hrtimers: Handle CPU state correctly on hotplug
commit 2f8dea1692eef2b7ba6a256246ed82c365fdc686 upstream.
Consider a scenario where a CPU transitions from CPUHP_ONLINE to halfway
through a CPU hotunplug down to CPUHP_HRTIMERS_PREPARE, and then back to
CPUHP_ONLINE:
Since hrtimers_prepare_cpu() does not run, cpu_base.hres_active remains set
to 1 throughout. However, during a CPU unplug operation, the tick and the
clockevents are shut down at CPUHP_AP_TICK_DYING. On return to the online
state, for instance CFS incorrectly assumes that the hrtick is already
active, and the chance of the clockevent device to transition to oneshot
mode is also lost forever for the CPU, unless it goes back to a lower state
than CPUHP_HRTIMERS_PREPARE once.
This round-trip reveals another issue; cpu_base.online is not set to 1
after the transition, which appears as a WARN_ON_ONCE in enqueue_hrtimer().
Aside of that, the bulk of the per CPU state is not reset either, which
means there are dangling pointers in the worst case.
Address this by adding a corresponding startup() callback, which resets the
stale per CPU state and sets the online flag.
[ tglx: Make the new callback unconditionally available, remove the online
modification in the prepare() callback and clear the remaining
state in the starting callback instead of the prepare callback ]
BUG=b/396831046
TEST=presubmit
RELEASE_NOTE=Fixed CVE-2024-57951 in the Linux kernel.
cos-patch: security-high
Fixes: 5c0930ccaad5 ("hrtimers: Push pending hrtimers away from outgoing CPU earlier")
Change-Id: I2865d81a1c0266427c82805f2111854bb333e782
Signed-off-by: Koichiro Den <koichiro.den@canonical.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/20241220134421.3809834-1-koichiro.den@canonical.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Kernel CVE Triage Automation <cloud-image-kernel-cve-triage-automation@prod.google.com>
Reviewed-on: https://cos-review.googlesource.com/c/third_party/kernel/+/93239
Reviewed-by: Michael Kochera <kochera@google.com>
Reviewed-by: Shuo Yang <gshuoy@google.com>
Tested-by: Cusky Presubmit Bot <presubmit@cos-infra-prod.iam.gserviceaccount.com>
diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
index 254d4a8..8f77bb0 100644
--- a/include/linux/hrtimer.h
+++ b/include/linux/hrtimer.h
@@ -532,6 +532,7 @@ extern void __init hrtimers_init(void);
extern void sysrq_timer_list_show(void);
int hrtimers_prepare_cpu(unsigned int cpu);
+int hrtimers_cpu_starting(unsigned int cpu);
#ifdef CONFIG_HOTPLUG_CPU
int hrtimers_cpu_dying(unsigned int cpu);
#else
diff --git a/kernel/cpu.c b/kernel/cpu.c
index acf16e3..eac96f4 100644
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -1766,7 +1766,7 @@ static struct cpuhp_step cpuhp_hp_states[] = {
},
[CPUHP_AP_HRTIMERS_DYING] = {
.name = "hrtimers:dying",
- .startup.single = NULL,
+ .startup.single = hrtimers_cpu_starting,
.teardown.single = hrtimers_cpu_dying,
},
diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c
index bdd9041..9e91f69 100644
--- a/kernel/time/hrtimer.c
+++ b/kernel/time/hrtimer.c
@@ -2176,6 +2176,15 @@ int hrtimers_prepare_cpu(unsigned int cpu)
}
cpu_base->cpu = cpu;
+ hrtimer_cpu_base_init_expiry_lock(cpu_base);
+ return 0;
+}
+
+int hrtimers_cpu_starting(unsigned int cpu)
+{
+ struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases);
+
+ /* Clear out any left over state from a CPU down operation */
cpu_base->active_bases = 0;
cpu_base->hres_active = 0;
cpu_base->hang_detected = 0;
@@ -2184,7 +2193,6 @@ int hrtimers_prepare_cpu(unsigned int cpu)
cpu_base->expires_next = KTIME_MAX;
cpu_base->softirq_expires_next = KTIME_MAX;
cpu_base->online = 1;
- hrtimer_cpu_base_init_expiry_lock(cpu_base);
return 0;
}
@@ -2262,6 +2270,7 @@ int hrtimers_cpu_dying(unsigned int dying_cpu)
void __init hrtimers_init(void)
{
hrtimers_prepare_cpu(smp_processor_id());
+ hrtimers_cpu_starting(smp_processor_id());
open_softirq(HRTIMER_SOFTIRQ, hrtimer_run_softirq);
}
