test_expanded_reduction_cpu failure on pytorch/pytorch CI

[jansel]

We should fix this one, I suspect not installing Triton was hiding this failure.

    It started to happen as soon as dynamo ci was enabled, see https://hud.pytorch.org/pytorch/pytorch/commit/a0c2a7f2eda788a48f1d243940297f1467faf138

Originally posted by @malfet in #87157 (comment)

cc @ezyang @soumith @msaroufim @wconstab @ngimel @bdhirsh

[jansel]

Looks like test_dense_mask_index_cpu might have a similar issue.

I disabled the two tests in #87084

To reproduce you can enable the two tests and add the ciflow/periodic to your PR.