Re: [PATCH] x86: tsc: fix L2 guest always in tsc_early clocksource

Re: [PATCH] x86: tsc: fix L2 guest always in tsc_early clocksource

[Peter Zijlstra]

On Tue, Apr 24, 2018 at 09:08:54AM +0800, Peng Hao wrote:
> In L2 guest tsc_read_refs always return ULLONG_MAX, and that will
> call tsc_refine_calibration_work periodly. So L2 guest will read
> acpi timer port 0x608 periodly.
> The patch will let it out of "if(tsc_start == -1){}" infinite loop.

Help me out a little. What's an L2 guest?

And I'm assuming you don't have HPET and then end up wanting to use
PM-timer? Which is also busted and resturns crap results?

So if the whole TSC sync has no way of ever working; what is your patch
doing? How is the actual TSC calibrated?

Should you instead be using fast_calibrate() and passing in a known
value from the host? Thereby avoiding all this calibration nonsense
entirely.

[PATCH] x86: tsc: fix L2 guest always in tsc_early clocksource

[Peng Hao]

In L2 guest tsc_read_refs always return ULLONG_MAX, and that will
call tsc_refine_calibration_work periodly. So L2 guest will read
acpi timer port 0x608 periodly.
The patch will let it out of "if(tsc_start == -1){}" infinite loop.

Signed-off-by: Peng Hao <peng.hao2@zte.com.cn>
---
 arch/x86/kernel/tsc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index ef32297..828ef50 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -1236,7 +1236,7 @@ struct system_counterval_t convert_art_ns_to_tsc(u64 art_ns)
  */
 static void tsc_refine_calibration_work(struct work_struct *work)
 {
-	static u64 tsc_start = -1, ref_start;
+	static u64 tsc_start = 0, ref_start;
 	static int hpet;
 	u64 tsc_stop, ref_stop, delta;
 	unsigned long freq;
@@ -1251,7 +1251,7 @@ static void tsc_refine_calibration_work(struct work_struct *work)
 	 * delayed the first time we expire. So set the workqueue
 	 * again once we know timers are working.
 	 */
-	if (tsc_start == -1) {
+	if (tsc_start == 0) {
 		/*
 		 * Only set hpet once, to avoid mixing hardware
 		 * if the hpet becomes enabled later.
-- 
1.8.3.1