linux-kernel - Re: [PATCH 10/15] mm/swap: break the loop if matching device is found

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: <e8f23593-1ff7-7204-2b17-2e9901b65098@huawei.com>
Date:   Tue, 10 May 2022 10:10:47 +0800
From:   Miaohe Lin <linmiaohe@...wei.com>
To:     Andrew Morton <akpm@...ux-foundation.org>
CC:     <willy@...radead.org>, <vbabka@...e.cz>, <dhowells@...hat.com>,
        <neilb@...e.de>, <apopple@...dia.com>, <david@...hat.com>,
        <surenb@...gle.com>, <peterx@...hat.com>,
        <naoya.horiguchi@....com>, <linux-mm@...ck.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 10/15] mm/swap: break the loop if matching device is found

On 2022/5/10 5:16, Andrew Morton wrote:
> On Mon, 9 May 2022 21:14:11 +0800 Miaohe Lin <linmiaohe@...wei.com> wrote:
> 
>> We can break the loop if matching device is found to save some possible
>> cpu cycles because there should be only one matching device and there is
>> no need to continue if the matching one is already found.
>>
>> ...
>>
>> --- a/mm/swapfile.c
>> +++ b/mm/swapfile.c
>> @@ -1692,6 +1692,8 @@ int swap_type_of(dev_t device, sector_t offset)
>>  				spin_unlock(&swap_lock);
>>  				return type;
>>  			}
>> +
>> +			break;
>>  		}
>>  	}
>>  	spin_unlock(&swap_lock);
> 
> Are you sure?  If we have two S_ISREG swapfiles on the same device,
> don't they have the same sis->bdev?

Oh, I missed this use case. Sorry about it! :(

> 
> If not, why bother passing `offset' into this function at all?

Yes, you're right. 'offset' indicates the swap header location. Will drop this patch.

Thanks!

> .
>