Re: [PATCH v2] mips: mm: Call rcutree_report_cpu_starting() even earlier

[Maciej W. Rozycki]

On Thu, 9 Apr 2026, Stefan Wiehler wrote:

> rcutree_report_cpu_starting() must be called on secondary CPUs before
> allocating memory to avoid the following Lockdep-RCU splat when
> CONFIG_PROVE_RCU_LIST=y:
> 
>   WARNING: suspicious RCU usage
>   6.6.119-00d46e15c416-fct #1 Not tainted
>   -----------------------------
>   /kernel/locking/lockdep.c:3762 RCU-list traversed in non-reader section!!

 I have tried to verify your change here, but the warning does not trigger 
with 6.19 as released and CONFIG_SIBYTE_SWARM as the platform (2-way SMP).  

$ grep CONFIG_PROVE_RCU_LIST .config
CONFIG_PROVE_RCU_LIST=y
$ 

Am I missing anything here, anything extra to enable?

> See also commit 55702ec9603e ("mips/smp: Call
> rcutree_report_cpu_starting() earlier").
> 
> Fixes: 231ac951faba ("MIPS: mm: kmalloc tlb_vpn array to avoid stack overflow")

 This does not appear to be a valid commit hash upstream; this is commit 
841ecc979b18 AFAICS.

> diff --git a/arch/mips/kernel/smp.c b/arch/mips/kernel/smp.c
> index 4868e79f3b30..bdb47c70d4f5 100644
> --- a/arch/mips/kernel/smp.c
> +++ b/arch/mips/kernel/smp.c
> @@ -359,8 +359,8 @@ asmlinkage void start_secondary(void)
>  	unsigned int cpu = raw_smp_processor_id();
>  
>  	cpu_probe();
> -	per_cpu_trap_init(false);
>  	rcutree_report_cpu_starting(cpu);
> +	per_cpu_trap_init(false);
>  	mips_clockevent_init();
>  	mp_ops->init_secondary();
>  	cpu_report();

 As I noted in my previous message: is there any reason for cpu_probe() to 
precede this call?

  Maciej