unofficial mirror of libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Carl Edquist <edquist@cs.wisc.edu>
To: Zachary Santer <zsanter@gmail.com>
Cc: libc-alpha@sourceware.org, coreutils@gnu.org,
	"Pádraig Brady" <p@draigbrady.com>
Subject: Re: Modify buffering of standard streams via environment variables (not LD_PRELOAD)?
Date: Sat, 20 Apr 2024 16:45:34 -0500 (CDT)	[thread overview]
Message-ID: <6b1c46a9-95a6-9422-85aa-fa8035d6c429@cs.wisc.edu> (raw)
In-Reply-To: <CABkLJU+rnFaB6JEiadcowC7pY+FmZ3Zg7wO3gXtQayUoeQ-Y6g@mail.gmail.com>

On Sat, 20 Apr 2024, Zachary Santer wrote:

> This was actually in RHEL 7.

Oh.  In that case it might be worth looking into ...


> I don't know how buffering works when stdout and stderr get redirected 
> to the same pipe. You'd think, whatever it is, it would have to be smart 
> enough to keep them interleaved in the same order they were printed to 
> in. That in mind, I would assume they both get placed into the same 
> block buffer by default.

My take is always to try it and find out.  Though in this case I think the 
default (without using stdbuf) is that the program's stderr is output to 
the pipe immediately (ie, unbuffered) on each library call (fprintf(3), 
fputs(3), putc(3), fwrite(3)), while stdout is written to the pipe at 
block boundaries - even though fd 1 and 2 refer to the same pipe.

If you force line buffering for stdout and stderr, that is likely what you 
want, and it will interleave _lines_ in the order that they were printed.

However, stdout and stderr are still separate streams even if they refer 
to the same output file/pipe/device, so partial lines are not interleaved 
in the order that they were printed.

For example:

 	#include <stdio.h>

 	int main()
 	{
 		putc('a', stderr);
 		putc('1', stdout);
 		putc('b', stderr);
 		putc('2', stdout);
 		putc('c', stderr);
 		putc('3', stdout);
 		putc('\n', stderr);
 		putc('\n', stdout);

 		return 0;
 	}

will output "abc\n123\n" instead of "a1b2c3\n\n", even if you run it as

 	$ ./abc123 2>&1 | cat
or
 	$ stdbuf -oL -eL ./abc123 2>&1 | cat


...

Not that that's relevant for what you're doing :)

Carl


      reply	other threads:[~2024-04-20 21:44 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CABkLJULa8c0zr1BkzWLTpAxHBcpb15Xms0-Q2OOVCHiAHuL0uA@mail.gmail.com>
     [not found] ` <9831afe6-958a-fbd3-9434-05dd0c9b602a@draigBrady.com>
2024-03-10 15:29   ` RFE: enable buffering on null-terminated data Zachary Santer
2024-03-10 20:36     ` Carl Edquist
2024-03-11  3:48       ` Zachary Santer
2024-03-11 11:54         ` Carl Edquist
2024-03-11 15:12           ` Examples of concurrent coproc usage? Zachary Santer
2024-03-14  9:58             ` Carl Edquist
2024-03-17 19:40               ` Zachary Santer
2024-04-01 19:24               ` Chet Ramey
2024-04-01 19:31                 ` Chet Ramey
2024-04-02 16:22                   ` Carl Edquist
2024-04-03 13:54                     ` Chet Ramey
2024-04-03 14:32               ` Chet Ramey
2024-04-03 17:19                 ` Zachary Santer
2024-04-08 15:07                   ` Chet Ramey
2024-04-09  3:44                     ` Zachary Santer
2024-04-13 18:45                       ` Chet Ramey
2024-04-14  2:09                         ` Zachary Santer
2024-04-04 12:52                 ` Carl Edquist
2024-04-04 23:23                   ` Martin D Kealey
2024-04-08 19:50                     ` Chet Ramey
2024-04-09 14:46                       ` Zachary Santer
2024-04-13 18:51                         ` Chet Ramey
2024-04-09 15:58                       ` Carl Edquist
2024-04-13 20:10                         ` Chet Ramey
2024-04-14 18:43                           ` Zachary Santer
2024-04-15 18:55                             ` Chet Ramey
2024-04-15 17:01                           ` Carl Edquist
2024-04-17 14:20                             ` Chet Ramey
2024-04-20 22:04                               ` Carl Edquist
2024-04-22 16:06                                 ` Chet Ramey
2024-04-27 16:56                                   ` Carl Edquist
2024-04-28 17:50                                     ` Chet Ramey
2024-04-08 16:21                   ` Chet Ramey
2024-04-12 16:49                     ` Carl Edquist
2024-04-16 15:48                       ` Chet Ramey
2024-04-20 23:11                         ` Carl Edquist
2024-04-22 16:12                           ` Chet Ramey
2024-04-17 14:37               ` Chet Ramey
2024-04-20 22:04                 ` Carl Edquist
2024-03-12  3:34           ` RFE: enable buffering on null-terminated data Zachary Santer
2024-03-14 14:15             ` Carl Edquist
2024-03-18  0:12               ` Zachary Santer
2024-03-19  5:24                 ` Kaz Kylheku
2024-03-19 12:50                   ` Zachary Santer
2024-03-20  8:55                     ` Carl Edquist
2024-04-19  0:16                       ` Modify buffering of standard streams via environment variables (not LD_PRELOAD)? Zachary Santer
2024-04-19  9:32                         ` Pádraig Brady
2024-04-19 11:36                           ` Zachary Santer
2024-04-19 12:26                             ` Pádraig Brady
2024-04-19 16:11                               ` Zachary Santer
2024-04-20 16:00                         ` Carl Edquist
2024-04-20 20:00                           ` Zachary Santer
2024-04-20 21:45                             ` Carl Edquist [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/libc/involved.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=6b1c46a9-95a6-9422-85aa-fa8035d6c429@cs.wisc.edu \
    --to=edquist@cs.wisc.edu \
    --cc=coreutils@gnu.org \
    --cc=libc-alpha@sourceware.org \
    --cc=p@draigbrady.com \
    --cc=zsanter@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).