linux-kernel - Re: [PATCH v9 0/8] Parallel CPU bringup for x86

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: <87fsaym4gg.ffs@tglx>
Date:   Wed, 22 Feb 2023 10:31:27 +0100
From:   Thomas Gleixner <tglx@...utronix.de>
To:     David Woodhouse <dwmw2@...radead.org>,
        Oleksandr Natalenko <oleksandr@...alenko.name>
Cc:     Kim Phillips <kim.phillips@....com>,
        Usama Arif <usama.arif@...edance.com>, arjan@...ux.intel.com,
        mingo@...hat.com, bp@...en8.de, dave.hansen@...ux.intel.com,
        hpa@...or.com, x86@...nel.org, pbonzini@...hat.com,
        paulmck@...nel.org, linux-kernel@...r.kernel.org,
        kvm@...r.kernel.org, rcu@...r.kernel.org, mimoja@...oja.de,
        hewenliang4@...wei.com, thomas.lendacky@....com, seanjc@...gle.com,
        pmenzel@...gen.mpg.de, fam.zheng@...edance.com,
        punit.agrawal@...edance.com, simon.evans@...edance.com,
        liangma@...ngbit.com,
        "Limonciello, Mario" <Mario.Limonciello@....com>,
        Piotr Gorski <piotrgorski@...hyos.org>
Subject: Re: [PATCH v9 0/8] Parallel CPU bringup for x86_64

On Tue, Feb 21 2023 at 23:18, David Woodhouse wrote:
> On Tue, 2023-02-21 at 22:41 +0100, Thomas Gleixner wrote:
>> +
>> +       if (IS_ENABLED(CONFIG_64BIT) && IS_ENABLED(CONFIG_SMP))
>> +               smpboot_control = saved_smpboot_ctrl;
>>         return 0;
>>  }
>>  
>
> But wait, why is this giving it a dedicated temp_stack anyway? Why
> can't it use that CPU's idle thread stack like we usually do? I already
> made idle_thread_get() accessible from here. So we could do this...

Because this very CPU is still online and from the kernels POV is does
not go offline. It goes into the firmware blackhole and comes back
magically through the startup code.

That means this very CPUs indle thread stack is in use and the resume
path will scribble over it. Maybe you won't notice because it only
clobbers top of stack which is never used again because the idle thread
does not return. But correct is something different.

Thanks,

        tglx