On 17 August 2017 at 15:27, Hal Finkel <hfinkel at anl.gov> wrote:> Speaking of benchmarks, we might be able use the library, or some parts of > it, in our test suite for correctness and performance testing. I see some > stand-alone benchmarks that seem useful (e.g., > https://github.com/insideloop/InsideLoop/blob/master/il/benchmark/types/32-vs-64-bit-integers.cpp) > but also some of the numeric things (e.g. > https://github.com/insideloop/InsideLoop/blob/master/il/benchmark/memory/matrix_multiplication.cpp > which makes use of SIMD pragmas and vector intrinsics). Something with the > hash tables and multi-dimensional arrays could be useful too.That is certainly an interesting proposition. I think we can have it on both testing and benchmarking sets, but with different workloads, maybe? --renato
By the way, do you have any solution for an automatic performance regression checker ? I don’t have one yet. François Fayard> That is certainly an interesting proposition. I think we can have it > on both testing and benchmarking sets, but with different workloads, > maybe? > > --renato-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170817/f9466174/attachment.html>
> We are working on one, but it's far too green to announce it to a > wider audience: > https://git.linaro.org/people/masaki.arai/hcqc.git/Will check it out. Also, I have just seen that some people want to develop a kind of “performance sanitizer” is there anything useable yet? François Fayard -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170817/436baa79/attachment.html>
> On Aug 17, 2017, at 8:16 AM, Francois Fayard via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > By the way, do you have any solution for an automatic performance regression checker ?LLVMs answer to this is setting up CI jobs that submit result to an LNT server (http://llvm.org/docs/lnt/ <http://llvm.org/docs/lnt/>). It should also work for non-llvm projects. Though today you better think of it as "performance tracking" rather than "automatic regression checking". - Matthias -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170817/4df63b2a/attachment.html>