* Use fork instead of spawn
We no longer do in-place updating so no need for the spawn. The
advantage of a fork is that it preserves flags like --prof which you can
use to profile code-server.
Also I am not sure the comment about not being able to reload in place
with fork was even true to begin with.
* Refresh heartbeat patch
Seems to have gotten out of date a little.
* Propagate execArgv to extension host
This will let us profile the extension host.
Having NODE_OPTIONS set is unexpected and although the later flag should
override the previous flag it is not certain that will always be the
case.
Also some users are having issues with the 2 GB limit.
This is because Node uses SIGUSR1 to enable the debug listener so even
if you just want to restart code-server you end up enabling the debug
listener as well.
Opted to leave the SIGUSR1 handler in to avoid breaking existing
workflows even though it does mean even if you only want to enable the
debug listener you will end up restarting code-server as well. We could
consider removing it after a transition phase.
We pipe the child's stdout and stderr to the log file (and to the parent's streams) but since we used `inherit` for `stdio` this caused the child to use the parent's streams directly which made `child.stdout` and `child.stderr` non-existent and thus we had no file logging.
Using `pipe` creates stdin and stderr on the child.
Also move our memory default to the beginning of NODE_OPTIONS so it can
be overidden. The version of the flag with dashes seems to be the more
correct one now so use that instead of underscores.
Related: #2113.
I think having them combined and relying on if statements was getting
confusing especially if we want to add additional messages with
different payloads (which will soon be the case).
- Immediately create ipcMain so it doesn't have to be a function which I
think feels cleaner.
- Move exit handling to a separate function to compensate (otherwise
the VS Code CLI for example won't be able to exit on its own).
- New isChild prop that is clearer than checking for parentPid (IMO).
- Skip all the checks that aren't necessary for the child process (like
--help, --version, etc).
- Since we check if we're the child in entry go ahead and move the
wrap code into entry as well since that's basically what it does.
- Use a single catch at the end of the entry.
- Split out the VS Code CLI and existing instance code into separate
functions.