On Thu, Jun 12, 2025 at 10:58:02AM -0300, Jason Gunthorpe wrote: > On Wed, Jun 11, 2025 at 11:59:00PM -0700, Nicolin Chen wrote: > > > We can see the 64MB was rounded up to 512MB by ksys_mmap_pgoff() > > when being passed in to hugetlb_file_setup() at: > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/mm/mmap.c?h=v6.16-rc1#n594 > > " len = ALIGN(len, huge_page_size(hs)); " > > > > By looking at the comments here..: > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/fs/hugetlbfs/inode.c#n1521 > > " > > /* > > * Note that size should be aligned to proper hugepage size in caller side, > > * otherwise hugetlb_reserve_pages reserves one less hugepages than > > intended. > > */ > > struct file *hugetlb_file_setup(const char *name, size_t size, > > " > > > > ..I guess this function was supposed to fail the not-a-multiple > > case as you remarked? But it certainly can't do that, when that > > size passed in is already hugepage-aligned.. > > > > It feels like a kernel bug as you suspect :-/ > > Certainly is > > > And I just found one more weird thing... > > > > In iommufd.c selftest code, we have: > > "static __attribute__((constructor)) void setup_sizes(void)" > > where it does another pair of posix_memalign/mmap, although this > > one doesn't flag MAP_HUGETLB and shouldn't impact what is coming > > to the next... > > This could all just be more weirdness from the above, it doesn't > really make alot of sense. > > I think change things so the MAP_HUGETLB test all skip if > HUGEPAGE_SIZE < buffer_size and move on.. > > Can't run those tests on ARM64 64k which is unfortunate.. I thought > there were patches to give that config a 2M huge page size option > based on the new contiguous page support though? Maybe it was only THPS..
If the assumption is that this is most likely a kernel bug, shouldn't it be fixed properly rather than worked around? After all the job of a selftest is to detect bugs to be fixed. But I wasn't able to follow all of your discussions, so I may be missing something. If the test is broken on ARM64 64k in general then I am also wondering how it didn't fail before my change to the selftest harness. Thomas
