CVE-2026-31734sched_ext: Fix is_bpf_migration_disabled() false negative on non-PREEMPT_RCU

Official description

In the Linux kernel, the following vulnerability has been resolved: sched_ext: Fix is_bpf_migration_disabled() false negative on non-PREEMPT_RCU Since commit 8e4f0b1ebcf2 ("bpf: use rcu_read_lock_dont_migrate() for trampoline.c"), the BPF prolog (__bpf_prog_enter) calls migrate_disable() only when CONFIG_PREEMPT_RCU is enabled, via rcu_read_lock_dont_migrate(). Without CONFIG_PREEMPT_RCU, the prolog never touches migration_disabled, so migration_disabled == 1 always means the task is truly migration-disabled regardless of whether it is the current task. The old unconditional p == current check was a false negative in this case, potentially allowing a migration-disabled task to be dispatched to a remote CPU and triggering scx_error in task_can_run_on_remote_rq(). Only apply the p == current disambiguation when CONFIG_PREEMPT_RCU is enabled, where the ambiguity with the BPF prolog still exists.

Affected packages & fixed versions

Affected products (as reported)

References

• https://git.kernel.org/stable/c/72c43eb2e334febe93018cfb68ae828f55c6e49e
• https://git.kernel.org/stable/c/b4992a9446bb9a639007bfd32bf5c5a7e30199e5
• https://git.kernel.org/stable/c/0c4a59df370bea245695c00aaae6ae75747139bd