Search debugger first in selected CBS toolchain.

[ewaterlander]

For Core Build System local debug target. If there is no absolute path set in the Debugger tab of the launch configuration, try to find the debugger first in the selected toolchain. If the debugger is not found in the toolchain, let GdbLaunch search in PATH.
If an absolute path is set, GdbLaunch will use that.

Fixes #1008

[ewaterlander]

Hi @jonahgraham and @betamaxbandit ,

Here is my fix for #1008

[github-actions]

Test Results

   590 files   -  11     590 suites   - 11   13m 8s ⏱️ +20s
10 079 tests  - 121  10 056 ✅  - 121  23 💤 ±0  0 ❌ ±0 
10 117 runs   - 121  10 094 ✅  - 121  23 💤 ±0  0 ❌ ±0 

Results for commit 4086e29. ± Comparison against base commit 751b031.

This pull request removes 121 tests.

org.eclipse.cdt.debug.core.memory.tests.PlainTextTransportTest ‑ transport0ffff
org.eclipse.cdt.debug.core.memory.tests.PlainTextTransportTest ‑ transport10000
org.eclipse.cdt.debug.core.memory.tests.PlainTextTransportTest ‑ transport10001
org.eclipse.cdt.debug.core.memory.tests.RAWBinaryTransportTest ‑ transport0ffff
org.eclipse.cdt.debug.core.memory.tests.RAWBinaryTransportTest ‑ transport10000
org.eclipse.cdt.debug.core.memory.tests.RAWBinaryTransportTest ‑ transport10001
org.eclipse.cdt.debug.core.memory.tests.SRecordTransportTest ‑ transport0ffff
org.eclipse.cdt.debug.core.memory.tests.SRecordTransportTest ‑ transport10000
org.eclipse.cdt.debug.core.memory.tests.SRecordTransportTest ‑ transport10001
org.eclipse.cdt.dsf.gdb.multicorevisualizer.ui.test.PersistentSettingsManagerTest ‑ testGlobalParamsWithMultipleInstances
…

♻️ This comment has been updated with latest results.

[jonahgraham]

This pull request removes 121 tests.

Please see #1002 (comment) for discussion about this new feature that comments tests results. I don't know why less tests were run, but the change doesn't seem to be the cause.

[jonahgraham]

This pull request removes 121 tests.

Please see #1002 (comment) for discussion about this new feature that comments tests results. I don't know why less tests were run, but the change doesn't seem to be the cause.

Never mind - the difference is explained in the log, the maven run did not complete. I have restarted it.

[jonahgraham]

This looks reasonable to me. @betamaxbandit can you confirm if this fixes the regression you identified in #1008?

[betamaxbandit]

LGTM.

[betamaxbandit]

This looks reasonable to me. @betamaxbandit can you confirm if this fixes the regression you identified in #1008?

I haven't tested this.

@ewaterlander , did you have chance to test this fix on Windows?

[jonahgraham]

@betamaxbandit and I did a screen share and tested this on Windows

[jonahgraham]

Tests are failing, catastrophically. I don't understand why and can't see that it is related to this change. But I need to investigate a bit before I merge to make sure this isn't somehow exposing some weird bug.

[jonahgraham]

Tests are failing, catastrophically. I don't understand why and can't see that it is related to this change. But I need to investigate a bit before I merge to make sure this isn't somehow exposing some weird bug.

#1035 has the same issue, so I am confident not related to this change now.

[ewaterlander]

@betamaxbandit and I did a screen share and tested this on Windows

I did test it, but on Linux. I could see that the debugger was taken from the toolchain.