| Message ID | 20210914114348.15569-1-linmiaohe@huawei.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [v3] mm/page_isolation: fix potential missing call to unset_migratetype_isolate() | expand |
On 14.09.21 13:43, Miaohe Lin wrote: > In start_isolate_page_range() undo path, pfn_to_online_page() just checks > the first pfn in a pageblock while __first_valid_page() will traverse the > pageblock until the first online pfn is found. So we may miss the call to > unset_migratetype_isolate() in undo path and pages will remain isolated > unexpectedly. Fix this by calling undo_isolate_page_range() and this will > also help to simplify the code further. Note we shouldn't ever trigger it > because MAX_ORDER-1 aligned pfn ranges shouldn't contain memory holes now. > > Fixes: 2ce13640b3f4 ("mm: __first_valid_page skip over offline pages") > Signed-off-by: Miaohe Lin <linmiaohe@huawei.com> I read Michals reply, however, I am quite conservative with Fixes: tags. If there is nothing to fix, there is no BUG and the patch consequently merely a cleanup. I'd have gone with a patch description/subject as follows: " mm/page_isolation: cleanup start_isolate_page_range() We can heavily simplify the code by reusing undo_isolate_page_range(). Note that this also tackles a theoretical issue that would have been a real BUG before commit c5e79ef561b0 ("mm/memory_hotplug.c: don't allow to online/offline memory blocks with holes"). In start_isolate_page_range() undo path, pfn_to_online_page() just checks the first pfn in a pageblock while __first_valid_page() will traverse the pageblock until the first online pfn is found. So we may miss the call to unset_migratetype_isolate() in undo path and pages will remain isolated unexpectedly. Nowadays, start_isolate_page_range() never gets called on ranges that might contain memory holes. Consequently, this patch is not a fix but a cleanup. " Anyhow, whatever the other people prefer, no strong opinion. Reviewed-by: David Hildenbrand <david@redhat.com>
On 2021/9/15 2:13, David Hildenbrand wrote: > On 14.09.21 13:43, Miaohe Lin wrote: >> In start_isolate_page_range() undo path, pfn_to_online_page() just checks >> the first pfn in a pageblock while __first_valid_page() will traverse the >> pageblock until the first online pfn is found. So we may miss the call to >> unset_migratetype_isolate() in undo path and pages will remain isolated >> unexpectedly. Fix this by calling undo_isolate_page_range() and this will >> also help to simplify the code further. Note we shouldn't ever trigger it >> because MAX_ORDER-1 aligned pfn ranges shouldn't contain memory holes now. >> >> Fixes: 2ce13640b3f4 ("mm: __first_valid_page skip over offline pages") >> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com> > > I read Michals reply, however, I am quite conservative with Fixes: tags. If there is nothing to fix, there is no BUG and the patch consequently merely a cleanup. > > I'd have gone with a patch description/subject as follows: > > " > mm/page_isolation: cleanup start_isolate_page_range() > > We can heavily simplify the code by reusing undo_isolate_page_range(). > > Note that this also tackles a theoretical issue that would have been a real BUG before commit c5e79ef561b0 ("mm/memory_hotplug.c: don't allow to online/offline memory blocks with holes"). In start_isolate_page_range() undo path, pfn_to_online_page() just checks > the first pfn in a pageblock while __first_valid_page() will traverse the pageblock until the first online pfn is found. So we may miss the call to unset_migratetype_isolate() in undo path and pages will remain isolated unexpectedly. > > Nowadays, start_isolate_page_range() never gets called on ranges that might contain memory holes. Consequently, this patch is not a fix but a cleanup. > " > > Anyhow, whatever the other people prefer, no strong opinion. I have no preference too. But if this is preferred, I will do it. > > Reviewed-by: David Hildenbrand <david@redhat.com> Many thanks! :) >
diff --git a/mm/page_isolation.c b/mm/page_isolation.c index a95c2c6562d0..f93cc63d8fa1 100644 --- a/mm/page_isolation.c +++ b/mm/page_isolation.c @@ -183,7 +183,6 @@ int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn, unsigned migratetype, int flags) { unsigned long pfn; - unsigned long undo_pfn; struct page *page; BUG_ON(!IS_ALIGNED(start_pfn, pageblock_nr_pages)); @@ -193,25 +192,12 @@ int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn, pfn < end_pfn; pfn += pageblock_nr_pages) { page = __first_valid_page(pfn, pageblock_nr_pages); - if (page) { - if (set_migratetype_isolate(page, migratetype, flags)) { - undo_pfn = pfn; - goto undo; - } + if (page && set_migratetype_isolate(page, migratetype, flags)) { + undo_isolate_page_range(start_pfn, pfn, migratetype); + return -EBUSY; } } return 0; -undo: - for (pfn = start_pfn; - pfn < undo_pfn; - pfn += pageblock_nr_pages) { - struct page *page = pfn_to_online_page(pfn); - if (!page) - continue; - unset_migratetype_isolate(page, migratetype); - } - - return -EBUSY; } /*
In start_isolate_page_range() undo path, pfn_to_online_page() just checks the first pfn in a pageblock while __first_valid_page() will traverse the pageblock until the first online pfn is found. So we may miss the call to unset_migratetype_isolate() in undo path and pages will remain isolated unexpectedly. Fix this by calling undo_isolate_page_range() and this will also help to simplify the code further. Note we shouldn't ever trigger it because MAX_ORDER-1 aligned pfn ranges shouldn't contain memory holes now. Fixes: 2ce13640b3f4 ("mm: __first_valid_page skip over offline pages") Signed-off-by: Miaohe Lin <linmiaohe@huawei.com> --- v1->v2: Simplify the code further per David Hildenbrand. v2->v3: remove unneeded cc stable. --- mm/page_isolation.c | 20 +++----------------- 1 file changed, 3 insertions(+), 17 deletions(-)