Thread (10 messages) flat view 10 messages, 5 authors, 2021-09-28

Re: [PATCH v2 6/6] x86/kvm: add boot parameter for setting max number of vcpus per guest

From: Yao Yuan <hidden>
Date: 2021-09-06 00:45:17
Also in: kvm, lkml

On Fri, Sep 03, 2021 at 03:08:07PM +0200, Juergen Gross wrote:
quoted hunk ↗ jump to hunk
Today the maximum number of vcpus of a kvm guest is set via a #define
in a header file.

In order to support higher vcpu numbers for guests without generally
increasing the memory consumption of guests on the host especially on
very large systems add a boot parameter for specifying the number of
allowed vcpus for guests.

The default will still be the current setting of 288. The value 0 has
the special meaning to limit the number of possible vcpus to the
number of possible cpus of the host.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
 Documentation/admin-guide/kernel-parameters.txt | 7 +++++++
 arch/x86/include/asm/kvm_host.h                 | 5 ++++-
 arch/x86/kvm/x86.c                              | 9 ++++++++-
 3 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 37e194299311..b9641c9989ef 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -2435,6 +2435,13 @@
          feature (tagged TLBs) on capable Intel chips.
          Default is 1 (enabled)

+	kvm.max_vcpus=	[KVM,X86] Set the maximum allowed numbers of vcpus per
+			guest. The special value 0 sets the limit to the number
+			of physical cpus possible on the host (including not
+			yet hotplugged cpus). Higher values will result in
+			slightly higher memory consumption per guest.
+			Default: 288
+
  kvm.vcpu_id_add_bits=
          [KVM,X86] The vcpu-ids of guests are sparse, as they
          are constructed by bit-wise concatenation of the ids of
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 6c28d0800208..a4ab387b0e1c 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -38,7 +38,8 @@

 #define __KVM_HAVE_ARCH_VCPU_DEBUGFS

-#define KVM_MAX_VCPUS 288
+#define KVM_DEFAULT_MAX_VCPUS 288
+#define KVM_MAX_VCPUS max_vcpus
 #define KVM_SOFT_MAX_VCPUS 240
 #define KVM_MAX_VCPU_ID kvm_max_vcpu_id()
 /* memory slots that are not exposed to userspace */
@@ -1588,6 +1589,8 @@ extern u64  kvm_max_tsc_scaling_ratio;
 extern u64  kvm_default_tsc_scaling_ratio;
 /* bus lock detection supported? */
 extern bool kvm_has_bus_lock_exit;
+/* maximum number of vcpus per guest */
+extern unsigned int max_vcpus;
 /* maximum vcpu-id */
 unsigned int kvm_max_vcpu_id(void);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index ff142b6dd00c..49c3d91c559e 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -188,9 +188,13 @@ module_param(pi_inject_timer, bint, S_IRUGO | S_IWUSR);
 static int __read_mostly vcpu_id_add_bits = -1;
 module_param(vcpu_id_add_bits, int, S_IRUGO);

+unsigned int __read_mostly max_vcpus = KVM_DEFAULT_MAX_VCPUS;
+module_param(max_vcpus, uint, S_IRUGO);
+EXPORT_SYMBOL_GPL(max_vcpus);
+
 unsigned int kvm_max_vcpu_id(void)
 {
-	int n_bits = fls(KVM_MAX_VCPUS - 1);
+	int n_bits = fls(max_vcpus - 1);
A quesintion here: the parameter "vcpu_id_add_bits" also depends
on the "max_vcpus", we can't calculate the "vcpu_id_add_bits" from
"max_vcpus" because KVM has no topologically knowledge to determine
bits needed for each socket/core/thread level, right?
quoted hunk ↗ jump to hunk
  if (vcpu_id_add_bits < -1 || vcpu_id_add_bits > (32 - n_bits)) {
      pr_err("Invalid value of vcpu_id_add_bits=%d parameter!\n",
@@ -11033,6 +11037,9 @@ int kvm_arch_hardware_setup(void *opaque)
  if (boot_cpu_has(X86_FEATURE_XSAVES))
      rdmsrl(MSR_IA32_XSS, host_xss);

+	if (max_vcpus == 0)
+		max_vcpus = num_possible_cpus();
+
  kvm_pcpu_vcpu_mask = __alloc_percpu(KVM_VCPU_MASK_SZ,
                      sizeof(unsigned long));
  kvm_hv_vp_bitmap = __alloc_percpu(KVM_HV_VPMAP_SZ, sizeof(u64));
--
2.26.2
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help