> On Oct 21, 2019, at 1:55 PM, Nicolas Saenz Julienne <nsaenzjulienne@xxxxxxx> wrote: > > On Mon, 2019-10-21 at 13:25 -0400, Qian Cai wrote: >>> On Oct 21, 2019, at 1:01 PM, Nicolas Saenz Julienne <nsaenzjulienne@xxxxxxx> >>> wrote: >>> >>> Could you enable CMA debugging to see if anything interesting comes out of >>> it. >> >> I did but nothing interesting came out. Did you use the same config I gave? > > Yes, aside from enabling ZONE_DMA. > >> Also, it has those cmdline. >> >> page_poison=on page_owner=on numa_balancing=enable \ >> systemd.unified_cgroup_hierarchy=1 debug_guardpage_minorder=1 \ >> page_alloc.shuffle=1 > > No luck, still works for me even after adding those extra flags. IIRC most of > them (if not all) are not even parsed by the time CMA is configured. > > So, can you confirm the zones setup you're seeing is similar to this one: > > [ 0.000000][ T0] Zone ranges: > [ 0.000000][ T0] DMA [mem 0x00000000802f0000-0x00000000bfffffff] > [ 0.000000][ T0] DMA32 [mem 0x00000000c0000000-0x00000000ffffffff] > [ 0.000000][ T0] Normal [mem 0x0000000100000000-0x00000093fcffffff] > > Maybe your memory starts between 0xe0000000-0xffffffff. That would be > problematic (although somewhat unwarranted). I managed to get more information here, [ 0.000000] cma: dma_contiguous_reserve(limit c0000000) [ 0.000000] cma: dma_contiguous_reserve: reserving 64 MiB for global area [ 0.000000] cma: cma_declare_contiguous(size 0x0000000004000000, base 0x0000000000000000, limit 0x00000000c0000000 alignment 0x0000000000000000) [ 0.000000] cma: Failed to reserve 512 MiB Full dmesg: https://cailca.github.io/files/dmesg.txt
