Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Add support for triSYCL | 2017-05-05 | |
| | | | | | | Eigen is now able to use triSYCL with EIGEN_SYCL_TRISYCL and TRISYCL_INCLUDE_DIR options Fix contraction kernel with correct nd_item dimension | ||
* | Preserve file naming conventions | 2017-04-04 | |
| | |||
* | Fixing TensorArgMaxSycl.h; Removing warning related to the hardcoded type of ↵ | 2017-03-28 | |
| | | | | dims to be int in Argmax. | ||
* | Merged eigen/eigen into default | 2017-03-15 | |
|\ | |||
* | | Temporary: Disables cxx11_tensor_argmax_sycl test since it is causing zombie ↵ | 2017-03-15 | |
| | | | | | | | | thread | ||
| * | Make the non-blocking threadpool more flexible and less wasteful of CPU ↵ | 2017-03-09 | |
| | | | | | | | | | | | | | | | | | | | | | | | | 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(). | ||
* | | Adding TensorIndexTuple and TensorTupleReduceOP backend (ArgMax/Min) for ↵ | 2017-03-07 | |
| | | | | | | | | sycl; fixing the address space issue for const TensorMap; converting all discard_write to write due to data missmatch. | ||
* | | Adding sycl backend for TensorCustomOp; fixing the partial lhs modification ↵ | 2017-02-28 | |
| | | | | | | | | issue on sycl when the rhs is TensorContraction, reduction or convolution; Fixing the partial modification for memset when sycl backend is used. | ||
* | | Adding TensorVolumePatchOP.h for sycl | 2017-02-24 | |
| | | |||
* | | Adding Sycl Backend for TensorGenerator.h. | 2017-02-22 | |
| | | |||
* | | Reducing the number of warnings. | 2017-02-21 | |
| | | |||
* | | Adding Sycl backend for TensorImagePatchOP.h; adding Sycl backend for ↵ | 2017-02-20 | |
| | | | | | | | | TensorInflation.h. | ||
* | | Adding TensorLayoutSwapOp for sycl. | 2017-02-15 | |
| | | |||
* | | Adding TensorPatch.h for sycl backend. | 2017-02-15 | |
|/ | |||
* | Adding TensorChippingOP for sycl backend; fixing the index value in the ↵ | 2017-02-13 | |
| | | | | verification operation for cxx11_tensorChipping.cpp test | ||
* | Adding mean to TensorReductionSycl.h | 2017-02-07 | |
| | |||
* | Fixing TensorReductionSycl for min and max. | 2017-02-06 | |
| | |||
* | Converting ptrdiff_t type to int64_t type in cxx11_tensor_contract_sycl.cpp ↵ | 2017-02-01 | |
| | | | | in order to be the same as other tests. | ||
* | Reducing warnings in Sycl backend. | 2017-02-01 | |
| | |||
* | Merge latest changes from upstream | 2017-01-30 | |
|\ | |||
| * | Make NaN propagatation consistent between the pmax/pmin and ↵ | 2017-01-24 | |
| | | | | | | | | | | | | 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. | ||
* | | Adding non-deferrenciable pointer track for ComputeCpp backend; Adding ↵ | 2017-01-19 | |
| | | | | | | | | TensorConvolutionOp for ComputeCpp; fixing typos. modifying TensorDeviceSycl to use the LegacyPointer class. | ||
* | | Adding Tensor ReverseOp; TensorStriding; TensorConversionOp; Modifying ↵ | 2017-01-16 | |
| | | | | | | | | Tensor Contractsycl to be located in any place in the expression tree. | ||
| * | Simplified the way we link libxsmm | 2016-12-21 | |
| | | |||
| * | Added support for libxsmm in the eigen makefiles | 2016-12-21 | |
|/ | |||
* | Merged eigen/eigen into default | 2016-12-20 | |
|\ | |||
| * | Properly adjust precision when saving to Market format. | 2016-12-20 | |
| | | |||
| * | Speed up parsing of sparse Market file. | 2016-12-20 | |
| | | |||
* | | Added an OpenCL regression test | 2016-12-19 | |
| | | |||
| * | Fixed race condition in the tensor_shuffling_sycl test | 2016-12-19 | |
|/ | |||
* | Converting all parallel for lambda to functor in order to prevent kernel ↵ | 2016-12-16 | |
| | | | | duplication name error; adding tensorConcatinationOp backend for sycl. | ||
* | Merged eigen/eigen into default | 2016-12-14 | |
|\ | |||
* | | Adding asynchronous execution as it improves the performance. | 2016-12-14 | |
| | | |||
* | | Adding tensor contraction operation backend for Sycl; adding test for ↵ | 2016-12-14 | |
| | | | | | | | | contractionOp sycl backend; adding temporary solution to prevent memory leak in buffer; cleaning up cxx11_tensor_buildins_sycl.h | ||
| * | Introduce a portable EIGEN_SLEEP macro. | 2016-12-09 | |
| | | |||
| * | Reworked the threadpool cancellation mechanism to not depend on ↵ | 2016-12-09 | |
|/ | | | | pthread_cancel since it turns out that pthread_cancel doesn't work properly on numerous platforms. | ||
* | Added a message to ease the detection of platforms on which thread ↵ | 2016-12-08 | |
| | | | | cancellation isn't supported. | ||
* | Added support for thread cancellation on Linux | 2016-12-08 | |
| | |||
* | Added support for expm1 in Eigen. | 2016-12-02 | |
| | |||
* | Adding sycl backend for TensorPadding.h; disbaling __unit128 for sycl in ↵ | 2016-12-01 | |
| | | | | 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. | ||
* | Merged eigen/eigen into default | 2016-11-30 | |
|\ | |||
* | | Adding TensorShuffling backend for sycl; adding TensorReshaping backend for ↵ | 2016-11-29 | |
| | | | | | | | | sycl; cleaning up the sycl backend. | ||
| * | Call internal::array_prod to compute the total size of the tensor. | 2016-11-28 | |
| | | |||
* | | Merged eigen/eigen into default | 2016-11-26 | |
|\| | |||
* | | Fixing LLVM error on TensorMorphingSycl.h on GPU; fixing int64_t crash for ↵ | 2016-11-25 | |
| | | | | | | | | tensor_broadcast_sycl on GPU; adding get_sycl_supported_devices() on syclDevice.h. | ||
* | | Removing unsupported device from test case; cleaning the tensor device sycl. | 2016-11-23 | |
| | | |||
| * | Merged in tal500/eigen-eulerangles (pull request PR-237) | 2016-11-23 | |
| |\ | | | | | | | | | | Euler angles | ||
| * | | Extend polynomial solver unit tests to complexes | 2016-11-23 | |
|/ / | |||
* | | Added test for cwiseMin, cwiseMax and operator%. | 2016-11-19 | |
| | | |||
* | | Code cleanup | 2016-11-18 | |
| | |