GNU bug report logs - #62556
29.0.60; [PATCH] Fix regression when calling 'eshell-command' with a pipeline in the background

Previous Next

Package: emacs;

Reported by: Jim Porter <jporterbugs <at> gmail.com>

Date: Fri, 31 Mar 2023 04:19:02 UTC

Severity: normal

Tags: patch

Found in version 29.0.60

Done: Jim Porter <jporterbugs <at> gmail.com>

Bug is archived. No further changes may be made.

Full log


View this message in rfc822 format

From: Jim Porter <jporterbugs <at> gmail.com>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: 62556 <at> debbugs.gnu.org
Subject: bug#62556: 29.0.60; [PATCH] Fix regression when calling 'eshell-command' with a pipeline in the background
Date: Sat, 1 Apr 2023 00:36:33 -0700
On 3/31/2023 11:35 PM, Eli Zaretskii wrote:
> I fixed the first 2, but I don't understand what goes wrong in the
> last one and why:
[snip]
>    Test eshell-test/subcommand-reset-in-pipeline condition:
>        (ert-test-failed
>         ((should
> 	 (eshell-command-result--equal command
> 				       (eshell-test-command-result command)
> 				       result))
> 	:form
> 	(eshell-command-result--equal "*cat $<echo $eshell-in-pipeline-p | echo> | *cat" nil "first")
> 	:value nil :explanation
> 	(nonequal-result
> 	 (command "*cat $<echo $eshell-in-pipeline-p | echo> | *cat")
> 	 (result nil)
> 	 (expected "first"))))
>       FAILED  16/16  eshell-test/subcommand-reset-in-pipeline (1.796875 sec) at lisp/eshell/eshell-tests.el:80

Lars reported a similar issue a while back too: 
<https://lists.gnu.org/archive/html/emacs-devel/2022-09/msg00314.html>. 
I *think* it's due to a race condition somewhere, but I've had quite a 
bit of difficulty fixing it. I have a couple ideas though, but they're 
probably too invasive for the release branch.

> As another annoyance, I get this prompt, and must type "yes RET":
> 
>       passed   1/16  eshell-test/command-running-p (0.203125 sec)
>       passed   2/16  eshell-test/eshell-command/background (0.015625 sec)
>    Buffer "*Eshell Async Command Output*" has a running process; kill it? (yes or no) yes
>       passed   3/16  eshell-test/eshell-command/background-pipeline (2.453125 sec)
> 
> Any suggestions?

Does this patch fix it?

----------------------------------------

diff --git a/test/lisp/eshell/eshell-tests.el 
b/test/lisp/eshell/eshell-tests.el
index bf7ec0389f0..fbff51a8873 100644
--- a/test/lisp/eshell/eshell-tests.el
+++ b/test/lisp/eshell/eshell-tests.el
@@ -136,6 +136,8 @@ eshell-test/eshell-command/background
       ;; buffer.
       (eshell-command "*echo hi &")
       (with-current-buffer "*Eshell Async Command Output*"
+        (while (get-buffer-process (current-buffer))
+          (accept-process-output))
         (goto-char (point-min))
         (should (looking-at "\\[echo\\(\\.exe\\)?\\(<[0-9]+>\\)?\\]"))))))

@@ -149,6 +151,8 @@ eshell-test/eshell-command/background-pipeline
       ;; XXX: As above, we can't write to the current buffer here.
       (eshell-command "*echo hi | *cat &")
       (with-current-buffer "*Eshell Async Command Output*"
+        (while (get-buffer-process (current-buffer))
+          (accept-process-output))
         (goto-char (point-min))
         (should (looking-at "\\[cat\\(\\.exe\\)?\\(<[0-9]+>\\)?\\]"))))))






This bug report was last modified 2 years and 48 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.