On Thu, Apr 4, 2024 at 1:17 PM Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > > On Thu, 4 Apr 2024 16:25:15 +0000 Frank van der Linden <fvdl@xxxxxxxxxx> wrote: > > > The hugetlb_cma code passes 0 in the order_per_bit argument to > > cma_declare_contiguous_nid (the alignment, computed using the > > page order, is correctly passed in). > > > > This causes a bit in the cma allocation bitmap to always represent > > a 4k page, making the bitmaps potentially very large, and slower. > > > > So, correctly pass in the order instead. > > Ditto. Should we backport this? Can we somewhat quantify "potentially very", > and understand under what circumstances this might occur? It would create bitmaps that would be pretty big. E.g. for a 4k page size on x86, hugetlb_cma=64G would mean a bitmap size of (64G / 4k) / 8 == 2M. With HUGETLB_PAGE_ORDER as order_per_bit, as intended, this would be (64G / 2M) / 8 == 4k. So, that's quite a difference :) Also, this restricted the hugetlb_cma area to ((PAGE_SIZE << MAX_PAGE_ORDER) * 8) * PAGE_SIZE (e.g. 128G on x86) , since bitmap_alloc uses normal page allocation, and is thus restricted by MAX_PAGE_ORDER. Specifying anything about that would fail the CMA initialization. - Frank
