Skip to content

BUG: einsum with int8/uint8 input is buggy #24732

Closed

Description

@rgommers

These tests are failing on macOS x86-64 and with clang-cl on Windows after the move to Meson:

FAILED core/tests/test_einsum.py::TestEinsum::test_einsum_sums_int8 - AssertionError:
FAILED core/tests/test_einsum.py::TestEinsum::test_einsum_sums_uint8 - AssertionError:

It passes with Clang on macOS arm64, and with MSVC on Windows. It's also passing with and without SIMD optimizations. So it's very likely to be a Clang + x86-64 + exact compile flags used bug in the C code somewhere. This needs investigation.

Activity

  1. added a commit that references this issue on Jun 3, 2024
  2. added this to the 2.1.0 release milestone on Jun 5, 2024
  3. rgommers commented on Jun 5, 2024

    @rgommers
    MemberAuthor

    This fixed itself somewhere along the way; the tests have been re-enabled in gh-26602.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions