by accident (I shoulda checked!). The thing is, FreeType's makefiles don't respect environment variables. So, I did this: CC=/path/to/llvmgcc CXX=/path/to/llvmg++ configure make and was able to reproduce the problem. Not sure what's up with that but as Misha pointed out, llvmgcc isn't any

On Wed, 28 Jul 2004, Reid Spencer wrote: > 2. MODE OF OPERATION > ==================== > The driver will simply read its command line arguments, read its > configuration data, and invoke the compilation, linking, and > optimization tools necessary to complete the user's request. Its basic I'm not sure that I agree with this. Compilers need to be extremely predictable and

A new and improved C frontend is now available for experimentation. This is a C frontend that is based on the GCC AST representation as opposed to the old frontend which came from the RTL representation (which lost a lot of type information and caused other problems). I am still working on the new front-end, but it is remarkably stable for being so new. If you are interested in playing with it,

hi all, i surely overlooked something. but i am currently trying to do some benchmarking stuff with llvm. in my thesis i was using llvm-1.5 (i did some extension work, which i would have to frontport now). But unfortunately 1.5 is for some benchmarks unstable (at least on my box) such that llc sigsegvs sometimes, etc. So I wanted to give 1.9 a go and build that on the my intel mac book. I did