diff mbox

[v6,07/13] vfio/ccw: vfio based subchannel passthrough driver

Message ID 20170428130951.44771-8-bjsdjshi@linux.vnet.ibm.com (mailing list archive)
State New, archived
Headers show

Commit Message

Dong Jia Shi April 28, 2017, 1:09 p.m. UTC 
From: Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>

We use the IOMMU_TYPE1 of VFIO to realize the subchannels
passthrough, implement a vfio based subchannels passthrough
driver called "vfio-ccw".

Support qemu parameters in the style of:
"-device vfio-ccw,sysfsdev=$mdev_file_path,devno=xx.x.xxxx'

Signed-off-by: Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>
Signed-off-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
---
 default-configs/s390x-softmmu.mak |   1 +
 hw/vfio/Makefile.objs             |   1 +
 hw/vfio/ccw.c                     | 189 ++++++++++++++++++++++++++++++++++++++
 include/hw/vfio/vfio-common.h     |   1 +
 4 files changed, 192 insertions(+)
 create mode 100644 hw/vfio/ccw.c

Comments

Eric Auger May 1, 2017, 11:08 a.m. UTC | #1 
Hi Dong,

On 28/04/2017 15:09, Dong Jia Shi wrote:
> From: Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>
> 
> We use the IOMMU_TYPE1 of VFIO to realize the subchannels
> passthrough, implement a vfio based subchannels passthrough
> driver called "vfio-ccw".
> 
> Support qemu parameters in the style of:
> "-device vfio-ccw,sysfsdev=$mdev_file_path,devno=xx.x.xxxx'
> 
> Signed-off-by: Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>
> Signed-off-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
> ---
>  default-configs/s390x-softmmu.mak |   1 +
>  hw/vfio/Makefile.objs             |   1 +
>  hw/vfio/ccw.c                     | 189 ++++++++++++++++++++++++++++++++++++++
>  include/hw/vfio/vfio-common.h     |   1 +
>  4 files changed, 192 insertions(+)
>  create mode 100644 hw/vfio/ccw.c
> 
> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
> index 36e15de..5576b0a 100644
> --- a/default-configs/s390x-softmmu.mak
> +++ b/default-configs/s390x-softmmu.mak
> @@ -4,4 +4,5 @@ CONFIG_VIRTIO=y
>  CONFIG_SCLPCONSOLE=y
>  CONFIG_S390_FLIC=y
>  CONFIG_S390_FLIC_KVM=$(CONFIG_KVM)
> +CONFIG_VFIO_CCW=$(CONFIG_LINUX)
>  CONFIG_WDT_DIAG288=y
> diff --git a/hw/vfio/Makefile.objs b/hw/vfio/Makefile.objs
> index 05e7fbb..c3ab909 100644
> --- a/hw/vfio/Makefile.objs
> +++ b/hw/vfio/Makefile.objs
> @@ -1,6 +1,7 @@
>  ifeq ($(CONFIG_LINUX), y)
>  obj-$(CONFIG_SOFTMMU) += common.o
>  obj-$(CONFIG_PCI) += pci.o pci-quirks.o
> +obj-$(CONFIG_VFIO_CCW) += ccw.o
>  obj-$(CONFIG_SOFTMMU) += platform.o
>  obj-$(CONFIG_VFIO_XGMAC) += calxeda-xgmac.o
>  obj-$(CONFIG_VFIO_AMD_XGBE) += amd-xgbe.o
> diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c
> new file mode 100644
> index 0000000..cd4dfe8
> --- /dev/null
> +++ b/hw/vfio/ccw.c
> @@ -0,0 +1,189 @@
> +/*
> + * vfio based subchannel assignment support
> + *
> + * Copyright 2017 IBM Corp.
> + * Author(s): Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
> + *            Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>
> + *            Pierre Morel <pmorel@linux.vnet.ibm.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or(at
> + * your option) any version. See the COPYING file in the top-level
> + * directory.
> + */
> +
> +#include <linux/vfio.h>
> +#include <sys/ioctl.h>
> +
> +#include "qemu/osdep.h"
> +#include "qapi/error.h"
> +#include "hw/sysbus.h"
> +#include "hw/vfio/vfio.h"
> +#include "hw/vfio/vfio-common.h"
> +#include "hw/s390x/s390-ccw.h"
> +#include "hw/s390x/ccw-device.h"
> +
> +#define TYPE_VFIO_CCW "vfio-ccw"
> +typedef struct VFIOCCWDevice {
> +    S390CCWDevice cdev;
> +    VFIODevice vdev;
> +} VFIOCCWDevice;
> +
> +static void vfio_ccw_compute_needs_reset(VFIODevice *vdev)
> +{
> +    vdev->needs_reset = false;
> +}
> +
> +/*
> + * We don't need vfio_hot_reset_multi and vfio_eoi operations for
> + * vfio_ccw device now.
> + */
> +struct VFIODeviceOps vfio_ccw_ops = {
> +    .vfio_compute_needs_reset = vfio_ccw_compute_needs_reset,
> +};
> +
> +static void vfio_ccw_reset(DeviceState *dev)
> +{
> +    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
> +    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
> +    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
> +
> +    ioctl(vcdev->vdev.fd, VFIO_DEVICE_RESET);
> +}
> +
> +static void vfio_put_device(VFIOCCWDevice *vcdev)
> +{
> +    g_free(vcdev->vdev.name);
> +    vfio_put_base_device(&vcdev->vdev);
> +}
> +
> +static VFIOGroup *vfio_ccw_get_group(S390CCWDevice *cdev, Error **errp)
> +{
> +    char *tmp, group_path[PATH_MAX];
> +    ssize_t len;
> +    int groupid;
> +
> +    tmp = g_strdup_printf("/sys/bus/css/devices/%x.%x.%04x/%s/iommu_group",
> +                          cdev->hostid.cssid, cdev->hostid.ssid,
> +                          cdev->hostid.devid, cdev->mdevid);
May be worth replacing this static function by a common
vfio_get_device_group(VFIODevice *vbasedev, Error **errp) in common.c
that would be used by vfio/pci devices? It would use the sysfsdev
populated before.
> +    len = readlink(tmp, group_path, sizeof(group_path));
> +    g_free(tmp);
> +
> +    if (len <= 0 || len >= sizeof(group_path)) {
> +        error_setg(errp, "vfio: no iommu_group found");
may be good to align the error message with pci/platform. I noticed the
case where len==0 is handled as ENAMETOOLONG on pci/platform which looks
bad. Maybe this is why you changed it.
> +        return NULL;
> +    }
> +
> +    group_path[len] = 0;
> +
> +    if (sscanf(basename(group_path), "%d", &groupid) != 1) {
> +        error_setg(errp, "vfio: failed to read %s", group_path);
> +        return NULL;
> +    }
> +
> +    return vfio_get_group(groupid, &address_space_memory, errp);
> +}
> +
> +static void vfio_ccw_put_group(VFIOGroup *group)
Is it really needed?
> +{
> +    vfio_put_group(group);
> +}
> +
> +static void vfio_ccw_realize(DeviceState *dev, Error **errp)
> +{
> +    VFIODevice *vbasedev;
> +    VFIOGroup *group;
> +    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
> +    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
> +    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
> +    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
> +
> +    /* Call the class init function for subchannel. */
> +    if (cdc->realize) {
> +        cdc->realize(cdev, vcdev->vdev.sysfsdev, errp);
> +        if (*errp) {
use local err?

Thanks

Eric
> +            return;
> +        }
> +    }
> +
> +    group = vfio_ccw_get_group(cdev, errp);
> +    if (!group) {
> +        goto out_group_err;
> +    }
> +
> +    vcdev->vdev.ops = &vfio_ccw_ops;
> +    vcdev->vdev.type = VFIO_DEVICE_TYPE_CCW;
> +    vcdev->vdev.name = g_strdup_printf("%x.%x.%04x", cdev->hostid.cssid,
> +                                       cdev->hostid.ssid, cdev->hostid.devid);
> +    QLIST_FOREACH(vbasedev, &group->device_list, next) {
> +        if (strcmp(vbasedev->name, vcdev->vdev.name) == 0) {
> +            error_setg(errp, "vfio: subchannel %s has already been attached",
> +                       vcdev->vdev.name);
> +            goto out_device_err;
> +        }
> +    }
> +
> +    if (vfio_get_device(group, cdev->mdevid, &vcdev->vdev, errp)) {
> +        goto out_device_err;
> +    }
> +
> +    return;
> +
> +out_device_err:
> +    vfio_ccw_put_group(group);
> +out_group_err:
> +    if (cdc->unrealize) {
> +        cdc->unrealize(cdev, errp);
> +    }
> +}
> +
> +static void vfio_ccw_unrealize(DeviceState *dev, Error **errp)
> +{
> +    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
> +    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
> +    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
> +    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
> +    VFIOGroup *group = vcdev->vdev.group;
> +
> +    vfio_put_device(vcdev);
> +    vfio_put_group(group);
> +
> +    if (cdc->unrealize) {
> +        cdc->unrealize(cdev, errp);
> +    }
> +}
> +
> +static Property vfio_ccw_properties[] = {
> +    DEFINE_PROP_STRING("sysfsdev", VFIOCCWDevice, vdev.sysfsdev),
> +    DEFINE_PROP_END_OF_LIST(),
> +};
> +
> +static const VMStateDescription vfio_ccw_vmstate = {
> +    .name = TYPE_VFIO_CCW,
> +    .unmigratable = 1,
> +};
> +
> +static void vfio_ccw_class_init(ObjectClass *klass, void *data)
> +{
> +    DeviceClass *dc = DEVICE_CLASS(klass);
> +
> +    dc->props = vfio_ccw_properties;
> +    dc->vmsd = &vfio_ccw_vmstate;
> +    dc->desc = "VFIO-based subchannel assignment";
> +    dc->realize = vfio_ccw_realize;
> +    dc->unrealize = vfio_ccw_unrealize;
> +    dc->reset = vfio_ccw_reset;
> +}
> +
> +static const TypeInfo vfio_ccw_info = {
> +    .name = TYPE_VFIO_CCW,
> +    .parent = TYPE_S390_CCW,
> +    .instance_size = sizeof(VFIOCCWDevice),
> +    .class_init = vfio_ccw_class_init,
> +};
> +
> +static void register_vfio_ccw_type(void)
> +{
> +    type_register_static(&vfio_ccw_info);
> +}
> +
> +type_init(register_vfio_ccw_type)
> diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
> index c582de1..9521013 100644
> --- a/include/hw/vfio/vfio-common.h
> +++ b/include/hw/vfio/vfio-common.h
> @@ -45,6 +45,7 @@
>  enum {
>      VFIO_DEVICE_TYPE_PCI = 0,
>      VFIO_DEVICE_TYPE_PLATFORM = 1,
> +    VFIO_DEVICE_TYPE_CCW = 2,
>  };
>  
>  typedef struct VFIOMmap {
>
Dong Jia Shi May 2, 2017, 5:07 a.m. UTC | #2 
* Auger Eric <eric.auger@redhat.com> [2017-05-01 13:08:05 +0200]:

Hi Eric,

[...]

> > diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c
> > new file mode 100644
> > index 0000000..cd4dfe8
> > --- /dev/null
> > +++ b/hw/vfio/ccw.c
[...]

> > +static VFIOGroup *vfio_ccw_get_group(S390CCWDevice *cdev, Error **errp)
> > +{
> > +    char *tmp, group_path[PATH_MAX];
> > +    ssize_t len;
> > +    int groupid;
> > +
> > +    tmp = g_strdup_printf("/sys/bus/css/devices/%x.%x.%04x/%s/iommu_group",
> > +                          cdev->hostid.cssid, cdev->hostid.ssid,
> > +                          cdev->hostid.devid, cdev->mdevid);
> May be worth replacing this static function by a common
> vfio_get_device_group(VFIODevice *vbasedev, Error **errp) in common.c
> that would be used by vfio/pci devices? It would use the sysfsdev
> populated before.
Are you propsing that we introduce a common interface for all of the
devices which use the quite alike logic of populating the sysfsdev?

I put this on my todo list, and prefer to defer it to a time after this
set upstreamed.

> > +    len = readlink(tmp, group_path, sizeof(group_path));
> > +    g_free(tmp);
> > +
> > +    if (len <= 0 || len >= sizeof(group_path)) {
> > +        error_setg(errp, "vfio: no iommu_group found");
> may be good to align the error message with pci/platform. I noticed the
> case where len==0 is handled as ENAMETOOLONG on pci/platform which looks
> bad. Maybe this is why you changed it.
I copy this from pci.c. :>

And I agree with you that, the error message is better to be aligned if
it's possible to do that.

> > +        return NULL;
> > +    }
> > +
> > +    group_path[len] = 0;
> > +
> > +    if (sscanf(basename(group_path), "%d", &groupid) != 1) {
> > +        error_setg(errp, "vfio: failed to read %s", group_path);
> > +        return NULL;
> > +    }
> > +
> > +    return vfio_get_group(groupid, &address_space_memory, errp);
> > +}
> > +
> > +static void vfio_ccw_put_group(VFIOGroup *group)
> Is it really needed?
Ok. Removed.

> > +{
> > +    vfio_put_group(group);
> > +}
> > +
> > +static void vfio_ccw_realize(DeviceState *dev, Error **errp)
> > +{
> > +    VFIODevice *vbasedev;
> > +    VFIOGroup *group;
> > +    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
> > +    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
> > +    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
> > +    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
> > +
> > +    /* Call the class init function for subchannel. */
> > +    if (cdc->realize) {
> > +        cdc->realize(cdev, vcdev->vdev.sysfsdev, errp);
> > +        if (*errp) {
> use local err?
Ok.

> 
> Thanks
> 
> Eric
> > +            return;
> > +        }
> > +    }
> > +
> > +    group = vfio_ccw_get_group(cdev, errp);
> > +    if (!group) {
> > +        goto out_group_err;
> > +    }
> > +
> > +    vcdev->vdev.ops = &vfio_ccw_ops;
> > +    vcdev->vdev.type = VFIO_DEVICE_TYPE_CCW;
> > +    vcdev->vdev.name = g_strdup_printf("%x.%x.%04x", cdev->hostid.cssid,
> > +                                       cdev->hostid.ssid, cdev->hostid.devid);
> > +    QLIST_FOREACH(vbasedev, &group->device_list, next) {
> > +        if (strcmp(vbasedev->name, vcdev->vdev.name) == 0) {
> > +            error_setg(errp, "vfio: subchannel %s has already been attached",
> > +                       vcdev->vdev.name);
> > +            goto out_device_err;
> > +        }
> > +    }
> > +
> > +    if (vfio_get_device(group, cdev->mdevid, &vcdev->vdev, errp)) {
> > +        goto out_device_err;
> > +    }
> > +
> > +    return;
> > +
> > +out_device_err:
> > +    vfio_ccw_put_group(group);
> > +out_group_err:
> > +    if (cdc->unrealize) {
> > +        cdc->unrealize(cdev, errp);
> > +    }
> > +}
> > +
[...]
Eric Auger May 2, 2017, 6:48 a.m. UTC | #3 
Hi Dong,

On 02/05/2017 07:07, Dong Jia Shi wrote:
> * Auger Eric <eric.auger@redhat.com> [2017-05-01 13:08:05 +0200]:
> 
> Hi Eric,
> 
> [...]
> 
>>> diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c
>>> new file mode 100644
>>> index 0000000..cd4dfe8
>>> --- /dev/null
>>> +++ b/hw/vfio/ccw.c
> [...]
> 
>>> +static VFIOGroup *vfio_ccw_get_group(S390CCWDevice *cdev, Error **errp)
>>> +{
>>> +    char *tmp, group_path[PATH_MAX];
>>> +    ssize_t len;
>>> +    int groupid;
>>> +
>>> +    tmp = g_strdup_printf("/sys/bus/css/devices/%x.%x.%04x/%s/iommu_group",
>>> +                          cdev->hostid.cssid, cdev->hostid.ssid,
>>> +                          cdev->hostid.devid, cdev->mdevid);
>> May be worth replacing this static function by a common
>> vfio_get_device_group(VFIODevice *vbasedev, Error **errp) in common.c
>> that would be used by vfio/pci devices? It would use the sysfsdev
>> populated before.
> Are you propsing that we introduce a common interface for all of the
> devices which use the quite alike logic of populating the sysfsdev?
Yes that's what I meant.
> 
> I put this on my todo list, and prefer to defer it to a time after this
> set upstreamed.
OK no worries.
> 
>>> +    len = readlink(tmp, group_path, sizeof(group_path));
>>> +    g_free(tmp);
>>> +
>>> +    if (len <= 0 || len >= sizeof(group_path)) {
>>> +        error_setg(errp, "vfio: no iommu_group found");
>> may be good to align the error message with pci/platform. I noticed the
>> case where len==0 is handled as ENAMETOOLONG on pci/platform which looks
>> bad. Maybe this is why you changed it.
> I copy this from pci.c. :>
yep I know ;-)

Thanks

Eric
> 
> And I agree with you that, the error message is better to be aligned if
> it's possible to do that.
> 
>>> +        return NULL;
>>> +    }
>>> +
>>> +    group_path[len] = 0;
>>> +
>>> +    if (sscanf(basename(group_path), "%d", &groupid) != 1) {
>>> +        error_setg(errp, "vfio: failed to read %s", group_path);
>>> +        return NULL;
>>> +    }
>>> +
>>> +    return vfio_get_group(groupid, &address_space_memory, errp);
>>> +}
>>> +
>>> +static void vfio_ccw_put_group(VFIOGroup *group)
>> Is it really needed?
> Ok. Removed.
> 
>>> +{
>>> +    vfio_put_group(group);
>>> +}
>>> +
>>> +static void vfio_ccw_realize(DeviceState *dev, Error **errp)
>>> +{
>>> +    VFIODevice *vbasedev;
>>> +    VFIOGroup *group;
>>> +    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
>>> +    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
>>> +    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
>>> +    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
>>> +
>>> +    /* Call the class init function for subchannel. */
>>> +    if (cdc->realize) {
>>> +        cdc->realize(cdev, vcdev->vdev.sysfsdev, errp);
>>> +        if (*errp) {
>> use local err?
> Ok.
> 
>>
>> Thanks
>>
>> Eric
>>> +            return;
>>> +        }
>>> +    }
>>> +
>>> +    group = vfio_ccw_get_group(cdev, errp);
>>> +    if (!group) {
>>> +        goto out_group_err;
>>> +    }
>>> +
>>> +    vcdev->vdev.ops = &vfio_ccw_ops;
>>> +    vcdev->vdev.type = VFIO_DEVICE_TYPE_CCW;
>>> +    vcdev->vdev.name = g_strdup_printf("%x.%x.%04x", cdev->hostid.cssid,
>>> +                                       cdev->hostid.ssid, cdev->hostid.devid);
>>> +    QLIST_FOREACH(vbasedev, &group->device_list, next) {
>>> +        if (strcmp(vbasedev->name, vcdev->vdev.name) == 0) {
>>> +            error_setg(errp, "vfio: subchannel %s has already been attached",
>>> +                       vcdev->vdev.name);
>>> +            goto out_device_err;
>>> +        }
>>> +    }
>>> +
>>> +    if (vfio_get_device(group, cdev->mdevid, &vcdev->vdev, errp)) {
>>> +        goto out_device_err;
>>> +    }
>>> +
>>> +    return;
>>> +
>>> +out_device_err:
>>> +    vfio_ccw_put_group(group);
>>> +out_group_err:
>>> +    if (cdc->unrealize) {
>>> +        cdc->unrealize(cdev, errp);
>>> +    }
>>> +}
>>> +
> [...]
>
diff mbox

Patch

diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
index 36e15de..5576b0a 100644
--- a/default-configs/s390x-softmmu.mak
+++ b/default-configs/s390x-softmmu.mak
@@ -4,4 +4,5 @@  CONFIG_VIRTIO=y
 CONFIG_SCLPCONSOLE=y
 CONFIG_S390_FLIC=y
 CONFIG_S390_FLIC_KVM=$(CONFIG_KVM)
+CONFIG_VFIO_CCW=$(CONFIG_LINUX)
 CONFIG_WDT_DIAG288=y
diff --git a/hw/vfio/Makefile.objs b/hw/vfio/Makefile.objs
index 05e7fbb..c3ab909 100644
--- a/hw/vfio/Makefile.objs
+++ b/hw/vfio/Makefile.objs
@@ -1,6 +1,7 @@ 
 ifeq ($(CONFIG_LINUX), y)
 obj-$(CONFIG_SOFTMMU) += common.o
 obj-$(CONFIG_PCI) += pci.o pci-quirks.o
+obj-$(CONFIG_VFIO_CCW) += ccw.o
 obj-$(CONFIG_SOFTMMU) += platform.o
 obj-$(CONFIG_VFIO_XGMAC) += calxeda-xgmac.o
 obj-$(CONFIG_VFIO_AMD_XGBE) += amd-xgbe.o
diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c
new file mode 100644
index 0000000..cd4dfe8
--- /dev/null
+++ b/hw/vfio/ccw.c
@@ -0,0 +1,189 @@ 
+/*
+ * vfio based subchannel assignment support
+ *
+ * Copyright 2017 IBM Corp.
+ * Author(s): Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
+ *            Xiao Feng Ren <renxiaof@linux.vnet.ibm.com>
+ *            Pierre Morel <pmorel@linux.vnet.ibm.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or(at
+ * your option) any version. See the COPYING file in the top-level
+ * directory.
+ */
+
+#include <linux/vfio.h>
+#include <sys/ioctl.h>
+
+#include "qemu/osdep.h"
+#include "qapi/error.h"
+#include "hw/sysbus.h"
+#include "hw/vfio/vfio.h"
+#include "hw/vfio/vfio-common.h"
+#include "hw/s390x/s390-ccw.h"
+#include "hw/s390x/ccw-device.h"
+
+#define TYPE_VFIO_CCW "vfio-ccw"
+typedef struct VFIOCCWDevice {
+    S390CCWDevice cdev;
+    VFIODevice vdev;
+} VFIOCCWDevice;
+
+static void vfio_ccw_compute_needs_reset(VFIODevice *vdev)
+{
+    vdev->needs_reset = false;
+}
+
+/*
+ * We don't need vfio_hot_reset_multi and vfio_eoi operations for
+ * vfio_ccw device now.
+ */
+struct VFIODeviceOps vfio_ccw_ops = {
+    .vfio_compute_needs_reset = vfio_ccw_compute_needs_reset,
+};
+
+static void vfio_ccw_reset(DeviceState *dev)
+{
+    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
+    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
+    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
+
+    ioctl(vcdev->vdev.fd, VFIO_DEVICE_RESET);
+}
+
+static void vfio_put_device(VFIOCCWDevice *vcdev)
+{
+    g_free(vcdev->vdev.name);
+    vfio_put_base_device(&vcdev->vdev);
+}
+
+static VFIOGroup *vfio_ccw_get_group(S390CCWDevice *cdev, Error **errp)
+{
+    char *tmp, group_path[PATH_MAX];
+    ssize_t len;
+    int groupid;
+
+    tmp = g_strdup_printf("/sys/bus/css/devices/%x.%x.%04x/%s/iommu_group",
+                          cdev->hostid.cssid, cdev->hostid.ssid,
+                          cdev->hostid.devid, cdev->mdevid);
+    len = readlink(tmp, group_path, sizeof(group_path));
+    g_free(tmp);
+
+    if (len <= 0 || len >= sizeof(group_path)) {
+        error_setg(errp, "vfio: no iommu_group found");
+        return NULL;
+    }
+
+    group_path[len] = 0;
+
+    if (sscanf(basename(group_path), "%d", &groupid) != 1) {
+        error_setg(errp, "vfio: failed to read %s", group_path);
+        return NULL;
+    }
+
+    return vfio_get_group(groupid, &address_space_memory, errp);
+}
+
+static void vfio_ccw_put_group(VFIOGroup *group)
+{
+    vfio_put_group(group);
+}
+
+static void vfio_ccw_realize(DeviceState *dev, Error **errp)
+{
+    VFIODevice *vbasedev;
+    VFIOGroup *group;
+    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
+    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
+    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
+    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
+
+    /* Call the class init function for subchannel. */
+    if (cdc->realize) {
+        cdc->realize(cdev, vcdev->vdev.sysfsdev, errp);
+        if (*errp) {
+            return;
+        }
+    }
+
+    group = vfio_ccw_get_group(cdev, errp);
+    if (!group) {
+        goto out_group_err;
+    }
+
+    vcdev->vdev.ops = &vfio_ccw_ops;
+    vcdev->vdev.type = VFIO_DEVICE_TYPE_CCW;
+    vcdev->vdev.name = g_strdup_printf("%x.%x.%04x", cdev->hostid.cssid,
+                                       cdev->hostid.ssid, cdev->hostid.devid);
+    QLIST_FOREACH(vbasedev, &group->device_list, next) {
+        if (strcmp(vbasedev->name, vcdev->vdev.name) == 0) {
+            error_setg(errp, "vfio: subchannel %s has already been attached",
+                       vcdev->vdev.name);
+            goto out_device_err;
+        }
+    }
+
+    if (vfio_get_device(group, cdev->mdevid, &vcdev->vdev, errp)) {
+        goto out_device_err;
+    }
+
+    return;
+
+out_device_err:
+    vfio_ccw_put_group(group);
+out_group_err:
+    if (cdc->unrealize) {
+        cdc->unrealize(cdev, errp);
+    }
+}
+
+static void vfio_ccw_unrealize(DeviceState *dev, Error **errp)
+{
+    CcwDevice *ccw_dev = DO_UPCAST(CcwDevice, parent_obj, dev);
+    S390CCWDevice *cdev = DO_UPCAST(S390CCWDevice, parent_obj, ccw_dev);
+    VFIOCCWDevice *vcdev = DO_UPCAST(VFIOCCWDevice, cdev, cdev);
+    S390CCWDeviceClass *cdc = S390_CCW_DEVICE_GET_CLASS(cdev);
+    VFIOGroup *group = vcdev->vdev.group;
+
+    vfio_put_device(vcdev);
+    vfio_put_group(group);
+
+    if (cdc->unrealize) {
+        cdc->unrealize(cdev, errp);
+    }
+}
+
+static Property vfio_ccw_properties[] = {
+    DEFINE_PROP_STRING("sysfsdev", VFIOCCWDevice, vdev.sysfsdev),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static const VMStateDescription vfio_ccw_vmstate = {
+    .name = TYPE_VFIO_CCW,
+    .unmigratable = 1,
+};
+
+static void vfio_ccw_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+
+    dc->props = vfio_ccw_properties;
+    dc->vmsd = &vfio_ccw_vmstate;
+    dc->desc = "VFIO-based subchannel assignment";
+    dc->realize = vfio_ccw_realize;
+    dc->unrealize = vfio_ccw_unrealize;
+    dc->reset = vfio_ccw_reset;
+}
+
+static const TypeInfo vfio_ccw_info = {
+    .name = TYPE_VFIO_CCW,
+    .parent = TYPE_S390_CCW,
+    .instance_size = sizeof(VFIOCCWDevice),
+    .class_init = vfio_ccw_class_init,
+};
+
+static void register_vfio_ccw_type(void)
+{
+    type_register_static(&vfio_ccw_info);
+}
+
+type_init(register_vfio_ccw_type)
diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
index c582de1..9521013 100644
--- a/include/hw/vfio/vfio-common.h
+++ b/include/hw/vfio/vfio-common.h
@@ -45,6 +45,7 @@ 
 enum {
     VFIO_DEVICE_TYPE_PCI = 0,
     VFIO_DEVICE_TYPE_PLATFORM = 1,
+    VFIO_DEVICE_TYPE_CCW = 2,
 };
 
 typedef struct VFIOMmap {