how to improve that codelldb consumes a large amount of CPU

[jm-observer]

When I try to debug Lapce on Windows, codelldb consumes a large amount of CPU, and I have to wait for it to decrease before I can operate, which takes a long time. How can I improve this situation?

launch configuration

{
            "type": "lldb",
            "request": "launch",
            "name": "Debug",
            "program": "target\\fastdev\\lapce.exe",
            "args": [
                "-w"
            ],
            "terminal": "console",
            "cwd": "${workspaceFolder}"
        }

[vadimcn]

Which build target? If compiling for -msvc, be aware that LLDB's support for MS PDB debug info had always been sketchy. For example, it does not implement incremental loading of the debug info, which could explain what you are observing.

[jm-observer]

Which build target?

yes. it is stable-x86_64-pc-windows-msvc
I tried to remove all breakpoints, but it still takes a long time to launch. Is this normal?
If I only use LLDB to debug, can the result be comparable?

[vadimcn]

I tried to remove all breakpoints, but it still takes a long time to launch. Is this normal?

Depends on how large debug information is and your hardware.
It doesn't matter whether you set breakpoints or not, the debugger will still parse the debug info.

If I only use LLDB to debug, can the result be comparable?

Not sure what you mean - what else would you use LLDB for if not debugging?

[jm-observer]

thanks! Since I can debug Lapce normally in RustRover, I think there might be an issue elsewhere.