Cornelia Huck
2019-Jun-03 12:06 UTC
[PATCH v3 3/8] s390/cio: add basic protected virtualization support
On Wed, 29 May 2019 14:26:52 +0200 Michael Mueller <mimu at linux.ibm.com> wrote:> From: Halil Pasic <pasic at linux.ibm.com> > > As virtio-ccw devices are channel devices, we need to use the dma area > for any communication with the hypervisor."we need to use the dma area within the common I/O layer for any communication with the hypervisor. Note that we do not need to use that area for control blocks directly referenced by instructions, e.g. the orb." ...although I'm still not particularly confident about the actual distinction here? I'm trusting you that you actually have tested it, though :)> > It handles neither QDIO in the common code, nor any device type specific > stuff (like channel programs constructed by the DASD driver). > > An interesting side effect is that virtio structures are now going to > get allocated in 31 bit addressable storage. > > Signed-off-by: Halil Pasic <pasic at linux.ibm.com> > Reviewed-by: Sebastian Ott <sebott at linux.ibm.com> > Signed-off-by: Michael Mueller <mimu at linux.ibm.com> > --- > arch/s390/include/asm/ccwdev.h | 4 +++ > drivers/s390/cio/ccwreq.c | 9 +++--- > drivers/s390/cio/device.c | 67 +++++++++++++++++++++++++++++++++------- > drivers/s390/cio/device_fsm.c | 49 +++++++++++++++++------------ > drivers/s390/cio/device_id.c | 20 ++++++------ > drivers/s390/cio/device_ops.c | 21 +++++++++++-- > drivers/s390/cio/device_pgid.c | 22 +++++++------ > drivers/s390/cio/device_status.c | 24 +++++++------- > drivers/s390/cio/io_sch.h | 20 +++++++++--- > drivers/s390/virtio/virtio_ccw.c | 10 ------ > 10 files changed, 163 insertions(+), 83 deletions(-)(...)> @@ -1593,20 +1625,31 @@ struct ccw_device * __init ccw_device_create_console(struct ccw_driver *drv) > return ERR_CAST(sch); > > io_priv = kzalloc(sizeof(*io_priv), GFP_KERNEL | GFP_DMA); > - if (!io_priv) { > - put_device(&sch->dev); > - return ERR_PTR(-ENOMEM); > - } > + if (!io_priv) > + goto err_priv; > + io_priv->dma_area = dma_alloc_coherent(&sch->dev, > + sizeof(*io_priv->dma_area), > + &io_priv->dma_area_dma, GFP_KERNEL); > + if (!io_priv->dma_area) > + goto err_dma_area; > set_io_private(sch, io_priv); > cdev = io_subchannel_create_ccwdev(sch); > if (IS_ERR(cdev)) { > put_device(&sch->dev); > + dma_free_coherent(&sch->dev, sizeof(*io_priv->dma_area), > + io_priv->dma_area, io_priv->dma_area_dma); > kfree(io_priv);<pre-existing, not introduced by this patch> Shouldn't that branch do set_io_private(sch, NULL)? Not sure if any code would make use of it, but it's probably better to clean out references to freed objects. </pre-existing, not introduced by this patch>> return cdev; > } > cdev->drv = drv; > ccw_device_set_int_class(cdev); > return cdev; > + > +err_dma_area: > + kfree(io_priv); > +err_priv: > + put_device(&sch->dev); > + return ERR_PTR(-ENOMEM); > } > > void __init ccw_device_destroy_console(struct ccw_device *cdev)With the reservations above, Reviewed-by: Cornelia Huck <cohuck at redhat.com>
Michael Mueller
2019-Jun-03 12:45 UTC
[PATCH v3 3/8] s390/cio: add basic protected virtualization support
On 03.06.19 14:06, Cornelia Huck wrote:> On Wed, 29 May 2019 14:26:52 +0200 > Michael Mueller <mimu at linux.ibm.com> wrote: > >> From: Halil Pasic <pasic at linux.ibm.com> >> >> As virtio-ccw devices are channel devices, we need to use the dma area >> for any communication with the hypervisor. > > "we need to use the dma area within the common I/O layer for any > communication with the hypervisor. Note that we do not need to use that > area for control blocks directly referenced by instructions, e.g. the > orb."using this now> > ...although I'm still not particularly confident about the actual > distinction here? I'm trusting you that you actually have tested it, > though :) > >> >> It handles neither QDIO in the common code, nor any device type specific >> stuff (like channel programs constructed by the DASD driver). >> >> An interesting side effect is that virtio structures are now going to >> get allocated in 31 bit addressable storage. >> >> Signed-off-by: Halil Pasic <pasic at linux.ibm.com> >> Reviewed-by: Sebastian Ott <sebott at linux.ibm.com> >> Signed-off-by: Michael Mueller <mimu at linux.ibm.com> >> --- >> arch/s390/include/asm/ccwdev.h | 4 +++ >> drivers/s390/cio/ccwreq.c | 9 +++--- >> drivers/s390/cio/device.c | 67 +++++++++++++++++++++++++++++++++------- >> drivers/s390/cio/device_fsm.c | 49 +++++++++++++++++------------ >> drivers/s390/cio/device_id.c | 20 ++++++------ >> drivers/s390/cio/device_ops.c | 21 +++++++++++-- >> drivers/s390/cio/device_pgid.c | 22 +++++++------ >> drivers/s390/cio/device_status.c | 24 +++++++------- >> drivers/s390/cio/io_sch.h | 20 +++++++++--- >> drivers/s390/virtio/virtio_ccw.c | 10 ------ >> 10 files changed, 163 insertions(+), 83 deletions(-) > > (...) > >> @@ -1593,20 +1625,31 @@ struct ccw_device * __init ccw_device_create_console(struct ccw_driver *drv) >> return ERR_CAST(sch); >> >> io_priv = kzalloc(sizeof(*io_priv), GFP_KERNEL | GFP_DMA); >> - if (!io_priv) { >> - put_device(&sch->dev); >> - return ERR_PTR(-ENOMEM); >> - } >> + if (!io_priv) >> + goto err_priv; >> + io_priv->dma_area = dma_alloc_coherent(&sch->dev, >> + sizeof(*io_priv->dma_area), >> + &io_priv->dma_area_dma, GFP_KERNEL); >> + if (!io_priv->dma_area) >> + goto err_dma_area; >> set_io_private(sch, io_priv); >> cdev = io_subchannel_create_ccwdev(sch); >> if (IS_ERR(cdev)) { >> put_device(&sch->dev); >> + dma_free_coherent(&sch->dev, sizeof(*io_priv->dma_area), >> + io_priv->dma_area, io_priv->dma_area_dma); >> kfree(io_priv); > > <pre-existing, not introduced by this patch> > Shouldn't that branch do set_io_private(sch, NULL)? Not sure if any > code would make use of it, but it's probably better to clean out > references to freed objects.Added behind kfree(). I hope nobody asks for a separate patch. ;)> </pre-existing, not introduced by this patch> > >> return cdev; >> } >> cdev->drv = drv; >> ccw_device_set_int_class(cdev); >> return cdev; >> + >> +err_dma_area: >> + kfree(io_priv); >> +err_priv: >> + put_device(&sch->dev); >> + return ERR_PTR(-ENOMEM); >> } >> >> void __init ccw_device_destroy_console(struct ccw_device *cdev) > > With the reservations above, > Reviewed-by: Cornelia Huck <cohuck at redhat.com> >Thanks, Michael
Cornelia Huck
2019-Jun-03 13:42 UTC
[PATCH v3 3/8] s390/cio: add basic protected virtualization support
On Mon, 3 Jun 2019 14:45:03 +0200 Michael Mueller <mimu at linux.ibm.com> wrote:> On 03.06.19 14:06, Cornelia Huck wrote: > > On Wed, 29 May 2019 14:26:52 +0200 > > Michael Mueller <mimu at linux.ibm.com> wrote:> >> @@ -1593,20 +1625,31 @@ struct ccw_device * __init ccw_device_create_console(struct ccw_driver *drv) > >> return ERR_CAST(sch); > >> > >> io_priv = kzalloc(sizeof(*io_priv), GFP_KERNEL | GFP_DMA); > >> - if (!io_priv) { > >> - put_device(&sch->dev); > >> - return ERR_PTR(-ENOMEM); > >> - } > >> + if (!io_priv) > >> + goto err_priv; > >> + io_priv->dma_area = dma_alloc_coherent(&sch->dev, > >> + sizeof(*io_priv->dma_area), > >> + &io_priv->dma_area_dma, GFP_KERNEL); > >> + if (!io_priv->dma_area) > >> + goto err_dma_area; > >> set_io_private(sch, io_priv); > >> cdev = io_subchannel_create_ccwdev(sch); > >> if (IS_ERR(cdev)) { > >> put_device(&sch->dev); > >> + dma_free_coherent(&sch->dev, sizeof(*io_priv->dma_area), > >> + io_priv->dma_area, io_priv->dma_area_dma); > >> kfree(io_priv); > > > > <pre-existing, not introduced by this patch> > > Shouldn't that branch do set_io_private(sch, NULL)? Not sure if any > > code would make use of it, but it's probably better to clean out > > references to freed objects. > > Added behind kfree(). I hope nobody asks for a separate patch. ;)I would probably have added it just before the kfree, but I'm not asking for a separate patch ;)> > > </pre-existing, not introduced by this patch> > > > >> return cdev; > >> } > >> cdev->drv = drv; > >> ccw_device_set_int_class(cdev); > >> return cdev; > >> + > >> +err_dma_area: > >> + kfree(io_priv); > >> +err_priv: > >> + put_device(&sch->dev); > >> + return ERR_PTR(-ENOMEM); > >> } > >> > >> void __init ccw_device_destroy_console(struct ccw_device *cdev)
Reasonably Related Threads
- [PATCH v3 3/8] s390/cio: add basic protected virtualization support
- [PATCH v3 3/8] s390/cio: add basic protected virtualization support
- [PATCH v2 3/8] s390/cio: add basic protected virtualization support
- [PATCH v2 3/8] s390/cio: add basic protected virtualization support
- [PATCH v3 3/8] s390/cio: add basic protected virtualization support