Messages in this thread | | | Date | Fri, 18 Nov 2022 15:37:26 +0100 | Subject | Re: [PATCH v3 1/9] KVM: s390: Extend MEM_OP ioctl by storage key checked cmpxchg | From | Thomas Huth <> |
| |
On 18/11/2022 11.12, Heiko Carstens wrote: > On Thu, Nov 17, 2022 at 11:17:50PM +0100, Janis Schoetterl-Glausch wrote: >> User space can use the MEM_OP ioctl to make storage key checked reads >> and writes to the guest, however, it has no way of performing atomic, >> key checked, accesses to the guest. >> Extend the MEM_OP ioctl in order to allow for this, by adding a cmpxchg >> mode. For now, support this mode for absolute accesses only. >> >> This mode can be use, for example, to set the device-state-change >> indicator and the adapter-local-summary indicator atomically. >> >> Signed-off-by: Janis Schoetterl-Glausch <scgl@linux.ibm.com> >> --- >> include/uapi/linux/kvm.h | 5 ++ >> arch/s390/kvm/gaccess.h | 3 ++ >> arch/s390/kvm/gaccess.c | 101 +++++++++++++++++++++++++++++++++++++++ >> arch/s390/kvm/kvm-s390.c | 35 +++++++++++++- >> 4 files changed, 142 insertions(+), 2 deletions(-) >> >> diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h >> index 0d5d4419139a..1f36be5493e6 100644 >> --- a/include/uapi/linux/kvm.h >> +++ b/include/uapi/linux/kvm.h >> @@ -588,6 +588,8 @@ struct kvm_s390_mem_op { >> struct { >> __u8 ar; /* the access register number */ >> __u8 key; /* access key, ignored if flag unset */ >> + __u8 pad1[6]; /* ignored */ >> + __u64 old_p; /* ignored if flag unset */ > > Just one comment: the suffix "_p" for pointer is quite unusual within > the kernel. This also would be the first of its kind within kvm.h. > Usually there is either no suffix or "_addr". > So for consistency reasons I would suggest to change this to one of > the common variants. > > The code itself looks good from my point of view, even though for the > sake of simplicity I would have put the complete sign/zero extended > 128 bit old value into the structure, instead of having a pointer to > the value.
See https://lore.kernel.org/kvm/37197cfe-d109-332f-089b-266d7e8e23f8@redhat.com/ ... it would break the "IOW" definition of the ioctl. It can be done, but that confuses tools like valgrind, as far as I know. So I think the idea with the pointer is better in this case.
Thomas
| |