logtoprocess: connect all fds to /dev/null to avoid bad interaction with pager
We detected that pager is waiting for log-to-process script to finish, which
is annoying when adding a script on commandfinish that does an HTTP push.
There seems to be no workaround on the script side and it will make the
behavior on Linux/MacOS closer to the Windows behavior.
The drawback is that it makes the related tests more flaky as log-to-process
outputs are now really asynchronous.
If it's considered a BC change, another option would be to add a config option
for this new behavior. I personally think that the different behavior between
Windows and Linux is confusing and that it's a bug I would be fine with a new
config option.
Differential Revision: https://phab.mercurial-scm.org/D4816
--- a/hgext/logtoprocess.py Fri Nov 03 21:35:36 2017 +0100
+++ b/hgext/logtoprocess.py Fri Nov 03 21:11:07 2017 +0100
@@ -83,11 +83,16 @@
else:
newsession = {'start_new_session': True}
try:
- # connect stdin to devnull to make sure the subprocess can't
- # muck up that stream for mercurial.
+ # connect std* to devnull to make sure the subprocess can't
+ # muck up these stream for mercurial.
+ # Connect all the streams to be more close to Windows behavior
+ # and pager will wait for scripts to end if we don't do that
+ nullrfd = open(os.devnull, 'r')
+ nullwfd = open(os.devnull, 'w')
subprocess.Popen(
procutil.tonativestr(script),
- shell=True, stdin=open(os.devnull, 'r'),
+ shell=True, stdin=nullrfd,
+ stdout=nullwfd, stderr=nullwfd,
env=procutil.tonativeenv(env),
close_fds=True, **newsession)
finally:
--- a/tests/test-logtoprocess.t Fri Nov 03 21:35:36 2017 +0100
+++ b/tests/test-logtoprocess.t Fri Nov 03 21:11:07 2017 +0100
@@ -121,4 +121,4 @@
$ touch $TESTTMP/wait-for-touched
$ sleep 0.2
$ test -f $TESTTMP/touched && echo "SUCCESS Pager is not waiting on ltp" || echo "FAIL Pager is waiting on ltp"
- FAIL Pager is waiting on ltp
+ SUCCESS Pager is not waiting on ltp