netdev - Re: [patch net-next 10/42] mlxsw: spectrum_router: Add basic ipv4 router initialization

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: <20160701175851.GA11348@colbert.idosch.org>
Date:	Fri, 1 Jul 2016 20:58:51 +0300
From:	Ido Schimmel <idosch@...lanox.com>
To:	David Ahern <dsa@...ulusnetworks.com>
CC:	Jiri Pirko <jiri@...nulli.us>, <netdev@...r.kernel.org>,
	<davem@...emloft.net>, <yotamg@...lanox.com>, <eladr@...lanox.com>,
	<nogahf@...lanox.com>, <ogerlitz@...lanox.com>,
	<sfeldma@...il.com>, <roopa@...ulusnetworks.com>,
	<andy@...yhouse.net>, <tgraf@...g.ch>, <jhs@...atatu.com>,
	<linville@...driver.com>, <ivecera@...hat.com>
Subject: Re: [patch net-next 10/42] mlxsw: spectrum_router: Add basic ipv4
 router initialization

Fri, Jul 01, 2016 at 05:39:01PM IDT, dsa@...ulusnetworks.com wrote:
>On 7/1/16 8:04 AM, Jiri Pirko wrote:
>> diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
>> index 05d5fcc..c2ac037 100644
>> --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
>> +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
>> @@ -74,6 +74,8 @@
>>
>>  #define MLXSW_SP_CELL_FACTOR 2	/* 2 * cell_size / (IPG + cell_size + 1) */
>>
>> +#define MLXSW_SP_RIF_MAX 800
>
>At most 800 RIFs can be created?

Yes.

Do you have use cases that require more than that? If so, how many
router interfaces would be required? That would be a good feedback for
us.

>Why 800?

Currently, trying to configure a RIF with an higher index will result in
firmware errors. We plan to increase this number in the future (don't
expect it to be infinite...).

We are currently implementing a new mechanism in the driver that will
query these resources from the firmware, so we won't need to hardcode
them.