[dpdk-dev] doc: Update doc for vhost sample

Message ID 1425286674-17495-1-git-send-email-changchun.ouyang@intel.com (mailing list archive)
State Superseded, archived
Headers

Commit Message

Ouyang Changchun March 2, 2015, 8:57 a.m. UTC
  Add some contents for vhost sample.

Signed-off-by: Changchun Ouyang <changchun.ouyang@intel.com>
---
 doc/guides/sample_app_ug/vhost.rst | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)
  

Comments

John McNamara March 27, 2015, 12:21 p.m. UTC | #1
V3: Fix for code indentation issue preventing PDF build.

Ouyang Changchun (1):
  doc: Update doc for vhost sample

 doc/guides/sample_app_ug/vhost.rst | 55 ++++++++++++++++++++++++++++++++------
 1 file changed, 47 insertions(+), 8 deletions(-)
  

Patch

diff --git a/doc/guides/sample_app_ug/vhost.rst b/doc/guides/sample_app_ug/vhost.rst
index fa53db6..756838c 100644
--- a/doc/guides/sample_app_ug/vhost.rst
+++ b/doc/guides/sample_app_ug/vhost.rst
@@ -654,6 +654,35 @@  The number of free hugepages can be checked as follows:
 
 The command above indicates how many hugepages are free to support QEMU's allocation request.
 
+**Can user space VHOST work properly with the guest with 2M sized hug pages?**
+
+Yes, it can. The guest may have 2M or 1G sized huge pages file, the user space VHOST can work properly in both cases.
+
+**Can user space VHOST work with QEMU without '-mem-prealloc' option?**
+The current implementation work properly only when the guest memory is pre-allocated, so you are required to use the
+correct QEMU version(e.g. 1.6) which supports '-mem-prealloc'; And the option '-mem-prealloc' is required to be specified
+explicitly in QEMU command line.
+
+**Can user space VHOST work with QEMU version without shared memory mapping?**
+No, it can't, shared memory mapping is mandatory for user space VHOST to work properly with guest as user space VHOST
+need access the shared memory from guest to receive and transmit packets. You are required to make sure the QEMU version
+support shared memory mapping.
+
+**When using libvirt "virsh create" the qemu-wrap.py spawns a new process to run "qemu-kvm". This impacts the behavior
+of the "virsh destroy" which kills the process running "qemu-wrap.py" without actually destroying the VM (leaves the
+"qemu-kvm" process running).**
+This patch can fix this issue: http://dpdk.org/ml/archives/dev/2014-June/003607.html
+
+**On Ubuntu environment, QEMU fail to start a new guest normally with user space VHOST due to hug pages can't be
+allocated for the new guest.**
+The solution for this issue could be adding "-boot c" into QEMU command line to make sure the huge pages are allocated
+properly and then the guest startup normally. You may use "cat /proc/meminfo" to check if there is any change in value
+of HugePages_Total and HugePages_Free after the guest startup.
+
+**Why I get this logging message: "eventfd_link: module verification failed: signature and/or required key missing - tainting kernel"?**
+You can ignore the above logging message. The message occurs due to the new module eventfd_link is not a standard
+module of Linux, but it is necessary for user space VHOST current implementation(CUSE-based) to communicate with guest.
+
 Running DPDK in the Virtual Machine
 -----------------------------------