[SRU][R/N][PATCH 0/2] [UBUNTU 24.04] kernel: CPU hotplug unsupported by CPUMF (LP: #2165732)

[Massimiliano Pellizzer]

BugLink: https://bugs.launchpad.net/bugs/2165732

[Impact]

s390/cpum_cf: Handle CPU hotplug via prepare/dead callbacks
The command 'perf stat -e cycles -- <command>' crashes the kernel
when CPUs are hotplug added during that run.

Root cause is the allocation of struct cpu_cf_events at first
event initialization. The allocation is dynamic and the first
event that has task context creates such a structure for
each online CPU. This is not sufficient. CPUs may be offline
during event creation and can be set online during the
perf run time. For example commands

 # echo 0 > /sys/devices/system/cpu/cpu1/online
 # perf stat -e cycles -i -- stress-ng -t10s --matrix X
 # sleep 1
 # echo 1 > /sys/devices/system/cpu/cpu1/online

create an event for CPUs 0,2-X. Since the events are created with
task-context, the scheduler will eventually schedule the program
on CPU1. This CPU has not created and initialized any per
CPU event infrastructure as that CPU was not online at the time
of the perf invocation. Thus when the scheduler runs stress-ng
on CPU1, the function cpumf_pmu_add() refers to a NULL pointer:

 struct cpu_cf_events *cpuhw = this_cpu_cfhw();

This function call is invoked after the task stress-ng has been
made runnable on CPU1. And this_cpu_cfhw() returns NULL.

The result is a panic.

The issue arises only in per-task context when the CPUMF facility is
used and the scheduler picks a random CPU for such a process to run on.
The scheduler enables the CPUMF infrastructure via PMU callback
functions pmu::add() and pmu::del().

Introduce a CPU hotplug prepare/dead callback pair which creates and
removes the per CPU counter data while the CPU is offline. Count the
users which track every CPU (cpu == -1), that is perf_event_open()
events with task context and /dev/hwctr device sessions, in the new
counter cpu_cf_root::tskcnt, protected by pmc_reserve_mutex.
This ensures the infrastructure is available when
new CPU is selected to run the per-task context process.

In cpum_cf_free_root() and cpum_cf_free_cpu() ensure the reference
pointer to data structures is set to NULL before the data is freed
to prevent interrupt handlers to access stale data.

[Fix]

Backport commit:
- 337bd95507a1 ("s390/cpum_cf: Handle CPU hotplug via prepare/dead callbacks")

[Test Plan]

Run the following commands:
# echo 0 > /sys/devices/system/cpu/cpu1/online
# perf stat -e cycles -i -- stress-ng -t10s --matrix X
# sleep 1
# echo 1 > /sys/devices/system/cpu/cpu1/online

[Regression Potential]

The change is confined to arch/s390/kernel/perf_cpum_cf.c,
affecting only s390 systems using the CPU-Measurement Counter Facility PMU.
It reworks locking and reference counting and adds two new CPU hotplug callbacks
that now run on every hotplug transition on affected systems.
A bookkeeping mismatch in the new tskctx counter or a lock-ordering error
could reintroduce use-after-free or refcount bugs similar to the one being
fixed.