Rasmus Munk Larsen
344c2694a6
Make the non-blocking threadpool more flexible and less wasteful of CPU cycles for high-latency use-cases.
...
* Adds a hint to ThreadPool allowing us to turn off spin waiting. Currently each reader and record yielder op in a graph creates a threadpool with a thread that spins for 1000 iterations through the work stealing loop before yielding. This is wasteful for such ops that process I/O.
* This also changes the number of iterations through the steal loop to be inversely proportional to the number of threads. Since the time of each iteration is proportional to the number of threads, this yields roughly a constant spin time.
* Implement a separate worker loop for the num_threads == 1 case since there is no point in going through the expensive steal loop. Moreover, since Steal() calls PopBack() on the victim queues it might reverse the order in which ops are executed, compared to the order in which they are scheduled, which is usually counter-productive for the types of I/O workloads the single thread pools tend to be used for.
* Store num_threads in a member variable for simplicity and to avoid a data race between the thread creation loop and worker threads calling threads_.size().
2017-03-09 15:41:03 -08:00
Mehdi Goli
0d153ded29
Adding TensorChippingOP for sycl backend; fixing the index value in the verification operation for cxx11_tensorChipping.cpp test
2017-02-13 17:25:12 +00:00
Mehdi Goli
0ee97b60c2
Adding mean to TensorReductionSycl.h
2017-02-07 15:43:17 +00:00
Mehdi Goli
42bd5c4e7b
Fixing TensorReductionSycl for min and max.
2017-02-06 18:05:23 +00:00
Mehdi Goli
ff53050034
Converting ptrdiff_t type to int64_t type in cxx11_tensor_contract_sycl.cpp in order to be the same as other tests.
2017-02-01 15:36:03 +00:00
Mehdi Goli
bab29936a1
Reducing warnings in Sycl backend.
2017-02-01 15:29:53 +00:00
Benoit Steiner
fbc39fd02c
Merge latest changes from upstream
2017-01-30 15:25:57 -08:00
Rasmus Munk Larsen
5e144bbaa4
Make NaN propagatation consistent between the pmax/pmin and std::max/std::min. This makes the NaN propagation consistent between the scalar and vectorized code paths of Eigen's scalar_max_op and scalar_min_op.
...
See #1373 for details.
2017-01-24 13:32:50 -08:00
Mehdi Goli
6bdd15f572
Adding non-deferrenciable pointer track for ComputeCpp backend; Adding TensorConvolutionOp for ComputeCpp; fixing typos. modifying TensorDeviceSycl to use the LegacyPointer class.
2017-01-19 11:30:59 +00:00
Mehdi Goli
e46e722381
Adding Tensor ReverseOp; TensorStriding; TensorConversionOp; Modifying Tensor Contractsycl to be located in any place in the expression tree.
2017-01-16 13:58:49 +00:00
Benoit Steiner
0657228569
Simplified the way we link libxsmm
2016-12-21 14:40:08 -08:00
Benoit Steiner
c19fe5e9ed
Added support for libxsmm in the eigen makefiles
2016-12-21 10:43:40 -08:00
Benoit Steiner
0f577d4744
Merged eigen/eigen into default
2016-12-20 17:02:06 -08:00
Gael Guennebaud
e8d6862f14
Properly adjust precision when saving to Market format.
2016-12-20 22:10:33 +01:00
Gael Guennebaud
e2f4ee1c2b
Speed up parsing of sparse Market file.
2016-12-20 21:56:21 +01:00
Benoit Steiner
548ed30a1c
Added an OpenCL regression test
2016-12-19 18:56:26 -08:00
Benoit Steiner
27ceb43bf6
Fixed race condition in the tensor_shuffling_sycl test
2016-12-19 15:34:42 -08:00
Mehdi Goli
35bae513a0
Converting all parallel for lambda to functor in order to prevent kernel duplication name error; adding tensorConcatinationOp backend for sycl.
2016-12-16 19:46:45 +00:00
Benoit Steiner
9ff5d0f821
Merged eigen/eigen into default
2016-12-14 17:32:16 -08:00
Mehdi Goli
730eb9fe1c
Adding asynchronous execution as it improves the performance.
2016-12-14 17:38:53 +00:00
Mehdi Goli
2d4a091beb
Adding tensor contraction operation backend for Sycl; adding test for contractionOp sycl backend; adding temporary solution to prevent memory leak in buffer; cleaning up cxx11_tensor_buildins_sycl.h
2016-12-14 15:30:37 +00:00
Benoit Steiner
4deafd35b7
Introduce a portable EIGEN_SLEEP macro.
2016-12-09 14:52:15 -08:00
Benoit Steiner
2f5b7a199b
Reworked the threadpool cancellation mechanism to not depend on pthread_cancel since it turns out that pthread_cancel doesn't work properly on numerous platforms.
2016-12-09 13:05:14 -08:00
Benoit Steiner
3d59a47720
Added a message to ease the detection of platforms on which thread cancellation isn't supported.
2016-12-08 14:51:46 -08:00
Benoit Steiner
7bfff85355
Added support for thread cancellation on Linux
2016-12-08 08:12:49 -08:00
Srinivas Vasudevan
218764ee1f
Added support for expm1 in Eigen.
2016-12-02 14:13:01 -08:00
Mehdi Goli
79aa2b784e
Adding sycl backend for TensorPadding.h; disbaling __unit128 for sycl in TensorIntDiv.h; disabling cashsize for sycl in tensorDeviceDefault.h; adding sycl backend for StrideSliceOP ; removing sycl compiler warning for creating an array of size 0 in CXX11Meta.h; cleaning up the sycl backend code.
2016-12-01 13:02:27 +00:00
Benoit Steiner
fd1dc3363e
Merged eigen/eigen into default
2016-11-30 20:16:17 -08:00
Mehdi Goli
577ce78085
Adding TensorShuffling backend for sycl; adding TensorReshaping backend for sycl; cleaning up the sycl backend.
2016-11-29 15:30:42 +00:00
Benoit Steiner
3011dc94ef
Call internal::array_prod to compute the total size of the tensor.
2016-11-28 09:00:31 -08:00
Benoit Steiner
9f8fbd9434
Merged eigen/eigen into default
2016-11-26 11:28:25 -08:00
Mehdi Goli
7318daf887
Fixing LLVM error on TensorMorphingSycl.h on GPU; fixing int64_t crash for tensor_broadcast_sycl on GPU; adding get_sycl_supported_devices() on syclDevice.h.
2016-11-25 16:19:07 +00:00
Mehdi Goli
b8cc5635d5
Removing unsupported device from test case; cleaning the tensor device sycl.
2016-11-23 16:30:41 +00:00
Gael Guennebaud
7f6333c32b
Merged in tal500/eigen-eulerangles (pull request PR-237)
...
Euler angles
2016-11-23 15:17:38 +00:00
Gael Guennebaud
f12b368417
Extend polynomial solver unit tests to complexes
2016-11-23 16:05:45 +01:00
Luke Iwanski
af67335e0e
Added test for cwiseMin, cwiseMax and operator%.
2016-11-19 13:37:27 +00:00
Benoit Steiner
a357fe1fb9
Code cleanup
2016-11-18 16:58:09 -08:00
Benoit Steiner
1c6eafb46b
Updated cxx11_tensor_device_sycl to run only on the OpenCL devices available on the host
2016-11-18 16:43:27 -08:00
Benoit Steiner
ca754caa23
Only runs the cxx11_tensor_reduction_sycl on devices that are available.
2016-11-18 16:31:14 -08:00
Benoit Steiner
dc601d79d1
Added the ability to run test exclusively OpenCL devices that are listed by sycl::device::get_devices().
2016-11-18 16:26:50 -08:00
Benoit Steiner
b5e3285e16
Test broadcasting on OpenCL devices with 64 bit indexing
2016-11-18 13:44:20 -08:00
Benoit Steiner
7335c49204
Fixed the cxx11_tensor_device_sycl test
2016-11-18 12:37:13 -08:00
Mehdi Goli
15e226d7d3
adding Benoit changes on the TensorDeviceSycl.h
2016-11-18 16:34:54 +00:00
Mehdi Goli
622805a0c5
Modifying TensorDeviceSycl.h to always create buffer of type uint8_t and convert them to the actual type at the execution on the device; adding the queue interface class to separate the lifespan of sycl queue and buffers,created for that queue, from Eigen::SyclDevice; modifying sycl tests to support the evaluation of the results for both row major and column major data layout on all different devices that are supported by Sycl{CPU; GPU; and Host}.
2016-11-18 16:20:42 +00:00
Luke Iwanski
5159675c33
Added isnan, isfinite and isinf for SYCL device. Plus test for that.
2016-11-18 16:01:48 +00:00
Luke Iwanski
927bd62d2a
Now testing out (+=, =) in.FUNC() and out (+=, =) out.FUNC()
2016-11-18 11:16:42 +00:00
Benoit Steiner
553f50b246
Added a way to detect errors generated by the opencl device from the host
2016-11-17 21:51:48 -08:00
Benoit Steiner
4349fc640e
Created a test to check that the sycl runtime can successfully report errors (like ivision by 0).
...
Small cleanup
2016-11-17 20:27:54 -08:00
Benoit Steiner
004344cf54
Avoid calling log(0) or 1/0
2016-11-17 11:56:44 -08:00
Luke Iwanski
7878756dea
Fixed existing test.
2016-11-17 17:46:55 +00:00
