Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE
From: Ackerley Tng <hidden>
Date: 2023-03-08 00:13:35
Also in:
kvm, linux-arch, linux-doc, linux-fsdevel, linux-mm, lkml, qemu-devel
Possibly related (same subject, not in this thread)
- 2023-04-25 · Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE · Sean Christopherson <seanjc@google.com>
- 2023-04-18 · Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE · Ackerley Tng <hidden>
- 2023-03-07 · Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE · Sean Christopherson <seanjc@google.com>
- 2023-03-07 · Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE · Ackerley Tng <hidden>
- 2023-01-28 · Re: [PATCH v10 9/9] KVM: Enable and expose KVM_MEM_PRIVATE · Chao Peng <hidden>
Chao Peng [off-list ref] writes:
On Sat, Jan 14, 2023 at 12:01:01AM +0000, Sean Christopherson wrote:quoted
On Fri, Dec 02, 2022, Chao Peng wrote:...quoted
Strongly prefer to use similar logic to existing code that detects wraps:
quoted
mem->restricted_offset + mem->memory_size < mem->restricted_offset
quoted
This is also where I'd like to add the "gfn is aligned to offset" check, though my brain is too fried to figure that out right now.
Used count_trailing_zeros() for this TODO, unsure we have other better approach.
quoted hunk
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index afc8c26fa652..fd34c5f7cd2f 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c@@ -56,6 +56,7 @@ #include <asm/processor.h> #include <asm/ioctl.h> #include <linux/uaccess.h> +#include <linux/count_zeros.h>
quoted hunk
#include "coalesced_mmio.h" #include "async_pf.h"@@ -2087,6 +2088,19 @@ static bool kvm_check_memslot_overlap(structkvm_memslots *slots, int id, return false; }
+/*
+ * Return true when ALIGNMENT(offset) >= ALIGNMENT(gpa).
+ */
+static bool kvm_check_rmem_offset_alignment(u64 offset, u64 gpa)
+{
+ if (!offset)
+ return true;
+ if (!gpa)
+ return false;
+
+ return !!(count_trailing_zeros(offset) >= count_trailing_zeros(gpa));Perhaps we could do something like #define lowest_set_bit(val) (val & -val) and use return lowest_set_bit(offset) >= lowest_set_bit(gpa); Please help me to understand: why must ALIGNMENT(offset) >= ALIGNMENT(gpa)? Why is it not sufficient to have both gpa and offset be aligned to PAGE_SIZE?
quoted hunk
+} + /* * Allocate some memory and give it an address in the guest physical address * space.@@ -2128,7 +2142,8 @@ int __kvm_set_memory_region(struct kvm *kvm, if (mem->flags & KVM_MEM_PRIVATE && (mem->restrictedmem_offset & (PAGE_SIZE - 1) || mem->restrictedmem_offset + mem->memory_size <mem->restrictedmem_offset || - 0 /* TODO: require gfn be aligned with restricted offset */)) + !kvm_check_rmem_offset_alignment(mem->restrictedmem_offset, + mem->guest_phys_addr))) return -EINVAL; if (as_id >= kvm_arch_nr_memslot_as_ids(kvm) || id >= KVM_MEM_SLOTS_NUM) return -EINVAL;