On Thu, Sep 17, 2026 at 10:21:47PM +0200, Pavel Machek wrote:
Seems to be identical / very similar to crash we see, and we have full log:
It is the same one - the photo shows the same gather_bootmem_prealloc_parallel() fault as your LAVA log, and Peter's and Brett's bisects both land on it. Dropped from the 6.12 queue:
b1b7c045e808 ("mm/hugetlb: initialize gigantic bootmem hugepage struct pages earlier")
Thanks both of you for correlating these - having the full oops next to the bisect made this a quick call.