From: John Hubbard <jhubbard@nvidia.com>
To: <willy@infradead.org>
Cc: <akpm@linux-foundation.org>, <cai@lca.pw>, <jhubbard@nvidia.com>,
<kirill@shutemov.name>, <linux-mm@kvack.org>,
<rppt@linux.ibm.com>, <vbabka@suse.cz>,
<william.kucharski@oracle.com>,
"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Subject: [PATCH] mm, dump_page: do not crash with bad compound_page()
Date: Tue, 4 Aug 2020 11:39:43 -0700 [thread overview]
Message-ID: <20200804183943.1244828-1-jhubbard@nvidia.com> (raw)
In-Reply-To: <20200804161414.GG23808@casper.infradead.org>
If a compound page is being split while dump_page() is being run on that
page, we can end up calling compound_mapcount() on a page that is no
longer compound. This leads to a crash (already seen at least once in
the field), due to the VM_BUG_ON_PAGE() assertion inside
compound_mapcount().
(The above is from Matthew Wilcox's analysis of Qian Cai's bug report.)
In order to avoid this kind of crash, make dump_page() slightly more
robust, by providing a version of compound_page() that doesn't assert,
but just warns the first time that such a thing happens. And the first
time is usually enough.
For debug tools, we don't want to go *too* far in this direction, but
this is a simple small fix, and the crash has already been seen, so it's
a good trade-off.
Reported-by: Qian Cai <cai@lca.pw>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
---
include/linux/mm.h | 17 +++++++++++++++--
mm/debug.c | 4 ++--
2 files changed, 17 insertions(+), 4 deletions(-)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index dc7b87310c10..e3991fbb42c0 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -779,6 +779,13 @@ static inline void *kvcalloc(size_t n, size_t size, gfp_t flags)
extern void kvfree(const void *addr);
extern void kvfree_sensitive(const void *addr, size_t len);
+
+static inline int __compound_mapcount(struct page *page)
+{
+ page = compound_head(page);
+ return atomic_read(compound_mapcount_ptr(page)) + 1;
+}
+
/*
* Mapcount of compound page as a whole, does not include mapped sub-pages.
*
@@ -787,8 +794,14 @@ extern void kvfree_sensitive(const void *addr, size_t len);
static inline int compound_mapcount(struct page *page)
{
VM_BUG_ON_PAGE(!PageCompound(page), page);
- page = compound_head(page);
- return atomic_read(compound_mapcount_ptr(page)) + 1;
+ return __compound_mapcount(page);
+}
+
+static inline int dump_page_compound_mapcount(struct page *page)
+{
+ if (WARN_ON_ONCE(!PageCompound(page)))
+ return 0;
+ return __compound_mapcount(page);
}
/*
diff --git a/mm/debug.c b/mm/debug.c
index 4f376514744d..eab4244aabd8 100644
--- a/mm/debug.c
+++ b/mm/debug.c
@@ -91,7 +91,7 @@ void __dump_page(struct page *page, const char *reason)
"compound_mapcount:%d compound_pincount:%d\n",
page, page_ref_count(head), mapcount,
mapping, page_to_pgoff(page), head,
- compound_order(head), compound_mapcount(page),
+ compound_order(head), dump_page_compound_mapcount(page),
compound_pincount(page));
} else {
pr_warn("page:%px refcount:%d mapcount:%d mapping:%p "
@@ -99,7 +99,7 @@ void __dump_page(struct page *page, const char *reason)
"compound_mapcount:%d\n",
page, page_ref_count(head), mapcount,
mapping, page_to_pgoff(page), head,
- compound_order(head), compound_mapcount(page));
+ compound_order(head), dump_page_compound_mapcount(page));
}
else
pr_warn("page:%px refcount:%d mapcount:%d mapping:%p index:%#lx\n",
--
2.28.0
next prev parent reply other threads:[~2020-08-04 18:39 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-09 20:21 [PATCH v2 0/6] Improvements for dump_page() Matthew Wilcox (Oracle)
2020-07-09 20:21 ` [PATCH v2 1/6] mm: Handle page->mapping better in dump_page Matthew Wilcox (Oracle)
2020-07-11 1:30 ` John Hubbard
2020-07-11 1:32 ` John Hubbard
2020-07-14 11:06 ` Vlastimil Babka
2020-07-09 20:21 ` [PATCH v2 2/6] mm: Dump compound page information on a second line Matthew Wilcox (Oracle)
2020-07-11 1:35 ` John Hubbard
2020-07-14 12:03 ` Vlastimil Babka
2020-08-04 15:37 ` Qian Cai
2020-08-04 16:14 ` Matthew Wilcox
2020-08-04 18:39 ` John Hubbard [this message]
2020-08-04 18:48 ` [PATCH] mm, dump_page: do not crash with bad compound_page() Matthew Wilcox
2020-08-04 19:17 ` John Hubbard
2020-08-04 19:42 ` Matthew Wilcox
2020-07-09 20:21 ` [PATCH v2 3/6] mm: Print head flags in dump_page Matthew Wilcox (Oracle)
2020-07-11 1:40 ` John Hubbard
2020-07-14 12:06 ` Vlastimil Babka
2020-07-09 20:21 ` [PATCH v2 4/6] mm: Switch dump_page to get_kernel_nofault Matthew Wilcox (Oracle)
2020-07-14 12:11 ` Vlastimil Babka
2020-07-09 20:21 ` [PATCH v2 5/6] mm: Print the inode number in dump_page Matthew Wilcox (Oracle)
2020-07-11 2:04 ` John Hubbard
2020-07-14 12:18 ` Vlastimil Babka
2020-07-09 20:21 ` [PATCH v2 6/6] mm: Print hashed address of struct page Matthew Wilcox (Oracle)
2020-07-11 1:48 ` John Hubbard
2020-07-09 20:54 ` [PATCH v2 0/6] Improvements for dump_page() Mike Rapoport
2020-07-09 20:54 ` William Kucharski
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200804183943.1244828-1-jhubbard@nvidia.com \
--to=jhubbard@nvidia.com \
--cc=akpm@linux-foundation.org \
--cc=cai@lca.pw \
--cc=kirill.shutemov@linux.intel.com \
--cc=kirill@shutemov.name \
--cc=linux-mm@kvack.org \
--cc=rppt@linux.ibm.com \
--cc=vbabka@suse.cz \
--cc=william.kucharski@oracle.com \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).