keepalive: ensure `close_all()` actually closes all cached connections
While debugging why LFS blob downloads are getting corrupted with workers, I
noticed that prior to spinning up the workers, the ConnectionManager has 2
connections to the server and calling `KeepAliveHandler.close_all()` left one
behind. The reason is the value component of `self._cm.get_all().items()` is a
list, and `self._cm.remove()` modifies said list while the caller is iterating
over it. Now `get_all()` is a deep copy of both the dict and lists in all
cases.
= Test the known() protocol function =
Create a test repository:
$ hg init repo
$ cd repo
$ touch a ; hg add a ; hg ci -ma
$ touch b ; hg add b ; hg ci -mb
$ touch c ; hg add c ; hg ci -mc
$ hg log --template '{node}\n'
991a3460af53952d10ec8a295d3d2cc2e5fa9690
0e067c57feba1a5694ca4844f05588bb1bf82342
3903775176ed42b1458a6281db4a0ccf4d9f287a
$ cd ..
Test locally:
$ hg debugknown repo 991a3460af53952d10ec8a295d3d2cc2e5fa9690 0e067c57feba1a5694ca4844f05588bb1bf82342 3903775176ed42b1458a6281db4a0ccf4d9f287a
111
$ hg debugknown repo 000a3460af53952d10ec8a295d3d2cc2e5fa9690 0e067c57feba1a5694ca4844f05588bb1bf82342 0003775176ed42b1458a6281db4a0ccf4d9f287a
010
$ hg debugknown repo
Test via HTTP:
$ hg serve -R repo -p $HGPORT -d --pid-file=hg.pid -E error.log -A access.log
$ cat hg.pid >> $DAEMON_PIDS
$ hg debugknown http://localhost:$HGPORT/ 991a3460af53952d10ec8a295d3d2cc2e5fa9690 0e067c57feba1a5694ca4844f05588bb1bf82342 3903775176ed42b1458a6281db4a0ccf4d9f287a
111
$ hg debugknown http://localhost:$HGPORT/ 000a3460af53952d10ec8a295d3d2cc2e5fa9690 0e067c57feba1a5694ca4844f05588bb1bf82342 0003775176ed42b1458a6281db4a0ccf4d9f287a
010
$ hg debugknown http://localhost:$HGPORT/
$ cat error.log
$ killdaemons.py