netdev - Re: [PATCH net-next v3 3/3] net: phy: add Rust Asix PHY 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]

Message-ID: <bdfac30f-364f-4625-a808-fcffab2f75b4@lunn.ch>
Date: Mon, 9 Oct 2023 14:32:44 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Jiri Pirko <jiri@...nulli.us>
Cc: FUJITA Tomonori <fujita.tomonori@...il.com>, netdev@...r.kernel.org,
	rust-for-linux@...r.kernel.org, miguel.ojeda.sandonis@...il.com,
	greg@...ah.com, tmgross@...ch.edu
Subject: Re: [PATCH net-next v3 3/3] net: phy: add Rust Asix PHY driver

> Wait. So you just add rust driver as a duplicate of existing c driver?
> What's the point exactly to have 2 drivers for the same thing?

To tell the truth, i don't think the driver itself is very
important. The important thing has been the discussion along the way
to get to a driver.

For me as a Maintainer, the discussion about maintainability, how do
we make the build fail when C and Rust diverge is important.  So it
seems C enum are better than #defines for that. Maybe we will need to
replace some #define lists with enum in core code? But we also have a
lot of #defines which it would be good to be able to use.

It took a while to get the code to actually register the driver
instances. But this is something nearly every driver needs to do, so i
hope the ideas and maybe some of the actual code can be used in other
drivers.

It has become much clearer the Rust build system needs work. It is too
monolithic at the moment, and there are no good examples of kconfig
integration. 

Documentation is still an open question for me. I know Rust can
generate documentation from the code, but how do we integrate that
into the existing kernel documentation?

Within netdev, our own tooling is not ready for Rust. Our patchwork
instance did not recognise the patch was for netdev. That has been
fixed now. But i'm pretty sure the latest version of the code was not
built as part of our build testing. Jakub said the machine did not
have a Rust toolchain.  I also think because of the Rust build system
limitations, even if it did have the tools, i don't think with
allmodconfig it would try to built the rust code.

When i build it on my machine, i get a million warnings from i think
the linker. That is not going to be acceptable to Mainline, so we need
to investigate that.

I hope some sort of lessons learned, best practices and TODO list can
be distilled from the experience, to help guide the Rust Experiment.

	Andrew