Date:   Fri, 10 Dec 2021 09:03:32 -0800
From:   Luis Chamberlain <mcgrof@kernel.org>
To:     Petr Mladek <pmladek@suse.com>
Cc:     Aaron Tomlin <atomlin@redhat.com>,
        Christoph Lameter <cl@linux.com>,
        Miroslav Benes <mbenes@suse.cz>,
        Andrew Morton <akpm@linux-foundation.org>, jeyu@kernel.org,
        linux-kernel@vger.kernel.org, linux-modules@vger.kernel.org,
        atomlin@atomlin.com
Subject: Re: [RFC PATCH] module: Introduce module unload taint tracking
Message-ID: <YbOIZOsiOarBtLxF@bombadil.infradead.org>
References: <20211124173327.3878177-1-atomlin@redhat.com>
 <YbEZ4HgSYQEPuRmS@bombadil.infradead.org>
 <20211209153131.a54fdfbci4qnyy6h@ava.usersys.com>
 <YbKUUJUtjBk/n913@bombadil.infradead.org>
 <YbMlVFwBiRujKdEX@alley>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <YbMlVFwBiRujKdEX@alley>
Sender: Luis Chamberlain <mcgrof@infradead.org>
Precedence: bulk

On Fri, Dec 10, 2021 at 11:00:52AM +0100, Petr Mladek wrote:
> Please, do not remove records that a module was removed. IMHO, it
> might be useful to track all removed module, including the non-tainted
> ones.

Then we'd need two features. One modules removed, and another which
limits this to only tainted modules. On kernel-ci systems where I
try to reproduce issues with fstests or blktests I might be unloading
a module 10,000 times, and so for those systems I'd like to disable
the tracking of all modules removed, otherwise we'd end up with
-ENOMEM eventually.

> Module removal is always tricky and not much tested.

It is tricky but I have been trying to correct issues along that path
and given that fstests and blktests uses it heavily I want to start
dispelling the false narrative that this is not a common use case.

> The tain flags might be just shown as extra information in the output.

Yes!

  Luis