You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@arrow.apache.org by "Jörn Horstmann (Jira)" <ji...@apache.org> on 2020/12/15 10:22:00 UTC

[jira] [Created] (ARROW-10914) [Rust]: SIMD implementation of arithmetic kernels reads out of bounds

Jörn Horstmann created ARROW-10914:
--------------------------------------

             Summary: [Rust]: SIMD implementation of arithmetic kernels reads out of bounds
                 Key: ARROW-10914
                 URL: https://issues.apache.org/jira/browse/ARROW-10914
             Project: Apache Arrow
          Issue Type: Bug
          Components: Rust
            Reporter: Jörn Horstmann
            Assignee: Jörn Horstmann


The simd arithmetic kernels use the following pattern repeatedly:

{code}
    for i in (0..left.len()).step_by(lanes) { ... }
{code}

If len is not a multiple of the number of lanes, this would read out of bounds in the last iteration. Currently, all buffers have an additional padding of 64 bytes (equal to the simd width), which masks this problem in most tests. As soon as we use a slice of an array, it should however be reproducible even with this padding.

Even without a crash, the issue is detectable with valgrind:

{code}
==31106== Invalid read of size 32
==31106==    at 0x1ECEE1: arrow::compute::kernels::arithmetic::add::hfded8b2c06cf22de (in /home/joernhorstmann/Source/github/apache/arrow/rust/target/release/deps/arrow-205580f93d58d5a9)
==31106==    by 0x2650EF: arrow::compute::kernels::arithmetic::tests::test_arithmetic_kernel_should_not_rely_on_padding::hacb7c7921dc38e6a (in /home/joernhorstmann/Source/github/apache/arrow/rust/target/release/deps/arrow-205580f93d58d5a9)
{code}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)