* bug#5238: 23.1.50; gzip: stdout: Broken pipe
@ 2009-12-17 0:49 jidanni
2009-12-17 23:38 ` Juri Linkov
` (2 more replies)
0 siblings, 3 replies; 9+ messages in thread
From: jidanni @ 2009-12-17 0:49 UTC (permalink / raw)
To: emacs-pretest-bug; +Cc: rfrancoise
Gentlemen, both
$ emacs -Q -eval '(compile "locale -m")'
$ emacs -Q -nw -eval '(compile "locale -m")'
work, however the former has lots of "gzip: stdout: Broken pipe",
the latter doesn't.
$ emacs -Q -eval '(compile "strace locale -m")'
shows it happens around
clone(child_stack=0, flags=...
In GNU Emacs 23.1.50.1 (i486-pc-linux-gnu, GTK+ Version 2.18.3)
of 2009-11-29 on elegiac, modified by Debian
(emacs-snapshot package, version 1:20091128-1)
I notice emacs 23.1.1 is also affected.
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2009-12-17 0:49 bug#5238: 23.1.50; gzip: stdout: Broken pipe jidanni
@ 2009-12-17 23:38 ` Juri Linkov
2010-11-13 22:29 ` Chong Yidong
2010-11-14 1:10 ` jidanni
2 siblings, 0 replies; 9+ messages in thread
From: Juri Linkov @ 2009-12-17 23:38 UTC (permalink / raw)
To: jidanni; +Cc: 5238
> Gentlemen, both
> $ emacs -Q -eval '(compile "locale -m")'
> $ emacs -Q -nw -eval '(compile "locale -m")'
> work, however the former has lots of "gzip: stdout: Broken pipe",
> the latter doesn't.
> $ emacs -Q -eval '(compile "strace locale -m")'
> shows it happens around
> clone(child_stack=0, flags=...
A simpler test case:
1. emacs -Q
2. M-! locale -m RET
Fails for Emacs 23, but not for Emacs 22 or Emacs 23 run with -nw.
--
Juri Linkov
http://www.jurta.org/emacs/
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2009-12-17 0:49 bug#5238: 23.1.50; gzip: stdout: Broken pipe jidanni
2009-12-17 23:38 ` Juri Linkov
@ 2010-11-13 22:29 ` Chong Yidong
2010-11-14 22:00 ` Juri Linkov
2010-11-14 1:10 ` jidanni
2 siblings, 1 reply; 9+ messages in thread
From: Chong Yidong @ 2010-11-13 22:29 UTC (permalink / raw)
To: Juri Linkov; +Cc: 5238, jidanni
Juri Linkov <juri@jurta.org> wrote:
> > Gentlemen, both
> > $ emacs -Q -eval '(compile "locale -m")'
> > $ emacs -Q -nw -eval '(compile "locale -m")'
> > work, however the former has lots of "gzip: stdout: Broken pipe",
> > the latter doesn't.
> > $ emacs -Q -eval '(compile "strace locale -m")'
> > shows it happens around
> > clone(child_stack=0, flags=...
>
> A simpler test case:
>
> 1. emacs -Q
> 2. M-! locale -m RET
>
> Fails for Emacs 23, but not for Emacs 22 or Emacs 23 run with -nw.
I can't reproduce this anymore, so presumably this old bug is fixed.
Juri, could you confirm?
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2010-11-13 22:29 ` Chong Yidong
@ 2010-11-14 22:00 ` Juri Linkov
2011-02-11 14:30 ` Jan Djärv
0 siblings, 1 reply; 9+ messages in thread
From: Juri Linkov @ 2010-11-14 22:00 UTC (permalink / raw)
To: Chong Yidong; +Cc: 5238, jidanni
>> 1. emacs -Q
>> 2. M-! locale -m RET
>>
>> Fails for Emacs 23, but not for Emacs 22 or Emacs 23 run with -nw.
>
> I can't reproduce this anymore, so presumably this old bug is fixed.
> Juri, could you confirm?
I still can reproduce this bug on the newest Ubuntu version 10.10 and
GNU Emacs 24.0.50.1 (x86_64-pc-linux-gnu, GTK+ Version 2.22.0) of 2010-10-22
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2010-11-14 22:00 ` Juri Linkov
@ 2011-02-11 14:30 ` Jan Djärv
2011-02-12 16:02 ` Juri Linkov
2011-02-12 17:18 ` Andreas Schwab
0 siblings, 2 replies; 9+ messages in thread
From: Jan Djärv @ 2011-02-11 14:30 UTC (permalink / raw)
To: Juri Linkov; +Cc: Chong Yidong, 5238, jidanni
2010-11-14 23:00, Juri Linkov skrev:
>>> 1. emacs -Q
>>> 2. M-! locale -m RET
>>>
>>> Fails for Emacs 23, but not for Emacs 22 or Emacs 23 run with -nw.
>>
>> I can't reproduce this anymore, so presumably this old bug is fixed.
>> Juri, could you confirm?
>
> I still can reproduce this bug on the newest Ubuntu version 10.10 and
> GNU Emacs 24.0.50.1 (x86_64-pc-linux-gnu, GTK+ Version 2.22.0) of 2010-10-22
>
>
Your shell ignores SIGPIPE for interactive sessions. Gzip checks if SIGPIPE
is ignored, and if it is not, installs a signal handler that prints "gzip:
stdout: Broken pipe". Try
M-! bash -i -c 'locale -m' RET
That gives some job and terminal control errors from the shell, but no gzip
messages. I don't know if ignoring SIGPIPE for subprocesses is the right
thing for Emacs to do.
Jan D.
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2011-02-11 14:30 ` Jan Djärv
@ 2011-02-12 16:02 ` Juri Linkov
2011-02-12 17:18 ` Andreas Schwab
1 sibling, 0 replies; 9+ messages in thread
From: Juri Linkov @ 2011-02-12 16:02 UTC (permalink / raw)
To: Jan Djärv; +Cc: Chong Yidong, 5238, jidanni
> Your shell ignores SIGPIPE for interactive sessions. Gzip checks if
> SIGPIPE is ignored, and if it is not, installs a signal handler that prints
> "gzip: stdout: Broken pipe". Try
> M-! bash -i -c 'locale -m' RET
>
> That gives some job and terminal control errors from the shell, but no gzip
> messages. I don't know if ignoring SIGPIPE for subprocesses is the right
> thing for Emacs to do.
Since `M-x shell RET locale -m RET' doesn't display broken pipe errors,
then perhaps `shell-command' should do the same and ignore SIGPIPE,
because the output of `shell-command' and `shell' is expected to be the same.
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2011-02-11 14:30 ` Jan Djärv
2011-02-12 16:02 ` Juri Linkov
@ 2011-02-12 17:18 ` Andreas Schwab
2011-02-12 18:55 ` Andreas Schwab
1 sibling, 1 reply; 9+ messages in thread
From: Andreas Schwab @ 2011-02-12 17:18 UTC (permalink / raw)
To: Jan Djärv; +Cc: Chong Yidong, 5238, jidanni
Jan Djärv <jan.h.d@swipnet.se> writes:
> Gzip checks if SIGPIPE is ignored, and if it is not, installs a signal
> handler that prints "gzip: stdout: Broken pipe".
No it doesn't. The error only happens because gzip gets EPIPE while
writing to stdout, which can only happen when SIGPIPE is ignored, which
in fact Emacs does. It's the fault of libgtk.
> I don't know if ignoring SIGPIPE for subprocesses is the right thing
> for Emacs to do.
It isn't.
Andreas.
--
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756 01D3 44D5 214B 8276 4ED5
"And now for something completely different."
^ permalink raw reply [flat|nested] 9+ messages in thread
* bug#5238: 23.1.50; gzip: stdout: Broken pipe
2009-12-17 0:49 bug#5238: 23.1.50; gzip: stdout: Broken pipe jidanni
2009-12-17 23:38 ` Juri Linkov
2010-11-13 22:29 ` Chong Yidong
@ 2010-11-14 1:10 ` jidanni
2 siblings, 0 replies; 9+ messages in thread
From: jidanni @ 2010-11-14 1:10 UTC (permalink / raw)
To: cyd; +Cc: 5238
I still see it in
$ emacs-snapshot --version
GNU Emacs 24.0.50.1
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2011-02-12 18:55 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-12-17 0:49 bug#5238: 23.1.50; gzip: stdout: Broken pipe jidanni
2009-12-17 23:38 ` Juri Linkov
2010-11-13 22:29 ` Chong Yidong
2010-11-14 22:00 ` Juri Linkov
2011-02-11 14:30 ` Jan Djärv
2011-02-12 16:02 ` Juri Linkov
2011-02-12 17:18 ` Andreas Schwab
2011-02-12 18:55 ` Andreas Schwab
2010-11-14 1:10 ` jidanni
Code repositories for project(s) associated with this external index
https://git.savannah.gnu.org/cgit/emacs.git
https://git.savannah.gnu.org/cgit/emacs/org-mode.git
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.