From mboxrd@z Thu Jan  1 00:00:00 1970
From: Sheng Yang <sheng@linux.intel.com>
Subject: Re: [PATCH 08/10] KVM: Add assigned_device_msi_dispatch()
Date: Tue, 4 Nov 2008 19:17:32 +0800
Message-ID: <200811041917.33216.sheng@linux.intel.com>
References: <1225428647-27614-1-git-send-email-sheng@linux.intel.com> <1225428647-27614-9-git-send-email-sheng@linux.intel.com> <49102F2F.4060404@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Cc: kvm@vger.kernel.org
To: Avi Kivity <avi@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mga05.intel.com ([192.55.52.89]:17627 "EHLO
	fmsmga101.fm.intel.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org
	with ESMTP id S1752442AbYKDLU1 (ORCPT <rfc822;kvm@vger.kernel.org>);
	Tue, 4 Nov 2008 06:20:27 -0500
In-Reply-To: <49102F2F.4060404@redhat.com>
Content-Disposition: inline
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Tuesday 04 November 2008 19:17:03 Avi Kivity wrote:
> Sheng Yang wrote:
> > The function is used to dispatch MSI to lapic according to MSI message
> > address and message data.
> >
> > (Any way to replace macro GET_VAL_FROM_SHIFT()?)
> >
> >
> > +#ifdef CONFIG_X86
> > +#include <asm/msidef.h>
> > +#endif
>
> The include can be unconditional.

it's pity that IA64 don't have this head file... I will separate the msi.c and 
add this later.

>
> > +
> > +#ifdef CONFIG_X86
> > +static void assigned_device_msi_dispatch(struct kvm_assigned_dev_kernel
> > *dev) +{
> > +#define GET_VAL_FROM_SHIFT(v, s) (((v) & (1 << s)) >> s)
>
> test_bit() ?

Oh, indeed here. :)
>
> > +	int vcpu_id;
> > +	struct kvm_vcpu *vcpu;
> > +	struct kvm_ioapic *ioapic = ioapic_irqchip(dev->kvm);
> > +	u8 dest_id = (dev->guest_msi_addr & MSI_ADDR_DEST_ID_MASK)
> > +			>> MSI_ADDR_DEST_ID_SHIFT;
> > +	u8 vector = (dev->guest_msi_data & MSI_DATA_VECTOR_MASK)
> > +			>> MSI_DATA_VECTOR_SHIFT;
> > +	u8 dest_mode = GET_VAL_FROM_SHIFT(dev->guest_msi_addr,
> > +					  MSI_ADDR_DEST_MODE_SHIFT);
> > +	u8 trig_mode = GET_VAL_FROM_SHIFT(dev->guest_msi_data,
> > +					  MSI_DATA_TRIGGER_SHIFT);
> > +	u8 delivery_mode = GET_VAL_FROM_SHIFT(dev->guest_msi_data,
> > +					  MSI_DATA_DELIVERY_MODE_SHIFT);
>
> int is usually better than u8.

OK.
>
> > +	u32 deliver_bitmask;
> > +
> > +	case MSI_DATA_DELIVERY_FIXED_VAL:
> > +		for (vcpu_id = 0; deliver_bitmask != 0; vcpu_id++) {
>
> __ffs() and __clear_bit() can make this a bit faster.  Can be done later
> though.

OK, also would optimize IOAPIC side later.

-- 
regards
Yang, Sheng