stream-clone: disable gc for `_entries_walk` duration
The number of small container created turn Python in a gc-frenzy that seriously
impact performance.
This significantly boost performance. The following number comes from a large
private repository using perf::stream-locked-section:
base-line: 35.04 seconds
prev-change: 24.51 seconds (-30%)
this-change: 20.88 seconds (-40% from baseline; -15% from previous changes)
$ hg init a
$ cd a
$ hg init b
$ echo x > b/x
Should print nothing:
$ hg add b
$ hg st
$ echo y > b/y
$ hg st
Should fail:
$ hg st b/x
abort: path 'b/x' is inside nested repo 'b'
[255]
$ hg add b/x
abort: path 'b/x' is inside nested repo 'b'
[255]
Should fail:
$ hg add b b/x
abort: path 'b/x' is inside nested repo 'b'
[255]
$ hg st
Should arguably print nothing:
$ hg st b
$ echo a > a
$ hg ci -Ama a
Should fail:
$ hg mv a b
abort: path 'b/a' is inside nested repo 'b'
[255]
$ hg st
$ cd ..