lkml.org 
[lkml]   [2020]   [Nov]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRE: [PATCH v15 6/9] arm64/kvm: Add hypercall service for kvm ptp.
Date
Hi Marc,

> -----Original Message-----
> From: Marc Zyngier <maz@kernel.org>
> Sent: Monday, November 23, 2020 7:59 PM
> To: Jianyong Wu <Jianyong.Wu@arm.com>
> Cc: Justin He <Justin.He@arm.com>; kvm@vger.kernel.org;
> netdev@vger.kernel.org; richardcochran@gmail.com; linux-
> kernel@vger.kernel.org; sean.j.christopherson@intel.com; Steven Price
> <Steven.Price@arm.com>; Andre Przywara <Andre.Przywara@arm.com>;
> john.stultz@linaro.org; yangbo.lu@nxp.com; pbonzini@redhat.com;
> tglx@linutronix.de; nd <nd@arm.com>; will@kernel.org;
> kvmarm@lists.cs.columbia.edu; linux-arm-kernel@lists.infradead.org
> Subject: Re: [PATCH v15 6/9] arm64/kvm: Add hypercall service for kvm ptp.
>
> On 2020-11-23 10:44, Marc Zyngier wrote:
> > On 2020-11-11 06:22, Jianyong Wu wrote:
> >> ptp_kvm will get this service through SMCC call.
> >> The service offers wall time and cycle count of host to guest.
> >> The caller must specify whether they want the host cycle count or the
> >> difference between host cycle count and cntvoff.
> >>
> >> Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
> >> ---
> >> arch/arm64/kvm/hypercalls.c | 61
> >> +++++++++++++++++++++++++++++++++++++
> >> include/linux/arm-smccc.h | 17 +++++++++++
> >> 2 files changed, 78 insertions(+)
> >>
> >> diff --git a/arch/arm64/kvm/hypercalls.c
> >> b/arch/arm64/kvm/hypercalls.c index b9d8607083eb..f7d189563f3d
> 100644
> >> --- a/arch/arm64/kvm/hypercalls.c
> >> +++ b/arch/arm64/kvm/hypercalls.c
> >> @@ -9,6 +9,51 @@
> >> #include <kvm/arm_hypercalls.h>
> >> #include <kvm/arm_psci.h>
> >>
> >> +static void kvm_ptp_get_time(struct kvm_vcpu *vcpu, u64 *val) {
> >> + struct system_time_snapshot systime_snapshot;
> >> + u64 cycles = ~0UL;
> >> + u32 feature;
> >> +
> >> + /*
> >> + * system time and counter value must captured in the same
> >> + * time to keep consistency and precision.
> >> + */
> >> + ktime_get_snapshot(&systime_snapshot);
> >> +
> >> + // binding ptp_kvm clocksource to arm_arch_counter
> >> + if (systime_snapshot.cs_id != CSID_ARM_ARCH_COUNTER)
> >> + return;
> >> +
> >> + val[0] = upper_32_bits(systime_snapshot.real);
> >> + val[1] = lower_32_bits(systime_snapshot.real);
> >
> > What is the endianness of these values? I can't see it defined
> > anywhere, and this is likely not to work if guest and hypervisor don't
> > align.
>
> Scratch that. This is all passed via registers, so the endianness of the data is
> irrelevant. Please discard any comment about endianness I made in this
> review.
>
Yeah, these data process and transfer are no relationship with endianness. Thanks.

> The documentation aspect still requires to be beefed up.

So the endianness description will be "no endianness restriction".

Thanks
Jianyong

>
> Thanks,
>
> M.
> --
> Jazz is not dead. It just smells funny...

\
 
 \ /
  Last update: 2020-11-24 06:14    [W:0.095 / U:2.088 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site