Commit 36df6e3dbd7e ("cgroup: make css_rstat_updated nmi safe") used this_cpu_cmpxchg() for the lockless insertion, and therefore required both ARCH_HAVE_NMI_SAFE_CMPXCHG and ARCH_HAS_NMI_SAFE_THIS_CPU_OPS in the NMI guard: on archs without the latter, this_cpu_cmpxchg() falls back to "local_irq_save() + plain cmpxchg", and local_irq_save() cannot mask NMIs. Commit 3309b63a2281 ("cgroup: rstat: use LOCK CMPXCHG in css_rstat_updated") later replaced this_cpu_cmpxchg() with plain try_cmpxchg() to fix cross-CPU lockless-list corruption, but left the NMI guard untouched. After that switch, css_rstat_updated() no longer performs any this_cpu_*() RMW operations and only relies on the arch having NMI-safe cmpxchg, so ARCH_HAS_NMI_SAFE_THIS_CPU_OPS is no longer required in the guard. Relax the guard accordingly so that archs which have HAVE_NMI and ARCH_HAVE_NMI_SAFE_CMPXCHG but not ARCH_HAS_NMI_SAFE_THIS_CPU_OPS (e.g. sparc, powerpc on PPC64/BOOK3S) can benefit from the existing CONFIG_MEMCG_NMI_SAFETY_REQUIRES_ATOMIC path. Without this, the css is never queued in NMI on those archs, and the atomics staged by account_{slab,kmem}_nmi_safe() are not drained by flush_nmi_stats().
Unified · Split
| @@ -81,11 +81,10 @@ void __css_rstat_updated(struct cgroup_subsys_state *css, int cpu) · 80 unmodified lines | ||
| 81 | 81 | lockdep_assert_preemption_disabled(); |
| 82 | 82 | |
| 83 | 83 | /* |
| 84 | − * For archs withnot nmi safe cmpxchg or percpu ops support, ignore | |
| 85 | − * the requests from nmi context. | |
| 84 | + * The lockless insertion below relies on NMI-safe cmpxchg; | |
| 85 | + * bail out in NMI on archs that don't provide it. | |
| 86 | 86 | */ |
| 87 | − if ((!IS_ENABLED(CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG) || | |
| 88 | − !IS_ENABLED(CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS)) && in_nmi()) | |
| 87 | + if (!IS_ENABLED(CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG) && in_nmi()) | |
| 89 | 88 | return; |
| 90 | 89 | |
| 91 | 90 | rstatc = css_rstat_cpu(css, cpu); |