We're hiring!
*

NVMe: Officially faster for emulated controllers!

Helen Koike avatar

Helen Koike
June 13, 2017

Share this post:

Reading time:

The Doorbell Buffer Config command

When I last wrote about NVMe, the feature to improve NVMe performance over emulated environments was just a living discussion and a work in progress patch. However, it has now been officially released in the NVMe Specification Revision 1.3 under the name "Doorbell Buffer Config command", along with an implementation that is already in the mainline Linux kernel! \o/

You can already feel the difference in performance if you compile Kernel 4.12-rc1 (or later) and run it over a virtual machine hosted on Google Compute Engine. Google actually updated their hypervisor as soon as the feature was ratified by the NVMe working group, even before it was publicly released.

There were very few changes from the original proposal, I.e. opcodes, return values and now fancy names; the buffers (as described in my last post) are now called Shadow Doorbell and EventIdx buffers.

In short, the first one mimics the Doorbell registers in memory, allowing the emulated controller to fetch the Doorbell value when convenient instead of waiting for the Doorbell register to be written. For its part, the EventIdx provides a hint given by the emulated controller to tell the host if the Doorbell register needs to be updated (in case the emulated controller is not fetching the Doorbell value from the Shadow Doorbell buffer). You can check section 7.13 of the specification for an example of usage.

Results

The following test results were obtained in a machine of type n1-standard-4 (4 vCPUs, 15 GB memory) at Google Cloud Engine platform with Kernel 4.12.0-rc5 using the following command:

$ sudo fio --time_based --name=benchmark --runtime=30 \ --filename=/dev/nvme0n1 --nrfiles=1 --ioengine=libaio --iodepth=32 \ --direct=1 --invalidate=1 --verify=0 --verify_fatal=0 --numjobs=1 \ --rw=randread --blocksize=4k --randrepeat=0

Results (in Input/Ouput Operations per Second):
Without Shadow Doorbell and EventIdx buffers: 43.9K IOPS
With Shadow Doorbell and EventIdx buffers: 184K IOPS
Gain ~= 4 times

Screenshot - Without Shadow Doorbell and EventIdx buffers


Screenshot - With Shadow Doorbell and EventIdx buffers


Enjoy your enhanced numbers of IOPS! :D

 

Original post

Comments (0)


Add a Comment






Allowed tags: <b><i><br>Add a new comment:


Search the newsroom

Latest Blog Posts

Writing a Rust GPU kernel driver: a brief introduction on how GPU drivers work

06/08/2025

This second post in the Tyr series dives deeper into GPU driver internals by using the Vulkan-based VkCube application to explain how User…

A practical debugging guide for media driver developers

22/07/2025

Getting into kernel development can be daunting. There are layers upon layers of knowledge to master, but no clear roadmap, especially when…

Quick notes from the GStreamer Spring Hackfest 2025

15/07/2025

This past May, we met with the community at the GStreamer Spring Hackfest in Nice, France, and were able to make great strides, including…

PipeWire workshop 2025: Updates on video transport, Rust efforts, TSN networking, and Bluetooth support

03/07/2025

As part of the activities Embedded Recipes in Nice, France, Collabora hosted a PipeWire workshop/hackfest, an opportunity for attendees…

Coccinelle for Rust progress report

25/06/2025

In collaboration with Inria, the French Institute for Research in Computer Science and Automation, Tathagata Roy shares the progress made…

Linux Media Summit 2025 recap

23/06/2025

Last month in Nice, active media developers came together for the annual Linux Media Summit to exchange insights and tackle ongoing challenges…

Open Since 2005 logo

Our website only uses a strictly necessary session cookie provided by our CMS system. To find out more please follow this link.

Collabora Limited © 2005-2025. All rights reserved. Privacy Notice. Sitemap.