linux-kernel - Re: Linux 5.10-rc2 (docs build)

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: <20201102085841.5490f6da@coco.lan>
Date:   Mon, 2 Nov 2020 08:58:41 +0100
From:   Mauro Carvalho Chehab <mchehab@...nel.org>
To:     Randy Dunlap <rdunlap@...radead.org>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Jonathan Corbet <corbet@....net>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: Linux 5.10-rc2 (docs build)

Hi Randy,

Em Sun, 1 Nov 2020 16:01:54 -0800
Randy Dunlap <rdunlap@...radead.org> escreveu:

> Something broke the docs build (SEVERE):
> 
> 
> Sphinx parallel build error:
> docutils.utils.SystemMessage: /home/rdunlap/lnx/lnx-510-rc2/Documentation/ABI/testing/sysfs-bus-rapidio:2: (SEVERE/4) Title level inconsistent:
> 
> Attributes Common for All RapidIO Devices
> -----------------------------------------
> 
> 

On what tree did you notice this? Linux-next or upstream?

Using title markups on ABI files is problematic. As they'll
all be placed as a single document, the markups for level
1, level 2, etc should be identical on all files that use them.

A quick fix would be to replace the above to either:

	**Attributes Common for All RapidIO Devices**

or just:


	Attributes Common for All RapidIO Devices

I guess it should be easy to change the get_abi.pl script to
just ignore markups that matches something like:

	^[=-]+$

That should be more reliable long-term. I'll write such patch
and submit it.

> and then it stops/hangs. Does not terminate normally but is no longer
> executing.

That's a docutils/Sphinx bug. Once I wrote a dirty patch fixing it, 
but I ended losing the patch[1]. Basically, when severe errors occur there,
it tries to wait for all processes to die, but this never happens,
causing it to wait forever.


[1] I guess I did something similar to this very dirty hack:

--- sphinx_3.2.1/lib/python3.8/site-packages/docutils/parsers/rst/states.py	2020-11-02 08:50:21.602785381 +0100
+++ sphinx_3.2.1/lib/python3.8/site-packages/docutils/parsers/rst/states.py	2020-11-02 08:52:45.216612493 +0100
@@ -371,7 +371,7 @@ class RSTState(StateWS):
         error = self.reporter.severe(
             'Title level inconsistent:', nodes.literal_block('', sourcetext),
             line=lineno)
-        return error
+        sys.exit()
 
     def new_subsection(self, title, lineno, messages):
         """Append new subsection to document tree. On return, check level."""

Thanks,
Mauro