RE: [PATCH v2 2/4] drivers/clocksource/hyper-v: Introduce TSC PFN getter

From: Michael Kelley (LINUX)
Date: Wed Nov 02 2022 - 19:39:18 EST


From: Stanislav Kinsburskii <skinsburskii@xxxxxxxxxxxxxxxxxxx> Sent: Wednesday, November 2, 2022 3:07 PM
>
> And rework the code to use it instead of the physical address, which isn't
> required by itself.
>
> This is a cleanup and precursor patch for upcoming support for TSC page
> mapping into Microsoft Hypervisor root partition, where TSC PFN will be
> defined by the hypervisor and not by the kernel.
>
> Signed-off-by: Stanislav Kinsburskiy <stanislav.kinsburskiy@xxxxxxxxx>
> CC: "K. Y. Srinivasan" <kys@xxxxxxxxxxxxx>
> CC: Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>
> CC: Wei Liu <wei.liu@xxxxxxxxxx>
> CC: Dexuan Cui <decui@xxxxxxxxxxxxx>
> CC: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>
> CC: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> CC: linux-hyperv@xxxxxxxxxxxxxxx
> CC: linux-kernel@xxxxxxxxxxxxxxx
> ---
> drivers/clocksource/hyperv_timer.c | 14 +++++++++-----
> 1 file changed, 9 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/clocksource/hyperv_timer.c b/drivers/clocksource/hyperv_timer.c
> index c4dbf40a3d3e..d447bc99a399 100644
> --- a/drivers/clocksource/hyperv_timer.c
> +++ b/drivers/clocksource/hyperv_timer.c
> @@ -367,6 +367,12 @@ static union {
> } tsc_pg __aligned(PAGE_SIZE);
>
> static struct ms_hyperv_tsc_page *tsc_page = &tsc_pg.page;
> +static unsigned long tsc_pfn;
> +
> +static unsigned long hv_get_tsc_pfn(void)
> +{
> + return tsc_pfn;
> +}
>
> struct ms_hyperv_tsc_page *hv_get_tsc_page(void)
> {
> @@ -408,13 +414,12 @@ static void suspend_hv_clock_tsc(struct clocksource *arg)
>
> static void resume_hv_clock_tsc(struct clocksource *arg)
> {
> - phys_addr_t phys_addr = virt_to_phys(tsc_page);
> union hv_reference_tsc_msr tsc_msr;
>
> /* Re-enable the TSC page */
> tsc_msr.as_uint64 = hv_get_register(HV_REGISTER_REFERENCE_TSC);
> tsc_msr.enable = 1;
> - tsc_msr.pfn = __phys_to_pfn(phys_addr);
> + tsc_msr.pfn = tsc_pfn;
> hv_set_register(HV_REGISTER_REFERENCE_TSC, tsc_msr.as_uint64);
> }
>
> @@ -498,7 +503,6 @@ static __always_inline void hv_setup_sched_clock(void
> *sched_clock) {}
> static bool __init hv_init_tsc_clocksource(void)
> {
> union hv_reference_tsc_msr tsc_msr;
> - phys_addr_t phys_addr;
>
> if (!(ms_hyperv.features & HV_MSR_REFERENCE_TSC_AVAILABLE))
> return false;
> @@ -523,7 +527,7 @@ static bool __init hv_init_tsc_clocksource(void)
> }
>
> hv_read_reference_counter = read_hv_clock_tsc;
> - phys_addr = virt_to_phys(hv_get_tsc_page());
> + tsc_pfn = __phys_to_pfn(virt_to_phys(tsc_page));

I probably wasn't specific enough when I said "your choice" in my previous
comments. This sets tsc_pfn to a guest PFN. In a guest with page size
other than 4096, it will be wrong when stored into tsc_msr.pfn a few
lines below. hyperv_timer.c is currently architecture and page size
independent, and it needs to stay that way.

> /*
> * The Hyper-V TLFS specifies to preserve the value of reserved
> @@ -534,7 +538,7 @@ static bool __init hv_init_tsc_clocksource(void)
> */
> tsc_msr.as_uint64 = hv_get_register(HV_REGISTER_REFERENCE_TSC);
> tsc_msr.enable = 1;
> - tsc_msr.pfn = __phys_to_pfn(phys_addr);
> + tsc_msr.pfn = tsc_pfn;
> hv_set_register(HV_REGISTER_REFERENCE_TSC, tsc_msr.as_uint64);
>
> clocksource_register_hz(&hyperv_cs_tsc, NSEC_PER_SEC/100);
>