linux-kernel - Re: [PATCH 1/2] drivers/base/node: merge register_one_node() and register

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <3ad7a1b6-eaba-4e1e-aa27-b6c9260ddc68@linux.ibm.com>
Date: Thu, 25 Sep 2025 18:50:47 +0530
From: Donet Tom <donettom@...ux.ibm.com>
To: David Hildenbrand <david@...hat.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Oscar Salvador <osalvador@...e.de>
Cc: Ritesh Harjani <ritesh.list@...il.com>,
        Aboorva Devarajan <aboorvad@...ux.ibm.com>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org,
        Madhavan Srinivasan <maddy@...ux.ibm.com>,
        linuxppc-dev@...ts.ozlabs.org,
        Christophe Leroy
 <christophe.leroy@...roup.eu>,
        Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>,
        x86@...nel.org, Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Danilo Krummrich <dakr@...nel.org>, Mike Rapoport <rppt@...nel.org>,
        Dave Jiang <dave.jiang@...el.com>
Subject: Re: [PATCH 1/2] drivers/base/node: merge register_one_node() and
 register_node() to a single function.


On 9/25/25 2:24 PM, David Hildenbrand wrote:
> On 24.09.25 20:40, Donet Tom wrote:
>> register_one_node() and register_node() are small functions.
>> This patch merges them into a single function named register_node()
>> to improve code readability.
>>
>> No functional changes are introduced.
>>
>> Signed-off-by: Donet Tom <donettom@...ux.ibm.com>
>> ---
>
> [...]
>
>>   /**
>>    * unregister_node - unregister a node device
>>    * @node: node going away
>> @@ -869,7 +842,13 @@ void 
>> register_memory_blocks_under_node_hotplug(int nid, unsigned long 
>> start_pfn,
>>   }
>>   #endif /* CONFIG_MEMORY_HOTPLUG */
>>   -int register_one_node(int nid)
>> +/*
>
> We can directly convert this to proper kernel doc by using /**
>
Sure I will add it.


>
>> + * register_node - Setup a sysfs device for a node.
>> + * @nid - Node number to use when creating the device.
>> + *
>> + * Initialize and register the node device.
>
> and briefly describing what the return value means
>
> "Returns 0 on success, ..."
>
Sure

>> + */
>> +int register_node(int nid)
>>   {
>>       int error;
>>       int cpu;
>> @@ -880,14 +859,23 @@ int register_one_node(int nid)
>>           return -ENOMEM;
>>         INIT_LIST_HEAD(&node->access_list);
>> -    node_devices[nid] = node;
>>   -    error = register_node(node_devices[nid], nid);
>> +    node->dev.id = nid;
>> +    node->dev.bus = &node_subsys;
>> +    node->dev.release = node_device_release;
>> +    node->dev.groups = node_dev_groups;
>> +
>> +    error = device_register(&node->dev);
>>       if (error) {
>> -        node_devices[nid] = NULL;
>
> Wondering why we did have this temporary setting of the node_devices[] 
> in there. But I cannot immediately spot why it was required.


node_devices[] is used in many places to access node data.

In the previous code, immediately after allocating the node

structure, it was stored in node_devices[]. On error paths, we

were clearing the node structure entry from node_devices[].


With the new code, the node structure is now stored in node_devices[]

only after the device has been registered, and just before calling

register_cpu_under_node(), since node_devices[] is accessed inside

that function.


It is also used outside of node.c, in hugetlb_register_all_nodes() .

Do you think we could use a different mechanism instead of this?