Background ========== I once sent out the preparatory patches [1] but I dropped the last patch of using fallback nid with memory_add_physaddr_to_nid() due to some dependency.
After phys_addr_to_target_node() and memory_add_physaddr_to_nid() are stable now, it's time to fix the original bug on arm64 now. Compared with the last version [2], this version rebases the patch to latest v5.14-rc3 (s/kmem_start/range.start) Test ==== Tested on ThunderX2 host/qemu "-M virt" guest with a nvdimm device. The memblocks from the dax pmem device can be either hot-added or hot-removed on arm64 guest. Also passed the compilation test on x86. [1] https://www.mail-archive.com/[email protected]/msg2228771.html [2] https://lkml.org/lkml/2020/7/8/1546 Jia He (1): device-dax: use fallback nid when numa_node is invalid drivers/dax/kmem.c | 36 ++++++++++++++++++++---------------- 1 file changed, 20 insertions(+), 16 deletions(-) -- 2.17.1
