From mboxrd@z Thu Jan  1 00:00:00 1970
From: Alex Williamson <alex.williamson@redhat.com>
Subject: Re: [PATCH 3/5] VFIO: Support threaded interrupt handling on VFIO
Date: Wed, 16 Dec 2015 12:49:06 -0700
Message-ID: <1450295346.2674.50.camel@redhat.com>
References: <1449166972-8894-1-git-send-email-yunhong.jiang@linux.intel.com>
	 <1449166972-8894-4-git-send-email-yunhong.jiang@linux.intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org
To: Yunhong Jiang <yunhong.jiang@linux.intel.com>, pbonzini@redhat.com
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <1449166972-8894-4-git-send-email-yunhong.jiang@linux.intel.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

On Thu, 2015-12-03 at 10:22 -0800, Yunhong Jiang wrote:
> For VFIO device with MSI interrupt type, it's possible to handle the
> interrupt on hard interrupt context without invoking the interrupt
> thread. Handling the interrupt on hard interrupt context reduce the
> interrupt latency.
>=20
> Signed-off-by: Yunhong Jiang <yunhong.jiang@linux.intel.com>
> ---
> =C2=A0drivers/vfio/pci/vfio_pci_intrs.c | 39 ++++++++++++++++++++++++=
++++++++++-----
> =C2=A01 file changed, 34 insertions(+), 5 deletions(-)
>=20
> diff --git a/drivers/vfio/pci/vfio_pci_intrs.c b/drivers/vfio/pci/vfi=
o_pci_intrs.c
> index 3b3ba15558b7..108d335c5656 100644
> --- a/drivers/vfio/pci/vfio_pci_intrs.c
> +++ b/drivers/vfio/pci/vfio_pci_intrs.c
> @@ -236,12 +236,35 @@ static void vfio_intx_disable(struct vfio_pci_d=
evice *vdev)
> =C2=A0	kfree(vdev->ctx);
> =C2=A0}
> =C2=A0
> +static irqreturn_t vfio_msihandler(int irq, void *arg)
> +{
> +	struct vfio_pci_irq_ctx *ctx =3D arg;
> +	struct irq_bypass_producer *producer =3D &ctx->producer;
> +	struct irq_bypass_consumer *consumer;
> +	int ret =3D IRQ_HANDLED, idx;
> +
> +	idx =3D srcu_read_lock(&producer->srcu);
> +
> +	list_for_each_entry_rcu(consumer, &producer->consumers, sibling) {
> +		/*
> +		=C2=A0* Invoke the thread handler if any consumer would block, but
> +		=C2=A0* finish all consumes.
> +		=C2=A0*/
> +		if (consumer->handle_irq(consumer->irq_context) =3D=3D -EWOULDBLOC=
K)
> +			ret =3D IRQ_WAKE_THREAD;
> +		continue;
> +	}
> +
> +	srcu_read_unlock(&producer->srcu, idx);


There should be an irq bypass manager interface to abstract this.