Project Ortega BCM5719 LLVM https://git.raptorcs.com/git/bcm5719-llvm/atom?h=meklort-10.0.1 2020-01-16T12:55:05+00:00 2020-01-16T12:55:05+00:00 Benjamin Kramer benny.kra@googlemail.com 2020-01-15T16:51:51+00:00 urn:sha1:ac446302ca4145cdc89f377c0c364c29ee303be5 This reverts commit 4624a1e8ac8a3f69cc887403b976f538f587744a. Causing problems downstream. (cherry picked from commit 0133cc60e4e230ee2c176c23eff5aa2f4ee17a75) 2020-01-14T13:06:37+00:00 Benjamin Kramer benny.kra@googlemail.com 2020-01-14T13:06:12+00:00 urn:sha1:df186507e1d07c3ddba091a076ba7a33dbdc5867 2020-01-14T11:05:47+00:00 Tres Popp tpopp@google.com 2020-01-14T10:09:59+00:00 urn:sha1:4624a1e8ac8a3f69cc887403b976f538f587744a Summary: This is based on the use of code constantly checking for an attribute on a model and instead represents the distinct operaion with a different op. Instead, this op can be used to provide better filtering. Reviewers: herhut, mravishankar, antiagainst, rriddle Reviewed By: herhut, antiagainst, rriddle Subscribers: liufengdb, aartbik, jholewinski, mgorny, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, csigg, arpith-jacob, mgester, lucyrfox, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D72336 2020-01-11T16:54:39+00:00 River Riddle riverriddle@google.com 2020-01-11T16:54:04+00:00 urn:sha1:2bdf33cc4c733342fc83081bc7410ac5e9a24f55 Summary: These were temporary methods used to simplify the transition. Reviewed By: antiagainst Differential Revision: https://reviews.llvm.org/D72548 2020-01-09T09:06:00+00:00 Alex Zinenko zinenko@google.com 2020-01-07T19:00:54+00:00 urn:sha1:08778d8c4fd8a6519c7f27bfa6b09c47262cb844 Introduce a set of function that promote a memref argument of a `gpu.func` to workgroup memory using memory attribution. The promotion boils down to additional loops performing the copy from the original argument to the attributed memory in the beginning of the function, and back at the end of the function using all available threads. The loop bounds are specified so as to adapt to any size of the workgroup. These utilities are intended to compose with other existing utilities (loop coalescing and tiling) in cases where the distribution of work across threads is uneven, e.g. copying a 2D memref with only the threads along the "x" dimension. Similarly, specialization of the kernel to specific launch sizes should be implemented as a separate pass combining constant propagation and canonicalization. Introduce a simple attribute-driven pass to test the promotion transformation since we don't have a heuristic at the moment. Differential revision: https://reviews.llvm.org/D71904 2019-12-24T00:36:53+00:00 River Riddle riverriddle@google.com 2019-12-23T22:45:01+00:00 urn:sha1:e62a69561fb9d7b1013d2853da68d79a7907fead ValuePtr was a temporary typedef during the transition to a value-typed Value. PiperOrigin-RevId: 286945714 2019-12-24T00:26:15+00:00 River Riddle riverriddle@google.com 2019-12-23T21:05:38+00:00 urn:sha1:5d5bd2e1da29d976cb125dbb3cd097a5e42b2be4 This means that in-place, or root, updates need to use explicit calls to `startRootUpdate`, `finalizeRootUpdate`, and `cancelRootUpdate`. The major benefit of this change is that it enables in-place updates in DialectConversion, which simplifies the FuncOp pattern for example. The major downside to this is that the cases that *may* modify an operation in-place will need an explicit cancel on the failure branches(assuming that they started an update before attempting the transformation). PiperOrigin-RevId: 286933674 2019-12-23T23:33:37+00:00 Mehdi Amini aminim@google.com 2019-12-23T17:35:36+00:00 urn:sha1:56222a0694e4caf35e892d70591417c39fef1185 PiperOrigin-RevId: 286906740 2019-12-23T06:00:23+00:00 River Riddle riverriddle@google.com 2019-12-23T05:59:55+00:00 urn:sha1:35807bc4c5c9d8abc31ba0b2f955a82abf276e12 This is an initial step to refactoring the representation of OpResult as proposed in: https://groups.google.com/a/tensorflow.org/g/mlir/c/XXzzKhqqF_0/m/v6bKb08WCgAJ This change will make it much simpler to incrementally transition all of the existing code to use value-typed semantics. PiperOrigin-RevId: 286844725 2019-12-20T10:52:52+00:00 Christian Sigg csigg@google.com 2019-12-20T10:52:21+00:00 urn:sha1:42d46b4efa9d7b596c74c101a777f4464a8f1fd0 This will allow us to lower most of gpu.all_reduce (when all_reduce doesn't exist in the target dialect) within the GPU dialect, and only do target-specific lowering for the shuffle op. PiperOrigin-RevId: 286548256