linux-kernel - Re: [PATCH] clockevents/drivers/i8253: Do not zero timer counter in shutdown

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: <874j842bqq.ffs@tglx>
Date: Thu, 01 Aug 2024 23:07:25 +0200
From: Thomas Gleixner <tglx@...utronix.de>
To: David Woodhouse <dwmw2@...radead.org>, mikelley@...rosoft.com, kvm
 <kvm@...r.kernel.org>
Cc: bp@...en8.de, dave.hansen@...ux.intel.com, decui@...rosoft.com,
 haiyangz@...rosoft.com, kys@...rosoft.com, linux-hyperv@...r.kernel.org,
 linux-kernel@...r.kernel.org, lirongqing@...du.com, mingo@...hat.com,
 seanjc@...gle.com, wei.liu@...nel.org, x86@...nel.org
Subject: Re: [PATCH] clockevents/drivers/i8253: Do not zero timer counter in
 shutdown

On Thu, Aug 01 2024 at 16:22, David Woodhouse wrote:
> On Thu, 2024-08-01 at 10:00 +0100, David Woodhouse wrote:
>  bool __init pit_timer_init(void)
>  {
> -	if (!use_pit())
> +	if (!use_pit()) {
> +		if (!hypervisor_is_type(X86_HYPER_NATIVE)) {
> +			/*
> +			 * Don't just ignore the PIT. Ensure it's stopped,
> +			 * because VMMs otherwise steal CPU time just to
> +			 * pointlessly waggle the (masked) IRQ.
> +			 */
> +			raw_spin_lock(&i8253_lock);
> +			outb_p(0x30, PIT_MODE);
> +
> +			/*
> +			 * It's not entirely clear from the datasheet, but some
> +			 * virtual implementations don't stop until the counter
> +			 * is actually written.
> +			 */
> +			if (i8253_clear_counter_on_shutdown) {
> +				outb_p(0, PIT_CH0);
> +				outb_p(0, PIT_CH0);
> +			}
> +			raw_spin_unlock(&i8253_lock);
> +		}
>  		return false;
> +	}

That's just wrong. What we want is to have the underlying problem
fixed in the driver and then make:
  
>  	clockevent_i8253_init(true);

bool clockevent_i8253_init(bool enable, bool oneshot);

so it can invoke the shutdown sequence instead of registering the pile:

   if (!enable) {
      shutdown();
      return false;
   }
   ...
   return true;

and the call site becomes:

    if (!clockevent_i8253_init(use_pit(), true))
    	return false;

No?

Thanks,

        tglx