netdev - Re: [PATCH net-next v2] Add Mellanox BlueField Gigabit Ethernet driver

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:   Thu, 20 Aug 2020 15:48:25 -0700 (PDT)
From:   David Miller <davem@...emloft.net>
To:     davthompson@...dia.com
Cc:     kuba@...nel.org, dthompson@...lanox.com, netdev@...r.kernel.org,
        jiri@...lanox.com, Asmaa@...lanox.com
Subject: Re: [PATCH net-next v2] Add Mellanox BlueField Gigabit Ethernet
 driver

From: David Thompson <davthompson@...dia.com>
Date: Thu, 20 Aug 2020 18:51:39 +0000

> When I wrote the Kconfig definition I was thinking that "INET" is an
> obvious functional dependency for an Ethernet driver.

People can and should be able to use your driver even if ipv4/ipv6 is
disabled, don't you think?

> Yes, the mlxbf_gige silicon block needs to be programmed with the
> buffer's physical address so that the silicon logic can DMA incoming
> packet data into the buffer.  The kernel API "dma_alloc_coherent()"
> meets the driver's requirements in that it returns a CPU-useable address
> as well as a bus/physical address (used by silicon).

For descriptors and statistics blocks, coherent DMA memory makes sense.

For packet data, it does not.  These are streaming blocks of memory that
have their cohernecy managed by appropriate map/unmap/sync calls.