<?xml version='1.0' encoding='UTF-8'?>
<?xml-stylesheet href="/static/style.xsl" type="text/xsl"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" version="2.0">
  <channel>
    <title>Most recent entries from all</title>
    <link>https://vulnerability.circl.lu</link>
    <description>Contains only the most 10 recent entries.</description>
    <docs>http://www.rssboard.org/rss-specification</docs>
    <generator>python-feedgen</generator>
    <language>en</language>
    <lastBuildDate>Tue, 29 Sep 2026 02:00:01 +0000</lastBuildDate>
    <item>
      <title>CVE-2026-97533 — x86/mm/pat: Acquire init_mm read lock on attribute changes to avoid UAF</title>
      <link>https://vulnerability.circl.lu/vuln/cve-2026-97533</link>
      <description>&lt;p&gt;&lt;strong&gt;Affected:&lt;/strong&gt; Linux&lt;/p&gt;
&lt;p&gt;In the Linux kernel, the following vulnerability has been resolved:&lt;/p&gt;
&lt;p&gt;x86/mm/pat: Acquire init_mm read lock on attribute changes to avoid UAF&lt;/p&gt;
&lt;p&gt;A previous commit protected against races between ptdump and CPA collapse,
however one still exists between attribute changes and collapse as reported
by Denis V. Lunev (linked).&lt;/p&gt;
&lt;p&gt;When an attribute change arises, a lockless page table walker obtains a PTE
entry, which is later written to via set_pte_atomic():&lt;/p&gt;
&lt;p&gt;...
  -&amp;gt; change_page_attr_set_clr()
  -&amp;gt; __change_page_attr_set_clr()
  -&amp;gt; __change_page_attr()
	-&amp;gt; _lookup_address_cpa()
	-&amp;gt; lookup_address_in_pgd_attr()
	-&amp;gt; [ lockless page table walker ]
  -&amp;gt; set_pte_atomic()&lt;/p&gt;
&lt;p&gt;There is nothing preventing a concurrent CPA collapse which can free the
PTE that was retrieved here, resulting in a use-after-free.&lt;/p&gt;
&lt;p&gt;With the mmap write lock taken on init_mm over CPA collapse, resolve this
race by acquiring an mmap read lock on init_mm over
__change_page_attr_set_clr().&lt;/p&gt;
&lt;p&gt;This locks across the whole operation over which the walk and the PTE
entry write occurs, solving the race.&lt;/p&gt;
&lt;p&gt;It is safe to do this here, as no spinlocks are held upon entry to
__change_page_attr_set_clr().&lt;/p&gt;
&lt;p&gt;However, the lock must not be held over an allocation, as allocation can
trigger reclaim and shrinkers may call into CPA recursively, making
deadlocks possible (init_mm -&amp;gt; ... -&amp;gt; fs_reclaim -&amp;gt; init_mm).&lt;/p&gt;
&lt;p&gt;A page table is allocated when a huge page needs to be split:&lt;/p&gt;
&lt;p&gt;-&amp;gt; change_page_attr_set_clr()
  -&amp;gt; __change_page_attr_set_clr()…&lt;/p&gt;</description>
      <content:encoded>&lt;p&gt;&lt;strong&gt;Affected:&lt;/strong&gt; Linux&lt;/p&gt;
&lt;p&gt;In the Linux kernel, the following vulnerability has been resolved:&lt;/p&gt;
&lt;p&gt;x86/mm/pat: Acquire init_mm read lock on attribute changes to avoid UAF&lt;/p&gt;
&lt;p&gt;A previous commit protected against races between ptdump and CPA collapse,
however one still exists between attribute changes and collapse as reported
by Denis V. Lunev (linked).&lt;/p&gt;
&lt;p&gt;When an attribute change arises, a lockless page table walker obtains a PTE
entry, which is later written to via set_pte_atomic():&lt;/p&gt;
&lt;p&gt;...
  -&amp;gt; change_page_attr_set_clr()
  -&amp;gt; __change_page_attr_set_clr()
  -&amp;gt; __change_page_attr()
	-&amp;gt; _lookup_address_cpa()
	-&amp;gt; lookup_address_in_pgd_attr()
	-&amp;gt; [ lockless page table walker ]
  -&amp;gt; set_pte_atomic()&lt;/p&gt;
&lt;p&gt;There is nothing preventing a concurrent CPA collapse which can free the
PTE that was retrieved here, resulting in a use-after-free.&lt;/p&gt;
&lt;p&gt;With the mmap write lock taken on init_mm over CPA collapse, resolve this
race by acquiring an mmap read lock on init_mm over
__change_page_attr_set_clr().&lt;/p&gt;
&lt;p&gt;This locks across the whole operation over which the walk and the PTE
entry write occurs, solving the race.&lt;/p&gt;
&lt;p&gt;It is safe to do this here, as no spinlocks are held upon entry to
__change_page_attr_set_clr().&lt;/p&gt;
&lt;p&gt;However, the lock must not be held over an allocation, as allocation can
trigger reclaim and shrinkers may call into CPA recursively, making
deadlocks possible (init_mm -&amp;gt; ... -&amp;gt; fs_reclaim -&amp;gt; init_mm).&lt;/p&gt;
&lt;p&gt;A page table is allocated when a huge page needs to be split:&lt;/p&gt;
&lt;p&gt;-&amp;gt; change_page_attr_set_clr()
  -&amp;gt; __change_page_attr_set_clr()…&lt;/p&gt;</content:encoded>
      <guid isPermaLink="false">https://vulnerability.circl.lu/vuln/cve-2026-97533</guid>
    </item>
  </channel>
</rss>
