From: Jan Dakinevich <[email protected]> For L1TF mitigation vmx allocates 16 pages (4th order) and use them to move out the content of L1D cache. An attempt to use physically non-contiguous memory could break the algorithm.
https://jira.sw.ru/browse/HCI-133 Signed-off-by: Jan Dakinevich <[email protected]> Signed-off-by: Jan Dakinevich <[email protected]> (cherry-picked from a8c7017e8f9b07fec07a5a87e8d7e569c03c3752) Signed-off-by: Andrey Zhadchenko <[email protected]> --- arch/x86/kvm/vmx/vmx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 4411eaf..ff68628 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -256,7 +256,7 @@ static int vmx_setup_l1d_flush(enum vmx_l1d_flush_state l1tf) * This allocation for vmx_l1d_flush_pages is not tied to a VM * lifetime and so should not be charged to a memcg. */ - page = alloc_pages(GFP_KERNEL, L1D_CACHE_ORDER); + page = alloc_pages(GFP_KERNEL | __GFP_NOWARN, L1D_CACHE_ORDER); if (!page) return -ENOMEM; vmx_l1d_flush_pages = page_address(page); -- 1.8.3.1 _______________________________________________ Devel mailing list [email protected] https://lists.openvz.org/mailman/listinfo/devel
