When ->ovv_begin is called in LeiXSearch->do_query in the top-level
lei-daemon process, $lei->{pkt_op_p} still exists. We must make
sure we're exiting the correct process since lei->out can call
lei->fail and lei->fail calls lei->x_it.
As to avoiding how I caused ->ovv_begin failures to begin with,
that's for a much bigger change...
---
lib/PublicInbox/LEI.pm | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/lib/PublicInbox/LEI.pm b/lib/PublicInbox/LEI.pm
index 7bc7b2dc..e060bcbe 100644
--- a/lib/PublicInbox/LEI.pm
+++ b/lib/PublicInbox/LEI.pm
@@ -40,6 +40,7 @@ $GLP_PASS->configure(qw(gnu_getopt no_ignore_case auto_abbrev
pass_through));
our (%PATH2CFG, # persistent for socket daemon
$MDIR2CFGPATH, # /path/to/maildir => { /path/to/config => [ ino watches ] }
$OPT, # shared between optparse and opt_dash callback (for Getopt::Long)
+$daemon_pid
);
# TBD: this is a documentation mechanism to show a subcommand
@@ -486,7 +487,7 @@ sub x_it ($$) {
stop_pager($self);
if ($self->{pkt_op_p}) { # worker => lei-daemon
$self->{pkt_op_p}->pkt_do('x_it', $code);
- exit($code >> 8);
+ exit($code >> 8) if $$ != $daemon_pid;
} elsif ($self->{sock}) { # lei->daemon => lei(1) client
send($self->{sock}, "x_it $code", 0);
} elsif ($quit == \&CORE::exit) { # an admin (one-shot) command
@@ -1341,8 +1342,8 @@ sub lazy_start {
my $pid = fork;
return if $pid;
$0 = "lei-daemon $path";
- local %PATH2CFG;
- local $MDIR2CFGPATH;
+ local (%PATH2CFG, $MDIR2CFGPATH);
+ local $daemon_pid = $$;
$listener->blocking(0);
my $exit_code;
my $pil = PublicInbox::Listener->new($listener, \&accept_dispatch);
