lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Mon, 16 May 2022 09:29:10 +0800
From:   "Leizhen (ThunderTown)" <thunder.leizhen@...wei.com>
To:     <paulmck@...nel.org>
CC:     Frederic Weisbecker <frederic@...nel.org>,
        Neeraj Upadhyay <quic_neeraju@...cinc.com>,
        Josh Triplett <josh@...htriplett.org>,
        "Steven Rostedt" <rostedt@...dmis.org>,
        Mathieu Desnoyers <mathieu.desnoyers@...icios.com>,
        Lai Jiangshan <jiangshanlai@...il.com>,
        Joel Fernandes <joel@...lfernandes.org>, <rcu@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] rcu/nocb: Delete local variable 'need_rcu_nocb_mask' in
 rcu_init_nohz()



On 2022/5/13 23:05, Paul E. McKenney wrote:
> On Tue, Apr 26, 2022 at 03:36:26PM +0800, Zhen Lei wrote:
>> The local variable 'need_rcu_nocb_mask' is true only if CONFIG_NO_HZ_FULL
>> is defined. So branch "if (need_rcu_nocb_mask)" can be moved within the
>> scope of "#if defined(CONFIG_NO_HZ_FULL)". At this point, using variable
>> 'need_rcu_nocb_mask' is not necessary, so delete it.
>>
>> No functional changes, but the code looks a little more concise.
>>
>> Signed-off-by: Zhen Lei <thunder.leizhen@...wei.com>
> 
> First, please accept my apologies for the late reply and for the
> overly active spam filters.
> 
> One question below.
> 
> 							Thanx, Paul
> 
>> ---
>>  kernel/rcu/tree_nocb.h | 8 ++------
>>  1 file changed, 2 insertions(+), 6 deletions(-)
>>
>> diff --git a/kernel/rcu/tree_nocb.h b/kernel/rcu/tree_nocb.h
>> index 636d0546a4e932e..1e334e217f0afb7 100644
>> --- a/kernel/rcu/tree_nocb.h
>> +++ b/kernel/rcu/tree_nocb.h
>> @@ -1165,15 +1165,10 @@ EXPORT_SYMBOL_GPL(rcu_nocb_cpu_offload);
>>  void __init rcu_init_nohz(void)
>>  {
>>  	int cpu;
>> -	bool need_rcu_nocb_mask = false;
>>  	struct rcu_data *rdp;
>>  
>>  #if defined(CONFIG_NO_HZ_FULL)
>> -	if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask))
>> -		need_rcu_nocb_mask = true;
>> -#endif /* #if defined(CONFIG_NO_HZ_FULL) */
>> -
>> -	if (need_rcu_nocb_mask) {
> 
> Could you please test this on a kernel built with CONFIG_NO_HZ_FULL=n
> and CONFIG_RCU_NOCB_CPU=y?  If that works, please add an explanation

OK, I will test it.

> of why it works to the commit log above and repost the patch.

Sorry, If I had written pseudocode in the commit log, it would have been clear, as below:

-----------------before-----------------
        bool need_rcu_nocb_mask = false;

#if defined(CONFIG_NO_HZ_FULL)
        if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask))
                need_rcu_nocb_mask = true;      <----	//Move it here, then delete all 'need_rcu_nocb_mask' related code
#endif /* #if defined(CONFIG_NO_HZ_FULL) */          |
                                                     |
        if (need_rcu_nocb_mask) {                    |  //Can only be true above
                ...                             -----
        }

-----------------after-----------------
#if defined(CONFIG_NO_HZ_FULL)
        if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask)) {
                ...
	}
#endif /* #if defined(CONFIG_NO_HZ_FULL) */

I'm doing clean up on the premise that the original code is correct.


> 
>> +	if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask)) {
>>  		if (!cpumask_available(rcu_nocb_mask)) {
>>  			if (!zalloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL)) {
>>  				pr_info("rcu_nocb_mask allocation failed, callback offloading disabled.\n");
>> @@ -1182,6 +1177,7 @@ void __init rcu_init_nohz(void)
>>  		}
>>  		rcu_nocb_is_setup = true;
>>  	}
>> +#endif /* #if defined(CONFIG_NO_HZ_FULL) */
>>  
>>  	if (!rcu_nocb_is_setup)
>>  		return;
>> -- 
>> 2.26.0.106.g9fadedd
>>
> .
> 

-- 
Regards,
  Zhen Lei

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ