netdev - Re: [PATCH V3 3/6] vDPA: implement vq IRQ allocate/free helpers in vDPA core

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] [day] [month] [year] [list]

Message-ID: <a3c05c65-74bf-6250-5f4a-3b6cf60f4474@redhat.com>
Date:   Thu, 23 Jul 2020 12:19:42 +0800
From:   Jason Wang <jasowang@...hat.com>
To:     Zhu Lingshan <lingshan.zhu@...el.com>, alex.williamson@...hat.com,
        mst@...hat.com, pbonzini@...hat.com,
        sean.j.christopherson@...el.com, wanpengli@...cent.com
Cc:     virtualization@...ts.linux-foundation.org, netdev@...r.kernel.org,
        kvm@...r.kernel.org
Subject: Re: [PATCH V3 3/6] vDPA: implement vq IRQ allocate/free helpers in
 vDPA core


On 2020/7/22 下午6:08, Zhu Lingshan wrote:
> +/*
> + * Request irq for a vq, setup irq offloading if its a vhost_vdpa vq.
> + * This function should be only called through setting virtio DRIVER_OK.
> + * If you want to request irq during probe, you should use raw APIs
> + * like request_irq() or devm_request_irq().


This makes the API less flexibile. The reason is we store the irq in 
vhost-vdpa not vDPA.

I wonder whether the following looks better:

1) store irq in vdpa device
2) register producer when DRIVER_OK and unregister producer when 
!DRIVER_OK in vhost-vDPA
3) deal with the synchronization with SET_VRING_CALL
4) document that irq is not expected to be changed during DRIVER_OK

This can make sure the API works during driver probe, and we don't need 
the setup_irq and unsetup_irq method in vdpa_driver

Thanks


> + */
> +int vdpa_devm_request_irq(struct device *dev, struct vdpa_device *vdev,
> +			  unsigned int irq, irq_handler_t handler,
> +			  unsigned long irqflags, const char *devname, void *dev_id,
> +			  int qid)
> +{
> +	int ret;
> +
> +	ret = devm_request_irq(dev, irq, handler, irqflags, devname, dev_id);
> +	if (ret)
> +		dev_err(dev, "Failed to request irq for vq %d\n", qid);
> +	else
> +		vdpa_setup_irq(vdev, qid, irq);
> +
> +	return ret;
> +
> +}
> +EXPORT_SYMBOL_GPL(vdpa_devm_request_irq);