Skip to contentSkip to conversation

KVM: Fix dirty ring mmap incorrect size due to renaming accident

PATCH In review Peter Xu 2 participants 2 messages Jun 9, 2021

lore
Share & tools
lore

Peter Xu

author
#0 lore raw

PATCH 1/1 KVM: Fix dirty ring mmap incorrect size due to renaming accident

Found this when I wanted to try the per-vcpu dirty rate series out, then I found that it's not really working and it can quickly hang death a guest. I found strange errors (e.g. guest crash after migration) happens even without the per-vcpu dirty rate series.

When merging dirty ring, probably no one notice that the trivial renaming diff
[1] missed two existing references of kvm_dirty_ring_sizes; they do matter since otherwise we'll mmap() a shorter range of memory after the renaming.

I think it didn't SIGBUS for me easily simply because some other stuff within qemu mmap()ed right after the dirty rings (e.g. when testing 4096 slots, it aligned with one small page on x86), so when we access the rings we've been reading/writting to random memory elsewhere of qemu.

Fix the two sizes when map/unmap the shared dirty gfn memory.

[1] https://lore.kernel.org/qemu-devel/[email protected]/

Cc Hyman Huang <[email protected]>
Cc Paolo Bonzini <[email protected]>
Cc Dr. David Alan Gilbert <[email protected]>
---
 accel/kvm/kvm-all.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
accel/kvm/kvm-all.c +2 -2
index c7ec5388500..e5b10dd129c 100644
@@ -411,7 +411,7 @@ static int do_kvm_destroy_vcpu(CPUState *cpu)
     }
 
     if (cpu->kvm_dirty_gfns) {
-        ret = munmap(cpu->kvm_dirty_gfns, s->kvm_dirty_ring_size);
+        ret = munmap(cpu->kvm_dirty_gfns, s->kvm_dirty_ring_bytes);
         if (ret < 0) {
             goto err;
         }
@@ -495,7 +495,7 @@ int kvm_init_vcpu(CPUState *cpu, Error **errp)
 
     if (s->kvm_dirty_ring_size) {
         /* Use MAP_SHARED to share pages with the kernel */
-        cpu->kvm_dirty_gfns = mmap(NULL, s->kvm_dirty_ring_size,
+        cpu->kvm_dirty_gfns = mmap(NULL, s->kvm_dirty_ring_bytes,
                                    PROT_READ | PROT_WRITE, MAP_SHARED,
                                    cpu->kvm_fd,
                                    PAGE_SIZE * KVM_DIRTY_LOG_PAGE_OFFSET);
Signature · 1 line
2.31.1
#1 lore raw
Quoted 41 lines · Peter Xu

Found this when I wanted to try the per-vcpu dirty rate series out, then I found that it's not really working and it can quickly hang death a guest. I found strange errors (e.g. guest crash after migration) happens even without the per-vcpu dirty rate series.

When merging dirty ring, probably no one notice that the trivial renaming diff
[1] missed two existing references of kvm_dirty_ring_sizes; they do matter since otherwise we'll mmap() a shorter range of memory after the renaming.

I think it didn't SIGBUS for me easily simply because some other stuff within qemu mmap()ed right after the dirty rings (e.g. when testing 4096 slots, it aligned with one small page on x86), so when we access the rings we've been reading/writting to random memory elsewhere of qemu.

Fix the two sizes when map/unmap the shared dirty gfn memory.

[1] https://lore.kernel.org/qemu-devel/[email protected]/

Cc Hyman Huang <[email protected]>
Cc Paolo Bonzini <[email protected]>
Cc Dr. David Alan Gilbert <[email protected]>
---
  accel/kvm/kvm-all.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)
accel/kvm/kvm-all.c +1 -1
index c7ec5388500..e5b10dd129c 100644
@@ -411,7 +411,7 @@ static int do_kvm_destroy_vcpu(CPUState *cpu)
      }
 
      if (cpu->kvm_dirty_gfns) {
-        ret = munmap(cpu->kvm_dirty_gfns, s->kvm_dirty_ring_size);
+        ret = munmap(cpu->kvm_dirty_gfns, s->kvm_dirty_ring_bytes);
          if (ret < 0) {
              goto err;
          }

accel/kvm/kvm-all.c int kvm_init_vcpu(CPUState *cpu, Error **errp) · line 501

Quoted 8 lines
accel/kvm/kvm-all.c +1 -1
@@ -495,7 +495,7 @@ int kvm_init_vcpu(CPUState *cpu, Error **errp)
 
      if (s->kvm_dirty_ring_size) {
          /* Use MAP_SHARED to share pages with the kernel */
-        cpu->kvm_dirty_gfns = mmap(NULL, s->kvm_dirty_ring_size,
+        cpu->kvm_dirty_gfns = mmap(NULL, s->kvm_dirty_ring_bytes,
                                     PROT_READ | PROT_WRITE, MAP_SHARED,
                                     cpu->kvm_fd,
                                     PAGE_SIZE * KVM_DIRTY_LOG_PAGE_OFFSET);

Oops, sorry. I queued it.

Paolo