[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [nmh-workers] closefds() _before_ fork?
From: |
Ken Hornstein |
Subject: |
Re: [nmh-workers] closefds() _before_ fork? |
Date: |
Tue, 23 Apr 2019 12:00:49 -0400 |
>> It's not the only one, e.g. docs/historical/mh-6.8.5/uip/post.c has
>>
>> 2622 if (fd != 0)
>> 2623 (void) dup2 (fd, 0);
>> 2624 (void) freopen ("/dev/null", "w", stdout);
>> 2625 (void) freopen ("/dev/null", "w", stderr);
>> 2626 if (fd != 3) /* backwards compatible... */
>> 2627 (void) dup2 (fd, 3);
>> 2628 closefds (4);
>>
>> Note the comment.
>
>The *real* WTF there is that the code doesn't actually know if fd 3 is
>known to be open on a file that should, or if it's just a stray leaked fd.
Wow. So do any greybeards out there have any idea of why this is there??
Clearly this was a thing, but I have no idea why. I see that same code
in MH 5, so it's been that way for approximately forever.
--Ken
- [nmh-workers] closefds() _before_ fork?, Ken Hornstein, 2019/04/22
- Re: [nmh-workers] closefds() _before_ fork?, Ralph Corderoy, 2019/04/22
- Re: [nmh-workers] closefds() _before_ fork?, David Levine, 2019/04/22
- Re: [nmh-workers] closefds() _before_ fork?, Ken Hornstein, 2019/04/22
- Re: [nmh-workers] closefds() _before_ fork?, David Levine, 2019/04/22
- Re: [nmh-workers] closefds() _before_ fork?, Robert Elz, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Valdis Klētnieks, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Robert Elz, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Valdis Klētnieks, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Ralph Corderoy, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Paul Vixie, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Ken Hornstein, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Ralph Corderoy, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Ken Hornstein, 2019/04/23
- Re: [nmh-workers] closefds() _before_ fork?, Tom Lane, 2019/04/23