In the Linux kernel, the following vulnerability has been resolved:[EOL][EOL]ALSA: pcm: Disable bottom softirqs as part of spin_lock_irq() on PREEMPT_RT[EOL][EOL]snd_pcm_group_lock_irq() acquires a spinlock_t and disables interrupts[EOL]via spin_lock_irq(). This also implicitly disables the handling of[EOL]softirqs such as TIMER_SOFTIRQ.[EOL]On PREEMPT_RT softirqs are preemptible and spin_lock_irq() does not[EOL]disable them. That means a timer can be invoked during spin_lock_irq()[EOL]on the same CPU. Due to synchronisations reasons local_bh_disable() has[EOL]a per-CPU lock named softirq_ctrl.lock which synchronizes individual[EOL]softirq against each other.[EOL]syz-bot managed to trigger a lockdep report where softirq_ctrl.lock is[EOL]acquired in hrtimer_cancel() in addition to hrtimer_run_softirq(). This[EOL]is a possible deadlock.[EOL][EOL]The softirq_ctrl.lock can not be made part of spin_lock_irq() as this[EOL]would lead to too much synchronisation against individual threads on the[EOL]system. To avoid the possible deadlock, softirqs must be manually[EOL]disabled before the lock is acquired.[EOL][EOL]Disable softirqs before the lock is acquired on PREEMPT_RT.
