On 8/8/19 1:40 PM, Mina Almasry wrote:
Problem: Currently tasks attempting to allocate more hugetlb memory than is available get a failure at mmap/shmget time. This is thanks to Hugetlbfs Reservations [1]. However, if a task attempts to allocate hugetlb memory only more than its hugetlb_cgroup limit allows, the kernel will allow the mmap/shmget call, but will SIGBUS the task when it attempts to fault the memory in.
We have developers interested in using hugetlb_cgroups, and they have expressed dissatisfaction regarding this behavior. We'd like to improve this behavior such that tasks violating the hugetlb_cgroup limits get an error on mmap/shmget time, rather than getting SIGBUS'd when they try to fault the excess memory in.
The underlying problem is that today's hugetlb_cgroup accounting happens at hugetlb memory *fault* time, rather than at *reservation* time. Thus, enforcing the hugetlb_cgroup limit only happens at fault time, and the offending task gets SIGBUS'd.
Proposed Solution: A new page counter named hugetlb.xMB.reservation_[limit|usage]_in_bytes. This counter has slightly different semantics than hugetlb.xMB.[limit|usage]_in_bytes:
- While usage_in_bytes tracks all *faulted* hugetlb memory,
reservation_usage_in_bytes tracks all *reserved* hugetlb memory.
- If a task attempts to reserve more memory than limit_in_bytes allows,
the kernel will allow it to do so. But if a task attempts to reserve more memory than reservation_limit_in_bytes, the kernel will fail this reservation.
This proposal is implemented in this patch, with tests to verify functionality and show the usage.
Alternatives considered:
A new cgroup, instead of only a new page_counter attached to the existing hugetlb_cgroup. Adding a new cgroup seemed like a lot of code duplication with hugetlb_cgroup. Keeping hugetlb related page counters under hugetlb_cgroup seemed cleaner as well.
Instead of adding a new counter, we considered adding a sysctl that modifies the behavior of hugetlb.xMB.[limit|usage]_in_bytes, to do accounting at reservation time rather than fault time. Adding a new page_counter seems better as userspace could, if it wants, choose to enforce different cgroups differently: one via limit_in_bytes, and another via reservation_limit_in_bytes. This could be very useful if you're transitioning how hugetlb memory is partitioned on your system one cgroup at a time, for example. Also, someone may find usage for both limit_in_bytes and reservation_limit_in_bytes concurrently, and this approach gives them the option to do so.
Caveats:
- This support is implemented for cgroups-v1. I have not tried hugetlb_cgroups with cgroups v2, and AFAICT it's not supported yet. This is largely because we use cgroups-v1 for now. If required, I can add hugetlb_cgroup support to cgroups v2 in this patch or a follow up.
- Most complicated bit of this patch I believe is: where to store the pointer to the hugetlb_cgroup to uncharge at unreservation time? Normally the cgroup pointers hang off the struct page. But, with hugetlb_cgroup reservations, one task can reserve a specific page and another task may fault it in (I believe), so storing the pointer in struct page is not appropriate. Proposed approach here is to store the pointer in the resv_map. See patch for details.
Signed-off-by: Mina Almasry almasrymina@google.com
include/linux/hugetlb.h | 10 +- include/linux/hugetlb_cgroup.h | 19 +- mm/hugetlb.c | 256 ++++++++-- mm/hugetlb_cgroup.c | 153 +++++-
Is there a reason why all these changes are in a single patch? I can see these split in at least 2 or 3 patches with the test as a separate patch.
Makes it lot easier to review.
thanks, -- Shuah