linux-kernel - Re: [PATCH] KVM: x86: do not shadow apic global definition

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]

Date:   Sat, 30 Jul 2022 13:15:51 +0900
From:   Vincent MAILHOL <mailhol.vincent@...adoo.fr>
To:     Sean Christopherson <seanjc@...gle.com>
Cc:     kvm@...r.kernel.org, Paolo Bonzini <pbonzini@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
        "H . Peter Anvin" <hpa@...or.com>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] KVM: x86: do not shadow apic global definition

On Sat. 30 Jul. 2022 at 02:48, Sean Christopherson <seanjc@...gle.com> wrote:
> On Fri, Jul 29, 2022, Vincent Mailhol wrote:
> > arch/x86/include/asm/apic.h declares a global variable named `apic'.
> >
> > Many function arguments from arch/x86/kvm/lapic.h also uses the same
> > name and thus shadow the global declaration. For each case of
> > shadowing, rename the function argument from `apic' to `lapic'.
> >
> > This patch silences below -Wshadow warnings:
>
> This is just the tip of the iceberg, nearly every KVM x86 .c file has at least one
> "apic" variable.  arch/x86/kvm/lapic.c alone has nearly 100.  If this were the very
> last step before a kernel-wide (or even KVM-wide) enabling of -Wshadow then maybe
> it would be worth doing, but as it stands IMO it's unnecesary churn.

I would say the opposite: in terms of *volume*, warnings from apic.c
would be the tip of the iceberg and apic.h is the submerged part.

When the warning occurs in a header from the include directory, it
will spam some random files which include such header. This is
annoying when trying to triage W=2 warnings because you get totally
unrelated warning (and W=2 has some useful flags such as
-Wmaybe-uninitialized so there are some insensitive to check it).

My intent is only to silence the headers. I do not really care about
the -Wshadow on *.c files because it is local.

> What I would really love is to not have the global (and exported!) "apic", but
> properly solving that, i.e. not just a rename, would require a significant rework.

I double agree. I would also like to rename the global "apic" but I do
not think this is easily feasible.


Yours sincerely,
Vincent Mailhol