Hi Chenbo,
On 4/25/23 08:19, Xia, Chenbo wrote:
Hi Maxime,
-----Original Message-----
From: Maxime Coquelin <maxime.coque...@redhat.com>
Sent: Friday, March 31, 2023 11:43 PM
To: dev@dpdk.org; david.march...@redhat.com; Xia, Chenbo
<chenbo....@intel.com>; m...@redhat.com; f...@redhat.com;
jasow...@redhat.com; Liang, Cunming <cunming.li...@intel.com>; Xie, Yongji
<xieyon...@bytedance.com>; echau...@redhat.com; epere...@redhat.com;
amore...@redhat.com
Cc: Maxime Coquelin <maxime.coque...@redhat.com>
Subject: [RFC 07/27] vhost: change to single IOTLB cache per device
This patch simplifies IOTLB implementation and improves
IOTLB memory consumption by having a single IOTLB cache
per device, instead of having one per queue.
In order to not impact performance, it keeps an IOTLB lock
per virtqueue, so that there is no contention between
multiple queue trying to acquire it.
Signed-off-by: Maxime Coquelin <maxime.coque...@redhat.com>
---
lib/vhost/iotlb.c | 212 +++++++++++++++++++----------------------
lib/vhost/iotlb.h | 43 ++++++---
lib/vhost/vhost.c | 18 ++--
lib/vhost/vhost.h | 16 ++--
lib/vhost/vhost_user.c | 25 +++--
5 files changed, 160 insertions(+), 154 deletions(-)
[...]
diff --git a/lib/vhost/vhost_user.c b/lib/vhost/vhost_user.c
index d60e39b6bc..81ebef0137 100644
--- a/lib/vhost/vhost_user.c
+++ b/lib/vhost/vhost_user.c
@@ -7,7 +7,7 @@
* The vhost-user protocol connection is an external interface, so it
must be
* robust against invalid inputs.
*
- * This is important because the vhost-user frontend is only one step
removed
+* This is important because the vhost-user frontend is only one step
removed
This is changed by accident?
Yes, this will be fixed in v1.
Thanks,
Maxime
Thanks,
Chenbo
