Re: [PATCH v12 08/31] LoongArch: KVM: Implement vcpu handle exit interface

From: Tianrui Zhao
Date: Mon Jun 05 2023 - 22:13:38 EST




在 2023年06月05日 21:03, bibo, mao 写道:

在 2023/5/30 09:52, Tianrui Zhao 写道:
Implement vcpu handle exit interface, getting the exit code by ESTAT
register and using kvm exception vector to handle it.

Signed-off-by: Tianrui Zhao <zhaotianrui@xxxxxxxxxxx>
---
arch/loongarch/kvm/vcpu.c | 46 +++++++++++++++++++++++++++++++++++++++
1 file changed, 46 insertions(+)

diff --git a/arch/loongarch/kvm/vcpu.c b/arch/loongarch/kvm/vcpu.c
index eca8b96a3e6e..ddea480fa5b0 100644
--- a/arch/loongarch/kvm/vcpu.c
+++ b/arch/loongarch/kvm/vcpu.c
@@ -55,6 +55,52 @@ static void kvm_pre_enter_guest(struct kvm_vcpu *vcpu)
vcpu->arch.aux_inuse &= ~KVM_LARCH_CSR;
}
+/*
+ * Return 1 for resume guest and "<= 0" for resume host.
+ */
+static int _kvm_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu)
+{
+ unsigned long exst = vcpu->arch.host_estat;
+ u32 intr = exst & 0x1fff; /* ignore NMI */
+ u32 exccode = (exst & CSR_ESTAT_EXC) >> CSR_ESTAT_EXC_SHIFT;
+ int ret = RESUME_GUEST;
+
+ vcpu->mode = OUTSIDE_GUEST_MODE;
+
+ /* Set a default exit reason */
+ run->exit_reason = KVM_EXIT_UNKNOWN;
+ run->ready_for_interrupt_injection = 1;
Is ready_for_interrupt_injection used in qemu or kvm for LoongArch?
We do not handle ready_for_interrupt_injection when vcpu exit in qemu/kvm, and I will remove it.

Thanks
Tianrui Zhao

+
+ local_irq_enable();
+ guest_state_exit_irqoff();
+
+ trace_kvm_exit(vcpu, exccode);
+ if (exccode) {
+ ret = _kvm_handle_fault(vcpu, exccode);
+ } else {
+ WARN(!intr, "suspicious vm exiting");
how about comments like this?
WARN(!intr, "vm exiting with suspicious irq \n")

Regards
Bibo, Mao
Thanks, this comment should contain more information, I will fix it.

Thanks
Tianrui Zhao
+ ++vcpu->stat.int_exits;
+ }
+
+ cond_resched();
+ local_irq_disable();
+
+ if (ret == RESUME_HOST)
+ return ret;
+
+ /* Only check for signals if not already exiting to userspace */
+ if (signal_pending(current)) {
+ vcpu->run->exit_reason = KVM_EXIT_INTR;
+ ++vcpu->stat.signal_exits;
+ return -EINTR;
+ }
+
+ kvm_pre_enter_guest(vcpu);
+ trace_kvm_reenter(vcpu);
+ guest_state_enter_irqoff();
+ return RESUME_GUEST;
+}
+
int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu)
{
unsigned long timer_hz;