Skip to contentSkip to conversation

mdev based hardware virtio offloading support

PATCH Superseded Jason Wang 4 participants 43 messages Nov 6, 2019 – Nov 7, 2019

Subsystems (16)

v10 of 13 Superseded Latest is v13 →

lore
Share & tools
lore

Jason Wang

author
#0 lore raw

v10 cover mdev based hardware virtio offloading support

Hi all:

There are hardwares that can do virtio datapath offloading while having its own control path. This path tries to implement a mdev based unified API to support using kernel virtio driver to drive those devices. This is done by introducing a new mdev transport for virtio (virtio_mdev) and register itself as a new kind of mdev driver. Then it provides a unified way for kernel virtio driver to talk with mdev device implementation.

Though the series only contains kernel driver support, the goal is to make the transport generic enough to support userspace drivers. This means vhost-mdev[1] could be built on top as well by resuing the transport.

A sample driver is also implemented which simulate a virito-net loopback ethernet device on top of vringh + workqueue. This could be used as a reference implementation for real hardware driver.

Also a real IFC VF driver was also posted here[2] which is a good reference for vendors who is interested in their own virtio datapath offloading product.

Consider mdev framework only support VFIO device and driver right now, this series also extend it to support other types. This is done through introducing class id to the device and pairing it with id_talbe claimed by the driver. On top, this seris also decouple device specific ops out of the common ones for implementing class specific operations over mdev bus.

Pktgen test was done with virito-net + mvnet loop back device.

Please review.

[1] https://lkml.org/lkml/2019/11/5/424
[2] https://lkml.org/lkml/2019/11/5/227

Changes from V9:
- Tweak the help text for virito-mdev kconfig

Changes from V8:
- try silent checkpatch, some are still there becuase they were inherited from virtio_config_ops which needs to be resolved in an independent series
- tweak on the comment and doc
- remove VIRTIO_MDEV_F_VERSION_1 completely
- rename CONFIG_VIRTIO_MDEV_DEVICE to CONFIG_VIRTIO_MDEV

Changes from V7:
- drop {set|get}_mdev_features for virtio
- typo and comment style fixes

Changes from V6:
- rename ops files and compile guard

Changes from V5:
- use dev_warn() instead of WARN(1) when class id is not set
- validate id_table before trying to do matching between device and driver
- add wildcard for modpost script
- use unique name for id_table
- move get_mdev_features() to be the first member of virtio_device_ops and more comments for it
- typo fixes for the comments above virtio_mdev_ops

Changes from V4:
- keep mdev_set_class() for the device that doesn't use device ops
- use union for device ops pointer in mdev_device
- introduce class specific helper for getting is device ops
- use WARN_ON instead of BUG_ON in mdev_set_virtio_ops
- explain details of get_mdev_features() and get_vendor_id()
- distinguish the optional virito device ops from mandatory ones and make get_generation() optional
- rename vfio_mdev.h to vfio_mdev_ops.h, rename virito_mdev.h to virtio_mdev_ops.h
- don't abuse version fileds in virtio_mdev structure, use features instead
- fix warning during device remove
- style & docs tweaks and typo fixes

Changes from V3:
- document that class id (device ops) must be specified in create()
- add WARN() when trying to set class_id when it has already set
- add WARN() when class_id is not specified in create() and correctly return an error in this case
- correct the prototype of mdev_set_class() in the doc
- add documention of mdev_set_class()
- remove the unnecessary "class_id_fail" label when class id is not specified in create()
- convert id_table in vfio_mdev to const
- move mdev_set_class and its friends after mdev_uuid()
- suqash the patch of bus uevent into patch of introducing class id
- tweak the words in the docs per Cornelia suggestion
- tie class_id and device ops through class specific initialization routine like mdev_set_vfio_ops()
- typos fixes in the docs of virtio-mdev callbacks
- document the usage of virtqueues in struct virtio_mdev_device
- remove the useless vqs array in struct virtio_mdev_device
- rename MDEV_ID_XXX to MDEV_CLASS_ID_XXX

Changes from V2:
- fail when class_id is not specified
- drop the vringh patch
- match the doc to the code
- tweak the commit log
- move device_ops from parent to mdev device
- remove the unused MDEV_ID_VHOST

Changes from V1:
- move virtio_mdev.c to drivers/virtio
- store class_id in mdev_device instead of mdev_parent
- store device_ops in mdev_device instead of mdev_parent
- reorder the patch, vringh fix comes first
- really silent compiling warnings
- really switch to use u16 for class_id
- uevent and modpost support for mdev class_id
- vraious tweaks per comments from Parav

Changes from RFC-V2:
- silent compile warnings on some specific configuration
- use u16 instead u8 for class id
- reseve MDEV_ID_VHOST for future vhost-mdev work
- introduce "virtio" type for mvnet and make "vhost" type for future work
- add entries in MAINTAINER
- tweak and typos fixes in commit log

Changes from RFC-V1:
- rename device id to class id
- add docs for class id and device specific ops (device_ops)
- split device_ops into seperate headers
- drop the mdev_set_dma_ops()
- use device_ops to implement the transport API, then it's not a part of UAPI any more
- use GFP_ATOMIC in mvnet sample device and other tweaks
- set_vring_base/get_vring_base support for mvnet device

Jason Wang (6):
mdev: class id support
modpost: add support for mdev class id
mdev: introduce device specific ops
mdev: introduce virtio device and its device ops
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework

 .../driver-api/vfio-mediated-device.rst       |  38 +-
 MAINTAINERS                                   |   3 +
 drivers/gpu/drm/i915/gvt/kvmgt.c              |  17 +-
 drivers/s390/cio/vfio_ccw_ops.c               |  17 +-
 drivers/s390/crypto/vfio_ap_ops.c             |  13 +-
 drivers/vfio/mdev/mdev_core.c                 |  60 ++
 drivers/vfio/mdev/mdev_driver.c               |  25 +
 drivers/vfio/mdev/mdev_private.h              |   8 +
 drivers/vfio/mdev/vfio_mdev.c                 |  45 +-
 drivers/virtio/Kconfig                        |  13 +
 drivers/virtio/Makefile                       |   1 +
 drivers/virtio/virtio_mdev.c                  | 406 +++++++++++
 include/linux/mdev.h                          |  57 +-
 include/linux/mdev_vfio_ops.h                 |  52 ++
 include/linux/mdev_virtio_ops.h               | 147 ++++
 include/linux/mod_devicetable.h               |   8 +
 samples/Kconfig                               |  10 +
 samples/vfio-mdev/Makefile                    |   1 +
 samples/vfio-mdev/mbochs.c                    |  19 +-
 samples/vfio-mdev/mdpy.c                      |  19 +-
 samples/vfio-mdev/mtty.c                      |  17 +-
 samples/vfio-mdev/mvnet.c                     | 686 ++++++++++++++++++
 scripts/mod/devicetable-offsets.c             |   3 +
 scripts/mod/file2alias.c                      |  11 +
 24 files changed, 1585 insertions(+), 91 deletions(-)
 create mode 100644 drivers/virtio/virtio_mdev.c
 create mode 100644 include/linux/mdev_vfio_ops.h
 create mode 100644 include/linux/mdev_virtio_ops.h
 create mode 100644 samples/vfio-mdev/mvnet.c
Signature · 1 line
2.19.1
author
#1 lore raw

PATCH v10 1/6 mdev: class id support

Mdev bus only supports vfio driver right now, so it doesn't implement match method. But in the future, we may add drivers other than vfio, the first driver could be virtio-mdev. This means we need to add device class id support in bus match method to pair the mdev device and mdev driver correctly.

So this patch adds id_table to mdev_driver and class_id for mdev device with the match method for mdev bus.

---
 .../driver-api/vfio-mediated-device.rst       |  5 ++++
 drivers/gpu/drm/i915/gvt/kvmgt.c              |  1 +
 drivers/s390/cio/vfio_ccw_ops.c               |  1 +
 drivers/s390/crypto/vfio_ap_ops.c             |  1 +
 drivers/vfio/mdev/mdev_core.c                 | 17 +++++++++++++
 drivers/vfio/mdev/mdev_driver.c               | 25 +++++++++++++++++++
 drivers/vfio/mdev/mdev_private.h              |  1 +
 drivers/vfio/mdev/vfio_mdev.c                 |  6 +++++
 include/linux/mdev.h                          |  8 ++++++
 include/linux/mod_devicetable.h               |  8 ++++++
 samples/vfio-mdev/mbochs.c                    |  1 +
 samples/vfio-mdev/mdpy.c                      |  1 +
 samples/vfio-mdev/mtty.c                      |  1 +
 13 files changed, 76 insertions(+)
Documentation/driver-api/vfio-mediated-device.rst +5 -0
index 25eb7d5b834b..6709413bee29 100644
@@ -102,12 +102,14 @@ structure to represent a mediated device's driver::
       * @probe: called when new device created
       * @remove: called when device removed
       * @driver: device driver structure
+      * @id_table: the ids serviced by this driver
       */
      struct mdev_driver {
 	     const char *name;
 	     int  (*probe)  (struct device *dev);
 	     void (*remove) (struct device *dev);
 	     struct device_driver    driver;
+	     const struct mdev_class_id *id_table;
      };
 
 A mediated bus driver for mdev should use this structure in the function calls
@@ -170,6 +172,9 @@ that a driver should use to unregister itself with the mdev core driver::
 
 	extern void mdev_unregister_device(struct device *dev);
 
+It is also required to specify the class_id in create() callback through::
+
+	int mdev_set_class(struct mdev_device *mdev, u16 id);
 
 Mediated Device Management Interface Through sysfs
 ==================================================
drivers/gpu/drm/i915/gvt/kvmgt.c +1 -0
index 343d79c1cb7e..6420f0dbd31b 100644
@@ -678,6 +678,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 		     dev_name(mdev_dev(mdev)));
 	ret = 0;
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 out:
 	return ret;
 }
drivers/s390/cio/vfio_ccw_ops.c +1 -0
index f0d71ab77c50..cf2c013ae32f 100644
@@ -129,6 +129,7 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 			   private->sch->schid.ssid,
 			   private->sch->schid.sch_no);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
drivers/s390/crypto/vfio_ap_ops.c +1 -0
index 5c0f53c6dde7..07c31070afeb 100644
@@ -343,6 +343,7 @@ static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&matrix_mdev->node, &matrix_dev->mdev_list);
 	mutex_unlock(&matrix_dev->lock);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
drivers/vfio/mdev/mdev_core.c +17 -0
index b558d4cfd082..7bfa2e46e829 100644
@@ -45,6 +45,17 @@ void mdev_set_drvdata(struct mdev_device *mdev, void *data)
 }
 EXPORT_SYMBOL(mdev_set_drvdata);
 
+/*
+ * Specify the class for the mdev device, this must be called during
+ * create() callback.
+ */
+void mdev_set_class(struct mdev_device *mdev, u16 id)
+{
+	WARN_ON(mdev->class_id);
+	mdev->class_id = id;
+}
+EXPORT_SYMBOL(mdev_set_class);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
@@ -324,6 +335,12 @@ int mdev_device_create(struct kobject *kobj,
 	if (ret)
 		goto ops_create_fail;
 
+	if (!mdev->class_id) {
+		ret = -EINVAL;
+		dev_warn(dev, "mdev vendor driver failed to specify device class\n");
+		goto add_fail;
+	}
+
 	ret = device_add(&mdev->dev);
 	if (ret)
 		goto add_fail;
drivers/vfio/mdev/mdev_driver.c +25 -0
index 0d3223aee20b..ed06433693e8 100644
@@ -69,8 +69,33 @@ static int mdev_remove(struct device *dev)
 	return 0;
 }
 
+static int mdev_match(struct device *dev, struct device_driver *drv)
+{
+	unsigned int i;
+	struct mdev_device *mdev = to_mdev_device(dev);
+	struct mdev_driver *mdrv = to_mdev_driver(drv);
+	const struct mdev_class_id *ids = mdrv->id_table;
+
+	if (!ids)
+		return 0;
+
+	for (i = 0; ids[i].id; i++)
+		if (ids[i].id == mdev->class_id)
+			return 1;
+	return 0;
+}
+
+static int mdev_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+	struct mdev_device *mdev = to_mdev_device(dev);
+
+	return add_uevent_var(env, "MODALIAS=mdev:c%02X", mdev->class_id);
+}
+
 struct bus_type mdev_bus_type = {
 	.name		= "mdev",
+	.match		= mdev_match,
+	.uevent		= mdev_uevent,
 	.probe		= mdev_probe,
 	.remove		= mdev_remove,
 };
drivers/vfio/mdev/mdev_private.h +1 -0
index 7d922950caaf..c65f436c1869 100644
@@ -33,6 +33,7 @@ struct mdev_device {
 	struct kobject *type_kobj;
 	struct device *iommu_device;
 	bool active;
+	u16 class_id;
 };
 
 #define to_mdev_device(dev)	container_of(dev, struct mdev_device, dev)
drivers/vfio/mdev/vfio_mdev.c +6 -0
index 30964a4e0a28..38431e9ef7f5 100644
@@ -120,10 +120,16 @@ static void vfio_mdev_remove(struct device *dev)
 	vfio_del_group_dev(dev);
 }
 
+static const struct mdev_class_id vfio_id_table[] = {
+	{ MDEV_CLASS_ID_VFIO },
+	{ 0 },
+};
+
 static struct mdev_driver vfio_mdev_driver = {
 	.name	= "vfio_mdev",
 	.probe	= vfio_mdev_probe,
 	.remove	= vfio_mdev_remove,
+	.id_table = vfio_id_table,
 };
 
 static int __init vfio_mdev_init(void)
include/linux/mdev.h +8 -0
index 0ce30ca78db0..78b69d09eb54 100644
@@ -118,6 +118,7 @@ struct mdev_type_attribute mdev_type_attr_##_name =		\
  * @probe: called when new device created
  * @remove: called when device removed
  * @driver: device driver structure
+ * @id_table: the ids serviced by this driver
  *
  **/
 struct mdev_driver {
@@ -125,6 +126,7 @@ struct mdev_driver {
 	int  (*probe)(struct device *dev);
 	void (*remove)(struct device *dev);
 	struct device_driver driver;
+	const struct mdev_class_id *id_table;
 };
 
 #define to_mdev_driver(drv)	container_of(drv, struct mdev_driver, driver)
@@ -132,6 +134,7 @@ struct mdev_driver {
 void *mdev_get_drvdata(struct mdev_device *mdev);
 void mdev_set_drvdata(struct mdev_device *mdev, void *data);
 const guid_t *mdev_uuid(struct mdev_device *mdev);
+void mdev_set_class(struct mdev_device *mdev, u16 id);
 
 extern struct bus_type mdev_bus_type;
 
@@ -145,4 +148,9 @@ struct device *mdev_parent_dev(struct mdev_device *mdev);
 struct device *mdev_dev(struct mdev_device *mdev);
 struct mdev_device *mdev_from_dev(struct device *dev);
 
+enum {
+	MDEV_CLASS_ID_VFIO = 1,
+	/* New entries must be added here */
+};
+
 #endif /* MDEV_H */
include/linux/mod_devicetable.h +8 -0
index 5714fd35a83c..f32c6e44fb1a 100644
@@ -821,4 +821,12 @@ struct wmi_device_id {
 	const void *context;
 };
 
+/**
+ * struct mdev_class_id - MDEV device class identifier
+ * @id: Used to identify a specific class of device, e.g vfio-mdev device.
+ */
+struct mdev_class_id {
+	__u16 id;
+};
+
 #endif /* LINUX_MOD_DEVICETABLE_H */
samples/vfio-mdev/mbochs.c +1 -0
index ac5c8c17b1ff..115bc5074656 100644
@@ -561,6 +561,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 	mbochs_reset(mdev);
 
 	mbochs_used_mbytes += type->mbytes;
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 
 err_mem:
samples/vfio-mdev/mdpy.c +1 -0
index cc86bf6566e4..665614574d50 100644
@@ -269,6 +269,7 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 	mdpy_reset(mdev);
 
 	mdpy_count++;
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
samples/vfio-mdev/mtty.c +1 -0
index ce84a300a4da..90da12ff7fd9 100644
@@ -755,6 +755,7 @@ static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&mdev_state->next, &mdev_devices_list);
 	mutex_unlock(&mdev_list_lock);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
Signature · 1 line
2.19.1
author
#2 lore raw

PATCH v10 1/6 mdev: class id support

In reply to Jason Wang

Mdev bus only supports vfio driver right now, so it doesn't implement match method. But in the future, we may add drivers other than vfio, the first driver could be virtio-mdev. This means we need to add device class id support in bus match method to pair the mdev device and mdev driver correctly.

So this patch adds id_table to mdev_driver and class_id for mdev device with the match method for mdev bus.

---
 .../driver-api/vfio-mediated-device.rst       |  5 ++++
 drivers/gpu/drm/i915/gvt/kvmgt.c              |  1 +
 drivers/s390/cio/vfio_ccw_ops.c               |  1 +
 drivers/s390/crypto/vfio_ap_ops.c             |  1 +
 drivers/vfio/mdev/mdev_core.c                 | 17 +++++++++++++
 drivers/vfio/mdev/mdev_driver.c               | 25 +++++++++++++++++++
 drivers/vfio/mdev/mdev_private.h              |  1 +
 drivers/vfio/mdev/vfio_mdev.c                 |  6 +++++
 include/linux/mdev.h                          |  8 ++++++
 include/linux/mod_devicetable.h               |  8 ++++++
 samples/vfio-mdev/mbochs.c                    |  1 +
 samples/vfio-mdev/mdpy.c                      |  1 +
 samples/vfio-mdev/mtty.c                      |  1 +
 13 files changed, 76 insertions(+)
Documentation/driver-api/vfio-mediated-device.rst +5 -0
index 25eb7d5b834b..6709413bee29 100644
@@ -102,12 +102,14 @@ structure to represent a mediated device's driver::
       * @probe: called when new device created
       * @remove: called when device removed
       * @driver: device driver structure
+      * @id_table: the ids serviced by this driver
       */
      struct mdev_driver {
 	     const char *name;
 	     int  (*probe)  (struct device *dev);
 	     void (*remove) (struct device *dev);
 	     struct device_driver    driver;
+	     const struct mdev_class_id *id_table;
      };
 
 A mediated bus driver for mdev should use this structure in the function calls
@@ -170,6 +172,9 @@ that a driver should use to unregister itself with the mdev core driver::
 
 	extern void mdev_unregister_device(struct device *dev);
 
+It is also required to specify the class_id in create() callback through::
+
+	int mdev_set_class(struct mdev_device *mdev, u16 id);
 
 Mediated Device Management Interface Through sysfs
 ==================================================
drivers/gpu/drm/i915/gvt/kvmgt.c +1 -0
index 343d79c1cb7e..6420f0dbd31b 100644
@@ -678,6 +678,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 		     dev_name(mdev_dev(mdev)));
 	ret = 0;
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 out:
 	return ret;
 }
drivers/s390/cio/vfio_ccw_ops.c +1 -0
index f0d71ab77c50..cf2c013ae32f 100644
@@ -129,6 +129,7 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 			   private->sch->schid.ssid,
 			   private->sch->schid.sch_no);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
drivers/s390/crypto/vfio_ap_ops.c +1 -0
index 5c0f53c6dde7..07c31070afeb 100644
@@ -343,6 +343,7 @@ static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&matrix_mdev->node, &matrix_dev->mdev_list);
 	mutex_unlock(&matrix_dev->lock);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
drivers/vfio/mdev/mdev_core.c +17 -0
index b558d4cfd082..7bfa2e46e829 100644
@@ -45,6 +45,17 @@ void mdev_set_drvdata(struct mdev_device *mdev, void *data)
 }
 EXPORT_SYMBOL(mdev_set_drvdata);
 
+/*
+ * Specify the class for the mdev device, this must be called during
+ * create() callback.
+ */
+void mdev_set_class(struct mdev_device *mdev, u16 id)
+{
+	WARN_ON(mdev->class_id);
+	mdev->class_id = id;
+}
+EXPORT_SYMBOL(mdev_set_class);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
@@ -324,6 +335,12 @@ int mdev_device_create(struct kobject *kobj,
 	if (ret)
 		goto ops_create_fail;
 
+	if (!mdev->class_id) {
+		ret = -EINVAL;
+		dev_warn(dev, "mdev vendor driver failed to specify device class\n");
+		goto add_fail;
+	}
+
 	ret = device_add(&mdev->dev);
 	if (ret)
 		goto add_fail;
drivers/vfio/mdev/mdev_driver.c +25 -0
index 0d3223aee20b..ed06433693e8 100644
@@ -69,8 +69,33 @@ static int mdev_remove(struct device *dev)
 	return 0;
 }
 
+static int mdev_match(struct device *dev, struct device_driver *drv)
+{
+	unsigned int i;
+	struct mdev_device *mdev = to_mdev_device(dev);
+	struct mdev_driver *mdrv = to_mdev_driver(drv);
+	const struct mdev_class_id *ids = mdrv->id_table;
+
+	if (!ids)
+		return 0;
+
+	for (i = 0; ids[i].id; i++)
+		if (ids[i].id == mdev->class_id)
+			return 1;
+	return 0;
+}
+
+static int mdev_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+	struct mdev_device *mdev = to_mdev_device(dev);
+
+	return add_uevent_var(env, "MODALIAS=mdev:c%02X", mdev->class_id);
+}
+
 struct bus_type mdev_bus_type = {
 	.name		= "mdev",
+	.match		= mdev_match,
+	.uevent		= mdev_uevent,
 	.probe		= mdev_probe,
 	.remove		= mdev_remove,
 };
drivers/vfio/mdev/mdev_private.h +1 -0
index 7d922950caaf..c65f436c1869 100644
@@ -33,6 +33,7 @@ struct mdev_device {
 	struct kobject *type_kobj;
 	struct device *iommu_device;
 	bool active;
+	u16 class_id;
 };
 
 #define to_mdev_device(dev)	container_of(dev, struct mdev_device, dev)
drivers/vfio/mdev/vfio_mdev.c +6 -0
index 30964a4e0a28..38431e9ef7f5 100644
@@ -120,10 +120,16 @@ static void vfio_mdev_remove(struct device *dev)
 	vfio_del_group_dev(dev);
 }
 
+static const struct mdev_class_id vfio_id_table[] = {
+	{ MDEV_CLASS_ID_VFIO },
+	{ 0 },
+};
+
 static struct mdev_driver vfio_mdev_driver = {
 	.name	= "vfio_mdev",
 	.probe	= vfio_mdev_probe,
 	.remove	= vfio_mdev_remove,
+	.id_table = vfio_id_table,
 };
 
 static int __init vfio_mdev_init(void)
include/linux/mdev.h +8 -0
index 0ce30ca78db0..78b69d09eb54 100644
@@ -118,6 +118,7 @@ struct mdev_type_attribute mdev_type_attr_##_name =		\
  * @probe: called when new device created
  * @remove: called when device removed
  * @driver: device driver structure
+ * @id_table: the ids serviced by this driver
  *
  **/
 struct mdev_driver {
@@ -125,6 +126,7 @@ struct mdev_driver {
 	int  (*probe)(struct device *dev);
 	void (*remove)(struct device *dev);
 	struct device_driver driver;
+	const struct mdev_class_id *id_table;
 };
 
 #define to_mdev_driver(drv)	container_of(drv, struct mdev_driver, driver)
@@ -132,6 +134,7 @@ struct mdev_driver {
 void *mdev_get_drvdata(struct mdev_device *mdev);
 void mdev_set_drvdata(struct mdev_device *mdev, void *data);
 const guid_t *mdev_uuid(struct mdev_device *mdev);
+void mdev_set_class(struct mdev_device *mdev, u16 id);
 
 extern struct bus_type mdev_bus_type;
 
@@ -145,4 +148,9 @@ struct device *mdev_parent_dev(struct mdev_device *mdev);
 struct device *mdev_dev(struct mdev_device *mdev);
 struct mdev_device *mdev_from_dev(struct device *dev);
 
+enum {
+	MDEV_CLASS_ID_VFIO = 1,
+	/* New entries must be added here */
+};
+
 #endif /* MDEV_H */
include/linux/mod_devicetable.h +8 -0
index 5714fd35a83c..f32c6e44fb1a 100644
@@ -821,4 +821,12 @@ struct wmi_device_id {
 	const void *context;
 };
 
+/**
+ * struct mdev_class_id - MDEV device class identifier
+ * @id: Used to identify a specific class of device, e.g vfio-mdev device.
+ */
+struct mdev_class_id {
+	__u16 id;
+};
+
 #endif /* LINUX_MOD_DEVICETABLE_H */
samples/vfio-mdev/mbochs.c +1 -0
index ac5c8c17b1ff..115bc5074656 100644
@@ -561,6 +561,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 	mbochs_reset(mdev);
 
 	mbochs_used_mbytes += type->mbytes;
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 
 err_mem:
samples/vfio-mdev/mdpy.c +1 -0
index cc86bf6566e4..665614574d50 100644
@@ -269,6 +269,7 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 	mdpy_reset(mdev);
 
 	mdpy_count++;
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
samples/vfio-mdev/mtty.c +1 -0
index ce84a300a4da..90da12ff7fd9 100644
@@ -755,6 +755,7 @@ static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&mdev_state->next, &mdev_devices_list);
 	mutex_unlock(&mdev_list_lock);
 
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
 	return 0;
 }
 
Signature · 1 line
2.19.1
author
#3 lore raw

PATCH v10 2/6 modpost: add support for mdev class id

In reply to Jason Wang

Add support to parse mdev class id table.

---
 drivers/vfio/mdev/vfio_mdev.c     |  2 ++
 scripts/mod/devicetable-offsets.c |  3 +++
 scripts/mod/file2alias.c          | 11 +++++++++++
 3 files changed, 16 insertions(+)
drivers/vfio/mdev/vfio_mdev.c +2 -0
index 38431e9ef7f5..a6641cd8b5a3 100644
@@ -125,6 +125,8 @@ static const struct mdev_class_id vfio_id_table[] = {
 	{ 0 },
 };
 
+MODULE_DEVICE_TABLE(mdev, vfio_id_table);
+
 static struct mdev_driver vfio_mdev_driver = {
 	.name	= "vfio_mdev",
 	.probe	= vfio_mdev_probe,
scripts/mod/devicetable-offsets.c +3 -0
index 054405b90ba4..6cbb1062488a 100644
@@ -231,5 +231,8 @@ int main(void)
 	DEVID(wmi_device_id);
 	DEVID_FIELD(wmi_device_id, guid_string);
 
+	DEVID(mdev_class_id);
+	DEVID_FIELD(mdev_class_id, id);
+
 	return 0;
 }
scripts/mod/file2alias.c +11 -0
index c91eba751804..45f1c22f49be 100644
@@ -1335,6 +1335,16 @@ static int do_wmi_entry(const char *filename, void *symval, char *alias)
 	return 1;
 }
 
+/* looks like: "mdev:cN" */
+static int do_mdev_entry(const char *filename, void *symval, char *alias)
+{
+	DEF_FIELD(symval, mdev_class_id, id);
+
+	sprintf(alias, "mdev:c%02X", id);
+	add_wildcard(alias);
+	return 1;
+}
+
 /* Does namelen bytes of name exactly match the symbol? */
 static bool sym_is(const char *name, unsigned namelen, const char *symbol)
 {
@@ -1407,6 +1417,7 @@ static const struct devtable devtable[] = {
 	{"typec", SIZE_typec_device_id, do_typec_entry},
 	{"tee", SIZE_tee_client_device_id, do_tee_entry},
 	{"wmi", SIZE_wmi_device_id, do_wmi_entry},
+	{"mdev", SIZE_mdev_class_id, do_mdev_entry},
 };
 
 /* Create MODULE_ALIAS() statements.
Signature · 1 line
2.19.1
author
#4 lore raw

PATCH v10 2/6 modpost: add support for mdev class id

In reply to Jason Wang

Add support to parse mdev class id table.

---
 drivers/vfio/mdev/vfio_mdev.c     |  2 ++
 scripts/mod/devicetable-offsets.c |  3 +++
 scripts/mod/file2alias.c          | 11 +++++++++++
 3 files changed, 16 insertions(+)
drivers/vfio/mdev/vfio_mdev.c +2 -0
index 38431e9ef7f5..a6641cd8b5a3 100644
@@ -125,6 +125,8 @@ static const struct mdev_class_id vfio_id_table[] = {
 	{ 0 },
 };
 
+MODULE_DEVICE_TABLE(mdev, vfio_id_table);
+
 static struct mdev_driver vfio_mdev_driver = {
 	.name	= "vfio_mdev",
 	.probe	= vfio_mdev_probe,
scripts/mod/devicetable-offsets.c +3 -0
index 054405b90ba4..6cbb1062488a 100644
@@ -231,5 +231,8 @@ int main(void)
 	DEVID(wmi_device_id);
 	DEVID_FIELD(wmi_device_id, guid_string);
 
+	DEVID(mdev_class_id);
+	DEVID_FIELD(mdev_class_id, id);
+
 	return 0;
 }
scripts/mod/file2alias.c +11 -0
index c91eba751804..45f1c22f49be 100644
@@ -1335,6 +1335,16 @@ static int do_wmi_entry(const char *filename, void *symval, char *alias)
 	return 1;
 }
 
+/* looks like: "mdev:cN" */
+static int do_mdev_entry(const char *filename, void *symval, char *alias)
+{
+	DEF_FIELD(symval, mdev_class_id, id);
+
+	sprintf(alias, "mdev:c%02X", id);
+	add_wildcard(alias);
+	return 1;
+}
+
 /* Does namelen bytes of name exactly match the symbol? */
 static bool sym_is(const char *name, unsigned namelen, const char *symbol)
 {
@@ -1407,6 +1417,7 @@ static const struct devtable devtable[] = {
 	{"typec", SIZE_typec_device_id, do_typec_entry},
 	{"tee", SIZE_tee_client_device_id, do_tee_entry},
 	{"wmi", SIZE_wmi_device_id, do_wmi_entry},
+	{"mdev", SIZE_mdev_class_id, do_mdev_entry},
 };
 
 /* Create MODULE_ALIAS() statements.
Signature · 1 line
2.19.1
author
#5 lore raw

PATCH v10 3/6 mdev: introduce device specific ops

In reply to Jason Wang

Currently, except for the create and remove, the rest of mdev_parent_ops is designed for vfio-mdev driver only and may not help for kernel mdev driver. With the help of class id, this patch introduces device specific callbacks inside mdev_device structure. This allows different set of callback to be used by vfio-mdev and virtio-mdev.

---
 .../driver-api/vfio-mediated-device.rst       | 35 +++++++++----
 MAINTAINERS                                   |  1 +
 drivers/gpu/drm/i915/gvt/kvmgt.c              | 18 ++++---
 drivers/s390/cio/vfio_ccw_ops.c               | 18 ++++---
 drivers/s390/crypto/vfio_ap_ops.c             | 14 +++--
 drivers/vfio/mdev/mdev_core.c                 | 24 ++++++++-
 drivers/vfio/mdev/mdev_private.h              |  5 ++
 drivers/vfio/mdev/vfio_mdev.c                 | 37 ++++++-------
 include/linux/mdev.h                          | 43 ++++-----------
 include/linux/mdev_vfio_ops.h                 | 52 +++++++++++++++++++
 samples/vfio-mdev/mbochs.c                    | 20 ++++---
 samples/vfio-mdev/mdpy.c                      | 20 ++++---
 samples/vfio-mdev/mtty.c                      | 18 ++++---
 13 files changed, 206 insertions(+), 99 deletions(-)
 create mode 100644 include/linux/mdev_vfio_ops.h
Documentation/driver-api/vfio-mediated-device.rst +25 -10
index 6709413bee29..04d56884c357 100644
@@ -152,15 +152,6 @@ callbacks per mdev parent device, per mdev type, or any other categorization.
 Vendor drivers are expected to be fully asynchronous in this respect or
 provide their own internal resource protection.)
 
-The callbacks in the mdev_parent_ops structure are as follows:
-
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
-* read : read emulation callback
-* write: write emulation callback
-* mmap: mmap emulation callback
-
 A driver should use the mdev_parent_ops structure in the function call to
 register itself with the mdev core driver::
 
@@ -172,10 +163,34 @@ that a driver should use to unregister itself with the mdev core driver::
 
 	extern void mdev_unregister_device(struct device *dev);
 
-It is also required to specify the class_id in create() callback through::
+As multiple types of mediated devices may be supported, class id needs
+to be specified in the create() callback. This could be done
+explicitly for the device that interacts with the mdev device directly
+through::
 
 	int mdev_set_class(struct mdev_device *mdev, u16 id);
 
+For the device that uses the mdev bus for its operation, the class
+should provide helper function to set class id and device specific
+ops. E.g for vfio-mdev devices, the function to be called is::
+
+	int mdev_set_vfio_ops(struct mdev_device *mdev,
+                              const struct mdev_vfio_device_ops *vfio_ops);
+
+The class id (set by this function to MDEV_CLASS_ID_VFIO) is used to
+match a device with an mdev driver via its id table. The device
+specific callbacks (specified in *vfio_ops) are obtainable via
+mdev_get_vfio_ops() (for use by the mdev bus driver). A vfio-mdev
+device (class id MDEV_CLASS_ID_VFIO) uses the following
+device-specific ops:
+
+* open: open callback of vfio mediated device
+* close: close callback of vfio mediated device
+* ioctl: ioctl callback of vfio mediated device
+* read : read emulation callback
+* write: write emulation callback
+* mmap: mmap emulation callback
+
 Mediated Device Management Interface Through sysfs
 ==================================================
 
MAINTAINERS +1 -0
index cba1095547fd..f661d13344d6 100644
@@ -17121,6 +17121,7 @@ S:	Maintained
 F:	Documentation/driver-api/vfio-mediated-device.rst
 F:	drivers/vfio/mdev/
 F:	include/linux/mdev.h
+F:	include/linux/mdev_vfio_ops.h
 F:	samples/vfio-mdev/
 
 VFIO PLATFORM DRIVER
drivers/gpu/drm/i915/gvt/kvmgt.c +11 -7
index 6420f0dbd31b..662f3a672372 100644
@@ -42,6 +42,7 @@
 #include <linux/kvm_host.h>
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/debugfs.h>
 
 #include <linux/nospec.h>
@@ -643,6 +644,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
 	vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
 }
 
+static const struct mdev_vfio_device_ops intel_vfio_vgpu_dev_ops;
+
 static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct intel_vgpu *vgpu = NULL;
@@ -678,7 +681,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 		     dev_name(mdev_dev(mdev)));
 	ret = 0;
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &intel_vfio_vgpu_dev_ops);
 out:
 	return ret;
 }
@@ -1599,20 +1602,21 @@ static const struct attribute_group *intel_vgpu_groups[] = {
 	NULL,
 };
 
-static struct mdev_parent_ops intel_vgpu_ops = {
-	.mdev_attr_groups       = intel_vgpu_groups,
-	.create			= intel_vgpu_create,
-	.remove			= intel_vgpu_remove,
-
+static const struct mdev_vfio_device_ops intel_vfio_vgpu_dev_ops = {
 	.open			= intel_vgpu_open,
 	.release		= intel_vgpu_release,
-
 	.read			= intel_vgpu_read,
 	.write			= intel_vgpu_write,
 	.mmap			= intel_vgpu_mmap,
 	.ioctl			= intel_vgpu_ioctl,
 };
 
+static struct mdev_parent_ops intel_vgpu_ops = {
+	.mdev_attr_groups       = intel_vgpu_groups,
+	.create			= intel_vgpu_create,
+	.remove			= intel_vgpu_remove,
+};
+
 static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
 {
 	struct attribute **kvm_type_attrs;
drivers/s390/cio/vfio_ccw_ops.c +12 -6
index cf2c013ae32f..fa473ded71c7 100644
@@ -12,6 +12,7 @@
 
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/nospec.h>
 #include <linux/slab.h>
 
@@ -110,6 +111,8 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct vfio_ccw_private *private =
@@ -129,7 +132,7 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 			   private->sch->schid.ssid,
 			   private->sch->schid.sch_no);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -575,11 +578,7 @@ static ssize_t vfio_ccw_mdev_ioctl(struct mdev_device *mdev,
 	}
 }
 
-static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
-	.owner			= THIS_MODULE,
-	.supported_type_groups  = mdev_type_groups,
-	.create			= vfio_ccw_mdev_create,
-	.remove			= vfio_ccw_mdev_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= vfio_ccw_mdev_open,
 	.release		= vfio_ccw_mdev_release,
 	.read			= vfio_ccw_mdev_read,
@@ -587,6 +586,13 @@ static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
 	.ioctl			= vfio_ccw_mdev_ioctl,
 };
 
+static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
+	.owner			= THIS_MODULE,
+	.supported_type_groups  = mdev_type_groups,
+	.create			= vfio_ccw_mdev_create,
+	.remove			= vfio_ccw_mdev_remove,
+};
+
 int vfio_ccw_mdev_reg(struct subchannel *sch)
 {
 	return mdev_register_device(&sch->dev, &vfio_ccw_mdev_ops);
drivers/s390/crypto/vfio_ap_ops.c +10 -4
index 07c31070afeb..7bdc62393112 100644
@@ -16,6 +16,7 @@
 #include <linux/bitops.h>
 #include <linux/kvm_host.h>
 #include <linux/module.h>
+#include <linux/mdev_vfio_ops.h>
 #include <asm/kvm.h>
 #include <asm/zcrypt.h>
 
@@ -321,6 +322,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
 	matrix->adm_max = info->apxa ? info->Nd : 15;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct ap_matrix_mdev *matrix_mdev;
@@ -343,7 +346,7 @@ static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&matrix_mdev->node, &matrix_dev->mdev_list);
 	mutex_unlock(&matrix_dev->lock);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -1281,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
 	return ret;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
+	.open			= vfio_ap_mdev_open,
+	.release		= vfio_ap_mdev_release,
+	.ioctl			= vfio_ap_mdev_ioctl,
+};
+
 static const struct mdev_parent_ops vfio_ap_matrix_ops = {
 	.owner			= THIS_MODULE,
 	.supported_type_groups	= vfio_ap_mdev_type_groups,
 	.mdev_attr_groups	= vfio_ap_mdev_attr_groups,
 	.create			= vfio_ap_mdev_create,
 	.remove			= vfio_ap_mdev_remove,
-	.open			= vfio_ap_mdev_open,
-	.release		= vfio_ap_mdev_release,
-	.ioctl			= vfio_ap_mdev_ioctl,
 };
 
 int vfio_ap_mdev_register(void)
drivers/vfio/mdev/mdev_core.c +23 -1
index 7bfa2e46e829..4e70f19ac145 100644
@@ -47,7 +47,8 @@ EXPORT_SYMBOL(mdev_set_drvdata);
 
 /*
  * Specify the class for the mdev device, this must be called during
- * create() callback.
+ * create() callback explicitly or implicitly through the helpers
+ * provided by each class.
  */
 void mdev_set_class(struct mdev_device *mdev, u16 id)
 {
@@ -56,6 +57,27 @@ void mdev_set_class(struct mdev_device *mdev, u16 id)
 }
 EXPORT_SYMBOL(mdev_set_class);
 
+/*
+ * Specify the mdev device to be a VFIO mdev device, and set VFIO
+ * device ops for it. This must be called from the create() callback
+ * for VFIO mdev device.
+ */
+void mdev_set_vfio_ops(struct mdev_device *mdev,
+		       const struct mdev_vfio_device_ops *vfio_ops)
+{
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev->vfio_ops = vfio_ops;
+}
+EXPORT_SYMBOL(mdev_set_vfio_ops);
+
+/* Get the VFIO device ops for the mdev device. */
+const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev)
+{
+	WARN_ON(mdev->class_id != MDEV_CLASS_ID_VFIO);
+	return mdev->vfio_ops;
+}
+EXPORT_SYMBOL(mdev_get_vfio_ops);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
drivers/vfio/mdev/mdev_private.h +5 -0
index c65f436c1869..411227373625 100644
@@ -10,6 +10,8 @@
 #ifndef MDEV_PRIVATE_H
 #define MDEV_PRIVATE_H
 
+#include <linux/mdev_vfio_ops.h>
+
 int  mdev_bus_register(void);
 void mdev_bus_unregister(void);
 
@@ -34,6 +36,9 @@ struct mdev_device {
 	struct device *iommu_device;
 	bool active;
 	u16 class_id;
+	union {
+		const struct mdev_vfio_device_ops *vfio_ops;
+	};
 };
 
 #define to_mdev_device(dev)	container_of(dev, struct mdev_device, dev)
drivers/vfio/mdev/vfio_mdev.c +19 -18
index a6641cd8b5a3..c01e2194e4b3 100644
@@ -14,6 +14,7 @@
 #include <linux/slab.h>
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 
 #include "mdev_private.h"
 
@@ -24,16 +25,16 @@
 static int vfio_mdev_open(void *device_data)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 	int ret;
 
-	if (unlikely(!parent->ops->open))
+	if (unlikely(!ops->open))
 		return -EINVAL;
 
 	if (!try_module_get(THIS_MODULE))
 		return -ENODEV;
 
-	ret = parent->ops->open(mdev);
+	ret = ops->open(mdev);
 	if (ret)
 		module_put(THIS_MODULE);
 
@@ -43,10 +44,10 @@ static int vfio_mdev_open(void *device_data)
 static void vfio_mdev_release(void *device_data)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (likely(parent->ops->release))
-		parent->ops->release(mdev);
+	if (likely(ops->release))
+		ops->release(mdev);
 
 	module_put(THIS_MODULE);
 }
@@ -55,47 +56,47 @@ static long vfio_mdev_unlocked_ioctl(void *device_data,
 				     unsigned int cmd, unsigned long arg)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->ioctl))
+	if (unlikely(!ops->ioctl))
 		return -EINVAL;
 
-	return parent->ops->ioctl(mdev, cmd, arg);
+	return ops->ioctl(mdev, cmd, arg);
 }
 
 static ssize_t vfio_mdev_read(void *device_data, char __user *buf,
 			      size_t count, loff_t *ppos)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->read))
+	if (unlikely(!ops->read))
 		return -EINVAL;
 
-	return parent->ops->read(mdev, buf, count, ppos);
+	return ops->read(mdev, buf, count, ppos);
 }
 
 static ssize_t vfio_mdev_write(void *device_data, const char __user *buf,
 			       size_t count, loff_t *ppos)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->write))
+	if (unlikely(!ops->write))
 		return -EINVAL;
 
-	return parent->ops->write(mdev, buf, count, ppos);
+	return ops->write(mdev, buf, count, ppos);
 }
 
 static int vfio_mdev_mmap(void *device_data, struct vm_area_struct *vma)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->mmap))
+	if (unlikely(!ops->mmap))
 		return -EINVAL;
 
-	return parent->ops->mmap(mdev, vma);
+	return ops->mmap(mdev, vma);
 }
 
 static const struct vfio_device_ops vfio_mdev_dev_ops = {
include/linux/mdev.h +10 -33

68 lines not on this page · raw message

include/linux/mdev_vfio_ops.h new file +52 -0

52 lines not on this page · raw message

samples/vfio-mdev/mbochs.c +13 -7
index 115bc5074656..12963767ba37 100644
@@ -30,6 +30,7 @@
 #include <linux/iommu.h>
 #include <linux/sysfs.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <linux/dma-buf.h>
 #include <linux/highmem.h>
@@ -516,6 +517,8 @@ static int mbochs_reset(struct mdev_device *mdev)
 	return 0;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,7 +564,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 	mbochs_reset(mdev);
 
 	mbochs_used_mbytes += type->mbytes;
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 
 err_mem:
@@ -1419,12 +1422,7 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
-static const struct mdev_parent_ops mdev_fops = {
-	.owner			= THIS_MODULE,
-	.mdev_attr_groups	= mdev_dev_groups,
-	.supported_type_groups	= mdev_type_groups,
-	.create			= mbochs_create,
-	.remove			= mbochs_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= mbochs_open,
 	.release		= mbochs_close,
 	.read			= mbochs_read,
@@ -1433,6 +1431,14 @@ static const struct mdev_parent_ops mdev_fops = {
 	.mmap			= mbochs_mmap,
 };
 
+static const struct mdev_parent_ops mdev_fops = {
+	.owner			= THIS_MODULE,
+	.mdev_attr_groups	= mdev_dev_groups,
+	.supported_type_groups	= mdev_type_groups,
+	.create			= mbochs_create,
+	.remove			= mbochs_remove,
+};
+
 static const struct file_operations vd_fops = {
 	.owner		= THIS_MODULE,
 };
samples/vfio-mdev/mdpy.c +13 -7
index 665614574d50..50ee6c98b2af 100644
@@ -26,6 +26,7 @@
 #include <linux/iommu.h>
 #include <linux/sysfs.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <drm/drm_fourcc.h>
 #include "mdpy-defs.h"
@@ -226,6 +227,8 @@ static int mdpy_reset(struct mdev_device *mdev)
 	return 0;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,7 +272,7 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 	mdpy_reset(mdev);
 
 	mdpy_count++;
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -726,12 +729,7 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
-static const struct mdev_parent_ops mdev_fops = {
-	.owner			= THIS_MODULE,
-	.mdev_attr_groups	= mdev_dev_groups,
-	.supported_type_groups	= mdev_type_groups,
-	.create			= mdpy_create,
-	.remove			= mdpy_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= mdpy_open,
 	.release		= mdpy_close,
 	.read			= mdpy_read,
@@ -740,6 +738,14 @@ static const struct mdev_parent_ops mdev_fops = {
 	.mmap			= mdpy_mmap,
 };
 
+static const struct mdev_parent_ops mdev_fops = {
+	.owner			= THIS_MODULE,
+	.mdev_attr_groups	= mdev_dev_groups,
+	.supported_type_groups	= mdev_type_groups,
+	.create			= mdpy_create,
+	.remove			= mdpy_remove,
+};
+
 static const struct file_operations vd_fops = {
 	.owner		= THIS_MODULE,
 };
samples/vfio-mdev/mtty.c +12 -6
index 90da12ff7fd9..be476e7ad1f8 100644
@@ -27,6 +27,7 @@
 #include <linux/ctype.h>
 #include <linux/file.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <linux/serial.h>
 #include <uapi/linux/serial_reg.h>
@@ -708,6 +709,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
 	return ret;
 }
 
+static const struct mdev_vfio_device_ops vfio_dev_ops;
+
 static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct mdev_state *mdev_state;
@@ -755,7 +758,7 @@ static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&mdev_state->next, &mdev_devices_list);
 	mutex_unlock(&mdev_list_lock);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &vfio_dev_ops);
 	return 0;
 }
 
@@ -1388,6 +1391,14 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
+static const struct mdev_vfio_device_ops vfio_dev_ops = {
+	.open		= mtty_open,
+	.release	= mtty_close,
+	.read		= mtty_read,
+	.write		= mtty_write,
+	.ioctl		= mtty_ioctl,
+};
+
 static const struct mdev_parent_ops mdev_fops = {
 	.owner                  = THIS_MODULE,
 	.dev_attr_groups        = mtty_dev_groups,
@@ -1395,11 +1406,6 @@ static const struct mdev_parent_ops mdev_fops = {
 	.supported_type_groups  = mdev_type_groups,
 	.create                 = mtty_create,
 	.remove			= mtty_remove,
-	.open                   = mtty_open,
-	.release                = mtty_close,
-	.read                   = mtty_read,
-	.write                  = mtty_write,
-	.ioctl		        = mtty_ioctl,
 };
 
 static void mtty_device_release(struct device *dev)
Signature · 1 line
2.19.1
author
#6 lore raw

PATCH v10 3/6 mdev: introduce device specific ops

In reply to Jason Wang

Currently, except for the create and remove, the rest of mdev_parent_ops is designed for vfio-mdev driver only and may not help for kernel mdev driver. With the help of class id, this patch introduces device specific callbacks inside mdev_device structure. This allows different set of callback to be used by vfio-mdev and virtio-mdev.

---
 .../driver-api/vfio-mediated-device.rst       | 35 +++++++++----
 MAINTAINERS                                   |  1 +
 drivers/gpu/drm/i915/gvt/kvmgt.c              | 18 ++++---
 drivers/s390/cio/vfio_ccw_ops.c               | 18 ++++---
 drivers/s390/crypto/vfio_ap_ops.c             | 14 +++--
 drivers/vfio/mdev/mdev_core.c                 | 24 ++++++++-
 drivers/vfio/mdev/mdev_private.h              |  5 ++
 drivers/vfio/mdev/vfio_mdev.c                 | 37 ++++++-------
 include/linux/mdev.h                          | 43 ++++-----------
 include/linux/mdev_vfio_ops.h                 | 52 +++++++++++++++++++
 samples/vfio-mdev/mbochs.c                    | 20 ++++---
 samples/vfio-mdev/mdpy.c                      | 20 ++++---
 samples/vfio-mdev/mtty.c                      | 18 ++++---
 13 files changed, 206 insertions(+), 99 deletions(-)
 create mode 100644 include/linux/mdev_vfio_ops.h
Documentation/driver-api/vfio-mediated-device.rst +25 -10
index 6709413bee29..04d56884c357 100644
@@ -152,15 +152,6 @@ callbacks per mdev parent device, per mdev type, or any other categorization.
 Vendor drivers are expected to be fully asynchronous in this respect or
 provide their own internal resource protection.)
 
-The callbacks in the mdev_parent_ops structure are as follows:
-
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
-* read : read emulation callback
-* write: write emulation callback
-* mmap: mmap emulation callback
-
 A driver should use the mdev_parent_ops structure in the function call to
 register itself with the mdev core driver::
 
@@ -172,10 +163,34 @@ that a driver should use to unregister itself with the mdev core driver::
 
 	extern void mdev_unregister_device(struct device *dev);
 
-It is also required to specify the class_id in create() callback through::
+As multiple types of mediated devices may be supported, class id needs
+to be specified in the create() callback. This could be done
+explicitly for the device that interacts with the mdev device directly
+through::
 
 	int mdev_set_class(struct mdev_device *mdev, u16 id);
 
+For the device that uses the mdev bus for its operation, the class
+should provide helper function to set class id and device specific
+ops. E.g for vfio-mdev devices, the function to be called is::
+
+	int mdev_set_vfio_ops(struct mdev_device *mdev,
+                              const struct mdev_vfio_device_ops *vfio_ops);
+
+The class id (set by this function to MDEV_CLASS_ID_VFIO) is used to
+match a device with an mdev driver via its id table. The device
+specific callbacks (specified in *vfio_ops) are obtainable via
+mdev_get_vfio_ops() (for use by the mdev bus driver). A vfio-mdev
+device (class id MDEV_CLASS_ID_VFIO) uses the following
+device-specific ops:
+
+* open: open callback of vfio mediated device
+* close: close callback of vfio mediated device
+* ioctl: ioctl callback of vfio mediated device
+* read : read emulation callback
+* write: write emulation callback
+* mmap: mmap emulation callback
+
 Mediated Device Management Interface Through sysfs
 ==================================================
 
MAINTAINERS +1 -0
index cba1095547fd..f661d13344d6 100644
@@ -17121,6 +17121,7 @@ S:	Maintained
 F:	Documentation/driver-api/vfio-mediated-device.rst
 F:	drivers/vfio/mdev/
 F:	include/linux/mdev.h
+F:	include/linux/mdev_vfio_ops.h
 F:	samples/vfio-mdev/
 
 VFIO PLATFORM DRIVER
drivers/gpu/drm/i915/gvt/kvmgt.c +11 -7
index 6420f0dbd31b..662f3a672372 100644
@@ -42,6 +42,7 @@
 #include <linux/kvm_host.h>
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/debugfs.h>
 
 #include <linux/nospec.h>
@@ -643,6 +644,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
 	vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
 }
 
+static const struct mdev_vfio_device_ops intel_vfio_vgpu_dev_ops;
+
 static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct intel_vgpu *vgpu = NULL;
@@ -678,7 +681,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 		     dev_name(mdev_dev(mdev)));
 	ret = 0;
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &intel_vfio_vgpu_dev_ops);
 out:
 	return ret;
 }
@@ -1599,20 +1602,21 @@ static const struct attribute_group *intel_vgpu_groups[] = {
 	NULL,
 };
 
-static struct mdev_parent_ops intel_vgpu_ops = {
-	.mdev_attr_groups       = intel_vgpu_groups,
-	.create			= intel_vgpu_create,
-	.remove			= intel_vgpu_remove,
-
+static const struct mdev_vfio_device_ops intel_vfio_vgpu_dev_ops = {
 	.open			= intel_vgpu_open,
 	.release		= intel_vgpu_release,
-
 	.read			= intel_vgpu_read,
 	.write			= intel_vgpu_write,
 	.mmap			= intel_vgpu_mmap,
 	.ioctl			= intel_vgpu_ioctl,
 };
 
+static struct mdev_parent_ops intel_vgpu_ops = {
+	.mdev_attr_groups       = intel_vgpu_groups,
+	.create			= intel_vgpu_create,
+	.remove			= intel_vgpu_remove,
+};
+
 static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
 {
 	struct attribute **kvm_type_attrs;
drivers/s390/cio/vfio_ccw_ops.c +12 -6
index cf2c013ae32f..fa473ded71c7 100644
@@ -12,6 +12,7 @@
 
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/nospec.h>
 #include <linux/slab.h>
 
@@ -110,6 +111,8 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct vfio_ccw_private *private =
@@ -129,7 +132,7 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 			   private->sch->schid.ssid,
 			   private->sch->schid.sch_no);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -575,11 +578,7 @@ static ssize_t vfio_ccw_mdev_ioctl(struct mdev_device *mdev,
 	}
 }
 
-static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
-	.owner			= THIS_MODULE,
-	.supported_type_groups  = mdev_type_groups,
-	.create			= vfio_ccw_mdev_create,
-	.remove			= vfio_ccw_mdev_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= vfio_ccw_mdev_open,
 	.release		= vfio_ccw_mdev_release,
 	.read			= vfio_ccw_mdev_read,
@@ -587,6 +586,13 @@ static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
 	.ioctl			= vfio_ccw_mdev_ioctl,
 };
 
+static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
+	.owner			= THIS_MODULE,
+	.supported_type_groups  = mdev_type_groups,
+	.create			= vfio_ccw_mdev_create,
+	.remove			= vfio_ccw_mdev_remove,
+};
+
 int vfio_ccw_mdev_reg(struct subchannel *sch)
 {
 	return mdev_register_device(&sch->dev, &vfio_ccw_mdev_ops);
drivers/s390/crypto/vfio_ap_ops.c +10 -4
index 07c31070afeb..7bdc62393112 100644
@@ -16,6 +16,7 @@
 #include <linux/bitops.h>
 #include <linux/kvm_host.h>
 #include <linux/module.h>
+#include <linux/mdev_vfio_ops.h>
 #include <asm/kvm.h>
 #include <asm/zcrypt.h>
 
@@ -321,6 +322,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
 	matrix->adm_max = info->apxa ? info->Nd : 15;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct ap_matrix_mdev *matrix_mdev;
@@ -343,7 +346,7 @@ static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&matrix_mdev->node, &matrix_dev->mdev_list);
 	mutex_unlock(&matrix_dev->lock);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -1281,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
 	return ret;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
+	.open			= vfio_ap_mdev_open,
+	.release		= vfio_ap_mdev_release,
+	.ioctl			= vfio_ap_mdev_ioctl,
+};
+
 static const struct mdev_parent_ops vfio_ap_matrix_ops = {
 	.owner			= THIS_MODULE,
 	.supported_type_groups	= vfio_ap_mdev_type_groups,
 	.mdev_attr_groups	= vfio_ap_mdev_attr_groups,
 	.create			= vfio_ap_mdev_create,
 	.remove			= vfio_ap_mdev_remove,
-	.open			= vfio_ap_mdev_open,
-	.release		= vfio_ap_mdev_release,
-	.ioctl			= vfio_ap_mdev_ioctl,
 };
 
 int vfio_ap_mdev_register(void)
drivers/vfio/mdev/mdev_core.c +23 -1
index 7bfa2e46e829..4e70f19ac145 100644
@@ -47,7 +47,8 @@ EXPORT_SYMBOL(mdev_set_drvdata);
 
 /*
  * Specify the class for the mdev device, this must be called during
- * create() callback.
+ * create() callback explicitly or implicitly through the helpers
+ * provided by each class.
  */
 void mdev_set_class(struct mdev_device *mdev, u16 id)
 {
@@ -56,6 +57,27 @@ void mdev_set_class(struct mdev_device *mdev, u16 id)
 }
 EXPORT_SYMBOL(mdev_set_class);
 
+/*
+ * Specify the mdev device to be a VFIO mdev device, and set VFIO
+ * device ops for it. This must be called from the create() callback
+ * for VFIO mdev device.
+ */
+void mdev_set_vfio_ops(struct mdev_device *mdev,
+		       const struct mdev_vfio_device_ops *vfio_ops)
+{
+	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev->vfio_ops = vfio_ops;
+}
+EXPORT_SYMBOL(mdev_set_vfio_ops);
+
+/* Get the VFIO device ops for the mdev device. */
+const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev)
+{
+	WARN_ON(mdev->class_id != MDEV_CLASS_ID_VFIO);
+	return mdev->vfio_ops;
+}
+EXPORT_SYMBOL(mdev_get_vfio_ops);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
drivers/vfio/mdev/mdev_private.h +5 -0
index c65f436c1869..411227373625 100644
@@ -10,6 +10,8 @@
 #ifndef MDEV_PRIVATE_H
 #define MDEV_PRIVATE_H
 
+#include <linux/mdev_vfio_ops.h>
+
 int  mdev_bus_register(void);
 void mdev_bus_unregister(void);
 
@@ -34,6 +36,9 @@ struct mdev_device {
 	struct device *iommu_device;
 	bool active;
 	u16 class_id;
+	union {
+		const struct mdev_vfio_device_ops *vfio_ops;
+	};
 };
 
 #define to_mdev_device(dev)	container_of(dev, struct mdev_device, dev)
drivers/vfio/mdev/vfio_mdev.c +19 -18
index a6641cd8b5a3..c01e2194e4b3 100644
@@ -14,6 +14,7 @@
 #include <linux/slab.h>
 #include <linux/vfio.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 
 #include "mdev_private.h"
 
@@ -24,16 +25,16 @@
 static int vfio_mdev_open(void *device_data)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 	int ret;
 
-	if (unlikely(!parent->ops->open))
+	if (unlikely(!ops->open))
 		return -EINVAL;
 
 	if (!try_module_get(THIS_MODULE))
 		return -ENODEV;
 
-	ret = parent->ops->open(mdev);
+	ret = ops->open(mdev);
 	if (ret)
 		module_put(THIS_MODULE);
 
@@ -43,10 +44,10 @@ static int vfio_mdev_open(void *device_data)
 static void vfio_mdev_release(void *device_data)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (likely(parent->ops->release))
-		parent->ops->release(mdev);
+	if (likely(ops->release))
+		ops->release(mdev);
 
 	module_put(THIS_MODULE);
 }
@@ -55,47 +56,47 @@ static long vfio_mdev_unlocked_ioctl(void *device_data,
 				     unsigned int cmd, unsigned long arg)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->ioctl))
+	if (unlikely(!ops->ioctl))
 		return -EINVAL;
 
-	return parent->ops->ioctl(mdev, cmd, arg);
+	return ops->ioctl(mdev, cmd, arg);
 }
 
 static ssize_t vfio_mdev_read(void *device_data, char __user *buf,
 			      size_t count, loff_t *ppos)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->read))
+	if (unlikely(!ops->read))
 		return -EINVAL;
 
-	return parent->ops->read(mdev, buf, count, ppos);
+	return ops->read(mdev, buf, count, ppos);
 }
 
 static ssize_t vfio_mdev_write(void *device_data, const char __user *buf,
 			       size_t count, loff_t *ppos)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->write))
+	if (unlikely(!ops->write))
 		return -EINVAL;
 
-	return parent->ops->write(mdev, buf, count, ppos);
+	return ops->write(mdev, buf, count, ppos);
 }
 
 static int vfio_mdev_mmap(void *device_data, struct vm_area_struct *vma)
 {
 	struct mdev_device *mdev = device_data;
-	struct mdev_parent *parent = mdev->parent;
+	const struct mdev_vfio_device_ops *ops = mdev_get_vfio_ops(mdev);
 
-	if (unlikely(!parent->ops->mmap))
+	if (unlikely(!ops->mmap))
 		return -EINVAL;
 
-	return parent->ops->mmap(mdev, vma);
+	return ops->mmap(mdev, vma);
 }
 
 static const struct vfio_device_ops vfio_mdev_dev_ops = {
include/linux/mdev.h +10 -33

68 lines not on this page · raw message

include/linux/mdev_vfio_ops.h new file +52 -0

52 lines not on this page · raw message

samples/vfio-mdev/mbochs.c +13 -7
index 115bc5074656..12963767ba37 100644
@@ -30,6 +30,7 @@
 #include <linux/iommu.h>
 #include <linux/sysfs.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <linux/dma-buf.h>
 #include <linux/highmem.h>
@@ -516,6 +517,8 @@ static int mbochs_reset(struct mdev_device *mdev)
 	return 0;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,7 +564,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
 	mbochs_reset(mdev);
 
 	mbochs_used_mbytes += type->mbytes;
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 
 err_mem:
@@ -1419,12 +1422,7 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
-static const struct mdev_parent_ops mdev_fops = {
-	.owner			= THIS_MODULE,
-	.mdev_attr_groups	= mdev_dev_groups,
-	.supported_type_groups	= mdev_type_groups,
-	.create			= mbochs_create,
-	.remove			= mbochs_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= mbochs_open,
 	.release		= mbochs_close,
 	.read			= mbochs_read,
@@ -1433,6 +1431,14 @@ static const struct mdev_parent_ops mdev_fops = {
 	.mmap			= mbochs_mmap,
 };
 
+static const struct mdev_parent_ops mdev_fops = {
+	.owner			= THIS_MODULE,
+	.mdev_attr_groups	= mdev_dev_groups,
+	.supported_type_groups	= mdev_type_groups,
+	.create			= mbochs_create,
+	.remove			= mbochs_remove,
+};
+
 static const struct file_operations vd_fops = {
 	.owner		= THIS_MODULE,
 };
samples/vfio-mdev/mdpy.c +13 -7
index 665614574d50..50ee6c98b2af 100644
@@ -26,6 +26,7 @@
 #include <linux/iommu.h>
 #include <linux/sysfs.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <drm/drm_fourcc.h>
 #include "mdpy-defs.h"
@@ -226,6 +227,8 @@ static int mdpy_reset(struct mdev_device *mdev)
 	return 0;
 }
 
+static const struct mdev_vfio_device_ops mdev_vfio_ops;
+
 static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,7 +272,7 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
 	mdpy_reset(mdev);
 
 	mdpy_count++;
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &mdev_vfio_ops);
 	return 0;
 }
 
@@ -726,12 +729,7 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
-static const struct mdev_parent_ops mdev_fops = {
-	.owner			= THIS_MODULE,
-	.mdev_attr_groups	= mdev_dev_groups,
-	.supported_type_groups	= mdev_type_groups,
-	.create			= mdpy_create,
-	.remove			= mdpy_remove,
+static const struct mdev_vfio_device_ops mdev_vfio_ops = {
 	.open			= mdpy_open,
 	.release		= mdpy_close,
 	.read			= mdpy_read,
@@ -740,6 +738,14 @@ static const struct mdev_parent_ops mdev_fops = {
 	.mmap			= mdpy_mmap,
 };
 
+static const struct mdev_parent_ops mdev_fops = {
+	.owner			= THIS_MODULE,
+	.mdev_attr_groups	= mdev_dev_groups,
+	.supported_type_groups	= mdev_type_groups,
+	.create			= mdpy_create,
+	.remove			= mdpy_remove,
+};
+
 static const struct file_operations vd_fops = {
 	.owner		= THIS_MODULE,
 };
samples/vfio-mdev/mtty.c +12 -6
index 90da12ff7fd9..be476e7ad1f8 100644
@@ -27,6 +27,7 @@
 #include <linux/ctype.h>
 #include <linux/file.h>
 #include <linux/mdev.h>
+#include <linux/mdev_vfio_ops.h>
 #include <linux/pci.h>
 #include <linux/serial.h>
 #include <uapi/linux/serial_reg.h>
@@ -708,6 +709,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
 	return ret;
 }
 
+static const struct mdev_vfio_device_ops vfio_dev_ops;
+
 static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 {
 	struct mdev_state *mdev_state;
@@ -755,7 +758,7 @@ static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
 	list_add(&mdev_state->next, &mdev_devices_list);
 	mutex_unlock(&mdev_list_lock);
 
-	mdev_set_class(mdev, MDEV_CLASS_ID_VFIO);
+	mdev_set_vfio_ops(mdev, &vfio_dev_ops);
 	return 0;
 }
 
@@ -1388,6 +1391,14 @@ static struct attribute_group *mdev_type_groups[] = {
 	NULL,
 };
 
+static const struct mdev_vfio_device_ops vfio_dev_ops = {
+	.open		= mtty_open,
+	.release	= mtty_close,
+	.read		= mtty_read,
+	.write		= mtty_write,
+	.ioctl		= mtty_ioctl,
+};
+
 static const struct mdev_parent_ops mdev_fops = {
 	.owner                  = THIS_MODULE,
 	.dev_attr_groups        = mtty_dev_groups,
@@ -1395,11 +1406,6 @@ static const struct mdev_parent_ops mdev_fops = {
 	.supported_type_groups  = mdev_type_groups,
 	.create                 = mtty_create,
 	.remove			= mtty_remove,
-	.open                   = mtty_open,
-	.release                = mtty_close,
-	.read                   = mtty_read,
-	.write                  = mtty_write,
-	.ioctl		        = mtty_ioctl,
 };
 
 static void mtty_device_release(struct device *dev)
Signature · 1 line
2.19.1
author
#7 lore raw

PATCH v10 4/6 mdev: introduce virtio device and its device ops

In reply to Jason Wang

This patch implements basic support for mdev driver that supports virtio transport for kernel virtio driver.

---
 MAINTAINERS                      |   1 +
 drivers/vfio/mdev/mdev_core.c    |  21 +++++
 drivers/vfio/mdev/mdev_private.h |   2 +
 include/linux/mdev.h             |   6 ++
 include/linux/mdev_virtio_ops.h  | 147 +++++++++++++++++++++++++++++++
 5 files changed, 177 insertions(+)
 create mode 100644 include/linux/mdev_virtio_ops.h
MAINTAINERS +1 -0
index f661d13344d6..4997957443df 100644
@@ -17248,6 +17248,7 @@ F:	include/linux/virtio*.h
 F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
+F:	include/linux/mdev_virtio_ops.h
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
drivers/vfio/mdev/mdev_core.c +21 -0
index 4e70f19ac145..c58253404ed5 100644
@@ -78,6 +78,27 @@ const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev)
 }
 EXPORT_SYMBOL(mdev_get_vfio_ops);
 
+/*
+ * Specify the virtio device ops for the mdev device, this
+ * must be called during create() callback for virtio mdev device.
+ */
+void mdev_set_virtio_ops(struct mdev_device *mdev,
+			 const struct mdev_virtio_device_ops *virtio_ops)
+{
+	mdev_set_class(mdev, MDEV_CLASS_ID_VIRTIO);
+	mdev->virtio_ops = virtio_ops;
+}
+EXPORT_SYMBOL(mdev_set_virtio_ops);
+
+/* Get the virtio device ops for the mdev device. */
+const struct mdev_virtio_device_ops *
+mdev_get_virtio_ops(struct mdev_device *mdev)
+{
+	WARN_ON(mdev->class_id != MDEV_CLASS_ID_VIRTIO);
+	return mdev->virtio_ops;
+}
+EXPORT_SYMBOL(mdev_get_virtio_ops);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
drivers/vfio/mdev/mdev_private.h +2 -0
index 411227373625..2c74dd032409 100644
@@ -11,6 +11,7 @@
 #define MDEV_PRIVATE_H
 
 #include <linux/mdev_vfio_ops.h>
+#include <linux/mdev_virtio_ops.h>
 
 int  mdev_bus_register(void);
 void mdev_bus_unregister(void);
@@ -38,6 +39,7 @@ struct mdev_device {
 	u16 class_id;
 	union {
 		const struct mdev_vfio_device_ops *vfio_ops;
+		const struct mdev_virtio_device_ops *virtio_ops;
 	};
 };
 
include/linux/mdev.h +6 -0
index 9e37506d1987..f3d75a60c2b5 100644
@@ -17,6 +17,7 @@
 
 struct mdev_device;
 struct mdev_vfio_device_ops;
+struct mdev_virtio_device_ops;
 
 /*
  * Called by the parent device driver to set the device which represents
@@ -112,6 +113,10 @@ void mdev_set_class(struct mdev_device *mdev, u16 id);
 void mdev_set_vfio_ops(struct mdev_device *mdev,
 		       const struct mdev_vfio_device_ops *vfio_ops);
 const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev);
+void mdev_set_virtio_ops(struct mdev_device *mdev,
+			 const struct mdev_virtio_device_ops *virtio_ops);
+const struct mdev_virtio_device_ops *
+mdev_get_virtio_ops(struct mdev_device *mdev);
 
 extern struct bus_type mdev_bus_type;
 
@@ -127,6 +132,7 @@ struct mdev_device *mdev_from_dev(struct device *dev);
 
 enum {
 	MDEV_CLASS_ID_VFIO = 1,
+	MDEV_CLASS_ID_VIRTIO = 2,
 	/* New entries must be added here */
 };
 
include/linux/mdev_virtio_ops.h new file +147 -0

147 lines not on this page · raw message

Signature · 1 line
2.19.1
author
#8 lore raw

PATCH v10 4/6 mdev: introduce virtio device and its device ops

In reply to Jason Wang

This patch implements basic support for mdev driver that supports virtio transport for kernel virtio driver.

---
 MAINTAINERS                      |   1 +
 drivers/vfio/mdev/mdev_core.c    |  21 +++++
 drivers/vfio/mdev/mdev_private.h |   2 +
 include/linux/mdev.h             |   6 ++
 include/linux/mdev_virtio_ops.h  | 147 +++++++++++++++++++++++++++++++
 5 files changed, 177 insertions(+)
 create mode 100644 include/linux/mdev_virtio_ops.h
MAINTAINERS +1 -0
index f661d13344d6..4997957443df 100644
@@ -17248,6 +17248,7 @@ F:	include/linux/virtio*.h
 F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
+F:	include/linux/mdev_virtio_ops.h
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
drivers/vfio/mdev/mdev_core.c +21 -0
index 4e70f19ac145..c58253404ed5 100644
@@ -78,6 +78,27 @@ const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev)
 }
 EXPORT_SYMBOL(mdev_get_vfio_ops);
 
+/*
+ * Specify the virtio device ops for the mdev device, this
+ * must be called during create() callback for virtio mdev device.
+ */
+void mdev_set_virtio_ops(struct mdev_device *mdev,
+			 const struct mdev_virtio_device_ops *virtio_ops)
+{
+	mdev_set_class(mdev, MDEV_CLASS_ID_VIRTIO);
+	mdev->virtio_ops = virtio_ops;
+}
+EXPORT_SYMBOL(mdev_set_virtio_ops);
+
+/* Get the virtio device ops for the mdev device. */
+const struct mdev_virtio_device_ops *
+mdev_get_virtio_ops(struct mdev_device *mdev)
+{
+	WARN_ON(mdev->class_id != MDEV_CLASS_ID_VIRTIO);
+	return mdev->virtio_ops;
+}
+EXPORT_SYMBOL(mdev_get_virtio_ops);
+
 struct device *mdev_dev(struct mdev_device *mdev)
 {
 	return &mdev->dev;
drivers/vfio/mdev/mdev_private.h +2 -0
index 411227373625..2c74dd032409 100644
@@ -11,6 +11,7 @@
 #define MDEV_PRIVATE_H
 
 #include <linux/mdev_vfio_ops.h>
+#include <linux/mdev_virtio_ops.h>
 
 int  mdev_bus_register(void);
 void mdev_bus_unregister(void);
@@ -38,6 +39,7 @@ struct mdev_device {
 	u16 class_id;
 	union {
 		const struct mdev_vfio_device_ops *vfio_ops;
+		const struct mdev_virtio_device_ops *virtio_ops;
 	};
 };
 
include/linux/mdev.h +6 -0
index 9e37506d1987..f3d75a60c2b5 100644
@@ -17,6 +17,7 @@
 
 struct mdev_device;
 struct mdev_vfio_device_ops;
+struct mdev_virtio_device_ops;
 
 /*
  * Called by the parent device driver to set the device which represents
@@ -112,6 +113,10 @@ void mdev_set_class(struct mdev_device *mdev, u16 id);
 void mdev_set_vfio_ops(struct mdev_device *mdev,
 		       const struct mdev_vfio_device_ops *vfio_ops);
 const struct mdev_vfio_device_ops *mdev_get_vfio_ops(struct mdev_device *mdev);
+void mdev_set_virtio_ops(struct mdev_device *mdev,
+			 const struct mdev_virtio_device_ops *virtio_ops);
+const struct mdev_virtio_device_ops *
+mdev_get_virtio_ops(struct mdev_device *mdev);
 
 extern struct bus_type mdev_bus_type;
 
@@ -127,6 +132,7 @@ struct mdev_device *mdev_from_dev(struct device *dev);
 
 enum {
 	MDEV_CLASS_ID_VFIO = 1,
+	MDEV_CLASS_ID_VIRTIO = 2,
 	/* New entries must be added here */
 };
 
include/linux/mdev_virtio_ops.h new file +147 -0

147 lines not on this page · raw message

Signature · 1 line
2.19.1
author
#9 lore raw

PATCH v10 5/6 virtio: introduce a mdev based transport

In reply to Jason Wang

This patch introduces a new mdev transport for virtio. This is used to use kernel virtio driver to drive the mediated device that is capable of populating virtqueue directly.

A new virtio-mdev driver will be registered to the mdev bus, when a new virtio-mdev device is probed, it will register the device with mdev based config ops. This means it is a software transport between mdev driver and mdev device. The transport was implemented through virtio device specific ops.

---
 drivers/virtio/Kconfig       |  13 ++
 drivers/virtio/Makefile      |   1 +
 drivers/virtio/virtio_mdev.c | 406 +++++++++++++++++++++++++++++++++++
 3 files changed, 420 insertions(+)
 create mode 100644 drivers/virtio/virtio_mdev.c
drivers/virtio/Kconfig +13 -0
index 078615cf2afc..bf526ce0facc 100644
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
 
 	  If unsure, say Y.
 
+config VIRTIO_MDEV
+	tristate "MDEV driver for virtio devices"
+	depends on VFIO_MDEV && VIRTIO
+	default n
+	help
+	  This driver provides support for virtio based paravirtual
+	  device driver over MDEV bus. For this to be useful, you need
+	  an appropriate virtio mdev device implementation that
+	  operates on a physical device to allow the datapath of virtio
+	  to be offloaded to hardware.
+
+	  If unsure, say M.
+
 config VIRTIO_PMEM
 	tristate "Support for virtio pmem driver"
 	depends on VIRTIO
drivers/virtio/Makefile +1 -0
index 3a2b5c5dcf46..f2997b6c812f 100644
@@ -6,3 +6,4 @@ virtio_pci-y := virtio_pci_modern.o virtio_pci_common.o
 virtio_pci-$(CONFIG_VIRTIO_PCI_LEGACY) += virtio_pci_legacy.o
 obj-$(CONFIG_VIRTIO_BALLOON) += virtio_balloon.o
 obj-$(CONFIG_VIRTIO_INPUT) += virtio_input.o
+obj-$(CONFIG_VIRTIO_MDEV) += virtio_mdev.o
drivers/virtio/virtio_mdev.c new file +406 -0

406 lines not on this page · raw message

Signature · 1 line
2.19.1
author
#10 lore raw

PATCH v10 5/6 virtio: introduce a mdev based transport

In reply to Jason Wang

This patch introduces a new mdev transport for virtio. This is used to use kernel virtio driver to drive the mediated device that is capable of populating virtqueue directly.

A new virtio-mdev driver will be registered to the mdev bus, when a new virtio-mdev device is probed, it will register the device with mdev based config ops. This means it is a software transport between mdev driver and mdev device. The transport was implemented through virtio device specific ops.

---
 drivers/virtio/Kconfig       |  13 ++
 drivers/virtio/Makefile      |   1 +
 drivers/virtio/virtio_mdev.c | 406 +++++++++++++++++++++++++++++++++++
 3 files changed, 420 insertions(+)
 create mode 100644 drivers/virtio/virtio_mdev.c
drivers/virtio/Kconfig +13 -0
index 078615cf2afc..bf526ce0facc 100644
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
 
 	  If unsure, say Y.
 
+config VIRTIO_MDEV
+	tristate "MDEV driver for virtio devices"
+	depends on VFIO_MDEV && VIRTIO
+	default n
+	help
+	  This driver provides support for virtio based paravirtual
+	  device driver over MDEV bus. For this to be useful, you need
+	  an appropriate virtio mdev device implementation that
+	  operates on a physical device to allow the datapath of virtio
+	  to be offloaded to hardware.
+
+	  If unsure, say M.
+
 config VIRTIO_PMEM
 	tristate "Support for virtio pmem driver"
 	depends on VIRTIO
drivers/virtio/Makefile +1 -0
index 3a2b5c5dcf46..f2997b6c812f 100644
@@ -6,3 +6,4 @@ virtio_pci-y := virtio_pci_modern.o virtio_pci_common.o
 virtio_pci-$(CONFIG_VIRTIO_PCI_LEGACY) += virtio_pci_legacy.o
 obj-$(CONFIG_VIRTIO_BALLOON) += virtio_balloon.o
 obj-$(CONFIG_VIRTIO_INPUT) += virtio_input.o
+obj-$(CONFIG_VIRTIO_MDEV) += virtio_mdev.o
drivers/virtio/virtio_mdev.c new file +406 -0

406 lines not on this page · raw message

Signature · 1 line
2.19.1
author
#11 lore raw

PATCH v10 6/6 docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

---
 MAINTAINERS                |   1 +
 samples/Kconfig            |  10 +
 samples/vfio-mdev/Makefile |   1 +
 samples/vfio-mdev/mvnet.c  | 686 +++++++++++++++++++++++++++++++++++++
 4 files changed, 698 insertions(+)
 create mode 100644 samples/vfio-mdev/mvnet.c
MAINTAINERS +1 -0
index 4997957443df..6e9ad105a28f 100644
@@ -17249,6 +17249,7 @@ F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
 F:	include/linux/mdev_virtio_ops.h
+F:	samples/vfio-mdev/mvnet.c
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
samples/Kconfig +10 -0
index c8dacb4dda80..13a2443e18e0 100644
@@ -131,6 +131,16 @@ config SAMPLE_VFIO_MDEV_MDPY
 	  mediated device.  It is a simple framebuffer and supports
 	  the region display interface (VFIO_GFX_PLANE_TYPE_REGION).
 
+config SAMPLE_VIRTIO_MDEV_NET
+	tristate "Build VIRTIO net example mediated device sample code -- loadable modules only"
+	depends on VIRTIO_MDEV && VHOST_RING && m
+	help
+	  Build a networking sample device for use as a virtio
+	  mediated device. The device coopreates with virtio-mdev bus
+	  driver to present an virtio ethernet driver for
+	  kernel. It simply loopbacks all packets from its TX
+	  virtqueue to its RX virtqueue.
+
 config SAMPLE_VFIO_MDEV_MDPY_FB
 	tristate "Build VFIO mdpy example guest fbdev driver -- loadable module only"
 	depends on FB && m
samples/vfio-mdev/Makefile +1 -0
index 10d179c4fdeb..f34af90ed0a0 100644
@@ -3,3 +3,4 @@ obj-$(CONFIG_SAMPLE_VFIO_MDEV_MTTY) += mtty.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY) += mdpy.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY_FB) += mdpy-fb.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MBOCHS) += mbochs.o
+obj-$(CONFIG_SAMPLE_VIRTIO_MDEV_NET) += mvnet.o
samples/vfio-mdev/mvnet.c new file +686 -0

686 lines not on this page · raw message

Signature · 6 lines
2.19.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#12 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 6 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Quoted 6 lines

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Quoted 744 lines
---
 MAINTAINERS                |   1 +
 samples/Kconfig            |  10 +
 samples/vfio-mdev/Makefile |   1 +
 samples/vfio-mdev/mvnet.c  | 686 +++++++++++++++++++++++++++++++++++++
 4 files changed, 698 insertions(+)
 create mode 100644 samples/vfio-mdev/mvnet.c
MAINTAINERS +1 -0
index 4997957443df..6e9ad105a28f 100644
@@ -17249,6 +17249,7 @@ F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
 F:	include/linux/mdev_virtio_ops.h
+F:	samples/vfio-mdev/mvnet.c
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
samples/Kconfig +10 -0
index c8dacb4dda80..13a2443e18e0 100644
@@ -131,6 +131,16 @@ config SAMPLE_VFIO_MDEV_MDPY
 	  mediated device.  It is a simple framebuffer and supports
 	  the region display interface (VFIO_GFX_PLANE_TYPE_REGION).
 
+config SAMPLE_VIRTIO_MDEV_NET
+	tristate "Build VIRTIO net example mediated device sample code -- loadable modules only"
+	depends on VIRTIO_MDEV && VHOST_RING && m
+	help
+	  Build a networking sample device for use as a virtio
+	  mediated device. The device coopreates with virtio-mdev bus
+	  driver to present an virtio ethernet driver for
+	  kernel. It simply loopbacks all packets from its TX
+	  virtqueue to its RX virtqueue.
+
 config SAMPLE_VFIO_MDEV_MDPY_FB
 	tristate "Build VFIO mdpy example guest fbdev driver -- loadable module only"
 	depends on FB && m
samples/vfio-mdev/Makefile +1 -0
index 10d179c4fdeb..f34af90ed0a0 100644
@@ -3,3 +3,4 @@ obj-$(CONFIG_SAMPLE_VFIO_MDEV_MTTY) += mtty.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY) += mdpy.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY_FB) += mdpy-fb.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MBOCHS) += mbochs.o
+obj-$(CONFIG_SAMPLE_VIRTIO_MDEV_NET) += mvnet.o
samples/vfio-mdev/mvnet.c new file +686 -0
new file mode 100644
index 000000000000..a89aecfab68a
@@ -0,0 +1,686 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated virtual virtio-net device driver.
+ *
+ * Copyright (c) 2019, Red Hat Inc. All rights reserved.
+ *     Author: Jason Wang <[email protected]>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet
+ * # cd /sys/devices/virtual/mvnet/mvnet/mdev_supported_types/mvnet-virtio
+ * # echo "83b8f4f2-509f-382f-3c1e-e6bfe0fa1001" > ./create
+ * # cd devices/83b8f4f2-509f-382f-3c1e-e6bfe0fa1001
+ * # ls -d virtio0
+ * virtio0
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/poll.h>
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+#include <linux/uuid.h>
+#include <linux/iommu.h>
+#include <linux/sysfs.h>
+#include <linux/file.h>
+#include <linux/etherdevice.h>
+#include <linux/mdev.h>
+#include <linux/vringh.h>
+#include <linux/mdev_virtio_ops.h>
+#include <uapi/linux/virtio_config.h>
+#include <uapi/linux/virtio_net.h>
+
+#define VERSION_STRING  "0.1"
+#define DRIVER_AUTHOR   "Red Hat Corporation"
+
+#define MVNET_CLASS_NAME "mvnet"
+#define MVNET_NAME       "mvnet"
+
+/*
+ * Global Structures
+ */
+
+static struct mvnet_dev {
+	struct class	*vd_class;
+	struct idr	vd_idr;
+	struct device	dev;
+} mvnet_dev;
+
+struct mvnet_virtqueue {
+	struct vringh vring;
+	struct vringh_kiov iov;
+	unsigned short head;
+	bool ready;
+	u64 desc_addr;
+	u64 device_addr;
+	u64 driver_addr;
+	u32 num;
+	void *private;
+	irqreturn_t (*cb)(void *data);
+};
+
+#define MVNET_QUEUE_ALIGN PAGE_SIZE
+#define MVNET_QUEUE_MAX 256
+#define MVNET_DEVICE_ID 0x1
+#define MVNET_VENDOR_ID 0
+
+u64 mvnet_features = (1ULL << VIRTIO_F_ANY_LAYOUT) |
+		     (1ULL << VIRTIO_F_VERSION_1) |
+		     (1ULL << VIRTIO_F_IOMMU_PLATFORM);
+
+/* State of each mdev device */
+struct mvnet_state {
+	struct mvnet_virtqueue vqs[2];
+	struct work_struct work;
+	/* spinlock to synchronize virtqueue state */
+	spinlock_t lock;
+	struct mdev_device *mdev;
+	struct virtio_net_config config;
+	void *buffer;
+	u32 status;
+	u32 generation;
+	u64 features;
+	struct list_head next;
+};
+
+static struct mutex mdev_list_lock;
+static struct list_head mdev_devices_list;
+
+static void mvnet_queue_ready(struct mvnet_state *mvnet, unsigned int idx)
+{
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	int ret;
+
+	ret = vringh_init_kern(&vq->vring, mvnet_features, MVNET_QUEUE_MAX,
+			       false, (struct vring_desc *)vq->desc_addr,
+			       (struct vring_avail *)vq->driver_addr,
+			       (struct vring_used *)vq->device_addr);
+}
+
+static void mvnet_vq_reset(struct mvnet_virtqueue *vq)
+{
+	vq->ready = 0;
+	vq->desc_addr = 0;
+	vq->driver_addr = 0;
+	vq->device_addr = 0;
+	vq->cb = NULL;
+	vq->private = NULL;
+	vringh_init_kern(&vq->vring, mvnet_features, MVNET_QUEUE_MAX,
+			 false, 0, 0, 0);
+}
+
+static void mvnet_reset(struct mvnet_state *mvnet)
+{
+	int i;
+
+	for (i = 0; i < 2; i++)
+		mvnet_vq_reset(&mvnet->vqs[i]);
+
+	mvnet->features = 0;
+	mvnet->status = 0;
+	++mvnet->generation;
+}
+
+static void mvnet_work(struct work_struct *work)
+{
+	struct mvnet_state *mvnet = container_of(work, struct
+						 mvnet_state, work);
+	struct mvnet_virtqueue *txq = &mvnet->vqs[1];
+	struct mvnet_virtqueue *rxq = &mvnet->vqs[0];
+	size_t read, write, total_write;
+	int err;
+	int pkts = 0;
+
+	spin_lock(&mvnet->lock);
+
+	if (!txq->ready || !rxq->ready)
+		goto out;
+
+	while (true) {
+		total_write = 0;
+		err = vringh_getdesc_kern(&txq->vring, &txq->iov, NULL,
+					  &txq->head, GFP_ATOMIC);
+		if (err <= 0)
+			break;
+
+		err = vringh_getdesc_kern(&rxq->vring, NULL, &rxq->iov,
+					  &rxq->head, GFP_ATOMIC);
+		if (err <= 0) {
+			vringh_complete_kern(&txq->vring, txq->head, 0);
+			break;
+		}
+
+		while (true) {
+			read = vringh_iov_pull_kern(&txq->iov, mvnet->buffer,
+						    PAGE_SIZE);
+			if (read <= 0)
+				break;
+
+			write = vringh_iov_push_kern(&rxq->iov, mvnet->buffer,
+						     read);
+			if (write <= 0)
+				break;
+
+			total_write += write;
+		}
+
+		/* Make sure data is wrote before advancing index */
+		smp_wmb();
+
+		vringh_complete_kern(&txq->vring, txq->head, 0);
+		vringh_complete_kern(&rxq->vring, rxq->head, total_write);
+
+		/* Make sure used is visible before rasing the interrupt. */
+		smp_wmb();
+
+		local_bh_disable();
+		if (txq->cb)
+			txq->cb(txq->private);
+		if (rxq->cb)
+			rxq->cb(rxq->private);
+		local_bh_enable();
+
+		if (++pkts > 4) {
+			schedule_work(&mvnet->work);
+			goto out;
+		}
+	}
+
+out:
+	spin_unlock(&mvnet->lock);
+}
+
+static dma_addr_t mvnet_map_page(struct device *dev, struct page *page,
+				 unsigned long offset, size_t size,
+				 enum dma_data_direction dir,
+				 unsigned long attrs)
+{
+	/* Vringh can only use HVA */
+	return (dma_addr_t)(page_address(page) + offset);
+}
+
+static void mvnet_unmap_page(struct device *dev, dma_addr_t dma_addr,
+			     size_t size, enum dma_data_direction dir,
+			     unsigned long attrs)
+{
+}
+
+static void *mvnet_alloc_coherent(struct device *dev, size_t size,
+				  dma_addr_t *dma_addr, gfp_t flag,
+				  unsigned long attrs)
+{
+	void *addr = kmalloc(size, flag);
+
+	if (!addr)
+		*dma_addr = DMA_MAPPING_ERROR;
+	else
+		*dma_addr = (dma_addr_t)addr;
+
+	return addr;
+}
+
+static void mvnet_free_coherent(struct device *dev, size_t size,
+				void *vaddr, dma_addr_t dma_addr,
+				unsigned long attrs)
+{
+	kfree((void *)dma_addr);
+}
+
+static const struct dma_map_ops mvnet_dma_ops = {
+	.map_page = mvnet_map_page,
+	.unmap_page = mvnet_unmap_page,
+	.alloc = mvnet_alloc_coherent,
+	.free = mvnet_free_coherent,
+};
+
+static const struct mdev_virtio_device_ops mdev_virtio_ops;
+
+static int mvnet_create(struct kobject *kobj, struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet;
+	struct virtio_net_config *config;
+	struct device *dev = mdev_dev(mdev);
+
+	if (!mdev)
+		return -EINVAL;
+
+	mvnet = kzalloc(sizeof(*mvnet), GFP_KERNEL);
+	if (!mvnet)
+		return -ENOMEM;
+
+	mvnet->buffer = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!mvnet->buffer) {
+		kfree(mvnet);
+		return -ENOMEM;
+	}
+
+	config = &mvnet->config;
+	config->mtu = 1500;
+	config->status = VIRTIO_NET_S_LINK_UP;
+	eth_random_addr(config->mac);
+
+	INIT_WORK(&mvnet->work, mvnet_work);
+
+	spin_lock_init(&mvnet->lock);
+	mvnet->mdev = mdev;
+	mdev_set_drvdata(mdev, mvnet);
+
+	mutex_lock(&mdev_list_lock);
+	list_add(&mvnet->next, &mdev_devices_list);
+	mutex_unlock(&mdev_list_lock);
+
+	dev->coherent_dma_mask = DMA_BIT_MASK(64);
+	set_dma_ops(dev, &mvnet_dma_ops);
+
+	mdev_set_virtio_ops(mdev, &mdev_virtio_ops);
+
+	return 0;
+}
+
+static int mvnet_remove(struct mdev_device *mdev)
+{
+	struct mvnet_state *mds, *tmp_mds;
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	int ret = -EINVAL;
+
+	mutex_lock(&mdev_list_lock);
+	list_for_each_entry_safe(mds, tmp_mds, &mdev_devices_list, next) {
+		if (mvnet == mds) {
+			list_del(&mvnet->next);
+			mdev_set_drvdata(mdev, NULL);
+			kfree(mvnet->buffer);
+			kfree(mvnet);
+			ret = 0;
+			break;
+		}
+	}
+	mutex_unlock(&mdev_list_lock);
+
+	return ret;
+}
+
+static ssize_t
+sample_mvnet_dev_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
+{
+	if (mdev_from_dev(dev))
+		return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
+
+	return sprintf(buf, "\n");
+}
+
+static DEVICE_ATTR_RO(sample_mvnet_dev);
+
+static struct attribute *mvnet_dev_attrs[] = {
+	&dev_attr_sample_mvnet_dev.attr,
+	NULL,
+};
+
+static const struct attribute_group mvnet_dev_group = {
+	.name  = "mvnet_dev",
+	.attrs = mvnet_dev_attrs,
+};
+
+static const struct attribute_group *mvnet_dev_groups[] = {
+	&mvnet_dev_group,
+	NULL,
+};
+
+static ssize_t
+sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
+		     char *buf)
+{
+	if (mdev_from_dev(dev))
+		return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
+
+	return sprintf(buf, "\n");
+}
+
+static DEVICE_ATTR_RO(sample_mdev_dev);
+
+static struct attribute *mdev_dev_attrs[] = {
+	&dev_attr_sample_mdev_dev.attr,
+	NULL,
+};
+
+static const struct attribute_group mdev_dev_group = {
+	.name  = "vendor",
+	.attrs = mdev_dev_attrs,
+};
+
+static const struct attribute_group *mdev_dev_groups[] = {
+	&mdev_dev_group,
+	NULL,
+};
+
+#define MVNET_STRING_LEN 16
+
+static ssize_t
+name_show(struct kobject *kobj, struct device *dev, char *buf)
+{
+	char name[MVNET_STRING_LEN];
+	const char *name_str = "virtio-net";
+
+	snprintf(name, MVNET_STRING_LEN, "%s", dev_driver_string(dev));
+	if (!strcmp(kobj->name, name))
+		return sprintf(buf, "%s\n", name_str);
+
+	return -EINVAL;
+}
+
+static MDEV_TYPE_ATTR_RO(name);
+
+static ssize_t
+available_instances_show(struct kobject *kobj, struct device *dev, char *buf)
+{
+	return sprintf(buf, "%d\n", INT_MAX);
+}
+
+static MDEV_TYPE_ATTR_RO(available_instances);
+
+static ssize_t device_api_show(struct kobject *kobj, struct device *dev,
+			       char *buf)
+{
+	return sprintf(buf, "%s\n", VIRTIO_MDEV_DEVICE_API_STRING);
+}
+
+static MDEV_TYPE_ATTR_RO(device_api);
+
+static struct attribute *mdev_types_attrs[] = {
+	&mdev_type_attr_name.attr,
+	&mdev_type_attr_device_api.attr,
+	&mdev_type_attr_available_instances.attr,
+	NULL,
+};
+
+static struct attribute_group mdev_type_group = {
+	.name  = "virtio",
+	.attrs = mdev_types_attrs,
+};
+
+/* TBD: "vhost" type */
+
+static struct attribute_group *mdev_type_groups[] = {
+	&mdev_type_group,
+	NULL,
+};
+
+static int mvnet_set_vq_address(struct mdev_device *mdev, u16 idx,
+				u64 desc_area, u64 driver_area, u64 device_area)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->desc_addr = desc_area;
+	vq->driver_addr = driver_area;
+	vq->device_addr = device_area;
+
+	return 0;
+}
+
+static void mvnet_set_vq_num(struct mdev_device *mdev, u16 idx, u32 num)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->num = num;
+}
+
+static void mvnet_kick_vq(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	if (vq->ready)
+		schedule_work(&mvnet->work);
+}
+
+static void mvnet_set_vq_cb(struct mdev_device *mdev, u16 idx,
+			    struct virtio_mdev_callback *cb)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->cb = cb->callback;
+	vq->private = cb->private;
+}
+
+static void mvnet_set_vq_ready(struct mdev_device *mdev, u16 idx, bool ready)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	spin_lock(&mvnet->lock);
+	vq->ready = ready;
+	if (vq->ready)
+		mvnet_queue_ready(mvnet, idx);
+	spin_unlock(&mvnet->lock);
+}
+
+static bool mvnet_get_vq_ready(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	return vq->ready;
+}
+
+static int mvnet_set_vq_state(struct mdev_device *mdev, u16 idx, u64 state)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	spin_lock(&mvnet->lock);
+	vrh->last_avail_idx = state;
+	spin_unlock(&mvnet->lock);
+
+	return 0;
+}
+
+static u64 mvnet_get_vq_state(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	return vrh->last_avail_idx;
+}
+
+static u16 mvnet_get_vq_align(struct mdev_device *mdev)
+{
+	return MVNET_QUEUE_ALIGN;
+}
+
+static u64 mvnet_get_features(struct mdev_device *mdev)
+{
+	return mvnet_features;
+}
+
+static int mvnet_set_features(struct mdev_device *mdev, u64 features)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	/* DMA mapping must be done by driver */
+	if (!(features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)))
+		return -EINVAL;
+
+	mvnet->features = features & mvnet_features;
+
+	return 0;
+}
+
+static void mvnet_set_config_cb(struct mdev_device *mdev,
+				struct virtio_mdev_callback *cb)
+{
+	/* We don't support config interrupt */
+}
+
+static u16 mvnet_get_vq_num_max(struct mdev_device *mdev)
+{
+	return MVNET_QUEUE_MAX;
+}
+
+static u32 mvnet_get_device_id(struct mdev_device *mdev)
+{
+	return MVNET_DEVICE_ID;
+}
+
+static u32 mvnet_get_vendor_id(struct mdev_device *mdev)
+{
+	return MVNET_VENDOR_ID;
+}
+
+static u8 mvnet_get_status(struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	return mvnet->status;
+}
+
+static void mvnet_set_status(struct mdev_device *mdev, u8 status)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	mvnet->status = status;
+
+	if (status == 0) {
+		spin_lock(&mvnet->lock);
+		mvnet_reset(mvnet);
+		spin_unlock(&mvnet->lock);
+	}
+}
+
+static void mvnet_get_config(struct mdev_device *mdev, unsigned int offset,
+			     void *buf, unsigned int len)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	if (offset + len < sizeof(struct virtio_net_config))
+		memcpy(buf, &mvnet->config + offset, len);
+}
+
+static void mvnet_set_config(struct mdev_device *mdev, unsigned int offset,
+			     const void *buf, unsigned int len)
+{
+	/* No writable config supportted by mvnet */
+}
+
+static u32 mvnet_get_generation(struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	return mvnet->generation;
+}
+
+static const struct mdev_virtio_device_ops mdev_virtio_ops = {
+	.set_vq_address         = mvnet_set_vq_address,
+	.set_vq_num             = mvnet_set_vq_num,
+	.kick_vq                = mvnet_kick_vq,
+	.set_vq_cb              = mvnet_set_vq_cb,
+	.set_vq_ready           = mvnet_set_vq_ready,
+	.get_vq_ready           = mvnet_get_vq_ready,
+	.set_vq_state           = mvnet_set_vq_state,
+	.get_vq_state           = mvnet_get_vq_state,
+	.get_vq_align           = mvnet_get_vq_align,
+	.get_features           = mvnet_get_features,
+	.set_features           = mvnet_set_features,
+	.set_config_cb          = mvnet_set_config_cb,
+	.get_vq_num_max         = mvnet_get_vq_num_max,
+	.get_device_id          = mvnet_get_device_id,
+	.get_vendor_id          = mvnet_get_vendor_id,
+	.get_status             = mvnet_get_status,
+	.set_status             = mvnet_set_status,
+	.get_config             = mvnet_get_config,
+	.set_config             = mvnet_set_config,
+	.get_generation         = mvnet_get_generation,
+};
+
+static const struct mdev_parent_ops mdev_fops = {
+	.owner                  = THIS_MODULE,
+	.dev_attr_groups        = mvnet_dev_groups,
+	.mdev_attr_groups       = mdev_dev_groups,
+	.supported_type_groups  = mdev_type_groups,
+	.create                 = mvnet_create,
+	.remove			= mvnet_remove,
+};
+
+static void mvnet_device_release(struct device *dev)
+{
+	dev_dbg(dev, "mvnet: released\n");
+}
+
+static int __init mvnet_dev_init(void)
+{
+	int ret = 0;
+
+	pr_info("mvnet_dev: %s\n", __func__);
+
+	memset(&mvnet_dev, 0, sizeof(mvnet_dev));
+
+	idr_init(&mvnet_dev.vd_idr);
+
+	mvnet_dev.vd_class = class_create(THIS_MODULE, MVNET_CLASS_NAME);
+
+	if (IS_ERR(mvnet_dev.vd_class)) {
+		pr_err("Error: failed to register mvnet_dev class\n");
+		ret = PTR_ERR(mvnet_dev.vd_class);
+		goto failed1;
+	}
+
+	mvnet_dev.dev.class = mvnet_dev.vd_class;
+	mvnet_dev.dev.release = mvnet_device_release;
+	dev_set_name(&mvnet_dev.dev, "%s", MVNET_NAME);
+
+	ret = device_register(&mvnet_dev.dev);
+	if (ret)
+		goto failed2;
+
+	ret = mdev_register_device(&mvnet_dev.dev, &mdev_fops);
+	if (ret)
+		goto failed3;
+
+	mutex_init(&mdev_list_lock);
+	INIT_LIST_HEAD(&mdev_devices_list);
+
+	goto all_done;
+
+failed3:
+
+	device_unregister(&mvnet_dev.dev);
+failed2:
+	class_destroy(mvnet_dev.vd_class);
+
+failed1:
+all_done:
+	return ret;
+}
+
+static void __exit mvnet_dev_exit(void)
+{
+	mvnet_dev.dev.bus = NULL;
+	mdev_unregister_device(&mvnet_dev.dev);
+
+	device_unregister(&mvnet_dev.dev);
+	idr_destroy(&mvnet_dev.vd_idr);
+	class_destroy(mvnet_dev.vd_class);
+	mvnet_dev.vd_class = NULL;
+	pr_info("mvnet_dev: Unloaded!\n");
+}
+
+module_init(mvnet_dev_init)
+module_exit(mvnet_dev_exit)
+
+MODULE_LICENSE("GPL v2");
+MODULE_INFO(supported, "Simulate loopback ethernet device over mdev");
+MODULE_VERSION(VERSION_STRING);
+MODULE_AUTHOR(DRIVER_AUTHOR);
Signature · 1 line
2.19.1
author
#13 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

Quoted 2 lines
#14 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 20 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Quoted 5 lines

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 3 lines
Quoted 3 lines
author
#15 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

author
#16 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 30 lines · Jason Wang
Quoted 28 lines · Michael S. Tsirkin
Quoted 14 lines · Jason Wang
Quoted 13 lines · Michael S. Tsirkin
Quoted 12 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 8 lines
Quoted 3 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Quoted 8 lines
Quoted 1 line

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

#17 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 43 lines · Jason Wang
Quoted 41 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 8 lines
Quoted 1 line

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

Signature · 1 line
MST
author
#18 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 43 lines · Michael S. Tsirkin
Quoted 42 lines · Jason Wang
Quoted 40 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Quoted 7 lines
Quoted 5 lines

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

It has a prefix, so user will see "mvnet-virtio-loopback".

Thanks

Quoted 1 line
#19 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 50 lines · Jason Wang
Quoted 48 lines · Michael S. Tsirkin
Quoted 47 lines · Jason Wang
Quoted 33 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Quoted 8 lines
Quoted 2 lines

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

It has a prefix, so user will see "mvnet-virtio-loopback".

Thanks

yes but it's mvnet that is doing the loopback, not virtio

#20 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 50 lines · Jason Wang
Quoted 48 lines · Michael S. Tsirkin
Quoted 47 lines · Jason Wang
Quoted 33 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Quoted 8 lines
Quoted 2 lines

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

It has a prefix, so user will see "mvnet-virtio-loopback".

Thanks

yes but it's mvnet that is doing the loopback, not virtio

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

author
#21 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Michael S. Tsirkin
Quoted 43 lines · Michael S. Tsirkin
Quoted 42 lines · Jason Wang
Quoted 40 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Quoted 7 lines
Quoted 5 lines

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

It has a prefix, so user will see "mvnet-virtio-loopback".

Thanks

Quoted 1 line

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#22 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 43 lines · Jason Wang
Quoted 41 lines · Jason Wang
Quoted 24 lines · Michael S. Tsirkin
Quoted 23 lines · Jason Wang
Quoted 12 lines · Michael S. Tsirkin
Quoted 11 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device
over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device.
Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 8 lines
Quoted 1 line

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

Maybe but is virtio actually a loopback somehow? I thought we can bind a regular virtio device there, no?

Signature · 5 lines
MST
_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
author
#23 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 30 lines · Jason Wang
Quoted 28 lines · Michael S. Tsirkin
Quoted 14 lines · Jason Wang
Quoted 13 lines · Michael S. Tsirkin
Quoted 12 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 8 lines
Quoted 3 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Quoted 8 lines
Quoted 1 line

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

Or maybe it's better to rename the type of current mdev from 'virtio' to 'virtio-loopback'. Then we can add more types in the future.

Thanks

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#24 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 33 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

this can be a patch on top.
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author
#25 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 32 lines · Michael S. Tsirkin
Quoted 31 lines · Jason Wang
Quoted 29 lines · Michael S. Tsirkin
Quoted 14 lines · Jason Wang
Quoted 13 lines · Michael S. Tsirkin
Quoted 12 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 8 lines
Quoted 3 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 8 lines
Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

#26 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 40 lines · Jason Wang
Quoted 38 lines · Michael S. Tsirkin
Quoted 37 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 8 lines
Quoted 2 lines
Quoted 1 line

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

That would be a substantial change. Maybe drop 6/6 for now until we have a better handle on this?

author
#27 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

Quoted 39 lines · Michael S. Tsirkin
Quoted 38 lines · Jason Wang
Quoted 37 lines · Michael S. Tsirkin
Quoted 36 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 9 lines
Quoted 6 lines
Quoted 2 lines
Quoted 1 line

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

That would be a substantial change. Maybe drop 6/6 for now until we have a better handle on this?

Ok, consider the change should be small, I will post V11 where I can fix the typos spotted.

Thanks

author
#28 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Michael S. Tsirkin
Quoted 39 lines · Michael S. Tsirkin
Quoted 38 lines · Jason Wang
Quoted 37 lines · Michael S. Tsirkin
Quoted 36 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 9 lines
Quoted 6 lines
Quoted 2 lines
Quoted 1 line

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

That would be a substantial change. Maybe drop 6/6 for now until we have a better handle on this?

Ok, consider the change should be small, I will post V11 where I can fix the typos spotted.

Thanks

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#29 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 40 lines · Jason Wang
Quoted 38 lines · Michael S. Tsirkin
Quoted 37 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Quoted 8 lines
Quoted 2 lines
Quoted 1 line

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

That would be a substantial change. Maybe drop 6/6 for now until we have a better handle on this?

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

author
#30 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Michael S. Tsirkin
Quoted 32 lines · Michael S. Tsirkin
Quoted 31 lines · Jason Wang
Quoted 29 lines · Michael S. Tsirkin
Quoted 14 lines · Jason Wang
Quoted 13 lines · Michael S. Tsirkin
Quoted 12 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 8 lines
Quoted 3 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 8 lines
Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

this can be a patch on top.

Then maybe it's better just extend it to work as a normal networking device on top?

Thanks

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#31 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 33 lines · Jason Wang
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 7 lines

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

this can be a patch on top.
_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

author
#32 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Michael S. Tsirkin
Quoted 22 lines · Michael S. Tsirkin
Quoted 21 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Quoted 7 lines
Quoted 1 line

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

My understanding is that it was a sample driver in /doc. It should not be used in production environment. Otherwise we need to move it to driver/virtio.

But if you insist, I can post a V11.

Thanks

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#33 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 20 lines · Jason Wang
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Quoted 5 lines

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

That's a big change in an interface so it's a good reason to rename the driver at that point right?
Oherwise users of an old kernel would expect a stacked driver and get a loopback instead.

Or did I miss something?

Quoted 3 lines
Quoted 3 lines

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

author
#34 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Michael S. Tsirkin
Quoted 11 lines · Michael S. Tsirkin
Quoted 10 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

Quoted 5 lines
Quoted 2 lines

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Actually, I plan to extend it as another standard network interface for kernel. It could be either a standalone pseudo device or a stack device. Does this sounds good to you?

Thanks

Quoted 2 lines

_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

#35 lore raw

Re: [PATCH V10 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang
Quoted 6 lines · Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Quoted 6 lines

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

I'd prefer it that we call this something else, e.g. mvnet-loopback. Just so people don't expect a fully functional device somehow. Can be renamed when applying?

Quoted 744 lines
---
 MAINTAINERS                |   1 +
 samples/Kconfig            |  10 +
 samples/vfio-mdev/Makefile |   1 +
 samples/vfio-mdev/mvnet.c  | 686 +++++++++++++++++++++++++++++++++++++
 4 files changed, 698 insertions(+)
 create mode 100644 samples/vfio-mdev/mvnet.c
MAINTAINERS +1 -0
index 4997957443df..6e9ad105a28f 100644
@@ -17249,6 +17249,7 @@ F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
 F:	include/linux/mdev_virtio_ops.h
+F:	samples/vfio-mdev/mvnet.c
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
samples/Kconfig +10 -0
index c8dacb4dda80..13a2443e18e0 100644
@@ -131,6 +131,16 @@ config SAMPLE_VFIO_MDEV_MDPY
 	  mediated device.  It is a simple framebuffer and supports
 	  the region display interface (VFIO_GFX_PLANE_TYPE_REGION).
 
+config SAMPLE_VIRTIO_MDEV_NET
+	tristate "Build VIRTIO net example mediated device sample code -- loadable modules only"
+	depends on VIRTIO_MDEV && VHOST_RING && m
+	help
+	  Build a networking sample device for use as a virtio
+	  mediated device. The device coopreates with virtio-mdev bus
+	  driver to present an virtio ethernet driver for
+	  kernel. It simply loopbacks all packets from its TX
+	  virtqueue to its RX virtqueue.
+
 config SAMPLE_VFIO_MDEV_MDPY_FB
 	tristate "Build VFIO mdpy example guest fbdev driver -- loadable module only"
 	depends on FB && m
samples/vfio-mdev/Makefile +1 -0
index 10d179c4fdeb..f34af90ed0a0 100644
@@ -3,3 +3,4 @@ obj-$(CONFIG_SAMPLE_VFIO_MDEV_MTTY) += mtty.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY) += mdpy.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY_FB) += mdpy-fb.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MBOCHS) += mbochs.o
+obj-$(CONFIG_SAMPLE_VIRTIO_MDEV_NET) += mvnet.o
samples/vfio-mdev/mvnet.c new file +686 -0
new file mode 100644
index 000000000000..a89aecfab68a
@@ -0,0 +1,686 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated virtual virtio-net device driver.
+ *
+ * Copyright (c) 2019, Red Hat Inc. All rights reserved.
+ *     Author: Jason Wang <[email protected]>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet
+ * # cd /sys/devices/virtual/mvnet/mvnet/mdev_supported_types/mvnet-virtio
+ * # echo "83b8f4f2-509f-382f-3c1e-e6bfe0fa1001" > ./create
+ * # cd devices/83b8f4f2-509f-382f-3c1e-e6bfe0fa1001
+ * # ls -d virtio0
+ * virtio0
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/poll.h>
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+#include <linux/uuid.h>
+#include <linux/iommu.h>
+#include <linux/sysfs.h>
+#include <linux/file.h>
+#include <linux/etherdevice.h>
+#include <linux/mdev.h>
+#include <linux/vringh.h>
+#include <linux/mdev_virtio_ops.h>
+#include <uapi/linux/virtio_config.h>
+#include <uapi/linux/virtio_net.h>
+
+#define VERSION_STRING  "0.1"
+#define DRIVER_AUTHOR   "Red Hat Corporation"
+
+#define MVNET_CLASS_NAME "mvnet"
+#define MVNET_NAME       "mvnet"
+
+/*
+ * Global Structures
+ */
+
+static struct mvnet_dev {
+	struct class	*vd_class;
+	struct idr	vd_idr;
+	struct device	dev;
+} mvnet_dev;
+
+struct mvnet_virtqueue {
+	struct vringh vring;
+	struct vringh_kiov iov;
+	unsigned short head;
+	bool ready;
+	u64 desc_addr;
+	u64 device_addr;
+	u64 driver_addr;
+	u32 num;
+	void *private;
+	irqreturn_t (*cb)(void *data);
+};
+
+#define MVNET_QUEUE_ALIGN PAGE_SIZE
+#define MVNET_QUEUE_MAX 256
+#define MVNET_DEVICE_ID 0x1
+#define MVNET_VENDOR_ID 0
+
+u64 mvnet_features = (1ULL << VIRTIO_F_ANY_LAYOUT) |
+		     (1ULL << VIRTIO_F_VERSION_1) |
+		     (1ULL << VIRTIO_F_IOMMU_PLATFORM);
+
+/* State of each mdev device */
+struct mvnet_state {
+	struct mvnet_virtqueue vqs[2];
+	struct work_struct work;
+	/* spinlock to synchronize virtqueue state */
+	spinlock_t lock;
+	struct mdev_device *mdev;
+	struct virtio_net_config config;
+	void *buffer;
+	u32 status;
+	u32 generation;
+	u64 features;
+	struct list_head next;
+};
+
+static struct mutex mdev_list_lock;
+static struct list_head mdev_devices_list;
+
+static void mvnet_queue_ready(struct mvnet_state *mvnet, unsigned int idx)
+{
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	int ret;
+
+	ret = vringh_init_kern(&vq->vring, mvnet_features, MVNET_QUEUE_MAX,
+			       false, (struct vring_desc *)vq->desc_addr,
+			       (struct vring_avail *)vq->driver_addr,
+			       (struct vring_used *)vq->device_addr);
+}
+
+static void mvnet_vq_reset(struct mvnet_virtqueue *vq)
+{
+	vq->ready = 0;
+	vq->desc_addr = 0;
+	vq->driver_addr = 0;
+	vq->device_addr = 0;
+	vq->cb = NULL;
+	vq->private = NULL;
+	vringh_init_kern(&vq->vring, mvnet_features, MVNET_QUEUE_MAX,
+			 false, 0, 0, 0);
+}
+
+static void mvnet_reset(struct mvnet_state *mvnet)
+{
+	int i;
+
+	for (i = 0; i < 2; i++)
+		mvnet_vq_reset(&mvnet->vqs[i]);
+
+	mvnet->features = 0;
+	mvnet->status = 0;
+	++mvnet->generation;
+}
+
+static void mvnet_work(struct work_struct *work)
+{
+	struct mvnet_state *mvnet = container_of(work, struct
+						 mvnet_state, work);
+	struct mvnet_virtqueue *txq = &mvnet->vqs[1];
+	struct mvnet_virtqueue *rxq = &mvnet->vqs[0];
+	size_t read, write, total_write;
+	int err;
+	int pkts = 0;
+
+	spin_lock(&mvnet->lock);
+
+	if (!txq->ready || !rxq->ready)
+		goto out;
+
+	while (true) {
+		total_write = 0;
+		err = vringh_getdesc_kern(&txq->vring, &txq->iov, NULL,
+					  &txq->head, GFP_ATOMIC);
+		if (err <= 0)
+			break;
+
+		err = vringh_getdesc_kern(&rxq->vring, NULL, &rxq->iov,
+					  &rxq->head, GFP_ATOMIC);
+		if (err <= 0) {
+			vringh_complete_kern(&txq->vring, txq->head, 0);
+			break;
+		}
+
+		while (true) {
+			read = vringh_iov_pull_kern(&txq->iov, mvnet->buffer,
+						    PAGE_SIZE);
+			if (read <= 0)
+				break;
+
+			write = vringh_iov_push_kern(&rxq->iov, mvnet->buffer,
+						     read);
+			if (write <= 0)
+				break;
+
+			total_write += write;
+		}
+
+		/* Make sure data is wrote before advancing index */
+		smp_wmb();
+
+		vringh_complete_kern(&txq->vring, txq->head, 0);
+		vringh_complete_kern(&rxq->vring, rxq->head, total_write);
+
+		/* Make sure used is visible before rasing the interrupt. */
+		smp_wmb();
+
+		local_bh_disable();
+		if (txq->cb)
+			txq->cb(txq->private);
+		if (rxq->cb)
+			rxq->cb(rxq->private);
+		local_bh_enable();
+
+		if (++pkts > 4) {
+			schedule_work(&mvnet->work);
+			goto out;
+		}
+	}
+
+out:
+	spin_unlock(&mvnet->lock);
+}
+
+static dma_addr_t mvnet_map_page(struct device *dev, struct page *page,
+				 unsigned long offset, size_t size,
+				 enum dma_data_direction dir,
+				 unsigned long attrs)
+{
+	/* Vringh can only use HVA */
+	return (dma_addr_t)(page_address(page) + offset);
+}
+
+static void mvnet_unmap_page(struct device *dev, dma_addr_t dma_addr,
+			     size_t size, enum dma_data_direction dir,
+			     unsigned long attrs)
+{
+}
+
+static void *mvnet_alloc_coherent(struct device *dev, size_t size,
+				  dma_addr_t *dma_addr, gfp_t flag,
+				  unsigned long attrs)
+{
+	void *addr = kmalloc(size, flag);
+
+	if (!addr)
+		*dma_addr = DMA_MAPPING_ERROR;
+	else
+		*dma_addr = (dma_addr_t)addr;
+
+	return addr;
+}
+
+static void mvnet_free_coherent(struct device *dev, size_t size,
+				void *vaddr, dma_addr_t dma_addr,
+				unsigned long attrs)
+{
+	kfree((void *)dma_addr);
+}
+
+static const struct dma_map_ops mvnet_dma_ops = {
+	.map_page = mvnet_map_page,
+	.unmap_page = mvnet_unmap_page,
+	.alloc = mvnet_alloc_coherent,
+	.free = mvnet_free_coherent,
+};
+
+static const struct mdev_virtio_device_ops mdev_virtio_ops;
+
+static int mvnet_create(struct kobject *kobj, struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet;
+	struct virtio_net_config *config;
+	struct device *dev = mdev_dev(mdev);
+
+	if (!mdev)
+		return -EINVAL;
+
+	mvnet = kzalloc(sizeof(*mvnet), GFP_KERNEL);
+	if (!mvnet)
+		return -ENOMEM;
+
+	mvnet->buffer = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!mvnet->buffer) {
+		kfree(mvnet);
+		return -ENOMEM;
+	}
+
+	config = &mvnet->config;
+	config->mtu = 1500;
+	config->status = VIRTIO_NET_S_LINK_UP;
+	eth_random_addr(config->mac);
+
+	INIT_WORK(&mvnet->work, mvnet_work);
+
+	spin_lock_init(&mvnet->lock);
+	mvnet->mdev = mdev;
+	mdev_set_drvdata(mdev, mvnet);
+
+	mutex_lock(&mdev_list_lock);
+	list_add(&mvnet->next, &mdev_devices_list);
+	mutex_unlock(&mdev_list_lock);
+
+	dev->coherent_dma_mask = DMA_BIT_MASK(64);
+	set_dma_ops(dev, &mvnet_dma_ops);
+
+	mdev_set_virtio_ops(mdev, &mdev_virtio_ops);
+
+	return 0;
+}
+
+static int mvnet_remove(struct mdev_device *mdev)
+{
+	struct mvnet_state *mds, *tmp_mds;
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	int ret = -EINVAL;
+
+	mutex_lock(&mdev_list_lock);
+	list_for_each_entry_safe(mds, tmp_mds, &mdev_devices_list, next) {
+		if (mvnet == mds) {
+			list_del(&mvnet->next);
+			mdev_set_drvdata(mdev, NULL);
+			kfree(mvnet->buffer);
+			kfree(mvnet);
+			ret = 0;
+			break;
+		}
+	}
+	mutex_unlock(&mdev_list_lock);
+
+	return ret;
+}
+
+static ssize_t
+sample_mvnet_dev_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
+{
+	if (mdev_from_dev(dev))
+		return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
+
+	return sprintf(buf, "\n");
+}
+
+static DEVICE_ATTR_RO(sample_mvnet_dev);
+
+static struct attribute *mvnet_dev_attrs[] = {
+	&dev_attr_sample_mvnet_dev.attr,
+	NULL,
+};
+
+static const struct attribute_group mvnet_dev_group = {
+	.name  = "mvnet_dev",
+	.attrs = mvnet_dev_attrs,
+};
+
+static const struct attribute_group *mvnet_dev_groups[] = {
+	&mvnet_dev_group,
+	NULL,
+};
+
+static ssize_t
+sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
+		     char *buf)
+{
+	if (mdev_from_dev(dev))
+		return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
+
+	return sprintf(buf, "\n");
+}
+
+static DEVICE_ATTR_RO(sample_mdev_dev);
+
+static struct attribute *mdev_dev_attrs[] = {
+	&dev_attr_sample_mdev_dev.attr,
+	NULL,
+};
+
+static const struct attribute_group mdev_dev_group = {
+	.name  = "vendor",
+	.attrs = mdev_dev_attrs,
+};
+
+static const struct attribute_group *mdev_dev_groups[] = {
+	&mdev_dev_group,
+	NULL,
+};
+
+#define MVNET_STRING_LEN 16
+
+static ssize_t
+name_show(struct kobject *kobj, struct device *dev, char *buf)
+{
+	char name[MVNET_STRING_LEN];
+	const char *name_str = "virtio-net";
+
+	snprintf(name, MVNET_STRING_LEN, "%s", dev_driver_string(dev));
+	if (!strcmp(kobj->name, name))
+		return sprintf(buf, "%s\n", name_str);
+
+	return -EINVAL;
+}
+
+static MDEV_TYPE_ATTR_RO(name);
+
+static ssize_t
+available_instances_show(struct kobject *kobj, struct device *dev, char *buf)
+{
+	return sprintf(buf, "%d\n", INT_MAX);
+}
+
+static MDEV_TYPE_ATTR_RO(available_instances);
+
+static ssize_t device_api_show(struct kobject *kobj, struct device *dev,
+			       char *buf)
+{
+	return sprintf(buf, "%s\n", VIRTIO_MDEV_DEVICE_API_STRING);
+}
+
+static MDEV_TYPE_ATTR_RO(device_api);
+
+static struct attribute *mdev_types_attrs[] = {
+	&mdev_type_attr_name.attr,
+	&mdev_type_attr_device_api.attr,
+	&mdev_type_attr_available_instances.attr,
+	NULL,
+};
+
+static struct attribute_group mdev_type_group = {
+	.name  = "virtio",
+	.attrs = mdev_types_attrs,
+};
+
+/* TBD: "vhost" type */
+
+static struct attribute_group *mdev_type_groups[] = {
+	&mdev_type_group,
+	NULL,
+};
+
+static int mvnet_set_vq_address(struct mdev_device *mdev, u16 idx,
+				u64 desc_area, u64 driver_area, u64 device_area)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->desc_addr = desc_area;
+	vq->driver_addr = driver_area;
+	vq->device_addr = device_area;
+
+	return 0;
+}
+
+static void mvnet_set_vq_num(struct mdev_device *mdev, u16 idx, u32 num)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->num = num;
+}
+
+static void mvnet_kick_vq(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	if (vq->ready)
+		schedule_work(&mvnet->work);
+}
+
+static void mvnet_set_vq_cb(struct mdev_device *mdev, u16 idx,
+			    struct virtio_mdev_callback *cb)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	vq->cb = cb->callback;
+	vq->private = cb->private;
+}
+
+static void mvnet_set_vq_ready(struct mdev_device *mdev, u16 idx, bool ready)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	spin_lock(&mvnet->lock);
+	vq->ready = ready;
+	if (vq->ready)
+		mvnet_queue_ready(mvnet, idx);
+	spin_unlock(&mvnet->lock);
+}
+
+static bool mvnet_get_vq_ready(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+
+	return vq->ready;
+}
+
+static int mvnet_set_vq_state(struct mdev_device *mdev, u16 idx, u64 state)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	spin_lock(&mvnet->lock);
+	vrh->last_avail_idx = state;
+	spin_unlock(&mvnet->lock);
+
+	return 0;
+}
+
+static u64 mvnet_get_vq_state(struct mdev_device *mdev, u16 idx)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+	struct mvnet_virtqueue *vq = &mvnet->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	return vrh->last_avail_idx;
+}
+
+static u16 mvnet_get_vq_align(struct mdev_device *mdev)
+{
+	return MVNET_QUEUE_ALIGN;
+}
+
+static u64 mvnet_get_features(struct mdev_device *mdev)
+{
+	return mvnet_features;
+}
+
+static int mvnet_set_features(struct mdev_device *mdev, u64 features)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	/* DMA mapping must be done by driver */
+	if (!(features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)))
+		return -EINVAL;
+
+	mvnet->features = features & mvnet_features;
+
+	return 0;
+}
+
+static void mvnet_set_config_cb(struct mdev_device *mdev,
+				struct virtio_mdev_callback *cb)
+{
+	/* We don't support config interrupt */
+}
+
+static u16 mvnet_get_vq_num_max(struct mdev_device *mdev)
+{
+	return MVNET_QUEUE_MAX;
+}
+
+static u32 mvnet_get_device_id(struct mdev_device *mdev)
+{
+	return MVNET_DEVICE_ID;
+}
+
+static u32 mvnet_get_vendor_id(struct mdev_device *mdev)
+{
+	return MVNET_VENDOR_ID;
+}
+
+static u8 mvnet_get_status(struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	return mvnet->status;
+}
+
+static void mvnet_set_status(struct mdev_device *mdev, u8 status)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	mvnet->status = status;
+
+	if (status == 0) {
+		spin_lock(&mvnet->lock);
+		mvnet_reset(mvnet);
+		spin_unlock(&mvnet->lock);
+	}
+}
+
+static void mvnet_get_config(struct mdev_device *mdev, unsigned int offset,
+			     void *buf, unsigned int len)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	if (offset + len < sizeof(struct virtio_net_config))
+		memcpy(buf, &mvnet->config + offset, len);
+}
+
+static void mvnet_set_config(struct mdev_device *mdev, unsigned int offset,
+			     const void *buf, unsigned int len)
+{
+	/* No writable config supportted by mvnet */
+}
+
+static u32 mvnet_get_generation(struct mdev_device *mdev)
+{
+	struct mvnet_state *mvnet = mdev_get_drvdata(mdev);
+
+	return mvnet->generation;
+}
+
+static const struct mdev_virtio_device_ops mdev_virtio_ops = {
+	.set_vq_address         = mvnet_set_vq_address,
+	.set_vq_num             = mvnet_set_vq_num,
+	.kick_vq                = mvnet_kick_vq,
+	.set_vq_cb              = mvnet_set_vq_cb,
+	.set_vq_ready           = mvnet_set_vq_ready,
+	.get_vq_ready           = mvnet_get_vq_ready,
+	.set_vq_state           = mvnet_set_vq_state,
+	.get_vq_state           = mvnet_get_vq_state,
+	.get_vq_align           = mvnet_get_vq_align,
+	.get_features           = mvnet_get_features,
+	.set_features           = mvnet_set_features,
+	.set_config_cb          = mvnet_set_config_cb,
+	.get_vq_num_max         = mvnet_get_vq_num_max,
+	.get_device_id          = mvnet_get_device_id,
+	.get_vendor_id          = mvnet_get_vendor_id,
+	.get_status             = mvnet_get_status,
+	.set_status             = mvnet_set_status,
+	.get_config             = mvnet_get_config,
+	.set_config             = mvnet_set_config,
+	.get_generation         = mvnet_get_generation,
+};
+
+static const struct mdev_parent_ops mdev_fops = {
+	.owner                  = THIS_MODULE,
+	.dev_attr_groups        = mvnet_dev_groups,
+	.mdev_attr_groups       = mdev_dev_groups,
+	.supported_type_groups  = mdev_type_groups,
+	.create                 = mvnet_create,
+	.remove			= mvnet_remove,
+};
+
+static void mvnet_device_release(struct device *dev)
+{
+	dev_dbg(dev, "mvnet: released\n");
+}
+
+static int __init mvnet_dev_init(void)
+{
+	int ret = 0;
+
+	pr_info("mvnet_dev: %s\n", __func__);
+
+	memset(&mvnet_dev, 0, sizeof(mvnet_dev));
+
+	idr_init(&mvnet_dev.vd_idr);
+
+	mvnet_dev.vd_class = class_create(THIS_MODULE, MVNET_CLASS_NAME);
+
+	if (IS_ERR(mvnet_dev.vd_class)) {
+		pr_err("Error: failed to register mvnet_dev class\n");
+		ret = PTR_ERR(mvnet_dev.vd_class);
+		goto failed1;
+	}
+
+	mvnet_dev.dev.class = mvnet_dev.vd_class;
+	mvnet_dev.dev.release = mvnet_device_release;
+	dev_set_name(&mvnet_dev.dev, "%s", MVNET_NAME);
+
+	ret = device_register(&mvnet_dev.dev);
+	if (ret)
+		goto failed2;
+
+	ret = mdev_register_device(&mvnet_dev.dev, &mdev_fops);
+	if (ret)
+		goto failed3;
+
+	mutex_init(&mdev_list_lock);
+	INIT_LIST_HEAD(&mdev_devices_list);
+
+	goto all_done;
+
+failed3:
+
+	device_unregister(&mvnet_dev.dev);
+failed2:
+	class_destroy(mvnet_dev.vd_class);
+
+failed1:
+all_done:
+	return ret;
+}
+
+static void __exit mvnet_dev_exit(void)
+{
+	mvnet_dev.dev.bus = NULL;
+	mdev_unregister_device(&mvnet_dev.dev);
+
+	device_unregister(&mvnet_dev.dev);
+	idr_destroy(&mvnet_dev.vd_idr);
+	class_destroy(mvnet_dev.vd_class);
+	mvnet_dev.vd_class = NULL;
+	pr_info("mvnet_dev: Unloaded!\n");
+}
+
+module_init(mvnet_dev_init)
+module_exit(mvnet_dev_exit)
+
+MODULE_LICENSE("GPL v2");
+MODULE_INFO(supported, "Simulate loopback ethernet device over mdev");
+MODULE_VERSION(VERSION_STRING);
+MODULE_AUTHOR(DRIVER_AUTHOR);
Signature · 1 line
2.19.1
author
#36 lore raw

PATCH v10 6/6 docs: sample driver to demonstrate how to implement virtio-mdev framework

In reply to Jason Wang

This sample driver creates mdev device that simulate virtio net device over virtio mdev transport. The device is implemented through vringh and workqueue. A device specific dma ops is to make sure HVA is used directly as the IOVA. This should be sufficient for kernel virtio driver to work.

Only 'virtio' type is supported right now. I plan to add 'vhost' type on top which requires some virtual IOMMU implemented in this sample driver.

---
 MAINTAINERS                |   1 +
 samples/Kconfig            |  10 +
 samples/vfio-mdev/Makefile |   1 +
 samples/vfio-mdev/mvnet.c  | 686 +++++++++++++++++++++++++++++++++++++
 4 files changed, 698 insertions(+)
 create mode 100644 samples/vfio-mdev/mvnet.c
MAINTAINERS +1 -0
index 4997957443df..6e9ad105a28f 100644
@@ -17249,6 +17249,7 @@ F:	include/uapi/linux/virtio_*.h
 F:	drivers/crypto/virtio/
 F:	mm/balloon_compaction.c
 F:	include/linux/mdev_virtio_ops.h
+F:	samples/vfio-mdev/mvnet.c
 
 VIRTIO BLOCK AND SCSI DRIVERS
 M:	"Michael S. Tsirkin" <[email protected]>
samples/Kconfig +10 -0
index c8dacb4dda80..13a2443e18e0 100644
@@ -131,6 +131,16 @@ config SAMPLE_VFIO_MDEV_MDPY
 	  mediated device.  It is a simple framebuffer and supports
 	  the region display interface (VFIO_GFX_PLANE_TYPE_REGION).
 
+config SAMPLE_VIRTIO_MDEV_NET
+	tristate "Build VIRTIO net example mediated device sample code -- loadable modules only"
+	depends on VIRTIO_MDEV && VHOST_RING && m
+	help
+	  Build a networking sample device for use as a virtio
+	  mediated device. The device coopreates with virtio-mdev bus
+	  driver to present an virtio ethernet driver for
+	  kernel. It simply loopbacks all packets from its TX
+	  virtqueue to its RX virtqueue.
+
 config SAMPLE_VFIO_MDEV_MDPY_FB
 	tristate "Build VFIO mdpy example guest fbdev driver -- loadable module only"
 	depends on FB && m
samples/vfio-mdev/Makefile +1 -0
index 10d179c4fdeb..f34af90ed0a0 100644
@@ -3,3 +3,4 @@ obj-$(CONFIG_SAMPLE_VFIO_MDEV_MTTY) += mtty.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY) += mdpy.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY_FB) += mdpy-fb.o
 obj-$(CONFIG_SAMPLE_VFIO_MDEV_MBOCHS) += mbochs.o
+obj-$(CONFIG_SAMPLE_VIRTIO_MDEV_NET) += mvnet.o
samples/vfio-mdev/mvnet.c new file +686 -0

686 lines not on this page · raw message

Signature · 1 line
2.19.1
#37 lore raw
In reply to Jason Wang
Quoted 30 lines · On Wed, 6 Nov 2019 21:35:25 +0800 Jason Wang <[email protected]> wrote:

Hi all:

There are hardwares that can do virtio datapath offloading while having its own control path. This path tries to implement a mdev based unified API to support using kernel virtio driver to drive those devices. This is done by introducing a new mdev transport for virtio (virtio_mdev) and register itself as a new kind of mdev driver. Then it provides a unified way for kernel virtio driver to talk with mdev device implementation.

Though the series only contains kernel driver support, the goal is to make the transport generic enough to support userspace drivers. This means vhost-mdev[1] could be built on top as well by resuing the transport.

A sample driver is also implemented which simulate a virito-net loopback ethernet device on top of vringh + workqueue. This could be used as a reference implementation for real hardware driver.

Also a real IFC VF driver was also posted here[2] which is a good reference for vendors who is interested in their own virtio datapath offloading product.

Consider mdev framework only support VFIO device and driver right now, this series also extend it to support other types. This is done through introducing class id to the device and pairing it with id_talbe claimed by the driver. On top, this seris also decouple device specific ops out of the common ones for implementing class specific operations over mdev bus.

Quoted 3 lines

Pktgen test was done with virito-net + mvnet loop back device.

Please review.

All looking good to me now.

Quoted 3 lines

[1] https://lkml.org/lkml/2019/11/5/424
[2] https://lkml.org/lkml/2019/11/5/227

#38 lore raw
In reply to Jason Wang
Quoted 30 lines · On Wed, 6 Nov 2019 21:35:25 +0800 Jason Wang <[email protected]> wrote:

Hi all:

There are hardwares that can do virtio datapath offloading while having its own control path. This path tries to implement a mdev based unified API to support using kernel virtio driver to drive those devices. This is done by introducing a new mdev transport for virtio (virtio_mdev) and register itself as a new kind of mdev driver. Then it provides a unified way for kernel virtio driver to talk with mdev device implementation.

Though the series only contains kernel driver support, the goal is to make the transport generic enough to support userspace drivers. This means vhost-mdev[1] could be built on top as well by resuing the transport.

A sample driver is also implemented which simulate a virito-net loopback ethernet device on top of vringh + workqueue. This could be used as a reference implementation for real hardware driver.

Also a real IFC VF driver was also posted here[2] which is a good reference for vendors who is interested in their own virtio datapath offloading product.

Consider mdev framework only support VFIO device and driver right now, this series also extend it to support other types. This is done through introducing class id to the device and pairing it with id_talbe claimed by the driver. On top, this seris also decouple device specific ops out of the common ones for implementing class specific operations over mdev bus.

Quoted 3 lines

Pktgen test was done with virito-net + mvnet loop back device.

Please review.

All looking good to me now.

Quoted 3 lines

[1] https://lkml.org/lkml/2019/11/5/424
[2] https://lkml.org/lkml/2019/11/5/227

#39 lore raw

✗ Fi.CI.CHECKPATCH: warning for mdev based hardware virtio offloading support

In reply to Jason Wang
== Series Details ==

== Series Details ==

Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69063/
State : warning

== Summary ==

$ dim checkpatch origin/drm-tip
2e15c56da9d0 mdev: class id support
aa81f6ab1bbe modpost: add support for mdev class id
94e6d607158c mdev: introduce device specific ops
-:499: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#499: 
new file mode 100644

total: 0 errors, 1 warnings, 0 checks, 618 lines checked
b92fd4081624 mdev: introduce virtio device and its device ops
-:108: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #108:
new file mode 100644

total: 0 errors, 1 warnings, 0 checks, 219 lines checked
b59bcebe0b2b virtio: introduce a mdev based transport
-:53: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #53:
new file mode 100644

-:113: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned' #113: FILE: drivers/virtio/virtio_mdev.c:56:
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,

-:114: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned' #114: FILE: drivers/virtio/virtio_mdev.c:57:
+ void *buf, unsigned len)

-:122: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned' #122: FILE: drivers/virtio/virtio_mdev.c:65:
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,

-:123: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned' #123: FILE: drivers/virtio/virtio_mdev.c:66:
+ const void *buf, unsigned len)

-:303: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned' #303: FILE: drivers/virtio/virtio_mdev.c:246:
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,

total: 0 errors, 6 warnings, 0 checks, 429 lines checked
42f8bf2303b0 docs: sample driver to demonstrate how to implement virtio-mdev framework -:63: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #63:
new file mode 100644

total: 0 errors, 1 warnings, 0 checks, 713 lines checked

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#40 lore raw

✓ Fi.CI.BAT: success for mdev based hardware virtio offloading support

In reply to Jason Wang
== Series Details ==

== Series Details ==

Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69063/
State : success

== Summary ==

CI Bug Log - changes from CI_DRM_7270 -> Patchwork_15153
====================================================

Summary
-------

**SUCCESS**

No regressions found.

External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15153/index.html

Known issues
------------

Here are the changes found in Patchwork_15153 that come from known issues:

### IGT changes ###

#### Issues hit ####

* igt@kms_frontbuffer_tracking@basic:
- fi-icl-u2: [PASS][1] -> [FAIL][2] ([fdo#103167])
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7270/fi-icl-u2/igt@[email protected]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15153/fi-icl-u2/igt@[email protected]

{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).

[fdo#103167]: https://bugs.freedesktop.org/show_bug.cgi?id=103167
[fdo#107713]: https://bugs.freedesktop.org/show_bug.cgi?id=107713
[fdo#108569]: https://bugs.freedesktop.org/show_bug.cgi?id=108569
[fdo#108743]: https://bugs.freedesktop.org/show_bug.cgi?id=108743

Participating hosts (51 -> 44)
------------------------------

Missing (7): fi-ilk-m540 fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-ctg-p8600 fi-byt-clapper fi-bdw-samus

Build changes
-------------

* CI: CI-20190529 -> None
* Linux: CI_DRM_7270 -> Patchwork_15153

CI-20190529: 20190529
CI_DRM_7270: 8d83f3978cb2413ffceb2b8a9c1d507484d7ff08 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_5264: f21213012393bd8041ad93084ce29da088ef8426 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_15153: 42f8bf2303b0ea32b40b81645a89f3674dd44f12 @ git://anongit.freedesktop.org/gfx-ci/linux

== Linux commits ==

42f8bf2303b0 docs: sample driver to demonstrate how to implement virtio-mdev framework b59bcebe0b2b virtio: introduce a mdev based transport
b92fd4081624 mdev: introduce virtio device and its device ops 94e6d607158c mdev: introduce device specific ops
aa81f6ab1bbe modpost: add support for mdev class id
2e15c56da9d0 mdev: class id support

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15153/index.html
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer Acked-by Acked-by on the v10 series
#41 lore raw
In reply to Jason Wang
Quoted 23 lines · Jason Wang

Hi all:

There are hardwares that can do virtio datapath offloading while having its own control path. This path tries to implement a mdev based unified API to support using kernel virtio driver to drive those devices. This is done by introducing a new mdev transport for virtio (virtio_mdev) and register itself as a new kind of mdev driver. Then it provides a unified way for kernel virtio driver to talk with mdev device implementation.

Though the series only contains kernel driver support, the goal is to make the transport generic enough to support userspace drivers. This means vhost-mdev[1] could be built on top as well by resuing the transport.

A sample driver is also implemented which simulate a virito-net loopback ethernet device on top of vringh + workqueue. This could be used as a reference implementation for real hardware driver.

Also a real IFC VF driver was also posted here[2] which is a good reference for vendors who is interested in their own virtio datapath offloading product.

Quoted 8 lines

Consider mdev framework only support VFIO device and driver right now, this series also extend it to support other types. This is done through introducing class id to the device and pairing it with id_talbe claimed by the driver. On top, this seris also decouple device specific ops out of the common ones for implementing class specific operations over mdev bus.

Pktgen test was done with virito-net + mvnet loop back device.

Patches 1-5:

Patch 6: I'd like the module to be renamed to make it more obvious what it does. With that:

Quoted 134 lines

Please review.

[1] https://lkml.org/lkml/2019/11/5/424
[2] https://lkml.org/lkml/2019/11/5/227

Changes from V9:
- Tweak the help text for virito-mdev kconfig

Changes from V8:
- try silent checkpatch, some are still there becuase they were inherited from virtio_config_ops which needs to be resolved in an independent series
- tweak on the comment and doc
- remove VIRTIO_MDEV_F_VERSION_1 completely
- rename CONFIG_VIRTIO_MDEV_DEVICE to CONFIG_VIRTIO_MDEV

Changes from V7:
- drop {set|get}_mdev_features for virtio
- typo and comment style fixes

Changes from V6:
- rename ops files and compile guard

Changes from V5:
- use dev_warn() instead of WARN(1) when class id is not set
- validate id_table before trying to do matching between device and driver
- add wildcard for modpost script
- use unique name for id_table
- move get_mdev_features() to be the first member of virtio_device_ops and more comments for it
- typo fixes for the comments above virtio_mdev_ops

Changes from V4:
- keep mdev_set_class() for the device that doesn't use device ops
- use union for device ops pointer in mdev_device
- introduce class specific helper for getting is device ops
- use WARN_ON instead of BUG_ON in mdev_set_virtio_ops
- explain details of get_mdev_features() and get_vendor_id()
- distinguish the optional virito device ops from mandatory ones and make get_generation() optional
- rename vfio_mdev.h to vfio_mdev_ops.h, rename virito_mdev.h to virtio_mdev_ops.h
- don't abuse version fileds in virtio_mdev structure, use features instead
- fix warning during device remove
- style & docs tweaks and typo fixes

Changes from V3:
- document that class id (device ops) must be specified in create()
- add WARN() when trying to set class_id when it has already set
- add WARN() when class_id is not specified in create() and correctly return an error in this case
- correct the prototype of mdev_set_class() in the doc
- add documention of mdev_set_class()
- remove the unnecessary "class_id_fail" label when class id is not specified in create()
- convert id_table in vfio_mdev to const
- move mdev_set_class and its friends after mdev_uuid()
- suqash the patch of bus uevent into patch of introducing class id
- tweak the words in the docs per Cornelia suggestion
- tie class_id and device ops through class specific initialization routine like mdev_set_vfio_ops()
- typos fixes in the docs of virtio-mdev callbacks
- document the usage of virtqueues in struct virtio_mdev_device
- remove the useless vqs array in struct virtio_mdev_device
- rename MDEV_ID_XXX to MDEV_CLASS_ID_XXX

Changes from V2:
- fail when class_id is not specified
- drop the vringh patch
- match the doc to the code
- tweak the commit log
- move device_ops from parent to mdev device
- remove the unused MDEV_ID_VHOST

Changes from V1:
- move virtio_mdev.c to drivers/virtio
- store class_id in mdev_device instead of mdev_parent
- store device_ops in mdev_device instead of mdev_parent
- reorder the patch, vringh fix comes first
- really silent compiling warnings
- really switch to use u16 for class_id
- uevent and modpost support for mdev class_id
- vraious tweaks per comments from Parav

Changes from RFC-V2:
- silent compile warnings on some specific configuration
- use u16 instead u8 for class id
- reseve MDEV_ID_VHOST for future vhost-mdev work
- introduce "virtio" type for mvnet and make "vhost" type for future work
- add entries in MAINTAINER
- tweak and typos fixes in commit log

Changes from RFC-V1:
- rename device id to class id
- add docs for class id and device specific ops (device_ops)
- split device_ops into seperate headers
- drop the mdev_set_dma_ops()
- use device_ops to implement the transport API, then it's not a part of UAPI any more
- use GFP_ATOMIC in mvnet sample device and other tweaks
- set_vring_base/get_vring_base support for mvnet device

Jason Wang (6):
mdev: class id support
modpost: add support for mdev class id
mdev: introduce device specific ops
mdev: introduce virtio device and its device ops
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework

 .../driver-api/vfio-mediated-device.rst       |  38 +-
 MAINTAINERS                                   |   3 +
 drivers/gpu/drm/i915/gvt/kvmgt.c              |  17 +-
 drivers/s390/cio/vfio_ccw_ops.c               |  17 +-
 drivers/s390/crypto/vfio_ap_ops.c             |  13 +-
 drivers/vfio/mdev/mdev_core.c                 |  60 ++
 drivers/vfio/mdev/mdev_driver.c               |  25 +
 drivers/vfio/mdev/mdev_private.h              |   8 +
 drivers/vfio/mdev/vfio_mdev.c                 |  45 +-
 drivers/virtio/Kconfig                        |  13 +
 drivers/virtio/Makefile                       |   1 +
 drivers/virtio/virtio_mdev.c                  | 406 +++++++++++
 include/linux/mdev.h                          |  57 +-
 include/linux/mdev_vfio_ops.h                 |  52 ++
 include/linux/mdev_virtio_ops.h               | 147 ++++
 include/linux/mod_devicetable.h               |   8 +
 samples/Kconfig                               |  10 +
 samples/vfio-mdev/Makefile                    |   1 +
 samples/vfio-mdev/mbochs.c                    |  19 +-
 samples/vfio-mdev/mdpy.c                      |  19 +-
 samples/vfio-mdev/mtty.c                      |  17 +-
Quoted 11 lines
 samples/vfio-mdev/mvnet.c                     | 686 ++++++++++++++++++
 scripts/mod/devicetable-offsets.c             |   3 +
 scripts/mod/file2alias.c                      |  11 +
 24 files changed, 1585 insertions(+), 91 deletions(-)
 create mode 100644 drivers/virtio/virtio_mdev.c
 create mode 100644 include/linux/mdev_vfio_ops.h
 create mode 100644 include/linux/mdev_virtio_ops.h
 create mode 100644 samples/vfio-mdev/mvnet.c
Signature · 1 line
2.19.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

maintainer Acked-by Acked-by on the v10 series
#42 lore raw
In reply to Jason Wang
Quoted 23 lines · Jason Wang

Hi all:

There are hardwares that can do virtio datapath offloading while having its own control path. This path tries to implement a mdev based unified API to support using kernel virtio driver to drive those devices. This is done by introducing a new mdev transport for virtio (virtio_mdev) and register itself as a new kind of mdev driver. Then it provides a unified way for kernel virtio driver to talk with mdev device implementation.

Though the series only contains kernel driver support, the goal is to make the transport generic enough to support userspace drivers. This means vhost-mdev[1] could be built on top as well by resuing the transport.

A sample driver is also implemented which simulate a virito-net loopback ethernet device on top of vringh + workqueue. This could be used as a reference implementation for real hardware driver.

Also a real IFC VF driver was also posted here[2] which is a good reference for vendors who is interested in their own virtio datapath offloading product.

Quoted 8 lines

Consider mdev framework only support VFIO device and driver right now, this series also extend it to support other types. This is done through introducing class id to the device and pairing it with id_talbe claimed by the driver. On top, this seris also decouple device specific ops out of the common ones for implementing class specific operations over mdev bus.

Pktgen test was done with virito-net + mvnet loop back device.

Patches 1-5:

Patch 6: I'd like the module to be renamed to make it more obvious what it does. With that:

Quoted 145 lines

Please review.

[1] https://lkml.org/lkml/2019/11/5/424
[2] https://lkml.org/lkml/2019/11/5/227

Changes from V9:
- Tweak the help text for virito-mdev kconfig

Changes from V8:
- try silent checkpatch, some are still there becuase they were inherited from virtio_config_ops which needs to be resolved in an independent series
- tweak on the comment and doc
- remove VIRTIO_MDEV_F_VERSION_1 completely
- rename CONFIG_VIRTIO_MDEV_DEVICE to CONFIG_VIRTIO_MDEV

Changes from V7:
- drop {set|get}_mdev_features for virtio
- typo and comment style fixes

Changes from V6:
- rename ops files and compile guard

Changes from V5:
- use dev_warn() instead of WARN(1) when class id is not set
- validate id_table before trying to do matching between device and driver
- add wildcard for modpost script
- use unique name for id_table
- move get_mdev_features() to be the first member of virtio_device_ops and more comments for it
- typo fixes for the comments above virtio_mdev_ops

Changes from V4:
- keep mdev_set_class() for the device that doesn't use device ops
- use union for device ops pointer in mdev_device
- introduce class specific helper for getting is device ops
- use WARN_ON instead of BUG_ON in mdev_set_virtio_ops
- explain details of get_mdev_features() and get_vendor_id()
- distinguish the optional virito device ops from mandatory ones and make get_generation() optional
- rename vfio_mdev.h to vfio_mdev_ops.h, rename virito_mdev.h to virtio_mdev_ops.h
- don't abuse version fileds in virtio_mdev structure, use features instead
- fix warning during device remove
- style & docs tweaks and typo fixes

Changes from V3:
- document that class id (device ops) must be specified in create()
- add WARN() when trying to set class_id when it has already set
- add WARN() when class_id is not specified in create() and correctly return an error in this case
- correct the prototype of mdev_set_class() in the doc
- add documention of mdev_set_class()
- remove the unnecessary "class_id_fail" label when class id is not specified in create()
- convert id_table in vfio_mdev to const
- move mdev_set_class and its friends after mdev_uuid()
- suqash the patch of bus uevent into patch of introducing class id
- tweak the words in the docs per Cornelia suggestion
- tie class_id and device ops through class specific initialization routine like mdev_set_vfio_ops()
- typos fixes in the docs of virtio-mdev callbacks
- document the usage of virtqueues in struct virtio_mdev_device
- remove the useless vqs array in struct virtio_mdev_device
- rename MDEV_ID_XXX to MDEV_CLASS_ID_XXX

Changes from V2:
- fail when class_id is not specified
- drop the vringh patch
- match the doc to the code
- tweak the commit log
- move device_ops from parent to mdev device
- remove the unused MDEV_ID_VHOST

Changes from V1:
- move virtio_mdev.c to drivers/virtio
- store class_id in mdev_device instead of mdev_parent
- store device_ops in mdev_device instead of mdev_parent
- reorder the patch, vringh fix comes first
- really silent compiling warnings
- really switch to use u16 for class_id
- uevent and modpost support for mdev class_id
- vraious tweaks per comments from Parav

Changes from RFC-V2:
- silent compile warnings on some specific configuration
- use u16 instead u8 for class id
- reseve MDEV_ID_VHOST for future vhost-mdev work
- introduce "virtio" type for mvnet and make "vhost" type for future work
- add entries in MAINTAINER
- tweak and typos fixes in commit log

Changes from RFC-V1:
- rename device id to class id
- add docs for class id and device specific ops (device_ops)
- split device_ops into seperate headers
- drop the mdev_set_dma_ops()
- use device_ops to implement the transport API, then it's not a part of UAPI any more
- use GFP_ATOMIC in mvnet sample device and other tweaks
- set_vring_base/get_vring_base support for mvnet device

Jason Wang (6):
mdev: class id support
modpost: add support for mdev class id
mdev: introduce device specific ops
mdev: introduce virtio device and its device ops
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework

 .../driver-api/vfio-mediated-device.rst       |  38 +-
 MAINTAINERS                                   |   3 +
 drivers/gpu/drm/i915/gvt/kvmgt.c              |  17 +-
 drivers/s390/cio/vfio_ccw_ops.c               |  17 +-
 drivers/s390/crypto/vfio_ap_ops.c             |  13 +-
 drivers/vfio/mdev/mdev_core.c                 |  60 ++
 drivers/vfio/mdev/mdev_driver.c               |  25 +
 drivers/vfio/mdev/mdev_private.h              |   8 +
 drivers/vfio/mdev/vfio_mdev.c                 |  45 +-
 drivers/virtio/Kconfig                        |  13 +
 drivers/virtio/Makefile                       |   1 +
 drivers/virtio/virtio_mdev.c                  | 406 +++++++++++
 include/linux/mdev.h                          |  57 +-
 include/linux/mdev_vfio_ops.h                 |  52 ++
 include/linux/mdev_virtio_ops.h               | 147 ++++
 include/linux/mod_devicetable.h               |   8 +
 samples/Kconfig                               |  10 +
 samples/vfio-mdev/Makefile                    |   1 +
 samples/vfio-mdev/mbochs.c                    |  19 +-
 samples/vfio-mdev/mdpy.c                      |  19 +-
 samples/vfio-mdev/mtty.c                      |  17 +-
 samples/vfio-mdev/mvnet.c                     | 686 ++++++++++++++++++
 scripts/mod/devicetable-offsets.c             |   3 +
 scripts/mod/file2alias.c                      |  11 +
 24 files changed, 1585 insertions(+), 91 deletions(-)
 create mode 100644 drivers/virtio/virtio_mdev.c
 create mode 100644 include/linux/mdev_vfio_ops.h
 create mode 100644 include/linux/mdev_virtio_ops.h
 create mode 100644 samples/vfio-mdev/mvnet.c
Signature · 1 line
2.19.1