* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
@ 2017-12-14 16:46 John Wiegley
2017-12-15 8:39 ` Michael Albinus
2017-12-16 1:10 ` Alan Third
0 siblings, 2 replies; 25+ messages in thread
From: John Wiegley @ 2017-12-14 16:46 UTC (permalink / raw)
To: 29712
[-- Attachment #1: Type: text/plain, Size: 1642 bytes --]
In order to get the ERT tests to pass on my system, I need to disable the
following three tests:
--- i/test/lisp/eshell/em-ls-tests.el
+++ w/test/lisp/eshell/em-ls-tests.el
@@ -77,6 +77,7 @@
(ert-deftest em-ls-test-bug27844 ()
"Test for https://debbugs.gnu.org/27844 ."
+ (skip-unless (not (eq system-type 'darwin)))
(let ((orig eshell-ls-use-in-dired)
(dired-use-ls-dired 'unspecified)
buf insert-directory-program)
--- i/test/lisp/net/tramp-tests.el
+++ w/test/lisp/net/tramp-tests.el
@@ -3636,6 +3636,7 @@ tramp--test-shell-command-to-string-asynchronously
;; This test is inspired by Bug#27009.
(ert-deftest tramp-test32-environment-variables-and-port-numbers ()
"Check that two connections with separate ports are different."
+ (skip-unless (not (eq system-type 'darwin)))
(skip-unless (tramp--test-enabled))
;; We test it only for the mock-up connection; otherwise there might
;; be problems with the used ports.
--- i/test/src/fileio-tests.el
+++ w/test/src/fileio-tests.el
@@ -48,7 +48,7 @@ fileio-tests--symlink-failure
;; Some Windows versions don't support symlinks, and those which do
;; will pop up UAC elevation prompts, so we disable this test on
;; MS-Windows.
- (skip-unless (not (eq system-type 'windows-nt)))
+ (skip-unless (not (memq system-type '(darwin windows-nt))))
(should (equal nil (fileio-tests--symlink-failure))))
(ert-deftest fileio-tests--directory-file-name ()
--
John Wiegley GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com 60E1 46C4 BD1A 7AC1 4BA2
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 658 bytes --]
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-14 16:46 bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin John Wiegley
@ 2017-12-15 8:39 ` Michael Albinus
2017-12-16 1:02 ` Alan Third
2017-12-16 1:10 ` Alan Third
1 sibling, 1 reply; 25+ messages in thread
From: Michael Albinus @ 2017-12-15 8:39 UTC (permalink / raw)
To: 29712
"John Wiegley" <johnw@gnu.org> writes:
> In order to get the ERT tests to pass on my system, I need to disable the
> following three tests:
> --- i/test/lisp/net/tramp-tests.el
> +++ w/test/lisp/net/tramp-tests.el
> @@ -3636,6 +3636,7 @@ tramp--test-shell-command-to-string-asynchronously
> ;; This test is inspired by Bug#27009.
> (ert-deftest tramp-test32-environment-variables-and-port-numbers ()
> "Check that two connections with separate ports are different."
> + (skip-unless (not (eq system-type 'darwin)))
> (skip-unless (tramp--test-enabled))
> ;; We test it only for the mock-up connection; otherwise there might
> ;; be problems with the used ports.
Committed to the emacs-26 branch.
Best regards, Michael.
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-15 8:39 ` Michael Albinus
@ 2017-12-16 1:02 ` Alan Third
2017-12-16 10:23 ` Michael Albinus
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 1:02 UTC (permalink / raw)
To: Michael Albinus; +Cc: 29712
On Fri, Dec 15, 2017 at 09:39:28AM +0100, Michael Albinus wrote:
> "John Wiegley" <johnw@gnu.org> writes:
>
> > In order to get the ERT tests to pass on my system, I need to disable the
> > following three tests:
>
> > --- i/test/lisp/net/tramp-tests.el
> > +++ w/test/lisp/net/tramp-tests.el
> > @@ -3636,6 +3636,7 @@ tramp--test-shell-command-to-string-asynchronously
> > ;; This test is inspired by Bug#27009.
> > (ert-deftest tramp-test32-environment-variables-and-port-numbers ()
> > "Check that two connections with separate ports are different."
> > + (skip-unless (not (eq system-type 'darwin)))
> > (skip-unless (tramp--test-enabled))
> > ;; We test it only for the mock-up connection; otherwise there might
> > ;; be problems with the used ports.
>
> Committed to the emacs-26 branch.
This particular failure happens because echo under /bin/sh on macOS
doesn’t recognise the -n flag:
(string-equal "11111" "-n 11111
")
It claims to be:
$ /bin/sh --version
GNU bash, version 3.2.57(1)-release (x86_64-apple-darwin16)
Copyright (C) 2007 Free Software Foundation, Inc.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-14 16:46 bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin John Wiegley
2017-12-15 8:39 ` Michael Albinus
@ 2017-12-16 1:10 ` Alan Third
2017-12-16 8:34 ` Eli Zaretskii
1 sibling, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 1:10 UTC (permalink / raw)
To: 29712
On Thu, Dec 14, 2017 at 08:46:02AM -0800, John Wiegley wrote:
> In order to get the ERT tests to pass on my system, I need to disable the
> following three tests:
>
> --- i/test/lisp/eshell/em-ls-tests.el
> +++ w/test/lisp/eshell/em-ls-tests.el
> @@ -77,6 +77,7 @@
>
> (ert-deftest em-ls-test-bug27844 ()
> "Test for https://debbugs.gnu.org/27844 ."
> + (skip-unless (not (eq system-type 'darwin)))
> (let ((orig eshell-ls-use-in-dired)
> (dired-use-ls-dired 'unspecified)
> buf insert-directory-program)
I can’t get this test to fail. Which version of macOS are you running?
> --- i/test/src/fileio-tests.el
> +++ w/test/src/fileio-tests.el
> @@ -48,7 +48,7 @@ fileio-tests--symlink-failure
> ;; Some Windows versions don't support symlinks, and those which do
> ;; will pop up UAC elevation prompts, so we disable this test on
> ;; MS-Windows.
> - (skip-unless (not (eq system-type 'windows-nt)))
> + (skip-unless (not (memq system-type '(darwin windows-nt))))
> (should (equal nil (fileio-tests--symlink-failure))))
>
> (ert-deftest fileio-tests--directory-file-name ()
This one fails on the carriage return character. It appears to be
getting converted to a newline character. I can’t work out why, but
I’m sure it’s Emacs that’s doing it.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 1:10 ` Alan Third
@ 2017-12-16 8:34 ` Eli Zaretskii
2017-12-16 9:46 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 8:34 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 01:10:05 +0000
> From: Alan Third <alan@idiocy.org>
>
> > --- i/test/src/fileio-tests.el
> > +++ w/test/src/fileio-tests.el
> > @@ -48,7 +48,7 @@ fileio-tests--symlink-failure
> > ;; Some Windows versions don't support symlinks, and those which do
> > ;; will pop up UAC elevation prompts, so we disable this test on
> > ;; MS-Windows.
> > - (skip-unless (not (eq system-type 'windows-nt)))
> > + (skip-unless (not (memq system-type '(darwin windows-nt))))
> > (should (equal nil (fileio-tests--symlink-failure))))
> >
> > (ert-deftest fileio-tests--directory-file-name ()
>
> This one fails on the carriage return character. It appears to be
> getting converted to a newline character. I can’t work out why, but
> I’m sure it’s Emacs that’s doing it.
Does it help to bind coding-system-for-write to 'no-conversion in that
test?
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 8:34 ` Eli Zaretskii
@ 2017-12-16 9:46 ` Alan Third
2017-12-16 10:35 ` Eli Zaretskii
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 9:46 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712
On Sat, Dec 16, 2017 at 10:34:02AM +0200, Eli Zaretskii wrote:
> > Date: Sat, 16 Dec 2017 01:10:05 +0000
> > From: Alan Third <alan@idiocy.org>
> >
> > > --- i/test/src/fileio-tests.el
> > > +++ w/test/src/fileio-tests.el
> > > @@ -48,7 +48,7 @@ fileio-tests--symlink-failure
> > > ;; Some Windows versions don't support symlinks, and those which do
> > > ;; will pop up UAC elevation prompts, so we disable this test on
> > > ;; MS-Windows.
> > > - (skip-unless (not (eq system-type 'windows-nt)))
> > > + (skip-unless (not (memq system-type '(darwin windows-nt))))
> > > (should (equal nil (fileio-tests--symlink-failure))))
> > >
> > > (ert-deftest fileio-tests--directory-file-name ()
> >
> > This one fails on the carriage return character. It appears to be
> > getting converted to a newline character. I can’t work out why, but
> > I’m sure it’s Emacs that’s doing it.
>
> Does it help to bind coding-system-for-write to 'no-conversion in that
> test?
No, it makes no difference.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 1:02 ` Alan Third
@ 2017-12-16 10:23 ` Michael Albinus
2017-12-16 13:47 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Michael Albinus @ 2017-12-16 10:23 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
[-- Attachment #1: Type: text/plain, Size: 214 bytes --]
Alan Third <alan@idiocy.org> writes:
> This particular failure happens because echo under /bin/sh on macOS
> doesn’t recognise the -n flag:
Does the appended patch fixes this?
Best regards, Michael.
[-- Attachment #2: Type: text/plain, Size: 927 bytes --]
diff --git a/test/lisp/net/tramp-tests.el b/test/lisp/net/tramp-tests.el
index c1577008d3..ff2ab31074 100644
--- a/test/lisp/net/tramp-tests.el
+++ b/test/lisp/net/tramp-tests.el
@@ -3638,7 +3638,7 @@ tramp--test-shell-command-to-string-asynchronously
"Check that two connections with separate ports are different."
(skip-unless (tramp--test-enabled))
;; Bug#29712.
- (skip-unless (not (eq system-type 'darwin)))
+ ;; (skip-unless (not (eq system-type 'darwin)))
;; We test it only for the mock-up connection; otherwise there might
;; be problems with the used ports.
(skip-unless (and (eq tramp-syntax 'default)
@@ -3663,7 +3663,7 @@ tramp--test-shell-command-to-string-asynchronously
(format "%s=%d" envvar port)
tramp-remote-process-environment)))
(should
- (string-equal
+ (string-match
(number-to-string port)
(shell-command-to-string (format "echo -n $%s" envvar))))))
^ permalink raw reply related [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 9:46 ` Alan Third
@ 2017-12-16 10:35 ` Eli Zaretskii
2017-12-16 13:42 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 10:35 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 09:46:43 +0000
> From: Alan Third <alan@idiocy.org>
> Cc: 29712@debbugs.gnu.org
>
> > > > --- i/test/src/fileio-tests.el
> > > > +++ w/test/src/fileio-tests.el
> > > > @@ -48,7 +48,7 @@ fileio-tests--symlink-failure
> > > > ;; Some Windows versions don't support symlinks, and those which do
> > > > ;; will pop up UAC elevation prompts, so we disable this test on
> > > > ;; MS-Windows.
> > > > - (skip-unless (not (eq system-type 'windows-nt)))
> > > > + (skip-unless (not (memq system-type '(darwin windows-nt))))
> > > > (should (equal nil (fileio-tests--symlink-failure))))
> > > >
> > > > (ert-deftest fileio-tests--directory-file-name ()
> > >
> > > This one fails on the carriage return character. It appears to be
> > > getting converted to a newline character. I can’t work out why, but
> > > I’m sure it’s Emacs that’s doing it.
> >
> > Does it help to bind coding-system-for-write to 'no-conversion in that
> > test?
>
> No, it makes no difference.
What is the value of file-name-coding-system and
default-file-name-coding-system?
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 10:35 ` Eli Zaretskii
@ 2017-12-16 13:42 ` Alan Third
2017-12-16 14:15 ` Eli Zaretskii
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 13:42 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712
On Sat, Dec 16, 2017 at 12:35:53PM +0200, Eli Zaretskii wrote:
> > Date: Sat, 16 Dec 2017 09:46:43 +0000
> > From: Alan Third <alan@idiocy.org>
> > Cc: 29712@debbugs.gnu.org
> >
> > > > > --- i/test/src/fileio-tests.el
> > > > > +++ w/test/src/fileio-tests.el
> > > > > @@ -48,7 +48,7 @@ fileio-tests--symlink-failure
> > > > > ;; Some Windows versions don't support symlinks, and those which do
> > > > > ;; will pop up UAC elevation prompts, so we disable this test on
> > > > > ;; MS-Windows.
> > > > > - (skip-unless (not (eq system-type 'windows-nt)))
> > > > > + (skip-unless (not (memq system-type '(darwin windows-nt))))
> > > > > (should (equal nil (fileio-tests--symlink-failure))))
> > > > >
> > > > > (ert-deftest fileio-tests--directory-file-name ()
> > > >
> > > > This one fails on the carriage return character. It appears to be
> > > > getting converted to a newline character. I can’t work out why, but
> > > > I’m sure it’s Emacs that’s doing it.
> > >
> > > Does it help to bind coding-system-for-write to 'no-conversion in that
> > > test?
> >
> > No, it makes no difference.
>
> What is the value of file-name-coding-system and
> default-file-name-coding-system?
file-name-coding-system: utf-8-hfs
default-file-name-coding-system: utf-8-unix
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 10:23 ` Michael Albinus
@ 2017-12-16 13:47 ` Alan Third
2017-12-16 15:55 ` Michael Albinus
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 13:47 UTC (permalink / raw)
To: Michael Albinus; +Cc: 29712
On Sat, Dec 16, 2017 at 11:23:39AM +0100, Michael Albinus wrote:
> Alan Third <alan@idiocy.org> writes:
>
> > This particular failure happens because echo under /bin/sh on macOS
> > doesn’t recognise the -n flag:
>
> Does the appended patch fixes this?
Yes, thank you.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 13:42 ` Alan Third
@ 2017-12-16 14:15 ` Eli Zaretskii
2017-12-16 15:07 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 14:15 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 13:42:10 +0000
> From: Alan Third <alan@idiocy.org>
> Cc: 29712@debbugs.gnu.org
>
> > What is the value of file-name-coding-system and
> > default-file-name-coding-system?
>
> file-name-coding-system: utf-8-hfs
> default-file-name-coding-system: utf-8-unix
Does it help to bind file-name-coding-system to utf-8-hfs-unix in that
test?
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 14:15 ` Eli Zaretskii
@ 2017-12-16 15:07 ` Alan Third
2017-12-16 16:19 ` Eli Zaretskii
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 15:07 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712
On Sat, Dec 16, 2017 at 04:15:06PM +0200, Eli Zaretskii wrote:
> > Date: Sat, 16 Dec 2017 13:42:10 +0000
> > From: Alan Third <alan@idiocy.org>
> > Cc: 29712@debbugs.gnu.org
> >
> > > What is the value of file-name-coding-system and
> > > default-file-name-coding-system?
> >
> > file-name-coding-system: utf-8-hfs
> > default-file-name-coding-system: utf-8-unix
>
> Does it help to bind file-name-coding-system to utf-8-hfs-unix in that
> test?
Yes, the test passes.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 13:47 ` Alan Third
@ 2017-12-16 15:55 ` Michael Albinus
0 siblings, 0 replies; 25+ messages in thread
From: Michael Albinus @ 2017-12-16 15:55 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
Alan Third <alan@idiocy.org> writes:
Hi Alan,
> On Sat, Dec 16, 2017 at 11:23:39AM +0100, Michael Albinus wrote:
>> Alan Third <alan@idiocy.org> writes:
>>
>> > This particular failure happens because echo under /bin/sh on macOS
>> > doesn’t recognise the -n flag:
>>
>> Does the appended patch fixes this?
>
> Yes, thank you.
Thanks for checking, I've pushed it to the emacs-26 branch.
Best regards, Michael.
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 15:07 ` Alan Third
@ 2017-12-16 16:19 ` Eli Zaretskii
2017-12-16 16:42 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 16:19 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 15:07:32 +0000
> From: Alan Third <alan@idiocy.org>
> Cc: 29712@debbugs.gnu.org
>
> > Does it help to bind file-name-coding-system to utf-8-hfs-unix in that
> > test?
>
> Yes, the test passes.
Then I think this is the right fix.
Maybe we should do that by default on Darwin, as it makes little sense
to convert EOLs in file names? Did the changeset in 83f0d60 on master
did that for Darwin as well?
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 16:19 ` Eli Zaretskii
@ 2017-12-16 16:42 ` Alan Third
2017-12-16 16:47 ` Eli Zaretskii
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 16:42 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712
On Sat, Dec 16, 2017 at 06:19:54PM +0200, Eli Zaretskii wrote:
> > Date: Sat, 16 Dec 2017 15:07:32 +0000
> > From: Alan Third <alan@idiocy.org>
> > Cc: 29712@debbugs.gnu.org
> >
> > > Does it help to bind file-name-coding-system to utf-8-hfs-unix in that
> > > test?
> >
> > Yes, the test passes.
>
> Then I think this is the right fix.
>
> Maybe we should do that by default on Darwin, as it makes little sense
> to convert EOLs in file names? Did the changeset in 83f0d60 on master
> did that for Darwin as well?
No, it’s set in ns-win.el:
modified lisp/term/ns-win.el
@@ -354,7 +354,7 @@ ns-delete-working-text
;; Used prior to Emacs 25.
(define-coding-system-alias 'utf-8-nfd 'utf-8-hfs)
- (set-file-name-coding-system 'utf-8-hfs))
+ (set-file-name-coding-system 'utf-8-hfs-unix))
;;;; Inter-app communications support.
I made the change and everything looks ok; the other tests still pass.
I don’t know anything about coding systems, though, so I could be
missing something.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 16:42 ` Alan Third
@ 2017-12-16 16:47 ` Eli Zaretskii
2017-12-16 17:01 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 16:47 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 16:42:18 +0000
> From: Alan Third <alan@idiocy.org>
> Cc: 29712@debbugs.gnu.org
>
> modified lisp/term/ns-win.el
> @@ -354,7 +354,7 @@ ns-delete-working-text
> ;; Used prior to Emacs 25.
> (define-coding-system-alias 'utf-8-nfd 'utf-8-hfs)
>
> - (set-file-name-coding-system 'utf-8-hfs))
> + (set-file-name-coding-system 'utf-8-hfs-unix))
>
> ;;;; Inter-app communications support.
Such a change is OK for master (assuming I understood you correctly,
and the current master doesn't have anything equivalent to the above).
On the release branch, I'd suggest to bind file-name-coding-system to
utf-8-hfs-unix, and tell in the commit log message not to merge to
master.
> I made the change and everything looks ok; the other tests still pass.
> I don’t know anything about coding systems, though, so I could be
> missing something.
I don't think you need to fear anything. It's the right thing to do
on systems that allow CR in file names.
Thanks.
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 16:47 ` Eli Zaretskii
@ 2017-12-16 17:01 ` Alan Third
2017-12-16 17:19 ` Eli Zaretskii
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2017-12-16 17:01 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712
On Sat, Dec 16, 2017 at 06:47:00PM +0200, Eli Zaretskii wrote:
> On the release branch, I'd suggest to bind file-name-coding-system to
> utf-8-hfs-unix, and tell in the commit log message not to merge to
> master.
Do you mean bind it in the test?
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 17:01 ` Alan Third
@ 2017-12-16 17:19 ` Eli Zaretskii
2018-01-08 21:16 ` John Wiegley
0 siblings, 1 reply; 25+ messages in thread
From: Eli Zaretskii @ 2017-12-16 17:19 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
> Date: Sat, 16 Dec 2017 17:01:30 +0000
> From: Alan Third <alan@idiocy.org>
> Cc: 29712@debbugs.gnu.org
>
> On Sat, Dec 16, 2017 at 06:47:00PM +0200, Eli Zaretskii wrote:
> > On the release branch, I'd suggest to bind file-name-coding-system to
> > utf-8-hfs-unix, and tell in the commit log message not to merge to
> > master.
>
> Do you mean bind it in the test?
Yes, for the release branch only.
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2017-12-16 17:19 ` Eli Zaretskii
@ 2018-01-08 21:16 ` John Wiegley
2018-01-10 0:12 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: John Wiegley @ 2018-01-08 21:16 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: 29712, Alan Third
>>>>> "EZ" == Eli Zaretskii <eliz@gnu.org> writes:
EZ> Yes, for the release branch only.
I'm still seeing one failure with the current emacs-26:
--8<---------------cut here---------------start------------->8---
ELC lisp/eshell/em-ls-tests.elc
In end of data:
lisp/eshell/em-ls-tests.el:99:1:Warning: the following functions are not known
to be defined: dired-toggle-marks, dired-get-marked-files
GEN lisp/eshell/em-ls-tests.log
Running 4 tests (2018-01-08 21:05:14+0000)
passed 1/4 em-ls-test-bug27631
passed 2/4 em-ls-test-bug27817
passed 3/4 em-ls-test-bug27843
Test em-ls-test-bug27844 backtrace:
signal(ert-test-failed (((should (looking-at "subr\\.el")) :form (lo
ert-fail(((should (looking-at "subr\\.el")) :form (looking-at "subr\
(if (unwind-protect (setq value-19 (apply fn-17 args-18)) (setq form
(let (form-description-21) (if (unwind-protect (setq value-19 (apply
(let ((value-19 'ert-form-evaluation-aborted-20)) (let (form-descrip
(let* ((fn-17 (function looking-at)) (args-18 (condition-case err (l
(progn (customize-set-variable 'eshell-ls-use-in-dired t) (setq buf
(unwind-protect (progn (customize-set-variable 'eshell-ls-use-in-dir
(let ((orig eshell-ls-use-in-dired) (dired-use-ls-dired 'unspecified
(lambda nil (let ((orig eshell-ls-use-in-dired) (dired-use-ls-dired
ert--run-test-internal(#s(ert--test-execution-info :test #s(ert-test
ert-run-test(#s(ert-test :name em-ls-test-bug27844 :documentation "T
ert-run-or-rerun-test(#s(ert--stats :selector (not (tag :expensive-t
ert-run-tests((not (tag :expensive-test)) #f(compiled-function (even
ert-run-tests-batch((not (tag :expensive-test)))
ert-run-tests-batch-and-exit((not (tag :expensive-test)))
eval((ert-run-tests-batch-and-exit '(not (tag :expensive-test))))
command-line-1(("-L" ":." "-l" "ert" "-l" "lisp/eshell/em-ls-tests.e
command-line()
normal-top-level()
Test em-ls-test-bug27844 condition:
(ert-test-failed
((should
(looking-at "subr\\.el"))
:form
(looking-at "subr\\.el")
:value nil))
FAILED 4/4 em-ls-test-bug27844
Ran 4 tests, 3 results as expected, 1 unexpected (2018-01-08 21:05:15+0000)
1 unexpected results:
FAILED em-ls-test-bug27844
make[3]: *** [Makefile:159: lisp/eshell/em-ls-tests.log] Error 1
--8<---------------cut here---------------end--------------->8---
--
John Wiegley GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com 60E1 46C4 BD1A 7AC1 4BA2
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-08 21:16 ` John Wiegley
@ 2018-01-10 0:12 ` Alan Third
2018-01-29 18:43 ` John Wiegley
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2018-01-10 0:12 UTC (permalink / raw)
To: John Wiegley; +Cc: 29712
On Mon, Jan 08, 2018 at 01:16:29PM -0800, John Wiegley wrote:
> >>>>> "EZ" == Eli Zaretskii <eliz@gnu.org> writes:
>
> EZ> Yes, for the release branch only.
>
> I'm still seeing one failure with the current emacs-26:
>
> --8<---------------cut here---------------start------------->8---
> ELC lisp/eshell/em-ls-tests.elc
I can’t replicate this failure at all:
Running 4 tests (2018-01-09 23:55:30+0000)
passed 1/4 em-ls-test-bug27631
passed 2/4 em-ls-test-bug27817
passed 3/4 em-ls-test-bug27843
passed 4/4 em-ls-test-bug27844
Ran 4 tests, 4 results as expected (2018-01-09 23:55:30+0000)
Does
(dired (expand-file-name "lisp/subr.el" source-directory))
work?
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-10 0:12 ` Alan Third
@ 2018-01-29 18:43 ` John Wiegley
2018-01-30 21:44 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: John Wiegley @ 2018-01-29 18:43 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
>>>>> "AT" == Alan Third <alan@idiocy.org> writes:
AT> Does
AT> (dired (expand-file-name "lisp/subr.el" source-directory))
AT> work?
No, I get:
dired-readin-insert: Directory
/private/tmp/nix-build-emacs-26.0.90.drv-0/emacs-8d45000/lisp/subr.el
inaccessible or nonexistent
--
John Wiegley GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com 60E1 46C4 BD1A 7AC1 4BA2
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-29 18:43 ` John Wiegley
@ 2018-01-30 21:44 ` Alan Third
2018-01-30 22:40 ` John Wiegley
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2018-01-30 21:44 UTC (permalink / raw)
To: John Wiegley; +Cc: 29712
On Mon, Jan 29, 2018 at 10:43:15AM -0800, John Wiegley wrote:
> >>>>> "AT" == Alan Third <alan@idiocy.org> writes:
>
> AT> Does
> AT> (dired (expand-file-name "lisp/subr.el" source-directory))
> AT> work?
>
> No, I get:
>
> dired-readin-insert: Directory
> /private/tmp/nix-build-emacs-26.0.90.drv-0/emacs-8d45000/lisp/subr.el
> inaccessible or nonexistent
How are you running the tests? Is it from this nix build location?
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-30 21:44 ` Alan Third
@ 2018-01-30 22:40 ` John Wiegley
2018-01-31 20:03 ` Alan Third
0 siblings, 1 reply; 25+ messages in thread
From: John Wiegley @ 2018-01-30 22:40 UTC (permalink / raw)
To: Alan Third; +Cc: 29712
>>>>> "AT" == Alan Third <alan@idiocy.org> writes:
AT> How are you running the tests? Is it from this nix build location?
I run the test using this expression:
emacs26 = with pkgs; stdenv.lib.overrideDerivation
(emacs25.override { srcRepo = true; }) (attrs: rec {
name = "emacs-${version}${versionModifier}";
version = "26.0";
versionModifier = ".90";
buildInputs = emacs25.buildInputs ++ [ git ];
patches = lib.optionals stdenv.isDarwin
[ ./emacs/patches/at-fdcwd.patch
./emacs/patches/emacs-26.patch ];
CFLAGS = "-Ofast -momit-leaf-frame-pointer";
src = fetchgit {
url = https://git.savannah.gnu.org/git/emacs.git;
rev = "8d4500087f547e203cfba03f61dcbe641bf650de";
sha256 = "1zk9xm01v4chnxf9ns9c3kx2jal3lj88hadv5vp0zb8xr9vz4f31";
};
postPatch = ''
rm -fr .git
'';
postInstall = ''
mkdir -p $out/share/emacs/site-lisp
cp ${./emacs/site-start.el} $out/share/emacs/site-lisp/site-start.el
$out/bin/emacs --batch -f batch-byte-compile $out/share/emacs/site-lisp/site-start.el
rm -rf $out/var
rm -rf $out/share/emacs/${version}/site-lisp
for srcdir in src lisp lwlib ; do
dstdir=$out/share/emacs/${version}/$srcdir
mkdir -p $dstdir
find $srcdir -name "*.[chm]" -exec cp {} $dstdir \;
cp $srcdir/TAGS $dstdir
echo '((nil . ((tags-file-name . "TAGS"))))' > $dstdir/.dir-locals.el
done
'' + lib.optionalString stdenv.isDarwin ''
mkdir -p $out/Applications
mv nextstep/Emacs.app $out/Applications
'';
});
emacs26debug = pkgs.stdenv.lib.overrideDerivation emacs26 (attrs: rec {
name = "emacs-26.0.90-debug";
doCheck = true;
CFLAGS = "-O0 -g3";
configureFlags = [ "--with-modules" ] ++
[ "--with-ns" "--disable-ns-self-contained"
"--enable-checking=yes,glyphs"
"--enable-check-lisp-object-type" ];
});
This is in my Nix overlays. I think just build nixpkgs.emacs26debug.
--
John Wiegley GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com 60E1 46C4 BD1A 7AC1 4BA2
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-30 22:40 ` John Wiegley
@ 2018-01-31 20:03 ` Alan Third
2018-02-01 18:05 ` Glenn Morris
0 siblings, 1 reply; 25+ messages in thread
From: Alan Third @ 2018-01-31 20:03 UTC (permalink / raw)
To: John Wiegley; +Cc: 29712
On Tue, Jan 30, 2018 at 02:40:08PM -0800, John Wiegley wrote:
> >>>>> "AT" == Alan Third <alan@idiocy.org> writes:
>
> AT> How are you running the tests? Is it from this nix build location?
>
> I run the test using this expression:
>
<snip>
>
> This is in my Nix overlays. I think just build nixpkgs.emacs26debug.
I don’t know anything about nix, but I suspect it’s running make check
from a different location than it built emacs from.
Basically this test just checks whether a file exists and it uses the
built‐in variable source-directory, which must be set at compile time,
as the root. Presumably nix builds in a temporary location which it
removes before running make check.
If that’s true I can think of two ways round it:
1. Create a temporary file and then check for its existence (e.g. in
/tmp).
2. Skip the test if the source directory doesn’t exist.
--
Alan Third
^ permalink raw reply [flat|nested] 25+ messages in thread
* bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin
2018-01-31 20:03 ` Alan Third
@ 2018-02-01 18:05 ` Glenn Morris
0 siblings, 0 replies; 25+ messages in thread
From: Glenn Morris @ 2018-02-01 18:05 UTC (permalink / raw)
To: Alan Third; +Cc: 29712, John Wiegley
Alan Third wrote:
> If that's true I can think of two ways round it:
>
> 1. Create a temporary file and then check for its existence (e.g. in
> /tmp).
> 2. Skip the test if the source directory doesn't exist.
3. Use a file that will exist, like doc-directory/DOC.
(Seems odd that none of the other tests using source-directory fail.)
^ permalink raw reply [flat|nested] 25+ messages in thread
end of thread, other threads:[~2018-02-01 18:05 UTC | newest]
Thread overview: 25+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-12-14 16:46 bug#29712: 26.0.90; Three emacs-26 elisp tests are failing on darwin John Wiegley
2017-12-15 8:39 ` Michael Albinus
2017-12-16 1:02 ` Alan Third
2017-12-16 10:23 ` Michael Albinus
2017-12-16 13:47 ` Alan Third
2017-12-16 15:55 ` Michael Albinus
2017-12-16 1:10 ` Alan Third
2017-12-16 8:34 ` Eli Zaretskii
2017-12-16 9:46 ` Alan Third
2017-12-16 10:35 ` Eli Zaretskii
2017-12-16 13:42 ` Alan Third
2017-12-16 14:15 ` Eli Zaretskii
2017-12-16 15:07 ` Alan Third
2017-12-16 16:19 ` Eli Zaretskii
2017-12-16 16:42 ` Alan Third
2017-12-16 16:47 ` Eli Zaretskii
2017-12-16 17:01 ` Alan Third
2017-12-16 17:19 ` Eli Zaretskii
2018-01-08 21:16 ` John Wiegley
2018-01-10 0:12 ` Alan Third
2018-01-29 18:43 ` John Wiegley
2018-01-30 21:44 ` Alan Third
2018-01-30 22:40 ` John Wiegley
2018-01-31 20:03 ` Alan Third
2018-02-01 18:05 ` Glenn Morris
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
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).