Messages in this thread | | | From | Cong Wang <> | Subject | [PATCH 0/3] iommu: reduce spinlock contention on fast path | Date | Wed, 20 Nov 2019 16:13:45 -0800 |
| |
This patchset contains three small optimizations for the global spinlock contention in IOVA cache. Our memcache perf test shows this reduced its p999 latency down by 45% on AMD when IOMMU is enabled.
Cong Wang (3): iommu: match the original algorithm iommu: optimize iova_magazine_free_pfns() iommu: avoid taking iova_rbtree_lock twice
--- drivers/iommu/iova.c | 44 +++++++++++++++++++++++++------------------- 1 file changed, 25 insertions(+), 19 deletions(-)
-- 2.21.0
| |