Re: [PATCH v2 6/7] locking/spinlocks, paravirt, hyperv: Correct the hv_nopvspin case

From: Zhenzhong Duan
Date: Thu Jun 27 2019 - 20:54:18 EST

On 2019/6/28 6:28, Sasha Levin wrote:
On Mon, Jun 24, 2019 at 08:02:58PM +0800, Zhenzhong Duan wrote:
With the boot parameter "hv_nopvspin" specified a Hyperv guest should
not make use of paravirt spinlocks, but behave as if running on bare
metal. This is not true, however, as the qspinlock code will fall back
to a test-and-set scheme when it is detecting a hypervisor.

In order to avoid this disable the virt_spin_lock_key.

Same change for XEN is already in Commit e6fd28eb3522
("locking/spinlocks, paravirt, xen: Correct the xen_nopvspin case")

Signed-off-by: Zhenzhong Duan <zhenzhong.duan@xxxxxxxxxx>
Cc: Waiman Long <longman@xxxxxxxxxx>
Cc: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Cc: "K. Y. Srinivasan" <kys@xxxxxxxxxxxxx>
Cc: Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>
Cc: Stephen Hemminger <sthemmin@xxxxxxxxxxxxx>
Cc: Sasha Levin <sashal@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: linux-hyperv@xxxxxxxxxxxxxxx
arch/x86/hyperv/hv_spinlock.c | 3 +++
1 file changed, 3 insertions(+)

diff --git a/arch/x86/hyperv/hv_spinlock.c b/arch/x86/hyperv/hv_spinlock.c
index 07f21a0..d90b4b0 100644
--- a/arch/x86/hyperv/hv_spinlock.c
+++ b/arch/x86/hyperv/hv_spinlock.c
@@ -64,6 +64,9 @@ __visible bool hv_vcpu_is_preempted(int vcpu)

void __init hv_init_spinlocks(void)
+ÂÂÂ if (unlikely(!hv_pvspin))
+ÂÂÂÂÂÂÂ static_branch_disable(&virt_spin_lock_key);

This should be combined in the conditional under it, which already
attempts to disable PV spinlocks, note how hv_pvspin is checked there.
hc_pvspin isn't the only reason we would disable PV spinlocks on hyperv.

In virt_spin_lock() there is a comment as below. The test-and-set spinlock

is an optimization to hypervisor platform when PV spinlock is unsupported.

ÂÂÂÂÂÂÂÂ * On hypervisors without PARAVIRT_SPINLOCKS support we fall
ÂÂÂÂÂÂÂÂ * back to a Test-and-Set spinlock, because fair locks have
ÂÂÂÂÂÂÂÂ * horrible lock 'holder' preemption issues.

So my understanding is:

If hv_pvspin=0 by command line, we want to behave as if running on bare metal(the fair locks path).

Though there is performance regression, but it's not that important when we use hv_pvspin=0.

If PV spinlock is disabled by other reasons, we prefer the optimization path.

Also, there's no need for the unlikely() here, it's only getting called

Ok, I'll removed it.