Re: [PATCH v3 1/5] KVM: exit halt polling on need_resched() for both book3s and generic halt-polling

[David Matlack <dmatlack@google.com>]

On Mon, May 17, 2021 at 7:01 AM Wanpeng Li <kernellwp@gmail.com> wrote:
>
> From: Wanpeng Li <wanpengli@tencent.com>
>
> Inspired by commit 262de4102c7bb8 (kvm: exit halt polling on need_resched()
> as well), CFS_BANDWIDTH throttling will use resched_task() when there is just
> one task to get the task to block. It was likely allowing VMs to overrun their
> quota when halt polling. Due to PPC implements an arch specific halt polling
> logic, we should add the need_resched() checking there as well. This
> patch adds a helper function that to be shared between book3s and generic
> halt-polling loop.
>
> Cc: Ben Segall <bsegall@google.com>
> Cc: Venkatesh Srinivas <venkateshs@chromium.org>
> Cc: Jim Mattson <jmattson@google.com>
> Cc: David Matlack <dmatlack@google.com>
> Cc: Paul Mackerras <paulus@ozlabs.org>
> Cc: Suraj Jitindar Singh <sjitindarsingh@gmail.com>
> Signed-off-by: Wanpeng Li <wanpengli@tencent.com>

Reviewed-by: David Matlack <dmatlack@google.com>

> ---
> v2 -> v3:
>  * add a helper function
> v1 -> v2:
>  * update patch description
>
>  arch/powerpc/kvm/book3s_hv.c | 2 +-
>  include/linux/kvm_host.h     | 2 ++
>  virt/kvm/kvm_main.c          | 9 +++++++--
>  3 files changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
> index 28a80d240b76..360165df345b 100644
> --- a/arch/powerpc/kvm/book3s_hv.c
> +++ b/arch/powerpc/kvm/book3s_hv.c
> @@ -3936,7 +3936,7 @@ static void kvmppc_vcore_blocked(struct kvmppc_vcore *vc)
>                                 break;
>                         }
>                         cur = ktime_get();
> -               } while (single_task_running() && ktime_before(cur, stop));
> +               } while (kvm_vcpu_can_block(cur, stop));
>
>                 spin_lock(&vc->lock);
>                 vc->vcore_state = VCORE_INACTIVE;
> diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> index 2f34487e21f2..bf4fd60c4699 100644
> --- a/include/linux/kvm_host.h
> +++ b/include/linux/kvm_host.h
> @@ -1583,4 +1583,6 @@ static inline void kvm_handle_signal_exit(struct kvm_vcpu *vcpu)
>  /* Max number of entries allowed for each kvm dirty ring */
>  #define  KVM_DIRTY_RING_MAX_ENTRIES  65536
>
> +bool kvm_vcpu_can_block(ktime_t cur, ktime_t stop);
> +
>  #endif
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 6b4feb92dc79..c81080667fd1 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2945,6 +2945,12 @@ update_halt_poll_stats(struct kvm_vcpu *vcpu, u64 poll_ns, bool waited)
>                 vcpu->stat.halt_poll_success_ns += poll_ns;
>  }
>
> +
> +bool kvm_vcpu_can_block(ktime_t cur, ktime_t stop)

nit: kvm_vcpu_can_poll() would be a more accurate name for this function.

> +{
> +       return single_task_running() && !need_resched() && ktime_before(cur, stop);
> +}
> +
>  /*
>   * The vCPU has executed a HLT instruction with in-kernel mode enabled.
>   */
> @@ -2973,8 +2979,7 @@ void kvm_vcpu_block(struct kvm_vcpu *vcpu)
>                                 goto out;
>                         }
>                         poll_end = cur = ktime_get();
> -               } while (single_task_running() && !need_resched() &&
> -                        ktime_before(cur, stop));
> +               } while (kvm_vcpu_can_block(cur, stop));
>         }
>
>         prepare_to_rcuwait(&vcpu->wait);
> --
> 2.25.1
>