Handling the ContinuedEvent

[asimgunes]

This update implements handling mechanism for "resume" gdb signal and redirect the "resume" signal as "ContinuedEvent" as described in the Debug Adapter Protocol.

This implementation is related to the issue #332

[jonahgraham]

@asimgunes this looks promising - it over sends ContinuedEvent, which is not strictly a bug, but may be undesired. From the DAP docs:

Please note: a debug adapter is not expected to send this event in response to a request that implies that execution continues, e.g. launch or continue.

It is only necessary to send a continued event if there was no previous request that implied this.

See here that VSCode already sends a simulated continued event on operations like next.

We could avoid sending these unnecessary ContinuedEvents to avoid unnecessary updates to the UI and related code.

The flip-side of this issue is that we aren't testing that we are sending continued events when no previous request implies one. To achieve this, instead of doing dc.continueRequest to continue, send a CLI command for continue either by using cdt-gdb-tests/executeCommand as you did previously here, or by sending an evaluate request with the > prefix to make it a GDB command like here

[jonahgraham]

This won't work on Windows in non-async case. see

cdt-gdb-adapter/src/integration-tests/breakpoints.spec.ts

Lines 103 to 106 in ffeb53a

	if (os.platform() === 'win32' && (!isRemoteTest || !gdbAsync)) {
	// win32 host can only pause remote + mi-async targets
	this.skip();
	}

[asimgunes]

I made a similar implementation, thanks for the suggestion.

[jonahgraham]

@asimgunes I have completed a review

[asimgunes]

Hi @jonahgraham ,

I updated the test scenarios and split the non-stop and all-stop cases in multithread environment. Also, switched to use cdt-gdb-tests/executeCommand for sending the continue commands.

Is it possible to re-check the pull-request again?

[jonahgraham]

This LGTM, apart from thread id == 1, see line comments

[asimgunes]

Hi @jonahgraham ,

I made the suggested changes, could you please review again when you are suitable?

Could you please check the update when you are suitable?

[jonahgraham]

The version with this fix is published to NPM https://www.npmjs.com/package/cdt-gdb-adapter/v/1.0.1-next.20240927165134.44fda62.0