linux-kernel - Re: [PATCH v9 3/8] x86/split_lock: Introduce flag X86_FEATURE_SLD_FATAL and drop sld

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]

Message-ID: <CALCETrXwtj5rhVM6YYNEDeDqT3eKFNkGFCgSB_hUd7aOYBFXmw@mail.gmail.com>
Date:   Sat, 9 May 2020 22:14:02 -0700
From:   Andy Lutomirski <luto@...nel.org>
To:     Xiaoyao Li <xiaoyao.li@...el.com>
Cc:     Paolo Bonzini <pbonzini@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Sean Christopherson <sean.j.christopherson@...el.com>,
        kvm list <kvm@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>, X86 ML <x86@...nel.org>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        "H. Peter Anvin" <hpa@...or.com>,
        Andy Lutomirski <luto@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Arvind Sankar <nivedita@...m.mit.edu>,
        Tony Luck <tony.luck@...el.com>,
        Fenghua Yu <fenghua.yu@...el.com>
Subject: Re: [PATCH v9 3/8] x86/split_lock: Introduce flag X86_FEATURE_SLD_FATAL
 and drop sld_state

On Fri, May 8, 2020 at 8:03 PM Xiaoyao Li <xiaoyao.li@...el.com> wrote:
>
> Introduce a synthetic feature flag X86_FEATURE_SLD_FATAL, which means
> kernel is in sld_fatal mode if set.
>
> Now sld_state is not needed any more that the state of SLD can be
> inferred from X86_FEATURE_SPLIT_LOCK_DETECT and X86_FEATURE_SLD_FATAL.

Is it too much to ask for Intel to actually allocate and define a
CPUID bit that means "this CPU *always* sends #AC on a split lock"?
This would be a pure documentation change, but it would make this
architectural rather than something that each hypervisor needs to hack
up.

Meanwhile, I don't see why adding a cpufeature flag is worthwhile to
avoid a less bizarre global variable.  There's no performance issue
here, and the old code looked a lot more comprehensible than the new
code.