# [PATCH] Fix start_command closing cmd->out/in regardless of cmd->close_out/in 4 messages from 2007-11-18 to 2007-11-19. Participants: Ping Yin, Junio C Hamano, Johannes Sixt. Thread: https://gitlist.dev/t/10920 ## Ping Yin, 2007-11-18 17:36 Subject: [PATCH] Fix start_command closing cmd->out/in regardless of cmd->close_out/in Message-ID: <1195407366-1610-1-git-send-email-pkufranky@gmail.com> URL: https://gitlist.dev/e/1195407366-1610-1-git-send-email-pkufranky%40gmail.com ``` When 'FILE *fp' is assigned to child_process.out and then start_command or run_command is run, the standard output of the child process is expected to be outputed to fp. However, sometimes fp is not expected to be closed since further IO may be still performmed on fp. --- run-command.c | 4 ---- 1 files changed, 0 insertions(+), 4 deletions(-) diff --git a/run-command.c b/run-command.c index 476d00c..4e5f58d 100644 --- a/run-command.c +++ b/run-command.c @@ -115,13 +115,9 @@ int start_command(struct child_process *cmd) if (need_in) close(fdin[0]); - else if (cmd->in) - close(cmd->in); if (need_out) close(fdout[1]); - else if (cmd->out > 1) - close(cmd->out); if (need_err) close(fderr[1]); -- 1.5.3.5.1876.g7ba19-dirty ``` ## Junio C Hamano, 2007-11-18 18:52 Subject: Re: [PATCH] Fix start_command closing cmd->out/in regardless of cmd->close_out/in Message-ID: <7vmytbv11m.fsf@gitster.siamese.dyndns.org> URL: https://gitlist.dev/e/7vmytbv11m.fsf%40gitster.siamese.dyndns.org In-Reply-To: <1195407366-1610-1-git-send-email-pkufranky@gmail.com> ``` Ping Yin writes: > When 'FILE *fp' is assigned to child_process.out and then start_command or > run_command is run, the standard output of the child process is expected to > be outputed to fp. However, sometimes fp is not expected to be closed since > further IO may be still performmed on fp. Could you clarify "However, sometimes" with a concrete codepath that has this problem in the proposed commit log message, and Sign-off your patch please? ``` ## Johannes Sixt, 2007-11-19 07:39 Subject: Re: [PATCH] Fix start_command closing cmd->out/in regardless of cmd->close_out/in Message-ID: <47413DB9.9030306@viscovery.net> URL: https://gitlist.dev/e/47413DB9.9030306%40viscovery.net In-Reply-To: <1195407366-1610-1-git-send-email-pkufranky@gmail.com> ``` Ping Yin schrieb: > When 'FILE *fp' is assigned to child_process.out and then start_command or > run_command is run, the standard output of the child process is expected to > be outputed to fp. However, sometimes fp is not expected to be closed since > further IO may be still performmed on fp. > --- > run-command.c | 4 ---- > 1 files changed, 0 insertions(+), 4 deletions(-) > > diff --git a/run-command.c b/run-command.c > index 476d00c..4e5f58d 100644 > --- a/run-command.c > +++ b/run-command.c > @@ -115,13 +115,9 @@ int start_command(struct child_process *cmd) > > if (need_in) > close(fdin[0]); > - else if (cmd->in) > - close(cmd->in); > > if (need_out) > close(fdout[1]); > - else if (cmd->out > 1) > - close(cmd->out); > > if (need_err) > close(fderr[1]); This is dangerous! You have to audit all current callers whether they close cmd->in or cmd->out (if they don't need the fd anymore). Otherwise you risk to keep a writable pipe end open and then the reader hangs, waiting for input that will never arrive. -- Hannes ``` ## Junio C Hamano, 2007-11-19 08:46 Subject: Re: [PATCH] Fix start_command closing cmd->out/in regardless of cmd->close_out/in Message-ID: <7vfxz2oc6e.fsf@gitster.siamese.dyndns.org> URL: https://gitlist.dev/e/7vfxz2oc6e.fsf%40gitster.siamese.dyndns.org In-Reply-To: <47413DB9.9030306@viscovery.net> ``` Johannes Sixt writes: > Ping Yin schrieb: >> When 'FILE *fp' is assigned to child_process.out and then start_command or >> run_command is run, the standard output of the child process is expected to >> be outputed to fp. However, sometimes fp is not expected to be closed since >> further IO may be still performmed on fp. >> --- >> run-command.c | 4 ---- >> 1 files changed, 0 insertions(+), 4 deletions(-) >> >> diff --git a/run-command.c b/run-command.c >> index 476d00c..4e5f58d 100644 >> --- a/run-command.c >> +++ b/run-command.c >> @@ -115,13 +115,9 @@ int start_command(struct child_process *cmd) >> if (need_in) >> close(fdin[0]); >> - else if (cmd->in) >> - close(cmd->in); >> if (need_out) >> close(fdout[1]); >> - else if (cmd->out > 1) >> - close(cmd->out); >> if (need_err) >> close(fderr[1]); > > This is dangerous! You have to audit all current callers whether they > close cmd->in or cmd->out (if they don't need the fd > anymore). I am reasonably sure that they are already relying on these auto closing of the file descriptors. Funny that somebody falls into the trap the day after we discussed it on another thread. ```