netdev - Re: [PATCH 29/34] net: smc: fix opencoded find_and_set_bit() in smc_wr_tx_get_free_slot

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]

Date: Tue, 21 Nov 2023 16:39:40 +0100
From: Alexandra Winter <wintera@...ux.ibm.com>
To: Yury Norov <yury.norov@...il.com>
Cc: linux-kernel@...r.kernel.org, Karsten Graul <kgraul@...ux.ibm.com>,
        Wenjia Zhang <wenjia@...ux.ibm.com>, Jan Karcher <jaka@...ux.ibm.com>,
        "D. Wythe" <alibuda@...ux.alibaba.com>,
        Tony Lu <tonylu@...ux.alibaba.com>, Wen Gu <guwen@...ux.alibaba.com>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>, linux-s390@...r.kernel.org,
        netdev@...r.kernel.org, Jan Kara <jack@...e.cz>,
        Mirsad Todorovac <mirsad.todorovac@....unizg.hr>,
        Matthew Wilcox <willy@...radead.org>,
        Rasmus Villemoes <linux@...musvillemoes.dk>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Maxim Kuvyrkov <maxim.kuvyrkov@...aro.org>,
        Alexey Klimov <klimov.linux@...il.com>
Subject: Re: [PATCH 29/34] net: smc: fix opencoded find_and_set_bit() in
 smc_wr_tx_get_free_slot_index()



On 21.11.23 14:41, Yury Norov wrote:
> On Mon, Nov 20, 2023 at 09:43:54AM +0100, Alexandra Winter wrote:
>>
>>
>> On 18.11.23 16:51, Yury Norov wrote:
>>> The function opencodes find_and_set_bit() with a for_each() loop. Fix
>>> it, and make the whole function a simple almost one-liner.
>>>
>>> Signed-off-by: Yury Norov <yury.norov@...il.com>
>>> ---
>>>  net/smc/smc_wr.c | 10 +++-------
>>>  1 file changed, 3 insertions(+), 7 deletions(-)
>>>
>>> diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c
>>> index 0021065a600a..b6f0cfc52788 100644
>>> --- a/net/smc/smc_wr.c
>>> +++ b/net/smc/smc_wr.c
>>> @@ -170,15 +170,11 @@ void smc_wr_tx_cq_handler(struct ib_cq *ib_cq, void *cq_context)
>>>  
>>>  static inline int smc_wr_tx_get_free_slot_index(struct smc_link *link, u32 *idx)
>>>  {
>>> -	*idx = link->wr_tx_cnt;
>>>  	if (!smc_link_sendable(link))
>>>  		return -ENOLINK;
>>> -	for_each_clear_bit(*idx, link->wr_tx_mask, link->wr_tx_cnt) {
>>> -		if (!test_and_set_bit(*idx, link->wr_tx_mask))
>>> -			return 0;
>>> -	}
>>> -	*idx = link->wr_tx_cnt;
>>> -	return -EBUSY;
>>> +
>>> +	*idx = find_and_set_bit(link->wr_tx_mask, link->wr_tx_cnt);
>>> +	return *idx < link->wr_tx_cnt ? 0 : -EBUSY;
>>>  }
>>>  
>>>  /**
>>
>>
>> My understanding is that you can omit the lines with
>>> -	*idx = link->wr_tx_cnt;
>> because they only apply to the error paths and you checked that the calling function
>> does not use the idx variable in the error cases. Do I understand this correct?
>>
>> If so the removal of these 2 lines is not related to your change of using find_and_set_bit(),
>> do I understand that correctly?
>>
>> If so, it may be worth mentioning that in the commit message.
> 
> I'll add:
> 
>         If find_and_set_bit() doesn't acquire a bit, it returns
>         ->wr_tx_cnt, and so explicit initialization of *idx with
>         the same value is unneeded.
> 
> Makes sense?
> 

Makes sense for the -EBUSY case, thank you. 
It does not explain that you also removed the line for the -ENOLINK case 
(which is ok, because the caller has also initialized it to link->wr_tx_cnt)