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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Fri, 03 Feb 2023 10:55:46 -0700
From:   Jonathan Corbet <corbet@....net>
To:     Carlos Bilbao <carlos.bilbao@....com>, ojeda@...nel.org,
        jani.nikula@...ux.intel.com, rdunlap@...radead.org
Cc:     linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org,
        konstantin@...uxfoundation.org, bilbao@...edu,
        Carlos Bilbao <carlos.bilbao@....com>
Subject: Re: [PATCH v6 0/2] docs: Integrate rustdoc into Rust documentation

Carlos Bilbao <carlos.bilbao@....com> writes:

> Include HTML output generated with rustdoc into the Linux kernel
> documentation on Rust.
>
> Carlos Bilbao:
>  docs: Move rustdoc output, cross-reference it
>  docs: Integrate rustdoc generation into htmldocs
>
So I spent a fair while messing with this, really hoping to get it
merged for 6.3, but I think it's still just too fragile.

- With a too-new version of bindgen it kills the htmldocs build
  entirely.  If we can't build the Rust docs successfully with the
  toolchain that the user has installed then so be it, but it shouldn't
  wreck the docs build entirely.

- It still throws the user into the "make oldconfig" process, which is
  unexpected for a docs build.  This happens even on a system with no
  Rust installed at all.

- I dumped my Fedora toolchain and did the curl|bash ritual to set up
  the proper versions with rustup.  It would be nice to have some better
  documentation on how to do that ... *how* to install bindgen may not
  be obvious to all users.  It also failed until I installed lld
  separately.

On a system with the right tool versions, and with a kernel already
configured to build the Rust support, it worked nicely, though.

I'd really like to see a solution that (1) doesn't kill the docs build
process if the tool versions are wrong, and (2) doesn't force the
creation of a kernel configuration.  What are our chances of getting
there?

Thanks,

jon

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ