Enable interpretive execution of zPCI instructions + adapter interruption
forwarding for s390x KVM vfio-pci. This is done by triggering a routine
when the VFIO group is associated with the KVM guest, transmitting to
firmware a special token (GISA designation) to enable that specific guest
for interpretive execution on that zPCI device. Load/store interpreation
enablement is then controlled by userspace (based upon whether or not a
SHM bit is placed in the virtual function handle). Adapter Event
Notification interpretation is controlled from userspace via a new KVM
ioctl.
By allowing intepretation of zPCI instructions and firmware delivery of
interrupts to guests, we can reduce the frequency of guest SIE exits for
zPCI.
From the perspective of guest configuration, you passthrough zPCI devices
in the same manner as before, with intepretation support being used by
default if available in kernel+qemu.
Some minor changes are already suggested for the QEMU series, but for
now the v5 one is still compatible (with one exception: in
linux-headers/linux/kvm.h KVM_CAP_S390_ZPCI_OP is now changed to 216)
https://lore.kernel.org/kvm/[email protected]/
Changelog v6->v7:
- Add some more R-b tags
- Mark a few routines static (kernel bot)
- Add a few lockdep_assert_held calls (Christian)
- Add a new condition under which locked_vm is defined (Christian)
- After discussions with Jason, let's go back to using a Kconfig option
to control vfio-pci-zdev inclusion so we can avoid some symbol_get
messiness. The name was chosen so that we can re-introduce zdev
for non-KVM environments in the future, should it be needed.
- Since we re-introduce a Kconfig option, we can go back to using this in
various places instead of CONFIG_VFIO_PCI.
- Expose separate interfaces for register and unregister of kvm (Jason)
- Cleanup KVM association during zdev device_close; if the device fd is
closed before we receive the kvm notifier, we will never see the
notification because we unregister it, but cleanup must still be done.
Conversely, if the kvm fd closes before the device is closed, we
should get the NULL notifier from kvm_destroy_devices since the notifier
has not yet been unregistered.
- As a result of the above, we no longer need to clear the pci list during
vm exit. But to be safe let's WARN_ON_ONCE if it's found to be non-empty
during free_vm.
- Add kzdev_lock to protect the kzdev
- uapi KVM_CAP_S390_ZPCI_OP moved to 216 (previously 214)
Matthew Rosato (22):
s390/sclp: detect the zPCI load/store interpretation facility
s390/sclp: detect the AISII facility
s390/sclp: detect the AENI facility
s390/sclp: detect the AISI facility
s390/airq: pass more TPI info to airq handlers
s390/airq: allow for airq structure that uses an input vector
s390/pci: externalize the SIC operation controls and routine
s390/pci: stash associated GISA designation
s390/pci: stash dtsm and maxstbl
vfio/pci: introduce CONFIG_VFIO_PCI_ZDEV_KVM
KVM: s390: pci: add basic kvm_zdev structure
KVM: s390: pci: do initial setup for AEN interpretation
KVM: s390: pci: enable host forwarding of Adapter Event Notifications
KVM: s390: mechanism to enable guest zPCI Interpretation
KVM: s390: pci: provide routines for enabling/disabling interrupt
forwarding
KVM: s390: pci: add routines to start/stop interpretive execution
vfio-pci/zdev: add open/close device hooks
vfio-pci/zdev: add function handle to clp base capability
vfio-pci/zdev: different maxstbl for interpreted devices
KVM: s390: add KVM_S390_ZPCI_OP to manage guest zPCI devices
KVM: s390: introduce CPU feature for zPCI Interpretation
MAINTAINERS: additional files related kvm s390 pci passthrough
Documentation/virt/kvm/api.rst | 45 ++
MAINTAINERS | 1 +
arch/s390/include/asm/airq.h | 7 +-
arch/s390/include/asm/kvm_host.h | 26 ++
arch/s390/include/asm/pci.h | 13 +
arch/s390/include/asm/pci_clp.h | 9 +-
arch/s390/include/asm/pci_insn.h | 29 +-
arch/s390/include/asm/sclp.h | 4 +
arch/s390/include/asm/tpi.h | 13 +
arch/s390/include/uapi/asm/kvm.h | 1 +
arch/s390/kvm/Makefile | 1 +
arch/s390/kvm/interrupt.c | 96 ++++-
arch/s390/kvm/kvm-s390.c | 94 ++++-
arch/s390/kvm/kvm-s390.h | 10 +
arch/s390/kvm/pci.c | 686 +++++++++++++++++++++++++++++++
arch/s390/kvm/pci.h | 88 ++++
arch/s390/pci/pci.c | 16 +
arch/s390/pci/pci_clp.c | 7 +
arch/s390/pci/pci_insn.c | 4 +-
arch/s390/pci/pci_irq.c | 48 ++-
drivers/s390/char/sclp_early.c | 4 +
drivers/s390/cio/airq.c | 12 +-
drivers/s390/cio/qdio_thinint.c | 6 +-
drivers/s390/crypto/ap_bus.c | 9 +-
drivers/s390/virtio/virtio_ccw.c | 6 +-
drivers/vfio/pci/Kconfig | 11 +
drivers/vfio/pci/Makefile | 2 +-
drivers/vfio/pci/vfio_pci_core.c | 2 +
drivers/vfio/pci/vfio_pci_zdev.c | 65 ++-
include/linux/sched/user.h | 3 +-
include/linux/vfio_pci_core.h | 12 +-
include/uapi/linux/kvm.h | 31 ++
include/uapi/linux/vfio_zdev.h | 7 +
33 files changed, 1313 insertions(+), 55 deletions(-)
create mode 100644 arch/s390/kvm/pci.c
create mode 100644 arch/s390/kvm/pci.h
--
2.27.0
When doing load/store interpretation, the maximum store block length is
determined by the underlying firmware, not the host kernel API. Reflect
that in the associated Query PCI Function Group clp capability and let
userspace decide which is appropriate to present to the guest.
Reviewed-by: Pierre Morel <[email protected]>
Signed-off-by: Matthew Rosato <[email protected]>
---
drivers/vfio/pci/vfio_pci_zdev.c | 6 ++++--
include/uapi/linux/vfio_zdev.h | 4 ++++
2 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/vfio/pci/vfio_pci_zdev.c b/drivers/vfio/pci/vfio_pci_zdev.c
index c7557f80ec64..067383f84cd8 100644
--- a/drivers/vfio/pci/vfio_pci_zdev.c
+++ b/drivers/vfio/pci/vfio_pci_zdev.c
@@ -45,14 +45,16 @@ static int zpci_group_cap(struct zpci_dev *zdev, struct vfio_info_cap *caps)
{
struct vfio_device_info_cap_zpci_group cap = {
.header.id = VFIO_DEVICE_INFO_CAP_ZPCI_GROUP,
- .header.version = 1,
+ .header.version = 2,
.dasm = zdev->dma_mask,
.msi_addr = zdev->msi_addr,
.flags = VFIO_DEVICE_INFO_ZPCI_FLAG_REFRESH,
.mui = zdev->fmb_update,
.noi = zdev->max_msi,
.maxstbl = ZPCI_MAX_WRITE_SIZE,
- .version = zdev->version
+ .version = zdev->version,
+ .reserved = 0,
+ .imaxstbl = zdev->maxstbl
};
return vfio_info_add_capability(caps, &cap.header, sizeof(cap));
diff --git a/include/uapi/linux/vfio_zdev.h b/include/uapi/linux/vfio_zdev.h
index 78c022af3d29..77f2aff1f27e 100644
--- a/include/uapi/linux/vfio_zdev.h
+++ b/include/uapi/linux/vfio_zdev.h
@@ -50,6 +50,10 @@ struct vfio_device_info_cap_zpci_group {
__u16 noi; /* Maximum number of MSIs */
__u16 maxstbl; /* Maximum Store Block Length */
__u8 version; /* Supported PCI Version */
+ /* End of version 1 */
+ __u8 reserved;
+ __u16 imaxstbl; /* Maximum Interpreted Store Block Length */
+ /* End of version 2 */
};
/**
--
2.27.0