* [PATCH V13 0/6] mdev based hardware virtio offloading support
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, 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.
Consider mdev framework only support VFIO device and driver right now,
this series also extend it to support other types. This is done
through decoupling VFIO specific bits out of mdev core and make mdev
an independent module that allows to be used by multiple types of
buses.
Pktgen test was done with virito-net + mvnet loop back device.
Please review.
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
Changes from V12:
- rebase on top of gvt-linux.git
Changes from V11:
- decouple VFIO specific bits out of mdev core
- make mdev an indepdent module to allow buses other than VFIO mdev
- allow structure composition of mdev through specifiy the size of
mdev structure
- introduce mdev_vfio structure and store the VFIO specific callbacks
there
- don't use "mdev" bus for virtio, use a new "mdev_virtio" bus, and
store the virtio specific callbacks in mdev_virtio structure.
- do the class_id, matching on top of "mdev_virtio" bus
Changes from V10:
- rename mvnet to mvnet_loopback
- fix typo in the help text for sample Kconfig
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: make mdev bus agnostic
mdev: split out VFIO bus specific parent ops
mdev: move to drivers/
mdev: introduce mediated virtio bus
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework
.../driver-api/vfio-mediated-device.rst | 94 ++-
MAINTAINERS | 11 +-
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 24 +-
drivers/mdev/Kconfig | 29 +
drivers/mdev/Makefile | 7 +
drivers/{vfio => }/mdev/mdev_core.c | 107 ++-
drivers/{vfio => }/mdev/mdev_driver.c | 29 +-
drivers/{vfio => }/mdev/mdev_private.h | 18 +-
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
drivers/mdev/vfio.c | 76 ++
drivers/mdev/virtio.c | 126 ++++
drivers/s390/cio/vfio_ccw_ops.c | 23 +-
drivers/s390/crypto/vfio_ap_ops.c | 34 +-
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 11 -
drivers/vfio/mdev/Makefile | 4 -
drivers/vfio/mdev/vfio_mdev.c | 45 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
drivers/virtio/Kconfig | 13 +
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++
include/linux/mdev.h | 49 +-
include/linux/mdev_vfio.h | 68 ++
include/linux/mdev_virtio.h | 168 +++++
include/linux/mod_devicetable.h | 8 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mbochs.c | 26 +-
samples/vfio-mdev/mdpy.c | 27 +-
samples/vfio-mdev/mtty.c | 24 +-
samples/vfio-mdev/mvnet_loopback.c | 690 ++++++++++++++++++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +
35 files changed, 1923 insertions(+), 235 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (76%)
rename drivers/{vfio => }/mdev/mdev_driver.c (83%)
rename drivers/{vfio => }/mdev/mdev_private.h (81%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
create mode 100644 drivers/mdev/vfio.c
create mode 100644 drivers/mdev/virtio.c
create mode 100644 drivers/virtio/virtio_mdev.c
create mode 100644 include/linux/mdev_vfio.h
create mode 100644 include/linux/mdev_virtio.h
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
--
2.19.1
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH V13 0/6] mdev based hardware virtio offloading support
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
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 decoupling VFIO specific bits out of mdev core and make mdev
an independent module that allows to be used by multiple types of
buses.
Pktgen test was done with virito-net + mvnet loop back device.
Please review.
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
Changes from V12:
- rebase on top of gvt-linux.git
Changes from V11:
- decouple VFIO specific bits out of mdev core
- make mdev an indepdent module to allow buses other than VFIO mdev
- allow structure composition of mdev through specifiy the size of
mdev structure
- introduce mdev_vfio structure and store the VFIO specific callbacks
there
- don't use "mdev" bus for virtio, use a new "mdev_virtio" bus, and
store the virtio specific callbacks in mdev_virtio structure.
- do the class_id, matching on top of "mdev_virtio" bus
Changes from V10:
- rename mvnet to mvnet_loopback
- fix typo in the help text for sample Kconfig
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: make mdev bus agnostic
mdev: split out VFIO bus specific parent ops
mdev: move to drivers/
mdev: introduce mediated virtio bus
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework
.../driver-api/vfio-mediated-device.rst | 94 ++-
MAINTAINERS | 11 +-
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 24 +-
drivers/mdev/Kconfig | 29 +
drivers/mdev/Makefile | 7 +
drivers/{vfio => }/mdev/mdev_core.c | 107 ++-
drivers/{vfio => }/mdev/mdev_driver.c | 29 +-
drivers/{vfio => }/mdev/mdev_private.h | 18 +-
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
drivers/mdev/vfio.c | 76 ++
drivers/mdev/virtio.c | 126 ++++
drivers/s390/cio/vfio_ccw_ops.c | 23 +-
drivers/s390/crypto/vfio_ap_ops.c | 34 +-
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 11 -
drivers/vfio/mdev/Makefile | 4 -
drivers/vfio/mdev/vfio_mdev.c | 45 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
drivers/virtio/Kconfig | 13 +
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++
include/linux/mdev.h | 49 +-
include/linux/mdev_vfio.h | 68 ++
include/linux/mdev_virtio.h | 168 +++++
include/linux/mod_devicetable.h | 8 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mbochs.c | 26 +-
samples/vfio-mdev/mdpy.c | 27 +-
samples/vfio-mdev/mtty.c | 24 +-
samples/vfio-mdev/mvnet_loopback.c | 690 ++++++++++++++++++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +
35 files changed, 1923 insertions(+), 235 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (76%)
rename drivers/{vfio => }/mdev/mdev_driver.c (83%)
rename drivers/{vfio => }/mdev/mdev_private.h (81%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
create mode 100644 drivers/mdev/vfio.c
create mode 100644 drivers/mdev/virtio.c
create mode 100644 drivers/virtio/virtio_mdev.c
create mode 100644 include/linux/mdev_vfio.h
create mode 100644 include/linux/mdev_virtio.h
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 0/6] mdev based hardware virtio offloading support
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
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 decoupling VFIO specific bits out of mdev core and make mdev
an independent module that allows to be used by multiple types of
buses.
Pktgen test was done with virito-net + mvnet loop back device.
Please review.
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
Changes from V12:
- rebase on top of gvt-linux.git
Changes from V11:
- decouple VFIO specific bits out of mdev core
- make mdev an indepdent module to allow buses other than VFIO mdev
- allow structure composition of mdev through specifiy the size of
mdev structure
- introduce mdev_vfio structure and store the VFIO specific callbacks
there
- don't use "mdev" bus for virtio, use a new "mdev_virtio" bus, and
store the virtio specific callbacks in mdev_virtio structure.
- do the class_id, matching on top of "mdev_virtio" bus
Changes from V10:
- rename mvnet to mvnet_loopback
- fix typo in the help text for sample Kconfig
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: make mdev bus agnostic
mdev: split out VFIO bus specific parent ops
mdev: move to drivers/
mdev: introduce mediated virtio bus
virtio: introduce a mdev based transport
docs: sample driver to demonstrate how to implement virtio-mdev
framework
.../driver-api/vfio-mediated-device.rst | 94 ++-
MAINTAINERS | 11 +-
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 24 +-
drivers/mdev/Kconfig | 29 +
drivers/mdev/Makefile | 7 +
drivers/{vfio => }/mdev/mdev_core.c | 107 ++-
drivers/{vfio => }/mdev/mdev_driver.c | 29 +-
drivers/{vfio => }/mdev/mdev_private.h | 18 +-
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
drivers/mdev/vfio.c | 76 ++
drivers/mdev/virtio.c | 126 ++++
drivers/s390/cio/vfio_ccw_ops.c | 23 +-
drivers/s390/crypto/vfio_ap_ops.c | 34 +-
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 11 -
drivers/vfio/mdev/Makefile | 4 -
drivers/vfio/mdev/vfio_mdev.c | 45 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
drivers/virtio/Kconfig | 13 +
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++
include/linux/mdev.h | 49 +-
include/linux/mdev_vfio.h | 68 ++
include/linux/mdev_virtio.h | 168 +++++
include/linux/mod_devicetable.h | 8 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mbochs.c | 26 +-
samples/vfio-mdev/mdpy.c | 27 +-
samples/vfio-mdev/mtty.c | 24 +-
samples/vfio-mdev/mvnet_loopback.c | 690 ++++++++++++++++++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +
35 files changed, 1923 insertions(+), 235 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (76%)
rename drivers/{vfio => }/mdev/mdev_driver.c (83%)
rename drivers/{vfio => }/mdev/mdev_private.h (81%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
create mode 100644 drivers/mdev/vfio.c
create mode 100644 drivers/mdev/virtio.c
create mode 100644 drivers/virtio/virtio_mdev.c
create mode 100644 include/linux/mdev_vfio.h
create mode 100644 include/linux/mdev_virtio.h
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH V13 1/6] mdev: make mdev bus agnostic
2019-11-18 10:59 ` Jason Wang
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, Jason Wang
Current mdev is tied to a VFIO specific "mdev" bus. This prevent mdev
from being used by other types of API/buses. So this patch tries to make
mdev bus agnostic through making a mdev core a thin module:
- decouple VFIO bus specific bits from mdev_core.c to mdev_vfio.c and
introduce mdev_vfio module
- require to specify the type of bus when registering mdev device and
mdev driver
With those modifications mdev become a generic module that could be
used by multiple types of virtual buses and devices.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 68 ++++++------
MAINTAINERS | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 8 +-
drivers/s390/cio/vfio_ccw_ops.c | 6 +-
drivers/s390/crypto/vfio_ap_ops.c | 21 ++--
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 17 ++-
drivers/vfio/mdev/Makefile | 4 +-
drivers/vfio/mdev/mdev_core.c | 104 +++++++++++++-----
drivers/vfio/mdev/mdev_driver.c | 29 ++---
drivers/vfio/mdev/mdev_private.h | 13 ++-
drivers/vfio/mdev/mdev_vfio.c | 48 ++++++++
drivers/vfio/mdev/vfio_mdev.c | 5 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
include/linux/mdev.h | 16 ++-
include/linux/mdev_vfio.h | 25 +++++
samples/vfio-mdev/mbochs.c | 8 +-
samples/vfio-mdev/mdpy.c | 8 +-
samples/vfio-mdev/mtty.c | 8 +-
19 files changed, 269 insertions(+), 128 deletions(-)
create mode 100644 drivers/vfio/mdev/mdev_vfio.c
create mode 100644 include/linux/mdev_vfio.h
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 25eb7d5b834b..1887d27a565e 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -49,35 +49,37 @@ devices as examples, as these devices are the first devices to use this module::
+---------------+
| |
- | +-----------+ | mdev_register_driver() +--------------+
- | | | +<------------------------+ |
- | | mdev | | | |
- | | bus | +------------------------>+ vfio_mdev.ko |<-> VFIO user
- | | driver | | probe()/remove() | | APIs
- | | | | +--------------+
- | +-----------+ |
+ | MDEV CORE | mdev_register_driver() +--------------+
+ | MODULE +<------------------------+ |
+ | mdev.ko | | |
+ | +------------------------>+ vfio_mdev.ko |<-> VFIO user
+ | | probe()/remove() | | APIs
+ | | +--------------+
+ +---+-------+---+
+ | /|\
+ | |
+callbacks| | mdev_register_device()
+ | | mdev_register_bus()
+ \|/ |
+ +---+-------+---+
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | nvidia.ko |<-> physical
+ | +----------------------------->+ | device
+ | MDEV VFIO | callbacks +--------------+
+ | Physical |
+ | device | mdev_vfio_register_device() +--------------+
+ | interface |<-----------------------------+ |
+ | | | i915.ko |<-> physical
+ | mdev_vfio.ko +----------------------------->+ | device
+ | | callbacks +--------------+
+ | |
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | ccw_device.ko|<-> physical
+ | +----------------------------->+ | device
+ | | callbacks +--------------+
| |
- | MDEV CORE |
- | MODULE |
- | mdev.ko |
- | +-----------+ | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | nvidia.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | Physical | |
- | | device | | mdev_register_device() +--------------+
- | | interface | |<------------------------+ |
- | | | | | i915.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | | |
- | | | | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | ccw_device.ko|<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | +-----------+ |
+---------------+
@@ -116,7 +118,8 @@ to register and unregister itself with the core driver:
* Register::
extern int mdev_register_driver(struct mdev_driver *drv,
- struct module *owner);
+ struct module *owner,
+ struct bus_type *bus);
* Unregister::
@@ -159,11 +162,12 @@ The callbacks in the mdev_parent_ops structure are as follows:
* 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::
+A driver should use the mdev_parent_ops structure and bus type in the
+function call to register itself with the mdev core driver::
extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops);
+ const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
However, the mdev_parent_ops structure is not required in the function call
that a driver should use to unregister itself with the mdev core driver::
diff --git a/MAINTAINERS b/MAINTAINERS
index 34ef0cf30ece..6d590afb62c3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17139,6 +17139,7 @@ S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
F: drivers/vfio/mdev/
F: include/linux/mdev.h
+F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
VFIO PLATFORM DRIVER
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 04a5a0d90823..afdb3de5ce2f 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -41,7 +41,7 @@
#include <linux/uuid.h>
#include <linux/kvm_host.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/debugfs.h>
#include <linux/nospec.h>
@@ -1554,7 +1554,7 @@ static ssize_t
vgpu_id_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
if (mdev) {
struct intel_vgpu *vgpu = (struct intel_vgpu *)
@@ -1606,12 +1606,12 @@ static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
return -EFAULT;
intel_vgpu_ops.supported_type_groups = kvm_vgpu_type_groups;
- return mdev_register_device(dev, &intel_vgpu_ops);
+ return mdev_vfio_register_device(dev, &intel_vgpu_ops);
}
static void kvmgt_host_exit(struct device *dev)
{
- mdev_unregister_device(dev);
+ mdev_vfio_unregister_device(dev);
}
static int kvmgt_page_track_add(unsigned long handle, u64 gfn)
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index f0d71ab77c50..791b8b0eb027 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -11,7 +11,7 @@
*/
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/nospec.h>
#include <linux/slab.h>
@@ -588,10 +588,10 @@ static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
int vfio_ccw_mdev_reg(struct subchannel *sch)
{
- return mdev_register_device(&sch->dev, &vfio_ccw_mdev_ops);
+ return mdev_vfio_register_device(&sch->dev, &vfio_ccw_mdev_ops);
}
void vfio_ccw_mdev_unreg(struct subchannel *sch)
{
- mdev_unregister_device(&sch->dev);
+ mdev_vfio_unregister_device(&sch->dev);
}
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 5c0f53c6dde7..78048e670374 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -602,7 +602,7 @@ static ssize_t assign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of adapter */
@@ -668,7 +668,7 @@ static ssize_t unassign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of adapter */
@@ -748,7 +748,7 @@ static ssize_t assign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_apqi = matrix_mdev->matrix.aqm_max;
@@ -810,7 +810,7 @@ static ssize_t unassign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of domain */
@@ -854,7 +854,7 @@ static ssize_t assign_control_domain_store(struct device *dev,
{
int ret;
unsigned long id;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of control domain */
@@ -903,7 +903,7 @@ static ssize_t unassign_control_domain_store(struct device *dev,
{
int ret;
unsigned long domid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -933,7 +933,7 @@ static ssize_t control_domains_show(struct device *dev,
int nchars = 0;
int n;
char *bufpos = buf;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -952,7 +952,7 @@ static DEVICE_ATTR_RO(control_domains);
static ssize_t matrix_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
char *bufpos = buf;
unsigned long apid;
@@ -1295,10 +1295,11 @@ int vfio_ap_mdev_register(void)
{
atomic_set(&matrix_dev->available_instances, MAX_ZDEV_ENTRIES_EXT);
- return mdev_register_device(&matrix_dev->device, &vfio_ap_matrix_ops);
+ return mdev_vfio_register_device(&matrix_dev->device,
+ &vfio_ap_matrix_ops);
}
void vfio_ap_mdev_unregister(void)
{
- mdev_unregister_device(&matrix_dev->device);
+ mdev_vfio_unregister_device(&matrix_dev->device);
}
diff --git a/drivers/s390/crypto/vfio_ap_private.h b/drivers/s390/crypto/vfio_ap_private.h
index f46dde56b464..4e37e0e3433a 100644
--- a/drivers/s390/crypto/vfio_ap_private.h
+++ b/drivers/s390/crypto/vfio_ap_private.h
@@ -14,7 +14,7 @@
#include <linux/types.h>
#include <linux/device.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/delay.h>
#include <linux/mutex.h>
#include <linux/kvm_host.h>
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 5da27f2100f9..2e07ca915a96 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,15 +1,24 @@
-# SPDX-License-Identifier: GPL-2.0-only
-config VFIO_MDEV
+config MDEV
tristate "Mediated device driver framework"
- depends on VFIO
default n
help
Provides a framework to virtualize devices.
- See Documentation/driver-api/vfio-mediated-device.rst for more details.
If you don't know what do here, say N.
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
+
+
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index 101516fdf375..e9675501271a 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,6 @@
-# SPDX-License-Identifier: GPL-2.0-only
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-obj-$(CONFIG_VFIO_MDEV) += mdev.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index b558d4cfd082..e1272a40c521 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -22,11 +22,13 @@
static LIST_HEAD(parent_list);
static DEFINE_MUTEX(parent_list_lock);
-static struct class_compat *mdev_bus_compat_class;
static LIST_HEAD(mdev_list);
static DEFINE_MUTEX(mdev_list_lock);
+static LIST_HEAD(class_compat_list);
+static DEFINE_MUTEX(compat_list_lock);
+
struct device *mdev_parent_dev(struct mdev_device *mdev)
{
return mdev->parent->dev;
@@ -51,9 +53,9 @@ struct device *mdev_dev(struct mdev_device *mdev)
}
EXPORT_SYMBOL(mdev_dev);
-struct mdev_device *mdev_from_dev(struct device *dev)
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus)
{
- return dev_is_mdev(dev) ? to_mdev_device(dev) : NULL;
+ return dev_is_mdev(dev, bus) ? to_mdev_device(dev) : NULL;
}
EXPORT_SYMBOL(mdev_from_dev);
@@ -122,7 +124,9 @@ static void mdev_device_remove_common(struct mdev_device *mdev)
static int mdev_device_remove_cb(struct device *dev, void *data)
{
- if (dev_is_mdev(dev)) {
+ struct bus_type *bus = data;
+
+ if (dev_is_mdev(dev, bus)) {
struct mdev_device *mdev;
mdev = to_mdev_device(dev);
@@ -131,6 +135,41 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
return 0;
}
+static struct mdev_class_compat *get_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat;
+
+ list_for_each_entry(mdev_class_compat, &class_compat_list, next) {
+ if (mdev_class_compat->bus == bus)
+ return mdev_class_compat;
+ }
+
+ return NULL;
+}
+
+static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat = get_class_compat(bus);
+ char class_name[64];
+
+ if (mdev_class_compat)
+ return mdev_class_compat->class_compat;
+
+ mdev_class_compat = kmalloc(sizeof(*mdev_class_compat), GFP_KERNEL);
+ if (!mdev_class_compat)
+ return NULL;
+ snprintf(class_name, 64, "%s_bus", bus->name);
+ mdev_class_compat->class_compat = class_compat_register(class_name);
+ if (!mdev_class_compat->class_compat) {
+ kfree(mdev_class_compat);
+ return NULL;
+ }
+ mdev_class_compat->bus = bus;
+ list_add(&mdev_class_compat->next, &class_compat_list);
+
+ return mdev_class_compat->class_compat;
+}
+
/*
* mdev_register_device : Register a device
* @dev: device structure representing parent device.
@@ -139,12 +178,14 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
* Add device to list of registered parent devices.
* Returns a negative value on error, otherwise 0.
*/
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus)
{
int ret;
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=registered";
char *envp[] = { env_string, NULL };
+ struct class_compat *class_compat;
/* check for mandatory ops */
if (!ops || !ops->create || !ops->remove || !ops->supported_type_groups)
@@ -175,20 +216,21 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
parent->dev = dev;
parent->ops = ops;
+ parent->bus = bus;
- if (!mdev_bus_compat_class) {
- mdev_bus_compat_class = class_compat_register("mdev_bus");
- if (!mdev_bus_compat_class) {
- ret = -ENOMEM;
- goto add_dev_err;
- }
+ mutex_lock(&compat_list_lock);
+ class_compat = mdev_alloc_class_compat(bus);
+ mutex_unlock(&compat_list_lock);
+ if (!class_compat) {
+ ret = -ENOMEM;
+ goto add_dev_err;
}
ret = parent_create_sysfs_files(parent);
if (ret)
goto add_dev_err;
- ret = class_compat_create_link(mdev_bus_compat_class, dev, NULL);
+ ret = class_compat_create_link(class_compat, dev, NULL);
if (ret)
dev_warn(dev, "Failed to create compatibility class link\n");
@@ -223,6 +265,7 @@ void mdev_unregister_device(struct device *dev)
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=unregistered";
char *envp[] = { env_string, NULL };
+ struct mdev_class_compat *mdev_class_compat;
mutex_lock(&parent_list_lock);
parent = __find_parent_device(dev);
@@ -238,9 +281,13 @@ void mdev_unregister_device(struct device *dev)
down_write(&parent->unreg_sem);
- class_compat_remove_link(mdev_bus_compat_class, dev, NULL);
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(parent->bus);
+ WARN_ON(!mdev_class_compat);
+ class_compat_remove_link(mdev_class_compat->class_compat, dev, NULL);
+ mutex_unlock(&compat_list_lock);
- device_for_each_child(dev, NULL, mdev_device_remove_cb);
+ device_for_each_child(dev, parent->bus, mdev_device_remove_cb);
parent_remove_sysfs_files(parent);
up_write(&parent->unreg_sem);
@@ -314,7 +361,7 @@ int mdev_device_create(struct kobject *kobj,
device_initialize(&mdev->dev);
mdev->dev.parent = dev;
- mdev->dev.bus = &mdev_bus_type;
+ mdev->dev.bus = parent->bus;
mdev->dev.release = mdev_device_release;
dev_set_name(&mdev->dev, "%pUl", uuid);
mdev->dev.groups = parent->ops->mdev_attr_groups;
@@ -404,24 +451,29 @@ struct device *mdev_get_iommu_device(struct device *dev)
}
EXPORT_SYMBOL(mdev_get_iommu_device);
-static int __init mdev_init(void)
+int mdev_register_bus(struct bus_type *bus)
{
- return mdev_bus_register();
+ return bus_register(bus);
}
+EXPORT_SYMBOL(mdev_register_bus);
-static void __exit mdev_exit(void)
+void mdev_unregister_bus(struct bus_type *bus)
{
- if (mdev_bus_compat_class)
- class_compat_unregister(mdev_bus_compat_class);
-
- mdev_bus_unregister();
+ struct mdev_class_compat *mdev_class_compat;
+
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(bus);
+ if (mdev_class_compat) {
+ list_del(&mdev_class_compat->next);
+ class_compat_unregister(mdev_class_compat->class_compat);
+ kfree(mdev_class_compat);
+ }
+ bus_unregister(bus);
+ mutex_unlock(&compat_list_lock);
}
-
-module_init(mdev_init)
-module_exit(mdev_exit)
+EXPORT_SYMBOL(mdev_unregister_bus);
MODULE_VERSION(DRIVER_VERSION);
MODULE_LICENSE("GPL v2");
MODULE_AUTHOR(DRIVER_AUTHOR);
MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/vfio/mdev/mdev_driver.c
index 0d3223aee20b..c3a2ac023712 100644
--- a/drivers/vfio/mdev/mdev_driver.c
+++ b/drivers/vfio/mdev/mdev_driver.c
@@ -10,6 +10,7 @@
#include <linux/device.h>
#include <linux/iommu.h>
#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -37,7 +38,7 @@ static void mdev_detach_iommu(struct mdev_device *mdev)
dev_info(&mdev->dev, "MDEV: detaching iommu\n");
}
-static int mdev_probe(struct device *dev)
+int mdev_probe(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -55,8 +56,9 @@ static int mdev_probe(struct device *dev)
return ret;
}
+EXPORT_SYMBOL(mdev_probe);
-static int mdev_remove(struct device *dev)
+int mdev_remove(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -68,26 +70,22 @@ static int mdev_remove(struct device *dev)
return 0;
}
-
-struct bus_type mdev_bus_type = {
- .name = "mdev",
- .probe = mdev_probe,
- .remove = mdev_remove,
-};
-EXPORT_SYMBOL_GPL(mdev_bus_type);
+EXPORT_SYMBOL(mdev_remove);
/**
* mdev_register_driver - register a new MDEV driver
* @drv: the driver to register
* @owner: module owner of driver to be registered
+ * @bus: but that the driver wants to attach
*
* Returns a negative value on error, otherwise 0.
**/
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner)
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus)
{
/* initialize common driver fields */
drv->driver.name = drv->name;
- drv->driver.bus = &mdev_bus_type;
+ drv->driver.bus = bus;
drv->driver.owner = owner;
/* register with core */
@@ -105,12 +103,3 @@ void mdev_unregister_driver(struct mdev_driver *drv)
}
EXPORT_SYMBOL(mdev_unregister_driver);
-int mdev_bus_register(void)
-{
- return bus_register(&mdev_bus_type);
-}
-
-void mdev_bus_unregister(void)
-{
- bus_unregister(&mdev_bus_type);
-}
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 7d922950caaf..298d7a0f493a 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -10,12 +10,10 @@
#ifndef MDEV_PRIVATE_H
#define MDEV_PRIVATE_H
-int mdev_bus_register(void);
-void mdev_bus_unregister(void);
-
struct mdev_parent {
struct device *dev;
const struct mdev_parent_ops *ops;
+ struct bus_type *bus;
struct kref ref;
struct list_head next;
struct kset *mdev_types_kset;
@@ -35,8 +33,15 @@ struct mdev_device {
bool active;
};
+struct mdev_class_compat {
+ struct class_compat *class_compat;
+ struct bus_type *bus;
+ struct list_head next;
+};
+
+
#define to_mdev_device(dev) container_of(dev, struct mdev_device, dev)
-#define dev_is_mdev(d) ((d)->bus == &mdev_bus_type)
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
struct mdev_type {
struct kobject kobj;
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
new file mode 100644
index 000000000000..f9d1191b9982
--- /dev/null
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -0,0 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0-only
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev_vfio.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VFIO bus"
+
+struct bus_type mdev_vfio_bus_type = {
+ .name = "mdev",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+};
+EXPORT_SYMBOL(mdev_vfio_bus_type);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_vfio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_vfio_bus_type);
+}
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+}
+EXPORT_SYMBOL(mdev_vfio_register_device);
+
+void mdev_vfio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_vfio_unregister_device);
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 30964a4e0a28..16e9ebe30d4a 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -13,7 +13,7 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -128,7 +128,8 @@ static struct mdev_driver vfio_mdev_driver = {
static int __init vfio_mdev_init(void)
{
- return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE);
+ return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE,
+ &mdev_vfio_bus_type);
}
static void __exit vfio_mdev_exit(void)
diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index d864277ea16f..f35523f822eb 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -34,7 +34,7 @@
#include <linux/uaccess.h>
#include <linux/vfio.h>
#include <linux/workqueue.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/notifier.h>
#include <linux/dma-iommu.h>
#include <linux/irqdomain.h>
@@ -1405,10 +1405,10 @@ static bool vfio_bus_is_mdev(struct bus_type *bus)
struct bus_type *mdev_bus;
bool ret = false;
- mdev_bus = symbol_get(mdev_bus_type);
+ mdev_bus = symbol_get(mdev_vfio_bus_type);
if (mdev_bus) {
ret = (bus == mdev_bus);
- symbol_put(mdev_bus_type);
+ symbol_put(mdev_vfio_bus_type);
}
return ret;
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index 0ce30ca78db0..ee2410246b3c 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -133,16 +133,22 @@ 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);
-extern struct bus_type mdev_bus_type;
-
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
void mdev_unregister_device(struct device *dev);
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner);
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
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);
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus);
+
+int mdev_probe(struct device *dev);
+int mdev_remove(struct device *dev);
+
+int mdev_register_bus(struct bus_type *bus);
+void mdev_unregister_bus(struct bus_type *bus);
#endif /* MDEV_H */
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
new file mode 100644
index 000000000000..446a7537e3fb
--- /dev/null
+++ b/include/linux/mdev_vfio.h
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VFIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef MDEV_VFIO_H
+#define MDEV_VFIO_H
+
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_vfio_bus_type;
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_vfio_unregister_device(struct device *dev);
+
+static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_vfio_bus_type);
+}
+
+#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index ac5c8c17b1ff..f041d58324b1 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -29,7 +29,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/dma-buf.h>
#include <linux/highmem.h>
@@ -1332,7 +1332,7 @@ static ssize_t
memory_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%d MB\n", mdev_state->type->mbytes);
@@ -1468,7 +1468,7 @@ static int __init mbochs_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mbochs_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mbochs_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1487,7 +1487,7 @@ static int __init mbochs_dev_init(void)
static void __exit mbochs_dev_exit(void)
{
mbochs_dev.bus = NULL;
- mdev_unregister_device(&mbochs_dev);
+ mdev_vfio_unregister_device(&mbochs_dev);
device_unregister(&mbochs_dev);
cdev_del(&mbochs_cdev);
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index cc86bf6566e4..9c32fe3795ad 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -25,7 +25,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <drm/drm_fourcc.h>
#include "mdpy-defs.h"
@@ -639,7 +639,7 @@ static ssize_t
resolution_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%dx%d\n",
@@ -775,7 +775,7 @@ static int __init mdpy_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mdpy_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mdpy_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -794,7 +794,7 @@ static int __init mdpy_dev_init(void)
static void __exit mdpy_dev_exit(void)
{
mdpy_dev.bus = NULL;
- mdev_unregister_device(&mdpy_dev);
+ mdev_vfio_unregister_device(&mdpy_dev);
device_unregister(&mdpy_dev);
cdev_del(&mdpy_cdev);
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index ce84a300a4da..6e4e6339e0f1 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -26,7 +26,7 @@
#include <linux/sysfs.h>
#include <linux/ctype.h>
#include <linux/file.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/serial.h>
#include <uapi/linux/serial_reg.h>
@@ -1285,7 +1285,7 @@ static ssize_t
sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- if (mdev_from_dev(dev))
+ if (vfio_mdev_from_dev(dev))
return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
return sprintf(buf, "\n");
@@ -1445,7 +1445,7 @@ static int __init mtty_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mtty_dev.dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mtty_dev.dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1471,7 +1471,7 @@ static int __init mtty_dev_init(void)
static void __exit mtty_dev_exit(void)
{
mtty_dev.dev.bus = NULL;
- mdev_unregister_device(&mtty_dev.dev);
+ mdev_vfio_unregister_device(&mtty_dev.dev);
device_unregister(&mtty_dev.dev);
idr_destroy(&mtty_dev.vd_idr);
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 1/6] mdev: make mdev bus agnostic
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
Current mdev is tied to a VFIO specific "mdev" bus. This prevent mdev
from being used by other types of API/buses. So this patch tries to make
mdev bus agnostic through making a mdev core a thin module:
- decouple VFIO bus specific bits from mdev_core.c to mdev_vfio.c and
introduce mdev_vfio module
- require to specify the type of bus when registering mdev device and
mdev driver
With those modifications mdev become a generic module that could be
used by multiple types of virtual buses and devices.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 68 ++++++------
MAINTAINERS | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 8 +-
drivers/s390/cio/vfio_ccw_ops.c | 6 +-
drivers/s390/crypto/vfio_ap_ops.c | 21 ++--
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 17 ++-
drivers/vfio/mdev/Makefile | 4 +-
drivers/vfio/mdev/mdev_core.c | 104 +++++++++++++-----
drivers/vfio/mdev/mdev_driver.c | 29 ++---
drivers/vfio/mdev/mdev_private.h | 13 ++-
drivers/vfio/mdev/mdev_vfio.c | 48 ++++++++
drivers/vfio/mdev/vfio_mdev.c | 5 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
include/linux/mdev.h | 16 ++-
include/linux/mdev_vfio.h | 25 +++++
samples/vfio-mdev/mbochs.c | 8 +-
samples/vfio-mdev/mdpy.c | 8 +-
samples/vfio-mdev/mtty.c | 8 +-
19 files changed, 269 insertions(+), 128 deletions(-)
create mode 100644 drivers/vfio/mdev/mdev_vfio.c
create mode 100644 include/linux/mdev_vfio.h
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 25eb7d5b834b..1887d27a565e 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -49,35 +49,37 @@ devices as examples, as these devices are the first devices to use this module::
+---------------+
| |
- | +-----------+ | mdev_register_driver() +--------------+
- | | | +<------------------------+ |
- | | mdev | | | |
- | | bus | +------------------------>+ vfio_mdev.ko |<-> VFIO user
- | | driver | | probe()/remove() | | APIs
- | | | | +--------------+
- | +-----------+ |
+ | MDEV CORE | mdev_register_driver() +--------------+
+ | MODULE +<------------------------+ |
+ | mdev.ko | | |
+ | +------------------------>+ vfio_mdev.ko |<-> VFIO user
+ | | probe()/remove() | | APIs
+ | | +--------------+
+ +---+-------+---+
+ | /|\
+ | |
+callbacks| | mdev_register_device()
+ | | mdev_register_bus()
+ \|/ |
+ +---+-------+---+
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | nvidia.ko |<-> physical
+ | +----------------------------->+ | device
+ | MDEV VFIO | callbacks +--------------+
+ | Physical |
+ | device | mdev_vfio_register_device() +--------------+
+ | interface |<-----------------------------+ |
+ | | | i915.ko |<-> physical
+ | mdev_vfio.ko +----------------------------->+ | device
+ | | callbacks +--------------+
+ | |
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | ccw_device.ko|<-> physical
+ | +----------------------------->+ | device
+ | | callbacks +--------------+
| |
- | MDEV CORE |
- | MODULE |
- | mdev.ko |
- | +-----------+ | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | nvidia.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | Physical | |
- | | device | | mdev_register_device() +--------------+
- | | interface | |<------------------------+ |
- | | | | | i915.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | | |
- | | | | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | ccw_device.ko|<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | +-----------+ |
+---------------+
@@ -116,7 +118,8 @@ to register and unregister itself with the core driver:
* Register::
extern int mdev_register_driver(struct mdev_driver *drv,
- struct module *owner);
+ struct module *owner,
+ struct bus_type *bus);
* Unregister::
@@ -159,11 +162,12 @@ The callbacks in the mdev_parent_ops structure are as follows:
* 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::
+A driver should use the mdev_parent_ops structure and bus type in the
+function call to register itself with the mdev core driver::
extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops);
+ const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
However, the mdev_parent_ops structure is not required in the function call
that a driver should use to unregister itself with the mdev core driver::
diff --git a/MAINTAINERS b/MAINTAINERS
index 34ef0cf30ece..6d590afb62c3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17139,6 +17139,7 @@ S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
F: drivers/vfio/mdev/
F: include/linux/mdev.h
+F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
VFIO PLATFORM DRIVER
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 04a5a0d90823..afdb3de5ce2f 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -41,7 +41,7 @@
#include <linux/uuid.h>
#include <linux/kvm_host.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/debugfs.h>
#include <linux/nospec.h>
@@ -1554,7 +1554,7 @@ static ssize_t
vgpu_id_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
if (mdev) {
struct intel_vgpu *vgpu = (struct intel_vgpu *)
@@ -1606,12 +1606,12 @@ static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
return -EFAULT;
intel_vgpu_ops.supported_type_groups = kvm_vgpu_type_groups;
- return mdev_register_device(dev, &intel_vgpu_ops);
+ return mdev_vfio_register_device(dev, &intel_vgpu_ops);
}
static void kvmgt_host_exit(struct device *dev)
{
- mdev_unregister_device(dev);
+ mdev_vfio_unregister_device(dev);
}
static int kvmgt_page_track_add(unsigned long handle, u64 gfn)
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index f0d71ab77c50..791b8b0eb027 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -11,7 +11,7 @@
*/
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/nospec.h>
#include <linux/slab.h>
@@ -588,10 +588,10 @@ static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
int vfio_ccw_mdev_reg(struct subchannel *sch)
{
- return mdev_register_device(&sch->dev, &vfio_ccw_mdev_ops);
+ return mdev_vfio_register_device(&sch->dev, &vfio_ccw_mdev_ops);
}
void vfio_ccw_mdev_unreg(struct subchannel *sch)
{
- mdev_unregister_device(&sch->dev);
+ mdev_vfio_unregister_device(&sch->dev);
}
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 5c0f53c6dde7..78048e670374 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -602,7 +602,7 @@ static ssize_t assign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of adapter */
@@ -668,7 +668,7 @@ static ssize_t unassign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of adapter */
@@ -748,7 +748,7 @@ static ssize_t assign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_apqi = matrix_mdev->matrix.aqm_max;
@@ -810,7 +810,7 @@ static ssize_t unassign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of domain */
@@ -854,7 +854,7 @@ static ssize_t assign_control_domain_store(struct device *dev,
{
int ret;
unsigned long id;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of control domain */
@@ -903,7 +903,7 @@ static ssize_t unassign_control_domain_store(struct device *dev,
{
int ret;
unsigned long domid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -933,7 +933,7 @@ static ssize_t control_domains_show(struct device *dev,
int nchars = 0;
int n;
char *bufpos = buf;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -952,7 +952,7 @@ static DEVICE_ATTR_RO(control_domains);
static ssize_t matrix_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
char *bufpos = buf;
unsigned long apid;
@@ -1295,10 +1295,11 @@ int vfio_ap_mdev_register(void)
{
atomic_set(&matrix_dev->available_instances, MAX_ZDEV_ENTRIES_EXT);
- return mdev_register_device(&matrix_dev->device, &vfio_ap_matrix_ops);
+ return mdev_vfio_register_device(&matrix_dev->device,
+ &vfio_ap_matrix_ops);
}
void vfio_ap_mdev_unregister(void)
{
- mdev_unregister_device(&matrix_dev->device);
+ mdev_vfio_unregister_device(&matrix_dev->device);
}
diff --git a/drivers/s390/crypto/vfio_ap_private.h b/drivers/s390/crypto/vfio_ap_private.h
index f46dde56b464..4e37e0e3433a 100644
--- a/drivers/s390/crypto/vfio_ap_private.h
+++ b/drivers/s390/crypto/vfio_ap_private.h
@@ -14,7 +14,7 @@
#include <linux/types.h>
#include <linux/device.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/delay.h>
#include <linux/mutex.h>
#include <linux/kvm_host.h>
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 5da27f2100f9..2e07ca915a96 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,15 +1,24 @@
-# SPDX-License-Identifier: GPL-2.0-only
-config VFIO_MDEV
+config MDEV
tristate "Mediated device driver framework"
- depends on VFIO
default n
help
Provides a framework to virtualize devices.
- See Documentation/driver-api/vfio-mediated-device.rst for more details.
If you don't know what do here, say N.
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
+
+
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index 101516fdf375..e9675501271a 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,6 @@
-# SPDX-License-Identifier: GPL-2.0-only
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-obj-$(CONFIG_VFIO_MDEV) += mdev.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index b558d4cfd082..e1272a40c521 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -22,11 +22,13 @@
static LIST_HEAD(parent_list);
static DEFINE_MUTEX(parent_list_lock);
-static struct class_compat *mdev_bus_compat_class;
static LIST_HEAD(mdev_list);
static DEFINE_MUTEX(mdev_list_lock);
+static LIST_HEAD(class_compat_list);
+static DEFINE_MUTEX(compat_list_lock);
+
struct device *mdev_parent_dev(struct mdev_device *mdev)
{
return mdev->parent->dev;
@@ -51,9 +53,9 @@ struct device *mdev_dev(struct mdev_device *mdev)
}
EXPORT_SYMBOL(mdev_dev);
-struct mdev_device *mdev_from_dev(struct device *dev)
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus)
{
- return dev_is_mdev(dev) ? to_mdev_device(dev) : NULL;
+ return dev_is_mdev(dev, bus) ? to_mdev_device(dev) : NULL;
}
EXPORT_SYMBOL(mdev_from_dev);
@@ -122,7 +124,9 @@ static void mdev_device_remove_common(struct mdev_device *mdev)
static int mdev_device_remove_cb(struct device *dev, void *data)
{
- if (dev_is_mdev(dev)) {
+ struct bus_type *bus = data;
+
+ if (dev_is_mdev(dev, bus)) {
struct mdev_device *mdev;
mdev = to_mdev_device(dev);
@@ -131,6 +135,41 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
return 0;
}
+static struct mdev_class_compat *get_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat;
+
+ list_for_each_entry(mdev_class_compat, &class_compat_list, next) {
+ if (mdev_class_compat->bus == bus)
+ return mdev_class_compat;
+ }
+
+ return NULL;
+}
+
+static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat = get_class_compat(bus);
+ char class_name[64];
+
+ if (mdev_class_compat)
+ return mdev_class_compat->class_compat;
+
+ mdev_class_compat = kmalloc(sizeof(*mdev_class_compat), GFP_KERNEL);
+ if (!mdev_class_compat)
+ return NULL;
+ snprintf(class_name, 64, "%s_bus", bus->name);
+ mdev_class_compat->class_compat = class_compat_register(class_name);
+ if (!mdev_class_compat->class_compat) {
+ kfree(mdev_class_compat);
+ return NULL;
+ }
+ mdev_class_compat->bus = bus;
+ list_add(&mdev_class_compat->next, &class_compat_list);
+
+ return mdev_class_compat->class_compat;
+}
+
/*
* mdev_register_device : Register a device
* @dev: device structure representing parent device.
@@ -139,12 +178,14 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
* Add device to list of registered parent devices.
* Returns a negative value on error, otherwise 0.
*/
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus)
{
int ret;
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=registered";
char *envp[] = { env_string, NULL };
+ struct class_compat *class_compat;
/* check for mandatory ops */
if (!ops || !ops->create || !ops->remove || !ops->supported_type_groups)
@@ -175,20 +216,21 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
parent->dev = dev;
parent->ops = ops;
+ parent->bus = bus;
- if (!mdev_bus_compat_class) {
- mdev_bus_compat_class = class_compat_register("mdev_bus");
- if (!mdev_bus_compat_class) {
- ret = -ENOMEM;
- goto add_dev_err;
- }
+ mutex_lock(&compat_list_lock);
+ class_compat = mdev_alloc_class_compat(bus);
+ mutex_unlock(&compat_list_lock);
+ if (!class_compat) {
+ ret = -ENOMEM;
+ goto add_dev_err;
}
ret = parent_create_sysfs_files(parent);
if (ret)
goto add_dev_err;
- ret = class_compat_create_link(mdev_bus_compat_class, dev, NULL);
+ ret = class_compat_create_link(class_compat, dev, NULL);
if (ret)
dev_warn(dev, "Failed to create compatibility class link\n");
@@ -223,6 +265,7 @@ void mdev_unregister_device(struct device *dev)
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=unregistered";
char *envp[] = { env_string, NULL };
+ struct mdev_class_compat *mdev_class_compat;
mutex_lock(&parent_list_lock);
parent = __find_parent_device(dev);
@@ -238,9 +281,13 @@ void mdev_unregister_device(struct device *dev)
down_write(&parent->unreg_sem);
- class_compat_remove_link(mdev_bus_compat_class, dev, NULL);
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(parent->bus);
+ WARN_ON(!mdev_class_compat);
+ class_compat_remove_link(mdev_class_compat->class_compat, dev, NULL);
+ mutex_unlock(&compat_list_lock);
- device_for_each_child(dev, NULL, mdev_device_remove_cb);
+ device_for_each_child(dev, parent->bus, mdev_device_remove_cb);
parent_remove_sysfs_files(parent);
up_write(&parent->unreg_sem);
@@ -314,7 +361,7 @@ int mdev_device_create(struct kobject *kobj,
device_initialize(&mdev->dev);
mdev->dev.parent = dev;
- mdev->dev.bus = &mdev_bus_type;
+ mdev->dev.bus = parent->bus;
mdev->dev.release = mdev_device_release;
dev_set_name(&mdev->dev, "%pUl", uuid);
mdev->dev.groups = parent->ops->mdev_attr_groups;
@@ -404,24 +451,29 @@ struct device *mdev_get_iommu_device(struct device *dev)
}
EXPORT_SYMBOL(mdev_get_iommu_device);
-static int __init mdev_init(void)
+int mdev_register_bus(struct bus_type *bus)
{
- return mdev_bus_register();
+ return bus_register(bus);
}
+EXPORT_SYMBOL(mdev_register_bus);
-static void __exit mdev_exit(void)
+void mdev_unregister_bus(struct bus_type *bus)
{
- if (mdev_bus_compat_class)
- class_compat_unregister(mdev_bus_compat_class);
-
- mdev_bus_unregister();
+ struct mdev_class_compat *mdev_class_compat;
+
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(bus);
+ if (mdev_class_compat) {
+ list_del(&mdev_class_compat->next);
+ class_compat_unregister(mdev_class_compat->class_compat);
+ kfree(mdev_class_compat);
+ }
+ bus_unregister(bus);
+ mutex_unlock(&compat_list_lock);
}
-
-module_init(mdev_init)
-module_exit(mdev_exit)
+EXPORT_SYMBOL(mdev_unregister_bus);
MODULE_VERSION(DRIVER_VERSION);
MODULE_LICENSE("GPL v2");
MODULE_AUTHOR(DRIVER_AUTHOR);
MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/vfio/mdev/mdev_driver.c
index 0d3223aee20b..c3a2ac023712 100644
--- a/drivers/vfio/mdev/mdev_driver.c
+++ b/drivers/vfio/mdev/mdev_driver.c
@@ -10,6 +10,7 @@
#include <linux/device.h>
#include <linux/iommu.h>
#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -37,7 +38,7 @@ static void mdev_detach_iommu(struct mdev_device *mdev)
dev_info(&mdev->dev, "MDEV: detaching iommu\n");
}
-static int mdev_probe(struct device *dev)
+int mdev_probe(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -55,8 +56,9 @@ static int mdev_probe(struct device *dev)
return ret;
}
+EXPORT_SYMBOL(mdev_probe);
-static int mdev_remove(struct device *dev)
+int mdev_remove(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -68,26 +70,22 @@ static int mdev_remove(struct device *dev)
return 0;
}
-
-struct bus_type mdev_bus_type = {
- .name = "mdev",
- .probe = mdev_probe,
- .remove = mdev_remove,
-};
-EXPORT_SYMBOL_GPL(mdev_bus_type);
+EXPORT_SYMBOL(mdev_remove);
/**
* mdev_register_driver - register a new MDEV driver
* @drv: the driver to register
* @owner: module owner of driver to be registered
+ * @bus: but that the driver wants to attach
*
* Returns a negative value on error, otherwise 0.
**/
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner)
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus)
{
/* initialize common driver fields */
drv->driver.name = drv->name;
- drv->driver.bus = &mdev_bus_type;
+ drv->driver.bus = bus;
drv->driver.owner = owner;
/* register with core */
@@ -105,12 +103,3 @@ void mdev_unregister_driver(struct mdev_driver *drv)
}
EXPORT_SYMBOL(mdev_unregister_driver);
-int mdev_bus_register(void)
-{
- return bus_register(&mdev_bus_type);
-}
-
-void mdev_bus_unregister(void)
-{
- bus_unregister(&mdev_bus_type);
-}
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 7d922950caaf..298d7a0f493a 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -10,12 +10,10 @@
#ifndef MDEV_PRIVATE_H
#define MDEV_PRIVATE_H
-int mdev_bus_register(void);
-void mdev_bus_unregister(void);
-
struct mdev_parent {
struct device *dev;
const struct mdev_parent_ops *ops;
+ struct bus_type *bus;
struct kref ref;
struct list_head next;
struct kset *mdev_types_kset;
@@ -35,8 +33,15 @@ struct mdev_device {
bool active;
};
+struct mdev_class_compat {
+ struct class_compat *class_compat;
+ struct bus_type *bus;
+ struct list_head next;
+};
+
+
#define to_mdev_device(dev) container_of(dev, struct mdev_device, dev)
-#define dev_is_mdev(d) ((d)->bus == &mdev_bus_type)
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
struct mdev_type {
struct kobject kobj;
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
new file mode 100644
index 000000000000..f9d1191b9982
--- /dev/null
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -0,0 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0-only
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev_vfio.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VFIO bus"
+
+struct bus_type mdev_vfio_bus_type = {
+ .name = "mdev",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+};
+EXPORT_SYMBOL(mdev_vfio_bus_type);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_vfio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_vfio_bus_type);
+}
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+}
+EXPORT_SYMBOL(mdev_vfio_register_device);
+
+void mdev_vfio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_vfio_unregister_device);
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 30964a4e0a28..16e9ebe30d4a 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -13,7 +13,7 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -128,7 +128,8 @@ static struct mdev_driver vfio_mdev_driver = {
static int __init vfio_mdev_init(void)
{
- return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE);
+ return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE,
+ &mdev_vfio_bus_type);
}
static void __exit vfio_mdev_exit(void)
diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index d864277ea16f..f35523f822eb 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -34,7 +34,7 @@
#include <linux/uaccess.h>
#include <linux/vfio.h>
#include <linux/workqueue.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/notifier.h>
#include <linux/dma-iommu.h>
#include <linux/irqdomain.h>
@@ -1405,10 +1405,10 @@ static bool vfio_bus_is_mdev(struct bus_type *bus)
struct bus_type *mdev_bus;
bool ret = false;
- mdev_bus = symbol_get(mdev_bus_type);
+ mdev_bus = symbol_get(mdev_vfio_bus_type);
if (mdev_bus) {
ret = (bus == mdev_bus);
- symbol_put(mdev_bus_type);
+ symbol_put(mdev_vfio_bus_type);
}
return ret;
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index 0ce30ca78db0..ee2410246b3c 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -133,16 +133,22 @@ 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);
-extern struct bus_type mdev_bus_type;
-
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
void mdev_unregister_device(struct device *dev);
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner);
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
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);
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus);
+
+int mdev_probe(struct device *dev);
+int mdev_remove(struct device *dev);
+
+int mdev_register_bus(struct bus_type *bus);
+void mdev_unregister_bus(struct bus_type *bus);
#endif /* MDEV_H */
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
new file mode 100644
index 000000000000..446a7537e3fb
--- /dev/null
+++ b/include/linux/mdev_vfio.h
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VFIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef MDEV_VFIO_H
+#define MDEV_VFIO_H
+
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_vfio_bus_type;
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_vfio_unregister_device(struct device *dev);
+
+static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_vfio_bus_type);
+}
+
+#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index ac5c8c17b1ff..f041d58324b1 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -29,7 +29,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/dma-buf.h>
#include <linux/highmem.h>
@@ -1332,7 +1332,7 @@ static ssize_t
memory_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%d MB\n", mdev_state->type->mbytes);
@@ -1468,7 +1468,7 @@ static int __init mbochs_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mbochs_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mbochs_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1487,7 +1487,7 @@ static int __init mbochs_dev_init(void)
static void __exit mbochs_dev_exit(void)
{
mbochs_dev.bus = NULL;
- mdev_unregister_device(&mbochs_dev);
+ mdev_vfio_unregister_device(&mbochs_dev);
device_unregister(&mbochs_dev);
cdev_del(&mbochs_cdev);
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index cc86bf6566e4..9c32fe3795ad 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -25,7 +25,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <drm/drm_fourcc.h>
#include "mdpy-defs.h"
@@ -639,7 +639,7 @@ static ssize_t
resolution_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%dx%d\n",
@@ -775,7 +775,7 @@ static int __init mdpy_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mdpy_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mdpy_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -794,7 +794,7 @@ static int __init mdpy_dev_init(void)
static void __exit mdpy_dev_exit(void)
{
mdpy_dev.bus = NULL;
- mdev_unregister_device(&mdpy_dev);
+ mdev_vfio_unregister_device(&mdpy_dev);
device_unregister(&mdpy_dev);
cdev_del(&mdpy_cdev);
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index ce84a300a4da..6e4e6339e0f1 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -26,7 +26,7 @@
#include <linux/sysfs.h>
#include <linux/ctype.h>
#include <linux/file.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/serial.h>
#include <uapi/linux/serial_reg.h>
@@ -1285,7 +1285,7 @@ static ssize_t
sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- if (mdev_from_dev(dev))
+ if (vfio_mdev_from_dev(dev))
return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
return sprintf(buf, "\n");
@@ -1445,7 +1445,7 @@ static int __init mtty_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mtty_dev.dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mtty_dev.dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1471,7 +1471,7 @@ static int __init mtty_dev_init(void)
static void __exit mtty_dev_exit(void)
{
mtty_dev.dev.bus = NULL;
- mdev_unregister_device(&mtty_dev.dev);
+ mdev_vfio_unregister_device(&mtty_dev.dev);
device_unregister(&mtty_dev.dev);
idr_destroy(&mtty_dev.vd_idr);
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 1/6] mdev: make mdev bus agnostic
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
Current mdev is tied to a VFIO specific "mdev" bus. This prevent mdev
from being used by other types of API/buses. So this patch tries to make
mdev bus agnostic through making a mdev core a thin module:
- decouple VFIO bus specific bits from mdev_core.c to mdev_vfio.c and
introduce mdev_vfio module
- require to specify the type of bus when registering mdev device and
mdev driver
With those modifications mdev become a generic module that could be
used by multiple types of virtual buses and devices.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 68 ++++++------
MAINTAINERS | 1 +
drivers/gpu/drm/i915/gvt/kvmgt.c | 8 +-
drivers/s390/cio/vfio_ccw_ops.c | 6 +-
drivers/s390/crypto/vfio_ap_ops.c | 21 ++--
drivers/s390/crypto/vfio_ap_private.h | 2 +-
drivers/vfio/mdev/Kconfig | 17 ++-
drivers/vfio/mdev/Makefile | 4 +-
drivers/vfio/mdev/mdev_core.c | 104 +++++++++++++-----
drivers/vfio/mdev/mdev_driver.c | 29 ++---
drivers/vfio/mdev/mdev_private.h | 13 ++-
drivers/vfio/mdev/mdev_vfio.c | 48 ++++++++
drivers/vfio/mdev/vfio_mdev.c | 5 +-
drivers/vfio/vfio_iommu_type1.c | 6 +-
include/linux/mdev.h | 16 ++-
include/linux/mdev_vfio.h | 25 +++++
samples/vfio-mdev/mbochs.c | 8 +-
samples/vfio-mdev/mdpy.c | 8 +-
samples/vfio-mdev/mtty.c | 8 +-
19 files changed, 269 insertions(+), 128 deletions(-)
create mode 100644 drivers/vfio/mdev/mdev_vfio.c
create mode 100644 include/linux/mdev_vfio.h
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 25eb7d5b834b..1887d27a565e 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -49,35 +49,37 @@ devices as examples, as these devices are the first devices to use this module::
+---------------+
| |
- | +-----------+ | mdev_register_driver() +--------------+
- | | | +<------------------------+ |
- | | mdev | | | |
- | | bus | +------------------------>+ vfio_mdev.ko |<-> VFIO user
- | | driver | | probe()/remove() | | APIs
- | | | | +--------------+
- | +-----------+ |
+ | MDEV CORE | mdev_register_driver() +--------------+
+ | MODULE +<------------------------+ |
+ | mdev.ko | | |
+ | +------------------------>+ vfio_mdev.ko |<-> VFIO user
+ | | probe()/remove() | | APIs
+ | | +--------------+
+ +---+-------+---+
+ | /|\
+ | |
+callbacks| | mdev_register_device()
+ | | mdev_register_bus()
+ \|/ |
+ +---+-------+---+
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | nvidia.ko |<-> physical
+ | +----------------------------->+ | device
+ | MDEV VFIO | callbacks +--------------+
+ | Physical |
+ | device | mdev_vfio_register_device() +--------------+
+ | interface |<-----------------------------+ |
+ | | | i915.ko |<-> physical
+ | mdev_vfio.ko +----------------------------->+ | device
+ | | callbacks +--------------+
+ | |
+ | | mdev_vfio_register_device() +--------------+
+ | +<-----------------------------+ |
+ | | | ccw_device.ko|<-> physical
+ | +----------------------------->+ | device
+ | | callbacks +--------------+
| |
- | MDEV CORE |
- | MODULE |
- | mdev.ko |
- | +-----------+ | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | nvidia.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | Physical | |
- | | device | | mdev_register_device() +--------------+
- | | interface | |<------------------------+ |
- | | | | | i915.ko |<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | | | |
- | | | | mdev_register_device() +--------------+
- | | | +<------------------------+ |
- | | | | | ccw_device.ko|<-> physical
- | | | +------------------------>+ | device
- | | | | callbacks +--------------+
- | +-----------+ |
+---------------+
@@ -116,7 +118,8 @@ to register and unregister itself with the core driver:
* Register::
extern int mdev_register_driver(struct mdev_driver *drv,
- struct module *owner);
+ struct module *owner,
+ struct bus_type *bus);
* Unregister::
@@ -159,11 +162,12 @@ The callbacks in the mdev_parent_ops structure are as follows:
* 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::
+A driver should use the mdev_parent_ops structure and bus type in the
+function call to register itself with the mdev core driver::
extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops);
+ const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
However, the mdev_parent_ops structure is not required in the function call
that a driver should use to unregister itself with the mdev core driver::
diff --git a/MAINTAINERS b/MAINTAINERS
index 34ef0cf30ece..6d590afb62c3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17139,6 +17139,7 @@ S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
F: drivers/vfio/mdev/
F: include/linux/mdev.h
+F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
VFIO PLATFORM DRIVER
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 04a5a0d90823..afdb3de5ce2f 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -41,7 +41,7 @@
#include <linux/uuid.h>
#include <linux/kvm_host.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/debugfs.h>
#include <linux/nospec.h>
@@ -1554,7 +1554,7 @@ static ssize_t
vgpu_id_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
if (mdev) {
struct intel_vgpu *vgpu = (struct intel_vgpu *)
@@ -1606,12 +1606,12 @@ static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
return -EFAULT;
intel_vgpu_ops.supported_type_groups = kvm_vgpu_type_groups;
- return mdev_register_device(dev, &intel_vgpu_ops);
+ return mdev_vfio_register_device(dev, &intel_vgpu_ops);
}
static void kvmgt_host_exit(struct device *dev)
{
- mdev_unregister_device(dev);
+ mdev_vfio_unregister_device(dev);
}
static int kvmgt_page_track_add(unsigned long handle, u64 gfn)
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index f0d71ab77c50..791b8b0eb027 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -11,7 +11,7 @@
*/
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/nospec.h>
#include <linux/slab.h>
@@ -588,10 +588,10 @@ static const struct mdev_parent_ops vfio_ccw_mdev_ops = {
int vfio_ccw_mdev_reg(struct subchannel *sch)
{
- return mdev_register_device(&sch->dev, &vfio_ccw_mdev_ops);
+ return mdev_vfio_register_device(&sch->dev, &vfio_ccw_mdev_ops);
}
void vfio_ccw_mdev_unreg(struct subchannel *sch)
{
- mdev_unregister_device(&sch->dev);
+ mdev_vfio_unregister_device(&sch->dev);
}
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 5c0f53c6dde7..78048e670374 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -602,7 +602,7 @@ static ssize_t assign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of adapter */
@@ -668,7 +668,7 @@ static ssize_t unassign_adapter_store(struct device *dev,
{
int ret;
unsigned long apid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of adapter */
@@ -748,7 +748,7 @@ static ssize_t assign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_apqi = matrix_mdev->matrix.aqm_max;
@@ -810,7 +810,7 @@ static ssize_t unassign_domain_store(struct device *dev,
{
int ret;
unsigned long apqi;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow un-assignment of domain */
@@ -854,7 +854,7 @@ static ssize_t assign_control_domain_store(struct device *dev,
{
int ret;
unsigned long id;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
/* If the guest is running, disallow assignment of control domain */
@@ -903,7 +903,7 @@ static ssize_t unassign_control_domain_store(struct device *dev,
{
int ret;
unsigned long domid;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -933,7 +933,7 @@ static ssize_t control_domains_show(struct device *dev,
int nchars = 0;
int n;
char *bufpos = buf;
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
unsigned long max_domid = matrix_mdev->matrix.adm_max;
@@ -952,7 +952,7 @@ static DEVICE_ATTR_RO(control_domains);
static ssize_t matrix_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
char *bufpos = buf;
unsigned long apid;
@@ -1295,10 +1295,11 @@ int vfio_ap_mdev_register(void)
{
atomic_set(&matrix_dev->available_instances, MAX_ZDEV_ENTRIES_EXT);
- return mdev_register_device(&matrix_dev->device, &vfio_ap_matrix_ops);
+ return mdev_vfio_register_device(&matrix_dev->device,
+ &vfio_ap_matrix_ops);
}
void vfio_ap_mdev_unregister(void)
{
- mdev_unregister_device(&matrix_dev->device);
+ mdev_vfio_unregister_device(&matrix_dev->device);
}
diff --git a/drivers/s390/crypto/vfio_ap_private.h b/drivers/s390/crypto/vfio_ap_private.h
index f46dde56b464..4e37e0e3433a 100644
--- a/drivers/s390/crypto/vfio_ap_private.h
+++ b/drivers/s390/crypto/vfio_ap_private.h
@@ -14,7 +14,7 @@
#include <linux/types.h>
#include <linux/device.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/delay.h>
#include <linux/mutex.h>
#include <linux/kvm_host.h>
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 5da27f2100f9..2e07ca915a96 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,15 +1,24 @@
-# SPDX-License-Identifier: GPL-2.0-only
-config VFIO_MDEV
+config MDEV
tristate "Mediated device driver framework"
- depends on VFIO
default n
help
Provides a framework to virtualize devices.
- See Documentation/driver-api/vfio-mediated-device.rst for more details.
If you don't know what do here, say N.
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
+
+
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index 101516fdf375..e9675501271a 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,6 @@
-# SPDX-License-Identifier: GPL-2.0-only
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-obj-$(CONFIG_VFIO_MDEV) += mdev.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index b558d4cfd082..e1272a40c521 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -22,11 +22,13 @@
static LIST_HEAD(parent_list);
static DEFINE_MUTEX(parent_list_lock);
-static struct class_compat *mdev_bus_compat_class;
static LIST_HEAD(mdev_list);
static DEFINE_MUTEX(mdev_list_lock);
+static LIST_HEAD(class_compat_list);
+static DEFINE_MUTEX(compat_list_lock);
+
struct device *mdev_parent_dev(struct mdev_device *mdev)
{
return mdev->parent->dev;
@@ -51,9 +53,9 @@ struct device *mdev_dev(struct mdev_device *mdev)
}
EXPORT_SYMBOL(mdev_dev);
-struct mdev_device *mdev_from_dev(struct device *dev)
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus)
{
- return dev_is_mdev(dev) ? to_mdev_device(dev) : NULL;
+ return dev_is_mdev(dev, bus) ? to_mdev_device(dev) : NULL;
}
EXPORT_SYMBOL(mdev_from_dev);
@@ -122,7 +124,9 @@ static void mdev_device_remove_common(struct mdev_device *mdev)
static int mdev_device_remove_cb(struct device *dev, void *data)
{
- if (dev_is_mdev(dev)) {
+ struct bus_type *bus = data;
+
+ if (dev_is_mdev(dev, bus)) {
struct mdev_device *mdev;
mdev = to_mdev_device(dev);
@@ -131,6 +135,41 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
return 0;
}
+static struct mdev_class_compat *get_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat;
+
+ list_for_each_entry(mdev_class_compat, &class_compat_list, next) {
+ if (mdev_class_compat->bus == bus)
+ return mdev_class_compat;
+ }
+
+ return NULL;
+}
+
+static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
+{
+ struct mdev_class_compat *mdev_class_compat = get_class_compat(bus);
+ char class_name[64];
+
+ if (mdev_class_compat)
+ return mdev_class_compat->class_compat;
+
+ mdev_class_compat = kmalloc(sizeof(*mdev_class_compat), GFP_KERNEL);
+ if (!mdev_class_compat)
+ return NULL;
+ snprintf(class_name, 64, "%s_bus", bus->name);
+ mdev_class_compat->class_compat = class_compat_register(class_name);
+ if (!mdev_class_compat->class_compat) {
+ kfree(mdev_class_compat);
+ return NULL;
+ }
+ mdev_class_compat->bus = bus;
+ list_add(&mdev_class_compat->next, &class_compat_list);
+
+ return mdev_class_compat->class_compat;
+}
+
/*
* mdev_register_device : Register a device
* @dev: device structure representing parent device.
@@ -139,12 +178,14 @@ static int mdev_device_remove_cb(struct device *dev, void *data)
* Add device to list of registered parent devices.
* Returns a negative value on error, otherwise 0.
*/
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus)
{
int ret;
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=registered";
char *envp[] = { env_string, NULL };
+ struct class_compat *class_compat;
/* check for mandatory ops */
if (!ops || !ops->create || !ops->remove || !ops->supported_type_groups)
@@ -175,20 +216,21 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops)
parent->dev = dev;
parent->ops = ops;
+ parent->bus = bus;
- if (!mdev_bus_compat_class) {
- mdev_bus_compat_class = class_compat_register("mdev_bus");
- if (!mdev_bus_compat_class) {
- ret = -ENOMEM;
- goto add_dev_err;
- }
+ mutex_lock(&compat_list_lock);
+ class_compat = mdev_alloc_class_compat(bus);
+ mutex_unlock(&compat_list_lock);
+ if (!class_compat) {
+ ret = -ENOMEM;
+ goto add_dev_err;
}
ret = parent_create_sysfs_files(parent);
if (ret)
goto add_dev_err;
- ret = class_compat_create_link(mdev_bus_compat_class, dev, NULL);
+ ret = class_compat_create_link(class_compat, dev, NULL);
if (ret)
dev_warn(dev, "Failed to create compatibility class link\n");
@@ -223,6 +265,7 @@ void mdev_unregister_device(struct device *dev)
struct mdev_parent *parent;
char *env_string = "MDEV_STATE=unregistered";
char *envp[] = { env_string, NULL };
+ struct mdev_class_compat *mdev_class_compat;
mutex_lock(&parent_list_lock);
parent = __find_parent_device(dev);
@@ -238,9 +281,13 @@ void mdev_unregister_device(struct device *dev)
down_write(&parent->unreg_sem);
- class_compat_remove_link(mdev_bus_compat_class, dev, NULL);
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(parent->bus);
+ WARN_ON(!mdev_class_compat);
+ class_compat_remove_link(mdev_class_compat->class_compat, dev, NULL);
+ mutex_unlock(&compat_list_lock);
- device_for_each_child(dev, NULL, mdev_device_remove_cb);
+ device_for_each_child(dev, parent->bus, mdev_device_remove_cb);
parent_remove_sysfs_files(parent);
up_write(&parent->unreg_sem);
@@ -314,7 +361,7 @@ int mdev_device_create(struct kobject *kobj,
device_initialize(&mdev->dev);
mdev->dev.parent = dev;
- mdev->dev.bus = &mdev_bus_type;
+ mdev->dev.bus = parent->bus;
mdev->dev.release = mdev_device_release;
dev_set_name(&mdev->dev, "%pUl", uuid);
mdev->dev.groups = parent->ops->mdev_attr_groups;
@@ -404,24 +451,29 @@ struct device *mdev_get_iommu_device(struct device *dev)
}
EXPORT_SYMBOL(mdev_get_iommu_device);
-static int __init mdev_init(void)
+int mdev_register_bus(struct bus_type *bus)
{
- return mdev_bus_register();
+ return bus_register(bus);
}
+EXPORT_SYMBOL(mdev_register_bus);
-static void __exit mdev_exit(void)
+void mdev_unregister_bus(struct bus_type *bus)
{
- if (mdev_bus_compat_class)
- class_compat_unregister(mdev_bus_compat_class);
-
- mdev_bus_unregister();
+ struct mdev_class_compat *mdev_class_compat;
+
+ mutex_lock(&compat_list_lock);
+ mdev_class_compat = get_class_compat(bus);
+ if (mdev_class_compat) {
+ list_del(&mdev_class_compat->next);
+ class_compat_unregister(mdev_class_compat->class_compat);
+ kfree(mdev_class_compat);
+ }
+ bus_unregister(bus);
+ mutex_unlock(&compat_list_lock);
}
-
-module_init(mdev_init)
-module_exit(mdev_exit)
+EXPORT_SYMBOL(mdev_unregister_bus);
MODULE_VERSION(DRIVER_VERSION);
MODULE_LICENSE("GPL v2");
MODULE_AUTHOR(DRIVER_AUTHOR);
MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/vfio/mdev/mdev_driver.c
index 0d3223aee20b..c3a2ac023712 100644
--- a/drivers/vfio/mdev/mdev_driver.c
+++ b/drivers/vfio/mdev/mdev_driver.c
@@ -10,6 +10,7 @@
#include <linux/device.h>
#include <linux/iommu.h>
#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -37,7 +38,7 @@ static void mdev_detach_iommu(struct mdev_device *mdev)
dev_info(&mdev->dev, "MDEV: detaching iommu\n");
}
-static int mdev_probe(struct device *dev)
+int mdev_probe(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -55,8 +56,9 @@ static int mdev_probe(struct device *dev)
return ret;
}
+EXPORT_SYMBOL(mdev_probe);
-static int mdev_remove(struct device *dev)
+int mdev_remove(struct device *dev)
{
struct mdev_driver *drv = to_mdev_driver(dev->driver);
struct mdev_device *mdev = to_mdev_device(dev);
@@ -68,26 +70,22 @@ static int mdev_remove(struct device *dev)
return 0;
}
-
-struct bus_type mdev_bus_type = {
- .name = "mdev",
- .probe = mdev_probe,
- .remove = mdev_remove,
-};
-EXPORT_SYMBOL_GPL(mdev_bus_type);
+EXPORT_SYMBOL(mdev_remove);
/**
* mdev_register_driver - register a new MDEV driver
* @drv: the driver to register
* @owner: module owner of driver to be registered
+ * @bus: but that the driver wants to attach
*
* Returns a negative value on error, otherwise 0.
**/
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner)
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus)
{
/* initialize common driver fields */
drv->driver.name = drv->name;
- drv->driver.bus = &mdev_bus_type;
+ drv->driver.bus = bus;
drv->driver.owner = owner;
/* register with core */
@@ -105,12 +103,3 @@ void mdev_unregister_driver(struct mdev_driver *drv)
}
EXPORT_SYMBOL(mdev_unregister_driver);
-int mdev_bus_register(void)
-{
- return bus_register(&mdev_bus_type);
-}
-
-void mdev_bus_unregister(void)
-{
- bus_unregister(&mdev_bus_type);
-}
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 7d922950caaf..298d7a0f493a 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -10,12 +10,10 @@
#ifndef MDEV_PRIVATE_H
#define MDEV_PRIVATE_H
-int mdev_bus_register(void);
-void mdev_bus_unregister(void);
-
struct mdev_parent {
struct device *dev;
const struct mdev_parent_ops *ops;
+ struct bus_type *bus;
struct kref ref;
struct list_head next;
struct kset *mdev_types_kset;
@@ -35,8 +33,15 @@ struct mdev_device {
bool active;
};
+struct mdev_class_compat {
+ struct class_compat *class_compat;
+ struct bus_type *bus;
+ struct list_head next;
+};
+
+
#define to_mdev_device(dev) container_of(dev, struct mdev_device, dev)
-#define dev_is_mdev(d) ((d)->bus == &mdev_bus_type)
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
struct mdev_type {
struct kobject kobj;
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
new file mode 100644
index 000000000000..f9d1191b9982
--- /dev/null
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -0,0 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0-only
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev_vfio.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VFIO bus"
+
+struct bus_type mdev_vfio_bus_type = {
+ .name = "mdev",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+};
+EXPORT_SYMBOL(mdev_vfio_bus_type);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_vfio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_vfio_bus_type);
+}
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+}
+EXPORT_SYMBOL(mdev_vfio_register_device);
+
+void mdev_vfio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_vfio_unregister_device);
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_SOFTDEP("post: vfio_mdev");
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 30964a4e0a28..16e9ebe30d4a 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -13,7 +13,7 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/vfio.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include "mdev_private.h"
@@ -128,7 +128,8 @@ static struct mdev_driver vfio_mdev_driver = {
static int __init vfio_mdev_init(void)
{
- return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE);
+ return mdev_register_driver(&vfio_mdev_driver, THIS_MODULE,
+ &mdev_vfio_bus_type);
}
static void __exit vfio_mdev_exit(void)
diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c
index d864277ea16f..f35523f822eb 100644
--- a/drivers/vfio/vfio_iommu_type1.c
+++ b/drivers/vfio/vfio_iommu_type1.c
@@ -34,7 +34,7 @@
#include <linux/uaccess.h>
#include <linux/vfio.h>
#include <linux/workqueue.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/notifier.h>
#include <linux/dma-iommu.h>
#include <linux/irqdomain.h>
@@ -1405,10 +1405,10 @@ static bool vfio_bus_is_mdev(struct bus_type *bus)
struct bus_type *mdev_bus;
bool ret = false;
- mdev_bus = symbol_get(mdev_bus_type);
+ mdev_bus = symbol_get(mdev_vfio_bus_type);
if (mdev_bus) {
ret = (bus == mdev_bus);
- symbol_put(mdev_bus_type);
+ symbol_put(mdev_vfio_bus_type);
}
return ret;
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index 0ce30ca78db0..ee2410246b3c 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -133,16 +133,22 @@ 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);
-extern struct bus_type mdev_bus_type;
-
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus);
void mdev_unregister_device(struct device *dev);
-int mdev_register_driver(struct mdev_driver *drv, struct module *owner);
+int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
+ struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
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);
+struct mdev_device *mdev_from_dev(struct device *dev, struct bus_type *bus);
+
+int mdev_probe(struct device *dev);
+int mdev_remove(struct device *dev);
+
+int mdev_register_bus(struct bus_type *bus);
+void mdev_unregister_bus(struct bus_type *bus);
#endif /* MDEV_H */
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
new file mode 100644
index 000000000000..446a7537e3fb
--- /dev/null
+++ b/include/linux/mdev_vfio.h
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VFIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef MDEV_VFIO_H
+#define MDEV_VFIO_H
+
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_vfio_bus_type;
+
+int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_vfio_unregister_device(struct device *dev);
+
+static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_vfio_bus_type);
+}
+
+#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index ac5c8c17b1ff..f041d58324b1 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -29,7 +29,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/dma-buf.h>
#include <linux/highmem.h>
@@ -1332,7 +1332,7 @@ static ssize_t
memory_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%d MB\n", mdev_state->type->mbytes);
@@ -1468,7 +1468,7 @@ static int __init mbochs_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mbochs_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mbochs_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1487,7 +1487,7 @@ static int __init mbochs_dev_init(void)
static void __exit mbochs_dev_exit(void)
{
mbochs_dev.bus = NULL;
- mdev_unregister_device(&mbochs_dev);
+ mdev_vfio_unregister_device(&mbochs_dev);
device_unregister(&mbochs_dev);
cdev_del(&mbochs_cdev);
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index cc86bf6566e4..9c32fe3795ad 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -25,7 +25,7 @@
#include <linux/vfio.h>
#include <linux/iommu.h>
#include <linux/sysfs.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <drm/drm_fourcc.h>
#include "mdpy-defs.h"
@@ -639,7 +639,7 @@ static ssize_t
resolution_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- struct mdev_device *mdev = mdev_from_dev(dev);
+ struct mdev_device *mdev = vfio_mdev_from_dev(dev);
struct mdev_state *mdev_state = mdev_get_drvdata(mdev);
return sprintf(buf, "%dx%d\n",
@@ -775,7 +775,7 @@ static int __init mdpy_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mdpy_dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mdpy_dev, &mdev_fops);
if (ret)
goto failed3;
@@ -794,7 +794,7 @@ static int __init mdpy_dev_init(void)
static void __exit mdpy_dev_exit(void)
{
mdpy_dev.bus = NULL;
- mdev_unregister_device(&mdpy_dev);
+ mdev_vfio_unregister_device(&mdpy_dev);
device_unregister(&mdpy_dev);
cdev_del(&mdpy_cdev);
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index ce84a300a4da..6e4e6339e0f1 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -26,7 +26,7 @@
#include <linux/sysfs.h>
#include <linux/ctype.h>
#include <linux/file.h>
-#include <linux/mdev.h>
+#include <linux/mdev_vfio.h>
#include <linux/pci.h>
#include <linux/serial.h>
#include <uapi/linux/serial_reg.h>
@@ -1285,7 +1285,7 @@ static ssize_t
sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
- if (mdev_from_dev(dev))
+ if (vfio_mdev_from_dev(dev))
return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
return sprintf(buf, "\n");
@@ -1445,7 +1445,7 @@ static int __init mtty_dev_init(void)
if (ret)
goto failed2;
- ret = mdev_register_device(&mtty_dev.dev, &mdev_fops);
+ ret = mdev_vfio_register_device(&mtty_dev.dev, &mdev_fops);
if (ret)
goto failed3;
@@ -1471,7 +1471,7 @@ static int __init mtty_dev_init(void)
static void __exit mtty_dev_exit(void)
{
mtty_dev.dev.bus = NULL;
- mdev_unregister_device(&mtty_dev.dev);
+ mdev_vfio_unregister_device(&mtty_dev.dev);
device_unregister(&mtty_dev.dev);
idr_destroy(&mtty_dev.vd_idr);
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 2/6] mdev: split out VFIO bus specific parent ops
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, Jason Wang
The only thing left for generalizing mdev is the VFIO specific parent
ops. This is basically the open/release/read/write/ioctl/mmap.
To support this, mdev core is extend to support a specific size
of structure during create, this will allow to compose mdev structure
into mdev vfio structure and place the VFIO specific callbacks there
like:
struct mdev_vfio {
struct mdev_device mdev;
const struct mdev_vfio_ops *ops;
};
Helpers for setting and getting the ops were introduced to support
mdev vfio device to set ops and vfio mdev driver to use the ops.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 34 +++++++++------
drivers/gpu/drm/i915/gvt/kvmgt.c | 16 ++++---
drivers/s390/cio/vfio_ccw_ops.c | 17 +++++---
drivers/s390/crypto/vfio_ap_ops.c | 13 ++++--
drivers/vfio/mdev/mdev_core.c | 5 ++-
drivers/vfio/mdev/mdev_private.h | 5 +++
drivers/vfio/mdev/mdev_vfio.c | 30 ++++++++++++-
drivers/vfio/mdev/vfio_mdev.c | 38 ++++++++--------
include/linux/mdev.h | 37 ----------------
include/linux/mdev_vfio.h | 43 +++++++++++++++++++
samples/vfio-mdev/mbochs.c | 18 +++++---
samples/vfio-mdev/mdpy.c | 19 +++++---
samples/vfio-mdev/mtty.c | 16 ++++---
13 files changed, 189 insertions(+), 102 deletions(-)
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 1887d27a565e..9045584e4ea3 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -153,26 +153,36 @@ 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:
+A driver should use the mdev_parent_ops structure in the function call
+to register itself with the mdev core driver::
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
+ extern int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+
+However, the mdev_parent_ops structure is not required in the function call
+that a driver should use to unregister itself with the mdev core driver::
+
+ extern void mdev_vfio_unregister_device(struct device *dev);
+
+The VFIO specific callbacks is abstracted in mdev_vfio_ops structure
+are as follows:
+
+* 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
-A driver should use the mdev_parent_ops structure and bus type in the
-function call to register itself with the mdev core driver::
+During the creation of VFIO mediated device, mdev_vfio_ops need to be
+specified::
- extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops,
- struct bus_type *bus);
+ void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
-However, the mdev_parent_ops structure is not required in the function call
-that a driver should use to unregister itself with the mdev core driver::
+Those callbacks could be fetched by drivers through::
- extern void mdev_unregister_device(struct device *dev);
+ const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
Mediated Device Management Interface Through sysfs
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index afdb3de5ce2f..e72c36174035 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -643,6 +643,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
}
+static const struct mdev_vfio_ops intel_mdev_vfio_ops;
+
static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct intel_vgpu *vgpu = NULL;
@@ -678,6 +680,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
dev_name(mdev_dev(mdev)));
ret = 0;
+ mdev_vfio_set_ops(mdev, &intel_mdev_vfio_ops);
out:
return ret;
}
@@ -1581,20 +1584,21 @@ 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_ops intel_mdev_vfio_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;
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index 791b8b0eb027..811f0a3b1903 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -110,6 +110,8 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops vfio_ccw_mdev_vfio_ops;
+
static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct vfio_ccw_private *private =
@@ -129,6 +131,8 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
private->sch->schid.ssid,
private->sch->schid.sch_no);
+ mdev_vfio_set_ops(mdev, &vfio_ccw_mdev_vfio_ops);
+
return 0;
}
@@ -574,16 +578,19 @@ 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_ops vfio_ccw_mdev_vfio_ops = {
.open = vfio_ccw_mdev_open,
.release = vfio_ccw_mdev_release,
.read = vfio_ccw_mdev_read,
.write = vfio_ccw_mdev_write,
.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)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 78048e670374..0649c68287d7 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -321,6 +321,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
matrix->adm_max = info->apxa ? info->Nd : 15;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_ops;
+
static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct ap_matrix_mdev *matrix_mdev;
@@ -343,6 +345,8 @@ 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_vfio_set_ops(mdev, &vfio_ap_matrix_mdev_ops);
+
return 0;
}
@@ -1280,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
return ret;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_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)
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index e1272a40c521..c6bc67bf63fa 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -179,7 +179,7 @@ static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
* Returns a negative value on error, otherwise 0.
*/
int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus)
+ struct bus_type *bus, size_t dev_size)
{
int ret;
struct mdev_parent *parent;
@@ -217,6 +217,7 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
parent->dev = dev;
parent->ops = ops;
parent->bus = bus;
+ parent->dev_size = dev_size;
mutex_lock(&compat_list_lock);
class_compat = mdev_alloc_class_compat(bus);
@@ -339,7 +340,7 @@ int mdev_device_create(struct kobject *kobj,
}
}
- mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
+ mdev = kzalloc(parent->dev_size, GFP_KERNEL);
if (!mdev) {
mutex_unlock(&mdev_list_lock);
ret = -ENOMEM;
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 298d7a0f493a..012ab80719e9 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -20,6 +20,7 @@ struct mdev_parent {
struct list_head type_list;
/* Synchronize device creation/removal with parent unregistration */
struct rw_semaphore unreg_sem;
+ size_t dev_size;
};
struct mdev_device {
@@ -66,4 +67,8 @@ int mdev_device_create(struct kobject *kobj,
struct device *dev, const guid_t *uuid);
int mdev_device_remove(struct device *dev);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus, size_t dev_size);
+void mdev_unregister_device(struct device *dev);
+
#endif /* MDEV_PRIVATE_H */
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
index f9d1191b9982..44e116074f88 100644
--- a/drivers/vfio/mdev/mdev_vfio.c
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -4,6 +4,8 @@
#include <linux/device.h>
#include <linux/mdev_vfio.h>
+#include "mdev_private.h"
+
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "Jason Wang"
#define DRIVER_DESC "Mediated VFIO bus"
@@ -15,6 +17,31 @@ struct bus_type mdev_vfio_bus_type = {
};
EXPORT_SYMBOL(mdev_vfio_bus_type);
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
+
+struct mdev_vfio_device {
+ struct mdev_device mdev;
+ const struct mdev_vfio_ops *ops;
+};
+
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ mdev_vfio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_vfio_set_ops);
+
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ return mdev_vfio->ops;
+}
+EXPORT_SYMBOL(mdev_vfio_get_ops);
+
static int __init mdev_init(void)
{
return mdev_register_bus(&mdev_vfio_bus_type);
@@ -28,7 +55,8 @@ static void __exit mdev_exit(void)
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops)
{
- return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type,
+ sizeof(struct mdev_vfio_device));
}
EXPORT_SYMBOL(mdev_vfio_register_device);
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 16e9ebe30d4a..8b42a4b3f161 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -24,16 +24,16 @@
static int vfio_mdev_open(void *device_data)
{
struct mdev_device *mdev = device_data;
- struct mdev_parent *parent = mdev->parent;
+ const struct mdev_vfio_ops *ops = mdev_vfio_get_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 +43,10 @@ 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_ops *ops = mdev_vfio_get_ops(mdev);
- if (likely(parent->ops->release))
- parent->ops->release(mdev);
+ if (likely(ops->release))
+ ops->release(mdev);
module_put(THIS_MODULE);
}
@@ -55,47 +55,47 @@ 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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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 = {
@@ -110,7 +110,7 @@ static const struct vfio_device_ops vfio_mdev_dev_ops = {
static int vfio_mdev_probe(struct device *dev)
{
- struct mdev_device *mdev = to_mdev_device(dev);
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_vfio_bus_type);
return vfio_add_group_dev(dev, &vfio_mdev_dev_ops, mdev);
}
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index ee2410246b3c..25554e55bcee 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -48,30 +48,6 @@ struct device *mdev_get_iommu_device(struct device *dev);
* @mdev: mdev_device device structure which is being
* destroyed
* Returns integer: success (0) or error (< 0)
- * @open: Open mediated device.
- * @mdev: mediated device.
- * Returns integer: success (0) or error (< 0)
- * @release: release mediated device
- * @mdev: mediated device.
- * @read: Read emulation callback
- * @mdev: mediated device structure
- * @buf: read buffer
- * @count: number of bytes to read
- * @ppos: address.
- * Retuns number on bytes read on success or error.
- * @write: Write emulation callback
- * @mdev: mediated device structure
- * @buf: write buffer
- * @count: number of bytes to be written
- * @ppos: address.
- * Retuns number on bytes written on success or error.
- * @ioctl: IOCTL callback
- * @mdev: mediated device structure
- * @cmd: ioctl command
- * @arg: arguments to ioctl
- * @mmap: mmap callback
- * @mdev: mediated device structure
- * @vma: vma structure
* Parent device that support mediated device should be registered with mdev
* module with mdev_parent_ops structure.
**/
@@ -83,15 +59,6 @@ struct mdev_parent_ops {
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
- int (*open)(struct mdev_device *mdev);
- void (*release)(struct mdev_device *mdev);
- ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
- size_t count, loff_t *ppos);
- ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
- size_t count, loff_t *ppos);
- long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
- unsigned long arg);
- int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
/* interface for exporting mdev supported type attributes */
@@ -133,10 +100,6 @@ 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);
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus);
-void mdev_unregister_device(struct device *dev);
-
int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
index 446a7537e3fb..243ad85019a6 100644
--- a/include/linux/mdev_vfio.h
+++ b/include/linux/mdev_vfio.h
@@ -13,6 +13,45 @@
extern struct bus_type mdev_vfio_bus_type;
+/* VFIO mdev ops
+ *
+ * @open: Open mediated device.
+ * @mdev: mediated device.
+ * Returns integer: success (0) or error (< 0)
+ * @release: release mediated device
+ * @mdev: mediated device.
+ * @read: Read emulation callback
+ * @mdev: mediated device structure
+ * @buf: read buffer
+ * @count: number of bytes to read
+ * @ppos: address.
+ * Retuns number on bytes read on success or error.
+ * @write: Write emulation callback
+ * @mdev: mediated device structure
+ * @buf: write buffer
+ * @count: number of bytes to be written
+ * @ppos: address.
+ * Retuns number on bytes written on success or error.
+ * @ioctl: IOCTL callback
+ * @mdev: mediated device structure
+ * @cmd: ioctl command
+ * @arg: arguments to ioctl
+ * @mmap: mmap callback
+ * @mdev: mediated device structure
+ * @vma: vma structure
+ */
+struct mdev_vfio_ops {
+ int (*open)(struct mdev_device *mdev);
+ void (*release)(struct mdev_device *mdev);
+ ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
+ size_t count, loff_t *ppos);
+ ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
+ size_t count, loff_t *ppos);
+ long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
+ unsigned long arg);
+ int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
+};
+
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops);
void mdev_vfio_unregister_device(struct device *dev);
@@ -22,4 +61,8 @@ static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
return mdev_from_dev(dev, &mdev_vfio_bus_type);
}
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
+
#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index f041d58324b1..b2ba32b5fed2 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -516,6 +516,8 @@ static int mbochs_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,6 +563,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
mbochs_reset(mdev);
mbochs_used_mbytes += type->mbytes;
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
err_mem:
@@ -1418,12 +1421,7 @@ 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_ops mdev_ops = {
.open = mbochs_open,
.release = mbochs_close,
.read = mbochs_read,
@@ -1432,6 +1430,14 @@ 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,
};
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 9c32fe3795ad..d26fd94b4783 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -226,6 +226,8 @@ static int mdpy_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,6 +271,8 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
mdpy_reset(mdev);
mdpy_count++;
+
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -725,12 +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_ops mdev_ops = {
.open = mdpy_open,
.release = mdpy_close,
.read = mdpy_read,
@@ -739,6 +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,
};
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index 6e4e6339e0f1..3f0c6506199a 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -708,6 +708,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
return ret;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct mdev_state *mdev_state;
@@ -755,6 +757,7 @@ 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_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -1387,6 +1390,14 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops mdev_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,
@@ -1394,11 +1405,6 @@ 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)
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 2/6] mdev: split out VFIO bus specific parent ops
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam
The only thing left for generalizing mdev is the VFIO specific parent
ops. This is basically the open/release/read/write/ioctl/mmap.
To support this, mdev core is extend to support a specific size
of structure during create, this will allow to compose mdev structure
into mdev vfio structure and place the VFIO specific callbacks there
like:
struct mdev_vfio {
struct mdev_device mdev;
const struct mdev_vfio_ops *ops;
};
Helpers for setting and getting the ops were introduced to support
mdev vfio device to set ops and vfio mdev driver to use the ops.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 34 +++++++++------
drivers/gpu/drm/i915/gvt/kvmgt.c | 16 ++++---
drivers/s390/cio/vfio_ccw_ops.c | 17 +++++---
drivers/s390/crypto/vfio_ap_ops.c | 13 ++++--
drivers/vfio/mdev/mdev_core.c | 5 ++-
drivers/vfio/mdev/mdev_private.h | 5 +++
drivers/vfio/mdev/mdev_vfio.c | 30 ++++++++++++-
drivers/vfio/mdev/vfio_mdev.c | 38 ++++++++--------
include/linux/mdev.h | 37 ----------------
include/linux/mdev_vfio.h | 43 +++++++++++++++++++
samples/vfio-mdev/mbochs.c | 18 +++++---
samples/vfio-mdev/mdpy.c | 19 +++++---
samples/vfio-mdev/mtty.c | 16 ++++---
13 files changed, 189 insertions(+), 102 deletions(-)
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 1887d27a565e..9045584e4ea3 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -153,26 +153,36 @@ 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:
+A driver should use the mdev_parent_ops structure in the function call
+to register itself with the mdev core driver::
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
+ extern int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+
+However, the mdev_parent_ops structure is not required in the function call
+that a driver should use to unregister itself with the mdev core driver::
+
+ extern void mdev_vfio_unregister_device(struct device *dev);
+
+The VFIO specific callbacks is abstracted in mdev_vfio_ops structure
+are as follows:
+
+* 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
-A driver should use the mdev_parent_ops structure and bus type in the
-function call to register itself with the mdev core driver::
+During the creation of VFIO mediated device, mdev_vfio_ops need to be
+specified::
- extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops,
- struct bus_type *bus);
+ void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
-However, the mdev_parent_ops structure is not required in the function call
-that a driver should use to unregister itself with the mdev core driver::
+Those callbacks could be fetched by drivers through::
- extern void mdev_unregister_device(struct device *dev);
+ const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
Mediated Device Management Interface Through sysfs
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index afdb3de5ce2f..e72c36174035 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -643,6 +643,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
}
+static const struct mdev_vfio_ops intel_mdev_vfio_ops;
+
static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct intel_vgpu *vgpu = NULL;
@@ -678,6 +680,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
dev_name(mdev_dev(mdev)));
ret = 0;
+ mdev_vfio_set_ops(mdev, &intel_mdev_vfio_ops);
out:
return ret;
}
@@ -1581,20 +1584,21 @@ 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_ops intel_mdev_vfio_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;
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index 791b8b0eb027..811f0a3b1903 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -110,6 +110,8 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops vfio_ccw_mdev_vfio_ops;
+
static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct vfio_ccw_private *private =
@@ -129,6 +131,8 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
private->sch->schid.ssid,
private->sch->schid.sch_no);
+ mdev_vfio_set_ops(mdev, &vfio_ccw_mdev_vfio_ops);
+
return 0;
}
@@ -574,16 +578,19 @@ 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_ops vfio_ccw_mdev_vfio_ops = {
.open = vfio_ccw_mdev_open,
.release = vfio_ccw_mdev_release,
.read = vfio_ccw_mdev_read,
.write = vfio_ccw_mdev_write,
.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)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 78048e670374..0649c68287d7 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -321,6 +321,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
matrix->adm_max = info->apxa ? info->Nd : 15;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_ops;
+
static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct ap_matrix_mdev *matrix_mdev;
@@ -343,6 +345,8 @@ 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_vfio_set_ops(mdev, &vfio_ap_matrix_mdev_ops);
+
return 0;
}
@@ -1280,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
return ret;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_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)
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index e1272a40c521..c6bc67bf63fa 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -179,7 +179,7 @@ static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
* Returns a negative value on error, otherwise 0.
*/
int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus)
+ struct bus_type *bus, size_t dev_size)
{
int ret;
struct mdev_parent *parent;
@@ -217,6 +217,7 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
parent->dev = dev;
parent->ops = ops;
parent->bus = bus;
+ parent->dev_size = dev_size;
mutex_lock(&compat_list_lock);
class_compat = mdev_alloc_class_compat(bus);
@@ -339,7 +340,7 @@ int mdev_device_create(struct kobject *kobj,
}
}
- mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
+ mdev = kzalloc(parent->dev_size, GFP_KERNEL);
if (!mdev) {
mutex_unlock(&mdev_list_lock);
ret = -ENOMEM;
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 298d7a0f493a..012ab80719e9 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -20,6 +20,7 @@ struct mdev_parent {
struct list_head type_list;
/* Synchronize device creation/removal with parent unregistration */
struct rw_semaphore unreg_sem;
+ size_t dev_size;
};
struct mdev_device {
@@ -66,4 +67,8 @@ int mdev_device_create(struct kobject *kobj,
struct device *dev, const guid_t *uuid);
int mdev_device_remove(struct device *dev);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus, size_t dev_size);
+void mdev_unregister_device(struct device *dev);
+
#endif /* MDEV_PRIVATE_H */
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
index f9d1191b9982..44e116074f88 100644
--- a/drivers/vfio/mdev/mdev_vfio.c
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -4,6 +4,8 @@
#include <linux/device.h>
#include <linux/mdev_vfio.h>
+#include "mdev_private.h"
+
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "Jason Wang"
#define DRIVER_DESC "Mediated VFIO bus"
@@ -15,6 +17,31 @@ struct bus_type mdev_vfio_bus_type = {
};
EXPORT_SYMBOL(mdev_vfio_bus_type);
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
+
+struct mdev_vfio_device {
+ struct mdev_device mdev;
+ const struct mdev_vfio_ops *ops;
+};
+
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ mdev_vfio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_vfio_set_ops);
+
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ return mdev_vfio->ops;
+}
+EXPORT_SYMBOL(mdev_vfio_get_ops);
+
static int __init mdev_init(void)
{
return mdev_register_bus(&mdev_vfio_bus_type);
@@ -28,7 +55,8 @@ static void __exit mdev_exit(void)
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops)
{
- return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type,
+ sizeof(struct mdev_vfio_device));
}
EXPORT_SYMBOL(mdev_vfio_register_device);
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 16e9ebe30d4a..8b42a4b3f161 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -24,16 +24,16 @@
static int vfio_mdev_open(void *device_data)
{
struct mdev_device *mdev = device_data;
- struct mdev_parent *parent = mdev->parent;
+ const struct mdev_vfio_ops *ops = mdev_vfio_get_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 +43,10 @@ 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_ops *ops = mdev_vfio_get_ops(mdev);
- if (likely(parent->ops->release))
- parent->ops->release(mdev);
+ if (likely(ops->release))
+ ops->release(mdev);
module_put(THIS_MODULE);
}
@@ -55,47 +55,47 @@ 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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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 = {
@@ -110,7 +110,7 @@ static const struct vfio_device_ops vfio_mdev_dev_ops = {
static int vfio_mdev_probe(struct device *dev)
{
- struct mdev_device *mdev = to_mdev_device(dev);
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_vfio_bus_type);
return vfio_add_group_dev(dev, &vfio_mdev_dev_ops, mdev);
}
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index ee2410246b3c..25554e55bcee 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -48,30 +48,6 @@ struct device *mdev_get_iommu_device(struct device *dev);
* @mdev: mdev_device device structure which is being
* destroyed
* Returns integer: success (0) or error (< 0)
- * @open: Open mediated device.
- * @mdev: mediated device.
- * Returns integer: success (0) or error (< 0)
- * @release: release mediated device
- * @mdev: mediated device.
- * @read: Read emulation callback
- * @mdev: mediated device structure
- * @buf: read buffer
- * @count: number of bytes to read
- * @ppos: address.
- * Retuns number on bytes read on success or error.
- * @write: Write emulation callback
- * @mdev: mediated device structure
- * @buf: write buffer
- * @count: number of bytes to be written
- * @ppos: address.
- * Retuns number on bytes written on success or error.
- * @ioctl: IOCTL callback
- * @mdev: mediated device structure
- * @cmd: ioctl command
- * @arg: arguments to ioctl
- * @mmap: mmap callback
- * @mdev: mediated device structure
- * @vma: vma structure
* Parent device that support mediated device should be registered with mdev
* module with mdev_parent_ops structure.
**/
@@ -83,15 +59,6 @@ struct mdev_parent_ops {
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
- int (*open)(struct mdev_device *mdev);
- void (*release)(struct mdev_device *mdev);
- ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
- size_t count, loff_t *ppos);
- ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
- size_t count, loff_t *ppos);
- long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
- unsigned long arg);
- int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
/* interface for exporting mdev supported type attributes */
@@ -133,10 +100,6 @@ 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);
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus);
-void mdev_unregister_device(struct device *dev);
-
int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
index 446a7537e3fb..243ad85019a6 100644
--- a/include/linux/mdev_vfio.h
+++ b/include/linux/mdev_vfio.h
@@ -13,6 +13,45 @@
extern struct bus_type mdev_vfio_bus_type;
+/* VFIO mdev ops
+ *
+ * @open: Open mediated device.
+ * @mdev: mediated device.
+ * Returns integer: success (0) or error (< 0)
+ * @release: release mediated device
+ * @mdev: mediated device.
+ * @read: Read emulation callback
+ * @mdev: mediated device structure
+ * @buf: read buffer
+ * @count: number of bytes to read
+ * @ppos: address.
+ * Retuns number on bytes read on success or error.
+ * @write: Write emulation callback
+ * @mdev: mediated device structure
+ * @buf: write buffer
+ * @count: number of bytes to be written
+ * @ppos: address.
+ * Retuns number on bytes written on success or error.
+ * @ioctl: IOCTL callback
+ * @mdev: mediated device structure
+ * @cmd: ioctl command
+ * @arg: arguments to ioctl
+ * @mmap: mmap callback
+ * @mdev: mediated device structure
+ * @vma: vma structure
+ */
+struct mdev_vfio_ops {
+ int (*open)(struct mdev_device *mdev);
+ void (*release)(struct mdev_device *mdev);
+ ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
+ size_t count, loff_t *ppos);
+ ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
+ size_t count, loff_t *ppos);
+ long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
+ unsigned long arg);
+ int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
+};
+
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops);
void mdev_vfio_unregister_device(struct device *dev);
@@ -22,4 +61,8 @@ static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
return mdev_from_dev(dev, &mdev_vfio_bus_type);
}
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
+
#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index f041d58324b1..b2ba32b5fed2 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -516,6 +516,8 @@ static int mbochs_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,6 +563,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
mbochs_reset(mdev);
mbochs_used_mbytes += type->mbytes;
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
err_mem:
@@ -1418,12 +1421,7 @@ 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_ops mdev_ops = {
.open = mbochs_open,
.release = mbochs_close,
.read = mbochs_read,
@@ -1432,6 +1430,14 @@ 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,
};
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 9c32fe3795ad..d26fd94b4783 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -226,6 +226,8 @@ static int mdpy_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,6 +271,8 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
mdpy_reset(mdev);
mdpy_count++;
+
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -725,12 +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_ops mdev_ops = {
.open = mdpy_open,
.release = mdpy_close,
.read = mdpy_read,
@@ -739,6 +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,
};
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index 6e4e6339e0f1..3f0c6506199a 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -708,6 +708,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
return ret;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct mdev_state *mdev_state;
@@ -755,6 +757,7 @@ 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_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -1387,6 +1390,14 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops mdev_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,
@@ -1394,11 +1405,6 @@ 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)
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 2/6] mdev: split out VFIO bus specific parent ops
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
The only thing left for generalizing mdev is the VFIO specific parent
ops. This is basically the open/release/read/write/ioctl/mmap.
To support this, mdev core is extend to support a specific size
of structure during create, this will allow to compose mdev structure
into mdev vfio structure and place the VFIO specific callbacks there
like:
struct mdev_vfio {
struct mdev_device mdev;
const struct mdev_vfio_ops *ops;
};
Helpers for setting and getting the ops were introduced to support
mdev vfio device to set ops and vfio mdev driver to use the ops.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 34 +++++++++------
drivers/gpu/drm/i915/gvt/kvmgt.c | 16 ++++---
drivers/s390/cio/vfio_ccw_ops.c | 17 +++++---
drivers/s390/crypto/vfio_ap_ops.c | 13 ++++--
drivers/vfio/mdev/mdev_core.c | 5 ++-
drivers/vfio/mdev/mdev_private.h | 5 +++
drivers/vfio/mdev/mdev_vfio.c | 30 ++++++++++++-
drivers/vfio/mdev/vfio_mdev.c | 38 ++++++++--------
include/linux/mdev.h | 37 ----------------
include/linux/mdev_vfio.h | 43 +++++++++++++++++++
samples/vfio-mdev/mbochs.c | 18 +++++---
samples/vfio-mdev/mdpy.c | 19 +++++---
samples/vfio-mdev/mtty.c | 16 ++++---
13 files changed, 189 insertions(+), 102 deletions(-)
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 1887d27a565e..9045584e4ea3 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -153,26 +153,36 @@ 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:
+A driver should use the mdev_parent_ops structure in the function call
+to register itself with the mdev core driver::
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
+ extern int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+
+However, the mdev_parent_ops structure is not required in the function call
+that a driver should use to unregister itself with the mdev core driver::
+
+ extern void mdev_vfio_unregister_device(struct device *dev);
+
+The VFIO specific callbacks is abstracted in mdev_vfio_ops structure
+are as follows:
+
+* 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
-A driver should use the mdev_parent_ops structure and bus type in the
-function call to register itself with the mdev core driver::
+During the creation of VFIO mediated device, mdev_vfio_ops need to be
+specified::
- extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops,
- struct bus_type *bus);
+ void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
-However, the mdev_parent_ops structure is not required in the function call
-that a driver should use to unregister itself with the mdev core driver::
+Those callbacks could be fetched by drivers through::
- extern void mdev_unregister_device(struct device *dev);
+ const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
Mediated Device Management Interface Through sysfs
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index afdb3de5ce2f..e72c36174035 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -643,6 +643,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
}
+static const struct mdev_vfio_ops intel_mdev_vfio_ops;
+
static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct intel_vgpu *vgpu = NULL;
@@ -678,6 +680,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
dev_name(mdev_dev(mdev)));
ret = 0;
+ mdev_vfio_set_ops(mdev, &intel_mdev_vfio_ops);
out:
return ret;
}
@@ -1581,20 +1584,21 @@ 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_ops intel_mdev_vfio_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;
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index 791b8b0eb027..811f0a3b1903 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -110,6 +110,8 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops vfio_ccw_mdev_vfio_ops;
+
static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct vfio_ccw_private *private =
@@ -129,6 +131,8 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
private->sch->schid.ssid,
private->sch->schid.sch_no);
+ mdev_vfio_set_ops(mdev, &vfio_ccw_mdev_vfio_ops);
+
return 0;
}
@@ -574,16 +578,19 @@ 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_ops vfio_ccw_mdev_vfio_ops = {
.open = vfio_ccw_mdev_open,
.release = vfio_ccw_mdev_release,
.read = vfio_ccw_mdev_read,
.write = vfio_ccw_mdev_write,
.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)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 78048e670374..0649c68287d7 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -321,6 +321,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
matrix->adm_max = info->apxa ? info->Nd : 15;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_ops;
+
static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct ap_matrix_mdev *matrix_mdev;
@@ -343,6 +345,8 @@ 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_vfio_set_ops(mdev, &vfio_ap_matrix_mdev_ops);
+
return 0;
}
@@ -1280,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
return ret;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_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)
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index e1272a40c521..c6bc67bf63fa 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -179,7 +179,7 @@ static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
* Returns a negative value on error, otherwise 0.
*/
int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus)
+ struct bus_type *bus, size_t dev_size)
{
int ret;
struct mdev_parent *parent;
@@ -217,6 +217,7 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
parent->dev = dev;
parent->ops = ops;
parent->bus = bus;
+ parent->dev_size = dev_size;
mutex_lock(&compat_list_lock);
class_compat = mdev_alloc_class_compat(bus);
@@ -339,7 +340,7 @@ int mdev_device_create(struct kobject *kobj,
}
}
- mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
+ mdev = kzalloc(parent->dev_size, GFP_KERNEL);
if (!mdev) {
mutex_unlock(&mdev_list_lock);
ret = -ENOMEM;
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 298d7a0f493a..012ab80719e9 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -20,6 +20,7 @@ struct mdev_parent {
struct list_head type_list;
/* Synchronize device creation/removal with parent unregistration */
struct rw_semaphore unreg_sem;
+ size_t dev_size;
};
struct mdev_device {
@@ -66,4 +67,8 @@ int mdev_device_create(struct kobject *kobj,
struct device *dev, const guid_t *uuid);
int mdev_device_remove(struct device *dev);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus, size_t dev_size);
+void mdev_unregister_device(struct device *dev);
+
#endif /* MDEV_PRIVATE_H */
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
index f9d1191b9982..44e116074f88 100644
--- a/drivers/vfio/mdev/mdev_vfio.c
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -4,6 +4,8 @@
#include <linux/device.h>
#include <linux/mdev_vfio.h>
+#include "mdev_private.h"
+
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "Jason Wang"
#define DRIVER_DESC "Mediated VFIO bus"
@@ -15,6 +17,31 @@ struct bus_type mdev_vfio_bus_type = {
};
EXPORT_SYMBOL(mdev_vfio_bus_type);
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
+
+struct mdev_vfio_device {
+ struct mdev_device mdev;
+ const struct mdev_vfio_ops *ops;
+};
+
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ mdev_vfio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_vfio_set_ops);
+
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ return mdev_vfio->ops;
+}
+EXPORT_SYMBOL(mdev_vfio_get_ops);
+
static int __init mdev_init(void)
{
return mdev_register_bus(&mdev_vfio_bus_type);
@@ -28,7 +55,8 @@ static void __exit mdev_exit(void)
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops)
{
- return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type,
+ sizeof(struct mdev_vfio_device));
}
EXPORT_SYMBOL(mdev_vfio_register_device);
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 16e9ebe30d4a..8b42a4b3f161 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -24,16 +24,16 @@
static int vfio_mdev_open(void *device_data)
{
struct mdev_device *mdev = device_data;
- struct mdev_parent *parent = mdev->parent;
+ const struct mdev_vfio_ops *ops = mdev_vfio_get_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 +43,10 @@ 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_ops *ops = mdev_vfio_get_ops(mdev);
- if (likely(parent->ops->release))
- parent->ops->release(mdev);
+ if (likely(ops->release))
+ ops->release(mdev);
module_put(THIS_MODULE);
}
@@ -55,47 +55,47 @@ 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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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 = {
@@ -110,7 +110,7 @@ static const struct vfio_device_ops vfio_mdev_dev_ops = {
static int vfio_mdev_probe(struct device *dev)
{
- struct mdev_device *mdev = to_mdev_device(dev);
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_vfio_bus_type);
return vfio_add_group_dev(dev, &vfio_mdev_dev_ops, mdev);
}
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index ee2410246b3c..25554e55bcee 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -48,30 +48,6 @@ struct device *mdev_get_iommu_device(struct device *dev);
* @mdev: mdev_device device structure which is being
* destroyed
* Returns integer: success (0) or error (< 0)
- * @open: Open mediated device.
- * @mdev: mediated device.
- * Returns integer: success (0) or error (< 0)
- * @release: release mediated device
- * @mdev: mediated device.
- * @read: Read emulation callback
- * @mdev: mediated device structure
- * @buf: read buffer
- * @count: number of bytes to read
- * @ppos: address.
- * Retuns number on bytes read on success or error.
- * @write: Write emulation callback
- * @mdev: mediated device structure
- * @buf: write buffer
- * @count: number of bytes to be written
- * @ppos: address.
- * Retuns number on bytes written on success or error.
- * @ioctl: IOCTL callback
- * @mdev: mediated device structure
- * @cmd: ioctl command
- * @arg: arguments to ioctl
- * @mmap: mmap callback
- * @mdev: mediated device structure
- * @vma: vma structure
* Parent device that support mediated device should be registered with mdev
* module with mdev_parent_ops structure.
**/
@@ -83,15 +59,6 @@ struct mdev_parent_ops {
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
- int (*open)(struct mdev_device *mdev);
- void (*release)(struct mdev_device *mdev);
- ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
- size_t count, loff_t *ppos);
- ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
- size_t count, loff_t *ppos);
- long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
- unsigned long arg);
- int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
/* interface for exporting mdev supported type attributes */
@@ -133,10 +100,6 @@ 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);
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus);
-void mdev_unregister_device(struct device *dev);
-
int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
index 446a7537e3fb..243ad85019a6 100644
--- a/include/linux/mdev_vfio.h
+++ b/include/linux/mdev_vfio.h
@@ -13,6 +13,45 @@
extern struct bus_type mdev_vfio_bus_type;
+/* VFIO mdev ops
+ *
+ * @open: Open mediated device.
+ * @mdev: mediated device.
+ * Returns integer: success (0) or error (< 0)
+ * @release: release mediated device
+ * @mdev: mediated device.
+ * @read: Read emulation callback
+ * @mdev: mediated device structure
+ * @buf: read buffer
+ * @count: number of bytes to read
+ * @ppos: address.
+ * Retuns number on bytes read on success or error.
+ * @write: Write emulation callback
+ * @mdev: mediated device structure
+ * @buf: write buffer
+ * @count: number of bytes to be written
+ * @ppos: address.
+ * Retuns number on bytes written on success or error.
+ * @ioctl: IOCTL callback
+ * @mdev: mediated device structure
+ * @cmd: ioctl command
+ * @arg: arguments to ioctl
+ * @mmap: mmap callback
+ * @mdev: mediated device structure
+ * @vma: vma structure
+ */
+struct mdev_vfio_ops {
+ int (*open)(struct mdev_device *mdev);
+ void (*release)(struct mdev_device *mdev);
+ ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
+ size_t count, loff_t *ppos);
+ ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
+ size_t count, loff_t *ppos);
+ long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
+ unsigned long arg);
+ int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
+};
+
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops);
void mdev_vfio_unregister_device(struct device *dev);
@@ -22,4 +61,8 @@ static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
return mdev_from_dev(dev, &mdev_vfio_bus_type);
}
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
+
#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index f041d58324b1..b2ba32b5fed2 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -516,6 +516,8 @@ static int mbochs_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,6 +563,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
mbochs_reset(mdev);
mbochs_used_mbytes += type->mbytes;
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
err_mem:
@@ -1418,12 +1421,7 @@ 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_ops mdev_ops = {
.open = mbochs_open,
.release = mbochs_close,
.read = mbochs_read,
@@ -1432,6 +1430,14 @@ 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,
};
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 9c32fe3795ad..d26fd94b4783 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -226,6 +226,8 @@ static int mdpy_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,6 +271,8 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
mdpy_reset(mdev);
mdpy_count++;
+
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -725,12 +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_ops mdev_ops = {
.open = mdpy_open,
.release = mdpy_close,
.read = mdpy_read,
@@ -739,6 +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,
};
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index 6e4e6339e0f1..3f0c6506199a 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -708,6 +708,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
return ret;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct mdev_state *mdev_state;
@@ -755,6 +757,7 @@ 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_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -1387,6 +1390,14 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops mdev_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,
@@ -1394,11 +1405,6 @@ 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)
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 2/6] mdev: split out VFIO bus specific parent ops
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
The only thing left for generalizing mdev is the VFIO specific parent
ops. This is basically the open/release/read/write/ioctl/mmap.
To support this, mdev core is extend to support a specific size
of structure during create, this will allow to compose mdev structure
into mdev vfio structure and place the VFIO specific callbacks there
like:
struct mdev_vfio {
struct mdev_device mdev;
const struct mdev_vfio_ops *ops;
};
Helpers for setting and getting the ops were introduced to support
mdev vfio device to set ops and vfio mdev driver to use the ops.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
.../driver-api/vfio-mediated-device.rst | 34 +++++++++------
drivers/gpu/drm/i915/gvt/kvmgt.c | 16 ++++---
drivers/s390/cio/vfio_ccw_ops.c | 17 +++++---
drivers/s390/crypto/vfio_ap_ops.c | 13 ++++--
drivers/vfio/mdev/mdev_core.c | 5 ++-
drivers/vfio/mdev/mdev_private.h | 5 +++
drivers/vfio/mdev/mdev_vfio.c | 30 ++++++++++++-
drivers/vfio/mdev/vfio_mdev.c | 38 ++++++++--------
include/linux/mdev.h | 37 ----------------
include/linux/mdev_vfio.h | 43 +++++++++++++++++++
samples/vfio-mdev/mbochs.c | 18 +++++---
samples/vfio-mdev/mdpy.c | 19 +++++---
samples/vfio-mdev/mtty.c | 16 ++++---
13 files changed, 189 insertions(+), 102 deletions(-)
diff --git a/Documentation/driver-api/vfio-mediated-device.rst b/Documentation/driver-api/vfio-mediated-device.rst
index 1887d27a565e..9045584e4ea3 100644
--- a/Documentation/driver-api/vfio-mediated-device.rst
+++ b/Documentation/driver-api/vfio-mediated-device.rst
@@ -153,26 +153,36 @@ 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:
+A driver should use the mdev_parent_ops structure in the function call
+to register itself with the mdev core driver::
-* open: open callback of mediated device
-* close: close callback of mediated device
-* ioctl: ioctl callback of mediated device
+ extern int mdev_vfio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+
+However, the mdev_parent_ops structure is not required in the function call
+that a driver should use to unregister itself with the mdev core driver::
+
+ extern void mdev_vfio_unregister_device(struct device *dev);
+
+The VFIO specific callbacks is abstracted in mdev_vfio_ops structure
+are as follows:
+
+* 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
-A driver should use the mdev_parent_ops structure and bus type in the
-function call to register itself with the mdev core driver::
+During the creation of VFIO mediated device, mdev_vfio_ops need to be
+specified::
- extern int mdev_register_device(struct device *dev,
- const struct mdev_parent_ops *ops,
- struct bus_type *bus);
+ void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
-However, the mdev_parent_ops structure is not required in the function call
-that a driver should use to unregister itself with the mdev core driver::
+Those callbacks could be fetched by drivers through::
- extern void mdev_unregister_device(struct device *dev);
+ const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
Mediated Device Management Interface Through sysfs
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index afdb3de5ce2f..e72c36174035 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -643,6 +643,8 @@ static void kvmgt_put_vfio_device(void *vgpu)
vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
}
+static const struct mdev_vfio_ops intel_mdev_vfio_ops;
+
static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct intel_vgpu *vgpu = NULL;
@@ -678,6 +680,7 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
dev_name(mdev_dev(mdev)));
ret = 0;
+ mdev_vfio_set_ops(mdev, &intel_mdev_vfio_ops);
out:
return ret;
}
@@ -1581,20 +1584,21 @@ 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_ops intel_mdev_vfio_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;
diff --git a/drivers/s390/cio/vfio_ccw_ops.c b/drivers/s390/cio/vfio_ccw_ops.c
index 791b8b0eb027..811f0a3b1903 100644
--- a/drivers/s390/cio/vfio_ccw_ops.c
+++ b/drivers/s390/cio/vfio_ccw_ops.c
@@ -110,6 +110,8 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops vfio_ccw_mdev_vfio_ops;
+
static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct vfio_ccw_private *private =
@@ -129,6 +131,8 @@ static int vfio_ccw_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
private->sch->schid.ssid,
private->sch->schid.sch_no);
+ mdev_vfio_set_ops(mdev, &vfio_ccw_mdev_vfio_ops);
+
return 0;
}
@@ -574,16 +578,19 @@ 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_ops vfio_ccw_mdev_vfio_ops = {
.open = vfio_ccw_mdev_open,
.release = vfio_ccw_mdev_release,
.read = vfio_ccw_mdev_read,
.write = vfio_ccw_mdev_write,
.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)
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 78048e670374..0649c68287d7 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -321,6 +321,8 @@ static void vfio_ap_matrix_init(struct ap_config_info *info,
matrix->adm_max = info->apxa ? info->Nd : 15;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_ops;
+
static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct ap_matrix_mdev *matrix_mdev;
@@ -343,6 +345,8 @@ 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_vfio_set_ops(mdev, &vfio_ap_matrix_mdev_ops);
+
return 0;
}
@@ -1280,15 +1284,18 @@ static ssize_t vfio_ap_mdev_ioctl(struct mdev_device *mdev,
return ret;
}
+static const struct mdev_vfio_ops vfio_ap_matrix_mdev_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)
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
index e1272a40c521..c6bc67bf63fa 100644
--- a/drivers/vfio/mdev/mdev_core.c
+++ b/drivers/vfio/mdev/mdev_core.c
@@ -179,7 +179,7 @@ static struct class_compat *mdev_alloc_class_compat(struct bus_type *bus)
* Returns a negative value on error, otherwise 0.
*/
int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus)
+ struct bus_type *bus, size_t dev_size)
{
int ret;
struct mdev_parent *parent;
@@ -217,6 +217,7 @@ int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
parent->dev = dev;
parent->ops = ops;
parent->bus = bus;
+ parent->dev_size = dev_size;
mutex_lock(&compat_list_lock);
class_compat = mdev_alloc_class_compat(bus);
@@ -339,7 +340,7 @@ int mdev_device_create(struct kobject *kobj,
}
}
- mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
+ mdev = kzalloc(parent->dev_size, GFP_KERNEL);
if (!mdev) {
mutex_unlock(&mdev_list_lock);
ret = -ENOMEM;
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/vfio/mdev/mdev_private.h
index 298d7a0f493a..012ab80719e9 100644
--- a/drivers/vfio/mdev/mdev_private.h
+++ b/drivers/vfio/mdev/mdev_private.h
@@ -20,6 +20,7 @@ struct mdev_parent {
struct list_head type_list;
/* Synchronize device creation/removal with parent unregistration */
struct rw_semaphore unreg_sem;
+ size_t dev_size;
};
struct mdev_device {
@@ -66,4 +67,8 @@ int mdev_device_create(struct kobject *kobj,
struct device *dev, const guid_t *uuid);
int mdev_device_remove(struct device *dev);
+int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
+ struct bus_type *bus, size_t dev_size);
+void mdev_unregister_device(struct device *dev);
+
#endif /* MDEV_PRIVATE_H */
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/vfio/mdev/mdev_vfio.c
index f9d1191b9982..44e116074f88 100644
--- a/drivers/vfio/mdev/mdev_vfio.c
+++ b/drivers/vfio/mdev/mdev_vfio.c
@@ -4,6 +4,8 @@
#include <linux/device.h>
#include <linux/mdev_vfio.h>
+#include "mdev_private.h"
+
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "Jason Wang"
#define DRIVER_DESC "Mediated VFIO bus"
@@ -15,6 +17,31 @@ struct bus_type mdev_vfio_bus_type = {
};
EXPORT_SYMBOL(mdev_vfio_bus_type);
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
+
+struct mdev_vfio_device {
+ struct mdev_device mdev;
+ const struct mdev_vfio_ops *ops;
+};
+
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ mdev_vfio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_vfio_set_ops);
+
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_vfio_device *mdev_vfio = to_vfio_mdev_device(mdev);
+
+ return mdev_vfio->ops;
+}
+EXPORT_SYMBOL(mdev_vfio_get_ops);
+
static int __init mdev_init(void)
{
return mdev_register_bus(&mdev_vfio_bus_type);
@@ -28,7 +55,8 @@ static void __exit mdev_exit(void)
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops)
{
- return mdev_register_device(dev, ops, &mdev_vfio_bus_type);
+ return mdev_register_device(dev, ops, &mdev_vfio_bus_type,
+ sizeof(struct mdev_vfio_device));
}
EXPORT_SYMBOL(mdev_vfio_register_device);
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 16e9ebe30d4a..8b42a4b3f161 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -24,16 +24,16 @@
static int vfio_mdev_open(void *device_data)
{
struct mdev_device *mdev = device_data;
- struct mdev_parent *parent = mdev->parent;
+ const struct mdev_vfio_ops *ops = mdev_vfio_get_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 +43,10 @@ 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_ops *ops = mdev_vfio_get_ops(mdev);
- if (likely(parent->ops->release))
- parent->ops->release(mdev);
+ if (likely(ops->release))
+ ops->release(mdev);
module_put(THIS_MODULE);
}
@@ -55,47 +55,47 @@ 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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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_ops *ops = mdev_vfio_get_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 = {
@@ -110,7 +110,7 @@ static const struct vfio_device_ops vfio_mdev_dev_ops = {
static int vfio_mdev_probe(struct device *dev)
{
- struct mdev_device *mdev = to_mdev_device(dev);
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_vfio_bus_type);
return vfio_add_group_dev(dev, &vfio_mdev_dev_ops, mdev);
}
diff --git a/include/linux/mdev.h b/include/linux/mdev.h
index ee2410246b3c..25554e55bcee 100644
--- a/include/linux/mdev.h
+++ b/include/linux/mdev.h
@@ -48,30 +48,6 @@ struct device *mdev_get_iommu_device(struct device *dev);
* @mdev: mdev_device device structure which is being
* destroyed
* Returns integer: success (0) or error (< 0)
- * @open: Open mediated device.
- * @mdev: mediated device.
- * Returns integer: success (0) or error (< 0)
- * @release: release mediated device
- * @mdev: mediated device.
- * @read: Read emulation callback
- * @mdev: mediated device structure
- * @buf: read buffer
- * @count: number of bytes to read
- * @ppos: address.
- * Retuns number on bytes read on success or error.
- * @write: Write emulation callback
- * @mdev: mediated device structure
- * @buf: write buffer
- * @count: number of bytes to be written
- * @ppos: address.
- * Retuns number on bytes written on success or error.
- * @ioctl: IOCTL callback
- * @mdev: mediated device structure
- * @cmd: ioctl command
- * @arg: arguments to ioctl
- * @mmap: mmap callback
- * @mdev: mediated device structure
- * @vma: vma structure
* Parent device that support mediated device should be registered with mdev
* module with mdev_parent_ops structure.
**/
@@ -83,15 +59,6 @@ struct mdev_parent_ops {
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
- int (*open)(struct mdev_device *mdev);
- void (*release)(struct mdev_device *mdev);
- ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
- size_t count, loff_t *ppos);
- ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
- size_t count, loff_t *ppos);
- long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
- unsigned long arg);
- int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
/* interface for exporting mdev supported type attributes */
@@ -133,10 +100,6 @@ 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);
-int mdev_register_device(struct device *dev, const struct mdev_parent_ops *ops,
- struct bus_type *bus);
-void mdev_unregister_device(struct device *dev);
-
int mdev_register_driver(struct mdev_driver *drv, struct module *owner,
struct bus_type *bus);
void mdev_unregister_driver(struct mdev_driver *drv);
diff --git a/include/linux/mdev_vfio.h b/include/linux/mdev_vfio.h
index 446a7537e3fb..243ad85019a6 100644
--- a/include/linux/mdev_vfio.h
+++ b/include/linux/mdev_vfio.h
@@ -13,6 +13,45 @@
extern struct bus_type mdev_vfio_bus_type;
+/* VFIO mdev ops
+ *
+ * @open: Open mediated device.
+ * @mdev: mediated device.
+ * Returns integer: success (0) or error (< 0)
+ * @release: release mediated device
+ * @mdev: mediated device.
+ * @read: Read emulation callback
+ * @mdev: mediated device structure
+ * @buf: read buffer
+ * @count: number of bytes to read
+ * @ppos: address.
+ * Retuns number on bytes read on success or error.
+ * @write: Write emulation callback
+ * @mdev: mediated device structure
+ * @buf: write buffer
+ * @count: number of bytes to be written
+ * @ppos: address.
+ * Retuns number on bytes written on success or error.
+ * @ioctl: IOCTL callback
+ * @mdev: mediated device structure
+ * @cmd: ioctl command
+ * @arg: arguments to ioctl
+ * @mmap: mmap callback
+ * @mdev: mediated device structure
+ * @vma: vma structure
+ */
+struct mdev_vfio_ops {
+ int (*open)(struct mdev_device *mdev);
+ void (*release)(struct mdev_device *mdev);
+ ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
+ size_t count, loff_t *ppos);
+ ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
+ size_t count, loff_t *ppos);
+ long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
+ unsigned long arg);
+ int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
+};
+
int mdev_vfio_register_device(struct device *dev,
const struct mdev_parent_ops *ops);
void mdev_vfio_unregister_device(struct device *dev);
@@ -22,4 +61,8 @@ static inline struct mdev_device *vfio_mdev_from_dev(struct device *dev)
return mdev_from_dev(dev, &mdev_vfio_bus_type);
}
+void mdev_vfio_set_ops(struct mdev_device *mdev,
+ const struct mdev_vfio_ops *ops);
+const struct mdev_vfio_ops *mdev_vfio_get_ops(struct mdev_device *mdev);
+
#endif
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
index f041d58324b1..b2ba32b5fed2 100644
--- a/samples/vfio-mdev/mbochs.c
+++ b/samples/vfio-mdev/mbochs.c
@@ -516,6 +516,8 @@ static int mbochs_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mbochs_type *type = mbochs_find_type(kobj);
@@ -561,6 +563,7 @@ static int mbochs_create(struct kobject *kobj, struct mdev_device *mdev)
mbochs_reset(mdev);
mbochs_used_mbytes += type->mbytes;
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
err_mem:
@@ -1418,12 +1421,7 @@ 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_ops mdev_ops = {
.open = mbochs_open,
.release = mbochs_close,
.read = mbochs_read,
@@ -1432,6 +1430,14 @@ 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,
};
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
index 9c32fe3795ad..d26fd94b4783 100644
--- a/samples/vfio-mdev/mdpy.c
+++ b/samples/vfio-mdev/mdpy.c
@@ -226,6 +226,8 @@ static int mdpy_reset(struct mdev_device *mdev)
return 0;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
{
const struct mdpy_type *type = mdpy_find_type(kobj);
@@ -269,6 +271,8 @@ static int mdpy_create(struct kobject *kobj, struct mdev_device *mdev)
mdpy_reset(mdev);
mdpy_count++;
+
+ mdev_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -725,12 +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_ops mdev_ops = {
.open = mdpy_open,
.release = mdpy_close,
.read = mdpy_read,
@@ -739,6 +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,
};
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
index 6e4e6339e0f1..3f0c6506199a 100644
--- a/samples/vfio-mdev/mtty.c
+++ b/samples/vfio-mdev/mtty.c
@@ -708,6 +708,8 @@ static ssize_t mdev_access(struct mdev_device *mdev, u8 *buf, size_t count,
return ret;
}
+static const struct mdev_vfio_ops mdev_ops;
+
static int mtty_create(struct kobject *kobj, struct mdev_device *mdev)
{
struct mdev_state *mdev_state;
@@ -755,6 +757,7 @@ 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_vfio_set_ops(mdev, &mdev_ops);
return 0;
}
@@ -1387,6 +1390,14 @@ static struct attribute_group *mdev_type_groups[] = {
NULL,
};
+static const struct mdev_vfio_ops mdev_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,
@@ -1394,11 +1405,6 @@ 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)
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 3/6] mdev: move to drivers/
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, Jason Wang
Mdev now is nothing VFIO specific, let's move it to upper
directory.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 7 +++++--
drivers/Kconfig | 2 ++
drivers/Makefile | 1 +
drivers/mdev/Kconfig | 19 ++++++++++++++++++
drivers/mdev/Makefile | 5 +++++
drivers/{vfio => }/mdev/mdev_core.c | 0
drivers/{vfio => }/mdev/mdev_driver.c | 0
drivers/{vfio => }/mdev/mdev_private.h | 0
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
.../{vfio/mdev/mdev_vfio.c => mdev/vfio.c} | 0
drivers/vfio/mdev/Kconfig | 20 -------------------
drivers/vfio/mdev/Makefile | 4 ----
drivers/vfio/mdev/vfio_mdev.c | 2 --
13 files changed, 32 insertions(+), 28 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (100%)
rename drivers/{vfio => }/mdev/mdev_driver.c (100%)
rename drivers/{vfio => }/mdev/mdev_private.h (100%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
rename drivers/{vfio/mdev/mdev_vfio.c => mdev/vfio.c} (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 6d590afb62c3..5d7e8badf58c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17129,15 +17129,18 @@ T: git git://github.com/awilliam/linux-vfio.git
S: Maintained
F: Documentation/driver-api/vfio.rst
F: drivers/vfio/
+F: drivers/mdev/vfio.c
F: include/linux/vfio.h
F: include/uapi/linux/vfio.h
-VFIO MEDIATED DEVICE DRIVERS
+MEDIATED DEVICE DRIVERS
+M: Alex Williamson <alex.williamson@redhat.com>
M: Kirti Wankhede <kwankhede@nvidia.com>
+R: Cornelia Huck <cohuck@redhat.com>
L: kvm@vger.kernel.org
S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
-F: drivers/vfio/mdev/
+F: drivers/mdev
F: include/linux/mdev.h
F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 8befa53f43be..3e2839048fe6 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -228,4 +228,6 @@ source "drivers/interconnect/Kconfig"
source "drivers/counter/Kconfig"
+source "drivers/mdev/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index aaef17cc6512..592e23f2e629 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -186,3 +186,4 @@ obj-$(CONFIG_SIOX) += siox/
obj-$(CONFIG_GNSS) += gnss/
obj-$(CONFIG_INTERCONNECT) += interconnect/
obj-$(CONFIG_COUNTER) += counter/
+obj-$(CONFIG_MDEV) += mdev/
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
new file mode 100644
index 000000000000..4561f2d4178f
--- /dev/null
+++ b/drivers/mdev/Kconfig
@@ -0,0 +1,19 @@
+
+config MDEV
+ tristate "Mediated device driver framework"
+ default n
+ help
+ Provides a framework to virtualize devices.
+
+ If you don't know what do here, say N.
+
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
new file mode 100644
index 000000000000..0b749e7f8ff4
--- /dev/null
+++ b/drivers/mdev/Makefile
@@ -0,0 +1,5 @@
+
+mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
+mdev_vfio-y := vfio.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/mdev/mdev_core.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_core.c
rename to drivers/mdev/mdev_core.c
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/mdev/mdev_driver.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_driver.c
rename to drivers/mdev/mdev_driver.c
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/mdev/mdev_private.h
similarity index 100%
rename from drivers/vfio/mdev/mdev_private.h
rename to drivers/mdev/mdev_private.h
diff --git a/drivers/vfio/mdev/mdev_sysfs.c b/drivers/mdev/mdev_sysfs.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_sysfs.c
rename to drivers/mdev/mdev_sysfs.c
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/mdev/vfio.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_vfio.c
rename to drivers/mdev/vfio.c
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 2e07ca915a96..9a9234c3e00e 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,24 +1,4 @@
-config MDEV
- tristate "Mediated device driver framework"
- default n
- help
- Provides a framework to virtualize devices.
-
- If you don't know what do here, say N.
-
-config VFIO_MDEV
- tristate "VFIO Mediated device driver"
- depends on VFIO && MDEV
- default n
- help
- Proivdes a mediated BUS for userspace driver through VFIO
- framework. See Documentation/vfio-mediated-device.txt for
- more details.
-
- If you don't know what do here, say N.
-
-
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index e9675501271a..e2a92df3089e 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,2 @@
-mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-
-obj-$(CONFIG_MDEV) += mdev.o
-obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 8b42a4b3f161..4eada31e9287 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -15,8 +15,6 @@
#include <linux/vfio.h>
#include <linux/mdev_vfio.h>
-#include "mdev_private.h"
-
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "NVIDIA Corporation"
#define DRIVER_DESC "VFIO based driver for Mediated device"
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam
Mdev now is nothing VFIO specific, let's move it to upper
directory.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 7 +++++--
drivers/Kconfig | 2 ++
drivers/Makefile | 1 +
drivers/mdev/Kconfig | 19 ++++++++++++++++++
drivers/mdev/Makefile | 5 +++++
drivers/{vfio => }/mdev/mdev_core.c | 0
drivers/{vfio => }/mdev/mdev_driver.c | 0
drivers/{vfio => }/mdev/mdev_private.h | 0
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
.../{vfio/mdev/mdev_vfio.c => mdev/vfio.c} | 0
drivers/vfio/mdev/Kconfig | 20 -------------------
drivers/vfio/mdev/Makefile | 4 ----
drivers/vfio/mdev/vfio_mdev.c | 2 --
13 files changed, 32 insertions(+), 28 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (100%)
rename drivers/{vfio => }/mdev/mdev_driver.c (100%)
rename drivers/{vfio => }/mdev/mdev_private.h (100%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
rename drivers/{vfio/mdev/mdev_vfio.c => mdev/vfio.c} (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 6d590afb62c3..5d7e8badf58c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17129,15 +17129,18 @@ T: git git://github.com/awilliam/linux-vfio.git
S: Maintained
F: Documentation/driver-api/vfio.rst
F: drivers/vfio/
+F: drivers/mdev/vfio.c
F: include/linux/vfio.h
F: include/uapi/linux/vfio.h
-VFIO MEDIATED DEVICE DRIVERS
+MEDIATED DEVICE DRIVERS
+M: Alex Williamson <alex.williamson@redhat.com>
M: Kirti Wankhede <kwankhede@nvidia.com>
+R: Cornelia Huck <cohuck@redhat.com>
L: kvm@vger.kernel.org
S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
-F: drivers/vfio/mdev/
+F: drivers/mdev
F: include/linux/mdev.h
F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 8befa53f43be..3e2839048fe6 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -228,4 +228,6 @@ source "drivers/interconnect/Kconfig"
source "drivers/counter/Kconfig"
+source "drivers/mdev/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index aaef17cc6512..592e23f2e629 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -186,3 +186,4 @@ obj-$(CONFIG_SIOX) += siox/
obj-$(CONFIG_GNSS) += gnss/
obj-$(CONFIG_INTERCONNECT) += interconnect/
obj-$(CONFIG_COUNTER) += counter/
+obj-$(CONFIG_MDEV) += mdev/
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
new file mode 100644
index 000000000000..4561f2d4178f
--- /dev/null
+++ b/drivers/mdev/Kconfig
@@ -0,0 +1,19 @@
+
+config MDEV
+ tristate "Mediated device driver framework"
+ default n
+ help
+ Provides a framework to virtualize devices.
+
+ If you don't know what do here, say N.
+
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
new file mode 100644
index 000000000000..0b749e7f8ff4
--- /dev/null
+++ b/drivers/mdev/Makefile
@@ -0,0 +1,5 @@
+
+mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
+mdev_vfio-y := vfio.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/mdev/mdev_core.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_core.c
rename to drivers/mdev/mdev_core.c
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/mdev/mdev_driver.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_driver.c
rename to drivers/mdev/mdev_driver.c
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/mdev/mdev_private.h
similarity index 100%
rename from drivers/vfio/mdev/mdev_private.h
rename to drivers/mdev/mdev_private.h
diff --git a/drivers/vfio/mdev/mdev_sysfs.c b/drivers/mdev/mdev_sysfs.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_sysfs.c
rename to drivers/mdev/mdev_sysfs.c
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/mdev/vfio.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_vfio.c
rename to drivers/mdev/vfio.c
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 2e07ca915a96..9a9234c3e00e 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,24 +1,4 @@
-config MDEV
- tristate "Mediated device driver framework"
- default n
- help
- Provides a framework to virtualize devices.
-
- If you don't know what do here, say N.
-
-config VFIO_MDEV
- tristate "VFIO Mediated device driver"
- depends on VFIO && MDEV
- default n
- help
- Proivdes a mediated BUS for userspace driver through VFIO
- framework. See Documentation/vfio-mediated-device.txt for
- more details.
-
- If you don't know what do here, say N.
-
-
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index e9675501271a..e2a92df3089e 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,2 @@
-mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-
-obj-$(CONFIG_MDEV) += mdev.o
-obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 8b42a4b3f161..4eada31e9287 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -15,8 +15,6 @@
#include <linux/vfio.h>
#include <linux/mdev_vfio.h>
-#include "mdev_private.h"
-
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "NVIDIA Corporation"
#define DRIVER_DESC "VFIO based driver for Mediated device"
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
Mdev now is nothing VFIO specific, let's move it to upper
directory.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 7 +++++--
drivers/Kconfig | 2 ++
drivers/Makefile | 1 +
drivers/mdev/Kconfig | 19 ++++++++++++++++++
drivers/mdev/Makefile | 5 +++++
drivers/{vfio => }/mdev/mdev_core.c | 0
drivers/{vfio => }/mdev/mdev_driver.c | 0
drivers/{vfio => }/mdev/mdev_private.h | 0
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
.../{vfio/mdev/mdev_vfio.c => mdev/vfio.c} | 0
drivers/vfio/mdev/Kconfig | 20 -------------------
drivers/vfio/mdev/Makefile | 4 ----
drivers/vfio/mdev/vfio_mdev.c | 2 --
13 files changed, 32 insertions(+), 28 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (100%)
rename drivers/{vfio => }/mdev/mdev_driver.c (100%)
rename drivers/{vfio => }/mdev/mdev_private.h (100%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
rename drivers/{vfio/mdev/mdev_vfio.c => mdev/vfio.c} (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 6d590afb62c3..5d7e8badf58c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17129,15 +17129,18 @@ T: git git://github.com/awilliam/linux-vfio.git
S: Maintained
F: Documentation/driver-api/vfio.rst
F: drivers/vfio/
+F: drivers/mdev/vfio.c
F: include/linux/vfio.h
F: include/uapi/linux/vfio.h
-VFIO MEDIATED DEVICE DRIVERS
+MEDIATED DEVICE DRIVERS
+M: Alex Williamson <alex.williamson@redhat.com>
M: Kirti Wankhede <kwankhede@nvidia.com>
+R: Cornelia Huck <cohuck@redhat.com>
L: kvm@vger.kernel.org
S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
-F: drivers/vfio/mdev/
+F: drivers/mdev
F: include/linux/mdev.h
F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 8befa53f43be..3e2839048fe6 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -228,4 +228,6 @@ source "drivers/interconnect/Kconfig"
source "drivers/counter/Kconfig"
+source "drivers/mdev/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index aaef17cc6512..592e23f2e629 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -186,3 +186,4 @@ obj-$(CONFIG_SIOX) += siox/
obj-$(CONFIG_GNSS) += gnss/
obj-$(CONFIG_INTERCONNECT) += interconnect/
obj-$(CONFIG_COUNTER) += counter/
+obj-$(CONFIG_MDEV) += mdev/
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
new file mode 100644
index 000000000000..4561f2d4178f
--- /dev/null
+++ b/drivers/mdev/Kconfig
@@ -0,0 +1,19 @@
+
+config MDEV
+ tristate "Mediated device driver framework"
+ default n
+ help
+ Provides a framework to virtualize devices.
+
+ If you don't know what do here, say N.
+
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
new file mode 100644
index 000000000000..0b749e7f8ff4
--- /dev/null
+++ b/drivers/mdev/Makefile
@@ -0,0 +1,5 @@
+
+mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
+mdev_vfio-y := vfio.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/mdev/mdev_core.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_core.c
rename to drivers/mdev/mdev_core.c
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/mdev/mdev_driver.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_driver.c
rename to drivers/mdev/mdev_driver.c
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/mdev/mdev_private.h
similarity index 100%
rename from drivers/vfio/mdev/mdev_private.h
rename to drivers/mdev/mdev_private.h
diff --git a/drivers/vfio/mdev/mdev_sysfs.c b/drivers/mdev/mdev_sysfs.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_sysfs.c
rename to drivers/mdev/mdev_sysfs.c
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/mdev/vfio.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_vfio.c
rename to drivers/mdev/vfio.c
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 2e07ca915a96..9a9234c3e00e 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,24 +1,4 @@
-config MDEV
- tristate "Mediated device driver framework"
- default n
- help
- Provides a framework to virtualize devices.
-
- If you don't know what do here, say N.
-
-config VFIO_MDEV
- tristate "VFIO Mediated device driver"
- depends on VFIO && MDEV
- default n
- help
- Proivdes a mediated BUS for userspace driver through VFIO
- framework. See Documentation/vfio-mediated-device.txt for
- more details.
-
- If you don't know what do here, say N.
-
-
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index e9675501271a..e2a92df3089e 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,2 @@
-mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-
-obj-$(CONFIG_MDEV) += mdev.o
-obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 8b42a4b3f161..4eada31e9287 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -15,8 +15,6 @@
#include <linux/vfio.h>
#include <linux/mdev_vfio.h>
-#include "mdev_private.h"
-
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "NVIDIA Corporation"
#define DRIVER_DESC "VFIO based driver for Mediated device"
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
Mdev now is nothing VFIO specific, let's move it to upper
directory.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 7 +++++--
drivers/Kconfig | 2 ++
drivers/Makefile | 1 +
drivers/mdev/Kconfig | 19 ++++++++++++++++++
drivers/mdev/Makefile | 5 +++++
drivers/{vfio => }/mdev/mdev_core.c | 0
drivers/{vfio => }/mdev/mdev_driver.c | 0
drivers/{vfio => }/mdev/mdev_private.h | 0
drivers/{vfio => }/mdev/mdev_sysfs.c | 0
.../{vfio/mdev/mdev_vfio.c => mdev/vfio.c} | 0
drivers/vfio/mdev/Kconfig | 20 -------------------
drivers/vfio/mdev/Makefile | 4 ----
drivers/vfio/mdev/vfio_mdev.c | 2 --
13 files changed, 32 insertions(+), 28 deletions(-)
create mode 100644 drivers/mdev/Kconfig
create mode 100644 drivers/mdev/Makefile
rename drivers/{vfio => }/mdev/mdev_core.c (100%)
rename drivers/{vfio => }/mdev/mdev_driver.c (100%)
rename drivers/{vfio => }/mdev/mdev_private.h (100%)
rename drivers/{vfio => }/mdev/mdev_sysfs.c (100%)
rename drivers/{vfio/mdev/mdev_vfio.c => mdev/vfio.c} (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 6d590afb62c3..5d7e8badf58c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17129,15 +17129,18 @@ T: git git://github.com/awilliam/linux-vfio.git
S: Maintained
F: Documentation/driver-api/vfio.rst
F: drivers/vfio/
+F: drivers/mdev/vfio.c
F: include/linux/vfio.h
F: include/uapi/linux/vfio.h
-VFIO MEDIATED DEVICE DRIVERS
+MEDIATED DEVICE DRIVERS
+M: Alex Williamson <alex.williamson@redhat.com>
M: Kirti Wankhede <kwankhede@nvidia.com>
+R: Cornelia Huck <cohuck@redhat.com>
L: kvm@vger.kernel.org
S: Maintained
F: Documentation/driver-api/vfio-mediated-device.rst
-F: drivers/vfio/mdev/
+F: drivers/mdev
F: include/linux/mdev.h
F: include/linux/mdev_vfio.h
F: samples/vfio-mdev/
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 8befa53f43be..3e2839048fe6 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -228,4 +228,6 @@ source "drivers/interconnect/Kconfig"
source "drivers/counter/Kconfig"
+source "drivers/mdev/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index aaef17cc6512..592e23f2e629 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -186,3 +186,4 @@ obj-$(CONFIG_SIOX) += siox/
obj-$(CONFIG_GNSS) += gnss/
obj-$(CONFIG_INTERCONNECT) += interconnect/
obj-$(CONFIG_COUNTER) += counter/
+obj-$(CONFIG_MDEV) += mdev/
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
new file mode 100644
index 000000000000..4561f2d4178f
--- /dev/null
+++ b/drivers/mdev/Kconfig
@@ -0,0 +1,19 @@
+
+config MDEV
+ tristate "Mediated device driver framework"
+ default n
+ help
+ Provides a framework to virtualize devices.
+
+ If you don't know what do here, say N.
+
+config VFIO_MDEV
+ tristate "VFIO Mediated device driver"
+ depends on VFIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for userspace driver through VFIO
+ framework. See Documentation/vfio-mediated-device.txt for
+ more details.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
new file mode 100644
index 000000000000..0b749e7f8ff4
--- /dev/null
+++ b/drivers/mdev/Makefile
@@ -0,0 +1,5 @@
+
+mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
+mdev_vfio-y := vfio.o
+obj-$(CONFIG_MDEV) += mdev.o
+obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/mdev/mdev_core.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_core.c
rename to drivers/mdev/mdev_core.c
diff --git a/drivers/vfio/mdev/mdev_driver.c b/drivers/mdev/mdev_driver.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_driver.c
rename to drivers/mdev/mdev_driver.c
diff --git a/drivers/vfio/mdev/mdev_private.h b/drivers/mdev/mdev_private.h
similarity index 100%
rename from drivers/vfio/mdev/mdev_private.h
rename to drivers/mdev/mdev_private.h
diff --git a/drivers/vfio/mdev/mdev_sysfs.c b/drivers/mdev/mdev_sysfs.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_sysfs.c
rename to drivers/mdev/mdev_sysfs.c
diff --git a/drivers/vfio/mdev/mdev_vfio.c b/drivers/mdev/vfio.c
similarity index 100%
rename from drivers/vfio/mdev/mdev_vfio.c
rename to drivers/mdev/vfio.c
diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
index 2e07ca915a96..9a9234c3e00e 100644
--- a/drivers/vfio/mdev/Kconfig
+++ b/drivers/vfio/mdev/Kconfig
@@ -1,24 +1,4 @@
-config MDEV
- tristate "Mediated device driver framework"
- default n
- help
- Provides a framework to virtualize devices.
-
- If you don't know what do here, say N.
-
-config VFIO_MDEV
- tristate "VFIO Mediated device driver"
- depends on VFIO && MDEV
- default n
- help
- Proivdes a mediated BUS for userspace driver through VFIO
- framework. See Documentation/vfio-mediated-device.txt for
- more details.
-
- If you don't know what do here, say N.
-
-
config VFIO_MDEV_DEVICE
tristate "VFIO driver for Mediated devices"
depends on VFIO && VFIO_MDEV
diff --git a/drivers/vfio/mdev/Makefile b/drivers/vfio/mdev/Makefile
index e9675501271a..e2a92df3089e 100644
--- a/drivers/vfio/mdev/Makefile
+++ b/drivers/vfio/mdev/Makefile
@@ -1,6 +1,2 @@
-mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
-
-obj-$(CONFIG_MDEV) += mdev.o
-obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
obj-$(CONFIG_VFIO_MDEV_DEVICE) += vfio_mdev.o
diff --git a/drivers/vfio/mdev/vfio_mdev.c b/drivers/vfio/mdev/vfio_mdev.c
index 8b42a4b3f161..4eada31e9287 100644
--- a/drivers/vfio/mdev/vfio_mdev.c
+++ b/drivers/vfio/mdev/vfio_mdev.c
@@ -15,8 +15,6 @@
#include <linux/vfio.h>
#include <linux/mdev_vfio.h>
-#include "mdev_private.h"
-
#define DRIVER_VERSION "0.1"
#define DRIVER_AUTHOR "NVIDIA Corporation"
#define DRIVER_DESC "VFIO based driver for Mediated device"
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 4/6] mdev: introduce mediated virtio bus
2019-11-18 10:59 ` Jason Wang
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, Jason Wang
This patch implements a mediated virtio bus over mdev framework. This
will be used by the future virtio-mdev and vhost-mdev on top to allow
driver from either userspace or kernel to control the device which is
capable of offloading virtio datapath.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 2 +
drivers/mdev/Kconfig | 10 ++
drivers/mdev/Makefile | 2 +
drivers/mdev/virtio.c | 126 +++++++++++++++++++++++
include/linux/mdev_virtio.h | 163 ++++++++++++++++++++++++++++++
include/linux/mod_devicetable.h | 8 ++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +++
8 files changed, 326 insertions(+)
create mode 100644 drivers/mdev/virtio.c
create mode 100644 include/linux/mdev_virtio.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 5d7e8badf58c..e1b57c84f249 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17269,6 +17269,8 @@ 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.h
+F: drivers/mdev/virtio.c
VIRTIO BLOCK AND SCSI DRIVERS
M: "Michael S. Tsirkin" <mst@redhat.com>
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
index 4561f2d4178f..cd84d4670552 100644
--- a/drivers/mdev/Kconfig
+++ b/drivers/mdev/Kconfig
@@ -17,3 +17,13 @@ config VFIO_MDEV
more details.
If you don't know what do here, say N.
+
+config MDEV_VIRTIO
+ tristate "Mediated VIRTIO bus"
+ depends on VIRTIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for virtio. It could be used by
+ either kenrel driver or userspace driver.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
index 0b749e7f8ff4..eb14031c9944 100644
--- a/drivers/mdev/Makefile
+++ b/drivers/mdev/Makefile
@@ -1,5 +1,7 @@
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
mdev_vfio-y := vfio.o
+mdev_virtio-y := virtio.o
obj-$(CONFIG_MDEV) += mdev.o
obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
+obj-$(CONFIG_MDEV_VIRTIO) += mdev_virtio.o
diff --git a/drivers/mdev/virtio.c b/drivers/mdev/virtio.c
new file mode 100644
index 000000000000..25de329615c4
--- /dev/null
+++ b/drivers/mdev/virtio.c
@@ -0,0 +1,126 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated VIRTIO bus
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev.h>
+#include <linux/mdev_virtio.h>
+#include <linux/mod_devicetable.h>
+
+#include "mdev_private.h"
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VIRTIO bus"
+
+struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_device {
+ struct mdev_device mdev;
+ const struct mdev_virtio_ops *ops;
+ u16 class_id;
+};
+
+#define to_mdev_virtio(mdev) container_of(mdev, \
+ struct mdev_virtio_device, mdev)
+#define to_mdev_virtio_drv(mdrv) container_of(mdrv, \
+ struct mdev_virtio_driver, drv)
+
+static int mdev_virtio_match(struct device *dev, struct device_driver *drv)
+{
+ unsigned int i;
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+ struct mdev_driver *mdrv = to_mdev_driver(drv);
+ struct mdev_virtio_driver *mdrv_virtio = to_mdev_virtio_drv(mdrv);
+ const struct mdev_virtio_class_id *ids = mdrv_virtio->id_table;
+
+ if (!ids)
+ return 0;
+
+ for (i = 0; ids[i].id; i++)
+ if (ids[i].id == mdev_virtio->class_id)
+ return 1;
+ return 0;
+}
+
+static int mdev_virtio_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return add_uevent_var(env, "MODALIAS=mdev_virtio:c%02X",
+ mdev_virtio->class_id);
+}
+
+struct bus_type mdev_virtio_bus_type = {
+ .name = "mdev_virtio",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+ .match = mdev_virtio_match,
+ .uevent = mdev_virtio_uevent,
+};
+EXPORT_SYMBOL(mdev_virtio_bus_type);
+
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->class_id = class_id;
+}
+EXPORT_SYMBOL(mdev_virtio_set_class_id);
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_virtio_bus_type,
+ sizeof(struct mdev_virtio_device));
+}
+EXPORT_SYMBOL(mdev_virtio_register_device);
+
+void mdev_virtio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_virtio_unregister_device);
+
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_virtio_set_ops);
+
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return mdev_virtio->ops;
+}
+EXPORT_SYMBOL(mdev_virtio_get_ops);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_virtio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_virtio_bus_type);
+}
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
new file mode 100644
index 000000000000..ef2dbb6c383a
--- /dev/null
+++ b/include/linux/mdev_virtio.h
@@ -0,0 +1,163 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VIRTIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef VIRTIO_MDEV_H
+#define VIRTIO_MDEV_H
+
+#include <linux/interrupt.h>
+#include <linux/mod_devicetable.h>
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_driver {
+ struct mdev_driver drv;
+ const struct mdev_virtio_class_id *id_table;
+};
+
+struct virtio_mdev_callback {
+ irqreturn_t (*callback)(void *data);
+ void *private;
+};
+
+/**
+ * struct mdev_virtio_device_ops - Structure to be registered for each
+ * mdev device to register the device for virtio/vhost drivers.
+ *
+ * The callbacks are mandatory unless explicitly mentioned.
+ *
+ * @set_vq_address: Set the address of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @desc_area: address of desc area
+ * @driver_area: address of driver area
+ * @device_area: address of device area
+ * Returns integer: success (0) or error (< 0)
+ * @set_vq_num: Set the size of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @num: the size of virtqueue
+ * @kick_vq: Kick the virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @set_vq_cb: Set the interrupt callback function for
+ * a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @cb: virtio-mdev interrupt callback structure
+ * @set_vq_ready: Set ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @ready: ready (true) not ready(false)
+ * @get_vq_ready: Get ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns boolean: ready (true) or not (false)
+ * @set_vq_state: Set the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @state: virtqueue state (last_avail_idx)
+ * Returns integer: success (0) or error (< 0)
+ * @get_vq_state: Get the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns virtqueue state (last_avail_idx)
+ * @get_vq_align: Get the virtqueue align requirement
+ * for the device
+ * @mdev: mediated device
+ * Returns virtqueue algin requirement
+ * @get_features: Get virtio features supported by the device
+ * @mdev: mediated device
+ * Returns the virtio features support by the
+ * device
+ * @set_features: Set virtio features supported by the driver
+ * @mdev: mediated device
+ * @features: feature support by the driver
+ * Returns integer: success (0) or error (< 0)
+ * @set_config_cb: Set the config interrupt callback
+ * @mdev: mediated device
+ * @cb: virtio-mdev interrupt callback structure
+ * @get_vq_num_max: Get the max size of virtqueue
+ * @mdev: mediated device
+ * Returns u16: max size of virtqueue
+ * @get_device_id: Get virtio device id
+ * @mdev: mediated device
+ * Returns u32: virtio device id
+ * @get_vendor_id: Get id for the vendor that provides this device
+ * @mdev: mediated device
+ * Returns u32: virtio vendor id
+ * @get_status: Get the device status
+ * @mdev: mediated device
+ * Returns u8: virtio device status
+ * @set_status: Set the device status
+ * @mdev: mediated device
+ * @status: virtio device status
+ * @get_config: Read from device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to read to
+ * @len: the length to read from
+ * configration space
+ * @set_config: Write to device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to write from
+ * @len: the length to write to
+ * configration space
+ * @get_generation: Get device config generaton (optional)
+ * @mdev: mediated device
+ * Returns u32: device generation
+ */
+struct mdev_virtio_ops {
+ /* Virtqueue ops */
+ int (*set_vq_address)(struct mdev_device *mdev,
+ u16 idx, u64 desc_area, u64 driver_area,
+ u64 device_area);
+ void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
+ void (*kick_vq)(struct mdev_device *mdev, u16 idx);
+ void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
+ struct virtio_mdev_callback *cb);
+ void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
+ bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
+ int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
+ u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
+
+ /* Device ops */
+ u16 (*get_vq_align)(struct mdev_device *mdev);
+ u64 (*get_features)(struct mdev_device *mdev);
+ int (*set_features)(struct mdev_device *mdev, u64 features);
+ void (*set_config_cb)(struct mdev_device *mdev,
+ struct virtio_mdev_callback *cb);
+ u16 (*get_vq_num_max)(struct mdev_device *mdev);
+ u32 (*get_device_id)(struct mdev_device *mdev);
+ u32 (*get_vendor_id)(struct mdev_device *mdev);
+ u8 (*get_status)(struct mdev_device *mdev);
+ void (*set_status)(struct mdev_device *mdev, u8 status);
+ void (*get_config)(struct mdev_device *mdev, unsigned int offset,
+ void *buf, unsigned int len);
+ void (*set_config)(struct mdev_device *mdev, unsigned int offset,
+ const void *buf, unsigned int len);
+ u32 (*get_generation)(struct mdev_device *mdev);
+};
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_virtio_unregister_device(struct device *dev);
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops);
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev);
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id);
+
+static inline struct mdev_device *mdev_virtio_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_virtio_bus_type);
+}
+
+#endif
diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
index 5714fd35a83c..59006c47ae8e 100644
--- a/include/linux/mod_devicetable.h
+++ b/include/linux/mod_devicetable.h
@@ -821,4 +821,12 @@ struct wmi_device_id {
const void *context;
};
+/**
+ * struct mdev_class_id - MDEV VIRTIO device class identifier
+ * @id: Used to identify a specific class of device, e.g vfio-mdev device.
+ */
+struct mdev_virtio_class_id {
+ __u16 id;
+};
+
#endif /* LINUX_MOD_DEVICETABLE_H */
diff --git a/scripts/mod/devicetable-offsets.c b/scripts/mod/devicetable-offsets.c
index 054405b90ba4..178fd7c70812 100644
--- a/scripts/mod/devicetable-offsets.c
+++ b/scripts/mod/devicetable-offsets.c
@@ -231,5 +231,8 @@ int main(void)
DEVID(wmi_device_id);
DEVID_FIELD(wmi_device_id, guid_string);
+ DEVID(mdev_virtio_class_id);
+ DEVID_FIELD(mdev_virtio_class_id, id);
+
return 0;
}
diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
index c91eba751804..1a9c1f591951 100644
--- a/scripts/mod/file2alias.c
+++ b/scripts/mod/file2alias.c
@@ -1335,6 +1335,17 @@ static int do_wmi_entry(const char *filename, void *symval, char *alias)
return 1;
}
+/* looks like: "mdev_virtio:cN" */
+static int do_mdev_virtio_entry(const char *filename, void *symval, char *alias)
+{
+ DEF_FIELD(symval, mdev_virtio_class_id, id);
+
+ sprintf(alias, "mdev_virtio: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 +1418,7 @@ 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_virtio", SIZE_mdev_virtio_class_id, do_mdev_virtio_entry},
};
/* Create MODULE_ALIAS() statements.
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
This patch implements a mediated virtio bus over mdev framework. This
will be used by the future virtio-mdev and vhost-mdev on top to allow
driver from either userspace or kernel to control the device which is
capable of offloading virtio datapath.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 2 +
drivers/mdev/Kconfig | 10 ++
drivers/mdev/Makefile | 2 +
drivers/mdev/virtio.c | 126 +++++++++++++++++++++++
include/linux/mdev_virtio.h | 163 ++++++++++++++++++++++++++++++
include/linux/mod_devicetable.h | 8 ++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +++
8 files changed, 326 insertions(+)
create mode 100644 drivers/mdev/virtio.c
create mode 100644 include/linux/mdev_virtio.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 5d7e8badf58c..e1b57c84f249 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17269,6 +17269,8 @@ 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.h
+F: drivers/mdev/virtio.c
VIRTIO BLOCK AND SCSI DRIVERS
M: "Michael S. Tsirkin" <mst@redhat.com>
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
index 4561f2d4178f..cd84d4670552 100644
--- a/drivers/mdev/Kconfig
+++ b/drivers/mdev/Kconfig
@@ -17,3 +17,13 @@ config VFIO_MDEV
more details.
If you don't know what do here, say N.
+
+config MDEV_VIRTIO
+ tristate "Mediated VIRTIO bus"
+ depends on VIRTIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for virtio. It could be used by
+ either kenrel driver or userspace driver.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
index 0b749e7f8ff4..eb14031c9944 100644
--- a/drivers/mdev/Makefile
+++ b/drivers/mdev/Makefile
@@ -1,5 +1,7 @@
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
mdev_vfio-y := vfio.o
+mdev_virtio-y := virtio.o
obj-$(CONFIG_MDEV) += mdev.o
obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
+obj-$(CONFIG_MDEV_VIRTIO) += mdev_virtio.o
diff --git a/drivers/mdev/virtio.c b/drivers/mdev/virtio.c
new file mode 100644
index 000000000000..25de329615c4
--- /dev/null
+++ b/drivers/mdev/virtio.c
@@ -0,0 +1,126 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated VIRTIO bus
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev.h>
+#include <linux/mdev_virtio.h>
+#include <linux/mod_devicetable.h>
+
+#include "mdev_private.h"
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VIRTIO bus"
+
+struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_device {
+ struct mdev_device mdev;
+ const struct mdev_virtio_ops *ops;
+ u16 class_id;
+};
+
+#define to_mdev_virtio(mdev) container_of(mdev, \
+ struct mdev_virtio_device, mdev)
+#define to_mdev_virtio_drv(mdrv) container_of(mdrv, \
+ struct mdev_virtio_driver, drv)
+
+static int mdev_virtio_match(struct device *dev, struct device_driver *drv)
+{
+ unsigned int i;
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+ struct mdev_driver *mdrv = to_mdev_driver(drv);
+ struct mdev_virtio_driver *mdrv_virtio = to_mdev_virtio_drv(mdrv);
+ const struct mdev_virtio_class_id *ids = mdrv_virtio->id_table;
+
+ if (!ids)
+ return 0;
+
+ for (i = 0; ids[i].id; i++)
+ if (ids[i].id == mdev_virtio->class_id)
+ return 1;
+ return 0;
+}
+
+static int mdev_virtio_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return add_uevent_var(env, "MODALIAS=mdev_virtio:c%02X",
+ mdev_virtio->class_id);
+}
+
+struct bus_type mdev_virtio_bus_type = {
+ .name = "mdev_virtio",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+ .match = mdev_virtio_match,
+ .uevent = mdev_virtio_uevent,
+};
+EXPORT_SYMBOL(mdev_virtio_bus_type);
+
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->class_id = class_id;
+}
+EXPORT_SYMBOL(mdev_virtio_set_class_id);
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_virtio_bus_type,
+ sizeof(struct mdev_virtio_device));
+}
+EXPORT_SYMBOL(mdev_virtio_register_device);
+
+void mdev_virtio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_virtio_unregister_device);
+
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_virtio_set_ops);
+
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return mdev_virtio->ops;
+}
+EXPORT_SYMBOL(mdev_virtio_get_ops);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_virtio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_virtio_bus_type);
+}
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
new file mode 100644
index 000000000000..ef2dbb6c383a
--- /dev/null
+++ b/include/linux/mdev_virtio.h
@@ -0,0 +1,163 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VIRTIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef VIRTIO_MDEV_H
+#define VIRTIO_MDEV_H
+
+#include <linux/interrupt.h>
+#include <linux/mod_devicetable.h>
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_driver {
+ struct mdev_driver drv;
+ const struct mdev_virtio_class_id *id_table;
+};
+
+struct virtio_mdev_callback {
+ irqreturn_t (*callback)(void *data);
+ void *private;
+};
+
+/**
+ * struct mdev_virtio_device_ops - Structure to be registered for each
+ * mdev device to register the device for virtio/vhost drivers.
+ *
+ * The callbacks are mandatory unless explicitly mentioned.
+ *
+ * @set_vq_address: Set the address of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @desc_area: address of desc area
+ * @driver_area: address of driver area
+ * @device_area: address of device area
+ * Returns integer: success (0) or error (< 0)
+ * @set_vq_num: Set the size of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @num: the size of virtqueue
+ * @kick_vq: Kick the virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @set_vq_cb: Set the interrupt callback function for
+ * a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @cb: virtio-mdev interrupt callback structure
+ * @set_vq_ready: Set ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @ready: ready (true) not ready(false)
+ * @get_vq_ready: Get ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns boolean: ready (true) or not (false)
+ * @set_vq_state: Set the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @state: virtqueue state (last_avail_idx)
+ * Returns integer: success (0) or error (< 0)
+ * @get_vq_state: Get the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns virtqueue state (last_avail_idx)
+ * @get_vq_align: Get the virtqueue align requirement
+ * for the device
+ * @mdev: mediated device
+ * Returns virtqueue algin requirement
+ * @get_features: Get virtio features supported by the device
+ * @mdev: mediated device
+ * Returns the virtio features support by the
+ * device
+ * @set_features: Set virtio features supported by the driver
+ * @mdev: mediated device
+ * @features: feature support by the driver
+ * Returns integer: success (0) or error (< 0)
+ * @set_config_cb: Set the config interrupt callback
+ * @mdev: mediated device
+ * @cb: virtio-mdev interrupt callback structure
+ * @get_vq_num_max: Get the max size of virtqueue
+ * @mdev: mediated device
+ * Returns u16: max size of virtqueue
+ * @get_device_id: Get virtio device id
+ * @mdev: mediated device
+ * Returns u32: virtio device id
+ * @get_vendor_id: Get id for the vendor that provides this device
+ * @mdev: mediated device
+ * Returns u32: virtio vendor id
+ * @get_status: Get the device status
+ * @mdev: mediated device
+ * Returns u8: virtio device status
+ * @set_status: Set the device status
+ * @mdev: mediated device
+ * @status: virtio device status
+ * @get_config: Read from device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to read to
+ * @len: the length to read from
+ * configration space
+ * @set_config: Write to device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to write from
+ * @len: the length to write to
+ * configration space
+ * @get_generation: Get device config generaton (optional)
+ * @mdev: mediated device
+ * Returns u32: device generation
+ */
+struct mdev_virtio_ops {
+ /* Virtqueue ops */
+ int (*set_vq_address)(struct mdev_device *mdev,
+ u16 idx, u64 desc_area, u64 driver_area,
+ u64 device_area);
+ void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
+ void (*kick_vq)(struct mdev_device *mdev, u16 idx);
+ void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
+ struct virtio_mdev_callback *cb);
+ void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
+ bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
+ int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
+ u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
+
+ /* Device ops */
+ u16 (*get_vq_align)(struct mdev_device *mdev);
+ u64 (*get_features)(struct mdev_device *mdev);
+ int (*set_features)(struct mdev_device *mdev, u64 features);
+ void (*set_config_cb)(struct mdev_device *mdev,
+ struct virtio_mdev_callback *cb);
+ u16 (*get_vq_num_max)(struct mdev_device *mdev);
+ u32 (*get_device_id)(struct mdev_device *mdev);
+ u32 (*get_vendor_id)(struct mdev_device *mdev);
+ u8 (*get_status)(struct mdev_device *mdev);
+ void (*set_status)(struct mdev_device *mdev, u8 status);
+ void (*get_config)(struct mdev_device *mdev, unsigned int offset,
+ void *buf, unsigned int len);
+ void (*set_config)(struct mdev_device *mdev, unsigned int offset,
+ const void *buf, unsigned int len);
+ u32 (*get_generation)(struct mdev_device *mdev);
+};
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_virtio_unregister_device(struct device *dev);
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops);
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev);
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id);
+
+static inline struct mdev_device *mdev_virtio_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_virtio_bus_type);
+}
+
+#endif
diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
index 5714fd35a83c..59006c47ae8e 100644
--- a/include/linux/mod_devicetable.h
+++ b/include/linux/mod_devicetable.h
@@ -821,4 +821,12 @@ struct wmi_device_id {
const void *context;
};
+/**
+ * struct mdev_class_id - MDEV VIRTIO device class identifier
+ * @id: Used to identify a specific class of device, e.g vfio-mdev device.
+ */
+struct mdev_virtio_class_id {
+ __u16 id;
+};
+
#endif /* LINUX_MOD_DEVICETABLE_H */
diff --git a/scripts/mod/devicetable-offsets.c b/scripts/mod/devicetable-offsets.c
index 054405b90ba4..178fd7c70812 100644
--- a/scripts/mod/devicetable-offsets.c
+++ b/scripts/mod/devicetable-offsets.c
@@ -231,5 +231,8 @@ int main(void)
DEVID(wmi_device_id);
DEVID_FIELD(wmi_device_id, guid_string);
+ DEVID(mdev_virtio_class_id);
+ DEVID_FIELD(mdev_virtio_class_id, id);
+
return 0;
}
diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
index c91eba751804..1a9c1f591951 100644
--- a/scripts/mod/file2alias.c
+++ b/scripts/mod/file2alias.c
@@ -1335,6 +1335,17 @@ static int do_wmi_entry(const char *filename, void *symval, char *alias)
return 1;
}
+/* looks like: "mdev_virtio:cN" */
+static int do_mdev_virtio_entry(const char *filename, void *symval, char *alias)
+{
+ DEF_FIELD(symval, mdev_virtio_class_id, id);
+
+ sprintf(alias, "mdev_virtio: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 +1418,7 @@ 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_virtio", SIZE_mdev_virtio_class_id, do_mdev_virtio_entry},
};
/* Create MODULE_ALIAS() statements.
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
This patch implements a mediated virtio bus over mdev framework. This
will be used by the future virtio-mdev and vhost-mdev on top to allow
driver from either userspace or kernel to control the device which is
capable of offloading virtio datapath.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 2 +
drivers/mdev/Kconfig | 10 ++
drivers/mdev/Makefile | 2 +
drivers/mdev/virtio.c | 126 +++++++++++++++++++++++
include/linux/mdev_virtio.h | 163 ++++++++++++++++++++++++++++++
include/linux/mod_devicetable.h | 8 ++
scripts/mod/devicetable-offsets.c | 3 +
scripts/mod/file2alias.c | 12 +++
8 files changed, 326 insertions(+)
create mode 100644 drivers/mdev/virtio.c
create mode 100644 include/linux/mdev_virtio.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 5d7e8badf58c..e1b57c84f249 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17269,6 +17269,8 @@ 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.h
+F: drivers/mdev/virtio.c
VIRTIO BLOCK AND SCSI DRIVERS
M: "Michael S. Tsirkin" <mst@redhat.com>
diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
index 4561f2d4178f..cd84d4670552 100644
--- a/drivers/mdev/Kconfig
+++ b/drivers/mdev/Kconfig
@@ -17,3 +17,13 @@ config VFIO_MDEV
more details.
If you don't know what do here, say N.
+
+config MDEV_VIRTIO
+ tristate "Mediated VIRTIO bus"
+ depends on VIRTIO && MDEV
+ default n
+ help
+ Proivdes a mediated BUS for virtio. It could be used by
+ either kenrel driver or userspace driver.
+
+ If you don't know what do here, say N.
diff --git a/drivers/mdev/Makefile b/drivers/mdev/Makefile
index 0b749e7f8ff4..eb14031c9944 100644
--- a/drivers/mdev/Makefile
+++ b/drivers/mdev/Makefile
@@ -1,5 +1,7 @@
mdev-y := mdev_core.o mdev_sysfs.o mdev_driver.o
mdev_vfio-y := vfio.o
+mdev_virtio-y := virtio.o
obj-$(CONFIG_MDEV) += mdev.o
obj-$(CONFIG_VFIO_MDEV) += mdev_vfio.o
+obj-$(CONFIG_MDEV_VIRTIO) += mdev_virtio.o
diff --git a/drivers/mdev/virtio.c b/drivers/mdev/virtio.c
new file mode 100644
index 000000000000..25de329615c4
--- /dev/null
+++ b/drivers/mdev/virtio.c
@@ -0,0 +1,126 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated VIRTIO bus
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#include <linux/module.h>
+#include <linux/uuid.h>
+#include <linux/device.h>
+#include <linux/mdev.h>
+#include <linux/mdev_virtio.h>
+#include <linux/mod_devicetable.h>
+
+#include "mdev_private.h"
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Jason Wang"
+#define DRIVER_DESC "Mediated VIRTIO bus"
+
+struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_device {
+ struct mdev_device mdev;
+ const struct mdev_virtio_ops *ops;
+ u16 class_id;
+};
+
+#define to_mdev_virtio(mdev) container_of(mdev, \
+ struct mdev_virtio_device, mdev)
+#define to_mdev_virtio_drv(mdrv) container_of(mdrv, \
+ struct mdev_virtio_driver, drv)
+
+static int mdev_virtio_match(struct device *dev, struct device_driver *drv)
+{
+ unsigned int i;
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+ struct mdev_driver *mdrv = to_mdev_driver(drv);
+ struct mdev_virtio_driver *mdrv_virtio = to_mdev_virtio_drv(mdrv);
+ const struct mdev_virtio_class_id *ids = mdrv_virtio->id_table;
+
+ if (!ids)
+ return 0;
+
+ for (i = 0; ids[i].id; i++)
+ if (ids[i].id == mdev_virtio->class_id)
+ return 1;
+ return 0;
+}
+
+static int mdev_virtio_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+ struct mdev_device *mdev = mdev_from_dev(dev, &mdev_virtio_bus_type);
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return add_uevent_var(env, "MODALIAS=mdev_virtio:c%02X",
+ mdev_virtio->class_id);
+}
+
+struct bus_type mdev_virtio_bus_type = {
+ .name = "mdev_virtio",
+ .probe = mdev_probe,
+ .remove = mdev_remove,
+ .match = mdev_virtio_match,
+ .uevent = mdev_virtio_uevent,
+};
+EXPORT_SYMBOL(mdev_virtio_bus_type);
+
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->class_id = class_id;
+}
+EXPORT_SYMBOL(mdev_virtio_set_class_id);
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops)
+{
+ return mdev_register_device(dev, ops, &mdev_virtio_bus_type,
+ sizeof(struct mdev_virtio_device));
+}
+EXPORT_SYMBOL(mdev_virtio_register_device);
+
+void mdev_virtio_unregister_device(struct device *dev)
+{
+ return mdev_unregister_device(dev);
+}
+EXPORT_SYMBOL(mdev_virtio_unregister_device);
+
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ mdev_virtio->ops = ops;
+}
+EXPORT_SYMBOL(mdev_virtio_set_ops);
+
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev)
+{
+ struct mdev_virtio_device *mdev_virtio = to_mdev_virtio(mdev);
+
+ return mdev_virtio->ops;
+}
+EXPORT_SYMBOL(mdev_virtio_get_ops);
+
+static int __init mdev_init(void)
+{
+ return mdev_register_bus(&mdev_virtio_bus_type);
+}
+
+static void __exit mdev_exit(void)
+{
+ mdev_unregister_bus(&mdev_virtio_bus_type);
+}
+
+module_init(mdev_init)
+module_exit(mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
new file mode 100644
index 000000000000..ef2dbb6c383a
--- /dev/null
+++ b/include/linux/mdev_virtio.h
@@ -0,0 +1,163 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VIRTIO Mediated device definition
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ */
+
+#ifndef VIRTIO_MDEV_H
+#define VIRTIO_MDEV_H
+
+#include <linux/interrupt.h>
+#include <linux/mod_devicetable.h>
+#include <linux/mdev.h>
+
+extern struct bus_type mdev_virtio_bus_type;
+
+struct mdev_virtio_driver {
+ struct mdev_driver drv;
+ const struct mdev_virtio_class_id *id_table;
+};
+
+struct virtio_mdev_callback {
+ irqreturn_t (*callback)(void *data);
+ void *private;
+};
+
+/**
+ * struct mdev_virtio_device_ops - Structure to be registered for each
+ * mdev device to register the device for virtio/vhost drivers.
+ *
+ * The callbacks are mandatory unless explicitly mentioned.
+ *
+ * @set_vq_address: Set the address of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @desc_area: address of desc area
+ * @driver_area: address of driver area
+ * @device_area: address of device area
+ * Returns integer: success (0) or error (< 0)
+ * @set_vq_num: Set the size of virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @num: the size of virtqueue
+ * @kick_vq: Kick the virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @set_vq_cb: Set the interrupt callback function for
+ * a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @cb: virtio-mdev interrupt callback structure
+ * @set_vq_ready: Set ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @ready: ready (true) not ready(false)
+ * @get_vq_ready: Get ready status for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns boolean: ready (true) or not (false)
+ * @set_vq_state: Set the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * @state: virtqueue state (last_avail_idx)
+ * Returns integer: success (0) or error (< 0)
+ * @get_vq_state: Get the state for a virtqueue
+ * @mdev: mediated device
+ * @idx: virtqueue index
+ * Returns virtqueue state (last_avail_idx)
+ * @get_vq_align: Get the virtqueue align requirement
+ * for the device
+ * @mdev: mediated device
+ * Returns virtqueue algin requirement
+ * @get_features: Get virtio features supported by the device
+ * @mdev: mediated device
+ * Returns the virtio features support by the
+ * device
+ * @set_features: Set virtio features supported by the driver
+ * @mdev: mediated device
+ * @features: feature support by the driver
+ * Returns integer: success (0) or error (< 0)
+ * @set_config_cb: Set the config interrupt callback
+ * @mdev: mediated device
+ * @cb: virtio-mdev interrupt callback structure
+ * @get_vq_num_max: Get the max size of virtqueue
+ * @mdev: mediated device
+ * Returns u16: max size of virtqueue
+ * @get_device_id: Get virtio device id
+ * @mdev: mediated device
+ * Returns u32: virtio device id
+ * @get_vendor_id: Get id for the vendor that provides this device
+ * @mdev: mediated device
+ * Returns u32: virtio vendor id
+ * @get_status: Get the device status
+ * @mdev: mediated device
+ * Returns u8: virtio device status
+ * @set_status: Set the device status
+ * @mdev: mediated device
+ * @status: virtio device status
+ * @get_config: Read from device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to read to
+ * @len: the length to read from
+ * configration space
+ * @set_config: Write to device specific configuration space
+ * @mdev: mediated device
+ * @offset: offset from the beginning of
+ * configuration space
+ * @buf: buffer used to write from
+ * @len: the length to write to
+ * configration space
+ * @get_generation: Get device config generaton (optional)
+ * @mdev: mediated device
+ * Returns u32: device generation
+ */
+struct mdev_virtio_ops {
+ /* Virtqueue ops */
+ int (*set_vq_address)(struct mdev_device *mdev,
+ u16 idx, u64 desc_area, u64 driver_area,
+ u64 device_area);
+ void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
+ void (*kick_vq)(struct mdev_device *mdev, u16 idx);
+ void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
+ struct virtio_mdev_callback *cb);
+ void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
+ bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
+ int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
+ u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
+
+ /* Device ops */
+ u16 (*get_vq_align)(struct mdev_device *mdev);
+ u64 (*get_features)(struct mdev_device *mdev);
+ int (*set_features)(struct mdev_device *mdev, u64 features);
+ void (*set_config_cb)(struct mdev_device *mdev,
+ struct virtio_mdev_callback *cb);
+ u16 (*get_vq_num_max)(struct mdev_device *mdev);
+ u32 (*get_device_id)(struct mdev_device *mdev);
+ u32 (*get_vendor_id)(struct mdev_device *mdev);
+ u8 (*get_status)(struct mdev_device *mdev);
+ void (*set_status)(struct mdev_device *mdev, u8 status);
+ void (*get_config)(struct mdev_device *mdev, unsigned int offset,
+ void *buf, unsigned int len);
+ void (*set_config)(struct mdev_device *mdev, unsigned int offset,
+ const void *buf, unsigned int len);
+ u32 (*get_generation)(struct mdev_device *mdev);
+};
+
+int mdev_virtio_register_device(struct device *dev,
+ const struct mdev_parent_ops *ops);
+void mdev_virtio_unregister_device(struct device *dev);
+void mdev_virtio_set_ops(struct mdev_device *mdev,
+ const struct mdev_virtio_ops *ops);
+const struct mdev_virtio_ops *mdev_virtio_get_ops(struct mdev_device *mdev);
+void mdev_virtio_set_class_id(struct mdev_device *mdev, u16 class_id);
+
+static inline struct mdev_device *mdev_virtio_from_dev(struct device *dev)
+{
+ return mdev_from_dev(dev, &mdev_virtio_bus_type);
+}
+
+#endif
diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
index 5714fd35a83c..59006c47ae8e 100644
--- a/include/linux/mod_devicetable.h
+++ b/include/linux/mod_devicetable.h
@@ -821,4 +821,12 @@ struct wmi_device_id {
const void *context;
};
+/**
+ * struct mdev_class_id - MDEV VIRTIO device class identifier
+ * @id: Used to identify a specific class of device, e.g vfio-mdev device.
+ */
+struct mdev_virtio_class_id {
+ __u16 id;
+};
+
#endif /* LINUX_MOD_DEVICETABLE_H */
diff --git a/scripts/mod/devicetable-offsets.c b/scripts/mod/devicetable-offsets.c
index 054405b90ba4..178fd7c70812 100644
--- a/scripts/mod/devicetable-offsets.c
+++ b/scripts/mod/devicetable-offsets.c
@@ -231,5 +231,8 @@ int main(void)
DEVID(wmi_device_id);
DEVID_FIELD(wmi_device_id, guid_string);
+ DEVID(mdev_virtio_class_id);
+ DEVID_FIELD(mdev_virtio_class_id, id);
+
return 0;
}
diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
index c91eba751804..1a9c1f591951 100644
--- a/scripts/mod/file2alias.c
+++ b/scripts/mod/file2alias.c
@@ -1335,6 +1335,17 @@ static int do_wmi_entry(const char *filename, void *symval, char *alias)
return 1;
}
+/* looks like: "mdev_virtio:cN" */
+static int do_mdev_virtio_entry(const char *filename, void *symval, char *alias)
+{
+ DEF_FIELD(symval, mdev_virtio_class_id, id);
+
+ sprintf(alias, "mdev_virtio: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 +1418,7 @@ 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_virtio", SIZE_mdev_virtio_class_id, do_mdev_virtio_entry},
};
/* Create MODULE_ALIAS() statements.
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 5/6] virtio: introduce a mdev based transport
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, 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
bus_ops of mdev parent.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
drivers/virtio/Kconfig | 13 ++
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++++++++++++++++++++++++++
include/linux/mdev_virtio.h | 5 +
4 files changed, 428 insertions(+)
create mode 100644 drivers/virtio/virtio_mdev.c
diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index 078615cf2afc..6a89b3de97d3 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
If unsure, say Y.
+config VIRTIO_MDEV
+ tristate "MDEV driver for virtio devices"
+ depends on 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
diff --git a/drivers/virtio/Makefile b/drivers/virtio/Makefile
index 3a2b5c5dcf46..f2997b6c812f 100644
--- a/drivers/virtio/Makefile
+++ b/drivers/virtio/Makefile
@@ -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
diff --git a/drivers/virtio/virtio_mdev.c b/drivers/virtio/virtio_mdev.c
new file mode 100644
index 000000000000..7fdb42f055df
--- /dev/null
+++ b/drivers/virtio/virtio_mdev.c
@@ -0,0 +1,409 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VIRTIO based driver for Mediated device
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uuid.h>
+#include <linux/virtio.h>
+#include <linux/mdev_virtio.h>
+#include <linux/virtio_config.h>
+#include <linux/virtio_ring.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Red Hat Corporation"
+#define DRIVER_DESC "VIRTIO based driver for Mediated device"
+
+#define to_virtio_mdev_device(dev) \
+ container_of(dev, struct virtio_mdev_device, vdev)
+
+struct virtio_mdev_device {
+ struct virtio_device vdev;
+ struct mdev_device *mdev;
+ u64 features;
+
+ /* The lock to protect virtqueue list */
+ spinlock_t lock;
+ /* List of virtio_mdev_vq_info */
+ struct list_head virtqueues;
+};
+
+struct virtio_mdev_vq_info {
+ /* the actual virtqueue */
+ struct virtqueue *vq;
+
+ /* the list node for the virtqueues list */
+ struct list_head node;
+};
+
+static struct mdev_device *vm_get_mdev(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return mdev;
+}
+
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
+ void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->get_config(mdev, offset, buf, len);
+}
+
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
+ const void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->set_config(mdev, offset, buf, len);
+}
+
+static u32 virtio_mdev_generation(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+
+ if (ops->get_generation)
+ return ops->get_generation(mdev);
+
+ return 0;
+}
+
+static u8 virtio_mdev_get_status(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_status(mdev);
+}
+
+static void virtio_mdev_set_status(struct virtio_device *vdev, u8 status)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, status);
+}
+
+static void virtio_mdev_reset(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, 0);
+}
+
+static bool virtio_mdev_notify(struct virtqueue *vq)
+{
+ struct mdev_device *mdev = vm_get_mdev(vq->vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->kick_vq(mdev, vq->index);
+
+ return true;
+}
+
+static irqreturn_t virtio_mdev_config_cb(void *private)
+{
+ struct virtio_mdev_device *vm_dev = private;
+
+ virtio_config_changed(&vm_dev->vdev);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t virtio_mdev_virtqueue_cb(void *private)
+{
+ struct virtio_mdev_vq_info *info = private;
+
+ return vring_interrupt(0, info->vq);
+}
+
+static struct virtqueue *
+virtio_mdev_setup_vq(struct virtio_device *vdev, unsigned int index,
+ void (*callback)(struct virtqueue *vq),
+ const char *name, bool ctx)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info;
+ struct virtio_mdev_callback cb;
+ struct virtqueue *vq;
+ u64 desc_addr, driver_addr, device_addr;
+ unsigned long flags;
+ u32 align, num;
+ int err;
+
+ if (!name)
+ return NULL;
+
+ /* Queue shouldn't already be set up. */
+ if (ops->get_vq_ready(mdev, index))
+ return ERR_PTR(-ENOENT);
+
+ /* Allocate and fill out our active queue description */
+ info = kmalloc(sizeof(*info), GFP_KERNEL);
+ if (!info)
+ return ERR_PTR(-ENOMEM);
+
+ num = ops->get_vq_num_max(mdev);
+ if (num == 0) {
+ err = -ENOENT;
+ goto error_new_virtqueue;
+ }
+
+ /* Create the vring */
+ align = ops->get_vq_align(mdev);
+ vq = vring_create_virtqueue(index, num, align, vdev,
+ true, true, ctx,
+ virtio_mdev_notify, callback, name);
+ if (!vq) {
+ err = -ENOMEM;
+ goto error_new_virtqueue;
+ }
+
+ /* Setup virtqueue callback */
+ cb.callback = virtio_mdev_virtqueue_cb;
+ cb.private = info;
+ ops->set_vq_cb(mdev, index, &cb);
+ ops->set_vq_num(mdev, index, virtqueue_get_vring_size(vq));
+
+ desc_addr = virtqueue_get_desc_addr(vq);
+ driver_addr = virtqueue_get_avail_addr(vq);
+ device_addr = virtqueue_get_used_addr(vq);
+
+ if (ops->set_vq_address(mdev, index,
+ desc_addr, driver_addr,
+ device_addr)) {
+ err = -EINVAL;
+ goto err_vq;
+ }
+
+ ops->set_vq_ready(mdev, index, 1);
+
+ vq->priv = info;
+ info->vq = vq;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_add(&info->node, &vm_dev->virtqueues);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ return vq;
+
+err_vq:
+ vring_del_virtqueue(vq);
+error_new_virtqueue:
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+ kfree(info);
+ return ERR_PTR(err);
+}
+
+static void virtio_mdev_del_vq(struct virtqueue *vq)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vq->vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info = vq->priv;
+ unsigned int index = vq->index;
+ unsigned long flags;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_del(&info->node);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ /* Select and deactivate the queue */
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+
+ vring_del_virtqueue(vq);
+
+ kfree(info);
+}
+
+static void virtio_mdev_del_vqs(struct virtio_device *vdev)
+{
+ struct virtqueue *vq, *n;
+
+ list_for_each_entry_safe(vq, n, &vdev->vqs, list)
+ virtio_mdev_del_vq(vq);
+}
+
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+ struct virtqueue *vqs[],
+ vq_callback_t *callbacks[],
+ const char * const names[],
+ const bool *ctx,
+ struct irq_affinity *desc)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_callback cb;
+ int i, err, queue_idx = 0;
+
+ for (i = 0; i < nvqs; ++i) {
+ if (!names[i]) {
+ vqs[i] = NULL;
+ continue;
+ }
+
+ vqs[i] = virtio_mdev_setup_vq(vdev, queue_idx++,
+ callbacks[i], names[i], ctx ?
+ ctx[i] : false);
+ if (IS_ERR(vqs[i])) {
+ err = PTR_ERR(vqs[i]);
+ goto err_setup_vq;
+ }
+ }
+
+ cb.callback = virtio_mdev_config_cb;
+ cb.private = vm_dev;
+ ops->set_config_cb(mdev, &cb);
+
+ return 0;
+
+err_setup_vq:
+ virtio_mdev_del_vqs(vdev);
+ return err;
+}
+
+static u64 virtio_mdev_get_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_features(mdev);
+}
+
+static int virtio_mdev_finalize_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ /* Give virtio_ring a chance to accept features. */
+ vring_transport_features(vdev);
+
+ return ops->set_features(mdev, vdev->features);
+}
+
+static const char *virtio_mdev_bus_name(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return dev_name(mdev_dev(mdev));
+}
+
+static const struct virtio_config_ops virtio_mdev_config_ops = {
+ .get = virtio_mdev_get,
+ .set = virtio_mdev_set,
+ .generation = virtio_mdev_generation,
+ .get_status = virtio_mdev_get_status,
+ .set_status = virtio_mdev_set_status,
+ .reset = virtio_mdev_reset,
+ .find_vqs = virtio_mdev_find_vqs,
+ .del_vqs = virtio_mdev_del_vqs,
+ .get_features = virtio_mdev_get_features,
+ .finalize_features = virtio_mdev_finalize_features,
+ .bus_name = virtio_mdev_bus_name,
+};
+
+static void virtio_mdev_release_dev(struct device *_d)
+{
+ struct virtio_device *vdev =
+ container_of(_d, struct virtio_device, dev);
+ struct virtio_mdev_device *vm_dev =
+ container_of(vdev, struct virtio_mdev_device, vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ devm_kfree(mdev_dev(mdev), vm_dev);
+}
+
+static int virtio_mdev_probe(struct device *dev)
+{
+ struct mdev_device *mdev = mdev_virtio_from_dev(dev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_device *vm_dev;
+ int rc;
+
+ vm_dev = devm_kzalloc(dev, sizeof(*vm_dev), GFP_KERNEL);
+ if (!vm_dev)
+ return -ENOMEM;
+
+ vm_dev->vdev.dev.parent = dev;
+ vm_dev->vdev.dev.release = virtio_mdev_release_dev;
+ vm_dev->vdev.config = &virtio_mdev_config_ops;
+ vm_dev->mdev = mdev;
+ INIT_LIST_HEAD(&vm_dev->virtqueues);
+ spin_lock_init(&vm_dev->lock);
+
+ vm_dev->vdev.id.device = ops->get_device_id(mdev);
+ if (vm_dev->vdev.id.device == 0)
+ return -ENODEV;
+
+ vm_dev->vdev.id.vendor = ops->get_vendor_id(mdev);
+ rc = register_virtio_device(&vm_dev->vdev);
+ if (rc)
+ put_device(dev);
+ else
+ dev_set_drvdata(dev, vm_dev);
+
+ return rc;
+}
+
+static void virtio_mdev_remove(struct device *dev)
+{
+ struct virtio_mdev_device *vm_dev = dev_get_drvdata(dev);
+
+ unregister_virtio_device(&vm_dev->vdev);
+}
+
+static const struct mdev_virtio_class_id virtio_id_table[] = {
+ { MDEV_VIRTIO_CLASS_ID_VIRTIO },
+ { 0 },
+};
+
+MODULE_DEVICE_TABLE(mdev_virtio, virtio_id_table);
+
+static struct mdev_virtio_driver virtio_mdev_driver = {
+ .drv = {
+ .name = "virtio_mdev",
+ .probe = virtio_mdev_probe,
+ .remove = virtio_mdev_remove,
+ },
+ .id_table = virtio_id_table,
+};
+
+static int __init virtio_mdev_init(void)
+{
+ return mdev_register_driver(&virtio_mdev_driver.drv, THIS_MODULE,
+ &mdev_virtio_bus_type);
+}
+
+static void __exit virtio_mdev_exit(void)
+{
+ mdev_unregister_driver(&virtio_mdev_driver.drv);
+}
+
+module_init(virtio_mdev_init)
+module_exit(virtio_mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
index ef2dbb6c383a..5f75f3cf59e1 100644
--- a/include/linux/mdev_virtio.h
+++ b/include/linux/mdev_virtio.h
@@ -25,6 +25,11 @@ struct virtio_mdev_callback {
void *private;
};
+enum {
+ MDEV_VIRTIO_CLASS_ID_VIRTIO = 1,
+ /* New entries must be added here */
+};
+
/**
* struct mdev_virtio_device_ops - Structure to be registered for each
* mdev device to register the device for virtio/vhost drivers.
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 5/6] virtio: introduce a mdev based transport
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam
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
bus_ops of mdev parent.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
drivers/virtio/Kconfig | 13 ++
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++++++++++++++++++++++++++
include/linux/mdev_virtio.h | 5 +
4 files changed, 428 insertions(+)
create mode 100644 drivers/virtio/virtio_mdev.c
diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index 078615cf2afc..6a89b3de97d3 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
If unsure, say Y.
+config VIRTIO_MDEV
+ tristate "MDEV driver for virtio devices"
+ depends on 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
diff --git a/drivers/virtio/Makefile b/drivers/virtio/Makefile
index 3a2b5c5dcf46..f2997b6c812f 100644
--- a/drivers/virtio/Makefile
+++ b/drivers/virtio/Makefile
@@ -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
diff --git a/drivers/virtio/virtio_mdev.c b/drivers/virtio/virtio_mdev.c
new file mode 100644
index 000000000000..7fdb42f055df
--- /dev/null
+++ b/drivers/virtio/virtio_mdev.c
@@ -0,0 +1,409 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VIRTIO based driver for Mediated device
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uuid.h>
+#include <linux/virtio.h>
+#include <linux/mdev_virtio.h>
+#include <linux/virtio_config.h>
+#include <linux/virtio_ring.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Red Hat Corporation"
+#define DRIVER_DESC "VIRTIO based driver for Mediated device"
+
+#define to_virtio_mdev_device(dev) \
+ container_of(dev, struct virtio_mdev_device, vdev)
+
+struct virtio_mdev_device {
+ struct virtio_device vdev;
+ struct mdev_device *mdev;
+ u64 features;
+
+ /* The lock to protect virtqueue list */
+ spinlock_t lock;
+ /* List of virtio_mdev_vq_info */
+ struct list_head virtqueues;
+};
+
+struct virtio_mdev_vq_info {
+ /* the actual virtqueue */
+ struct virtqueue *vq;
+
+ /* the list node for the virtqueues list */
+ struct list_head node;
+};
+
+static struct mdev_device *vm_get_mdev(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return mdev;
+}
+
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
+ void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->get_config(mdev, offset, buf, len);
+}
+
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
+ const void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->set_config(mdev, offset, buf, len);
+}
+
+static u32 virtio_mdev_generation(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+
+ if (ops->get_generation)
+ return ops->get_generation(mdev);
+
+ return 0;
+}
+
+static u8 virtio_mdev_get_status(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_status(mdev);
+}
+
+static void virtio_mdev_set_status(struct virtio_device *vdev, u8 status)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, status);
+}
+
+static void virtio_mdev_reset(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, 0);
+}
+
+static bool virtio_mdev_notify(struct virtqueue *vq)
+{
+ struct mdev_device *mdev = vm_get_mdev(vq->vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->kick_vq(mdev, vq->index);
+
+ return true;
+}
+
+static irqreturn_t virtio_mdev_config_cb(void *private)
+{
+ struct virtio_mdev_device *vm_dev = private;
+
+ virtio_config_changed(&vm_dev->vdev);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t virtio_mdev_virtqueue_cb(void *private)
+{
+ struct virtio_mdev_vq_info *info = private;
+
+ return vring_interrupt(0, info->vq);
+}
+
+static struct virtqueue *
+virtio_mdev_setup_vq(struct virtio_device *vdev, unsigned int index,
+ void (*callback)(struct virtqueue *vq),
+ const char *name, bool ctx)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info;
+ struct virtio_mdev_callback cb;
+ struct virtqueue *vq;
+ u64 desc_addr, driver_addr, device_addr;
+ unsigned long flags;
+ u32 align, num;
+ int err;
+
+ if (!name)
+ return NULL;
+
+ /* Queue shouldn't already be set up. */
+ if (ops->get_vq_ready(mdev, index))
+ return ERR_PTR(-ENOENT);
+
+ /* Allocate and fill out our active queue description */
+ info = kmalloc(sizeof(*info), GFP_KERNEL);
+ if (!info)
+ return ERR_PTR(-ENOMEM);
+
+ num = ops->get_vq_num_max(mdev);
+ if (num == 0) {
+ err = -ENOENT;
+ goto error_new_virtqueue;
+ }
+
+ /* Create the vring */
+ align = ops->get_vq_align(mdev);
+ vq = vring_create_virtqueue(index, num, align, vdev,
+ true, true, ctx,
+ virtio_mdev_notify, callback, name);
+ if (!vq) {
+ err = -ENOMEM;
+ goto error_new_virtqueue;
+ }
+
+ /* Setup virtqueue callback */
+ cb.callback = virtio_mdev_virtqueue_cb;
+ cb.private = info;
+ ops->set_vq_cb(mdev, index, &cb);
+ ops->set_vq_num(mdev, index, virtqueue_get_vring_size(vq));
+
+ desc_addr = virtqueue_get_desc_addr(vq);
+ driver_addr = virtqueue_get_avail_addr(vq);
+ device_addr = virtqueue_get_used_addr(vq);
+
+ if (ops->set_vq_address(mdev, index,
+ desc_addr, driver_addr,
+ device_addr)) {
+ err = -EINVAL;
+ goto err_vq;
+ }
+
+ ops->set_vq_ready(mdev, index, 1);
+
+ vq->priv = info;
+ info->vq = vq;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_add(&info->node, &vm_dev->virtqueues);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ return vq;
+
+err_vq:
+ vring_del_virtqueue(vq);
+error_new_virtqueue:
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+ kfree(info);
+ return ERR_PTR(err);
+}
+
+static void virtio_mdev_del_vq(struct virtqueue *vq)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vq->vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info = vq->priv;
+ unsigned int index = vq->index;
+ unsigned long flags;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_del(&info->node);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ /* Select and deactivate the queue */
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+
+ vring_del_virtqueue(vq);
+
+ kfree(info);
+}
+
+static void virtio_mdev_del_vqs(struct virtio_device *vdev)
+{
+ struct virtqueue *vq, *n;
+
+ list_for_each_entry_safe(vq, n, &vdev->vqs, list)
+ virtio_mdev_del_vq(vq);
+}
+
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+ struct virtqueue *vqs[],
+ vq_callback_t *callbacks[],
+ const char * const names[],
+ const bool *ctx,
+ struct irq_affinity *desc)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_callback cb;
+ int i, err, queue_idx = 0;
+
+ for (i = 0; i < nvqs; ++i) {
+ if (!names[i]) {
+ vqs[i] = NULL;
+ continue;
+ }
+
+ vqs[i] = virtio_mdev_setup_vq(vdev, queue_idx++,
+ callbacks[i], names[i], ctx ?
+ ctx[i] : false);
+ if (IS_ERR(vqs[i])) {
+ err = PTR_ERR(vqs[i]);
+ goto err_setup_vq;
+ }
+ }
+
+ cb.callback = virtio_mdev_config_cb;
+ cb.private = vm_dev;
+ ops->set_config_cb(mdev, &cb);
+
+ return 0;
+
+err_setup_vq:
+ virtio_mdev_del_vqs(vdev);
+ return err;
+}
+
+static u64 virtio_mdev_get_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_features(mdev);
+}
+
+static int virtio_mdev_finalize_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ /* Give virtio_ring a chance to accept features. */
+ vring_transport_features(vdev);
+
+ return ops->set_features(mdev, vdev->features);
+}
+
+static const char *virtio_mdev_bus_name(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return dev_name(mdev_dev(mdev));
+}
+
+static const struct virtio_config_ops virtio_mdev_config_ops = {
+ .get = virtio_mdev_get,
+ .set = virtio_mdev_set,
+ .generation = virtio_mdev_generation,
+ .get_status = virtio_mdev_get_status,
+ .set_status = virtio_mdev_set_status,
+ .reset = virtio_mdev_reset,
+ .find_vqs = virtio_mdev_find_vqs,
+ .del_vqs = virtio_mdev_del_vqs,
+ .get_features = virtio_mdev_get_features,
+ .finalize_features = virtio_mdev_finalize_features,
+ .bus_name = virtio_mdev_bus_name,
+};
+
+static void virtio_mdev_release_dev(struct device *_d)
+{
+ struct virtio_device *vdev =
+ container_of(_d, struct virtio_device, dev);
+ struct virtio_mdev_device *vm_dev =
+ container_of(vdev, struct virtio_mdev_device, vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ devm_kfree(mdev_dev(mdev), vm_dev);
+}
+
+static int virtio_mdev_probe(struct device *dev)
+{
+ struct mdev_device *mdev = mdev_virtio_from_dev(dev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_device *vm_dev;
+ int rc;
+
+ vm_dev = devm_kzalloc(dev, sizeof(*vm_dev), GFP_KERNEL);
+ if (!vm_dev)
+ return -ENOMEM;
+
+ vm_dev->vdev.dev.parent = dev;
+ vm_dev->vdev.dev.release = virtio_mdev_release_dev;
+ vm_dev->vdev.config = &virtio_mdev_config_ops;
+ vm_dev->mdev = mdev;
+ INIT_LIST_HEAD(&vm_dev->virtqueues);
+ spin_lock_init(&vm_dev->lock);
+
+ vm_dev->vdev.id.device = ops->get_device_id(mdev);
+ if (vm_dev->vdev.id.device == 0)
+ return -ENODEV;
+
+ vm_dev->vdev.id.vendor = ops->get_vendor_id(mdev);
+ rc = register_virtio_device(&vm_dev->vdev);
+ if (rc)
+ put_device(dev);
+ else
+ dev_set_drvdata(dev, vm_dev);
+
+ return rc;
+}
+
+static void virtio_mdev_remove(struct device *dev)
+{
+ struct virtio_mdev_device *vm_dev = dev_get_drvdata(dev);
+
+ unregister_virtio_device(&vm_dev->vdev);
+}
+
+static const struct mdev_virtio_class_id virtio_id_table[] = {
+ { MDEV_VIRTIO_CLASS_ID_VIRTIO },
+ { 0 },
+};
+
+MODULE_DEVICE_TABLE(mdev_virtio, virtio_id_table);
+
+static struct mdev_virtio_driver virtio_mdev_driver = {
+ .drv = {
+ .name = "virtio_mdev",
+ .probe = virtio_mdev_probe,
+ .remove = virtio_mdev_remove,
+ },
+ .id_table = virtio_id_table,
+};
+
+static int __init virtio_mdev_init(void)
+{
+ return mdev_register_driver(&virtio_mdev_driver.drv, THIS_MODULE,
+ &mdev_virtio_bus_type);
+}
+
+static void __exit virtio_mdev_exit(void)
+{
+ mdev_unregister_driver(&virtio_mdev_driver.drv);
+}
+
+module_init(virtio_mdev_init)
+module_exit(virtio_mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
index ef2dbb6c383a..5f75f3cf59e1 100644
--- a/include/linux/mdev_virtio.h
+++ b/include/linux/mdev_virtio.h
@@ -25,6 +25,11 @@ struct virtio_mdev_callback {
void *private;
};
+enum {
+ MDEV_VIRTIO_CLASS_ID_VIRTIO = 1,
+ /* New entries must be added here */
+};
+
/**
* struct mdev_virtio_device_ops - Structure to be registered for each
* mdev device to register the device for virtio/vhost drivers.
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 5/6] virtio: introduce a mdev based transport
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
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
bus_ops of mdev parent.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
drivers/virtio/Kconfig | 13 ++
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++++++++++++++++++++++++++
include/linux/mdev_virtio.h | 5 +
4 files changed, 428 insertions(+)
create mode 100644 drivers/virtio/virtio_mdev.c
diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index 078615cf2afc..6a89b3de97d3 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
If unsure, say Y.
+config VIRTIO_MDEV
+ tristate "MDEV driver for virtio devices"
+ depends on 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
diff --git a/drivers/virtio/Makefile b/drivers/virtio/Makefile
index 3a2b5c5dcf46..f2997b6c812f 100644
--- a/drivers/virtio/Makefile
+++ b/drivers/virtio/Makefile
@@ -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
diff --git a/drivers/virtio/virtio_mdev.c b/drivers/virtio/virtio_mdev.c
new file mode 100644
index 000000000000..7fdb42f055df
--- /dev/null
+++ b/drivers/virtio/virtio_mdev.c
@@ -0,0 +1,409 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VIRTIO based driver for Mediated device
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uuid.h>
+#include <linux/virtio.h>
+#include <linux/mdev_virtio.h>
+#include <linux/virtio_config.h>
+#include <linux/virtio_ring.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Red Hat Corporation"
+#define DRIVER_DESC "VIRTIO based driver for Mediated device"
+
+#define to_virtio_mdev_device(dev) \
+ container_of(dev, struct virtio_mdev_device, vdev)
+
+struct virtio_mdev_device {
+ struct virtio_device vdev;
+ struct mdev_device *mdev;
+ u64 features;
+
+ /* The lock to protect virtqueue list */
+ spinlock_t lock;
+ /* List of virtio_mdev_vq_info */
+ struct list_head virtqueues;
+};
+
+struct virtio_mdev_vq_info {
+ /* the actual virtqueue */
+ struct virtqueue *vq;
+
+ /* the list node for the virtqueues list */
+ struct list_head node;
+};
+
+static struct mdev_device *vm_get_mdev(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return mdev;
+}
+
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
+ void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->get_config(mdev, offset, buf, len);
+}
+
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
+ const void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->set_config(mdev, offset, buf, len);
+}
+
+static u32 virtio_mdev_generation(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+
+ if (ops->get_generation)
+ return ops->get_generation(mdev);
+
+ return 0;
+}
+
+static u8 virtio_mdev_get_status(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_status(mdev);
+}
+
+static void virtio_mdev_set_status(struct virtio_device *vdev, u8 status)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, status);
+}
+
+static void virtio_mdev_reset(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, 0);
+}
+
+static bool virtio_mdev_notify(struct virtqueue *vq)
+{
+ struct mdev_device *mdev = vm_get_mdev(vq->vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->kick_vq(mdev, vq->index);
+
+ return true;
+}
+
+static irqreturn_t virtio_mdev_config_cb(void *private)
+{
+ struct virtio_mdev_device *vm_dev = private;
+
+ virtio_config_changed(&vm_dev->vdev);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t virtio_mdev_virtqueue_cb(void *private)
+{
+ struct virtio_mdev_vq_info *info = private;
+
+ return vring_interrupt(0, info->vq);
+}
+
+static struct virtqueue *
+virtio_mdev_setup_vq(struct virtio_device *vdev, unsigned int index,
+ void (*callback)(struct virtqueue *vq),
+ const char *name, bool ctx)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info;
+ struct virtio_mdev_callback cb;
+ struct virtqueue *vq;
+ u64 desc_addr, driver_addr, device_addr;
+ unsigned long flags;
+ u32 align, num;
+ int err;
+
+ if (!name)
+ return NULL;
+
+ /* Queue shouldn't already be set up. */
+ if (ops->get_vq_ready(mdev, index))
+ return ERR_PTR(-ENOENT);
+
+ /* Allocate and fill out our active queue description */
+ info = kmalloc(sizeof(*info), GFP_KERNEL);
+ if (!info)
+ return ERR_PTR(-ENOMEM);
+
+ num = ops->get_vq_num_max(mdev);
+ if (num == 0) {
+ err = -ENOENT;
+ goto error_new_virtqueue;
+ }
+
+ /* Create the vring */
+ align = ops->get_vq_align(mdev);
+ vq = vring_create_virtqueue(index, num, align, vdev,
+ true, true, ctx,
+ virtio_mdev_notify, callback, name);
+ if (!vq) {
+ err = -ENOMEM;
+ goto error_new_virtqueue;
+ }
+
+ /* Setup virtqueue callback */
+ cb.callback = virtio_mdev_virtqueue_cb;
+ cb.private = info;
+ ops->set_vq_cb(mdev, index, &cb);
+ ops->set_vq_num(mdev, index, virtqueue_get_vring_size(vq));
+
+ desc_addr = virtqueue_get_desc_addr(vq);
+ driver_addr = virtqueue_get_avail_addr(vq);
+ device_addr = virtqueue_get_used_addr(vq);
+
+ if (ops->set_vq_address(mdev, index,
+ desc_addr, driver_addr,
+ device_addr)) {
+ err = -EINVAL;
+ goto err_vq;
+ }
+
+ ops->set_vq_ready(mdev, index, 1);
+
+ vq->priv = info;
+ info->vq = vq;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_add(&info->node, &vm_dev->virtqueues);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ return vq;
+
+err_vq:
+ vring_del_virtqueue(vq);
+error_new_virtqueue:
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+ kfree(info);
+ return ERR_PTR(err);
+}
+
+static void virtio_mdev_del_vq(struct virtqueue *vq)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vq->vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info = vq->priv;
+ unsigned int index = vq->index;
+ unsigned long flags;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_del(&info->node);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ /* Select and deactivate the queue */
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+
+ vring_del_virtqueue(vq);
+
+ kfree(info);
+}
+
+static void virtio_mdev_del_vqs(struct virtio_device *vdev)
+{
+ struct virtqueue *vq, *n;
+
+ list_for_each_entry_safe(vq, n, &vdev->vqs, list)
+ virtio_mdev_del_vq(vq);
+}
+
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+ struct virtqueue *vqs[],
+ vq_callback_t *callbacks[],
+ const char * const names[],
+ const bool *ctx,
+ struct irq_affinity *desc)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_callback cb;
+ int i, err, queue_idx = 0;
+
+ for (i = 0; i < nvqs; ++i) {
+ if (!names[i]) {
+ vqs[i] = NULL;
+ continue;
+ }
+
+ vqs[i] = virtio_mdev_setup_vq(vdev, queue_idx++,
+ callbacks[i], names[i], ctx ?
+ ctx[i] : false);
+ if (IS_ERR(vqs[i])) {
+ err = PTR_ERR(vqs[i]);
+ goto err_setup_vq;
+ }
+ }
+
+ cb.callback = virtio_mdev_config_cb;
+ cb.private = vm_dev;
+ ops->set_config_cb(mdev, &cb);
+
+ return 0;
+
+err_setup_vq:
+ virtio_mdev_del_vqs(vdev);
+ return err;
+}
+
+static u64 virtio_mdev_get_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_features(mdev);
+}
+
+static int virtio_mdev_finalize_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ /* Give virtio_ring a chance to accept features. */
+ vring_transport_features(vdev);
+
+ return ops->set_features(mdev, vdev->features);
+}
+
+static const char *virtio_mdev_bus_name(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return dev_name(mdev_dev(mdev));
+}
+
+static const struct virtio_config_ops virtio_mdev_config_ops = {
+ .get = virtio_mdev_get,
+ .set = virtio_mdev_set,
+ .generation = virtio_mdev_generation,
+ .get_status = virtio_mdev_get_status,
+ .set_status = virtio_mdev_set_status,
+ .reset = virtio_mdev_reset,
+ .find_vqs = virtio_mdev_find_vqs,
+ .del_vqs = virtio_mdev_del_vqs,
+ .get_features = virtio_mdev_get_features,
+ .finalize_features = virtio_mdev_finalize_features,
+ .bus_name = virtio_mdev_bus_name,
+};
+
+static void virtio_mdev_release_dev(struct device *_d)
+{
+ struct virtio_device *vdev =
+ container_of(_d, struct virtio_device, dev);
+ struct virtio_mdev_device *vm_dev =
+ container_of(vdev, struct virtio_mdev_device, vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ devm_kfree(mdev_dev(mdev), vm_dev);
+}
+
+static int virtio_mdev_probe(struct device *dev)
+{
+ struct mdev_device *mdev = mdev_virtio_from_dev(dev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_device *vm_dev;
+ int rc;
+
+ vm_dev = devm_kzalloc(dev, sizeof(*vm_dev), GFP_KERNEL);
+ if (!vm_dev)
+ return -ENOMEM;
+
+ vm_dev->vdev.dev.parent = dev;
+ vm_dev->vdev.dev.release = virtio_mdev_release_dev;
+ vm_dev->vdev.config = &virtio_mdev_config_ops;
+ vm_dev->mdev = mdev;
+ INIT_LIST_HEAD(&vm_dev->virtqueues);
+ spin_lock_init(&vm_dev->lock);
+
+ vm_dev->vdev.id.device = ops->get_device_id(mdev);
+ if (vm_dev->vdev.id.device == 0)
+ return -ENODEV;
+
+ vm_dev->vdev.id.vendor = ops->get_vendor_id(mdev);
+ rc = register_virtio_device(&vm_dev->vdev);
+ if (rc)
+ put_device(dev);
+ else
+ dev_set_drvdata(dev, vm_dev);
+
+ return rc;
+}
+
+static void virtio_mdev_remove(struct device *dev)
+{
+ struct virtio_mdev_device *vm_dev = dev_get_drvdata(dev);
+
+ unregister_virtio_device(&vm_dev->vdev);
+}
+
+static const struct mdev_virtio_class_id virtio_id_table[] = {
+ { MDEV_VIRTIO_CLASS_ID_VIRTIO },
+ { 0 },
+};
+
+MODULE_DEVICE_TABLE(mdev_virtio, virtio_id_table);
+
+static struct mdev_virtio_driver virtio_mdev_driver = {
+ .drv = {
+ .name = "virtio_mdev",
+ .probe = virtio_mdev_probe,
+ .remove = virtio_mdev_remove,
+ },
+ .id_table = virtio_id_table,
+};
+
+static int __init virtio_mdev_init(void)
+{
+ return mdev_register_driver(&virtio_mdev_driver.drv, THIS_MODULE,
+ &mdev_virtio_bus_type);
+}
+
+static void __exit virtio_mdev_exit(void)
+{
+ mdev_unregister_driver(&virtio_mdev_driver.drv);
+}
+
+module_init(virtio_mdev_init)
+module_exit(virtio_mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
index ef2dbb6c383a..5f75f3cf59e1 100644
--- a/include/linux/mdev_virtio.h
+++ b/include/linux/mdev_virtio.h
@@ -25,6 +25,11 @@ struct virtio_mdev_callback {
void *private;
};
+enum {
+ MDEV_VIRTIO_CLASS_ID_VIRTIO = 1,
+ /* New entries must be added here */
+};
+
/**
* struct mdev_virtio_device_ops - Structure to be registered for each
* mdev device to register the device for virtio/vhost drivers.
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 5/6] virtio: introduce a mdev based transport
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
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
bus_ops of mdev parent.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
drivers/virtio/Kconfig | 13 ++
drivers/virtio/Makefile | 1 +
drivers/virtio/virtio_mdev.c | 409 +++++++++++++++++++++++++++++++++++
include/linux/mdev_virtio.h | 5 +
4 files changed, 428 insertions(+)
create mode 100644 drivers/virtio/virtio_mdev.c
diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index 078615cf2afc..6a89b3de97d3 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -43,6 +43,19 @@ config VIRTIO_PCI_LEGACY
If unsure, say Y.
+config VIRTIO_MDEV
+ tristate "MDEV driver for virtio devices"
+ depends on 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
diff --git a/drivers/virtio/Makefile b/drivers/virtio/Makefile
index 3a2b5c5dcf46..f2997b6c812f 100644
--- a/drivers/virtio/Makefile
+++ b/drivers/virtio/Makefile
@@ -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
diff --git a/drivers/virtio/virtio_mdev.c b/drivers/virtio/virtio_mdev.c
new file mode 100644
index 000000000000..7fdb42f055df
--- /dev/null
+++ b/drivers/virtio/virtio_mdev.c
@@ -0,0 +1,409 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VIRTIO based driver for Mediated device
+ *
+ * Copyright (c) 2019, Red Hat. All rights reserved.
+ * Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uuid.h>
+#include <linux/virtio.h>
+#include <linux/mdev_virtio.h>
+#include <linux/virtio_config.h>
+#include <linux/virtio_ring.h>
+
+#define DRIVER_VERSION "0.1"
+#define DRIVER_AUTHOR "Red Hat Corporation"
+#define DRIVER_DESC "VIRTIO based driver for Mediated device"
+
+#define to_virtio_mdev_device(dev) \
+ container_of(dev, struct virtio_mdev_device, vdev)
+
+struct virtio_mdev_device {
+ struct virtio_device vdev;
+ struct mdev_device *mdev;
+ u64 features;
+
+ /* The lock to protect virtqueue list */
+ spinlock_t lock;
+ /* List of virtio_mdev_vq_info */
+ struct list_head virtqueues;
+};
+
+struct virtio_mdev_vq_info {
+ /* the actual virtqueue */
+ struct virtqueue *vq;
+
+ /* the list node for the virtqueues list */
+ struct list_head node;
+};
+
+static struct mdev_device *vm_get_mdev(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return mdev;
+}
+
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
+ void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->get_config(mdev, offset, buf, len);
+}
+
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
+ const void *buf, unsigned len)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->set_config(mdev, offset, buf, len);
+}
+
+static u32 virtio_mdev_generation(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+
+ if (ops->get_generation)
+ return ops->get_generation(mdev);
+
+ return 0;
+}
+
+static u8 virtio_mdev_get_status(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_status(mdev);
+}
+
+static void virtio_mdev_set_status(struct virtio_device *vdev, u8 status)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, status);
+}
+
+static void virtio_mdev_reset(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->set_status(mdev, 0);
+}
+
+static bool virtio_mdev_notify(struct virtqueue *vq)
+{
+ struct mdev_device *mdev = vm_get_mdev(vq->vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ ops->kick_vq(mdev, vq->index);
+
+ return true;
+}
+
+static irqreturn_t virtio_mdev_config_cb(void *private)
+{
+ struct virtio_mdev_device *vm_dev = private;
+
+ virtio_config_changed(&vm_dev->vdev);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t virtio_mdev_virtqueue_cb(void *private)
+{
+ struct virtio_mdev_vq_info *info = private;
+
+ return vring_interrupt(0, info->vq);
+}
+
+static struct virtqueue *
+virtio_mdev_setup_vq(struct virtio_device *vdev, unsigned int index,
+ void (*callback)(struct virtqueue *vq),
+ const char *name, bool ctx)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info;
+ struct virtio_mdev_callback cb;
+ struct virtqueue *vq;
+ u64 desc_addr, driver_addr, device_addr;
+ unsigned long flags;
+ u32 align, num;
+ int err;
+
+ if (!name)
+ return NULL;
+
+ /* Queue shouldn't already be set up. */
+ if (ops->get_vq_ready(mdev, index))
+ return ERR_PTR(-ENOENT);
+
+ /* Allocate and fill out our active queue description */
+ info = kmalloc(sizeof(*info), GFP_KERNEL);
+ if (!info)
+ return ERR_PTR(-ENOMEM);
+
+ num = ops->get_vq_num_max(mdev);
+ if (num == 0) {
+ err = -ENOENT;
+ goto error_new_virtqueue;
+ }
+
+ /* Create the vring */
+ align = ops->get_vq_align(mdev);
+ vq = vring_create_virtqueue(index, num, align, vdev,
+ true, true, ctx,
+ virtio_mdev_notify, callback, name);
+ if (!vq) {
+ err = -ENOMEM;
+ goto error_new_virtqueue;
+ }
+
+ /* Setup virtqueue callback */
+ cb.callback = virtio_mdev_virtqueue_cb;
+ cb.private = info;
+ ops->set_vq_cb(mdev, index, &cb);
+ ops->set_vq_num(mdev, index, virtqueue_get_vring_size(vq));
+
+ desc_addr = virtqueue_get_desc_addr(vq);
+ driver_addr = virtqueue_get_avail_addr(vq);
+ device_addr = virtqueue_get_used_addr(vq);
+
+ if (ops->set_vq_address(mdev, index,
+ desc_addr, driver_addr,
+ device_addr)) {
+ err = -EINVAL;
+ goto err_vq;
+ }
+
+ ops->set_vq_ready(mdev, index, 1);
+
+ vq->priv = info;
+ info->vq = vq;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_add(&info->node, &vm_dev->virtqueues);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ return vq;
+
+err_vq:
+ vring_del_virtqueue(vq);
+error_new_virtqueue:
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+ kfree(info);
+ return ERR_PTR(err);
+}
+
+static void virtio_mdev_del_vq(struct virtqueue *vq)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vq->vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_vq_info *info = vq->priv;
+ unsigned int index = vq->index;
+ unsigned long flags;
+
+ spin_lock_irqsave(&vm_dev->lock, flags);
+ list_del(&info->node);
+ spin_unlock_irqrestore(&vm_dev->lock, flags);
+
+ /* Select and deactivate the queue */
+ ops->set_vq_ready(mdev, index, 0);
+ WARN_ON(ops->get_vq_ready(mdev, index));
+
+ vring_del_virtqueue(vq);
+
+ kfree(info);
+}
+
+static void virtio_mdev_del_vqs(struct virtio_device *vdev)
+{
+ struct virtqueue *vq, *n;
+
+ list_for_each_entry_safe(vq, n, &vdev->vqs, list)
+ virtio_mdev_del_vq(vq);
+}
+
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
+ struct virtqueue *vqs[],
+ vq_callback_t *callbacks[],
+ const char * const names[],
+ const bool *ctx,
+ struct irq_affinity *desc)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_callback cb;
+ int i, err, queue_idx = 0;
+
+ for (i = 0; i < nvqs; ++i) {
+ if (!names[i]) {
+ vqs[i] = NULL;
+ continue;
+ }
+
+ vqs[i] = virtio_mdev_setup_vq(vdev, queue_idx++,
+ callbacks[i], names[i], ctx ?
+ ctx[i] : false);
+ if (IS_ERR(vqs[i])) {
+ err = PTR_ERR(vqs[i]);
+ goto err_setup_vq;
+ }
+ }
+
+ cb.callback = virtio_mdev_config_cb;
+ cb.private = vm_dev;
+ ops->set_config_cb(mdev, &cb);
+
+ return 0;
+
+err_setup_vq:
+ virtio_mdev_del_vqs(vdev);
+ return err;
+}
+
+static u64 virtio_mdev_get_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ return ops->get_features(mdev);
+}
+
+static int virtio_mdev_finalize_features(struct virtio_device *vdev)
+{
+ struct mdev_device *mdev = vm_get_mdev(vdev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+
+ /* Give virtio_ring a chance to accept features. */
+ vring_transport_features(vdev);
+
+ return ops->set_features(mdev, vdev->features);
+}
+
+static const char *virtio_mdev_bus_name(struct virtio_device *vdev)
+{
+ struct virtio_mdev_device *vm_dev = to_virtio_mdev_device(vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ return dev_name(mdev_dev(mdev));
+}
+
+static const struct virtio_config_ops virtio_mdev_config_ops = {
+ .get = virtio_mdev_get,
+ .set = virtio_mdev_set,
+ .generation = virtio_mdev_generation,
+ .get_status = virtio_mdev_get_status,
+ .set_status = virtio_mdev_set_status,
+ .reset = virtio_mdev_reset,
+ .find_vqs = virtio_mdev_find_vqs,
+ .del_vqs = virtio_mdev_del_vqs,
+ .get_features = virtio_mdev_get_features,
+ .finalize_features = virtio_mdev_finalize_features,
+ .bus_name = virtio_mdev_bus_name,
+};
+
+static void virtio_mdev_release_dev(struct device *_d)
+{
+ struct virtio_device *vdev =
+ container_of(_d, struct virtio_device, dev);
+ struct virtio_mdev_device *vm_dev =
+ container_of(vdev, struct virtio_mdev_device, vdev);
+ struct mdev_device *mdev = vm_dev->mdev;
+
+ devm_kfree(mdev_dev(mdev), vm_dev);
+}
+
+static int virtio_mdev_probe(struct device *dev)
+{
+ struct mdev_device *mdev = mdev_virtio_from_dev(dev);
+ const struct mdev_virtio_ops *ops = mdev_virtio_get_ops(mdev);
+ struct virtio_mdev_device *vm_dev;
+ int rc;
+
+ vm_dev = devm_kzalloc(dev, sizeof(*vm_dev), GFP_KERNEL);
+ if (!vm_dev)
+ return -ENOMEM;
+
+ vm_dev->vdev.dev.parent = dev;
+ vm_dev->vdev.dev.release = virtio_mdev_release_dev;
+ vm_dev->vdev.config = &virtio_mdev_config_ops;
+ vm_dev->mdev = mdev;
+ INIT_LIST_HEAD(&vm_dev->virtqueues);
+ spin_lock_init(&vm_dev->lock);
+
+ vm_dev->vdev.id.device = ops->get_device_id(mdev);
+ if (vm_dev->vdev.id.device == 0)
+ return -ENODEV;
+
+ vm_dev->vdev.id.vendor = ops->get_vendor_id(mdev);
+ rc = register_virtio_device(&vm_dev->vdev);
+ if (rc)
+ put_device(dev);
+ else
+ dev_set_drvdata(dev, vm_dev);
+
+ return rc;
+}
+
+static void virtio_mdev_remove(struct device *dev)
+{
+ struct virtio_mdev_device *vm_dev = dev_get_drvdata(dev);
+
+ unregister_virtio_device(&vm_dev->vdev);
+}
+
+static const struct mdev_virtio_class_id virtio_id_table[] = {
+ { MDEV_VIRTIO_CLASS_ID_VIRTIO },
+ { 0 },
+};
+
+MODULE_DEVICE_TABLE(mdev_virtio, virtio_id_table);
+
+static struct mdev_virtio_driver virtio_mdev_driver = {
+ .drv = {
+ .name = "virtio_mdev",
+ .probe = virtio_mdev_probe,
+ .remove = virtio_mdev_remove,
+ },
+ .id_table = virtio_id_table,
+};
+
+static int __init virtio_mdev_init(void)
+{
+ return mdev_register_driver(&virtio_mdev_driver.drv, THIS_MODULE,
+ &mdev_virtio_bus_type);
+}
+
+static void __exit virtio_mdev_exit(void)
+{
+ mdev_unregister_driver(&virtio_mdev_driver.drv);
+}
+
+module_init(virtio_mdev_init)
+module_exit(virtio_mdev_exit)
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
diff --git a/include/linux/mdev_virtio.h b/include/linux/mdev_virtio.h
index ef2dbb6c383a..5f75f3cf59e1 100644
--- a/include/linux/mdev_virtio.h
+++ b/include/linux/mdev_virtio.h
@@ -25,6 +25,11 @@ struct virtio_mdev_callback {
void *private;
};
+enum {
+ MDEV_VIRTIO_CLASS_ID_VIRTIO = 1,
+ /* New entries must be added here */
+};
+
/**
* struct mdev_virtio_device_ops - Structure to be registered for each
* mdev device to register the device for virtio/vhost drivers.
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 10:59 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher, 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.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 1 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
4 files changed, 702 insertions(+)
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
diff --git a/MAINTAINERS b/MAINTAINERS
index e1b57c84f249..36f9fe9034be 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17246,6 +17246,7 @@ F: net/vmw_vsock/virtio_transport.c
F: drivers/net/vsockmon.c
F: drivers/vhost/vsock.c
F: tools/testing/vsock/
+F: samples/vfio-mdev/mvnet_loopback.c
VIRTIO CONSOLE DRIVER
M: Amit Shah <amit@kernel.org>
diff --git a/samples/Kconfig b/samples/Kconfig
index c8dacb4dda80..1bef029cc977 100644
--- a/samples/Kconfig
+++ b/samples/Kconfig
@@ -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_LOOPBACK
+ tristate "Build loopback VIRTIO net example mediated device sample code -- loadable modules only"
+ depends on MDEV_VIRTIO && VHOST_RING && m
+ help
+ Build a networking sample device for use as a virtio
+ mediated device. The device cooperates 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
diff --git a/samples/vfio-mdev/Makefile b/samples/vfio-mdev/Makefile
index 10d179c4fdeb..817618569848 100644
--- a/samples/vfio-mdev/Makefile
+++ b/samples/vfio-mdev/Makefile
@@ -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_LOOPBACK) += mvnet_loopback.o
diff --git a/samples/vfio-mdev/mvnet_loopback.c b/samples/vfio-mdev/mvnet_loopback.c
new file mode 100644
index 000000000000..79059a177f39
--- /dev/null
+++ b/samples/vfio-mdev/mvnet_loopback.c
@@ -0,0 +1,690 @@
+// 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 <jasowang@redhat.com>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet_loopback
+ * # cd /sys/devices/virtual/mvnet_loopback/mvnet_loopback/ \
+ * mdev_supported_types/mvnet_loopback-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.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_loopback"
+#define MVNET_NAME "mvnet_loopback"
+
+#define VIRTIO_MDEV_DEVICE_API_STRING "virtio-mdev"
+
+/*
+ * 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_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_virtio_set_ops(mdev, &mdev_virtio_ops);
+ mdev_virtio_set_class_id(mdev, MDEV_VIRTIO_CLASS_ID_VIRTIO);
+
+ 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_virtio_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_virtio_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_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_virtio_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_virtio_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);
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam
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.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 1 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
4 files changed, 702 insertions(+)
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
diff --git a/MAINTAINERS b/MAINTAINERS
index e1b57c84f249..36f9fe9034be 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17246,6 +17246,7 @@ F: net/vmw_vsock/virtio_transport.c
F: drivers/net/vsockmon.c
F: drivers/vhost/vsock.c
F: tools/testing/vsock/
+F: samples/vfio-mdev/mvnet_loopback.c
VIRTIO CONSOLE DRIVER
M: Amit Shah <amit@kernel.org>
diff --git a/samples/Kconfig b/samples/Kconfig
index c8dacb4dda80..1bef029cc977 100644
--- a/samples/Kconfig
+++ b/samples/Kconfig
@@ -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_LOOPBACK
+ tristate "Build loopback VIRTIO net example mediated device sample code -- loadable modules only"
+ depends on MDEV_VIRTIO && VHOST_RING && m
+ help
+ Build a networking sample device for use as a virtio
+ mediated device. The device cooperates 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
diff --git a/samples/vfio-mdev/Makefile b/samples/vfio-mdev/Makefile
index 10d179c4fdeb..817618569848 100644
--- a/samples/vfio-mdev/Makefile
+++ b/samples/vfio-mdev/Makefile
@@ -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_LOOPBACK) += mvnet_loopback.o
diff --git a/samples/vfio-mdev/mvnet_loopback.c b/samples/vfio-mdev/mvnet_loopback.c
new file mode 100644
index 000000000000..79059a177f39
--- /dev/null
+++ b/samples/vfio-mdev/mvnet_loopback.c
@@ -0,0 +1,690 @@
+// 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 <jasowang@redhat.com>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet_loopback
+ * # cd /sys/devices/virtual/mvnet_loopback/mvnet_loopback/ \
+ * mdev_supported_types/mvnet_loopback-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.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_loopback"
+#define MVNET_NAME "mvnet_loopback"
+
+#define VIRTIO_MDEV_DEVICE_API_STRING "virtio-mdev"
+
+/*
+ * 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_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_virtio_set_ops(mdev, &mdev_virtio_ops);
+ mdev_virtio_set_class_id(mdev, MDEV_VIRTIO_CLASS_ID_VIRTIO);
+
+ 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_virtio_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_virtio_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_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_virtio_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_virtio_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);
--
2.19.1
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, kevin.tian, rob.miller, hch, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
zhi.a.wang, farman, parav, gor, cunming.liang, rodrigo.vivi,
xiao.w.wang, freude, stefanha, zhihong.wang, akrowiak, jiri,
netdev, cohuck, oberpar, maxime.coquelin, aadam, lingshan.zhu
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.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 1 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
4 files changed, 702 insertions(+)
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
diff --git a/MAINTAINERS b/MAINTAINERS
index e1b57c84f249..36f9fe9034be 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17246,6 +17246,7 @@ F: net/vmw_vsock/virtio_transport.c
F: drivers/net/vsockmon.c
F: drivers/vhost/vsock.c
F: tools/testing/vsock/
+F: samples/vfio-mdev/mvnet_loopback.c
VIRTIO CONSOLE DRIVER
M: Amit Shah <amit@kernel.org>
diff --git a/samples/Kconfig b/samples/Kconfig
index c8dacb4dda80..1bef029cc977 100644
--- a/samples/Kconfig
+++ b/samples/Kconfig
@@ -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_LOOPBACK
+ tristate "Build loopback VIRTIO net example mediated device sample code -- loadable modules only"
+ depends on MDEV_VIRTIO && VHOST_RING && m
+ help
+ Build a networking sample device for use as a virtio
+ mediated device. The device cooperates 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
diff --git a/samples/vfio-mdev/Makefile b/samples/vfio-mdev/Makefile
index 10d179c4fdeb..817618569848 100644
--- a/samples/vfio-mdev/Makefile
+++ b/samples/vfio-mdev/Makefile
@@ -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_LOOPBACK) += mvnet_loopback.o
diff --git a/samples/vfio-mdev/mvnet_loopback.c b/samples/vfio-mdev/mvnet_loopback.c
new file mode 100644
index 000000000000..79059a177f39
--- /dev/null
+++ b/samples/vfio-mdev/mvnet_loopback.c
@@ -0,0 +1,690 @@
+// 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 <jasowang@redhat.com>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet_loopback
+ * # cd /sys/devices/virtual/mvnet_loopback/mvnet_loopback/ \
+ * mdev_supported_types/mvnet_loopback-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.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_loopback"
+#define MVNET_NAME "mvnet_loopback"
+
+#define VIRTIO_MDEV_DEVICE_API_STRING "virtio-mdev"
+
+/*
+ * 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_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_virtio_set_ops(mdev, &mdev_virtio_ops);
+ mdev_virtio_set_class_id(mdev, MDEV_VIRTIO_CLASS_ID_VIRTIO);
+
+ 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_virtio_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_virtio_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_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_virtio_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_virtio_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);
--
2.19.1
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 10:59 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-18 10:59 UTC (permalink / raw)
To: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, sebott, airlied,
Jason Wang, heiko.carstens, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, farman,
parav, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, akrowiak, jiri, netdev, cohuck, oberpar,
maxime.coquelin, aadam, lingshan.zhu
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.
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
MAINTAINERS | 1 +
samples/Kconfig | 10 +
samples/vfio-mdev/Makefile | 1 +
samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
4 files changed, 702 insertions(+)
create mode 100644 samples/vfio-mdev/mvnet_loopback.c
diff --git a/MAINTAINERS b/MAINTAINERS
index e1b57c84f249..36f9fe9034be 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17246,6 +17246,7 @@ F: net/vmw_vsock/virtio_transport.c
F: drivers/net/vsockmon.c
F: drivers/vhost/vsock.c
F: tools/testing/vsock/
+F: samples/vfio-mdev/mvnet_loopback.c
VIRTIO CONSOLE DRIVER
M: Amit Shah <amit@kernel.org>
diff --git a/samples/Kconfig b/samples/Kconfig
index c8dacb4dda80..1bef029cc977 100644
--- a/samples/Kconfig
+++ b/samples/Kconfig
@@ -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_LOOPBACK
+ tristate "Build loopback VIRTIO net example mediated device sample code -- loadable modules only"
+ depends on MDEV_VIRTIO && VHOST_RING && m
+ help
+ Build a networking sample device for use as a virtio
+ mediated device. The device cooperates 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
diff --git a/samples/vfio-mdev/Makefile b/samples/vfio-mdev/Makefile
index 10d179c4fdeb..817618569848 100644
--- a/samples/vfio-mdev/Makefile
+++ b/samples/vfio-mdev/Makefile
@@ -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_LOOPBACK) += mvnet_loopback.o
diff --git a/samples/vfio-mdev/mvnet_loopback.c b/samples/vfio-mdev/mvnet_loopback.c
new file mode 100644
index 000000000000..79059a177f39
--- /dev/null
+++ b/samples/vfio-mdev/mvnet_loopback.c
@@ -0,0 +1,690 @@
+// 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 <jasowang@redhat.com>
+ *
+ * Sample driver that creates mdev device that simulates ethernet loopback
+ * device.
+ *
+ * Usage:
+ *
+ * # modprobe virtio_mdev
+ * # modprobe mvnet_loopback
+ * # cd /sys/devices/virtual/mvnet_loopback/mvnet_loopback/ \
+ * mdev_supported_types/mvnet_loopback-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.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_loopback"
+#define MVNET_NAME "mvnet_loopback"
+
+#define VIRTIO_MDEV_DEVICE_API_STRING "virtio-mdev"
+
+/*
+ * 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_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_virtio_set_ops(mdev, &mdev_virtio_ops);
+ mdev_virtio_set_class_id(mdev, MDEV_VIRTIO_CLASS_ID_VIRTIO);
+
+ 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_virtio_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_virtio_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_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_virtio_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_virtio_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);
--
2.19.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply related [flat|nested] 100+ messages in thread
* ✗ Fi.CI.CHECKPATCH: warning for mdev based hardware virtio offloading support
@ 2019-11-18 13:35 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 13:35 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
67f3ea3cda70 mdev: make mdev bus agnostic
-:648: CHECK:LINE_SPACING: Please don't use multiple blank lines
#648: FILE: drivers/vfio/mdev/mdev_private.h:42:
+
+
-:651: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'bus' - possible side-effects?
#651: FILE: drivers/vfio/mdev/mdev_private.h:44:
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
-:651: CHECK:MACRO_ARG_PRECEDENCE: Macro argument 'bus' may be better as '(bus)' to avoid precedence issues
#651: FILE: drivers/vfio/mdev/mdev_private.h:44:
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
-:656: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#656:
new file mode 100644
total: 0 errors, 1 warnings, 3 checks, 776 lines checked
f69d8d9902fd mdev: split out VFIO bus specific parent ops
-:285: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'mdev' - possible side-effects?
#285: FILE: drivers/vfio/mdev/mdev_vfio.c:20:
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
total: 0 errors, 0 warnings, 1 checks, 585 lines checked
8544fcd816bc mdev: move to drivers/
-:57: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#57:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 92 lines checked
fe3f935905d1 mdev: introduce mediated virtio bus
-:57: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#57:
new file mode 100644
-:91: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'mdev' - possible side-effects?
#91: FILE: drivers/mdev/virtio.c:30:
+#define to_mdev_virtio(mdev) container_of(mdev, \
+ struct mdev_virtio_device, mdev)
-:405: CHECK:LINE_SPACING: Please don't use multiple blank lines
#405: FILE: scripts/mod/file2alias.c:1348:
+
+
total: 0 errors, 1 warnings, 2 checks, 361 lines checked
bbb2f0eac428 virtio: introduce a mdev based transport
-:52: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#52:
new file mode 100644
-:111: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#111: FILE: drivers/virtio/virtio_mdev.c:55:
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
-:112: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#112: FILE: drivers/virtio/virtio_mdev.c:56:
+ void *buf, unsigned len)
-:120: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#120: FILE: drivers/virtio/virtio_mdev.c:64:
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
-:121: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#121: FILE: drivers/virtio/virtio_mdev.c:65:
+ const void *buf, unsigned len)
-:134: CHECK:LINE_SPACING: Please don't use multiple blank lines
#134: FILE: drivers/virtio/virtio_mdev.c:78:
+
+
-:302: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#302: FILE: drivers/virtio/virtio_mdev.c:246:
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
total: 0 errors, 6 warnings, 1 checks, 443 lines checked
8a37245a6260 docs: sample driver to demonstrate how to implement virtio-mdev framework
-:62: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#62:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 717 lines checked
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for mdev based hardware virtio offloading support
@ 2019-11-18 13:35 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 13:35 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
67f3ea3cda70 mdev: make mdev bus agnostic
-:648: CHECK:LINE_SPACING: Please don't use multiple blank lines
#648: FILE: drivers/vfio/mdev/mdev_private.h:42:
+
+
-:651: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'bus' - possible side-effects?
#651: FILE: drivers/vfio/mdev/mdev_private.h:44:
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
-:651: CHECK:MACRO_ARG_PRECEDENCE: Macro argument 'bus' may be better as '(bus)' to avoid precedence issues
#651: FILE: drivers/vfio/mdev/mdev_private.h:44:
+#define dev_is_mdev(d, bus) ((d)->bus == bus)
-:656: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#656:
new file mode 100644
total: 0 errors, 1 warnings, 3 checks, 776 lines checked
f69d8d9902fd mdev: split out VFIO bus specific parent ops
-:285: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'mdev' - possible side-effects?
#285: FILE: drivers/vfio/mdev/mdev_vfio.c:20:
+#define to_vfio_mdev_device(mdev) container_of(mdev, \
+ struct mdev_vfio_device, mdev)
total: 0 errors, 0 warnings, 1 checks, 585 lines checked
8544fcd816bc mdev: move to drivers/
-:57: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#57:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 92 lines checked
fe3f935905d1 mdev: introduce mediated virtio bus
-:57: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#57:
new file mode 100644
-:91: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'mdev' - possible side-effects?
#91: FILE: drivers/mdev/virtio.c:30:
+#define to_mdev_virtio(mdev) container_of(mdev, \
+ struct mdev_virtio_device, mdev)
-:405: CHECK:LINE_SPACING: Please don't use multiple blank lines
#405: FILE: scripts/mod/file2alias.c:1348:
+
+
total: 0 errors, 1 warnings, 2 checks, 361 lines checked
bbb2f0eac428 virtio: introduce a mdev based transport
-:52: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#52:
new file mode 100644
-:111: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#111: FILE: drivers/virtio/virtio_mdev.c:55:
+static void virtio_mdev_get(struct virtio_device *vdev, unsigned offset,
-:112: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#112: FILE: drivers/virtio/virtio_mdev.c:56:
+ void *buf, unsigned len)
-:120: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#120: FILE: drivers/virtio/virtio_mdev.c:64:
+static void virtio_mdev_set(struct virtio_device *vdev, unsigned offset,
-:121: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#121: FILE: drivers/virtio/virtio_mdev.c:65:
+ const void *buf, unsigned len)
-:134: CHECK:LINE_SPACING: Please don't use multiple blank lines
#134: FILE: drivers/virtio/virtio_mdev.c:78:
+
+
-:302: WARNING:UNSPECIFIED_INT: Prefer 'unsigned int' to bare use of 'unsigned'
#302: FILE: drivers/virtio/virtio_mdev.c:246:
+static int virtio_mdev_find_vqs(struct virtio_device *vdev, unsigned nvqs,
total: 0 errors, 6 warnings, 1 checks, 443 lines checked
8a37245a6260 docs: sample driver to demonstrate how to implement virtio-mdev framework
-:62: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#62:
new file mode 100644
total: 0 errors, 1 warnings, 0 checks, 717 lines checked
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 13:41 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 13:41 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> +struct bus_type mdev_virtio_bus_type;
> +
> +struct mdev_virtio_device {
> + struct mdev_device mdev;
> + const struct mdev_virtio_ops *ops;
> + u16 class_id;
> +};
This seems to share nothing with mdev (ie mdev-vfio), why is it on the
same bus?
We went over this recently with Greg and he seemed pretty clear on
this..
> +struct mdev_virtio_ops {
> + /* Virtqueue ops */
> + int (*set_vq_address)(struct mdev_device *mdev,
> + u16 idx, u64 desc_area, u64 driver_area,
> + u64 device_area);
> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
> + struct virtio_mdev_callback *cb);
> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
> +
> + /* Device ops */
> + u16 (*get_vq_align)(struct mdev_device *mdev);
> + u64 (*get_features)(struct mdev_device *mdev);
> + int (*set_features)(struct mdev_device *mdev, u64 features);
> + void (*set_config_cb)(struct mdev_device *mdev,
> + struct virtio_mdev_callback *cb);
> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
> + u32 (*get_device_id)(struct mdev_device *mdev);
> + u32 (*get_vendor_id)(struct mdev_device *mdev);
> + u8 (*get_status)(struct mdev_device *mdev);
> + void (*set_status)(struct mdev_device *mdev, u8 status);
> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
> + void *buf, unsigned int len);
> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
> + const void *buf, unsigned int len);
> + u32 (*get_generation)(struct mdev_device *mdev);
> +};
Why aren't all of these 'struct mdev_device_virtio *' ?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 13:41 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 13:41 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
zhi.a.wang, farman, Parav Pandit, gor, intel-gfx,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, hch, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> +struct bus_type mdev_virtio_bus_type;
> +
> +struct mdev_virtio_device {
> + struct mdev_device mdev;
> + const struct mdev_virtio_ops *ops;
> + u16 class_id;
> +};
This seems to share nothing with mdev (ie mdev-vfio), why is it on the
same bus?
We went over this recently with Greg and he seemed pretty clear on
this..
> +struct mdev_virtio_ops {
> + /* Virtqueue ops */
> + int (*set_vq_address)(struct mdev_device *mdev,
> + u16 idx, u64 desc_area, u64 driver_area,
> + u64 device_area);
> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
> + struct virtio_mdev_callback *cb);
> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
> +
> + /* Device ops */
> + u16 (*get_vq_align)(struct mdev_device *mdev);
> + u64 (*get_features)(struct mdev_device *mdev);
> + int (*set_features)(struct mdev_device *mdev, u64 features);
> + void (*set_config_cb)(struct mdev_device *mdev,
> + struct virtio_mdev_callback *cb);
> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
> + u32 (*get_device_id)(struct mdev_device *mdev);
> + u32 (*get_vendor_id)(struct mdev_device *mdev);
> + u8 (*get_status)(struct mdev_device *mdev);
> + void (*set_status)(struct mdev_device *mdev, u8 status);
> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
> + void *buf, unsigned int len);
> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
> + const void *buf, unsigned int len);
> + u32 (*get_generation)(struct mdev_device *mdev);
> +};
Why aren't all of these 'struct mdev_device_virtio *' ?
Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 13:41 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 13:41 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> +struct bus_type mdev_virtio_bus_type;
> +
> +struct mdev_virtio_device {
> + struct mdev_device mdev;
> + const struct mdev_virtio_ops *ops;
> + u16 class_id;
> +};
This seems to share nothing with mdev (ie mdev-vfio), why is it on the
same bus?
We went over this recently with Greg and he seemed pretty clear on
this..
> +struct mdev_virtio_ops {
> + /* Virtqueue ops */
> + int (*set_vq_address)(struct mdev_device *mdev,
> + u16 idx, u64 desc_area, u64 driver_area,
> + u64 device_area);
> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
> + struct virtio_mdev_callback *cb);
> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
> +
> + /* Device ops */
> + u16 (*get_vq_align)(struct mdev_device *mdev);
> + u64 (*get_features)(struct mdev_device *mdev);
> + int (*set_features)(struct mdev_device *mdev, u64 features);
> + void (*set_config_cb)(struct mdev_device *mdev,
> + struct virtio_mdev_callback *cb);
> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
> + u32 (*get_device_id)(struct mdev_device *mdev);
> + u32 (*get_vendor_id)(struct mdev_device *mdev);
> + u8 (*get_status)(struct mdev_device *mdev);
> + void (*set_status)(struct mdev_device *mdev, u8 status);
> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
> + void *buf, unsigned int len);
> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
> + const void *buf, unsigned int len);
> + u32 (*get_generation)(struct mdev_device *mdev);
> +};
Why aren't all of these 'struct mdev_device_virtio *' ?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* ✗ Fi.CI.DOCS: warning for mdev based hardware virtio offloading support
@ 2019-11-18 13:51 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 13:51 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : warning
== Summary ==
$ make htmldocs 2>&1 > /dev/null | grep i915
| | | i915.ko |<-> physical
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* [Intel-gfx] ✗ Fi.CI.DOCS: warning for mdev based hardware virtio offloading support
@ 2019-11-18 13:51 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 13:51 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : warning
== Summary ==
$ make htmldocs 2>&1 > /dev/null | grep i915
| | | i915.ko |<-> physical
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* ✗ Fi.CI.BAT: failure for mdev based hardware virtio offloading support
@ 2019-11-18 14:25 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 14:25 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_7361 -> Patchwork_15313
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with Patchwork_15313 absolutely need to be
verified manually.
If you think the reported changes have nothing to do with the changes
introduced in Patchwork_15313, please notify your bug team to allow them
to document this new failure mode, which will reduce false positives in CI.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/index.html
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_15313:
### IGT changes ###
#### Possible regressions ####
* igt@i915_module_load@reload:
- fi-apl-guc: [PASS][1] -> [DMESG-WARN][2]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-apl-guc/igt@i915_module_load@reload.html
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-apl-guc/igt@i915_module_load@reload.html
Known issues
------------
Here are the changes found in Patchwork_15313 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@i915_pm_rpm@module-reload:
- fi-skl-6770hq: [PASS][3] -> [DMESG-WARN][4] ([fdo#112261])
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-skl-6770hq/igt@i915_pm_rpm@module-reload.html
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-skl-6770hq/igt@i915_pm_rpm@module-reload.html
- fi-skl-lmem: [PASS][5] -> [DMESG-WARN][6] ([fdo#112261])
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-skl-lmem/igt@i915_pm_rpm@module-reload.html
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-skl-lmem/igt@i915_pm_rpm@module-reload.html
* igt@kms_frontbuffer_tracking@basic:
- fi-icl-guc: [PASS][7] -> [FAIL][8] ([fdo#103167])
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-icl-guc/igt@kms_frontbuffer_tracking@basic.html
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-icl-guc/igt@kms_frontbuffer_tracking@basic.html
- fi-hsw-peppy: [PASS][9] -> [DMESG-WARN][10] ([fdo#102614])
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-peppy/igt@kms_frontbuffer_tracking@basic.html
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-peppy/igt@kms_frontbuffer_tracking@basic.html
#### Possible fixes ####
* igt@i915_selftest@live_blt:
- fi-hsw-4770r: [DMESG-FAIL][11] -> [PASS][12]
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-4770r/igt@i915_selftest@live_blt.html
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-4770r/igt@i915_selftest@live_blt.html
- fi-hsw-peppy: [DMESG-FAIL][13] ([fdo#112147]) -> [PASS][14]
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-peppy/igt@i915_selftest@live_blt.html
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-peppy/igt@i915_selftest@live_blt.html
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[fdo#102614]: https://bugs.freedesktop.org/show_bug.cgi?id=102614
[fdo#103167]: https://bugs.freedesktop.org/show_bug.cgi?id=103167
[fdo#109964]: https://bugs.freedesktop.org/show_bug.cgi?id=109964
[fdo#112147]: https://bugs.freedesktop.org/show_bug.cgi?id=112147
[fdo#112260]: https://bugs.freedesktop.org/show_bug.cgi?id=112260
[fdo#112261]: https://bugs.freedesktop.org/show_bug.cgi?id=112261
[fdo#112298]: https://bugs.freedesktop.org/show_bug.cgi?id=112298
Participating hosts (50 -> 44)
------------------------------
Additional (1): fi-icl-y
Missing (7): fi-tgl-u 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_7361 -> Patchwork_15313
CI-20190529: 20190529
CI_DRM_7361: 9dff9fddfefbd9c185b84b30e24a78687dce62c8 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_5292: ea9cd47fdb72c16d5ec84c04a85122c451c30025 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_15313: 8a37245a6260e49cd01d8d2e346a2caf8f7c2462 @ git://anongit.freedesktop.org/gfx-ci/linux
== Linux commits ==
8a37245a6260 docs: sample driver to demonstrate how to implement virtio-mdev framework
bbb2f0eac428 virtio: introduce a mdev based transport
fe3f935905d1 mdev: introduce mediated virtio bus
8544fcd816bc mdev: move to drivers/
f69d8d9902fd mdev: split out VFIO bus specific parent ops
67f3ea3cda70 mdev: make mdev bus agnostic
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/index.html
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* [Intel-gfx] ✗ Fi.CI.BAT: failure for mdev based hardware virtio offloading support
@ 2019-11-18 14:25 ` Patchwork
0 siblings, 0 replies; 100+ messages in thread
From: Patchwork @ 2019-11-18 14:25 UTC (permalink / raw)
To: Jason Wang; +Cc: intel-gfx
== Series Details ==
Series: mdev based hardware virtio offloading support
URL : https://patchwork.freedesktop.org/series/69621/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_7361 -> Patchwork_15313
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with Patchwork_15313 absolutely need to be
verified manually.
If you think the reported changes have nothing to do with the changes
introduced in Patchwork_15313, please notify your bug team to allow them
to document this new failure mode, which will reduce false positives in CI.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/index.html
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_15313:
### IGT changes ###
#### Possible regressions ####
* igt@i915_module_load@reload:
- fi-apl-guc: [PASS][1] -> [DMESG-WARN][2]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-apl-guc/igt@i915_module_load@reload.html
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-apl-guc/igt@i915_module_load@reload.html
Known issues
------------
Here are the changes found in Patchwork_15313 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@i915_pm_rpm@module-reload:
- fi-skl-6770hq: [PASS][3] -> [DMESG-WARN][4] ([fdo#112261])
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-skl-6770hq/igt@i915_pm_rpm@module-reload.html
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-skl-6770hq/igt@i915_pm_rpm@module-reload.html
- fi-skl-lmem: [PASS][5] -> [DMESG-WARN][6] ([fdo#112261])
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-skl-lmem/igt@i915_pm_rpm@module-reload.html
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-skl-lmem/igt@i915_pm_rpm@module-reload.html
* igt@kms_frontbuffer_tracking@basic:
- fi-icl-guc: [PASS][7] -> [FAIL][8] ([fdo#103167])
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-icl-guc/igt@kms_frontbuffer_tracking@basic.html
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-icl-guc/igt@kms_frontbuffer_tracking@basic.html
- fi-hsw-peppy: [PASS][9] -> [DMESG-WARN][10] ([fdo#102614])
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-peppy/igt@kms_frontbuffer_tracking@basic.html
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-peppy/igt@kms_frontbuffer_tracking@basic.html
#### Possible fixes ####
* igt@i915_selftest@live_blt:
- fi-hsw-4770r: [DMESG-FAIL][11] -> [PASS][12]
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-4770r/igt@i915_selftest@live_blt.html
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-4770r/igt@i915_selftest@live_blt.html
- fi-hsw-peppy: [DMESG-FAIL][13] ([fdo#112147]) -> [PASS][14]
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_7361/fi-hsw-peppy/igt@i915_selftest@live_blt.html
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/fi-hsw-peppy/igt@i915_selftest@live_blt.html
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[fdo#102614]: https://bugs.freedesktop.org/show_bug.cgi?id=102614
[fdo#103167]: https://bugs.freedesktop.org/show_bug.cgi?id=103167
[fdo#109964]: https://bugs.freedesktop.org/show_bug.cgi?id=109964
[fdo#112147]: https://bugs.freedesktop.org/show_bug.cgi?id=112147
[fdo#112260]: https://bugs.freedesktop.org/show_bug.cgi?id=112260
[fdo#112261]: https://bugs.freedesktop.org/show_bug.cgi?id=112261
[fdo#112298]: https://bugs.freedesktop.org/show_bug.cgi?id=112298
Participating hosts (50 -> 44)
------------------------------
Additional (1): fi-icl-y
Missing (7): fi-tgl-u 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_7361 -> Patchwork_15313
CI-20190529: 20190529
CI_DRM_7361: 9dff9fddfefbd9c185b84b30e24a78687dce62c8 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_5292: ea9cd47fdb72c16d5ec84c04a85122c451c30025 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_15313: 8a37245a6260e49cd01d8d2e346a2caf8f7c2462 @ git://anongit.freedesktop.org/gfx-ci/linux
== Linux commits ==
8a37245a6260 docs: sample driver to demonstrate how to implement virtio-mdev framework
bbb2f0eac428 virtio: introduce a mdev based transport
fe3f935905d1 mdev: introduce mediated virtio bus
8544fcd816bc mdev: move to drivers/
f69d8d9902fd mdev: split out VFIO bus specific parent ops
67f3ea3cda70 mdev: make mdev bus agnostic
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_15313/index.html
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 15:17 ` Greg KH
-1 siblings, 0 replies; 100+ messages in thread
From: Greg KH @ 2019-11-18 15:17 UTC (permalink / raw)
To: Jason Wang
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie, jgg,
netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher
On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
> +}
We used to have documentation in the kernel source tree that said that
whenever anyone did this, I got to make fun of them. Unfortunately that
has been removed.
Think about what you did right here. You silenced a kernel runtime
warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
doing the above because "I am smarter than the kernel, I will silence it
by putting an empty release function in there."
{sigh}
Did you ever think _why_ we took the time and effort to add that warning
there? It wasn't just so that people can circumvent it, it is to
PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
give people a _chance_ to get things right instead of having you just
live with a silent memory leak.
After 13 versions of this series, basic things like this are still here?
Who is reviewing this thing?
{ugh}
Also, see the other conversations we are having about a "virtual" bus
and devices. I do not want to have two different ways of doing the same
thing in the kernel at the same time please. Please work together with
the Intel developers to solve this in a unified way, as you both
need/want the same thing here.
Neither this, nor the other proposal can be accepted until you all agree
on the design and implementation.
/me goes off to find a nice fruity drink with an umbrella.
greg k-h
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:17 ` Greg KH
0 siblings, 0 replies; 100+ messages in thread
From: Greg KH @ 2019-11-18 15:17 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, haotian.wang, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, zhi.a.wang,
farman, parav, gor, intel-gfx, kevin.tian, alex.williamson,
xiao.w.wang, freude, stefanha, zhihong.wang, rodrigo.vivi,
intel-gvt-dev, hch, akrowiak, jiri, tiw
On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
> +}
We used to have documentation in the kernel source tree that said that
whenever anyone did this, I got to make fun of them. Unfortunately that
has been removed.
Think about what you did right here. You silenced a kernel runtime
warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
doing the above because "I am smarter than the kernel, I will silence it
by putting an empty release function in there."
{sigh}
Did you ever think _why_ we took the time and effort to add that warning
there? It wasn't just so that people can circumvent it, it is to
PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
give people a _chance_ to get things right instead of having you just
live with a silent memory leak.
After 13 versions of this series, basic things like this are still here?
Who is reviewing this thing?
{ugh}
Also, see the other conversations we are having about a "virtual" bus
and devices. I do not want to have two different ways of doing the same
thing in the kernel at the same time please. Please work together with
the Intel developers to solve this in a unified way, as you both
need/want the same thing here.
Neither this, nor the other proposal can be accepted until you all agree
on the design and implementation.
/me goes off to find a nice fruity drink with an umbrella.
greg k-h
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:17 ` Greg KH
0 siblings, 0 replies; 100+ messages in thread
From: Greg KH @ 2019-11-18 15:17 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, haotian.wang, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, zhi.a.wang,
farman, parav, gor, intel-gfx, kevin.tian, alex.williamson,
xiao.w.wang, freude, stefanha, zhihong.wang, rodrigo.vivi,
intel-gvt-dev, hch, akrowiak, jiri, tiwei.bie, netdev, cohuck,
linux-kernel, maxime.coquelin, aadam, lingshan.zhu
On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
> +}
We used to have documentation in the kernel source tree that said that
whenever anyone did this, I got to make fun of them. Unfortunately that
has been removed.
Think about what you did right here. You silenced a kernel runtime
warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
doing the above because "I am smarter than the kernel, I will silence it
by putting an empty release function in there."
{sigh}
Did you ever think _why_ we took the time and effort to add that warning
there? It wasn't just so that people can circumvent it, it is to
PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
give people a _chance_ to get things right instead of having you just
live with a silent memory leak.
After 13 versions of this series, basic things like this are still here?
Who is reviewing this thing?
{ugh}
Also, see the other conversations we are having about a "virtual" bus
and devices. I do not want to have two different ways of doing the same
thing in the kernel at the same time please. Please work together with
the Intel developers to solve this in a unified way, as you both
need/want the same thing here.
Neither this, nor the other proposal can be accepted until you all agree
on the design and implementation.
/me goes off to find a nice fruity drink with an umbrella.
greg k-h
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:17 ` Greg KH
0 siblings, 0 replies; 100+ messages in thread
From: Greg KH @ 2019-11-18 15:17 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, haotian.wang, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, farman,
parav, gor, intel-gfx, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, hch, akrowiak, jiri, tiwei.bie,
netdev, cohuck, linux-kernel, maxime.coquelin, aadam,
lingshan.zhu
On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
> +}
We used to have documentation in the kernel source tree that said that
whenever anyone did this, I got to make fun of them. Unfortunately that
has been removed.
Think about what you did right here. You silenced a kernel runtime
warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
doing the above because "I am smarter than the kernel, I will silence it
by putting an empty release function in there."
{sigh}
Did you ever think _why_ we took the time and effort to add that warning
there? It wasn't just so that people can circumvent it, it is to
PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
give people a _chance_ to get things right instead of having you just
live with a silent memory leak.
After 13 versions of this series, basic things like this are still here?
Who is reviewing this thing?
{ugh}
Also, see the other conversations we are having about a "virtual" bus
and devices. I do not want to have two different ways of doing the same
thing in the kernel at the same time please. Please work together with
the Intel developers to solve this in a unified way, as you both
need/want the same thing here.
Neither this, nor the other proposal can be accepted until you all agree
on the design and implementation.
/me goes off to find a nice fruity drink with an umbrella.
greg k-h
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-18 15:45 ` Cornelia Huck
-1 siblings, 0 replies; 100+ messages in thread
From: Cornelia Huck @ 2019-11-18 15:45 UTC (permalink / raw)
To: Jason Wang
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg, netdev, maxime.coquelin, cunming.liang,
zhihong.wang, rob.miller, xiao.w.wang, haotian.wang, zhenyuw,
zhi.a.wang, jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied,
daniel, farman, pasic, sebott, oberpar, heiko.carstens, gor,
borntraeger, akrowiak, freude, lingshan.zhu, eperezma, lulu,
parav, christophe.de.dinechin, kevin.tian, stefanha, rdunlap,
hch, aadam, jakub.kicinski, jiri, jeffrey.t.kirsher
On Mon, 18 Nov 2019 18:59:23 +0800
Jason Wang <jasowang@redhat.com> wrote:
[Note: I have not looked into the reworked architecture of this *at all*
so far; just something that I noted...]
> 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.
>
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
> MAINTAINERS | 1 +
> samples/Kconfig | 10 +
> samples/vfio-mdev/Makefile | 1 +
> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
> 4 files changed, 702 insertions(+)
> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>
> +static struct mvnet_dev {
> + struct class *vd_class;
> + struct idr vd_idr;
> + struct device dev;
> +} mvnet_dev;
This structure embeds a struct device (a reference-counted structure),
yet it is a static variable. This is giving a bad example to potential
implementers; just allocate it dynamically.
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
And that also means you need a proper release function here, of
course.
> +}
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:45 ` Cornelia Huck
0 siblings, 0 replies; 100+ messages in thread
From: Cornelia Huck @ 2019-11-18 15:45 UTC (permalink / raw)
To: Jason Wang
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg, netdev, maxime.coquelin, cunming.liang,
zhihong.wang, rob.miller, xiao.w.wang, haotian.wang, zhenyuw,
zhi.a.wang, jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied,
daniel, farman, pasic, sebott, oberpar, heiko.carstens, gor
On Mon, 18 Nov 2019 18:59:23 +0800
Jason Wang <jasowang@redhat.com> wrote:
[Note: I have not looked into the reworked architecture of this *at all*
so far; just something that I noted...]
> 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.
>
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
> MAINTAINERS | 1 +
> samples/Kconfig | 10 +
> samples/vfio-mdev/Makefile | 1 +
> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
> 4 files changed, 702 insertions(+)
> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>
> +static struct mvnet_dev {
> + struct class *vd_class;
> + struct idr vd_idr;
> + struct device dev;
> +} mvnet_dev;
This structure embeds a struct device (a reference-counted structure),
yet it is a static variable. This is giving a bad example to potential
implementers; just allocate it dynamically.
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
And that also means you need a proper release function here, of
course.
> +}
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:45 ` Cornelia Huck
0 siblings, 0 replies; 100+ messages in thread
From: Cornelia Huck @ 2019-11-18 15:45 UTC (permalink / raw)
To: Jason Wang
Cc: jakub.kicinski, christophe.de.dinechin, kvm, mst, airlied,
heiko.carstens, haotian.wang, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, zhi.a.wang,
farman, parav, gor, intel-gfx, kevin.tian, alex.williamson,
xiao.w.wang, freude, stefanha, zhihong.wang, rodrigo.vivi,
intel-gvt-dev, hch, akrowiak, aadam, jiri, tiwei.bie, gregkh,
rdunlap, linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On Mon, 18 Nov 2019 18:59:23 +0800
Jason Wang <jasowang@redhat.com> wrote:
[Note: I have not looked into the reworked architecture of this *at all*
so far; just something that I noted...]
> 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.
>
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
> MAINTAINERS | 1 +
> samples/Kconfig | 10 +
> samples/vfio-mdev/Makefile | 1 +
> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
> 4 files changed, 702 insertions(+)
> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>
> +static struct mvnet_dev {
> + struct class *vd_class;
> + struct idr vd_idr;
> + struct device dev;
> +} mvnet_dev;
This structure embeds a struct device (a reference-counted structure),
yet it is a static variable. This is giving a bad example to potential
implementers; just allocate it dynamically.
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
And that also means you need a proper release function here, of
course.
> +}
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-18 15:45 ` Cornelia Huck
0 siblings, 0 replies; 100+ messages in thread
From: Cornelia Huck @ 2019-11-18 15:45 UTC (permalink / raw)
To: Jason Wang
Cc: jakub.kicinski, christophe.de.dinechin, kvm, mst, airlied,
heiko.carstens, haotian.wang, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, farman,
parav, gor, intel-gfx, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam, jiri,
tiwei.bie, gregkh, rdunlap, linux-kernel, maxime.coquelin,
netdev, lingshan.zhu
On Mon, 18 Nov 2019 18:59:23 +0800
Jason Wang <jasowang@redhat.com> wrote:
[Note: I have not looked into the reworked architecture of this *at all*
so far; just something that I noted...]
> 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.
>
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
> MAINTAINERS | 1 +
> samples/Kconfig | 10 +
> samples/vfio-mdev/Makefile | 1 +
> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
> 4 files changed, 702 insertions(+)
> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>
> +static struct mvnet_dev {
> + struct class *vd_class;
> + struct idr vd_idr;
> + struct device dev;
> +} mvnet_dev;
This structure embeds a struct device (a reference-counted structure),
yet it is a static variable. This is giving a bad example to potential
implementers; just allocate it dynamically.
> +static void mvnet_device_release(struct device *dev)
> +{
> + dev_dbg(dev, "mvnet: released\n");
And that also means you need a proper release function here, of
course.
> +}
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:27 ` Michael S. Tsirkin
0 siblings, 0 replies; 100+ messages in thread
From: Michael S. Tsirkin @ 2019-11-18 20:27 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger@de.ibm.com
On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > +struct bus_type mdev_virtio_bus_type;
> > +
> > +struct mdev_virtio_device {
> > + struct mdev_device mdev;
> > + const struct mdev_virtio_ops *ops;
> > + u16 class_id;
> > +};
>
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
I must be missing something - which bus do they share?
--
MST
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:27 ` Michael S. Tsirkin
0 siblings, 0 replies; 100+ messages in thread
From: Michael S. Tsirkin @ 2019-11-18 20:27 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, kevin.tian, dri-devel,
oberpar, kwankhede, rob.miller, linux-s390, sebott, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
intel-gfx, zhi.a.wang, farman, Parav Pandit, gor, cunming.liang,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > +struct bus_type mdev_virtio_bus_type;
> > +
> > +struct mdev_virtio_device {
> > + struct mdev_device mdev;
> > + const struct mdev_virtio_ops *ops;
> > + u16 class_id;
> > +};
>
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
I must be missing something - which bus do they share?
--
MST
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:27 ` Michael S. Tsirkin
0 siblings, 0 replies; 100+ messages in thread
From: Michael S. Tsirkin @ 2019-11-18 20:27 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, intel-gfx, farman,
Parav Pandit, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > +struct bus_type mdev_virtio_bus_type;
> > +
> > +struct mdev_virtio_device {
> > + struct mdev_device mdev;
> > + const struct mdev_virtio_ops *ops;
> > + u16 class_id;
> > +};
>
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
I must be missing something - which bus do they share?
--
MST
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:28 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 20:28 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger@de.ibm.com
On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > +struct bus_type mdev_virtio_bus_type;
> > > +
> > > +struct mdev_virtio_device {
> > > + struct mdev_device mdev;
> > > + const struct mdev_virtio_ops *ops;
> > > + u16 class_id;
> > > +};
> >
> > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > same bus?
>
> I must be missing something - which bus do they share?
mdev_bus_type ?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:28 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 20:28 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, kevin.tian, dri-devel,
oberpar, kwankhede, rob.miller, linux-s390, sebott, lulu,
eperezma, pasic, borntraeger, haotian.wang, jeffrey.t.kirsher,
intel-gfx, zhi.a.wang, farman, Parav Pandit, gor, cunming.liang,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > +struct bus_type mdev_virtio_bus_type;
> > > +
> > > +struct mdev_virtio_device {
> > > + struct mdev_device mdev;
> > > + const struct mdev_virtio_ops *ops;
> > > + u16 class_id;
> > > +};
> >
> > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > same bus?
>
> I must be missing something - which bus do they share?
mdev_bus_type ?
Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-18 20:28 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-18 20:28 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, Jason Wang, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, intel-gfx, farman,
Parav Pandit, gor, cunming.liang, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > +struct bus_type mdev_virtio_bus_type;
> > > +
> > > +struct mdev_virtio_device {
> > > + struct mdev_device mdev;
> > > + const struct mdev_virtio_ops *ops;
> > > + u16 class_id;
> > > +};
> >
> > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > same bus?
>
> I must be missing something - which bus do they share?
mdev_bus_type ?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:40 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:40 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger
On 2019/11/18 下午9:41, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>> +struct bus_type mdev_virtio_bus_type;
>> +
>> +struct mdev_virtio_device {
>> + struct mdev_device mdev;
>> + const struct mdev_virtio_ops *ops;
>> + u16 class_id;
>> +};
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
>
> We went over this recently with Greg and he seemed pretty clear on
> this..
Mdev-vfio is not on this bus. The class_id here is used for distinguish
userspace virtio driver (vhost-mdev) and kernel virtio driver (virtio-mdev).
Parent can choose to create a type of "vhost" device then vhost-mdev
driver is matched, or "virtio" device then virtio-mdev driver is matched.
>
>> +struct mdev_virtio_ops {
>> + /* Virtqueue ops */
>> + int (*set_vq_address)(struct mdev_device *mdev,
>> + u16 idx, u64 desc_area, u64 driver_area,
>> + u64 device_area);
>> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
>> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
>> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
>> + struct virtio_mdev_callback *cb);
>> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
>> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
>> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
>> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
>> +
>> + /* Device ops */
>> + u16 (*get_vq_align)(struct mdev_device *mdev);
>> + u64 (*get_features)(struct mdev_device *mdev);
>> + int (*set_features)(struct mdev_device *mdev, u64 features);
>> + void (*set_config_cb)(struct mdev_device *mdev,
>> + struct virtio_mdev_callback *cb);
>> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
>> + u32 (*get_device_id)(struct mdev_device *mdev);
>> + u32 (*get_vendor_id)(struct mdev_device *mdev);
>> + u8 (*get_status)(struct mdev_device *mdev);
>> + void (*set_status)(struct mdev_device *mdev, u8 status);
>> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
>> + void *buf, unsigned int len);
>> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
>> + const void *buf, unsigned int len);
>> + u32 (*get_generation)(struct mdev_device *mdev);
>> +};
> Why aren't all of these 'struct mdev_device_virtio *' ?
>
> Jason
It can simplify the assignment of those ops in mdev device implementation.
Thanks
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:40 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:40 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
zhi.a.wang, farman, Parav Pandit, gor, intel-gfx,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, hch, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On 2019/11/18 下午9:41, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>> +struct bus_type mdev_virtio_bus_type;
>> +
>> +struct mdev_virtio_device {
>> + struct mdev_device mdev;
>> + const struct mdev_virtio_ops *ops;
>> + u16 class_id;
>> +};
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
>
> We went over this recently with Greg and he seemed pretty clear on
> this..
Mdev-vfio is not on this bus. The class_id here is used for distinguish
userspace virtio driver (vhost-mdev) and kernel virtio driver (virtio-mdev).
Parent can choose to create a type of "vhost" device then vhost-mdev
driver is matched, or "virtio" device then virtio-mdev driver is matched.
>
>> +struct mdev_virtio_ops {
>> + /* Virtqueue ops */
>> + int (*set_vq_address)(struct mdev_device *mdev,
>> + u16 idx, u64 desc_area, u64 driver_area,
>> + u64 device_area);
>> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
>> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
>> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
>> + struct virtio_mdev_callback *cb);
>> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
>> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
>> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
>> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
>> +
>> + /* Device ops */
>> + u16 (*get_vq_align)(struct mdev_device *mdev);
>> + u64 (*get_features)(struct mdev_device *mdev);
>> + int (*set_features)(struct mdev_device *mdev, u64 features);
>> + void (*set_config_cb)(struct mdev_device *mdev,
>> + struct virtio_mdev_callback *cb);
>> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
>> + u32 (*get_device_id)(struct mdev_device *mdev);
>> + u32 (*get_vendor_id)(struct mdev_device *mdev);
>> + u8 (*get_status)(struct mdev_device *mdev);
>> + void (*set_status)(struct mdev_device *mdev, u8 status);
>> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
>> + void *buf, unsigned int len);
>> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
>> + const void *buf, unsigned int len);
>> + u32 (*get_generation)(struct mdev_device *mdev);
>> +};
> Why aren't all of these 'struct mdev_device_virtio *' ?
>
> Jason
It can simplify the assignment of those ops in mdev device implementation.
Thanks
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:40 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:40 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/18 下午9:41, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>> +struct bus_type mdev_virtio_bus_type;
>> +
>> +struct mdev_virtio_device {
>> + struct mdev_device mdev;
>> + const struct mdev_virtio_ops *ops;
>> + u16 class_id;
>> +};
> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> same bus?
>
> We went over this recently with Greg and he seemed pretty clear on
> this..
Mdev-vfio is not on this bus. The class_id here is used for distinguish
userspace virtio driver (vhost-mdev) and kernel virtio driver (virtio-mdev).
Parent can choose to create a type of "vhost" device then vhost-mdev
driver is matched, or "virtio" device then virtio-mdev driver is matched.
>
>> +struct mdev_virtio_ops {
>> + /* Virtqueue ops */
>> + int (*set_vq_address)(struct mdev_device *mdev,
>> + u16 idx, u64 desc_area, u64 driver_area,
>> + u64 device_area);
>> + void (*set_vq_num)(struct mdev_device *mdev, u16 idx, u32 num);
>> + void (*kick_vq)(struct mdev_device *mdev, u16 idx);
>> + void (*set_vq_cb)(struct mdev_device *mdev, u16 idx,
>> + struct virtio_mdev_callback *cb);
>> + void (*set_vq_ready)(struct mdev_device *mdev, u16 idx, bool ready);
>> + bool (*get_vq_ready)(struct mdev_device *mdev, u16 idx);
>> + int (*set_vq_state)(struct mdev_device *mdev, u16 idx, u64 state);
>> + u64 (*get_vq_state)(struct mdev_device *mdev, u16 idx);
>> +
>> + /* Device ops */
>> + u16 (*get_vq_align)(struct mdev_device *mdev);
>> + u64 (*get_features)(struct mdev_device *mdev);
>> + int (*set_features)(struct mdev_device *mdev, u64 features);
>> + void (*set_config_cb)(struct mdev_device *mdev,
>> + struct virtio_mdev_callback *cb);
>> + u16 (*get_vq_num_max)(struct mdev_device *mdev);
>> + u32 (*get_device_id)(struct mdev_device *mdev);
>> + u32 (*get_vendor_id)(struct mdev_device *mdev);
>> + u8 (*get_status)(struct mdev_device *mdev);
>> + void (*set_status)(struct mdev_device *mdev, u8 status);
>> + void (*get_config)(struct mdev_device *mdev, unsigned int offset,
>> + void *buf, unsigned int len);
>> + void (*set_config)(struct mdev_device *mdev, unsigned int offset,
>> + const void *buf, unsigned int len);
>> + u32 (*get_generation)(struct mdev_device *mdev);
>> +};
> Why aren't all of these 'struct mdev_device_virtio *' ?
>
> Jason
It can simplify the assignment of those ops in mdev device implementation.
Thanks
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:41 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:41 UTC (permalink / raw)
To: Jason Gunthorpe, Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>> +struct bus_type mdev_virtio_bus_type;
>>>> +
>>>> +struct mdev_virtio_device {
>>>> + struct mdev_device mdev;
>>>> + const struct mdev_virtio_ops *ops;
>>>> + u16 class_id;
>>>> +};
>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>> same bus?
>> I must be missing something - which bus do they share?
> mdev_bus_type ?
>
> Jason
Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the
same bus.
Thanks
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:41 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:41 UTC (permalink / raw)
To: Jason Gunthorpe, Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
zhi.a.wang, farman, Parav Pandit, gor, intel-gfx,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, akrowiak, aadam, Jiri Pirko,
tiwei.bie, gregkh, cohuck, linux-kernel, maxime.coquelin, netdev,
lingshan.zhu
On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>> +struct bus_type mdev_virtio_bus_type;
>>>> +
>>>> +struct mdev_virtio_device {
>>>> + struct mdev_device mdev;
>>>> + const struct mdev_virtio_ops *ops;
>>>> + u16 class_id;
>>>> +};
>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>> same bus?
>> I must be missing something - which bus do they share?
> mdev_bus_type ?
>
> Jason
Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the
same bus.
Thanks
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 2:41 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 2:41 UTC (permalink / raw)
To: Jason Gunthorpe, Michael S. Tsirkin
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, hch,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>> +struct bus_type mdev_virtio_bus_type;
>>>> +
>>>> +struct mdev_virtio_device {
>>>> + struct mdev_device mdev;
>>>> + const struct mdev_virtio_ops *ops;
>>>> + u16 class_id;
>>>> +};
>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>> same bus?
>> I must be missing something - which bus do they share?
> mdev_bus_type ?
>
> Jason
Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the
same bus.
Thanks
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
2019-11-18 15:17 ` Greg KH
(?)
(?)
@ 2019-11-19 3:03 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:03 UTC (permalink / raw)
To: Greg KH
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie, jgg,
netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, rdunlap, hch,
aadam, jakub.kicinski, jiri, jeffrey.t.kirsher
On 2019/11/18 下午11:17, Greg KH wrote:
> On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
>> +}
> We used to have documentation in the kernel source tree that said that
> whenever anyone did this, I got to make fun of them. Unfortunately that
> has been removed.
>
> Think about what you did right here. You silenced a kernel runtime
> warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
> doing the above because "I am smarter than the kernel, I will silence it
> by putting an empty release function in there."
>
> {sigh}
>
> Did you ever think _why_ we took the time and effort to add that warning
> there? It wasn't just so that people can circumvent it, it is to
> PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
> give people a _chance_ to get things right instead of having you just
> live with a silent memory leak.
>
> After 13 versions of this series, basic things like this are still here?
> Who is reviewing this thing?
Apologize that static structure is used here, will fix them with dynamic
one. I just borrow the codes from other vfio-mdev samples without too
much thought here ...
>
> {ugh}
>
> Also, see the other conversations we are having about a "virtual" bus
> and devices. I do not want to have two different ways of doing the same
> thing in the kernel at the same time please. Please work together with
> the Intel developers to solve this in a unified way, as you both
> need/want the same thing here.
Sure, some functions looks similar, but the "virtual" bus does not
contain a management interface and it's not clear that how it can be
used by userspace driver. For this series, sysfs/GUID based management
interface is reused and we had a concrete example of how it would be
used by userspace driver[1] and a real hardware driver implementation[2].
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
>
> Neither this, nor the other proposal can be accepted until you all agree
> on the design and implementation.
Yes.
Thanks
>
> /me goes off to find a nice fruity drink with an umbrella.
>
> greg k-h
>
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:03 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:03 UTC (permalink / raw)
To: Greg KH
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie, jgg,
netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor
On 2019/11/18 下午11:17, Greg KH wrote:
> On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
>> +}
> We used to have documentation in the kernel source tree that said that
> whenever anyone did this, I got to make fun of them. Unfortunately that
> has been removed.
>
> Think about what you did right here. You silenced a kernel runtime
> warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
> doing the above because "I am smarter than the kernel, I will silence it
> by putting an empty release function in there."
>
> {sigh}
>
> Did you ever think _why_ we took the time and effort to add that warning
> there? It wasn't just so that people can circumvent it, it is to
> PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
> give people a _chance_ to get things right instead of having you just
> live with a silent memory leak.
>
> After 13 versions of this series, basic things like this are still here?
> Who is reviewing this thing?
Apologize that static structure is used here, will fix them with dynamic
one. I just borrow the codes from other vfio-mdev samples without too
much thought here ...
>
> {ugh}
>
> Also, see the other conversations we are having about a "virtual" bus
> and devices. I do not want to have two different ways of doing the same
> thing in the kernel at the same time please. Please work together with
> the Intel developers to solve this in a unified way, as you both
> need/want the same thing here.
Sure, some functions looks similar, but the "virtual" bus does not
contain a management interface and it's not clear that how it can be
used by userspace driver. For this series, sysfs/GUID based management
interface is reused and we had a concrete example of how it would be
used by userspace driver[1] and a real hardware driver implementation[2].
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
>
> Neither this, nor the other proposal can be accepted until you all agree
> on the design and implementation.
Yes.
Thanks
>
> /me goes off to find a nice fruity drink with an umbrella.
>
> greg k-h
>
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:03 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:03 UTC (permalink / raw)
To: Greg KH
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, haotian.wang, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, zhi.a.wang,
farman, parav, gor, intel-gfx, kevin.tian, alex.williamson,
xiao.w.wang, freude, stefanha, zhihong.wang, rodrigo.vivi,
intel-gvt-dev, hch, akrowiak, jiri, tiwei.bie, netdev, cohuck,
linux-kernel, maxime.coquelin, aadam, lingshan.zhu
On 2019/11/18 下午11:17, Greg KH wrote:
> On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
>> +}
> We used to have documentation in the kernel source tree that said that
> whenever anyone did this, I got to make fun of them. Unfortunately that
> has been removed.
>
> Think about what you did right here. You silenced a kernel runtime
> warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
> doing the above because "I am smarter than the kernel, I will silence it
> by putting an empty release function in there."
>
> {sigh}
>
> Did you ever think _why_ we took the time and effort to add that warning
> there? It wasn't just so that people can circumvent it, it is to
> PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
> give people a _chance_ to get things right instead of having you just
> live with a silent memory leak.
>
> After 13 versions of this series, basic things like this are still here?
> Who is reviewing this thing?
Apologize that static structure is used here, will fix them with dynamic
one. I just borrow the codes from other vfio-mdev samples without too
much thought here ...
>
> {ugh}
>
> Also, see the other conversations we are having about a "virtual" bus
> and devices. I do not want to have two different ways of doing the same
> thing in the kernel at the same time please. Please work together with
> the Intel developers to solve this in a unified way, as you both
> need/want the same thing here.
Sure, some functions looks similar, but the "virtual" bus does not
contain a management interface and it's not clear that how it can be
used by userspace driver. For this series, sysfs/GUID based management
interface is reused and we had a concrete example of how it would be
used by userspace driver[1] and a real hardware driver implementation[2].
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
>
> Neither this, nor the other proposal can be accepted until you all agree
> on the design and implementation.
Yes.
Thanks
>
> /me goes off to find a nice fruity drink with an umbrella.
>
> greg k-h
>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:03 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:03 UTC (permalink / raw)
To: Greg KH
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, haotian.wang, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, farman,
parav, gor, intel-gfx, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, hch, akrowiak, jiri, tiwei.bie,
netdev, cohuck, linux-kernel, maxime.coquelin, aadam,
lingshan.zhu
On 2019/11/18 下午11:17, Greg KH wrote:
> On Mon, Nov 18, 2019 at 06:59:23PM +0800, Jason Wang wrote:
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
>> +}
> We used to have documentation in the kernel source tree that said that
> whenever anyone did this, I got to make fun of them. Unfortunately that
> has been removed.
>
> Think about what you did right here. You silenced a kernel runtime
> warning that said something like "ERROR! NO RELEASE FUNCTION FOUND!" by
> doing the above because "I am smarter than the kernel, I will silence it
> by putting an empty release function in there."
>
> {sigh}
>
> Did you ever think _why_ we took the time and effort to add that warning
> there? It wasn't just so that people can circumvent it, it is to
> PREVENT A MAJOR BUG IN YOUR DESIGN! We are trying to be nice here and
> give people a _chance_ to get things right instead of having you just
> live with a silent memory leak.
>
> After 13 versions of this series, basic things like this are still here?
> Who is reviewing this thing?
Apologize that static structure is used here, will fix them with dynamic
one. I just borrow the codes from other vfio-mdev samples without too
much thought here ...
>
> {ugh}
>
> Also, see the other conversations we are having about a "virtual" bus
> and devices. I do not want to have two different ways of doing the same
> thing in the kernel at the same time please. Please work together with
> the Intel developers to solve this in a unified way, as you both
> need/want the same thing here.
Sure, some functions looks similar, but the "virtual" bus does not
contain a management interface and it's not clear that how it can be
used by userspace driver. For this series, sysfs/GUID based management
interface is reused and we had a concrete example of how it would be
used by userspace driver[1] and a real hardware driver implementation[2].
[1] https://lkml.org/lkml/2019/11/7/62
[2] https://lkml.org/lkml/2019/11/12/215
>
> Neither this, nor the other proposal can be accepted until you all agree
> on the design and implementation.
Yes.
Thanks
>
> /me goes off to find a nice fruity drink with an umbrella.
>
> greg k-h
>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
2019-11-18 15:45 ` Cornelia Huck
(?)
(?)
@ 2019-11-19 3:04 ` Jason Wang
-1 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:04 UTC (permalink / raw)
To: Cornelia Huck
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg, netdev, maxime.coquelin, cunming.liang,
zhihong.wang, rob.miller, xiao.w.wang, haotian.wang, zhenyuw,
zhi.a.wang, jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied,
daniel, farman, pasic, sebott, oberpar, heiko.carstens, gor,
borntraeger, akrowiak, freude, lingshan.zhu, eperezma, lulu,
parav, christophe.de.dinechin, kevin.tian, stefanha, rdunlap,
hch, aadam, jakub.kicinski, jiri, jeffrey.t.kirsher
On 2019/11/18 下午11:45, Cornelia Huck wrote:
> On Mon, 18 Nov 2019 18:59:23 +0800
> Jason Wang <jasowang@redhat.com> wrote:
>
> [Note: I have not looked into the reworked architecture of this *at all*
> so far; just something that I noted...]
>
>> 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.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>> MAINTAINERS | 1 +
>> samples/Kconfig | 10 +
>> samples/vfio-mdev/Makefile | 1 +
>> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
>> 4 files changed, 702 insertions(+)
>> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>>
>> +static struct mvnet_dev {
>> + struct class *vd_class;
>> + struct idr vd_idr;
>> + struct device dev;
>> +} mvnet_dev;
> This structure embeds a struct device (a reference-counted structure),
> yet it is a static variable. This is giving a bad example to potential
> implementers; just allocate it dynamically.
Yes, as spotted by Greg.
>
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
> And that also means you need a proper release function here, of
> course.
Right.
Thanks
>
>> +}
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:04 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:04 UTC (permalink / raw)
To: Cornelia Huck
Cc: kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg, netdev, maxime.coquelin, cunming.liang,
zhihong.wang, rob.miller, xiao.w.wang, haotian.wang, zhenyuw,
zhi.a.wang, jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied,
daniel, farman, pasic, sebott, oberpar, heiko.carstens, gor
On 2019/11/18 下午11:45, Cornelia Huck wrote:
> On Mon, 18 Nov 2019 18:59:23 +0800
> Jason Wang <jasowang@redhat.com> wrote:
>
> [Note: I have not looked into the reworked architecture of this *at all*
> so far; just something that I noted...]
>
>> 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.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>> MAINTAINERS | 1 +
>> samples/Kconfig | 10 +
>> samples/vfio-mdev/Makefile | 1 +
>> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
>> 4 files changed, 702 insertions(+)
>> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>>
>> +static struct mvnet_dev {
>> + struct class *vd_class;
>> + struct idr vd_idr;
>> + struct device dev;
>> +} mvnet_dev;
> This structure embeds a struct device (a reference-counted structure),
> yet it is a static variable. This is giving a bad example to potential
> implementers; just allocate it dynamically.
Yes, as spotted by Greg.
>
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
> And that also means you need a proper release function here, of
> course.
Right.
Thanks
>
>> +}
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:04 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:04 UTC (permalink / raw)
To: Cornelia Huck
Cc: jakub.kicinski, christophe.de.dinechin, kvm, mst, airlied,
heiko.carstens, haotian.wang, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, zhi.a.wang,
farman, parav, gor, intel-gfx, kevin.tian, alex.williamson,
xiao.w.wang, freude, stefanha, zhihong.wang, rodrigo.vivi,
intel-gvt-dev, hch, akrowiak, aadam, jiri, tiwei.bie, gregkh,
rdunlap, linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On 2019/11/18 下午11:45, Cornelia Huck wrote:
> On Mon, 18 Nov 2019 18:59:23 +0800
> Jason Wang <jasowang@redhat.com> wrote:
>
> [Note: I have not looked into the reworked architecture of this *at all*
> so far; just something that I noted...]
>
>> 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.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>> MAINTAINERS | 1 +
>> samples/Kconfig | 10 +
>> samples/vfio-mdev/Makefile | 1 +
>> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
>> 4 files changed, 702 insertions(+)
>> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>>
>> +static struct mvnet_dev {
>> + struct class *vd_class;
>> + struct idr vd_idr;
>> + struct device dev;
>> +} mvnet_dev;
> This structure embeds a struct device (a reference-counted structure),
> yet it is a static variable. This is giving a bad example to potential
> implementers; just allocate it dynamically.
Yes, as spotted by Greg.
>
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
> And that also means you need a proper release function here, of
> course.
Right.
Thanks
>
>> +}
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 3:04 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 3:04 UTC (permalink / raw)
To: Cornelia Huck
Cc: jakub.kicinski, christophe.de.dinechin, kvm, mst, airlied,
heiko.carstens, haotian.wang, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, jgg, jeffrey.t.kirsher, cunming.liang, farman,
parav, gor, intel-gfx, xiao.w.wang, freude, stefanha,
zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam, jiri,
tiwei.bie, gregkh, rdunlap, linux-kernel, maxime.coquelin,
netdev, lingshan.zhu
On 2019/11/18 下午11:45, Cornelia Huck wrote:
> On Mon, 18 Nov 2019 18:59:23 +0800
> Jason Wang <jasowang@redhat.com> wrote:
>
> [Note: I have not looked into the reworked architecture of this *at all*
> so far; just something that I noted...]
>
>> 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.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>> MAINTAINERS | 1 +
>> samples/Kconfig | 10 +
>> samples/vfio-mdev/Makefile | 1 +
>> samples/vfio-mdev/mvnet_loopback.c | 690 +++++++++++++++++++++++++++++
>> 4 files changed, 702 insertions(+)
>> create mode 100644 samples/vfio-mdev/mvnet_loopback.c
>>
>> +static struct mvnet_dev {
>> + struct class *vd_class;
>> + struct idr vd_idr;
>> + struct device dev;
>> +} mvnet_dev;
> This structure embeds a struct device (a reference-counted structure),
> yet it is a static variable. This is giving a bad example to potential
> implementers; just allocate it dynamically.
Yes, as spotted by Greg.
>
>> +static void mvnet_device_release(struct device *dev)
>> +{
>> + dev_dbg(dev, "mvnet: released\n");
> And that also means you need a proper release function here, of
> course.
Right.
Thanks
>
>> +}
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 1/6] mdev: make mdev bus agnostic
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-19 3:08 ` Randy Dunlap
-1 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:08 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, hch, aadam,
jakub.kicinski, jiri, jeffrey.t.kirsher
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
> index 5da27f2100f9..2e07ca915a96 100644
> --- a/drivers/vfio/mdev/Kconfig
> +++ b/drivers/vfio/mdev/Kconfig
> @@ -1,15 +1,24 @@
> -# SPDX-License-Identifier: GPL-2.0-only
>
> -config VFIO_MDEV
> +config MDEV
> tristate "Mediated device driver framework"
> - depends on VFIO
> default n
> help
> Provides a framework to virtualize devices.
> - See Documentation/driver-api/vfio-mediated-device.rst for more details.
>
> If you don't know what do here, say N.
>
Hi,
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
The depends and default lines should be indented with tab, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
> +
> +
> config VFIO_MDEV_DEVICE
> tristate "VFIO driver for Mediated devices"
> depends on VFIO && VFIO_MDEV
--
~Randy
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 1/6] mdev: make mdev bus agnostic
@ 2019-11-19 3:08 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:08 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, hch, aadam,
jakub.kicinski
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
> index 5da27f2100f9..2e07ca915a96 100644
> --- a/drivers/vfio/mdev/Kconfig
> +++ b/drivers/vfio/mdev/Kconfig
> @@ -1,15 +1,24 @@
> -# SPDX-License-Identifier: GPL-2.0-only
>
> -config VFIO_MDEV
> +config MDEV
> tristate "Mediated device driver framework"
> - depends on VFIO
> default n
> help
> Provides a framework to virtualize devices.
> - See Documentation/driver-api/vfio-mediated-device.rst for more details.
>
> If you don't know what do here, say N.
>
Hi,
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
The depends and default lines should be indented with tab, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
> +
> +
> config VFIO_MDEV_DEVICE
> tristate "VFIO driver for Mediated devices"
> depends on VFIO && VFIO_MDEV
--
~Randy
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 1/6] mdev: make mdev bus agnostic
@ 2019-11-19 3:08 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:08 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, kevin.tian, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, zhi.a.wang,
farman, parav, gor, cunming.liang, rodrigo.vivi, xiao.w.wang,
freude, stefanha, zhihong.wang, akrowiak, jiri, netdev, cohuck,
oberpar, maxime.coquelin, aadam, lingshan.zhu
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
> index 5da27f2100f9..2e07ca915a96 100644
> --- a/drivers/vfio/mdev/Kconfig
> +++ b/drivers/vfio/mdev/Kconfig
> @@ -1,15 +1,24 @@
> -# SPDX-License-Identifier: GPL-2.0-only
>
> -config VFIO_MDEV
> +config MDEV
> tristate "Mediated device driver framework"
> - depends on VFIO
> default n
> help
> Provides a framework to virtualize devices.
> - See Documentation/driver-api/vfio-mediated-device.rst for more details.
>
> If you don't know what do here, say N.
>
Hi,
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
The depends and default lines should be indented with tab, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
> +
> +
> config VFIO_MDEV_DEVICE
> tristate "VFIO driver for Mediated devices"
> depends on VFIO && VFIO_MDEV
--
~Randy
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 1/6] mdev: make mdev bus agnostic
@ 2019-11-19 3:08 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:08 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, rob.miller, hch, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, farman, parav, gor,
cunming.liang, xiao.w.wang, freude, stefanha, zhihong.wang,
akrowiak, jiri, netdev, cohuck, oberpar, maxime.coquelin, aadam,
lingshan.zhu
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/vfio/mdev/Kconfig b/drivers/vfio/mdev/Kconfig
> index 5da27f2100f9..2e07ca915a96 100644
> --- a/drivers/vfio/mdev/Kconfig
> +++ b/drivers/vfio/mdev/Kconfig
> @@ -1,15 +1,24 @@
> -# SPDX-License-Identifier: GPL-2.0-only
>
> -config VFIO_MDEV
> +config MDEV
> tristate "Mediated device driver framework"
> - depends on VFIO
> default n
> help
> Provides a framework to virtualize devices.
> - See Documentation/driver-api/vfio-mediated-device.rst for more details.
>
> If you don't know what do here, say N.
>
Hi,
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
The depends and default lines should be indented with tab, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
> +
> +
> config VFIO_MDEV_DEVICE
> tristate "VFIO driver for Mediated devices"
> depends on VFIO && VFIO_MDEV
--
~Randy
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 3/6] mdev: move to drivers/
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-19 3:11 ` Randy Dunlap
-1 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:11 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, hch, aadam,
jakub.kicinski, jiri, jeffrey.t.kirsher
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> new file mode 100644
> index 000000000000..4561f2d4178f
> --- /dev/null
> +++ b/drivers/mdev/Kconfig
> @@ -0,0 +1,19 @@
> +
> +config MDEV
> + tristate "Mediated device driver framework"
> + default n
> + help
> + Provides a framework to virtualize devices.
> +
> + If you don't know what do here, say N.
> +
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
Use tab on the 2 lines above, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
--
~Randy
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-19 3:11 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:11 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, hch, aadam,
jakub.kicinski
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> new file mode 100644
> index 000000000000..4561f2d4178f
> --- /dev/null
> +++ b/drivers/mdev/Kconfig
> @@ -0,0 +1,19 @@
> +
> +config MDEV
> + tristate "Mediated device driver framework"
> + default n
> + help
> + Provides a framework to virtualize devices.
> +
> + If you don't know what do here, say N.
> +
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
Use tab on the 2 lines above, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
--
~Randy
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-19 3:11 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:11 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, kevin.tian, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, zhi.a.wang,
farman, parav, gor, cunming.liang, rodrigo.vivi, xiao.w.wang,
freude, stefanha, zhihong.wang, akrowiak, jiri, netdev, cohuck,
oberpar, maxime.coquelin, aadam, lingshan.zhu
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> new file mode 100644
> index 000000000000..4561f2d4178f
> --- /dev/null
> +++ b/drivers/mdev/Kconfig
> @@ -0,0 +1,19 @@
> +
> +config MDEV
> + tristate "Mediated device driver framework"
> + default n
> + help
> + Provides a framework to virtualize devices.
> +
> + If you don't know what do here, say N.
> +
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
Use tab on the 2 lines above, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
--
~Randy
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 3/6] mdev: move to drivers/
@ 2019-11-19 3:11 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:11 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, rob.miller, hch, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, farman, parav, gor,
cunming.liang, xiao.w.wang, freude, stefanha, zhihong.wang,
akrowiak, jiri, netdev, cohuck, oberpar, maxime.coquelin, aadam,
lingshan.zhu
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> new file mode 100644
> index 000000000000..4561f2d4178f
> --- /dev/null
> +++ b/drivers/mdev/Kconfig
> @@ -0,0 +1,19 @@
> +
> +config MDEV
> + tristate "Mediated device driver framework"
> + default n
> + help
> + Provides a framework to virtualize devices.
> +
> + If you don't know what do here, say N.
> +
> +config VFIO_MDEV
> + tristate "VFIO Mediated device driver"
> + depends on VFIO && MDEV
> + default n
Use tab on the 2 lines above, not spaces.
> + help
> + Proivdes a mediated BUS for userspace driver through VFIO
Provides
> + framework. See Documentation/vfio-mediated-device.txt for
> + more details.
> +
> + If you don't know what do here, say N.
--
~Randy
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
2019-11-18 10:59 ` Jason Wang
(?)
(?)
@ 2019-11-19 3:13 ` Randy Dunlap
-1 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:13 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: netdev, cohuck, maxime.coquelin, cunming.liang, zhihong.wang,
rob.miller, xiao.w.wang, haotian.wang, zhenyuw, zhi.a.wang,
jani.nikula, joonas.lahtinen, rodrigo.vivi, airlied, daniel,
farman, pasic, sebott, oberpar, heiko.carstens, gor, borntraeger,
akrowiak, freude, lingshan.zhu, eperezma, lulu, parav,
christophe.de.dinechin, kevin.tian, stefanha, hch, aadam,
jakub.kicinski, jiri, jeffrey.t.kirsher
Hi,
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> index 4561f2d4178f..cd84d4670552 100644
> --- a/drivers/mdev/Kconfig
> +++ b/drivers/mdev/Kconfig
> @@ -17,3 +17,13 @@ config VFIO_MDEV
> more details.
>
> If you don't know what do here, say N.
> +
> +config MDEV_VIRTIO
> + tristate "Mediated VIRTIO bus"
> + depends on VIRTIO && MDEV
> + default n
> + help
> + Proivdes a mediated BUS for virtio. It could be used by
Provides
> + either kenrel driver or userspace driver.
kernel
> +
> + If you don't know what do here, say N.
All of these lines should be indented with one tab, not spaces.
--
~Randy
Reported-by: Randy Dunlap <rdunlap@infradead.org>
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 3:13 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:13 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, rob.miller, hch, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, farman, parav, gor,
cunming.liang, xiao.w.wang, freude, stefanha, zhihong.wang,
akrowiak, jiri, netdev, cohuck, oberpar, maxime.coquelin, aadam,
lingshan.zhu
Hi,
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> index 4561f2d4178f..cd84d4670552 100644
> --- a/drivers/mdev/Kconfig
> +++ b/drivers/mdev/Kconfig
> @@ -17,3 +17,13 @@ config VFIO_MDEV
> more details.
>
> If you don't know what do here, say N.
> +
> +config MDEV_VIRTIO
> + tristate "Mediated VIRTIO bus"
> + depends on VIRTIO && MDEV
> + default n
> + help
> + Proivdes a mediated BUS for virtio. It could be used by
Provides
> + either kenrel driver or userspace driver.
kernel
> +
> + If you don't know what do here, say N.
All of these lines should be indented with one tab, not spaces.
--
~Randy
Reported-by: Randy Dunlap <rdunlap@infradead.org>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 3:13 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:13 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, kevin.tian, rob.miller, hch, lulu, eperezma,
pasic, borntraeger, haotian.wang, jeffrey.t.kirsher, zhi.a.wang,
farman, parav, gor, cunming.liang, rodrigo.vivi, xiao.w.wang,
freude, stefanha, zhihong.wang, akrowiak, jiri, netdev, cohuck,
oberpar, maxime.coquelin, aadam, lingshan.zhu
Hi,
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> index 4561f2d4178f..cd84d4670552 100644
> --- a/drivers/mdev/Kconfig
> +++ b/drivers/mdev/Kconfig
> @@ -17,3 +17,13 @@ config VFIO_MDEV
> more details.
>
> If you don't know what do here, say N.
> +
> +config MDEV_VIRTIO
> + tristate "Mediated VIRTIO bus"
> + depends on VIRTIO && MDEV
> + default n
> + help
> + Proivdes a mediated BUS for virtio. It could be used by
Provides
> + either kenrel driver or userspace driver.
kernel
> +
> + If you don't know what do here, say N.
All of these lines should be indented with one tab, not spaces.
--
~Randy
Reported-by: Randy Dunlap <rdunlap@infradead.org>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 3:13 ` Randy Dunlap
0 siblings, 0 replies; 100+ messages in thread
From: Randy Dunlap @ 2019-11-19 3:13 UTC (permalink / raw)
To: Jason Wang, kvm, linux-s390, linux-kernel, dri-devel, intel-gfx,
intel-gvt-dev, kwankhede, alex.williamson, mst, tiwei.bie,
gregkh, jgg
Cc: jakub.kicinski, christophe.de.dinechin, sebott, airlied,
heiko.carstens, rob.miller, hch, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, farman, parav, gor,
cunming.liang, xiao.w.wang, freude, stefanha, zhihong.wang,
akrowiak, jiri, netdev, cohuck, oberpar, maxime.coquelin, aadam,
lingshan.zhu
Hi,
On 11/18/19 2:59 AM, Jason Wang wrote:
> diff --git a/drivers/mdev/Kconfig b/drivers/mdev/Kconfig
> index 4561f2d4178f..cd84d4670552 100644
> --- a/drivers/mdev/Kconfig
> +++ b/drivers/mdev/Kconfig
> @@ -17,3 +17,13 @@ config VFIO_MDEV
> more details.
>
> If you don't know what do here, say N.
> +
> +config MDEV_VIRTIO
> + tristate "Mediated VIRTIO bus"
> + depends on VIRTIO && MDEV
> + default n
> + help
> + Proivdes a mediated BUS for virtio. It could be used by
Provides
> + either kenrel driver or userspace driver.
kernel
> +
> + If you don't know what do here, say N.
All of these lines should be indented with one tab, not spaces.
--
~Randy
Reported-by: Randy Dunlap <rdunlap@infradead.org>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 12:38 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 12:38 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>
> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> > On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> > > On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > > > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > > > +struct bus_type mdev_virtio_bus_type;
> > > > > +
> > > > > +struct mdev_virtio_device {
> > > > > + struct mdev_device mdev;
> > > > > + const struct mdev_virtio_ops *ops;
> > > > > + u16 class_id;
> > > > > +};
> > > > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > > > same bus?
> > > I must be missing something - which bus do they share?
> > mdev_bus_type ?
> >
> > Jason
>
>
> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
> bus.
That is even worse, why involve struct mdev_device at all then?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 12:38 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 12:38 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>
> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> > On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> > > On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > > > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > > > +struct bus_type mdev_virtio_bus_type;
> > > > > +
> > > > > +struct mdev_virtio_device {
> > > > > + struct mdev_device mdev;
> > > > > + const struct mdev_virtio_ops *ops;
> > > > > + u16 class_id;
> > > > > +};
> > > > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > > > same bus?
> > > I must be missing something - which bus do they share?
> > mdev_bus_type ?
> >
> > Jason
>
>
> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
> bus.
That is even worse, why involve struct mdev_device at all then?
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 12:40 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 12:40 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
> > Also, see the other conversations we are having about a "virtual" bus
> > and devices. I do not want to have two different ways of doing the same
> > thing in the kernel at the same time please. Please work together with
> > the Intel developers to solve this in a unified way, as you both
> > need/want the same thing here.
>
> Sure, some functions looks similar, but the "virtual" bus does not contain a
> management interface and it's not clear that how it can be used by userspace
> driver. For this series, sysfs/GUID based management interface is reused and
> we had a concrete example of how it would be used by userspace driver[1] and
> a real hardware driver implementation[2].
The lifecycle stuff should be re-used through a library of this guid
stuff, not by 'subclassing' mdev_device
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 12:40 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 12:40 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, Greg KH, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
> > Also, see the other conversations we are having about a "virtual" bus
> > and devices. I do not want to have two different ways of doing the same
> > thing in the kernel at the same time please. Please work together with
> > the Intel developers to solve this in a unified way, as you both
> > need/want the same thing here.
>
> Sure, some functions looks similar, but the "virtual" bus does not contain a
> management interface and it's not clear that how it can be used by userspace
> driver. For this series, sysfs/GUID based management interface is reused and
> we had a concrete example of how it would be used by userspace driver[1] and
> a real hardware driver implementation[2].
The lifecycle stuff should be re-used through a library of this guid
stuff, not by 'subclassing' mdev_device
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 14:02 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:02 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger
On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>> +
>>>>>> +struct mdev_virtio_device {
>>>>>> + struct mdev_device mdev;
>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>> + u16 class_id;
>>>>>> +};
>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>> same bus?
>>>> I must be missing something - which bus do they share?
>>> mdev_bus_type ?
>>>
>>> Jason
>>
>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>> bus.
> That is even worse, why involve struct mdev_device at all then?
>
> Jason
I don't quite get the question here.
My understanding for mdev is that it was a mediator between the driver
and physical device when it's hard to let them talk directly due to the
complexity of refactoring and maintenance. This is exact the case of
hardware that can offload virtio datapath but not control path. We want
to present a unified interface (standard virtio) instead of a vendor
specific interface, so a mediator level in the middle is a must. For
virtio driver, mediator present a full virtio compatible device. For
hardware, mediator will mediate the difference between the behavior
defined by virtio spec and real hardware.
And the reason why not inventing something new instead of existed mdev
is because mdev fits into the requirement of virtio-mdev very well:
1) mature framework which has been used by vGPU and other type for years
2) life cycle interface, have a unified interface for management instead
of a vendor specific one so less pain for management
3) device type management. In the case of virtio, user can choose to
create a vhost type of device or virtio type of device, or technically
it can choose which version or features of virtio device it want to create.
4) IOMMU support, mdev allows DMA isolation done at either parent level
or platform/bus level
5) vendor specific attributes
So in Parav's thread [1], if I understand correctly. The major concern
is the API multiplexer at a single mdev bus level. So comes to this
series which decouple VFIO and make mdev more generic to be suitable for
implementing a set of independent buses with similar functions.
Thanks
[1] https://www.spinics.net/lists/linux-rdma/msg85856.html
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 14:02 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:02 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger, haotian.wang,
jeffrey.t.kirsher, cunming.liang, zhi.a.wang, farman,
Parav Pandit, gor, intel-gfx, alex.williamson, xiao.w.wang,
freude, stefanha, zhihong.wang, rodrigo.vivi, intel-gvt-dev, hch,
akrowiak, aadam, Jiri Pirko, tiwei.bie, gregkh, cohuck,
linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>> +
>>>>>> +struct mdev_virtio_device {
>>>>>> + struct mdev_device mdev;
>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>> + u16 class_id;
>>>>>> +};
>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>> same bus?
>>>> I must be missing something - which bus do they share?
>>> mdev_bus_type ?
>>>
>>> Jason
>>
>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>> bus.
> That is even worse, why involve struct mdev_device at all then?
>
> Jason
I don't quite get the question here.
My understanding for mdev is that it was a mediator between the driver
and physical device when it's hard to let them talk directly due to the
complexity of refactoring and maintenance. This is exact the case of
hardware that can offload virtio datapath but not control path. We want
to present a unified interface (standard virtio) instead of a vendor
specific interface, so a mediator level in the middle is a must. For
virtio driver, mediator present a full virtio compatible device. For
hardware, mediator will mediate the difference between the behavior
defined by virtio spec and real hardware.
And the reason why not inventing something new instead of existed mdev
is because mdev fits into the requirement of virtio-mdev very well:
1) mature framework which has been used by vGPU and other type for years
2) life cycle interface, have a unified interface for management instead
of a vendor specific one so less pain for management
3) device type management. In the case of virtio, user can choose to
create a vhost type of device or virtio type of device, or technically
it can choose which version or features of virtio device it want to create.
4) IOMMU support, mdev allows DMA isolation done at either parent level
or platform/bus level
5) vendor specific attributes
So in Parav's thread [1], if I understand correctly. The major concern
is the API multiplexer at a single mdev bus level. So comes to this
series which decouple VFIO and make mdev more generic to be suitable for
implementing a set of independent buses with similar functions.
Thanks
[1] https://www.spinics.net/lists/linux-rdma/msg85856.html
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 14:02 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:02 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>> +
>>>>>> +struct mdev_virtio_device {
>>>>>> + struct mdev_device mdev;
>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>> + u16 class_id;
>>>>>> +};
>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>> same bus?
>>>> I must be missing something - which bus do they share?
>>> mdev_bus_type ?
>>>
>>> Jason
>>
>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>> bus.
> That is even worse, why involve struct mdev_device at all then?
>
> Jason
I don't quite get the question here.
My understanding for mdev is that it was a mediator between the driver
and physical device when it's hard to let them talk directly due to the
complexity of refactoring and maintenance. This is exact the case of
hardware that can offload virtio datapath but not control path. We want
to present a unified interface (standard virtio) instead of a vendor
specific interface, so a mediator level in the middle is a must. For
virtio driver, mediator present a full virtio compatible device. For
hardware, mediator will mediate the difference between the behavior
defined by virtio spec and real hardware.
And the reason why not inventing something new instead of existed mdev
is because mdev fits into the requirement of virtio-mdev very well:
1) mature framework which has been used by vGPU and other type for years
2) life cycle interface, have a unified interface for management instead
of a vendor specific one so less pain for management
3) device type management. In the case of virtio, user can choose to
create a vhost type of device or virtio type of device, or technically
it can choose which version or features of virtio device it want to create.
4) IOMMU support, mdev allows DMA isolation done at either parent level
or platform/bus level
5) vendor specific attributes
So in Parav's thread [1], if I understand correctly. The major concern
is the API multiplexer at a single mdev bus level. So comes to this
series which decouple VFIO and make mdev more generic to be suitable for
implementing a set of independent buses with similar functions.
Thanks
[1] https://www.spinics.net/lists/linux-rdma/msg85856.html
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 14:07 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:07 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger
On 2019/11/19 下午8:40, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
>>> Also, see the other conversations we are having about a "virtual" bus
>>> and devices. I do not want to have two different ways of doing the same
>>> thing in the kernel at the same time please. Please work together with
>>> the Intel developers to solve this in a unified way, as you both
>>> need/want the same thing here.
>> Sure, some functions looks similar, but the "virtual" bus does not contain a
>> management interface and it's not clear that how it can be used by userspace
>> driver. For this series, sysfs/GUID based management interface is reused and
>> we had a concrete example of how it would be used by userspace driver[1] and
>> a real hardware driver implementation[2].
> The lifecycle stuff should be re-used through a library of this guid
> stuff, not by 'subclassing' mdev_device
>
> Jason
But mdev provides more than lifecycle management: type management, IOMMU
support etc. And more could be added in the future.
Having a library that serves exactly for the case of mdev seems less
convenient than making mdev_device a 'parent class'.
Thanks
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 14:07 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:07 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, kevin.tian, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
zhi.a.wang, farman, Parav Pandit, gor, intel-gfx,
alex.williamson, xiao.w.wang, freude, stefanha, zhihong.wang,
rodrigo.vivi, intel-gvt-dev, hch, akrowiak, aadam, Jiri Pirko,
tiwei.bie, Greg KH, cohuck, linux-kernel, maxime.coquelin,
netdev, lingshan.zhu
On 2019/11/19 下午8:40, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
>>> Also, see the other conversations we are having about a "virtual" bus
>>> and devices. I do not want to have two different ways of doing the same
>>> thing in the kernel at the same time please. Please work together with
>>> the Intel developers to solve this in a unified way, as you both
>>> need/want the same thing here.
>> Sure, some functions looks similar, but the "virtual" bus does not contain a
>> management interface and it's not clear that how it can be used by userspace
>> driver. For this series, sysfs/GUID based management interface is reused and
>> we had a concrete example of how it would be used by userspace driver[1] and
>> a real hardware driver implementation[2].
> The lifecycle stuff should be re-used through a library of this guid
> stuff, not by 'subclassing' mdev_device
>
> Jason
But mdev provides more than lifecycle management: type management, IOMMU
support etc. And more could be added in the future.
Having a library that serves exactly for the case of mdev seems less
convenient than making mdev_device a 'parent class'.
Thanks
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 14:07 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-19 14:07 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, Greg KH, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 下午8:40, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
>>> Also, see the other conversations we are having about a "virtual" bus
>>> and devices. I do not want to have two different ways of doing the same
>>> thing in the kernel at the same time please. Please work together with
>>> the Intel developers to solve this in a unified way, as you both
>>> need/want the same thing here.
>> Sure, some functions looks similar, but the "virtual" bus does not contain a
>> management interface and it's not clear that how it can be used by userspace
>> driver. For this series, sysfs/GUID based management interface is reused and
>> we had a concrete example of how it would be used by userspace driver[1] and
>> a real hardware driver implementation[2].
> The lifecycle stuff should be re-used through a library of this guid
> stuff, not by 'subclassing' mdev_device
>
> Jason
But mdev provides more than lifecycle management: type management, IOMMU
support etc. And more could be added in the future.
Having a library that serves exactly for the case of mdev seems less
convenient than making mdev_device a 'parent class'.
Thanks
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 14:14 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 14:14 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
>
> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
> > > On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> > > > On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> > > > > On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > > > > > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > > > > > +struct bus_type mdev_virtio_bus_type;
> > > > > > > +
> > > > > > > +struct mdev_virtio_device {
> > > > > > > + struct mdev_device mdev;
> > > > > > > + const struct mdev_virtio_ops *ops;
> > > > > > > + u16 class_id;
> > > > > > > +};
> > > > > > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > > > > > same bus?
> > > > > I must be missing something - which bus do they share?
> > > > mdev_bus_type ?
> > > >
> > > > Jason
> > >
> > > Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
> > > bus.
> > That is even worse, why involve struct mdev_device at all then?
> >
> > Jason
>
>
> I don't quite get the question here.
In the driver model the bus_type and foo_device are closely
linked. Creating 'mdev_device' instances and overriding the bus_type
is a very abusive thing to do.
> My understanding for mdev is that it was a mediator between the driver and
> physical device when it's hard to let them talk directly due to the
> complexity of refactoring and maintenance.
Really, mdev is to support vfio with a backend other than PCI, nothing
more.
Abusing it for other things is not appropriate. ie creating an
instance and not filling in most of the vfio focused ops is an abusive
thing to do.
> hardware that can offload virtio datapath but not control path. We want to
> present a unified interface (standard virtio) instead of a vendor specific
> interface, so a mediator level in the middle is a must. For virtio driver,
> mediator present a full virtio compatible device. For hardware, mediator
> will mediate the difference between the behavior defined by virtio spec and
> real hardware.
If you need to bind to the VFIO driver then mdev is the right thing to
use, otherwise it is not.
It certainly should not be used to bind to random kernel drivers. This
problem is what this virtual bus idea Intel is working on might solve.
It seems the only thing people care about with mdev is the GUID
lifecycle stuff, but at the same time folks like Parav are saying they
don't want to use that lifecycle stuff and prefer devlink
instead.
Most likely, at least for virtio-net, everyone else will be able to
use devlink as well, making it much less clear if that GUID lifecycle
stuff is a good idea or not.
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-19 14:14 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 14:14 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
>
> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
> > > On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> > > > On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> > > > > On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> > > > > > On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> > > > > > > +struct bus_type mdev_virtio_bus_type;
> > > > > > > +
> > > > > > > +struct mdev_virtio_device {
> > > > > > > + struct mdev_device mdev;
> > > > > > > + const struct mdev_virtio_ops *ops;
> > > > > > > + u16 class_id;
> > > > > > > +};
> > > > > > This seems to share nothing with mdev (ie mdev-vfio), why is it on the
> > > > > > same bus?
> > > > > I must be missing something - which bus do they share?
> > > > mdev_bus_type ?
> > > >
> > > > Jason
> > >
> > > Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
> > > bus.
> > That is even worse, why involve struct mdev_device at all then?
> >
> > Jason
>
>
> I don't quite get the question here.
In the driver model the bus_type and foo_device are closely
linked. Creating 'mdev_device' instances and overriding the bus_type
is a very abusive thing to do.
> My understanding for mdev is that it was a mediator between the driver and
> physical device when it's hard to let them talk directly due to the
> complexity of refactoring and maintenance.
Really, mdev is to support vfio with a backend other than PCI, nothing
more.
Abusing it for other things is not appropriate. ie creating an
instance and not filling in most of the vfio focused ops is an abusive
thing to do.
> hardware that can offload virtio datapath but not control path. We want to
> present a unified interface (standard virtio) instead of a vendor specific
> interface, so a mediator level in the middle is a must. For virtio driver,
> mediator present a full virtio compatible device. For hardware, mediator
> will mediate the difference between the behavior defined by virtio spec and
> real hardware.
If you need to bind to the VFIO driver then mdev is the right thing to
use, otherwise it is not.
It certainly should not be used to bind to random kernel drivers. This
problem is what this virtual bus idea Intel is working on might solve.
It seems the only thing people care about with mdev is the GUID
lifecycle stuff, but at the same time folks like Parav are saying they
don't want to use that lifecycle stuff and prefer devlink
instead.
Most likely, at least for virtio-net, everyone else will be able to
use devlink as well, making it much less clear if that GUID lifecycle
stuff is a good idea or not.
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 16:39 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 16:39 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Tue, Nov 19, 2019 at 10:07:18PM +0800, Jason Wang wrote:
>
> On 2019/11/19 下午8:40, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
> > > > Also, see the other conversations we are having about a "virtual" bus
> > > > and devices. I do not want to have two different ways of doing the same
> > > > thing in the kernel at the same time please. Please work together with
> > > > the Intel developers to solve this in a unified way, as you both
> > > > need/want the same thing here.
> > > Sure, some functions looks similar, but the "virtual" bus does not contain a
> > > management interface and it's not clear that how it can be used by userspace
> > > driver. For this series, sysfs/GUID based management interface is reused and
> > > we had a concrete example of how it would be used by userspace driver[1] and
> > > a real hardware driver implementation[2].
> > The lifecycle stuff should be re-used through a library of this guid
> > stuff, not by 'subclassing' mdev_device
>
> But mdev provides more than lifecycle management: type management, IOMMU
> support etc. And more could be added in the future.
iommu support comes from dma_ops, not mdev.
mdev only provides vfio, and in this example there is no use of those
vfio ops - so it is not mdev.
> Having a library that serves exactly for the case of mdev seems less
> convenient than making mdev_device a 'parent class'.
It is the design pattern the device core uses
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework
@ 2019-11-19 16:39 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-19 16:39 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm, mst,
airlied, heiko.carstens, dri-devel, oberpar, kwankhede,
rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, Greg KH, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Tue, Nov 19, 2019 at 10:07:18PM +0800, Jason Wang wrote:
>
> On 2019/11/19 下午8:40, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 11:03:39AM +0800, Jason Wang wrote:
> > > > Also, see the other conversations we are having about a "virtual" bus
> > > > and devices. I do not want to have two different ways of doing the same
> > > > thing in the kernel at the same time please. Please work together with
> > > > the Intel developers to solve this in a unified way, as you both
> > > > need/want the same thing here.
> > > Sure, some functions looks similar, but the "virtual" bus does not contain a
> > > management interface and it's not clear that how it can be used by userspace
> > > driver. For this series, sysfs/GUID based management interface is reused and
> > > we had a concrete example of how it would be used by userspace driver[1] and
> > > a real hardware driver implementation[2].
> > The lifecycle stuff should be re-used through a library of this guid
> > stuff, not by 'subclassing' mdev_device
>
> But mdev provides more than lifecycle management: type management, IOMMU
> support etc. And more could be added in the future.
iommu support comes from dma_ops, not mdev.
mdev only provides vfio, and in this example there is no use of those
vfio ops - so it is not mdev.
> Having a library that serves exactly for the case of mdev seems less
> convenient than making mdev_device a 'parent class'.
It is the design pattern the device core uses
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 2:14 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-20 2:14 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
>> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
>>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>>>> +
>>>>>>>> +struct mdev_virtio_device {
>>>>>>>> + struct mdev_device mdev;
>>>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>>>> + u16 class_id;
>>>>>>>> +};
>>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>>>> same bus?
>>>>>> I must be missing something - which bus do they share?
>>>>> mdev_bus_type ?
>>>>>
>>>>> Jason
>>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>>>> bus.
>>> That is even worse, why involve struct mdev_device at all then?
>>>
>>> Jason
>>
>> I don't quite get the question here.
> In the driver model the bus_type and foo_device are closely
> linked.
I don't get the definition of "closely linked" here. Do you think the
bus and device implement virtual bus series are closely linked? If yes,
how did they achieve that?
> Creating 'mdev_device' instances and overriding the bus_type
> is a very abusive thing to do.
Ok, mdev_device (without this series) had:
struct mdev_device {
struct device dev;
struct mdev_parent *parent;
guid_t uuid;
void *driver_data;
struct list_head next;
struct kobject *type_kobj;
struct device *iommu_device;
bool active;
};
So it's nothing bus or VFIO specific. And what virtual bus had is:
struct virtbus_device {
const char *name;
int id;
const struct virtbus_dev_id *dev_id;
struct device dev;
void *data;
};
Are there any fundamental issues that you think mdev_device is abused? I
won't expect the answers are generic objects like kobj, iommu device
pointer etc.
>
>> My understanding for mdev is that it was a mediator between the driver and
>> physical device when it's hard to let them talk directly due to the
>> complexity of refactoring and maintenance.
> Really, mdev is to support vfio with a backend other than PCI, nothing
> more.
That partially explain why it was called mdev. So for virito, we want
standard virtio driver to talk with a backend other than virtio.
For the issue of PCI, actually the API is generic enough to support
device other than PCI, e.g AP bus.
>
> Abusing it for other things is not appropriate. ie creating an
> instance and not filling in most of the vfio focused ops is an abusive
> thing to do.
Well, it's only half of the mdev_parent_ops in mdev_parent, various
methods could be done do be more generic to avoid duplication of codes. No?
>
>> hardware that can offload virtio datapath but not control path. We want to
>> present a unified interface (standard virtio) instead of a vendor specific
>> interface, so a mediator level in the middle is a must. For virtio driver,
>> mediator present a full virtio compatible device. For hardware, mediator
>> will mediate the difference between the behavior defined by virtio spec and
>> real hardware.
> If you need to bind to the VFIO driver then mdev is the right thing to
> use, otherwise it is not.
>
> It certainly should not be used to bind to random kernel drivers. This
> problem is what this virtual bus idea Intel is working on might solve.
What do you mean by random here? With this series, we have dedicated bus
and dedicated driver with matching method to make sure the binding is
correct.
>
> It seems the only thing people care about with mdev is the GUID
> lifecycle stuff, but at the same time folks like Parav are saying they
> don't want to use that lifecycle stuff and prefer devlink
> instead.
I'm sure you will need to handle other issues besides GUID which had
been settled by mdev e.g IOMMU and types when starting to write a real
hardware driver.
>
> Most likely, at least for virtio-net, everyone else will be able to
> use devlink as well, making it much less clear if that GUID lifecycle
> stuff is a good idea or not.
This assumption is wrong, we hard already had at least two concrete
examples of vDPA device that doesn't use devlink:
- Intel IFC where virtio is done at VF level
- Ali Cloud ECS instance where virtio is done at PF level
Again, the device slicing is only part of our goal. The major goal is to
have a mediator level that can take over the virtio control path between
a standard virtio driver and a hardware who datapath is virtio
compatible but not control path.
Thanks
>
> Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 2:14 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-20 2:14 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger, haotian.wang,
jeffrey.t.kirsher, cunming.liang, zhi.a.wang, farman,
Parav Pandit, gor, intel-gfx, alex.williamson, xiao.w.wang,
freude, stefanha, zhihong.wang, rodrigo.vivi, intel-gvt-dev, hch,
akrowiak, aadam, Jiri Pirko, tiwei.bie, gregkh, cohuck,
linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
>> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
>>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>>>> +
>>>>>>>> +struct mdev_virtio_device {
>>>>>>>> + struct mdev_device mdev;
>>>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>>>> + u16 class_id;
>>>>>>>> +};
>>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>>>> same bus?
>>>>>> I must be missing something - which bus do they share?
>>>>> mdev_bus_type ?
>>>>>
>>>>> Jason
>>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>>>> bus.
>>> That is even worse, why involve struct mdev_device at all then?
>>>
>>> Jason
>>
>> I don't quite get the question here.
> In the driver model the bus_type and foo_device are closely
> linked.
I don't get the definition of "closely linked" here. Do you think the
bus and device implement virtual bus series are closely linked? If yes,
how did they achieve that?
> Creating 'mdev_device' instances and overriding the bus_type
> is a very abusive thing to do.
Ok, mdev_device (without this series) had:
struct mdev_device {
struct device dev;
struct mdev_parent *parent;
guid_t uuid;
void *driver_data;
struct list_head next;
struct kobject *type_kobj;
struct device *iommu_device;
bool active;
};
So it's nothing bus or VFIO specific. And what virtual bus had is:
struct virtbus_device {
const char *name;
int id;
const struct virtbus_dev_id *dev_id;
struct device dev;
void *data;
};
Are there any fundamental issues that you think mdev_device is abused? I
won't expect the answers are generic objects like kobj, iommu device
pointer etc.
>
>> My understanding for mdev is that it was a mediator between the driver and
>> physical device when it's hard to let them talk directly due to the
>> complexity of refactoring and maintenance.
> Really, mdev is to support vfio with a backend other than PCI, nothing
> more.
That partially explain why it was called mdev. So for virito, we want
standard virtio driver to talk with a backend other than virtio.
For the issue of PCI, actually the API is generic enough to support
device other than PCI, e.g AP bus.
>
> Abusing it for other things is not appropriate. ie creating an
> instance and not filling in most of the vfio focused ops is an abusive
> thing to do.
Well, it's only half of the mdev_parent_ops in mdev_parent, various
methods could be done do be more generic to avoid duplication of codes. No?
>
>> hardware that can offload virtio datapath but not control path. We want to
>> present a unified interface (standard virtio) instead of a vendor specific
>> interface, so a mediator level in the middle is a must. For virtio driver,
>> mediator present a full virtio compatible device. For hardware, mediator
>> will mediate the difference between the behavior defined by virtio spec and
>> real hardware.
> If you need to bind to the VFIO driver then mdev is the right thing to
> use, otherwise it is not.
>
> It certainly should not be used to bind to random kernel drivers. This
> problem is what this virtual bus idea Intel is working on might solve.
What do you mean by random here? With this series, we have dedicated bus
and dedicated driver with matching method to make sure the binding is
correct.
>
> It seems the only thing people care about with mdev is the GUID
> lifecycle stuff, but at the same time folks like Parav are saying they
> don't want to use that lifecycle stuff and prefer devlink
> instead.
I'm sure you will need to handle other issues besides GUID which had
been settled by mdev e.g IOMMU and types when starting to write a real
hardware driver.
>
> Most likely, at least for virtio-net, everyone else will be able to
> use devlink as well, making it much less clear if that GUID lifecycle
> stuff is a good idea or not.
This assumption is wrong, we hard already had at least two concrete
examples of vDPA device that doesn't use devlink:
- Intel IFC where virtio is done at VF level
- Ali Cloud ECS instance where virtio is done at PF level
Again, the device slicing is only part of our goal. The major goal is to
have a mediator level that can take over the virtio control path between
a standard virtio driver and a hardware who datapath is virtio
compatible but not control path.
Thanks
>
> Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 2:14 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-20 2:14 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
>> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
>>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
>>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
>>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
>>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
>>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
>>>>>>>> +struct bus_type mdev_virtio_bus_type;
>>>>>>>> +
>>>>>>>> +struct mdev_virtio_device {
>>>>>>>> + struct mdev_device mdev;
>>>>>>>> + const struct mdev_virtio_ops *ops;
>>>>>>>> + u16 class_id;
>>>>>>>> +};
>>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on the
>>>>>>> same bus?
>>>>>> I must be missing something - which bus do they share?
>>>>> mdev_bus_type ?
>>>>>
>>>>> Jason
>>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not the same
>>>> bus.
>>> That is even worse, why involve struct mdev_device at all then?
>>>
>>> Jason
>>
>> I don't quite get the question here.
> In the driver model the bus_type and foo_device are closely
> linked.
I don't get the definition of "closely linked" here. Do you think the
bus and device implement virtual bus series are closely linked? If yes,
how did they achieve that?
> Creating 'mdev_device' instances and overriding the bus_type
> is a very abusive thing to do.
Ok, mdev_device (without this series) had:
struct mdev_device {
struct device dev;
struct mdev_parent *parent;
guid_t uuid;
void *driver_data;
struct list_head next;
struct kobject *type_kobj;
struct device *iommu_device;
bool active;
};
So it's nothing bus or VFIO specific. And what virtual bus had is:
struct virtbus_device {
const char *name;
int id;
const struct virtbus_dev_id *dev_id;
struct device dev;
void *data;
};
Are there any fundamental issues that you think mdev_device is abused? I
won't expect the answers are generic objects like kobj, iommu device
pointer etc.
>
>> My understanding for mdev is that it was a mediator between the driver and
>> physical device when it's hard to let them talk directly due to the
>> complexity of refactoring and maintenance.
> Really, mdev is to support vfio with a backend other than PCI, nothing
> more.
That partially explain why it was called mdev. So for virito, we want
standard virtio driver to talk with a backend other than virtio.
For the issue of PCI, actually the API is generic enough to support
device other than PCI, e.g AP bus.
>
> Abusing it for other things is not appropriate. ie creating an
> instance and not filling in most of the vfio focused ops is an abusive
> thing to do.
Well, it's only half of the mdev_parent_ops in mdev_parent, various
methods could be done do be more generic to avoid duplication of codes. No?
>
>> hardware that can offload virtio datapath but not control path. We want to
>> present a unified interface (standard virtio) instead of a vendor specific
>> interface, so a mediator level in the middle is a must. For virtio driver,
>> mediator present a full virtio compatible device. For hardware, mediator
>> will mediate the difference between the behavior defined by virtio spec and
>> real hardware.
> If you need to bind to the VFIO driver then mdev is the right thing to
> use, otherwise it is not.
>
> It certainly should not be used to bind to random kernel drivers. This
> problem is what this virtual bus idea Intel is working on might solve.
What do you mean by random here? With this series, we have dedicated bus
and dedicated driver with matching method to make sure the binding is
correct.
>
> It seems the only thing people care about with mdev is the GUID
> lifecycle stuff, but at the same time folks like Parav are saying they
> don't want to use that lifecycle stuff and prefer devlink
> instead.
I'm sure you will need to handle other issues besides GUID which had
been settled by mdev e.g IOMMU and types when starting to write a real
hardware driver.
>
> Most likely, at least for virtio-net, everyone else will be able to
> use devlink as well, making it much less clear if that GUID lifecycle
> stuff is a good idea or not.
This assumption is wrong, we hard already had at least two concrete
examples of vDPA device that doesn't use devlink:
- Intel IFC where virtio is done at VF level
- Ali Cloud ECS instance where virtio is done at PF level
Again, the device slicing is only part of our goal. The major goal is to
have a mediator level that can take over the virtio control path between
a standard virtio driver and a hardware who datapath is virtio
compatible but not control path.
Thanks
>
> Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 13:49 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-20 13:49 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang
On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
> > > I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
> I don't get the definition of "closely linked" here. Do you think the bus
> and device implement virtual bus series are closely linked? If yes, how did
> they achieve that?
I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
not on some 'bar_bus', as that is how the driver core generally works.
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
What do mean? 'struct mdev_parent *parent' is the VFIO specific
stuff. I haven't figured out what the confusing mdev_parent is
supposed to be, or whhy the VFIO ops are linked to the parent or not
the device.. Honestly the whole mdev thing has a very strange take on
how to use the driver core.
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
> could be done do be more generic to avoid duplication of codes. No?
There are many ways to avoid duplicating code.
Taking something well defined, and bolting on something unrelated just
to share a bit of code is a very poor way to avoid code duplication.
> I'm sure you will need to handle other issues besides GUID which had been
> settled by mdev e.g IOMMU and types when starting to write a real hardware
> driver.
The iommu framework already handles that, the mdev stuff contributes
very little from what I can see.
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
> This assumption is wrong, we hard already had at least two concrete examples
> of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
Again, you don't explain why they couldn't use devlink.
Or, why do we need GUID lifecycle stuff when these PCI devices can
only create a single virtio and can just go ahead and do that as soon
as they are probed.
The GUID stuff was invented for slicing, which you say is not
happening in these cases.
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 13:49 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-20 13:49 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger, haotian.wang,
jeffrey.t.kirsher, cunming.liang, zhi.a.wang, farman,
Parav Pandit, gor, intel-gfx, alex.williamson, xiao.w.wang,
freude, stefanha, zhihong.wang, rodrigo.vivi, intel-gvt-dev, hch,
akrowiak, aadam, Jiri Pirko, tiwei.bie, gregkh, cohuck,
linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
> > > I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
> I don't get the definition of "closely linked" here. Do you think the bus
> and device implement virtual bus series are closely linked? If yes, how did
> they achieve that?
I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
not on some 'bar_bus', as that is how the driver core generally works.
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
What do mean? 'struct mdev_parent *parent' is the VFIO specific
stuff. I haven't figured out what the confusing mdev_parent is
supposed to be, or whhy the VFIO ops are linked to the parent or not
the device.. Honestly the whole mdev thing has a very strange take on
how to use the driver core.
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
> could be done do be more generic to avoid duplication of codes. No?
There are many ways to avoid duplicating code.
Taking something well defined, and bolting on something unrelated just
to share a bit of code is a very poor way to avoid code duplication.
> I'm sure you will need to handle other issues besides GUID which had been
> settled by mdev e.g IOMMU and types when starting to write a real hardware
> driver.
The iommu framework already handles that, the mdev stuff contributes
very little from what I can see.
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
> This assumption is wrong, we hard already had at least two concrete examples
> of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
Again, you don't explain why they couldn't use devlink.
Or, why do we need GUID lifecycle stuff when these PCI devices can
only create a single virtio and can just go ahead and do that as soon
as they are probed.
The GUID stuff was invented for slicing, which you say is not
happening in these cases.
Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-20 13:49 ` Jason Gunthorpe
0 siblings, 0 replies; 100+ messages in thread
From: Jason Gunthorpe @ 2019-11-20 13:49 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
> > > I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
> I don't get the definition of "closely linked" here. Do you think the bus
> and device implement virtual bus series are closely linked? If yes, how did
> they achieve that?
I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
not on some 'bar_bus', as that is how the driver core generally works.
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
What do mean? 'struct mdev_parent *parent' is the VFIO specific
stuff. I haven't figured out what the confusing mdev_parent is
supposed to be, or whhy the VFIO ops are linked to the parent or not
the device.. Honestly the whole mdev thing has a very strange take on
how to use the driver core.
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
> could be done do be more generic to avoid duplication of codes. No?
There are many ways to avoid duplicating code.
Taking something well defined, and bolting on something unrelated just
to share a bit of code is a very poor way to avoid code duplication.
> I'm sure you will need to handle other issues besides GUID which had been
> settled by mdev e.g IOMMU and types when starting to write a real hardware
> driver.
The iommu framework already handles that, the mdev stuff contributes
very little from what I can see.
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
> This assumption is wrong, we hard already had at least two concrete examples
> of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
Again, you don't explain why they couldn't use devlink.
Or, why do we need GUID lifecycle stuff when these PCI devices can
only create a single virtio and can just go ahead and do that as soon
as they are probed.
The GUID stuff was invented for slicing, which you say is not
happening in these cases.
Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-21 3:05 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-21 3:05 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger
On 2019/11/20 下午9:49, Jason Gunthorpe wrote:
> On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
>
>>>> I don't quite get the question here.
>>> In the driver model the bus_type and foo_device are closely
>>> linked.
>> I don't get the definition of "closely linked" here. Do you think the bus
>> and device implement virtual bus series are closely linked? If yes, how did
>> they achieve that?
> I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
> not on some 'bar_bus', as that is how the driver core generally works.
I fully agree with you here. But isn't that just what this patch did? We
had "mdev_virtio" device on "mdev_virtio" bus not "mdev_vfio" bus.
>
>>> Creating 'mdev_device' instances and overriding the bus_type
>>> is a very abusive thing to do.
>> Ok, mdev_device (without this series) had:
>>
>> struct mdev_device {
>> struct device dev;
>> struct mdev_parent *parent;
>> guid_t uuid;
>> void *driver_data;
>> struct list_head next;
>> struct kobject *type_kobj;
>> struct device *iommu_device;
>> bool active;
>> };
>>
>> So it's nothing bus or VFIO specific. And what virtual bus had is:
> What do mean? 'struct mdev_parent *parent' is the VFIO specific
> stuff. I haven't figured out what the confusing mdev_parent is
> supposed to be,
struct mdev_parent_ops {
struct module *owner;
const struct attribute_group **dev_attr_groups;
const struct attribute_group **mdev_attr_groups;
struct attribute_group **supported_type_groups;
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
int (*open)(struct mdev_device *mdev);
void (*release)(struct mdev_device *mdev);
ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
size_t count, loff_t *ppos);
ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
size_t count, loff_t *ppos);
long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
unsigned long arg);
int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
You can see that nothing is really VFIO specific here..
> or whhy the VFIO ops are linked to the parent or not
> the device..
I guess the answer the mdev_devices belongs to the same parent are
expected to have same ops.
> Honestly the whole mdev thing has a very strange take on
> how to use the driver core.
Suggestions are welcomed.
>
>>> Abusing it for other things is not appropriate. ie creating an
>>> instance and not filling in most of the vfio focused ops is an abusive
>>> thing to do.
>> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
>> could be done do be more generic to avoid duplication of codes. No?
> There are many ways to avoid duplicating code.
>
> Taking something well defined, and bolting on something unrelated just
> to share a bit of code is a very poor way to avoid code duplication.
We can have make the code better...
>> I'm sure you will need to handle other issues besides GUID which had been
>> settled by mdev e.g IOMMU and types when starting to write a real hardware
>> driver.
> The iommu framework already handles that, the mdev stuff contributes
> very little from what I can see.
Yes, but if we start from beginning to invent a new infrastructure and
we still need GUID, IOMMU, types. So it will be very similar to mdev
looks right now. So why not improve mdev?
>
>>> Most likely, at least for virtio-net, everyone else will be able to
>>> use devlink as well, making it much less clear if that GUID lifecycle
>>> stuff is a good idea or not.
>> This assumption is wrong, we hard already had at least two concrete examples
>> of vDPA device that doesn't use devlink:
>>
>> - Intel IFC where virtio is done at VF level
>> - Ali Cloud ECS instance where virtio is done at PF level
> Again, you don't explain why they couldn't use devlink.
Yes, they could, but of course for many reasons they won't use devlink.
Not only devlink, even netlink is not used or implemented in all type of
network devices.
>
> Or, why do we need GUID lifecycle stuff when these PCI devices can
> only create a single virtio and can just go ahead and do that as soon
> as they are probed.
>
> The GUID stuff was invented for slicing, which you say is not
> happening in these cases.
I think that's all about a consistent management interface, "slicing by
one" is still compatible.
Thanks
>
> Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-21 3:05 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-21 3:05 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, kevin.tian,
dri-devel, oberpar, kwankhede, rob.miller, linux-s390, sebott,
lulu, eperezma, pasic, borntraeger, haotian.wang,
jeffrey.t.kirsher, cunming.liang, zhi.a.wang, farman,
Parav Pandit, gor, intel-gfx, alex.williamson, xiao.w.wang,
freude, stefanha, zhihong.wang, rodrigo.vivi, intel-gvt-dev, hch,
akrowiak, aadam, Jiri Pirko, tiwei.bie, gregkh, cohuck,
linux-kernel, maxime.coquelin, netdev, lingshan.zhu
On 2019/11/20 下午9:49, Jason Gunthorpe wrote:
> On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
>
>>>> I don't quite get the question here.
>>> In the driver model the bus_type and foo_device are closely
>>> linked.
>> I don't get the definition of "closely linked" here. Do you think the bus
>> and device implement virtual bus series are closely linked? If yes, how did
>> they achieve that?
> I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
> not on some 'bar_bus', as that is how the driver core generally works.
I fully agree with you here. But isn't that just what this patch did? We
had "mdev_virtio" device on "mdev_virtio" bus not "mdev_vfio" bus.
>
>>> Creating 'mdev_device' instances and overriding the bus_type
>>> is a very abusive thing to do.
>> Ok, mdev_device (without this series) had:
>>
>> struct mdev_device {
>> struct device dev;
>> struct mdev_parent *parent;
>> guid_t uuid;
>> void *driver_data;
>> struct list_head next;
>> struct kobject *type_kobj;
>> struct device *iommu_device;
>> bool active;
>> };
>>
>> So it's nothing bus or VFIO specific. And what virtual bus had is:
> What do mean? 'struct mdev_parent *parent' is the VFIO specific
> stuff. I haven't figured out what the confusing mdev_parent is
> supposed to be,
struct mdev_parent_ops {
struct module *owner;
const struct attribute_group **dev_attr_groups;
const struct attribute_group **mdev_attr_groups;
struct attribute_group **supported_type_groups;
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
int (*open)(struct mdev_device *mdev);
void (*release)(struct mdev_device *mdev);
ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
size_t count, loff_t *ppos);
ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
size_t count, loff_t *ppos);
long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
unsigned long arg);
int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
You can see that nothing is really VFIO specific here..
> or whhy the VFIO ops are linked to the parent or not
> the device..
I guess the answer the mdev_devices belongs to the same parent are
expected to have same ops.
> Honestly the whole mdev thing has a very strange take on
> how to use the driver core.
Suggestions are welcomed.
>
>>> Abusing it for other things is not appropriate. ie creating an
>>> instance and not filling in most of the vfio focused ops is an abusive
>>> thing to do.
>> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
>> could be done do be more generic to avoid duplication of codes. No?
> There are many ways to avoid duplicating code.
>
> Taking something well defined, and bolting on something unrelated just
> to share a bit of code is a very poor way to avoid code duplication.
We can have make the code better...
>> I'm sure you will need to handle other issues besides GUID which had been
>> settled by mdev e.g IOMMU and types when starting to write a real hardware
>> driver.
> The iommu framework already handles that, the mdev stuff contributes
> very little from what I can see.
Yes, but if we start from beginning to invent a new infrastructure and
we still need GUID, IOMMU, types. So it will be very similar to mdev
looks right now. So why not improve mdev?
>
>>> Most likely, at least for virtio-net, everyone else will be able to
>>> use devlink as well, making it much less clear if that GUID lifecycle
>>> stuff is a good idea or not.
>> This assumption is wrong, we hard already had at least two concrete examples
>> of vDPA device that doesn't use devlink:
>>
>> - Intel IFC where virtio is done at VF level
>> - Ali Cloud ECS instance where virtio is done at PF level
> Again, you don't explain why they couldn't use devlink.
Yes, they could, but of course for many reasons they won't use devlink.
Not only devlink, even netlink is not used or implemented in all type of
network devices.
>
> Or, why do we need GUID lifecycle stuff when these PCI devices can
> only create a single virtio and can just go ahead and do that as soon
> as they are probed.
>
> The GUID stuff was invented for slicing, which you say is not
> happening in these cases.
I think that's all about a consistent management interface, "slicing by
one" is still compatible.
Thanks
>
> Jason
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-21 3:05 ` Jason Wang
0 siblings, 0 replies; 100+ messages in thread
From: Jason Wang @ 2019-11-21 3:05 UTC (permalink / raw)
To: Jason Gunthorpe
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, dri-devel, oberpar,
kwankhede, rob.miller, linux-s390, sebott, lulu, eperezma, pasic,
borntraeger, haotian.wang, jeffrey.t.kirsher, cunming.liang,
farman, Parav Pandit, gor, intel-gfx, xiao.w.wang, freude,
stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
On 2019/11/20 下午9:49, Jason Gunthorpe wrote:
> On Wed, Nov 20, 2019 at 10:14:26AM +0800, Jason Wang wrote:
>
>>>> I don't quite get the question here.
>>> In the driver model the bus_type and foo_device are closely
>>> linked.
>> I don't get the definition of "closely linked" here. Do you think the bus
>> and device implement virtual bus series are closely linked? If yes, how did
>> they achieve that?
> I mean if you have a 'foo_device' then it should be on a 'foo_bus' and
> not on some 'bar_bus', as that is how the driver core generally works.
I fully agree with you here. But isn't that just what this patch did? We
had "mdev_virtio" device on "mdev_virtio" bus not "mdev_vfio" bus.
>
>>> Creating 'mdev_device' instances and overriding the bus_type
>>> is a very abusive thing to do.
>> Ok, mdev_device (without this series) had:
>>
>> struct mdev_device {
>> struct device dev;
>> struct mdev_parent *parent;
>> guid_t uuid;
>> void *driver_data;
>> struct list_head next;
>> struct kobject *type_kobj;
>> struct device *iommu_device;
>> bool active;
>> };
>>
>> So it's nothing bus or VFIO specific. And what virtual bus had is:
> What do mean? 'struct mdev_parent *parent' is the VFIO specific
> stuff. I haven't figured out what the confusing mdev_parent is
> supposed to be,
struct mdev_parent_ops {
struct module *owner;
const struct attribute_group **dev_attr_groups;
const struct attribute_group **mdev_attr_groups;
struct attribute_group **supported_type_groups;
int (*create)(struct kobject *kobj, struct mdev_device *mdev);
int (*remove)(struct mdev_device *mdev);
int (*open)(struct mdev_device *mdev);
void (*release)(struct mdev_device *mdev);
ssize_t (*read)(struct mdev_device *mdev, char __user *buf,
size_t count, loff_t *ppos);
ssize_t (*write)(struct mdev_device *mdev, const char __user *buf,
size_t count, loff_t *ppos);
long (*ioctl)(struct mdev_device *mdev, unsigned int cmd,
unsigned long arg);
int (*mmap)(struct mdev_device *mdev, struct vm_area_struct *vma);
};
You can see that nothing is really VFIO specific here..
> or whhy the VFIO ops are linked to the parent or not
> the device..
I guess the answer the mdev_devices belongs to the same parent are
expected to have same ops.
> Honestly the whole mdev thing has a very strange take on
> how to use the driver core.
Suggestions are welcomed.
>
>>> Abusing it for other things is not appropriate. ie creating an
>>> instance and not filling in most of the vfio focused ops is an abusive
>>> thing to do.
>> Well, it's only half of the mdev_parent_ops in mdev_parent, various methods
>> could be done do be more generic to avoid duplication of codes. No?
> There are many ways to avoid duplicating code.
>
> Taking something well defined, and bolting on something unrelated just
> to share a bit of code is a very poor way to avoid code duplication.
We can have make the code better...
>> I'm sure you will need to handle other issues besides GUID which had been
>> settled by mdev e.g IOMMU and types when starting to write a real hardware
>> driver.
> The iommu framework already handles that, the mdev stuff contributes
> very little from what I can see.
Yes, but if we start from beginning to invent a new infrastructure and
we still need GUID, IOMMU, types. So it will be very similar to mdev
looks right now. So why not improve mdev?
>
>>> Most likely, at least for virtio-net, everyone else will be able to
>>> use devlink as well, making it much less clear if that GUID lifecycle
>>> stuff is a good idea or not.
>> This assumption is wrong, we hard already had at least two concrete examples
>> of vDPA device that doesn't use devlink:
>>
>> - Intel IFC where virtio is done at VF level
>> - Ali Cloud ECS instance where virtio is done at PF level
> Again, you don't explain why they couldn't use devlink.
Yes, they could, but of course for many reasons they won't use devlink.
Not only devlink, even netlink is not used or implemented in all type of
network devices.
>
> Or, why do we need GUID lifecycle stuff when these PCI devices can
> only create a single virtio and can just go ahead and do that as soon
> as they are probed.
>
> The GUID stuff was invented for slicing, which you say is not
> happening in these cases.
I think that's all about a consistent management interface, "slicing by
one" is still compatible.
Thanks
>
> Jason
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-26 12:07 ` Rob Miller
0 siblings, 0 replies; 100+ messages in thread
From: Rob Miller @ 2019-11-26 12:07 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, haotian.wang,
dri-devel, oberpar, kwankhede, linux-s390, sebott, lulu,
eperezma, pasic, borntraeger, Jason
[-- Attachment #1.1: Type: text/plain, Size: 6442 bytes --]
On Tue, Nov 19, 2019 at 9:16 PM Jason Wang <jasowang@redhat.com> wrote:
>
> On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
> >> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> >>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
> >>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> >>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> >>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> >>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> >>>>>>>> +struct bus_type mdev_virtio_bus_type;
> >>>>>>>> +
> >>>>>>>> +struct mdev_virtio_device {
> >>>>>>>> + struct mdev_device mdev;
> >>>>>>>> + const struct mdev_virtio_ops *ops;
> >>>>>>>> + u16 class_id;
> >>>>>>>> +};
> >>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on
> the
> >>>>>>> same bus?
> >>>>>> I must be missing something - which bus do they share?
> >>>>> mdev_bus_type ?
> >>>>>
> >>>>> Jason
> >>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not
> the same
> >>>> bus.
> >>> That is even worse, why involve struct mdev_device at all then?
> >>>
> >>> Jason
> >>
> >> I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
>
> I don't get the definition of "closely linked" here. Do you think the
> bus and device implement virtual bus series are closely linked? If yes,
> how did they achieve that?
>
>
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
RJM>] abusive is a subjective term. Looking at the whole context of the
vDPA framework, I still believe that extending the mdev API is preferable.
Without the framework, every vendor would have to "mediate" their own
devices, which would force us to effectively "duplicate" the mdev core code
and implement our own functionality on top. The core idea of VIRTIO is to
have a common interface and having a framework that also supports a lot of
commonality is fantastic, since we (hw vendors) too, really want to get out
of the business of crafting/verify/maintaining device drivers for every
version of Linux/Windows/... Heck, i',m hoping that a generic sample vDPA
parent driver (ie: sort of like Intel's IFCVF driver but even more so)
would be good enough for our product such that we (Brcm) don't have to
supply any driver.
>
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
>
> struct virtbus_device {
> const char *name;
> int id;
> const struct virtbus_dev_id *dev_id;
> struct device dev;
> void *data;
> };
>
> Are there any fundamental issues that you think mdev_device is abused? I
> won't expect the answers are generic objects like kobj, iommu device
> pointer etc.
>
>
> >
> >> My understanding for mdev is that it was a mediator between the driver
> and
> >> physical device when it's hard to let them talk directly due to the
> >> complexity of refactoring and maintenance.
> > Really, mdev is to support vfio with a backend other than PCI, nothing
> > more.
>
>
> That partially explain why it was called mdev. So for virito, we want
> standard virtio driver to talk with a backend other than virtio.
>
> For the issue of PCI, actually the API is generic enough to support
> device other than PCI, e.g AP bus.
>
>
> >
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various
> methods could be done do be more generic to avoid duplication of codes. No?
>
>
> >
> >> hardware that can offload virtio datapath but not control path. We want
> to
> >> present a unified interface (standard virtio) instead of a vendor
> specific
> >> interface, so a mediator level in the middle is a must. For virtio
> driver,
> >> mediator present a full virtio compatible device. For hardware, mediator
> >> will mediate the difference between the behavior defined by virtio spec
> and
> >> real hardware.
> > If you need to bind to the VFIO driver then mdev is the right thing to
> > use, otherwise it is not.
> >
> > It certainly should not be used to bind to random kernel drivers. This
> > problem is what this virtual bus idea Intel is working on might solve.
>
>
> What do you mean by random here? With this series, we have dedicated bus
> and dedicated driver with matching method to make sure the binding is
> correct.
>
RJM>] I think it's pretty clear that it's not random. The class id takes
care of the match and allows flexibility to choose vhost-mdev vs
vitrio-mdev, depending if the deployment is bare-metal vs virtualized.
>
>
> >
> > It seems the only thing people care about with mdev is the GUID
> > lifecycle stuff, but at the same time folks like Parav are saying they
> > don't want to use that lifecycle stuff and prefer devlink
> > instead.
>
>
> I'm sure you will need to handle other issues besides GUID which had
> been settled by mdev e.g IOMMU and types when starting to write a real
> hardware driver.
>
>
> >
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
>
> This assumption is wrong, we hard already had at least two concrete
> examples of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
>
> Again, the device slicing is only part of our goal. The major goal is to
> have a mediator level that can take over the virtio control path between
> a standard virtio driver and a hardware who datapath is virtio
> compatible but not control path.
>
> Thanks
>
>
> >
> > Jason
>
>
[-- Attachment #1.2: Type: text/html, Size: 8123 bytes --]
[-- Attachment #2: Type: text/plain, Size: 159 bytes --]
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-26 12:07 ` Rob Miller
0 siblings, 0 replies; 100+ messages in thread
From: Rob Miller @ 2019-11-26 12:07 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, haotian.wang,
dri-devel, oberpar, kwankhede, linux-s390, sebott, lulu,
eperezma, pasic, borntraeger, Jason Gunthorpe, jeffrey.t.kirsher,
cunming.liang, zhi.a.wang, farman, Parav Pandit, gor, intel-gfx,
kevin.tian, alex.williamson, xiao.w.wang, freude, stefanha,
zhihong.wang, rodrigo.vivi, intel-gvt-dev, hch, akrowiak, aadam,
Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
[-- Attachment #1.1: Type: text/plain, Size: 6442 bytes --]
On Tue, Nov 19, 2019 at 9:16 PM Jason Wang <jasowang@redhat.com> wrote:
>
> On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
> >> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> >>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
> >>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> >>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> >>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> >>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> >>>>>>>> +struct bus_type mdev_virtio_bus_type;
> >>>>>>>> +
> >>>>>>>> +struct mdev_virtio_device {
> >>>>>>>> + struct mdev_device mdev;
> >>>>>>>> + const struct mdev_virtio_ops *ops;
> >>>>>>>> + u16 class_id;
> >>>>>>>> +};
> >>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on
> the
> >>>>>>> same bus?
> >>>>>> I must be missing something - which bus do they share?
> >>>>> mdev_bus_type ?
> >>>>>
> >>>>> Jason
> >>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not
> the same
> >>>> bus.
> >>> That is even worse, why involve struct mdev_device at all then?
> >>>
> >>> Jason
> >>
> >> I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
>
> I don't get the definition of "closely linked" here. Do you think the
> bus and device implement virtual bus series are closely linked? If yes,
> how did they achieve that?
>
>
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
RJM>] abusive is a subjective term. Looking at the whole context of the
vDPA framework, I still believe that extending the mdev API is preferable.
Without the framework, every vendor would have to "mediate" their own
devices, which would force us to effectively "duplicate" the mdev core code
and implement our own functionality on top. The core idea of VIRTIO is to
have a common interface and having a framework that also supports a lot of
commonality is fantastic, since we (hw vendors) too, really want to get out
of the business of crafting/verify/maintaining device drivers for every
version of Linux/Windows/... Heck, i',m hoping that a generic sample vDPA
parent driver (ie: sort of like Intel's IFCVF driver but even more so)
would be good enough for our product such that we (Brcm) don't have to
supply any driver.
>
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
>
> struct virtbus_device {
> const char *name;
> int id;
> const struct virtbus_dev_id *dev_id;
> struct device dev;
> void *data;
> };
>
> Are there any fundamental issues that you think mdev_device is abused? I
> won't expect the answers are generic objects like kobj, iommu device
> pointer etc.
>
>
> >
> >> My understanding for mdev is that it was a mediator between the driver
> and
> >> physical device when it's hard to let them talk directly due to the
> >> complexity of refactoring and maintenance.
> > Really, mdev is to support vfio with a backend other than PCI, nothing
> > more.
>
>
> That partially explain why it was called mdev. So for virito, we want
> standard virtio driver to talk with a backend other than virtio.
>
> For the issue of PCI, actually the API is generic enough to support
> device other than PCI, e.g AP bus.
>
>
> >
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various
> methods could be done do be more generic to avoid duplication of codes. No?
>
>
> >
> >> hardware that can offload virtio datapath but not control path. We want
> to
> >> present a unified interface (standard virtio) instead of a vendor
> specific
> >> interface, so a mediator level in the middle is a must. For virtio
> driver,
> >> mediator present a full virtio compatible device. For hardware, mediator
> >> will mediate the difference between the behavior defined by virtio spec
> and
> >> real hardware.
> > If you need to bind to the VFIO driver then mdev is the right thing to
> > use, otherwise it is not.
> >
> > It certainly should not be used to bind to random kernel drivers. This
> > problem is what this virtual bus idea Intel is working on might solve.
>
>
> What do you mean by random here? With this series, we have dedicated bus
> and dedicated driver with matching method to make sure the binding is
> correct.
>
RJM>] I think it's pretty clear that it's not random. The class id takes
care of the match and allows flexibility to choose vhost-mdev vs
vitrio-mdev, depending if the deployment is bare-metal vs virtualized.
>
>
> >
> > It seems the only thing people care about with mdev is the GUID
> > lifecycle stuff, but at the same time folks like Parav are saying they
> > don't want to use that lifecycle stuff and prefer devlink
> > instead.
>
>
> I'm sure you will need to handle other issues besides GUID which had
> been settled by mdev e.g IOMMU and types when starting to write a real
> hardware driver.
>
>
> >
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
>
> This assumption is wrong, we hard already had at least two concrete
> examples of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
>
> Again, the device slicing is only part of our goal. The major goal is to
> have a mediator level that can take over the virtio control path between
> a standard virtio driver and a hardware who datapath is virtio
> compatible but not control path.
>
> Thanks
>
>
> >
> > Jason
>
>
[-- Attachment #1.2: Type: text/html, Size: 8123 bytes --]
[-- Attachment #2: Type: text/plain, Size: 159 bytes --]
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [Intel-gfx] [PATCH V13 4/6] mdev: introduce mediated virtio bus
@ 2019-11-26 12:07 ` Rob Miller
0 siblings, 0 replies; 100+ messages in thread
From: Rob Miller @ 2019-11-26 12:07 UTC (permalink / raw)
To: Jason Wang
Cc: rdunlap, jakub.kicinski, christophe.de.dinechin, kvm,
Michael S. Tsirkin, airlied, heiko.carstens, haotian.wang,
dri-devel, oberpar, kwankhede, linux-s390, sebott, lulu,
eperezma, pasic, borntraeger, Jason Gunthorpe, jeffrey.t.kirsher,
cunming.liang, farman, Parav Pandit, gor, intel-gfx, xiao.w.wang,
freude, stefanha, zhihong.wang, intel-gvt-dev, hch, akrowiak,
aadam, Jiri Pirko, tiwei.bie, gregkh, cohuck, linux-kernel,
maxime.coquelin, netdev, lingshan.zhu
[-- Attachment #1.1: Type: text/plain, Size: 6442 bytes --]
On Tue, Nov 19, 2019 at 9:16 PM Jason Wang <jasowang@redhat.com> wrote:
>
> On 2019/11/19 下午10:14, Jason Gunthorpe wrote:
> > On Tue, Nov 19, 2019 at 10:02:08PM +0800, Jason Wang wrote:
> >> On 2019/11/19 下午8:38, Jason Gunthorpe wrote:
> >>> On Tue, Nov 19, 2019 at 10:41:31AM +0800, Jason Wang wrote:
> >>>> On 2019/11/19 上午4:28, Jason Gunthorpe wrote:
> >>>>> On Mon, Nov 18, 2019 at 03:27:13PM -0500, Michael S. Tsirkin wrote:
> >>>>>> On Mon, Nov 18, 2019 at 01:41:00PM +0000, Jason Gunthorpe wrote:
> >>>>>>> On Mon, Nov 18, 2019 at 06:59:21PM +0800, Jason Wang wrote:
> >>>>>>>> +struct bus_type mdev_virtio_bus_type;
> >>>>>>>> +
> >>>>>>>> +struct mdev_virtio_device {
> >>>>>>>> + struct mdev_device mdev;
> >>>>>>>> + const struct mdev_virtio_ops *ops;
> >>>>>>>> + u16 class_id;
> >>>>>>>> +};
> >>>>>>> This seems to share nothing with mdev (ie mdev-vfio), why is it on
> the
> >>>>>>> same bus?
> >>>>>> I must be missing something - which bus do they share?
> >>>>> mdev_bus_type ?
> >>>>>
> >>>>> Jason
> >>>> Note: virtio has its own bus: mdev_virtio_bus_type. So they are not
> the same
> >>>> bus.
> >>> That is even worse, why involve struct mdev_device at all then?
> >>>
> >>> Jason
> >>
> >> I don't quite get the question here.
> > In the driver model the bus_type and foo_device are closely
> > linked.
>
>
> I don't get the definition of "closely linked" here. Do you think the
> bus and device implement virtual bus series are closely linked? If yes,
> how did they achieve that?
>
>
> > Creating 'mdev_device' instances and overriding the bus_type
> > is a very abusive thing to do.
>
RJM>] abusive is a subjective term. Looking at the whole context of the
vDPA framework, I still believe that extending the mdev API is preferable.
Without the framework, every vendor would have to "mediate" their own
devices, which would force us to effectively "duplicate" the mdev core code
and implement our own functionality on top. The core idea of VIRTIO is to
have a common interface and having a framework that also supports a lot of
commonality is fantastic, since we (hw vendors) too, really want to get out
of the business of crafting/verify/maintaining device drivers for every
version of Linux/Windows/... Heck, i',m hoping that a generic sample vDPA
parent driver (ie: sort of like Intel's IFCVF driver but even more so)
would be good enough for our product such that we (Brcm) don't have to
supply any driver.
>
>
> Ok, mdev_device (without this series) had:
>
> struct mdev_device {
> struct device dev;
> struct mdev_parent *parent;
> guid_t uuid;
> void *driver_data;
> struct list_head next;
> struct kobject *type_kobj;
> struct device *iommu_device;
> bool active;
> };
>
> So it's nothing bus or VFIO specific. And what virtual bus had is:
>
> struct virtbus_device {
> const char *name;
> int id;
> const struct virtbus_dev_id *dev_id;
> struct device dev;
> void *data;
> };
>
> Are there any fundamental issues that you think mdev_device is abused? I
> won't expect the answers are generic objects like kobj, iommu device
> pointer etc.
>
>
> >
> >> My understanding for mdev is that it was a mediator between the driver
> and
> >> physical device when it's hard to let them talk directly due to the
> >> complexity of refactoring and maintenance.
> > Really, mdev is to support vfio with a backend other than PCI, nothing
> > more.
>
>
> That partially explain why it was called mdev. So for virito, we want
> standard virtio driver to talk with a backend other than virtio.
>
> For the issue of PCI, actually the API is generic enough to support
> device other than PCI, e.g AP bus.
>
>
> >
> > Abusing it for other things is not appropriate. ie creating an
> > instance and not filling in most of the vfio focused ops is an abusive
> > thing to do.
>
>
> Well, it's only half of the mdev_parent_ops in mdev_parent, various
> methods could be done do be more generic to avoid duplication of codes. No?
>
>
> >
> >> hardware that can offload virtio datapath but not control path. We want
> to
> >> present a unified interface (standard virtio) instead of a vendor
> specific
> >> interface, so a mediator level in the middle is a must. For virtio
> driver,
> >> mediator present a full virtio compatible device. For hardware, mediator
> >> will mediate the difference between the behavior defined by virtio spec
> and
> >> real hardware.
> > If you need to bind to the VFIO driver then mdev is the right thing to
> > use, otherwise it is not.
> >
> > It certainly should not be used to bind to random kernel drivers. This
> > problem is what this virtual bus idea Intel is working on might solve.
>
>
> What do you mean by random here? With this series, we have dedicated bus
> and dedicated driver with matching method to make sure the binding is
> correct.
>
RJM>] I think it's pretty clear that it's not random. The class id takes
care of the match and allows flexibility to choose vhost-mdev vs
vitrio-mdev, depending if the deployment is bare-metal vs virtualized.
>
>
> >
> > It seems the only thing people care about with mdev is the GUID
> > lifecycle stuff, but at the same time folks like Parav are saying they
> > don't want to use that lifecycle stuff and prefer devlink
> > instead.
>
>
> I'm sure you will need to handle other issues besides GUID which had
> been settled by mdev e.g IOMMU and types when starting to write a real
> hardware driver.
>
>
> >
> > Most likely, at least for virtio-net, everyone else will be able to
> > use devlink as well, making it much less clear if that GUID lifecycle
> > stuff is a good idea or not.
>
>
> This assumption is wrong, we hard already had at least two concrete
> examples of vDPA device that doesn't use devlink:
>
> - Intel IFC where virtio is done at VF level
> - Ali Cloud ECS instance where virtio is done at PF level
>
> Again, the device slicing is only part of our goal. The major goal is to
> have a mediator level that can take over the virtio control path between
> a standard virtio driver and a hardware who datapath is virtio
> compatible but not control path.
>
> Thanks
>
>
> >
> > Jason
>
>
[-- Attachment #1.2: Type: text/html, Size: 8123 bytes --]
[-- Attachment #2: Type: text/plain, Size: 159 bytes --]
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
^ permalink raw reply [flat|nested] 100+ messages in thread
end of thread, other threads:[~2019-11-27 8:09 UTC | newest]
Thread overview: 100+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-11-18 10:59 [PATCH V13 0/6] mdev based hardware virtio offloading support Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` [PATCH V13 1/6] mdev: make mdev bus agnostic Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-19 3:08 ` Randy Dunlap
2019-11-19 3:08 ` [Intel-gfx] " Randy Dunlap
2019-11-19 3:08 ` Randy Dunlap
2019-11-19 3:08 ` Randy Dunlap
2019-11-18 10:59 ` [PATCH V13 2/6] mdev: split out VFIO bus specific parent ops Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` [PATCH V13 3/6] mdev: move to drivers/ Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-19 3:11 ` Randy Dunlap
2019-11-19 3:11 ` [Intel-gfx] " Randy Dunlap
2019-11-19 3:11 ` Randy Dunlap
2019-11-19 3:11 ` Randy Dunlap
2019-11-18 10:59 ` [PATCH V13 4/6] mdev: introduce mediated virtio bus Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 13:41 ` Jason Gunthorpe
2019-11-18 13:41 ` [Intel-gfx] " Jason Gunthorpe
2019-11-18 13:41 ` Jason Gunthorpe
2019-11-18 20:27 ` Michael S. Tsirkin
2019-11-18 20:27 ` [Intel-gfx] " Michael S. Tsirkin
2019-11-18 20:27 ` Michael S. Tsirkin
2019-11-18 20:28 ` Jason Gunthorpe
2019-11-18 20:28 ` [Intel-gfx] " Jason Gunthorpe
2019-11-18 20:28 ` Jason Gunthorpe
2019-11-19 2:41 ` Jason Wang
2019-11-19 2:41 ` [Intel-gfx] " Jason Wang
2019-11-19 2:41 ` Jason Wang
2019-11-19 12:38 ` Jason Gunthorpe
2019-11-19 12:38 ` [Intel-gfx] " Jason Gunthorpe
2019-11-19 14:02 ` Jason Wang
2019-11-19 14:02 ` [Intel-gfx] " Jason Wang
2019-11-19 14:02 ` Jason Wang
2019-11-19 14:14 ` Jason Gunthorpe
2019-11-19 14:14 ` [Intel-gfx] " Jason Gunthorpe
2019-11-20 2:14 ` Jason Wang
2019-11-20 2:14 ` [Intel-gfx] " Jason Wang
2019-11-20 2:14 ` Jason Wang
2019-11-20 13:49 ` Jason Gunthorpe
2019-11-20 13:49 ` [Intel-gfx] " Jason Gunthorpe
2019-11-20 13:49 ` Jason Gunthorpe
2019-11-21 3:05 ` Jason Wang
2019-11-21 3:05 ` [Intel-gfx] " Jason Wang
2019-11-21 3:05 ` Jason Wang
2019-11-26 12:07 ` Rob Miller
2019-11-26 12:07 ` [Intel-gfx] " Rob Miller
2019-11-26 12:07 ` Rob Miller
2019-11-19 2:40 ` Jason Wang
2019-11-19 2:40 ` [Intel-gfx] " Jason Wang
2019-11-19 2:40 ` Jason Wang
2019-11-19 3:13 ` Randy Dunlap
2019-11-19 3:13 ` [Intel-gfx] " Randy Dunlap
2019-11-19 3:13 ` Randy Dunlap
2019-11-19 3:13 ` Randy Dunlap
2019-11-18 10:59 ` [PATCH V13 5/6] virtio: introduce a mdev based transport Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` [PATCH V13 6/6] docs: sample driver to demonstrate how to implement virtio-mdev framework Jason Wang
2019-11-18 10:59 ` [Intel-gfx] " Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 10:59 ` Jason Wang
2019-11-18 15:17 ` Greg KH
2019-11-18 15:17 ` [Intel-gfx] " Greg KH
2019-11-18 15:17 ` Greg KH
2019-11-18 15:17 ` Greg KH
2019-11-19 3:03 ` Jason Wang
2019-11-19 3:03 ` [Intel-gfx] " Jason Wang
2019-11-19 3:03 ` Jason Wang
2019-11-19 3:03 ` Jason Wang
2019-11-19 12:40 ` Jason Gunthorpe
2019-11-19 12:40 ` [Intel-gfx] " Jason Gunthorpe
2019-11-19 14:07 ` Jason Wang
2019-11-19 14:07 ` [Intel-gfx] " Jason Wang
2019-11-19 14:07 ` Jason Wang
2019-11-19 16:39 ` Jason Gunthorpe
2019-11-19 16:39 ` [Intel-gfx] " Jason Gunthorpe
2019-11-18 15:45 ` Cornelia Huck
2019-11-18 15:45 ` [Intel-gfx] " Cornelia Huck
2019-11-18 15:45 ` Cornelia Huck
2019-11-18 15:45 ` Cornelia Huck
2019-11-19 3:04 ` Jason Wang
2019-11-19 3:04 ` [Intel-gfx] " Jason Wang
2019-11-19 3:04 ` Jason Wang
2019-11-19 3:04 ` Jason Wang
2019-11-18 13:35 ` ✗ Fi.CI.CHECKPATCH: warning for mdev based hardware virtio offloading support Patchwork
2019-11-18 13:35 ` [Intel-gfx] " Patchwork
2019-11-18 13:51 ` ✗ Fi.CI.DOCS: " Patchwork
2019-11-18 13:51 ` [Intel-gfx] " Patchwork
2019-11-18 14:25 ` ✗ Fi.CI.BAT: failure " Patchwork
2019-11-18 14:25 ` [Intel-gfx] " Patchwork
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.