linux-kernel - Re: [PATCH] s390/vfio-ap: do not use open locks during VFIO_GROUP_NOTIFY_SET

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: <20210628182248.GH4459@nvidia.com>
Date:   Mon, 28 Jun 2021 15:22:48 -0300
From:   Jason Gunthorpe <jgg@...dia.com>
To:     Tony Krowiak <akrowiak@...ux.ibm.com>
Cc:     linux-s390@...r.kernel.org, linux-kernel@...r.kernel.org,
        borntraeger@...ibm.com, cohuck@...hat.com,
        pasic@...ux.vnet.ibm.com, jjherne@...ux.ibm.com,
        alex.williamson@...hat.com, kwankhede@...dia.com,
        frankja@...ux.ibm.com, david@...hat.com, imbrenda@...ux.ibm.com,
        hca@...ux.ibm.com
Subject: Re: [PATCH] s390/vfio-ap: do not use open locks during
 VFIO_GROUP_NOTIFY_SET_KVM notification

On Mon, Jun 28, 2021 at 02:20:55PM -0400, Tony Krowiak wrote:
> 
> 
> On 6/28/21 1:34 PM, Jason Gunthorpe wrote:
> > On Fri, Jun 25, 2021 at 06:07:58PM -0400, Tony Krowiak wrote:
> > >   static void vfio_ap_mdev_unset_kvm(struct ap_matrix_mdev *matrix_mdev)
> > >   {
> > > +	mutex_lock(&matrix_dev->lock);
> > > +	if ((matrix_mdev->kvm) && (matrix_mdev->kvm->arch.crypto.crycbd)) {
> > >   		mutex_unlock(&matrix_dev->lock);
> > > +		down_write(&matrix_mdev->kvm->arch.crypto.pqap_hook_rwsem);
> > > +		matrix_mdev->kvm->arch.crypto.pqap_hook = NULL;
> > > +		up_write(&matrix_mdev->kvm->arch.crypto.pqap_hook_rwsem);
> > > +		kvm_arch_crypto_clear_masks(matrix_mdev->kvm);
> > >   		mutex_lock(&matrix_dev->lock);
> > >   		vfio_ap_mdev_reset_queues(matrix_mdev->mdev);
> > >   		kvm_put_kvm(matrix_mdev->kvm);
> > >   		matrix_mdev->kvm = NULL;
> > > +		mutex_unlock(&matrix_dev->lock);
> > >   	}
> > Doesn't a flow exit the function with matrix_dev->lock held he
> 
> How can that happen? What flow?

When the if isn't taken

> > Write it with 'success oriented flow'
> 
> I'm not sure what you mean, can you clarify this statement?

Basically, don't write the bulk of the function under an if statement

mutex_lock(&matrix_dev->lock);
if (!matrix_mdev->kvm || !matrix_mdev->kvm->arch.crypto.crycbd) {
    mutex_unlock(&matrix_dev->lock);
    return;
}

Jason