From mboxrd@z Thu Jan  1 00:00:00 1970
From: Glauber Costa <glommer@parallels.com>
Subject: Re: [patch 12/18] x86: vdso: pvclock gettime support
Date: Mon, 29 Oct 2012 18:59:35 +0400
Message-ID: <508E99D7.2050402@parallels.com>
References: <20121024131340.742340256@redhat.com> <20121024131621.840274360@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 7bit
Cc: <kvm@vger.kernel.org>, <johnstul@us.ibm.com>, <jeremy@goop.org>,
	<zamsden@gmail.com>, <gleb@redhat.com>, <avi@redhat.com>,
	<pbonzini@redhat.com>
To: Marcelo Tosatti <mtosatti@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx2.parallels.com ([64.131.90.16]:32838 "EHLO mx2.parallels.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752231Ab2J2O7o (ORCPT <rfc822;kvm@vger.kernel.org>);
	Mon, 29 Oct 2012 10:59:44 -0400
In-Reply-To: <20121024131621.840274360@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On 10/24/2012 05:13 PM, Marcelo Tosatti wrote:
> Improve performance of time system calls when using Linux pvclock, 
> by reading time info from fixmap visible copy of pvclock data.
> 
> Originally from Jeremy Fitzhardinge.
> 
> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
> 
> Index: vsyscall/arch/x86/vdso/vclock_gettime.c
> ===================================================================
> --- vsyscall.orig/arch/x86/vdso/vclock_gettime.c
> +++ vsyscall/arch/x86/vdso/vclock_gettime.c
> @@ -22,6 +22,7 @@
>  #include <asm/hpet.h>
>  #include <asm/unistd.h>
>  #include <asm/io.h>
> +#include <asm/pvclock.h>
>  
>  #define gtod (&VVAR(vsyscall_gtod_data))
>  
> @@ -62,6 +63,69 @@ static notrace cycle_t vread_hpet(void)
>  	return readl((const void __iomem *)fix_to_virt(VSYSCALL_HPET) + 0xf0);
>  }
>  
> +#ifdef CONFIG_PARAVIRT_CLOCK_VSYSCALL
> +
> +static notrace const struct pvclock_vsyscall_time_info *get_pvti(int cpu)
> +{
> +	const aligned_pvti_t *pvti_base;
> +	int idx = cpu / (PAGE_SIZE/PVTI_SIZE);
> +	int offset = cpu % (PAGE_SIZE/PVTI_SIZE);
> +
> +	BUG_ON(PVCLOCK_FIXMAP_BEGIN + idx > PVCLOCK_FIXMAP_END);
> +
> +	pvti_base = (aligned_pvti_t *)__fix_to_virt(PVCLOCK_FIXMAP_BEGIN+idx);
> +
> +	return &pvti_base[offset].info;
> +}
> +

Unless I am missing something, if gcc decides to not inline get_pvti,
this will break, right? I believe you need to mark that function with
__always_inline.

> +static notrace cycle_t vread_pvclock(int *mode)
> +{
> +	const struct pvclock_vsyscall_time_info *pvti;
> +	cycle_t ret;
> +	u64 last;
> +	u32 version;
> +	u32 migrate_count;
> +	u8 flags;
> +	unsigned cpu, cpu1;
> +
> +
> +	/*
> +	 * When looping to get a consistent (time-info, tsc) pair, we
> +	 * also need to deal with the possibility we can switch vcpus,
> +	 * so make sure we always re-fetch time-info for the current vcpu.
> +	 */
> +	do {
> +		cpu = __getcpu() & 0xfff;

Please wrap this 0xfff into something meaningful.

> +		pvti = get_pvti(cpu);
> +
> +		migrate_count = pvti->migrate_count;
> +
> +		version = __pvclock_read_cycles(&pvti->pvti, &ret, &flags);
> +
> +		/*
> +		 * Test we're still on the cpu as well as the version.
> +		 * We could have been migrated just after the first
> +		 * vgetcpu but before fetching the version, so we
> +		 * wouldn't notice a version change.
> +		 */
> +		cpu1 = __getcpu() & 0xfff;
> +	} while (unlikely(cpu != cpu1 ||
> +			  (pvti->pvti.version & 1) ||
> +			  pvti->pvti.version != version ||
> +			  pvti->migrate_count != migrate_count));
> +
> +	if (unlikely(!(flags & PVCLOCK_TSC_STABLE_BIT)))
> +		*mode = VCLOCK_NONE;
> +
> +	last = VVAR(vsyscall_gtod_data).clock.cycle_last;
> +
> +	if (likely(ret >= last))
> +		return ret;
> +

Please add a comment here referring to tsc.c, where an explanation of
this test lives. This is quite non-obvious for the non initiated.