You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It seems like there is no check in place to ensure the SubAckPacket is the response to the SubscribePacket that is sent. From what I see it could be just any SubAckPacket .
JaggerJo
changed the title
SubscribeResult Constructor crashes if broker does not send subAckPacket.ReasonCodes for all topics
SubscribeAsync / UnsubscribeAsync does not check PacketIdentifier of response, just uses first received package
Jan 17, 2025
🐛 Bug Report
It seems like there is no check in place to ensure the
SubAckPacket
is the response to theSubscribePacket
that is sent. From what I see it could be just anySubAckPacket
.hivemq-mqtt-client-dotnet/Source/HiveMQtt/Client/HiveMQClient.cs
Lines 307 to 345 in 04593d4
So sometimes the wrong
SubAckPacket
ends up in here paired with the wrongSubscribeOptions
. This leads to anArgumentOutOfRangeException
here:hivemq-mqtt-client-dotnet/Source/HiveMQtt/Client/Results/SubscribeResult.cs
Lines 27 to 43 in 04593d4
🔬 How To Reproduce
Steps to reproduce the behavior:
Call Subscribe a lot of times in a concurrent manner with a different number of topics.
Code sample
Environment
Where are you running/using this client?
Windows, Linux, MacOS.
Reproduces with HiveMQtt 0.24.0 and current master.
Screenshots
📈 Expected behavior
📎 Additional context
The text was updated successfully, but these errors were encountered: