[X][X/linux-kvm][SRU][PATCH 0/1] mm: mlock: avoid increase mm->locked_vm on mlock() when already mlock2(, MLOCK_ONFAULT)
po-hsu.lin at canonical.com
Tue Oct 23 08:44:51 UTC 2018
== Justification ==
When one vma was with flag VM_LOCKED|VM_LOCKONFAULT (by invoking
mlock2(,MLOCK_ONFAULT)), it can again be populated with mlock() with
VM_LOCKED flag only.
There is a hole in mlock_fixup() which increase mm->locked_vm twice even
the two operations are on the same vma and both with VM_LOCKED flags.
The issue can be reproduced by following code:
mlock2(p, 1024 * 64, MLOCK_ONFAULT); //VM_LOCKED|VM_LOCKONFAULT
mlock(p, 1024 * 64); //VM_LOCKED
Then check the increase VmLck field in /proc/pid/status(to 128k).
When vma is set with different vm_flags, and the new vm_flags is with
VM_LOCKED, it is not necessarily be a "new locked" vma.
There is a dedicated reproducer, the "mlock203" test in ubuntu_ltp_syscalls, you can see the failure for all the Ubuntu 4.4 kernel:
tst_test.c:1063: INFO: Timeout per run is 0h 05m 00s
mlock203.c:63: FAIL: Locking one memory again increased VmLck
== Fix ==
b155b4fd (mm: mlock: avoid increase mm->locked_vm on mlock() when already mlock2(,MLOCK_ONFAULT))
A test kernel for Xenial / Xenial-KVM could be found here:
== Regression Potential ==
Low, this patch prevents mm->locked_vm from increment just by adding an extra check to see if the old vm_flags is already VM_LOCKED.
== Test Case ==
Run the mlock203 test in ubuntu_ltp_syscalls test suite. And it will pass with the patched kernel.
Simon Guo (1):
mm: mlock: avoid increase mm->locked_vm on mlock() when already
mm/mlock.c | 3 +++
1 file changed, 3 insertions(+)
More information about the kernel-team