RE: [PATCH v7 04/10] hisi_acc_vfio_pci: add new vfio_pci driver for HiSilicon ACC devices

[Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>]

> -----Original Message-----
> From: John Garry
> Sent: 02 March 2022 19:03
> To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>;
> kvm@vger.kernel.org; linux-kernel@vger.kernel.org;
> linux-crypto@vger.kernel.org
> Cc: linux-pci@vger.kernel.org; alex.williamson@redhat.com; jgg@nvidia.com;
> cohuck@redhat.com; mgurtovoy@nvidia.com; yishaih@nvidia.com; Linuxarm
> <linuxarm@huawei.com>; liulongfang <liulongfang@huawei.com>; Zengtao (B)
> <prime.zeng@hisilicon.com>; Jonathan Cameron
> <jonathan.cameron@huawei.com>; Wangzhou (B) <wangzhou1@hisilicon.com>
> Subject: Re: [PATCH v7 04/10] hisi_acc_vfio_pci: add new vfio_pci driver for
> HiSilicon ACC devices
> 
> On 02/03/2022 17:28, Shameer Kolothum wrote:
> > +config HISI_ACC_VFIO_PCI
> > +	tristate "VFIO PCI support for HiSilicon ACC devices"
> > +	depends on (ARM64 && VFIO_PCI_CORE) || (COMPILE_TEST && 64BIT)
> 
> This means that we will have HISI_ACC_VFIO_PCI=y for COMPILE_TEST=y and
> 64BIT=y and VFIO_PCI_CORE=n, but ...
> 
> > +	help
> > +	  This provides generic PCI support for HiSilicon ACC devices
> > +	  using the VFIO framework.
> > +
> > +	  If you don't know what to do here, say N.
> > diff --git a/drivers/vfio/pci/hisilicon/Makefile
> b/drivers/vfio/pci/hisilicon/Makefile
> > new file mode 100644
> > index 000000000000..c66b3783f2f9
> > --- /dev/null
> > +++ b/drivers/vfio/pci/hisilicon/Makefile
> > @@ -0,0 +1,4 @@
> > +# SPDX-License-Identifier: GPL-2.0-only
> > +obj-$(CONFIG_HISI_ACC_VFIO_PCI) += hisi-acc-vfio-pci.o
> > +hisi-acc-vfio-pci-y := hisi_acc_vfio_pci.o
> > +
> > diff --git a/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> > new file mode 100644
> > index 000000000000..8129c3457b3b
> > --- /dev/null
> > +++ b/drivers/vfio/pci/hisilicon/hisi_acc_vfio_pci.c
> > @@ -0,0 +1,100 @@
> > +// SPDX-License-Identifier: GPL-2.0-only
> > +/*
> > + * Copyright (c) 2021, HiSilicon Ltd.
> > + */
> > +
> > +#include <linux/device.h>
> > +#include <linux/eventfd.h>
> > +#include <linux/file.h>
> > +#include <linux/hisi_acc_qm.h>
> > +#include <linux/interrupt.h>
> > +#include <linux/module.h>
> > +#include <linux/pci.h>
> > +#include <linux/vfio.h>
> > +#include <linux/vfio_pci_core.h>
> > +
> > +static int hisi_acc_vfio_pci_open_device(struct vfio_device *core_vdev)
> > +{
> > +	struct vfio_pci_core_device *vdev =
> > +		container_of(core_vdev, struct vfio_pci_core_device, vdev);
> > +	int ret;
> > +
> > +	ret = vfio_pci_core_enable(vdev);
> > +	if (ret)
> > +		return ret;
> > +
> > +	vfio_pci_core_finish_enable(vdev);
> 
> ... there does not seem to be a stub for vfio_pci_core_finish_enable(),
> so I don't think that we compile under the conditions described. I think
> that HISI_ACC_VFIO_PCI should always depends on HISI_ACC_VFIO_PCI
> 

That looks right. I do remember running a compile test with x86_64. May
be VFIO_PCI_CORE was somehow selected.

I will make the "depends on VFIO_PCI_CORE" common.

Thanks,
Shameer