linux-kernel - [PATCH 4.9 55/83] mm, gup: remove broken VM_BUG_ON

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sun,  9 Jun 2019 18:42:25 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Will Deacon <will.deacon@....com>,
        Punit Agrawal <punit.agrawal@....com>,
        Steve Capper <steve.capper@....com>,
        "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
        "Aneesh Kumar K.V" <aneesh.kumar@...ux.vnet.ibm.com>,
        Catalin Marinas <catalin.marinas@....com>,
        Naoya Horiguchi <n-horiguchi@...jp.nec.com>,
        Mark Rutland <mark.rutland@....com>,
        Hillf Danton <hillf.zj@...baba-inc.com>,
        Michal Hocko <mhocko@...e.com>,
        Mike Kravetz <mike.kravetz@...cle.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Ben Hutchings <ben.hutchings@...ethink.co.uk>
Subject: [PATCH 4.9 55/83] mm, gup: remove broken VM_BUG_ON_PAGE compound check for hugepages

From: Will Deacon <will.deacon@....com>

commit a3e328556d41bb61c55f9dfcc62d6a826ea97b85 upstream.

When operating on hugepages with DEBUG_VM enabled, the GUP code checks
the compound head for each tail page prior to calling
page_cache_add_speculative.  This is broken, because on the fast-GUP
path (where we don't hold any page table locks) we can be racing with a
concurrent invocation of split_huge_page_to_list.

split_huge_page_to_list deals with this race by using page_ref_freeze to
freeze the page and force concurrent GUPs to fail whilst the component
pages are modified.  This modification includes clearing the
compound_head field for the tail pages, so checking this prior to a
successful call to page_cache_add_speculative can lead to false
positives: In fact, page_cache_add_speculative *already* has this check
once the page refcount has been successfully updated, so we can simply
remove the broken calls to VM_BUG_ON_PAGE.

Link: http://lkml.kernel.org/r/20170522133604.11392-2-punit.agrawal@arm.com
Signed-off-by: Will Deacon <will.deacon@....com>
Signed-off-by: Punit Agrawal <punit.agrawal@....com>
Acked-by: Steve Capper <steve.capper@....com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@...ux.intel.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@...ux.vnet.ibm.com>
Cc: Catalin Marinas <catalin.marinas@....com>
Cc: Naoya Horiguchi <n-horiguchi@...jp.nec.com>
Cc: Mark Rutland <mark.rutland@....com>
Cc: Hillf Danton <hillf.zj@...baba-inc.com>
Cc: Michal Hocko <mhocko@...e.com>
Cc: Mike Kravetz <mike.kravetz@...cle.com>
Signed-off-by: Andrew Morton <akpm@...ux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@...ux-foundation.org>
Signed-off-by: Ben Hutchings <ben.hutchings@...ethink.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
 mm/gup.c |    3 ---
 1 file changed, 3 deletions(-)

--- a/mm/gup.c
+++ b/mm/gup.c
@@ -1316,7 +1316,6 @@ static int gup_huge_pmd(pmd_t orig, pmd_
 	head = pmd_page(orig);
 	page = head + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
 	do {
-		VM_BUG_ON_PAGE(compound_head(page) != head, page);
 		pages[*nr] = page;
 		(*nr)++;
 		page++;
@@ -1351,7 +1350,6 @@ static int gup_huge_pud(pud_t orig, pud_
 	head = pud_page(orig);
 	page = head + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
 	do {
-		VM_BUG_ON_PAGE(compound_head(page) != head, page);
 		pages[*nr] = page;
 		(*nr)++;
 		page++;
@@ -1387,7 +1385,6 @@ static int gup_huge_pgd(pgd_t orig, pgd_
 	head = pgd_page(orig);
 	page = head + ((addr & ~PGDIR_MASK) >> PAGE_SHIFT);
 	do {
-		VM_BUG_ON_PAGE(compound_head(page) != head, page);
 		pages[*nr] = page;
 		(*nr)++;
 		page++;