similar to: [LLVMdev] Compiling to NVPTX

Hi LLVM developers, I have successfully installed LLVM 2.9 on an ubuntu linux box, with a LLVM-based frontend support of gcc and g++. After installing successfully vmkit I have tried running my java applications, but it failed. I have installed the latest stable release of vmkit (0.29) using the introduction text of the LLVM website. The application does consist of two programs. The first

Puppet 3.3.1 // CentOS release 6.4 (Final) rpm -qa | grep puppet puppetlabs-release-6-7.noarch puppet-3.1.1-1.el6.noarch puppetdb-1.3.0-1.el6.noarch puppet-server-3.1.1-1.el6.noarch puppetdb-terminus-1.3.0-1.el6.noarch Installed from yum packages: Running Transaction Installing : puppetdb-1.3.0-1.el6.noarch

On 09.04.2010, at 11:28, nicolas geoffray wrote: > VMKit runs multiple threads (for example for GC, finalization, etc), > so I need the back trace of other threads as well :) Here it comes: Thread 4 (process 14442 thread 0x2003): #0 0x950be44e in __semwait_signal () #1 0x950e93e6 in _pthread_cond_wait () #2 0x950e8dcd in pthread_cond_wait$UNIX2003 () #3 0x000be2d2 in mvm::Cond::wait

Hi Timothy, I'm not sure what you mean by this working for other intrinsics, but in this case, I think you want the intrinsic name llvm.nvvm.read.ptx.sreg.tid.x. For me, this looks like: %x = call i32 @llvm.nvvm.read.ptx.sreg.tid.x() Pete On Fri, Mar 1, 2013 at 11:51 AM, Timothy Baldridge <tbaldridge at gmail.com> wrote: > I'm building this with llvm-c, and accessing these

The identifier INT_PTX_SREG_TID_X is the name of an instruction as the back-end sees it, and has very little to do with the name you should use in your IR. Your best bet is to look at the include/llvm/IR/IntrinsicsNVVM.td file and see the definitions for each intrinsic. Then, the name mapping is just: int_foo_bar -> llvm.foo.bar() int_ prefix becomes llvm., and all underscores turn into

I'm writing a JIT for my toy language, and I'm confused by one thing. How do I, with IR Builder, Insert a pointer constant? For the first version of my jit, I really only need LLVM to pass around opaque types and then pass those variables to C functions that return other opaque types. So for instance, I have a Int object that wraps the GNU GMP routines. A want to be able to tell IRBuilder

Hi everyone, Speed is the key here. I need to find the difference between a vector and its one-period lag (i.e. the difference between each value and the subsequent one in the vector). Let's say the vector contains 10 million random integers between 0 and 1,000. The solution vector will have 9,999,999 values, since their is no lag for the 1st observation. In R we have: #Set up input vector

I'd like to do some simple C# -> LLVM C API bindings for windows. Is there a place I can get pre-built LLVM .dlls for windows? Thanks Timothy -- “One of the main causes of the fall of the Roman Empire was that–lacking zero–they had no way to indicate successful termination of their C programs.” (Robert Firth) -------------- next part -------------- An HTML attachment was scrubbed...

Hi Konrad, I have added some diagnostic message when failing that early in the bootstrap in the VM. Please svn up and let me know what it prints. About the SIGSEGV you get with clojure.jar, I ran the code and got the same error. After investigation, it turns out that the methods you have in that .jar are quite big (50,000 bytecodes) and the static initializer of the core__init class just crashes

I'm working on implementing a simple functional language in LLVM. It's statically typed, but I'd like to a have some simple multimethods involved. Does LLVM have some intrinsics for handling something like C++ vtables? Or do I need to implement my own way of handling single-dispatch functions? Thanks, Timothy -- “One of the main causes of the fall of the Roman Empire was

I can't seem to get LLVMCreateTargetMachine to work on my box (MBP i7 OS X 10.8). What should the values of triple, cpu and features be? I'm just looking for a way to get this up and going for now, I don't need specific optimizations. No matter what I send the function, I always get nil back. I'm giving it LLVMGetFirstTarget as the target value. Thanks for the help, Timothy

I'm building this with llvm-c, and accessing these intrinsics via calling the intrinsic as if it were a function. class F_SREG<string OpStr, NVPTXRegClass regclassOut, Intrinsic IntOp> : NVPTXInst<(outs regclassOut:$dst), (ins), OpStr, [(set regclassOut:$dst, (IntOp))]>; def INT_PTX_SREG_TID_X : F_SREG<"mov.u32 \t$dst, %tid.x;",

Hi everybody, I have noticed that the new NVPTX backend requires new metadata to identify the kernels in the module: define void @metadata_kernel(float* %a) { ret void } !nvvm.annotations = !{!1} !1 = metadata !{void (float*)* @metadata_kernel, metadata !"kernel", i32 1} Is clang going to support the emission of this metadata soon ? Or do I have to write it on my own ? :) Thanks,

Hi, ' Not specific to 'R'. I search for patterns and found http://patternsinfp.wordpress.com/ which is too heavy for me. There is a 'Pragmatic Programmer' book on such patterns for Scala and Clojure. Is there anything for R ? I wanted to code this. Is there a functional pattern in R for multiple 'if' loops like this ? if(

The file also has a bunch of email headers stuck in the middle of it: ..... (QUALITY-OF-LIFE SCALE:1-5 4) (ACADEMIC-EMPHASIS HEALTH-SCIENCE) ) ------- ------- >From LEBOWITZ at cs.columbia.edu Mon Feb 22 20:53:02 1988 Received: from zodiac by meridian (5.52/4.7) Received: from Jessica.Stanford.EDU by ads.com (5.58/1.9) id AA04539; Mon, 22 Feb 88 20:59:59 PST Received: from

Hi all, For loads and stores i want to extract information about the number of indices accessed. For instance: struct S {int X, int *Y}; __global__ void kernel(int *A, int **B, struct S) {   int x = A[..][..]; // -> L: A[..][..]   int y = *B[2];   // -> L: B[0][2]   int z = S.y[..];  // -> L: S.1[..]   // etc.. } I am performing some preprocessing on IR to: 1. Move constant

*Hi, I am considering a language extension to Clang for optimizing GPU programs. This extension will allow the compiler to use different optimization strategies for divergent and non-divergent branches (to be explained below). We have observed significant performance gain by leveraging this proposed extension, so I want to discuss it here to see how the community likes/dislikes the idea. I will

This new metadata format is currently optional. The old ptx_kernel calling convention should still work. The only thing you should have to change when converting from PTX -> NVPTX is the address space map. The calling conventions and intrinsics should be compatible with both. > -----Original Message----- > From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu]

Hi Justin, In the PTX backend, "ret" instruction at the end of a ptx_kernel function is translated to "exit" instruction. A test case named exit.ll demos this. But in the NVPTX backend, it seems that you didn't do such a translation. Why do you choose this? Is this due to the changes of the NVIDA PTX itself? Thanks, Yabin -------------- next part -------------- An HTML

Hi, The range metadata can only be attached to LoadInst for now. I am considering extending its usage to IntrinsicInst so that the frontend can annotate the range of the return value of an intrinsic call. e.g., %a = call i32 @llvm.xxx(), !range !0 !0 = metadata !{ i32 0, i23 1024 } The motivation behind this extension is some optimizations we are working on for CUDA programs. Some special