netdev - Re: [net-next][PATCH 11/13] RDS: IB: add Fastreg MR (FRMR) detection support

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]

Date:	Mon, 22 Feb 2016 08:38:21 -0800
From:	santosh shilimkar <santosh.shilimkar@...cle.com>
To:	Bart Van Assche <bart.vanassche@...disk.com>,
	David Miller <davem@...emloft.net>
Cc:	netdev@...r.kernel.org, linux-rdma@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [net-next][PATCH 11/13] RDS: IB: add Fastreg MR (FRMR) detection
 support

On 2/22/2016 7:38 AM, Bart Van Assche wrote:
> On 02/21/16 19:36, David Miller wrote:
>> From: Santosh Shilimkar <santosh.shilimkar@...cle.com>
>> Date: Sat, 20 Feb 2016 03:30:02 -0800
>>
>>> @@ -54,6 +55,8 @@ module_param(rds_ib_mr_8k_pool_size, int, 0444);
>>>   MODULE_PARM_DESC(rds_ib_mr_8k_pool_size, " Max number of 8K mr per
>>> HCA");
>>>   module_param(rds_ib_retry_count, int, 0444);
>>>   MODULE_PARM_DESC(rds_ib_retry_count, " Number of hw retries before
>>> reporting an error");
>>> +module_param(prefer_frmr, bool, 0444);
>>> +MODULE_PARM_DESC(prefer_frmr, "Preferred MR method if both FMR and
>>> FRMR supported");
>>
>> Sorry, you're going to have to create a real run time method to configure
>> this parameter.
>>
>> I'm strongly against module parameters.
>>
>> Please don't go into details about why this might be difficult to do,
>> I'm totally not interested.  Doing things properly is sometimes not
>> easy, that's life.
>
Sure Dave. Will drop the parameter. The runtime detection is already
in place. When an HCA hardware supports both FMR and FRMR features,
parameter can be used as over-ride over a default selection.

> Hello Santosh,
>
> What is the purpose of the prefer_frmr kernel module parameter ? Is this
> a parameter that is useful to RDS users or is its only purpose to allow
> developers of the RDS module to test both the FMR and FRMR code paths on
> hardware that supports both MR methods ?
>
Right. Since FRMR in early phase still for RDS, it was useful on HCA's
which supports both registration methods. Its not a deal breaker so
am going to drop the parameter as mentioned above.

Regards,
Santosh