On 12/21/2017 11:49 PM, Michael S. Tsirkin wrote:
On Thu, Dec 21, 2017 at 10:21:20PM +0100, Maxime Coquelin wrote:
Sorry, I missed to fixup the feature mask in patch 4,
this is the only change in this v3.
The two changes in v2 are fixing the features mask
before the function rework, and reword of commit message
of patch "vhost-user-test: extract read-guest-mem test
from main loop" (Thanks Marc-André).
This series fixes two issues in vhost-user-test:
1. Setup virtqueues in all tests
2. Fix features mask for all but test_multiqueue()
The clean-ups comprises making read-guest-mem test consistent
with other tests by initializing the device in the qtest thread.
Also, some code factorization is done with regard to device
initialization so that all tests share the same init.
Looks like mingw tests fail with this.
No, the build issue with mingw is due to a regression in master.
Laurent posted a patch to fix it:
[PATCH] hw/i386/vmport: fix missing definitions with x86_64-w64-mingw32
Message-Id: <20171221211103.30311-1-laur...@vivier.eu>
Maxime
Maxime Coquelin (5):
vhost-user-test: fix features mask
vhost-user-test: extract read-guest-mem test from main loop
vhost-user-test: setup virtqueues in all tests
vhost-user-test: make features mask an init_virtio_dev() argument
vhost-user-test: use init_virtio_dev in multiqueue test
tests/vhost-user-test.c | 171 ++++++++++++++++++++++--------------------------
1 file changed, 79 insertions(+), 92 deletions(-)
--
2.14.3
