RE: [PATCH v3 07/10] clocksource/drivers/hyper-v: Handle vDSO differences inline

[Michael Kelley]

From: Michael Kelley <mikelley@xxxxxxxxxxxxx> Sent: Tuesday, March 2, 2021 3:07 PM
> 
> From: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx> Sent: Tuesday, March 2, 2021 2:14 PM
> >
> > On 02/03/2021 22:38, Michael Kelley wrote:
> > > While the driver for the Hyper-V Reference TSC and STIMERs is architecture
> > > neutral, vDSO is implemented for x86/x64, but not for ARM64.  Current code
> > > calls into utility functions under arch/x86 (and coming, under arch/arm64)
> > > to handle the difference.
> > >
> > > Change this approach to handle the difference inline based on whether
> > > VDSO_CLOCK_MODE_HVCLOCK is present.  The new approach removes code under
> > > arch/* since the difference is tied more to the specifics of the Linux
> > > implementation than to the architecture.
> > >
> > > No functional change.
> > >
> > > Signed-off-by: Michael Kelley <mikelley@xxxxxxxxxxxxx>
> > > Reviewed-by: Boqun Feng <boqun.feng@xxxxxxxxx>
> > > ---
> > >  arch/x86/include/asm/mshyperv.h    |  4 ----
> > >  drivers/clocksource/hyperv_timer.c | 10 ++++++++--
> > >  2 files changed, 8 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/arch/x86/include/asm/mshyperv.h b/arch/x86/include/asm/mshyperv.h
> > > index c10dd1c..4f566db 100644
> > > --- a/arch/x86/include/asm/mshyperv.h
> > > +++ b/arch/x86/include/asm/mshyperv.h
> > > @@ -27,10 +27,6 @@ static inline u64 hv_get_register(unsigned int reg)
> > >  	return value;
> > >  }
> > >
> > > -#define hv_set_clocksource_vdso(val) \
> > > -	((val).vdso_clock_mode = VDSO_CLOCKMODE_HVCLOCK)
> > > -#define hv_enable_vdso_clocksource() \
> > > -	vclocks_set_used(VDSO_CLOCKMODE_HVCLOCK);
> > >  #define hv_get_raw_timer() rdtsc_ordered()
> > >
> > >  /*
> > > diff --git a/drivers/clocksource/hyperv_timer.c b/drivers/clocksource/hyperv_timer.c
> > > index c73c127..06984fa 100644
> > > --- a/drivers/clocksource/hyperv_timer.c
> > > +++ b/drivers/clocksource/hyperv_timer.c
> > > @@ -370,11 +370,13 @@ static void resume_hv_clock_tsc(struct clocksource *arg)
> > >  	hv_set_register(HV_REGISTER_REFERENCE_TSC, tsc_msr);
> > >  }
> > >
> > > +#ifdef VDSO_CLOCKMODE_HVCLOCK
> > >  static int hv_cs_enable(struct clocksource *cs)
> > >  {
> > > -	hv_enable_vdso_clocksource();
> > > +	vclocks_set_used(VDSO_CLOCKMODE_HVCLOCK);
> > >  	return 0;
> > >  }
> > > +#endif
> >
> > We had a confusion here. The suggestion was to remove the #ifdef here
> > and add the __maybe_unused annotation to the function.
> 
> I wondered if maybe that's what you were getting at with your
> most recent comments.  But the code doesn't compile on ARM64
> with __maybe_unused instead of the #ifdef because
> VDSO_CLOCKMODE_HVCLOCK is undefined.
> 
> Michael
> 

Daniel -- any additional comments/feedback?  Getting your Ack on
this patch should be the last step to wrap up the patch series.

Thanks,

Michael

> >
> > >  static struct clocksource hyperv_cs_tsc = {
> > >  	.name	= "hyperv_clocksource_tsc_page",
> > > @@ -384,7 +386,12 @@ static int hv_cs_enable(struct clocksource *cs)
> > >  	.flags	= CLOCK_SOURCE_IS_CONTINUOUS,
> > >  	.suspend= suspend_hv_clock_tsc,
> > >  	.resume	= resume_hv_clock_tsc,
> > > +#ifdef VDSO_CLOCKMODE_HVCLOCK
> > >  	.enable = hv_cs_enable,
> > > +	.vdso_clock_mode = VDSO_CLOCKMODE_HVCLOCK,
> > > +#else
> > > +	.vdso_clock_mode = VDSO_CLOCKMODE_NONE,
> > > +#endif
> > >  };
> > >