search for implementation very slow in latest version

Geoff Swartz could you collect the following diagnostic information in your environment so we could look into this performance issue:

1. Rider logs:

• Use Help→Diagnostic Tools→Choose Trace Scenarios action and pick SearchAndNavigation and Goto trace scenarios
• Reproduce the issue (trigger implementation search and wait until it finishes)
• Use Help→Collect Logs and Diagnostic Data to generate the log archive
• Disable all previously enabled trace scenarios

2. Performance snapshot collected by following SUPPORT-A-1354 Rider Profiling Instructions (.NET process) using Profile Backend with Timeline.

You can send us the diagnostics via the Upload Service and let us know the Upload ID in this thread.

I haven't had this problem again since the latest version came out.

This started happening again.  Here's the upload id for the diagnostic zip file.  2025_09_22_VXrCFGFX4BbsqpNafFJHEQ

Geoff Swartz, it looks like the upload only contains the snapshot. Would it be possible to collect the logs with diagnostic traces as well, as mentioned earlier?

Ok I'll try it again.

Here's the latest upload - 2025_09_23_PeFm32mwQXr6nt7w1kLpkE

Thank you for the diagnostics. This issue will be investigated further in the scope of RIDER-130600 Go to Implementation is very slow. Please follow (★) the linked issue to monitor progress. 

Gswartz, we are currently investigating the cause of this issue in your environment, and we currently suspect that there might be an issue in the native part of .NET CLR. In order to investigate this further, we need to collect another snapshot with an additional profiling option enabled. This process is a bit involved, so please let me know if you encounter any issues or have any questions.

Please send us another snapshot collected by following these steps:

• Download and install JetBrains dotTrace. If you don’t have a suitable license (dotUltimate or All Product Pack), you can start a new 30-day free trial.
• Close all currently running Rider instances
• Launch dotTrace

• On the dotTrace Home screen, select the green plus sign button (Add a run configuration)

  

   

• In the New Run Configuration dialog, select the External .NET Process

  

• Pick any name you prefer for the configuration, e.g. Rider and click Save

  

• Back on the dotTrace home page, pick the newly created configuration from the list (1), pick Timeline profiling with Download symbol files enabled and disable Collect profiling data from the start

  

• Launch Rider

• You should see dotTrace detecting and automatically attaching to the Rider.Backend.exe process

  

• When you are ready to reproduce the issue, hit Start recording in dotTrace

• Wait until you Get snapshot button becomes active/available in dotTrace UI, indicated by the blue button color:

  

• Switch to Rider and reproduce the problem (trigger implementation search and wait until it finishes)
• Switch back to dotTrace and press Get snapshot
• Wait until dotTrace finishes collecting the snapshot. This might take a while depending on your environment.
• Once dotTrace finishes collecting the snapshot, it should automatically open in dotTrace Viewer

• When dotTrace Viewer opens the snapshot, use File → Save Snapshot

  

• Upload the generated snapshot to the Upload Service and let us know the Upload ID
• Close any running Rider and dotTrace instances on the affected system

If you have any questions about this procedure, please let me know.

Gswartz, our development team reviewed the initial snapshot containing limited profiling information. Unfortunately, it looks like collecting a new Timeline snapshot with symbols (as described above) will also not be enough to pinpoint the cause of this issue.

Instead, could you please collect a PerfView snapshot in the affected environment?

To collect a snapshot, please follow the steps below:

• Download a set of helper scripts designed to download and run PerfView by downloading JetBrainsCollectPerfviewSnapshot_v4.zip attached to the SUPPORT-A-2191 How to collect ReSharper performance snapshot with PerfView article
• Extract the archive into an arbitrary folder on your machine
• Open PowerShell script PerfView.ps1

• Update /CircularMB to by setting the value to 32000, e.g. /CircularMB:32000. Line 38 should look like this after the change:

  $argumentList += " /BufferSizeMB:256 /StackCompression /CircularMB:32000 /KernelEvents:ThreadTime /NoGui /NoNGenRundown collect"

• Launch Rider and open the affected project
• Open Windows Console/Terminal and run PerfView_until_S_pressed.cmd script. PerfView will be downloaded to your system automatically and should start tracing
• Switch back to Rider and reproduce the issue, wait until Rider finishes implementation lookup
• Switch back to a window running PerfView and press S to stop tracing
• After you stop PerfView tracing, you should see a newly generated PerfViewData.ETL.zip archive in the directory containing the scripts (JetBrainsCollectPerfviewSnapshot)
• Send us the generated PerfViewData.ETL.zip via the Upload Service and let me know the Upload ID in this thread

Let me know if you have any questions about this or encounter any issues.