diff mbox series

[PULL,v2,37/61] accel/kvm: Extract common KVM vCPU {creation,parking} code

Message ID 08c328682231b64878fc052a11091bea39577a6f.1721731723.git.mst@redhat.com
State New
Headers show
Series [PULL,v2,01/61] hw/virtio/virtio-crypto: Fix op_code assignment in virtio_crypto_create_asym_session | expand

Commit Message

Michael S. Tsirkin July 23, 2024, 10:58 a.m. UTC 
From: Salil Mehta <salil.mehta@huawei.com>

KVM vCPU creation is done once during the vCPU realization when Qemu vCPU thread
is spawned. This is common to all the architectures as of now.

Hot-unplug of vCPU results in destruction of the vCPU object in QOM but the
corresponding KVM vCPU object in the Host KVM is not destroyed as KVM doesn't
support vCPU removal. Therefore, its representative KVM vCPU object/context in
Qemu is parked.

Refactor architecture common logic so that some APIs could be reused by vCPU
Hotplug code of some architectures likes ARM, Loongson etc. Update new/old APIs
with trace events. New APIs qemu_{create,park,unpark}_vcpu() can be externally
called. No functional change is intended here.

Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Tested-by: Vishnu Pajjuri <vishnu@os.amperecomputing.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Xianglai Li <lixianglai@loongson.cn>
Tested-by: Miguel Luis <miguel.luis@oracle.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Vishnu Pajjuri <vishnu@os.amperecomputing.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Tested-by: Zhao Liu <zhao1.liu@intel.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20240716111502.202344-2-salil.mehta@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 accel/kvm/kvm-cpus.h   |  1 -
 include/sysemu/kvm.h   | 25 +++++++++++
 accel/kvm/kvm-all.c    | 95 ++++++++++++++++++++++++++++--------------
 accel/kvm/trace-events |  5 ++-
 4 files changed, 92 insertions(+), 34 deletions(-)

Comments

Peter Maydell July 25, 2024, 10:35 a.m. UTC | #1 
On Tue, 23 Jul 2024 at 11:58, Michael S. Tsirkin <mst@redhat.com> wrote:
>
> From: Salil Mehta <salil.mehta@huawei.com>
>
> KVM vCPU creation is done once during the vCPU realization when Qemu vCPU thread
> is spawned. This is common to all the architectures as of now.
>
> Hot-unplug of vCPU results in destruction of the vCPU object in QOM but the
> corresponding KVM vCPU object in the Host KVM is not destroyed as KVM doesn't
> support vCPU removal. Therefore, its representative KVM vCPU object/context in
> Qemu is parked.
>
> Refactor architecture common logic so that some APIs could be reused by vCPU
> Hotplug code of some architectures likes ARM, Loongson etc. Update new/old APIs
> with trace events. New APIs qemu_{create,park,unpark}_vcpu() can be externally
> called. No functional change is intended here.

Hi; Coverity points out an issue with this code (CID 1558552):

> +int kvm_unpark_vcpu(KVMState *s, unsigned long vcpu_id)
> +{
> +    struct KVMParkedVcpu *cpu;
> +    int kvm_fd = -ENOENT;
> +
> +    QLIST_FOREACH(cpu, &s->kvm_parked_vcpus, node) {
> +        if (cpu->vcpu_id == vcpu_id) {
> +            QLIST_REMOVE(cpu, node);
> +            kvm_fd = cpu->kvm_fd;
> +            g_free(cpu);
> +        }
> +    }

If you are going to remove an entry from a list as you
iterate over it, you can't use QLIST_FOREACH(), because
QLIST_FOREACH will look at the next pointer of the
iteration variable at the end of the loop when it
wants to advance to the next node. In this case we've
already freed 'cpu', so it would be reading freed memory.

Should we break out of the loop when we find the entry?

If we do need to continue iteration after removing the
list node, you need to use QLIST_FOREACH_SAFE() to do
the list iteration.

> -static int kvm_get_vcpu(KVMState *s, unsigned long vcpu_id)
> -{
> -    struct KVMParkedVcpu *cpu;
> -
> -    QLIST_FOREACH(cpu, &s->kvm_parked_vcpus, node) {
> -        if (cpu->vcpu_id == vcpu_id) {
> -            int kvm_fd;
> -
> -            QLIST_REMOVE(cpu, node);
> -            kvm_fd = cpu->kvm_fd;
> -            g_free(cpu);
> -            return kvm_fd;

In this old piece of code we were OK using QLIST_FOREACH
because we returned immediately we took the node off
the list and didn't continue the iteration.

> -        }
> -    }
> -
> -    return kvm_vm_ioctl(s, KVM_CREATE_VCPU, (void *)vcpu_id);
> -}

thanks
-- PMM
Peter Maydell July 25, 2024, 12:27 p.m. UTC | #2 
On Thu, 25 Jul 2024 at 13:05, Salil Mehta <salil.mehta@huawei.com> wrote:
>
> HI Peter,
>
> >  From: Peter Maydell <peter.maydell@linaro.org>
> >  Sent: Thursday, July 25, 2024 11:36 AM
> >  To: Michael S. Tsirkin <mst@redhat.com>
> >
> >  On Tue, 23 Jul 2024 at 11:58, Michael S. Tsirkin <mst@redhat.com> wrote:
> >  >
> >  > From: Salil Mehta <salil.mehta@huawei.com>
> >  >
> >  > KVM vCPU creation is done once during the vCPU realization when Qemu
> >  > vCPU thread is spawned. This is common to all the architectures as of now.
> >  >
> >  > Hot-unplug of vCPU results in destruction of the vCPU object in QOM
> >  > but the corresponding KVM vCPU object in the Host KVM is not destroyed
> >  > as KVM doesn't support vCPU removal. Therefore, its representative KVM
> >  > vCPU object/context in Qemu is parked.
> >  >
> >  > Refactor architecture common logic so that some APIs could be reused
> >  > by vCPU Hotplug code of some architectures likes ARM, Loongson etc.
> >  > Update new/old APIs with trace events. New APIs
> >  > qemu_{create,park,unpark}_vcpu() can be externally called. No functional
> >  change is intended here.
> >
> >  Hi; Coverity points out an issue with this code (CID 1558552):
> >
> >  > +int kvm_unpark_vcpu(KVMState *s, unsigned long vcpu_id) {
> >  > +    struct KVMParkedVcpu *cpu;
> >  > +    int kvm_fd = -ENOENT;
> >  > +
> >  > +    QLIST_FOREACH(cpu, &s->kvm_parked_vcpus, node) {
> >  > +        if (cpu->vcpu_id == vcpu_id) {
> >  > +            QLIST_REMOVE(cpu, node);
> >  > +            kvm_fd = cpu->kvm_fd;
> >  > +            g_free(cpu);
> >  > +        }
> >  > +    }
> >
> >  If you are going to remove an entry from a list as you iterate over it, you
> >  can't use QLIST_FOREACH(), because QLIST_FOREACH will look at the next
> >  pointer of the iteration variable at the end of the loop when it wants to
> >  advance to the next node. In this case we've already freed 'cpu', so it would
> >  be reading freed memory.
> >
> >  Should we break out of the loop when we find the entry?
>
>
> Thanks for identifying this. Yes, a  break is missing. Should I send a fix for this
> now or you can incorporate it?

The code is already in upstream git, so please send a patch
to fix the bug.

thanks
-- PMM
diff mbox series

Patch

diff --git a/accel/kvm/kvm-cpus.h b/accel/kvm/kvm-cpus.h
index ca40add32c..171b22fd29 100644
--- a/accel/kvm/kvm-cpus.h
+++ b/accel/kvm/kvm-cpus.h
@@ -22,5 +22,4 @@  bool kvm_supports_guest_debug(void);
 int kvm_insert_breakpoint(CPUState *cpu, int type, vaddr addr, vaddr len);
 int kvm_remove_breakpoint(CPUState *cpu, int type, vaddr addr, vaddr len);
 void kvm_remove_all_breakpoints(CPUState *cpu);
-
 #endif /* KVM_CPUS_H */
diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h
index c31d9c7356..c4a914b3d8 100644
--- a/include/sysemu/kvm.h
+++ b/include/sysemu/kvm.h
@@ -313,6 +313,31 @@  int kvm_create_device(KVMState *s, uint64_t type, bool test);
  */
 bool kvm_device_supported(int vmfd, uint64_t type);
 
+/**
+ * kvm_create_vcpu - Gets a parked KVM vCPU or creates a KVM vCPU
+ * @cpu: QOM CPUState object for which KVM vCPU has to be fetched/created.
+ *
+ * @returns: 0 when success, errno (<0) when failed.
+ */
+int kvm_create_vcpu(CPUState *cpu);
+
+/**
+ * kvm_park_vcpu - Park QEMU KVM vCPU context
+ * @cpu: QOM CPUState object for which QEMU KVM vCPU context has to be parked.
+ *
+ * @returns: none
+ */
+void kvm_park_vcpu(CPUState *cpu);
+
+/**
+ * kvm_unpark_vcpu - unpark QEMU KVM vCPU context
+ * @s: KVM State
+ * @vcpu_id: Architecture vCPU ID of the parked vCPU
+ *
+ * @returns: KVM fd
+ */
+int kvm_unpark_vcpu(KVMState *s, unsigned long vcpu_id);
+
 /* Arch specific hooks */
 
 extern const KVMCapabilityInfo kvm_arch_required_capabilities[];
diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
index 64bf47a033..0f110cce3e 100644
--- a/accel/kvm/kvm-all.c
+++ b/accel/kvm/kvm-all.c
@@ -340,14 +340,71 @@  err:
     return ret;
 }
 
+void kvm_park_vcpu(CPUState *cpu)
+{
+    struct KVMParkedVcpu *vcpu;
+
+    trace_kvm_park_vcpu(cpu->cpu_index, kvm_arch_vcpu_id(cpu));
+
+    vcpu = g_malloc0(sizeof(*vcpu));
+    vcpu->vcpu_id = kvm_arch_vcpu_id(cpu);
+    vcpu->kvm_fd = cpu->kvm_fd;
+    QLIST_INSERT_HEAD(&kvm_state->kvm_parked_vcpus, vcpu, node);
+}
+
+int kvm_unpark_vcpu(KVMState *s, unsigned long vcpu_id)
+{
+    struct KVMParkedVcpu *cpu;
+    int kvm_fd = -ENOENT;
+
+    QLIST_FOREACH(cpu, &s->kvm_parked_vcpus, node) {
+        if (cpu->vcpu_id == vcpu_id) {
+            QLIST_REMOVE(cpu, node);
+            kvm_fd = cpu->kvm_fd;
+            g_free(cpu);
+        }
+    }
+
+    trace_kvm_unpark_vcpu(vcpu_id, kvm_fd > 0 ? "unparked" : "!found parked");
+
+    return kvm_fd;
+}
+
+int kvm_create_vcpu(CPUState *cpu)
+{
+    unsigned long vcpu_id = kvm_arch_vcpu_id(cpu);
+    KVMState *s = kvm_state;
+    int kvm_fd;
+
+    /* check if the KVM vCPU already exist but is parked */
+    kvm_fd = kvm_unpark_vcpu(s, vcpu_id);
+    if (kvm_fd < 0) {
+        /* vCPU not parked: create a new KVM vCPU */
+        kvm_fd = kvm_vm_ioctl(s, KVM_CREATE_VCPU, vcpu_id);
+        if (kvm_fd < 0) {
+            error_report("KVM_CREATE_VCPU IOCTL failed for vCPU %lu", vcpu_id);
+            return kvm_fd;
+        }
+    }
+
+    cpu->kvm_fd = kvm_fd;
+    cpu->kvm_state = s;
+    cpu->vcpu_dirty = true;
+    cpu->dirty_pages = 0;
+    cpu->throttle_us_per_full = 0;
+
+    trace_kvm_create_vcpu(cpu->cpu_index, vcpu_id, kvm_fd);
+
+    return 0;
+}
+
 static int do_kvm_destroy_vcpu(CPUState *cpu)
 {
     KVMState *s = kvm_state;
     long mmap_size;
-    struct KVMParkedVcpu *vcpu = NULL;
     int ret = 0;
 
-    trace_kvm_destroy_vcpu();
+    trace_kvm_destroy_vcpu(cpu->cpu_index, kvm_arch_vcpu_id(cpu));
 
     ret = kvm_arch_destroy_vcpu(cpu);
     if (ret < 0) {
@@ -373,10 +430,7 @@  static int do_kvm_destroy_vcpu(CPUState *cpu)
         }
     }
 
-    vcpu = g_malloc0(sizeof(*vcpu));
-    vcpu->vcpu_id = kvm_arch_vcpu_id(cpu);
-    vcpu->kvm_fd = cpu->kvm_fd;
-    QLIST_INSERT_HEAD(&kvm_state->kvm_parked_vcpus, vcpu, node);
+    kvm_park_vcpu(cpu);
 err:
     return ret;
 }
@@ -389,24 +443,6 @@  void kvm_destroy_vcpu(CPUState *cpu)
     }
 }
 
-static int kvm_get_vcpu(KVMState *s, unsigned long vcpu_id)
-{
-    struct KVMParkedVcpu *cpu;
-
-    QLIST_FOREACH(cpu, &s->kvm_parked_vcpus, node) {
-        if (cpu->vcpu_id == vcpu_id) {
-            int kvm_fd;
-
-            QLIST_REMOVE(cpu, node);
-            kvm_fd = cpu->kvm_fd;
-            g_free(cpu);
-            return kvm_fd;
-        }
-    }
-
-    return kvm_vm_ioctl(s, KVM_CREATE_VCPU, (void *)vcpu_id);
-}
-
 int kvm_init_vcpu(CPUState *cpu, Error **errp)
 {
     KVMState *s = kvm_state;
@@ -415,19 +451,14 @@  int kvm_init_vcpu(CPUState *cpu, Error **errp)
 
     trace_kvm_init_vcpu(cpu->cpu_index, kvm_arch_vcpu_id(cpu));
 
-    ret = kvm_get_vcpu(s, kvm_arch_vcpu_id(cpu));
+    ret = kvm_create_vcpu(cpu);
     if (ret < 0) {
-        error_setg_errno(errp, -ret, "kvm_init_vcpu: kvm_get_vcpu failed (%lu)",
+        error_setg_errno(errp, -ret,
+                         "kvm_init_vcpu: kvm_create_vcpu failed (%lu)",
                          kvm_arch_vcpu_id(cpu));
         goto err;
     }
 
-    cpu->kvm_fd = ret;
-    cpu->kvm_state = s;
-    cpu->vcpu_dirty = true;
-    cpu->dirty_pages = 0;
-    cpu->throttle_us_per_full = 0;
-
     mmap_size = kvm_ioctl(s, KVM_GET_VCPU_MMAP_SIZE, 0);
     if (mmap_size < 0) {
         ret = mmap_size;
diff --git a/accel/kvm/trace-events b/accel/kvm/trace-events
index 681ccb667d..37626c1ac5 100644
--- a/accel/kvm/trace-events
+++ b/accel/kvm/trace-events
@@ -9,6 +9,10 @@  kvm_device_ioctl(int fd, int type, void *arg) "dev fd %d, type 0x%x, arg %p"
 kvm_failed_reg_get(uint64_t id, const char *msg) "Warning: Unable to retrieve ONEREG %" PRIu64 " from KVM: %s"
 kvm_failed_reg_set(uint64_t id, const char *msg) "Warning: Unable to set ONEREG %" PRIu64 " to KVM: %s"
 kvm_init_vcpu(int cpu_index, unsigned long arch_cpu_id) "index: %d id: %lu"
+kvm_create_vcpu(int cpu_index, unsigned long arch_cpu_id, int kvm_fd) "index: %d, id: %lu, kvm fd: %d"
+kvm_destroy_vcpu(int cpu_index, unsigned long arch_cpu_id) "index: %d id: %lu"
+kvm_park_vcpu(int cpu_index, unsigned long arch_cpu_id) "index: %d id: %lu"
+kvm_unpark_vcpu(unsigned long arch_cpu_id, const char *msg) "id: %lu %s"
 kvm_irqchip_commit_routes(void) ""
 kvm_irqchip_add_msi_route(char *name, int vector, int virq) "dev %s vector %d virq %d"
 kvm_irqchip_update_msi_route(int virq) "Updating MSI route virq=%d"
@@ -25,7 +29,6 @@  kvm_dirty_ring_reaper(const char *s) "%s"
 kvm_dirty_ring_reap(uint64_t count, int64_t t) "reaped %"PRIu64" pages (took %"PRIi64" us)"
 kvm_dirty_ring_reaper_kick(const char *reason) "%s"
 kvm_dirty_ring_flush(int finished) "%d"
-kvm_destroy_vcpu(void) ""
 kvm_failed_get_vcpu_mmap_size(void) ""
 kvm_cpu_exec(void) ""
 kvm_interrupt_exit_request(void) ""