From: "Michael S. Tsirkin" <mst@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Peter Maydell" <peter.maydell@linaro.org>,
"Marc-André Lureau" <marcandre.lureau@redhat.com>,
"Jason Wang" <jasowang@redhat.com>
Subject: [Qemu-devel] [PULL 28/41] vhost-user: keep vhost_net after a disconnection
Date: Fri, 29 Jul 2016 06:16:41 +0300 [thread overview]
Message-ID: <1469762011-7902-29-git-send-email-mst@redhat.com> (raw)
In-Reply-To: <1469762011-7902-1-git-send-email-mst@redhat.com>
From: Marc-André Lureau <marcandre.lureau@redhat.com>
Many code paths assume get_vhost_net() returns non-null.
Keep VhostUserState.vhost_net after a successful vhost_net_init(),
instead of freeing it in vhost_net_cleanup().
VhostUserState.vhost_net is thus freed before after being recreated or
on final vhost_user_cleanup() and there is no need to save the acked
features.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/net/vhost_net.c | 1 -
net/tap.c | 1 +
net/vhost-user.c | 36 +++++++++++++++++++-----------------
3 files changed, 20 insertions(+), 18 deletions(-)
diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c
index c11f69c..7b76591 100644
--- a/hw/net/vhost_net.c
+++ b/hw/net/vhost_net.c
@@ -378,7 +378,6 @@ void vhost_net_stop(VirtIODevice *dev, NetClientState *ncs,
void vhost_net_cleanup(struct vhost_net *net)
{
vhost_dev_cleanup(&net->dev);
- g_free(net);
}
int vhost_net_notify_migration_done(struct vhost_net *net, char* mac_addr)
diff --git a/net/tap.c b/net/tap.c
index 40a8c74..6abb962 100644
--- a/net/tap.c
+++ b/net/tap.c
@@ -312,6 +312,7 @@ static void tap_cleanup(NetClientState *nc)
if (s->vhost_net) {
vhost_net_cleanup(s->vhost_net);
+ g_free(s->vhost_net);
s->vhost_net = NULL;
}
diff --git a/net/vhost-user.c b/net/vhost-user.c
index 2cac32e..d2a984d 100644
--- a/net/vhost-user.c
+++ b/net/vhost-user.c
@@ -45,11 +45,6 @@ uint64_t vhost_user_get_acked_features(NetClientState *nc)
return s->acked_features;
}
-static int vhost_user_running(VhostUserState *s)
-{
- return (s->vhost_net) ? 1 : 0;
-}
-
static void vhost_user_stop(int queues, NetClientState *ncs[])
{
VhostUserState *s;
@@ -59,15 +54,14 @@ static void vhost_user_stop(int queues, NetClientState *ncs[])
assert(ncs[i]->info->type == NET_CLIENT_DRIVER_VHOST_USER);
s = DO_UPCAST(VhostUserState, nc, ncs[i]);
- if (!vhost_user_running(s)) {
- continue;
- }
if (s->vhost_net) {
/* save acked features */
- s->acked_features = vhost_net_get_acked_features(s->vhost_net);
+ uint64_t features = vhost_net_get_acked_features(s->vhost_net);
+ if (features) {
+ s->acked_features = features;
+ }
vhost_net_cleanup(s->vhost_net);
- s->vhost_net = NULL;
}
}
}
@@ -75,6 +69,7 @@ static void vhost_user_stop(int queues, NetClientState *ncs[])
static int vhost_user_start(int queues, NetClientState *ncs[])
{
VhostNetOptions options;
+ struct vhost_net *net = NULL;
VhostUserState *s;
int max_queues;
int i;
@@ -85,33 +80,39 @@ static int vhost_user_start(int queues, NetClientState *ncs[])
assert(ncs[i]->info->type == NET_CLIENT_DRIVER_VHOST_USER);
s = DO_UPCAST(VhostUserState, nc, ncs[i]);
- if (vhost_user_running(s)) {
- continue;
- }
options.net_backend = ncs[i];
options.opaque = s->chr;
options.busyloop_timeout = 0;
- s->vhost_net = vhost_net_init(&options);
- if (!s->vhost_net) {
+ net = vhost_net_init(&options);
+ if (!net) {
error_report("failed to init vhost_net for queue %d", i);
goto err;
}
if (i == 0) {
- max_queues = vhost_net_get_max_queues(s->vhost_net);
+ max_queues = vhost_net_get_max_queues(net);
if (queues > max_queues) {
error_report("you are asking more queues than supported: %d",
max_queues);
goto err;
}
}
+
+ if (s->vhost_net) {
+ vhost_net_cleanup(s->vhost_net);
+ g_free(s->vhost_net);
+ }
+ s->vhost_net = net;
}
return 0;
err:
- vhost_user_stop(i + 1, ncs);
+ if (net) {
+ vhost_net_cleanup(net);
+ }
+ vhost_user_stop(i, ncs);
return -1;
}
@@ -150,6 +151,7 @@ static void vhost_user_cleanup(NetClientState *nc)
if (s->vhost_net) {
vhost_net_cleanup(s->vhost_net);
+ g_free(s->vhost_net);
s->vhost_net = NULL;
}
if (s->chr) {
--
MST
next prev parent reply other threads:[~2016-07-29 3:16 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-07-29 3:14 [Qemu-devel] [PULL 00/41] pc, pci, virtio: cleanups, fixes Michael S. Tsirkin
2016-07-29 3:14 ` [Qemu-devel] [PULL 01/41] pcie: fix link active status bit migration Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 02/41] hw/pcie-root-port: Fix PCIe root port initialization Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 03/41] hw/pxb: declare pxb devices as not hot-pluggable Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 04/41] hw/acpi: fix a DSDT table issue when a pxb is present Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 05/41] acpi: refactor pxb crs computation Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 06/41] hw/apci: handle 64-bit MMIO regions correctly Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 07/41] hw/pci-bridge: Convert pxb initialization functions to Error Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 08/41] apb: convert init to realize Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 09/41] hw/virtio-pci: fix virtio behaviour Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 10/41] virtio: check vring descriptor buffer length Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 11/41] misc: indentation Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 12/41] vhost-user: minor simplification Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 13/41] vhost-user: disconnect on HUP Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 14/41] vhost: don't assume opaque is a fd, use backend cleanup Michael S. Tsirkin
2016-07-29 3:15 ` [Qemu-devel] [PULL 15/41] vhost: make vhost_log_put() idempotent Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 16/41] vhost: assert the log was cleaned up Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 17/41] vhost: fix cleanup on not fully initialized device Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 18/41] vhost: make vhost_dev_cleanup() idempotent Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 19/41] vhost-net: always call vhost_dev_cleanup() on failure Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 20/41] vhost: fix calling vhost_dev_cleanup() after vhost_dev_init() Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 21/41] vhost: do not assert() on vhost_ops failure Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 22/41] vhost: add missing VHOST_OPS_DEBUG Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 23/41] vhost: use error_report() instead of fprintf(stderr, ...) Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 24/41] qemu-char: fix qemu_chr_fe_set_msgfds() crash when disconnected Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 25/41] vhost-user: call set_msgfds unconditionally Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 26/41] vhost-user: check qemu_chr_fe_set_msgfds() return value Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 27/41] vhost-user: check vhost_user_{read, write}() " Michael S. Tsirkin
2016-07-29 3:16 ` Michael S. Tsirkin [this message]
2016-07-29 3:16 ` [Qemu-devel] [PULL 29/41] vhost-user: add get_vhost_net() assertions Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 30/41] Revert "vhost-net: do not crash if backend is not present" Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 31/41] vhost-net: vhost_migration_done is vhost-user specific Michael S. Tsirkin
2016-07-29 3:16 ` [Qemu-devel] [PULL 32/41] vhost: add assert() to check runtime behaviour Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 33/41] char: add chr_wait_connected callback Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 34/41] char: add and use tcp_chr_wait_connected Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 35/41] vhost-user: wait until backend init is completed Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 36/41] tests: plug some leaks in virtio-net-test Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 37/41] tests: fix vhost-user-test leak Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 38/41] vhost-user: add error report in vhost_user_write() Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 39/41] vhost: add vhost_net_set_backend() Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 40/41] vhost: do not update last avail idx on get_vring_base() failure Michael S. Tsirkin
2016-07-29 3:17 ` [Qemu-devel] [PULL 41/41] mptsas: Fix a migration compatible issue Michael S. Tsirkin
2016-07-29 11:36 ` [Qemu-devel] [PULL 00/41] pc, pci, virtio: cleanups, fixes Peter Maydell
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1469762011-7902-29-git-send-email-mst@redhat.com \
--to=mst@redhat.com \
--cc=jasowang@redhat.com \
--cc=marcandre.lureau@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).