From: Benjamin Segall Date: Thu, 29 Apr 2021 16:22:34 +0000 (+0000) Subject: kvm: exit halt polling on need_resched() as well X-Git-Tag: v5.10.79~4418 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=ce7639252357be0e1f93a77abd6d81f39c800a91;p=platform%2Fkernel%2Flinux-rpi.git kvm: exit halt polling on need_resched() as well commit 262de4102c7bb8e59f26a967a8ffe8cce85cc537 upstream. single_task_running() is usually more general than need_resched() but CFS_BANDWIDTH throttling will use resched_task() when there is just one task to get the task to block. This was causing long-need_resched warnings and was likely allowing VMs to overrun their quota when halt polling. Signed-off-by: Ben Segall Signed-off-by: Venkatesh Srinivas Message-Id: <20210429162233.116849-1-venkateshs@chromium.org> Signed-off-by: Paolo Bonzini Cc: stable@vger.kernel.org Reviewed-by: Jim Mattson Signed-off-by: Greg Kroah-Hartman --- diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index fafb061..f446c36 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2797,7 +2797,8 @@ void kvm_vcpu_block(struct kvm_vcpu *vcpu) goto out; } poll_end = cur = ktime_get(); - } while (single_task_running() && ktime_before(cur, stop)); + } while (single_task_running() && !need_resched() && + ktime_before(cur, stop)); } prepare_to_rcuwait(&vcpu->wait);