The patch titled
Subject: mm/memory-failure: split thp earlier in memory error handling
has been added to the -mm tree. Its filename is
mm-memory-failure-split-thp-earlier-in-memory-error-handling.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-memory-failure-split-thp-earlier-in-memory-error-handling.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-memory-failure-split-thp-earlier-in-memory-error-handling.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Subject: mm/memory-failure: split thp earlier in memory error handling
memory_failure() doesn't handle thp itself at this time and need to split
it before doing isolation. Currently thp is split in the middle of
hwpoison_user_mappings(), but there're corner cases where memory_failure()
wrongly tries to handle thp without splitting.
1) "non anonymous" thp, which is not a normal operating mode of thp,
but a memory error could hit a thp before anon_vma is initialized. In
such case, split_huge_page() fails and me_huge_page() (intended for
hugetlb) is called for thp, which triggers BUG_ON in page_hstate().
2) !PageLRU case, where hwpoison_user_mappings() returns with
SWAP_SUCCESS and the result is the same as case 1.
memory_failure() can't avoid splitting, so let's split it more earlier,
which also reduces code which are prepared for both of normal page and
thp.
Signed-off-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Cc: Andi Kleen <andi@xxxxxxxxxxxxxx>
Cc: Tony Luck <tony.luck@xxxxxxxxx>
Cc: "Kirill A. Shutemov" <kirill@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/memory-failure.c | 88 +++++++++++-------------------------------
1 file changed, 25 insertions(+), 63 deletions(-)
diff -puN mm/memory-failure.c~mm-memory-failure-split-thp-earlier-in-memory-error-handling mm/memory-failure.c
--- a/mm/memory-failure.c~mm-memory-failure-split-thp-earlier-in-memory-error-handling
+++ a/mm/memory-failure.c
@@ -898,7 +898,6 @@ static int hwpoison_user_mappings(struct
int ret;
int kill = 1, forcekill;
struct page *hpage = *hpagep;
- struct page *ppage;
/*
* Here we are interested only in user-mapped pages, so skip any
@@ -948,59 +947,6 @@ static int hwpoison_user_mappings(struct
}
/*
- * ppage: poisoned page
- * if p is regular page(4k page)
- * ppage == real poisoned page;
- * else p is hugetlb or THP, ppage == head page.
- */
- ppage = hpage;
-
- if (PageTransHuge(hpage)) {
- /*
- * Verify that this isn't a hugetlbfs head page, the check for
- * PageAnon is just for avoid tripping a split_huge_page
- * internal debug check, as split_huge_page refuses to deal with
- * anything that isn't an anon page. PageAnon can't go away fro
- * under us because we hold a refcount on the hpage, without a
- * refcount on the hpage. split_huge_page can't be safely called
- * in the first place, having a refcount on the tail isn't
- * enough * to be safe.
- */
- if (!PageHuge(hpage) && PageAnon(hpage)) {
- if (unlikely(split_huge_page(hpage))) {
- /*
- * FIXME: if splitting THP is failed, it is
- * better to stop the following operation rather
- * than causing panic by unmapping. System might
- * survive if the page is freed later.
- */
- printk(KERN_INFO
- "MCE %#lx: failed to split THP\n", pfn);
-
- BUG_ON(!PageHWPoison(p));
- return SWAP_FAIL;
- }
- /*
- * We pinned the head page for hwpoison handling,
- * now we split the thp and we are interested in
- * the hwpoisoned raw page, so move the refcount
- * to it. Similarly, page lock is shifted.
- */
- if (hpage != p) {
- if (!(flags & MF_COUNT_INCREASED)) {
- put_page(hpage);
- get_page(p);
- }
- lock_page(p);
- unlock_page(hpage);
- *hpagep = p;
- }
- /* THP is split, so ppage should be the real poisoned page. */
- ppage = p;
- }
- }
-
- /*
* First collect all the processes that have the page
* mapped in dirty form. This has to be done before try_to_unmap,
* because ttu takes the rmap data structures down.
@@ -1009,12 +955,12 @@ static int hwpoison_user_mappings(struct
* there's nothing that can be done.
*/
if (kill)
- collect_procs(ppage, &tokill, flags & MF_ACTION_REQUIRED);
+ collect_procs(hpage, &tokill, flags & MF_ACTION_REQUIRED);
- ret = try_to_unmap(ppage, ttu);
+ ret = try_to_unmap(hpage, ttu);
if (ret != SWAP_SUCCESS)
printk(KERN_ERR "MCE %#lx: failed to unmap page (mapcount=%d)\n",
- pfn, page_mapcount(ppage));
+ pfn, page_mapcount(hpage));
/*
* Now that the dirty bit has been propagated to the
@@ -1026,7 +972,7 @@ static int hwpoison_user_mappings(struct
* use a more force-full uncatchable kill to prevent
* any accesses to the poisoned memory.
*/
- forcekill = PageDirty(ppage) || (flags & MF_MUST_KILL);
+ forcekill = PageDirty(hpage) || (flags & MF_MUST_KILL);
kill_procs(&tokill, forcekill, trapno,
ret != SWAP_SUCCESS, p, pfn, flags);
@@ -1072,6 +1018,7 @@ int memory_failure(unsigned long pfn, in
struct page_state *ps;
struct page *p;
struct page *hpage;
+ struct page *orig_head;
int res;
unsigned int nr_pages;
unsigned long page_flags;
@@ -1087,7 +1034,7 @@ int memory_failure(unsigned long pfn, in
}
p = pfn_to_page(pfn);
- hpage = compound_head(p);
+ orig_head = hpage = compound_head(p);
if (TestSetPageHWPoison(p)) {
printk(KERN_ERR "MCE %#lx: already hardware poisoned\n", pfn);
return 0;
@@ -1150,6 +1097,21 @@ int memory_failure(unsigned long pfn, in
}
}
+ if (!PageHuge(p) && PageTransHuge(hpage)) {
+ if (!PageAnon(hpage)) {
+ pr_info("MCE: %#lx: non anonymous thp", pfn);
+ put_page(p);
+ return -EBUSY;
+ }
+ if (unlikely(split_huge_page(hpage))) {
+ pr_info("MCE: %#lx: thp split failed", pfn);
+ put_page(p);
+ return -EBUSY;
+ }
+ VM_BUG_ON_PAGE(!page_count(p), p);
+ hpage = compound_head(p);
+ }
+
/*
* We ignore non-LRU pages for good reasons.
* - PG_locked is only well defined for LRU pages and a few others
@@ -1159,9 +1121,9 @@ int memory_failure(unsigned long pfn, in
* walked by the page reclaim code, however that's not a big loss.
*/
if (!PageHuge(p)) {
- if (!PageLRU(hpage))
- shake_page(hpage, 0);
- if (!PageLRU(hpage)) {
+ if (!PageLRU(p))
+ shake_page(p, 0);
+ if (!PageLRU(p)) {
/*
* shake_page could have turned it free.
*/
@@ -1182,7 +1144,7 @@ int memory_failure(unsigned long pfn, in
* The page could have changed compound pages during the locking.
* If this happens just bail out.
*/
- if (compound_head(p) != hpage) {
+ if (PageCompound(p) && compound_head(p) != orig_head) {
action_result(pfn, MF_MSG_DIFFERENT_COMPOUND, MF_IGNORED);
res = -EBUSY;
goto out;
_
Patches currently in -mm which might be from n-horiguchi@xxxxxxxxxxxxx are
tools-vm-fix-page-flags-build.patch
mm-hwpoison-add-comment-describing-when-to-add-new-cases.patch
mm-hwpoison-remove-obsolete-notebook-todo-list.patch
memory-failure-export-page_type-and-action-result.patch
memory-failure-change-type-of-action_results-param-3-to-enum.patch
tracing-add-trace-event-for-memory-failure.patch
mm-memory-failure-split-thp-earlier-in-memory-error-handling.patch
mm-memory-failure-introduce-get_hwpoison_page-for-consistent-refcount-handling.patch
mm-soft-offline-dont-free-target-page-in-successful-page-migration.patch
mm-memory-failure-me_huge_page-does-nothing-for-thp.patch
page-flags-trivial-cleanup-for-pagetrans-helpers.patch
page-flags-introduce-page-flags-policies-wrt-compound-pages.patch
page-flags-define-pg_locked-behavior-on-compound-pages.patch
page-flags-define-behavior-of-fs-io-related-flags-on-compound-pages.patch
page-flags-define-behavior-of-lru-related-flags-on-compound-pages.patch
page-flags-define-behavior-slb-related-flags-on-compound-pages.patch
page-flags-define-behavior-of-xen-related-flags-on-compound-pages.patch
page-flags-define-pg_reserved-behavior-on-compound-pages.patch
page-flags-define-pg_swapbacked-behavior-on-compound-pages.patch
page-flags-define-pg_swapcache-behavior-on-compound-pages.patch
page-flags-define-pg_mlocked-behavior-on-compound-pages.patch
page-flags-define-pg_uncached-behavior-on-compound-pages.patch
page-flags-define-pg_uptodate-behavior-on-compound-pages.patch
page-flags-look-on-head-page-if-the-flag-is-encoded-in-page-mapping.patch
mm-sanitize-page-mapping-for-tail-pages.patch
do_shared_fault-check-that-mmap_sem-is-held.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
