CVE-2022-49406

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> block: Fix potential deadlock in blk_ia_range_sysfs_show()<br /> <br /> When being read, a sysfs attribute is already protected against removal<br /> with the kobject node active reference counter. As a result, in<br /> blk_ia_range_sysfs_show(), there is no need to take the queue sysfs<br /> lock when reading the value of a range attribute. Using the queue sysfs<br /> lock in this function creates a potential deadlock situation with the<br /> disk removal, something that a lockdep signals with a splat when the<br /> device is removed:<br /> <br /> [ 760.703551] Possible unsafe locking scenario:<br /> [ 760.703551]<br /> [ 760.703554] CPU0 CPU1<br /> [ 760.703556] ---- ----<br /> [ 760.703558] lock(&amp;q-&gt;sysfs_lock);<br /> [ 760.703565] lock(kn-&gt;active#385);<br /> [ 760.703573] lock(&amp;q-&gt;sysfs_lock);<br /> [ 760.703579] lock(kn-&gt;active#385);<br /> [ 760.703587]<br /> [ 760.703587] *** DEADLOCK ***<br /> <br /> Solve this by removing the mutex_lock()/mutex_unlock() calls from<br /> blk_ia_range_sysfs_show().