linux-kernel - Re: [PATCH 16/23] x86/bugs: Remove default case for fully switched enums

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: <20230905050828.2y5rx4qqq2zertk2@treble>
Date:   Mon, 4 Sep 2023 22:08:28 -0700
From:   Josh Poimboeuf <jpoimboe@...nel.org>
To:     Borislav Petkov <bp@...en8.de>
Cc:     x86@...nel.org, linux-kernel@...r.kernel.org,
        Peter Zijlstra <peterz@...radead.org>,
        Babu Moger <babu.moger@....com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Sean Christopherson <seanjc@...gle.com>, David.Kaplan@....com,
        Andrew Cooper <andrew.cooper3@...rix.com>,
        Nikolay Borisov <nik.borisov@...e.com>,
        gregkh@...uxfoundation.org, Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH 16/23] x86/bugs: Remove default case for fully switched
 enums

On Sat, Sep 02, 2023 at 11:02:16AM +0200, Borislav Petkov wrote:
> On Fri, Aug 25, 2023 at 12:01:47AM -0700, Josh Poimboeuf wrote:
> > For enum switch statements which handle all possible cases, remove the
> > default case so a compiler warning gets printed if one of the enums gets
> > accidentally omitted from the switch statement.
> > 
> > Signed-off-by: Josh Poimboeuf <jpoimboe@...nel.org>
> > ---
> >  arch/x86/kernel/cpu/bugs.c | 17 +++++++----------
> >  1 file changed, 7 insertions(+), 10 deletions(-)
> 
> You could just as well take care of the default: cases in
> update_srbds_msr(), retbleed_select_mitigation(), unpriv_ebpf_notify(),
> spectre_v2_parse_user_cmdline() and cpu_show_common() and get rid of
> them all in this file and have the compiler warn for all of them.

I tried that, but adding all the unused cases adds a LOT of noise.  For
example the switch statement in spectre_v2_parse_user_cmdline() has
eight unused enums.  'default' is much more compact and readable than a
big list of all the unused enums which aren't really relevant there.

And in other places it added confusion. e.g., "RETBLEED_MITIGATION_NONE
isn't possible here, why is it being checked?"

It's a balance between readability and robustness (which is itself
affected by readability).  So I just constrained the patch to switch
statements which already handle all possible cases (as described above).

-- 
Josh