Skip to contentSkip to conversation

KVM: Dynamically allocate "new" memslots from the get-go

bug report Dan Carpenter 2 participants 3 messages Dec 15, 2021 KVM

lore
Share & tools
lore
#0 lore raw

Hello Sean Christopherson,

This is a semi-automatic email about new static checker warnings.

The patch 244893fa2859: "KVM: Dynamically allocate "new" memslots from the get-go" from Dec 6, 2021, leads to the following Smatch complaint:

arch/x86/kvm/../../../virt/kvm/kvm_main.c:1526 kvm_prepare_memory_region()
warn: variable dereferenced before check 'new' (see line 1509)

arch/x86/kvm/../../../virt/kvm/kvm_main.c
1508 if (change != KVM_MR_DELETE) {
1509 if (!(new->flags & KVM_MEM_LOG_DIRTY_PAGES))
1510 new->dirty_bitmap = NULL;
1511 else if (old && old->dirty_bitmap)
1512 new->dirty_bitmap = old->dirty_bitmap;
1513 else if (!kvm->dirty_ring_size) {
1514 r = kvm_alloc_dirty_bitmap(new);
1515 if (r)
1516 return r;
1517
1518 if (kvm_dirty_log_manual_protect_and_init_set(kvm))

  1519					bitmap_set(new->dirty_bitmap, 0, new->npages);
                                                   ^^^^^
  1520			}
  1521		}
  1522	
  1523		r = kvm_arch_prepare_memory_region(kvm, old, new, change);
                                                             ^^^

Lots of unchecked dereferences

1524
1525 /* Free the bitmap on failure if it was allocated above. */
1526 if (r && new && new->dirty_bitmap && old && !old->dirty_bitmap)

                         ^^^

New check for NULL. Can this be NULL?

  1527			kvm_destroy_dirty_bitmap(new);
  1528	

regards,
dan carpenter

#1 lore raw
Quoted 3 lines · Dan Carpenter

Hello Sean Christopherson,

This is a semi-automatic email about new static checker warnings.

These are all ok, KVM is being clever and using implicit checks on pointers in select flows, while using explicit checks in others.

Quoted 6 lines

The patch 244893fa2859: "KVM: Dynamically allocate "new" memslots from the get-go" from Dec 6, 2021, leads to the following Smatch complaint:

arch/x86/kvm/../../../virt/kvm/kvm_main.c:1526 kvm_prepare_memory_region()
warn: variable dereferenced before check 'new' (see line 1509)

@new is guaranteed to be non-NULL in the !KVM_MR_DELETE case.

Quoted 12 lines

arch/x86/kvm/../../../virt/kvm/kvm_main.c
1508 if (change != KVM_MR_DELETE) {
1509 if (!(new->flags & KVM_MEM_LOG_DIRTY_PAGES))
1510 new->dirty_bitmap = NULL;
1511 else if (old && old->dirty_bitmap)
1512 new->dirty_bitmap = old->dirty_bitmap;
1513 else if (!kvm->dirty_ring_size) {
1514 r = kvm_alloc_dirty_bitmap(new);
1515 if (r)
1516 return r;
1517
1518 if (kvm_dirty_log_manual_protect_and_init_set(kvm))

Quoted 8 lines
  1519					bitmap_set(new->dirty_bitmap, 0, new->npages);
                                                   ^^^^^
  1520			}
  1521		}
  1522	
  1523		r = kvm_arch_prepare_memory_region(kvm, old, new, change);
                                                             ^^^

Lots of unchecked dereferences

There's no true dereference, architectures are responsible for ensuring @old and @new are non-NULL, either via explicit checks on the pointer or implicit checks on @change.

Quoted 5 lines

1524
1525 /* Free the bitmap on failure if it was allocated above. */
1526 if (r && new && new->dirty_bitmap && old && !old->dirty_bitmap)

                         ^^^

New check for NULL. Can this be NULL?

Yes, when change == KVM_MR_DELETE.

Quoted 5 lines
  1527			kvm_destroy_dirty_bitmap(new);
  1528	

regards,
dan carpenter

#2 lore raw
Quoted 3 lines · Sean Christopherson

There's no true dereference, architectures are responsible for ensuring @old and @new are non-NULL, either via explicit checks on the pointer or implicit checks on @change.

Ah... Okay. Thanks.

Smatch tries to track these implicit checks within a function but the relationships get lost at the function boundaries. So, for example, if there were three callers and two were non-NULL and one was NULL then Smatch could figure it out. But in this case there is only one caller so the data gets squished together and the relationship between "change" and "new" gets lost.

regards,
dan carpenter