netdev - Re: [PATCH net-next] rds: transport module should be auto loaded when transport is set

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]

Message-ID: <0237090f-ba68-f689-782f-3993b89fae15@oracle.com>
Date:   Tue, 2 Jun 2020 11:19:39 -0700
From:   Rao Shoaib <rao.shoaib@...cle.com>
To:     Leon Romanovsky <leon@...nel.org>
Cc:     David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
        linux-rdma@...r.kernel.org, somasundaram.krishnasamy@...cle.com
Subject: Re: [PATCH net-next] rds: transport module should be auto loaded when
 transport is set


On 6/1/20 11:17 PM, Leon Romanovsky wrote:
> On Mon, Jun 01, 2020 at 09:59:30AM -0700, Rao Shoaib wrote:
>> On 5/29/20 4:41 PM, David Miller wrote:
>>> From: rao.shoaib@...cle.com
>>> Date: Wed, 27 May 2020 01:17:42 -0700
>>>
>>>> diff --git a/include/uapi/linux/rds.h b/include/uapi/linux/rds.h
>>>> index cba368e55863..7273c681e6c1 100644
>>>> --- a/include/uapi/linux/rds.h
>>>> +++ b/include/uapi/linux/rds.h
>>>> @@ -64,7 +64,7 @@
>>>>    /* supported values for SO_RDS_TRANSPORT */
>>>>    #define	RDS_TRANS_IB	0
>>>> -#define	RDS_TRANS_IWARP	1
>>>> +#define	RDS_TRANS_GAP	1
>>>>    #define	RDS_TRANS_TCP	2
>>>>    #define RDS_TRANS_COUNT	3
>>>>    #define	RDS_TRANS_NONE	(~0)
>>> You can't break user facing UAPI like this, sorry.
>> I was hoping that this could be considered an exception as IWARP has been
>> deprecated for almost a decade and there is no current product using it.
>> With the change any old binary will continue to work, a new compilation fill
>> fail so that the code can be examined, otherwise we will never be able to
>> reuse this number.
>>
>> If the above is not acceptable I can revert this part of the change.
> Nothing prohibits you from adding the following lines:
>
> + /* don't use RDS_TRANS_IWARP - it is deprecated */
> + #define  RDS_TRANS_GAP RDS_TRANS_IWARP

Correct. That is what I was planning on doing in case I could not get 
rid of RDS_TRANS_IWARP. I will resubmit the patch.

Shoaib

>
>> Shoaib
>>