search for: divergence

TL;DR This RFC is a joint effort by Intel and Saarland University to bring the divergence analysis of the Region Vectorizer [1,2,3,4,5] (dubbed the vectorization analysis of RV) to LLVM. The implementation is available on github for feedback [0]. The existing divergence analysis infrastructure in LLVM has conceptual limitations (structured control, SCEV based). The new analysis reso...

...alf Karrenberg used in his thesis. > > > > *From:* llvm-dev [mailto:llvm-dev-bounces at lists.llvm.org] *On Behalf Of *Alexandre > Isoard via llvm-dev > *Sent:* Friday, July 14, 2017 20:53 > *To:* llvm-dev <llvm-dev at lists.llvm.org> > *Subject:* [llvm-dev] [SPIR/PTX] Divergence analysis for BasicBlocks > > > > Hello, > > > > It seems to me that our current DivergenceAnalysis does not save which > BasicBlocks may suffer from divergent control. Am I correct? > > > > I want to modify our DivergenceAnalysis to add a "bool > isCon...

Hello, It seems to me that our current DivergenceAnalysis does not save which BasicBlocks may suffer from divergent control. Am I correct? I want to modify our DivergenceAnalysis to add a "bool isControlDivergent(BasicBlock*) const" method and save in the divergence propagator the basicblock that are divergent. I am not sure that is ent...

On 25.10.2016 16:28, Nicolai Hähnle wrote: > But I fear that this path leads to eternal fuzziness. Let me try a > completely different approach to define what we need by augmenting the > semantics of IR with "divergence tokens". In addition to its usual > value, every IR value carries a "divergence set" of divergence tokens. > > The basic rule is: the divergence set of a value is (at least) the union > of the divergence sets of its operands. > > Every function input carries a uniqu...

Hi list! I have prepared a custom function (below) in order to calculate separability indices (Divergence, Bhattacharyya, Jeffries-Matusita, Transformed divergene) between two samples of (spectral land cover) classes. I need help to cross-compare results to verify that it works as expected (since I don't know of any other foss-tool that will give me quickly some results). Does anybody use anot...

...taking the intrinsics into account? > > Cheers, > Michael > The intrinsics certainly are correctly treated as divergent. Nothing would work otherwise. If I run the annotate pass or analysis on the examples it does the right thing and sees the load as divergent. $ opt -S -analyze -divergence -o - as1.ll Printing analysis 'Divergence Analysis' for function '_ZN5pacxx2v213genericKernelIZL12test_barrieriPPcE3$_0EEvT_': DIVERGENT: %6 = tail call i32 @llvm.amdgcn.workitem.id.x() #0, !range !11 DIVERGENT: %add.i.i.i.i.i = add nsw i32 %mul.i.i.i.i.i, %6 DIVERGENT: %idxprom....

> On Oct 24, 2016, at 4:15 PM, Nicolai Hähnle <nhaehnle at gmail.com> wrote: > > On 25.10.2016 01:11, Nicolai Hähnle wrote: >> On 24.10.2016 21:54, Mehdi Amini wrote: >>>> On Oct 24, 2016, at 12:38 PM, Nicolai Hähnle via llvm-dev >>>> <llvm-dev at lists.llvm.org> wrote: >>>> Some brain-storming on an issue with SPMD/SIMT backend

...ignificant 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 focus on the CUDA language and the PTX backend for now, but I believe this addition will benefit OpenCL and its backends too. Background on branch divergenceCUDA programs have a very different execution model for code with branches. A CUDA program is executed by an array of threads broken into groups called warps. A warp typically contains 32 threads, and all the threads in a warp execute instructions in lock-step, i.e., executing the same instruction a...

...ce gain by > leveraging this proposed extension, so I want to discuss it here to see how > the community likes/dislikes the idea. I will focus on the CUDA language > and the PTX backend for now, but I believe this addition will benefit > OpenCL and its backends too. Background on branch divergenceCUDA programs > have a very different execution model for code with branches. A CUDA > program is executed by an array of threads broken into groups called warps. > A warp typically contains 32 threads, and all the threads in a warp execute > instructions in lock-step, i.e., executing th...

...gt; leveraging this proposed extension, so I want to discuss it here to see how >> the community likes/dislikes the idea. I will focus on the CUDA language >> and the PTX backend for now, but I believe this addition will benefit >> OpenCL and its backends too. Background on branch divergenceCUDA programs >> have a very different execution model for code with branches. A CUDA >> program is executed by an array of threads broken into groups called warps. >> A warp typically contains 32 threads, and all the threads in a warp execute >> instructions in lock-step, i....

...llvm-dev <llvm-dev at lists.llvm.org> wrote: > On 25.10.2016 16:28, Nicolai Hähnle wrote: >> >> But I fear that this path leads to eternal fuzziness. Let me try a >> completely different approach to define what we need by augmenting the >> semantics of IR with "divergence tokens". In addition to its usual >> value, every IR value carries a "divergence set" of divergence tokens. >> >> The basic rule is: the divergence set of a value is (at least) the union >> of the divergence sets of its operands. >> >> Every functi...

Hi all, I have some wind data (U and V components) and I would like to compute Vorticity and Divergence of these fields. Is there any R function that can easily do that? Thanks in advance for any help Igor Oliveira CSAG, Dept. Environmental & Geographical Science, University of Cape Town, Private Bag X3, Rondebosch 7701. Tel.: +27 (0)21 650 5774 South Africa Fax: +27 (0)21 650 577...

We use Nutch and Lucene for our heavy duty text analysis jobs but I?m trying to use ferrret to do some experiments. I understood that Ferret used the same index format as lucene but I cannot look into a lucene index with ferret and cannot read a ferret index with luke (the lucene index browser). Am I doing somehting wrong or have the formats diverged? -------------- next part -------------- An

On the surface this seems pretty simple, but I flummoxed. I have two sets of numbers they bounce around zero, positive one and negative one. They have a relationship between them, where one diverges away from the other. I want create a second set of numbers that tracks that divergence. #Lets make some data like mine, kinda Firstset <- runif(100, min = -1 , max =1) Secondset <- runif(100, min = -1 , max =1) #So something like: Divergence <- abs (Firstset - Secondset) #but this doesn't work because when Firstset is at .5 and Secondset is at -.25 it returns .25 in...

I am trying to solve an integral in R. However, I am getting an error when I am trying to solve for that integral. The equation that I am trying to solve is as follows: $$ C_m = \frac{{abs{x}}e^{2x}}{\pi^{1/2}}\int_0^t t^{-3/2}e^{-x^2/t-t}dt $$ [image: enter image description here] The code that I am using is as follows: a <- seq(from=-10, by=0.5,length=100) ## Create a function to compute

> On Dec 5, 2017, at 13:53, Matt Arsenault <arsenm2 at gmail.com> wrote: > > > >> On Dec 5, 2017, at 02:51, Haidl, Michael via llvm-dev <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>> wrote: >> >> Hi dev list, >> >> I am currently exploring the integration of AMDGPU/ROCm into the PACXX project and observing some

I have R version 2.9.1 on my computer and the anlaysis is not working because I need to update to R version 2.12.0 the latest release. The person incharge of IT tried to download R version 2.12.0 but .exe file referenced in install isn't there - What might we be doing wrong? We have downloaded as tar.gz, uncompress then look for .exe file but not present. Many thanks, Morris

> misc.c provides the ability to override some functions, including the > allocation and printing. fftwrap.c uses speex_alloc, then calls > kiss_fftr_alloc, which calls kiss_fft_alloc, which calls KISS_FFT_MALLOC, > which is defined as malloc in kiss_fft.h. It would make it more consistent > to define KISS_FFT_MALLOC as speex_alloc. That is the only change that I > would

I tried to reduce the test case in https://bugs.llvm.org/show_bug.cgi?id=42706. Here it is crashing opt: $ ~/llvm-debug/bin/opt -use-gpu-divergence-analysis -divergence stripped.ll WARNING: You're attempting to print out a bitcode file. This is inadvisable as it may cause display problems. If you REALLY want to taste LLVM bitcode first-hand, you can force output with the `-f' option. opt: /home/jayfoad2/git/llvm-project/llvm/lib/Analy...

I'm using 7.0. I am attaching the reduced example. nvcc sync.cu -arch=sm_35 -ptx gives // .globl _Z3foov .visible .entry _Z3foov( ) { .reg .pred %p<2>; .reg .s32 %r<3>; mov.u32 %r1, %tid.x; and.b32 %r2, %r1, 1; setp.eq.b32 %p1, %r2, 1; @!%p1 bra BB7_2; bra.uni