Re: rocrand_5.5.1-1_source.changes ACCEPTED into unstable

[Christian Kastner <ckk@debian.org>]

On 2023-08-25 10:34, Christian Kastner wrote:
> My local test however should that while all tests passed, a handful of
> test took an abnormally high amount of time: 17s, where < 10ms are
> typical. That's a subject for another thread.

5.5.1-1 got picked up by CI [1]. Tests pass, but you can see the
outliers in the test log, e.g.

> 577s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/0
> 577s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/0 (0 ms)
> 577s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/1
> 577s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/1 (0 ms)
> 577s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/2
> 577s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/2 (1 ms)
> 577s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/3
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/3 (17052 ms)
> 594s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/4
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/4 (0 ms)
> 594s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/5
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/5 (2 ms)
> 594s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/6
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/6 (2 ms)
> 594s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/7
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/7 (5 ms)
> 594s [ RUN      ] hiprand/hiprand_api.hiprand_generate_uniform_test/8
> 594s [       OK ] hiprand/hiprand_api.hiprand_generate_uniform_test/8 (5 ms)

No investigation yet.

Best,
Christian

[1] https://ci.rocm.debian.net/packages/r/rocrand/unstable/amd64+gfx1030/