提交 46bea48a 编写于 作者: S Shakeel Butt 提交者: Paolo Bonzini

kvm, mm: account kvm related kmem slabs to kmemcg

The kvm slabs can consume a significant amount of system memory
and indeed in our production environment we have observed that
a lot of machines are spending significant amount of memory that
can not be left as system memory overhead. Also the allocations
from these slabs can be triggered directly by user space applications
which has access to kvm and thus a buggy application can leak
such memory. So, these caches should be accounted to kmemcg.
Signed-off-by: NShakeel Butt <shakeelb@google.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
上级 736fdf72
...@@ -5475,13 +5475,13 @@ int kvm_mmu_module_init(void) ...@@ -5475,13 +5475,13 @@ int kvm_mmu_module_init(void)
pte_list_desc_cache = kmem_cache_create("pte_list_desc", pte_list_desc_cache = kmem_cache_create("pte_list_desc",
sizeof(struct pte_list_desc), sizeof(struct pte_list_desc),
0, 0, NULL); 0, SLAB_ACCOUNT, NULL);
if (!pte_list_desc_cache) if (!pte_list_desc_cache)
goto nomem; goto nomem;
mmu_page_header_cache = kmem_cache_create("kvm_mmu_page_header", mmu_page_header_cache = kmem_cache_create("kvm_mmu_page_header",
sizeof(struct kvm_mmu_page), sizeof(struct kvm_mmu_page),
0, 0, NULL); 0, SLAB_ACCOUNT, NULL);
if (!mmu_page_header_cache) if (!mmu_page_header_cache)
goto nomem; goto nomem;
......
...@@ -4010,7 +4010,7 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align, ...@@ -4010,7 +4010,7 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
if (!vcpu_align) if (!vcpu_align)
vcpu_align = __alignof__(struct kvm_vcpu); vcpu_align = __alignof__(struct kvm_vcpu);
kvm_vcpu_cache = kmem_cache_create("kvm_vcpu", vcpu_size, vcpu_align, kvm_vcpu_cache = kmem_cache_create("kvm_vcpu", vcpu_size, vcpu_align,
0, NULL); SLAB_ACCOUNT, NULL);
if (!kvm_vcpu_cache) { if (!kvm_vcpu_cache) {
r = -ENOMEM; r = -ENOMEM;
goto out_free_3; goto out_free_3;
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册