Re: [PATCH] remote_daemon_stream: Hold an extra reference to stream in daemonStreamFilter

[Lance Liu <liu.lance.89@xxxxxxxxx>]

Yeah, your patch is perfectly fine for stream freed problem.

But I have reviewed code in daemonStreamEvent() and daemonStreamFilter() again, and I think there still two issue need to be reconsidered:

1. stream->ref only ++ in daemonStreamFilter, except for error occurred call daemonRemoveClientStream() lead to ref--, like code as follow. Though code won't be executed because of no VIR_STREAM_EVENT_HANGUP event taken place,

but it is not good code:
/* If we got HANGUP, we need to only send an empty
     * packet so the client sees an EOF and cleans up
     */
    if (!stream->closed && !stream->recvEOF &&
        (events & VIR_STREAM_EVENT_HANGUP)) {
        virNetMessagePtr msg;
        events &= ~(VIR_STREAM_EVENT_HANGUP);
        stream->tx = false;
        stream->recvEOF = true;
        if (!(msg = virNetMessageNew(false))) {
            daemonRemoveClientStream(client, stream);
            virNetServerClientClose(client);
            goto cleanup;
        }
        msg->cb = daemonStreamMessageFinished;
        msg->opaque = stream;
        stream->refs++;
        if (virNetServerProgramSendStreamData(stream->prog,
                                              client,
                                              msg,
                                              stream->procedure,
                                              stream->serial,
                                              "", 0) < 0) {
            virNetMessageFree(msg);
            daemonRemoveClientStream(client, stream);
            virNetServerClientClose(client);
            goto cleanup;
        }
    }

2. call virNetServerClientClose() is still inappropriate in  because the it may free the client resource which other threads need ref, may be replace it with virNetServerClientImmediateClose() is better,

the daemon can process client resource release unified

3. Segmentfault may still exists when another thread call remoteClientCloseFunc in virNetServerClientClose()(for example, when new force console session break down existed console session, and existed console session

's client close the session simutaneously, but remoteClientCloseFunc not lock(priv->lock), so the resource maybe released when daemonStreamEvent ref)

How do you see it?

Best Regards!

Michal Privoznik <mprivozn@xxxxxxxxxx> 于2019年11月26日周二 上午12:49写道：

In v5.9.0-273-g8ecab214de I've tried to fix a lock ordering
problem, but introduced a crasher. Problem is that because the
client lock is unlocked (in order to honour lock ordering) the
stream we are currently checking in daemonStreamFilter() might be
freed and thus stream->priv might not even exist when the control
get to virMutexLock() call.

To resolve this, grab an extra reference to the stream and handle
its cleanup should the refcounter reach zero after the deref.
If that's the case and we are the only ones holding a reference
to the stream, we MUST return a positive value to make
virNetServerClientDispatchRead() break its loop where it iterates
over filters. The problem is, if we did not do so, then
"filter = filter->next" line will read from a memory that was
just freed (freeing a stream also unregisters its filter).

Signed-off-by: Michal Privoznik <mprivozn@xxxxxxxxxx>
---

Reproducing this issue is very easy:

1) put sleep(5) right after virObjectUnlock(client) in the fist hunk,
2) virsh console --force $dom   and type something so that the stream
   has some data to process
3) while 2) is still running, run the same command from another terminal
4) observe libvirtd crash

 src/remote/remote_daemon_stream.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/src/remote/remote_daemon_stream.c b/src/remote/remote_daemon_stream.c
index 82cadb67ac..73e4d7befb 100644
--- a/src/remote/remote_daemon_stream.c
+++ b/src/remote/remote_daemon_stream.c
@@ -293,10 +293,25 @@ daemonStreamFilter(virNetServerClientPtr client,
     daemonClientStream *stream = opaque;
     int ret = 0;

+    /* We must honour lock ordering here. Client private data lock must
+     * be acquired before client lock. Bu we are already called with
+     * client locked. To avoid stream disappearing while we unlock
+     * everything, let's increase its refcounter. This has some
+     * implications though. */
+    stream->refs++;
     virObjectUnlock(client);
     virMutexLock(&stream->priv->lock);
     virObjectLock(client);

+    if (stream->refs == 1) {
+        /* So we are the only ones holding the reference to the stream.
+         * Return 1 to signal to the caller that we've processed the
+         * message. And to "process" means free. */
+        virNetMessageFree(msg);
+        ret = 1;
+        goto cleanup;
+    }
+
     if (msg->header.type != VIR_NET_STREAM &&
         msg->header.type != VIR_NET_STREAM_HOLE)
         goto cleanup;
@@ -318,6 +333,10 @@ daemonStreamFilter(virNetServerClientPtr client,

  cleanup:
     virMutexUnlock(&stream->priv->lock);
+    /* Don't pass client here, because client is locked here and this
+     * function might try to lock it again which would result in a
+     * deadlock. */
+    daemonFreeClientStream(NULL, stream);
     return ret;
 }

--
2.23.0

--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list