You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mina.apache.org by "Jiachen Lu (Jira)" <ji...@apache.org> on 2024/09/08 13:06:00 UTC

[jira] [Updated] (SSHD-1341) Deadlock in old version of SSHD

     [ https://issues.apache.org/jira/browse/SSHD-1341?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jiachen Lu updated SSHD-1341:
-----------------------------
    Affects Version/s: 2.4.0
                           (was: 0.14.0)
          Description: 
h2. In a nutshell
I used some static analysis tools to discover a potential deadlock in version 2.4.0 of mina-sshd. I would like to know if it is a false positive.
 
The related lock objects and trace are shown below.
 
I look forward to the community's assistance in confirming this issue. Thank you for your response. If you need more information, please feel free to contact me.
 
This deadlock might be similar to [SSHD-966|https://issues.apache.org/jira/browse/SSHD-966]
 
h2. Deadlock Pair
Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
 
./sshd-core/src/main/java/org/apache/sshd/common/session/helpers/AbstractSession.java
 
{code:java}
912    protected IoWriteFuture doWritePacket(Buffer buffer) throws IOException { 
913        // Synchronize all write requests as needed by the encoding algorithm 
914        // and also queue the write request in this synchronized block to ensure 915        // packets are sent in the correct order 
916        synchronized (encodeLock) { 
917            Buffer packet = resolveOutputPacket(buffer); 
918            IoSession networkSession = getIoSession(); 
919            return networkSession.writePacket(packet); 
920        }
921    }{code}
 
 
 
Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
 
./sshd-core/src/main/java/org/apache/sshd/server/forward/TcpipServerChannel.java
{code:java}
203            public void messageReceived(IoSession session, Readable message) throws Exception {
204                if (isClosing()) {
205                    if (debugEnabled) {
206                        log.debug("doInit({}) Ignoring write to channel in CLOSING state", TcpipServerChannel.this);
207                    }
208                } else {
209                    Buffer buffer = new ByteArrayBuffer(message.available() + Long.SIZE, false);
210                    buffer.putBuffer(message);
211                    out.write(buffer.array(), buffer.rpos(), buffer.available());
212 >                  out.flush();
213                }
214            } {code}
h2. Trace
h3. Trace1

  !image-2024-06-16-21-16-19-706.png!
h3. 
Trace2  !image-2024-06-16-21-16-41-229.png!
 

  was:
# SSHD Deadlock
## In a nutshell
 
I used some static analysis tools to discover a potential deadlock in version 0.14.x of mina-sshd. I would like to know if it is a false positive.
 
The related lock objects and trace are shown below.
 
I look forward to the community's assistance in confirming this issue. Thank you for your response. If you need more information, please feel free to contact me.
 
This deadlock might be similar to [SSHD-966](https://issues.apache.org/jira/browse/SSHD-966).
 
## Deadlock Pair
Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
 
./sshd-core/src/main/java/org/apache/sshd/common/session/AbstractSession.java
 
```java
551    protected IoWriteFuture doWritePacket(Buffer buffer) throws IOException {
552        // Synchronize all write requests as needed by the encoding algorithm
553        // and also queue the write request in this synchronized block to ensure
554        // packets are sent in the correct order
555 >      synchronized (encodeLock) {
556            encode(buffer);
557            return ioSession.write(buffer);
558        }
559    }
```
 
Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
 
./sshd-core/src/main/java/org/apache/sshd/common/forward/TcpipServerChannel.java
 
```java
113            public void messageReceived(IoSession session, Readable message) throws Exception {
114                if (isClosing()) {
115                    log.debug("Ignoring write to channel {} in CLOSING state", id);
116                } else {
117                    Buffer buffer = new Buffer();
118                    buffer.putBuffer(message);
119                    out.write(buffer.array(), buffer.rpos(), buffer.available());
120 >                  out.flush();
121                }
122            }
```
 
## Trace
### Trace1
 
!image-2024-06-16-21-16-19-706.png!
 
### Trace2
!image-2024-06-16-21-16-41-229.png!
 


> Deadlock in old version of SSHD
> -------------------------------
>
>                 Key: SSHD-1341
>                 URL: https://issues.apache.org/jira/browse/SSHD-1341
>             Project: MINA SSHD
>          Issue Type: Question
>    Affects Versions: 2.4.0
>            Reporter: Jiachen Lu
>            Priority: Minor
>         Attachments: image-2024-06-16-21-16-19-706.png, image-2024-06-16-21-16-41-229.png
>
>
> h2. In a nutshell
> I used some static analysis tools to discover a potential deadlock in version 2.4.0 of mina-sshd. I would like to know if it is a false positive.
>  
> The related lock objects and trace are shown below.
>  
> I look forward to the community's assistance in confirming this issue. Thank you for your response. If you need more information, please feel free to contact me.
>  
> This deadlock might be similar to [SSHD-966|https://issues.apache.org/jira/browse/SSHD-966]
>  
> h2. Deadlock Pair
> Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
>  
> ./sshd-core/src/main/java/org/apache/sshd/common/session/helpers/AbstractSession.java
>  
> {code:java}
> 912    protected IoWriteFuture doWritePacket(Buffer buffer) throws IOException { 
> 913        // Synchronize all write requests as needed by the encoding algorithm 
> 914        // and also queue the write request in this synchronized block to ensure 915        // packets are sent in the correct order 
> 916        synchronized (encodeLock) { 
> 917            Buffer packet = resolveOutputPacket(buffer); 
> 918            IoSession networkSession = getIoSession(); 
> 919            return networkSession.writePacket(packet); 
> 920        }
> 921    }{code}
>  
>  
>  
> Deadlock `synchronized void flush() throws IOException` (Trace 1) and `synchronized (encodeLock)` (Trace 2) acquire locks `this` in class `ChannelOutputStream` and `encodeLock` in reverse orders.
>  
> ./sshd-core/src/main/java/org/apache/sshd/server/forward/TcpipServerChannel.java
> {code:java}
> 203            public void messageReceived(IoSession session, Readable message) throws Exception {
> 204                if (isClosing()) {
> 205                    if (debugEnabled) {
> 206                        log.debug("doInit({}) Ignoring write to channel in CLOSING state", TcpipServerChannel.this);
> 207                    }
> 208                } else {
> 209                    Buffer buffer = new ByteArrayBuffer(message.available() + Long.SIZE, false);
> 210                    buffer.putBuffer(message);
> 211                    out.write(buffer.array(), buffer.rpos(), buffer.available());
> 212 >                  out.flush();
> 213                }
> 214            } {code}
> h2. Trace
> h3. Trace1
>   !image-2024-06-16-21-16-19-706.png!
> h3. 
> Trace2  !image-2024-06-16-21-16-41-229.png!
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@mina.apache.org
For additional commands, e-mail: dev-help@mina.apache.org