linux-kernel - Re: [PATCH V4 0/5] mlx5 ConnectX control misc 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: Wed, 14 Feb 2024 00:29:16 -0800
From: Christoph Hellwig <hch@...radead.org>
To: Saeed Mahameed <saeed@...nel.org>
Cc: Arnd Bergmann <arnd@...db.de>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Leon Romanovsky <leonro@...dia.com>,
	Jason Gunthorpe <jgg@...dia.com>, Jiri Pirko <jiri@...dia.com>,
	Leonid Bloch <lbloch@...dia.com>, Itay Avraham <itayavr@...dia.com>,
	Jakub Kicinski <kuba@...nel.org>,
	Saeed Mahameed <saeedm@...dia.com>,
	David Ahern <dsahern@...nel.org>,
	Aron Silverton <aron.silverton@...cle.com>,
	Christoph Hellwig <hch@...radead.org>,
	andrew.gospodarek@...adcom.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH V4 0/5] mlx5 ConnectX control misc driver

Chmining in late after a vacation last week, but I really do not
understand the discussion this cause.

Complex devices need diagnostics, and mlx5ctl and the (free software)
userspace tool provide that.  Given how it is a complex multi-subsystem
driver that exports at least RDMA, ethernet, nvme and virtio_blk
interfaces this functionality by definition can't fit into a single
subsystem.

So with my nvme co-maintainer hat on:

Acked-by: Christoph Hellwig <hch@....de>

to th concept (which is not a full code review!).

With my busy kernel contributor head on I have to voice my
dissatisfaction with the subsystem maintainer overreach that's causing
the troubles here.  I think all maintainers can and should voice the
opinions, be those technical or political, but trying to block a useful
feature without lots of precedence because it is vaguely related to the
subsystem is not helpful.  Note that this is absolutely not intended to
shut the discussion down - if we can find valid arguments why some of
this functionality should also be reported through a netdev API we
should continue that.  E.g. just because we have SCSI, NVMe or product
specific passthrough interface we still try to provide useful common
interface generically.