Prev: [PATCH] kvm: don not fail when cache is sufficient
Next: [PATCH 2/2] kvm: set new alias directly when old one is not valid
From: Lai Jiangshan on 17 Apr 2010 05:30

It's just time waste if we free a object and allocate it immediately.

Signed-off-by: Lai Jiangshan <laijs(a)cn.fujitsu.com>
---
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 24cd0ee..db8d7fe 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2479,15 +2479,8 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm,
rcu_assign_pointer(kvm->arch.aliases, aliases);
synchronize_srcu_expedited(&kvm->srcu);
kvm_mmu_zap_all(kvm);
- kfree(old_aliases);
-
- r = -ENOMEM;
- aliases = kzalloc(sizeof(struct kvm_mem_aliases), GFP_KERNEL);
- if (!aliases)
- goto out_unlock;
-
- memcpy(aliases, kvm->arch.aliases, sizeof(struct kvm_mem_aliases));

+ aliases = old_aliases;
p = &aliases->aliases[alias->slot];
p->base_gfn = alias->guest_phys_addr >> PAGE_SHIFT;
p->npages = alias->memory_size >> PAGE_SHIFT;
@@ -2505,7 +2498,6 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm,
kfree(old_aliases);
r = 0;

-out_unlock:
mutex_unlock(&kvm->slots_lock);
out:
return r;

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: [PATCH] kvm: don not fail when cache is sufficient
Next: [PATCH 2/2] kvm: set new alias directly when old one is not valid