Re: [PATCH RFC v1 07/10] target/loongarch: Implement kvm_arch_handle_exit

[Tianrui Zhao]

在 2023年04月20日 18:00, Philippe Mathieu-Daudé 写道:

On 20/4/23 11:36, Tianrui Zhao wrote:

Implement kvm_arch_handle_exit for loongarch. In this
function, the KVM_EXIT_LOONGARCH_IOCSR is handled,
we read or write the iocsr address space by the addr,
length and is_write argument in kvm_run.

Signed-off-by: Tianrui Zhao <zhaotian...@loongson.cn>
---
  target/loongarch/kvm.c | 24 +++++++++++++++++++++++-
  1 file changed, 23 insertions(+), 1 deletion(-)

diff --git a/target/loongarch/kvm.c b/target/loongarch/kvm.c
index f8772bbb27..4ce343d276 100644
--- a/target/loongarch/kvm.c
+++ b/target/loongarch/kvm.c
@@ -499,7 +499,29 @@ bool kvm_arch_cpu_check_are_resettable(void)
    int kvm_arch_handle_exit(CPUState *cs, struct kvm_run *run)
  {
-    return 0;
+    int ret = 0;
+    LoongArchCPU *cpu = LOONGARCH_CPU(cs);
+    CPULoongArchState *env = &cpu->env;
+    MemTxAttrs attrs = {};
+
+    attrs.requester_id = env_cpu(env)->cpu_index;
+
+    DPRINTF("%s\n", __func__);

Please use trace events instead of DPRINTF(), as we are trying to remove
these.

Thanks, I will replace it with trace events.

Thanks
Tianrui Zhao

+    switch (run->exit_reason) {
+    case KVM_EXIT_LOONGARCH_IOCSR:
+        address_space_rw(&env->address_space_iocsr,
+                         run->iocsr_io.phys_addr,
+                         attrs,
+                         run->iocsr_io.data,
+                         run->iocsr_io.len,
+                         run->iocsr_io.is_write);
+        break;
+    default:
+        ret = -1;

+ fprintf(stderr, "KVM: unknown exit reason %d\n", run->exit_reason);

Would warn_report() be more appropriate here?

Thanks, I will use warn_report() here.

Thanks
Tianrui Zhao

+        break;
+    }
+    return ret;
  }
    void kvm_arch_accel_class_init(ObjectClass *oc)