On Sat, Apr 18, 2020 at 23:57:54 +0800, Han Han wrote:
Signed-off-by: Han Han <hhan(a)redhat.com>
---
docs/formatdomain.html.in | 3 ++-
docs/news.xml | 9 +++++++++
docs/schemas/domaincommon.rng | 1 +
3 files changed, 12 insertions(+), 1 deletion(-)
diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in
index 00776668..2b279992 100644
--- a/docs/formatdomain.html.in
+++ b/docs/formatdomain.html.in
@@ -3747,7 +3747,8 @@
<li>
The optional <code>io</code> attribute controls specific
policies on I/O; qemu guests support "threads" and
- "native". <span class="since">Since
0.8.8</span>
+ "native"<span class="since">Since
0.8.8</span>, io_uring
Missing space after "native".
+ <span class="since">Since 6.3.0(QEMU
5.0.0)</span>.
Since 6.3.0 (QEMU 5.0)
</li>
<li>
The optional <code>ioeventfd</code> attribute allows users to
diff --git a/docs/news.xml b/docs/news.xml
index 4d0efd42..119f2cda 100644
--- a/docs/news.xml
+++ b/docs/news.xml
news.xml changes must be in a separate commit.
@@ -56,6 +56,15 @@
hotplug/unplug attempts, but this is often undesireable).
</description>
</change>
+ <change>
+ <summary>
+ qemu: support io mode io_uring to libvirt
... async IO mode 'io_uring' in libvirt
+ </summary>
+ <description>
+ From Linux 5.1 and QEMU 5.0.0, the io_uring aio interface was
+ introduced. Implement that to libvirt with the io value 'io_uring'.
I'd expect some description of what the new backend provides:
Add support for the 'io_uring' mode for asynchronous I/O introduced in
QEMU 5.0 which is designed to improve performance. (Requires at least
Linux 5.1).
Or something along that.
+ </description>
+ </change>
</section>
<section title="Improvements">
</section>