Given that dax / device-mapped pages are never subject to page allocations remove them from consideration by the soft-offline mechanism.
Reported-by: Naoya Horiguchi <[email protected]> Signed-off-by: Dan Williams <[email protected]> --- mm/memory-failure.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/mm/memory-failure.c b/mm/memory-failure.c index 9d142b9b86dc..988f977db3d2 100644 --- a/mm/memory-failure.c +++ b/mm/memory-failure.c @@ -1751,6 +1751,14 @@ int soft_offline_page(struct page *page, int flags) int ret; unsigned long pfn = page_to_pfn(page); + if (is_zone_device_page(page)) { + pr_debug_ratelimited("soft_offline: %#lx page is device page\n", + pfn); + if (flags & MF_COUNT_INCREASED) + put_page(page); + return -EIO; + } + if (PageHWPoison(page)) { pr_info("soft offline: %#lx page already poisoned\n", pfn); if (flags & MF_COUNT_INCREASED) _______________________________________________ Linux-nvdimm mailing list [email protected] https://lists.01.org/mailman/listinfo/linux-nvdimm
