Описание
In the Linux kernel, the following vulnerability has been resolved:
net: mvneta: re-enable percpu interrupt on resume
On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules NAPI poll, which calls enable_percpu_irq() on completion to unmask.
If suspend occurs while NAPI poll is pending (between disable_percpu_irq in the ISR and enable_percpu_irq in poll completion), the interrupt is never re-enabled:
- mvneta_percpu_isr: disable_percpu_irq() + napi_schedule() => MPIC masked, percpu_enabled cpumask bit cleared
- NAPI poll does not complete before suspend proceeds (on PREEMPT_RT this is highly likely since softirqs run in ksoftirqd which gets frozen; on non-RT it can happen when softirq processing is deferred to ksoftirqd)
- mvneta_stop_dev => napi_disable(): cancels the pending poll without execut...
In the Linux kernel, the following vulnerability has been resolved:
net: mvneta: re-enable percpu interrupt on resume
On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules NAPI poll, which calls enable_percpu_irq() on completion to unmask.
If suspend occurs while NAPI poll is pending (between disable_percpu_irq in the ISR and enable_percpu_irq in poll completion), the interrupt is never re-enabled:
- mvneta_percpu_isr: disable_percpu_irq() + napi_schedule() => MPIC masked, percpu_enabled cpumask bit cleared
- NAPI poll does not complete before suspend proceeds (on PREEMPT_RT this is highly likely since softirqs run in ksoftirqd which gets frozen; on non-RT it can happen when softirq processing is deferred to ksoftirqd)
- mvneta_stop_dev => napi_disable(): cancels the pending poll without executing the completion path
- suspend_device_irqs => IRQCHIP_MASK_ON_SUSPEND: masks MPIC (already masked, but records IRQS_SUSPENDED)
- Resume: mpic_resume checks irq_percpu_is_enabled() => false (bit was cleared in step 1) => skips unmask
- mvneta_start_dev only restores device-level INTR_NEW_MASK, does not touch the MPIC per-CPU mask
Result: MPIC per-CPU interrupt stays masked permanently. The NIC generates interrupts (INTR_NEW_CAUSE != 0) but the CPU never receives them, causing complete loss of network connectivity.
Fix by calling on_each_cpu(mvneta_percpu_enable) in the resume path to unconditionally unmask the MPIC per-CPU interrupt regardless of pre-suspend state.
Ссылки
- https://nvd.nist.gov/vuln/detail/CVE-2026-72409
- https://git.kernel.org/stable/c/1cc312dc8bc78fa24c80d5bc193dbf5b57a99cc6
- https://git.kernel.org/stable/c/5bdb33ff6e58bdc43632e98b30723eb65352d671
- https://git.kernel.org/stable/c/82c13027ed283b856017adee970dbfdffce5c6b8
- https://git.kernel.org/stable/c/8c7a489aa71d2693752b2e794a68bf672d16c829
- https://git.kernel.org/stable/c/b84dd48f9da1eb132bdc06a944423cd5a1641ef1
- https://git.kernel.org/stable/c/be626ac1faadd49c2cead9f9cd06ba8752d81563
- https://git.kernel.org/stable/c/bf88cd3b649bc3e638f1e8a77649581852747a68
- https://git.kernel.org/stable/c/fd398d6480987e4c84fff0aaab6b9d6642a93343
Связанные уязвимости
In the Linux kernel, the following vulnerability has been resolved: net: mvneta: re-enable percpu interrupt on resume On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules NAPI poll, which calls enable_percpu_irq() on completion to unmask. If suspend occurs while NAPI poll is pending (between disable_percpu_irq in the ISR and enable_percpu_irq in poll completion), the interrupt is never re-enabled: 1. mvneta_percpu_isr: disable_percpu_irq() + napi_schedule() => MPIC masked, percpu_enabled cpumask bit cleared 2. NAPI poll does not complete before suspend proceeds (on PREEMPT_RT this is highly likely since softirqs run in ksoftirqd which gets frozen; on non-RT it can happen when softirq processing is deferred to ksoftirqd) 3. mvneta_stop_dev => napi_disable(): cancels the pending poll without executing the completion path 4. suspend_...
In the Linux kernel, the following vulnerability has been resolved: net: mvneta: re-enable percpu interrupt on resume On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules NAPI poll, which calls enable_percpu_irq() on completion to unmask. If suspend occurs while NAPI poll is pending (between disable_percpu_irq in the ISR and enable_percpu_irq in poll completion), the interrupt is never re-enabled: 1. mvneta_percpu_isr: disable_percpu_irq() + napi_schedule() => MPIC masked, percpu_enabled cpumask bit cleared 2. NAPI poll does not complete before suspend proceeds (on PREEMPT_RT this is highly likely since softirqs run in ksoftirqd which gets frozen; on non-RT it can happen when softirq processing is deferred to ksoftirqd) 3. mvneta_stop_dev => napi_disable(): cancels the pending poll without executing
In the Linux kernel, the following vulnerability has been resolved: n ...