Hi Anders -
Can you look at this tomorrow?
Regression in test libhugetlbfs/truncate_above_4GB-2M-32 on arm32 that
showed up on mainline on March 23 and is in the 4.14 and 4.15 RCs that
Greg pushed today.
I suspect commit
is the culprit (because it is in all three bad tree diffs and has
hugetlbfs in the name..).
I tried reproducing locally on my x15 but I'm having trouble getting
the kernel installed.
If you're able, can you try reproducing the error and then dropping
the above commit and seeing if it is resolved? We'll have to get back
to Greg tomorrow about this so it would be nice to be able to report
the specific patch.