diff options
Diffstat (limited to 'mm/memory-failure.c')
| -rw-r--r-- | mm/memory-failure.c | 76 |
1 files changed, 38 insertions, 38 deletions
diff --git a/mm/memory-failure.c b/mm/memory-failure.c index 5b663eca1f29..9a285038d765 100644 --- a/mm/memory-failure.c +++ b/mm/memory-failure.c @@ -6,16 +6,16 @@ * High level machine check handler. Handles pages reported by the * hardware as being corrupted usually due to a multi-bit ECC memory or cache * failure. - * + * * In addition there is a "soft offline" entry point that allows stop using * not-yet-corrupted-by-suspicious pages without killing anything. * * Handles page cache pages in various states. The tricky part - * here is that we can access any page asynchronously in respect to - * other VM users, because memory failures could happen anytime and - * anywhere. This could violate some of their assumptions. This is why - * this code has to be extremely careful. Generally it tries to use - * normal locking rules, as in get the standard locks, even if that means + * here is that we can access any page asynchronously in respect to + * other VM users, because memory failures could happen anytime and + * anywhere. This could violate some of their assumptions. This is why + * this code has to be extremely careful. Generally it tries to use + * normal locking rules, as in get the standard locks, even if that means * the error handling takes potentially a long time. * * It can be very tempting to add handling for obscure cases here. @@ -25,12 +25,12 @@ * https://git.kernel.org/cgit/utils/cpu/mce/mce-test.git/ * - The case actually shows up as a frequent (top 10) page state in * tools/mm/page-types when running a real workload. - * + * * There are several operations here with exponential complexity because - * of unsuitable VM data structures. For example the operation to map back - * from RMAP chains to processes has to walk the complete process list and + * of unsuitable VM data structures. For example the operation to map back + * from RMAP chains to processes has to walk the complete process list and * has non linear complexity with the number. But since memory corruptions - * are rare we hope to get away with this. This avoids impacting the core + * are rare we hope to get away with this. This avoids impacting the core * VM. */ @@ -123,7 +123,6 @@ const struct attribute_group memory_failure_attr_group = { .attrs = memory_failure_attr, }; -#ifdef CONFIG_SYSCTL static struct ctl_table memory_failure_table[] = { { .procname = "memory_failure_early_kill", @@ -146,14 +145,6 @@ static struct ctl_table memory_failure_table[] = { { } }; -static int __init memory_failure_sysctl_init(void) -{ - register_sysctl_init("vm", memory_failure_table); - return 0; -} -late_initcall(memory_failure_sysctl_init); -#endif /* CONFIG_SYSCTL */ - /* * Return values: * 1: the page is dissolved (if needed) and taken off from buddy, @@ -395,6 +386,7 @@ static unsigned long dev_pagemap_mapping_shift(struct vm_area_struct *vma, pud_t *pud; pmd_t *pmd; pte_t *pte; + pte_t ptent; VM_BUG_ON_VMA(address == -EFAULT, vma); pgd = pgd_offset(vma->vm_mm, address); @@ -414,7 +406,10 @@ static unsigned long dev_pagemap_mapping_shift(struct vm_area_struct *vma, if (pmd_devmap(*pmd)) return PMD_SHIFT; pte = pte_offset_map(pmd, address); - if (pte_present(*pte) && pte_devmap(*pte)) + if (!pte) + return 0; + ptent = ptep_get(pte); + if (pte_present(ptent) && pte_devmap(ptent)) ret = PAGE_SHIFT; pte_unmap(pte); return ret; @@ -800,13 +795,13 @@ static int hwpoison_pte_range(pmd_t *pmdp, unsigned long addr, goto out; } - if (pmd_trans_unstable(pmdp)) - goto out; - mapped_pte = ptep = pte_offset_map_lock(walk->vma->vm_mm, pmdp, addr, &ptl); + if (!ptep) + goto out; + for (; addr != end; ptep++, addr += PAGE_SIZE) { - ret = check_hwpoisoned_entry(*ptep, addr, PAGE_SHIFT, + ret = check_hwpoisoned_entry(ptep_get(ptep), addr, PAGE_SHIFT, hwp->pfn, &hwp->tk); if (ret == 1) break; @@ -2441,6 +2436,8 @@ static int __init memory_failure_init(void) INIT_WORK(&mf_cpu->work, memory_failure_work_func); } + register_sysctl_init("vm", memory_failure_table); + return 0; } core_initcall(memory_failure_init); @@ -2469,7 +2466,7 @@ int unpoison_memory(unsigned long pfn) { struct folio *folio; struct page *p; - int ret = -EBUSY; + int ret = -EBUSY, ghp; unsigned long count = 1; bool huge = false; static DEFINE_RATELIMIT_STATE(unpoison_rs, DEFAULT_RATELIMIT_INTERVAL, @@ -2490,7 +2487,7 @@ int unpoison_memory(unsigned long pfn) goto unlock_mutex; } - if (!folio_test_hwpoison(folio)) { + if (!PageHWPoison(p)) { unpoison_pr_info("Unpoison: Page was already unpoisoned %#lx\n", pfn, &unpoison_rs); goto unlock_mutex; @@ -2502,6 +2499,13 @@ int unpoison_memory(unsigned long pfn) goto unlock_mutex; } + if (folio_test_slab(folio) || PageTable(&folio->page) || folio_test_reserved(folio)) + goto unlock_mutex; + + /* + * Note that folio->_mapcount is overloaded in SLAB, so the simple test + * in folio_mapped() has to be done after folio_test_slab() is checked. + */ if (folio_mapped(folio)) { unpoison_pr_info("Unpoison: Someone maps the hwpoison page %#lx\n", pfn, &unpoison_rs); @@ -2514,32 +2518,28 @@ int unpoison_memory(unsigned long pfn) goto unlock_mutex; } - if (folio_test_slab(folio) || PageTable(&folio->page) || folio_test_reserved(folio)) - goto unlock_mutex; - - ret = get_hwpoison_page(p, MF_UNPOISON); - if (!ret) { + ghp = get_hwpoison_page(p, MF_UNPOISON); + if (!ghp) { if (PageHuge(p)) { huge = true; count = folio_free_raw_hwp(folio, false); - if (count == 0) { - ret = -EBUSY; + if (count == 0) goto unlock_mutex; - } } ret = folio_test_clear_hwpoison(folio) ? 0 : -EBUSY; - } else if (ret < 0) { - if (ret == -EHWPOISON) { + } else if (ghp < 0) { + if (ghp == -EHWPOISON) { ret = put_page_back_buddy(p) ? 0 : -EBUSY; - } else + } else { + ret = ghp; unpoison_pr_info("Unpoison: failed to grab page %#lx\n", pfn, &unpoison_rs); + } } else { if (PageHuge(p)) { huge = true; count = folio_free_raw_hwp(folio, false); if (count == 0) { - ret = -EBUSY; folio_put(folio); goto unlock_mutex; } |