User Details
- User Since
- Jan 9 2020, 1:31 AM (66 w, 6 d)
Mon, Mar 29
Fri, Mar 26
Addressed comments.
Wed, Mar 24
Addressed comments.
Tue, Mar 23
Mar 17 2021
Mar 15 2021
Mar 11 2021
Adressed comments.
Mar 9 2021
Mar 5 2021
Feb 18 2021
Feb 12 2021
Removed corresponding ops in std dialect and adapted test cases.
Addressed comments.
Feb 11 2021
Feb 10 2021
Feb 8 2021
Jan 27 2021
Jan 20 2021
Jan 19 2021
Addressed comments.
Jan 18 2021
Jan 13 2021
Jan 11 2021
Dec 3 2020
Dec 1 2020
Addressed comments.
Nov 25 2020
Addressed comments.
Changed default max rank to 1 instead of 2 and changed condition to less or equal.
This way is more intuitive.
Nov 24 2020
Addressed comments.
Introduce transformation if the alloc operands are RankOps.
Nov 23 2020
Nov 18 2020
Nov 17 2020
Nov 13 2020
Nov 12 2020
Nov 5 2020
Addressed comments
Nov 3 2020
Oct 23 2020
Oct 22 2020
Addressed comments.
Oct 21 2020
Addressed comments.
Oct 19 2020
Addressed comments.
Aug 3 2020
Jul 31 2020
Changed check to exclude automatic allocation scope resources.
Jul 3 2020
Changed from a specialized MemRefCastOp version to a more general ViewLikeOpInterface version.
Adapt test case, since MemRefCastOp is currently not a ViewLikeOp.
Jul 2 2020
Apr 28 2020
Apr 27 2020
Apr 22 2020
Updated commit message and added a short summary.
Feb 6 2020
@rriddle I've tried to come up with a minimal example including a new dialect. However, the issue does not occur in simple cases on our testing machines. The easiest way to dive into this issue would be to move line 728 in LowerGPUOpsToNNVMOps.cc to 750:
@mehdi_amini I fully get your point and I totally agree that we should not hide underlying bugs. However, you can argue as soon as two pattern matchers have the same benefit their ordering will be undefinied. In this case, the only possible solution you have, is to change the pattern matcher benefit.
@herhut We follow your arguments that we should unbreak the build on Windows asap. As mentioned above, this does not solve the general non-determinism issue. However, even if the ordering would be deterministic, we should not rely on intrinsic magic that might be easily broken by another commit in the future.
Jan 30 2020
We built the project on a Windows machine and were able to reproduce the failed test. We figured out that the tests are right, but the error was caused by a non-deterministic pattern rewriter step. We fixed the issue in another PR: https://reviews.llvm.org/D73713.
This fix should work on all platforms (Windows/Linux).
Jan 29 2020
Fixed test cases and format issues.
Jan 28 2020
Jan 27 2020
Fixed clang format issue.
Renamed BinaryIntrinsicOp to BinarySameArgsIntrinsicOp.
Renamed TernaryIntrinsicOp to TernarySameArgsIntrinsicOp.