KVM: selftests: Use EPOLL in userfaultfd_util reader threads

Browse source

With multiple reader threads POLLing a single UFFD, the demand paging test
suffers from the thundering herd problem: performance degrades as the
number of reader threads is increased. Solve this issue [1] by switching
the the polling mechanism to EPOLL + EPOLLEXCLUSIVE.

Also, change the error-handling convention of uffd_handler_thread_fn.
Instead of just printing errors and returning early from the polling
loop, check for them via TEST_ASSERT(). "return NULL" is reserved for a
successful exit from uffd_handler_thread_fn, i.e. one triggered by a
write to the exit pipe.

Performance samples generated by the command in [2] are given below.

Num Reader Threads, Paging Rate (POLL), Paging Rate (EPOLL)
1      249k      185k
2      201k      235k
4      186k      155k
16     150k      217k
32     89k       198k

[1] Single-vCPU performance does suffer somewhat.
[2] ./demand_paging_test -u MINOR -s shmem -v 4 -o -r <num readers>

Signed-off-by: Anish Moorthy <amoorthy@google.com>
Acked-by: James Houghton <jthoughton@google.com>
Link: https://lore.kernel.org/r/20240215235405.368539-13-amoorthy@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

...

This commit is contained in:

Anish Moorthy 2024-02-15 23:54:03 +00:00 • committed by Sean Christopherson

parent df4ec5aada

commit 0cba6442e9

2 changed files with 35 additions and 40 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

1

tools/testing/selftests/kvm/demand_paging_test.c

View file

74

tools/testing/selftests/kvm/lib/userfaultfd_util.c

View file