On Fri, 2023-11-24 at 17:46 +0000, Greg Kroah-Hartman wrote:
6.5-stable review patch. If anyone has any objections, please let me know.
From: Juergen Gross jgross@suse.com
[ Upstream commit e64e7c74b99ec9e439abca75f522f4b98f220bd1 ]
xen_send_IPI_one() is being used by cpuhp_report_idle_dead() after it calls rcu_report_dead(), meaning that any RCU usage by xen_send_IPI_one() is a bad idea.
Unfortunately xen_send_IPI_one() is using notify_remote_via_irq() today, which is using irq_get_chip_data() via info_for_irq(). And irq_get_chip_data() in turn is using a maple-tree lookup requiring RCU.
Avoid this problem by caching the ipi event channels in another percpu variable, allowing the use notify_remote_via_evtchn() in xen_send_IPI_one().
IIRC, the result of this bug is a mostly harmless lockdep splat.
On the other hand, the result of *fixing* this bug is that lockdep doesn't whine, and lockdep keeps working. Only to later trigger a warning really early during cpu hotplug, causing the whole machine to go down with a triple-fault.
Can we have the backport of https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?... *first* and then this one?