mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Leizhen (ThunderTown)" <thunder.leizhen@huawei.com>
To: Thomas Gleixner <tglx@linutronix.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/5] debugobjects: Fix the misuse of global variables in fill_pool()
Date: Tue, 3 Sep 2024 15:00:11 +0800	[thread overview]
Message-ID: <659f0321-e567-ad48-4545-4a47a158d6c2@huawei.com> (raw)
In-Reply-To: <3bb35c94-dd54-33d4-b7ac-64f0d2b77c07@huawei.com>



On 2024/9/3 11:22, Leizhen (ThunderTown) wrote:
> 
> 
> On 2024/9/3 10:16, Leizhen (ThunderTown) wrote:
>>
>>
>> On 2024/9/3 0:22, Thomas Gleixner wrote:
>>> On Mon, Sep 02 2024 at 22:05, Zhen Lei wrote:
>>>
>>>> The global variable 'obj_pool_min_free' records the lowest historical
>>>> value of the number of nodes in the global list 'obj_pool', instead of
>>>> being used as the lowest threshold value. This may be a mistake and
>>>
>>> Maybe? It's either a bug or not.
>>
>> Yes, it's a bug, but I'm just learning this module, so I'm not confident.
>>
>>>
>>>> should be replaced with variable 'debug_objects_pool_min_level'.
>>>
>>> And if it's a bug then it has to be replaced.
>>
>> OK, I will update the commit message in V2.
>>
>>>
>>> This misses another minor issue:
>>>
>>> static int			obj_pool_min_free = ODEBUG_POOL_SIZE;
>>>
>>> static int  __data_racy		debug_objects_pool_min_level __read_mostly
>>> 				= ODEBUG_POOL_MIN_LEVEL;
>>>
>>> As debug_objects_pool_min_level is the minimum level to keep around and
>>> obj_pool_min_free is a statistics mechanism, __data_racy is misplaced
>>> too. The variables should swap their position, because
>>> debug_objects_pool_min_level is functional, but obj_pool_min_free is
>>> pure stats.
> 
> This principle covers a large number of variables. It should be sorted
> by the variable name before. It's better not to change the position this time.
> 
>>>
>>> Also debug_objects_pool_min_level and debug_objects_pool_size should
>>> be __ro_after_init.
>>
>> OK, How about modifying it like this? I further removed the __data_racy from
>> debug_objects_pool_min_level and debug_objects_pool_size, because __ro_after_init.
>>
>>
>> diff --git a/lib/debugobjects.c b/lib/debugobjects.c
>> index d3845705db955fa..816d3d968cd9f14 100644
>> --- a/lib/debugobjects.c
>> +++ b/lib/debugobjects.c
>> @@ -70,7 +70,8 @@ static HLIST_HEAD(obj_to_free);
>>   * made at debug_stats_show(). Both obj_pool_min_free and obj_pool_max_used
>>   * can be off.
>>   */
>> -static int                     obj_pool_min_free = ODEBUG_POOL_SIZE;
>> +static int __ro_after_init     debug_objects_pool_size = ODEBUG_POOL_SIZE;
>> +static int __ro_after_init     debug_objects_pool_min_level = ODEBUG_POOL_MIN_LEVEL;
>>  static int                     obj_pool_free = ODEBUG_POOL_SIZE;
>>  static int                     obj_pool_used;
>>  static int                     obj_pool_max_used;
>> @@ -84,10 +85,7 @@ static int __data_racy                       debug_objects_fixups __read_mostly;
>>  static int __data_racy                 debug_objects_warnings __read_mostly;
>>  static int __data_racy                 debug_objects_enabled __read_mostly
>>                                         = CONFIG_DEBUG_OBJECTS_ENABLE_DEFAULT;
>> -static int __data_racy                 debug_objects_pool_size __read_mostly
>> -                                       = ODEBUG_POOL_SIZE;
>> -static int __data_racy                 debug_objects_pool_min_level __read_mostly
>> -                                       = ODEBUG_POOL_MIN_LEVEL;
>> +static int __data_racy                 obj_pool_min_free = ODEBUG_POOL_SIZE;

Sorry, I rechecked it again. After this patch, obj_pool_min_free is referenced in the
same way as obj_pool_max_used. The only race point is located in debug_stats_show().
However, this reference point does not need to be included in the race analysis. So
there is no need to add __data_racy for obj_pool_min_free.

The final modification is as follows:

diff --git a/lib/debugobjects.c b/lib/debugobjects.c
index d3845705db955fa..721e207f85b6be7 100644
--- a/lib/debugobjects.c
+++ b/lib/debugobjects.c
@@ -84,9 +84,9 @@ static int __data_racy                        debug_objects_fixups __read_mostly;
 static int __data_racy                 debug_objects_warnings __read_mostly;
 static int __data_racy                 debug_objects_enabled __read_mostly
                                        = CONFIG_DEBUG_OBJECTS_ENABLE_DEFAULT;
-static int __data_racy                 debug_objects_pool_size __read_mostly
+static int                             debug_objects_pool_size __ro_after_init
                                        = ODEBUG_POOL_SIZE;
-static int __data_racy                 debug_objects_pool_min_level __read_mostly
+static int                             debug_objects_pool_min_level __ro_after_init
                                        = ODEBUG_POOL_MIN_LEVEL;

 static const struct debug_obj_descr *descr_test  __read_mostly;


>>
>>  static const struct debug_obj_descr *descr_test  __read_mostly;
>>  static struct kmem_cache       *obj_cache __ro_after_init;
>>
>>
>>>
>>>> Fixes: d26bf5056fc0 ("debugobjects: Reduce number of pool_lock acquisitions in fill_pool()")
>>>> Fixes: 36c4ead6f6df ("debugobjects: Add global free list and the counter")
>>>
>>> Nice detective work!
>>>
>>> Thanks,
>>>
>>>         tglx
>>> .
>>>
>>
> 

-- 
Regards,
  Zhen Lei

  reply	other threads:[~2024-09-03  7:00 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-02 14:05 [PATCH 0/5] debugobjects: Do some minor optimizations, fixes and cleaups Zhen Lei
2024-09-02 14:05 ` [PATCH 1/5] debugobjects: Fix the misuse of global variables in fill_pool() Zhen Lei
2024-09-02 16:22   ` Thomas Gleixner
2024-09-03  2:16     ` Leizhen (ThunderTown)
2024-09-03  3:22       ` Leizhen (ThunderTown)
2024-09-03  7:00         ` Leizhen (ThunderTown) [this message]
2024-09-03  9:37           ` Thomas Gleixner
2024-09-03 11:14             ` Leizhen (ThunderTown)
2024-09-03 11:43               ` Thomas Gleixner
2024-09-03 12:22                 ` Leizhen (ThunderTown)
2024-09-02 14:05 ` [PATCH 2/5] debugobjects: Remove redundant checks " Zhen Lei
2024-09-03  9:44   ` Thomas Gleixner
2024-09-03 11:23     ` Leizhen (ThunderTown)
2024-09-02 14:05 ` [PATCH 3/5] debugobjects: Don't start fill if there are remaining nodes locally Zhen Lei
2024-09-03  9:52   ` Thomas Gleixner
2024-09-03 12:06     ` Leizhen (ThunderTown)
2024-09-02 14:05 ` [PATCH 4/5] debugobjects: Use hlist_splice_init() to reduce lock conflicts Zhen Lei
2024-09-03 10:09   ` Thomas Gleixner
2024-09-03 12:14     ` Leizhen (ThunderTown)
2024-09-02 14:05 ` [PATCH 5/5] debugobjects: Delete a piece of redundant code Zhen Lei
2024-09-03 10:14   ` Thomas Gleixner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=659f0321-e567-ad48-4545-4a47a158d6c2@huawei.com \
    --to=thunder.leizhen@huawei.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=tglx@linutronix.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®