Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> 
> branch: master
> commit b5bbb29634eca42582c01e2fd30c867d22f53298
> Merge: a047fb8494c 01fb898420f
> Author: Eli Zaretskii <eliz@gnu.org>
> Commit: Eli Zaretskii <eliz@gnu.org>
> 
>     Merge from origin/emacs-29
>     
>     01fb898420f Simplify after adding internal function to enter a labele...
>     b741dc7fcde Add internal function to enter a labeled restriction

Since this commit(?) I have the following issue:

When I open an org file (e.g. simple table) the table and formula is
not highlighted. C-u C-x = shows

There are text properties here:
  fontified            t

instead of

There are text properties here:
  face                 org-table
  fontified            t

To reproduce:

% ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
  --with-mailutils --with-native-compilation
or
% ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
  --with-mailutils --with-native-compilation=aot
% make install
% /usr/local/emacs30/bin/emacs -Q --file ~/table.org

The issue does not occur if I

- build without --with-native-compilation(=aot)
- run emacs from the source tree: ./src/emacs -Q --file ~/table.org

table.org is a simple file created with

- C-x C-f ~/table.org
- org-table-create RET
- C-x C-s

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Mon, 07 Aug 2023 14:07:43 +0200, "Herbert J. Skuhra" wrote:
> 
> On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> > 
> > branch: master
> > commit b5bbb29634eca42582c01e2fd30c867d22f53298
> > Merge: a047fb8494c 01fb898420f
> > Author: Eli Zaretskii <eliz@gnu.org>
> > Commit: Eli Zaretskii <eliz@gnu.org>
> > 
> >     Merge from origin/emacs-29
> >     
> >     01fb898420f Simplify after adding internal function to enter a labele...
> >     b741dc7fcde Add internal function to enter a labeled restriction
> 
> Since this commit(?) I have the following issue:
> 
> When I open an org file (e.g. simple table) the table and formula is
> not highlighted. C-u C-x = shows
> 
> There are text properties here:
>   fontified            t
> 
> instead of
> 
> There are text properties here:
>   face                 org-table
>   fontified            t
> 
> To reproduce:
> 
> % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
>   --with-mailutils --with-native-compilation
> or
> % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
>   --with-mailutils --with-native-compilation=aot
> % make install
> % /usr/local/emacs30/bin/emacs -Q --file ~/table.org
> 
> The issue does not occur if I
> 
> - build without --with-native-compilation(=aot)
> - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> 
> table.org is a simple file created with
> 
> - C-x C-f ~/table.org
> - org-table-create RET
> - C-x C-s

Well, if I close an (re)open the same or another org file, the file is
correctly displayed. Any ideas? The problem still occurs in master
(6ab90effbe5dbda01a934c3d8eaccf4aa835b60d).

Thanks.

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> Date: Tue, 22 Aug 2023 14:32:15 +0200
> From: "Herbert J. Skuhra" <herbert@gojira.at>
> 
> On Mon, 07 Aug 2023 14:07:43 +0200, "Herbert J. Skuhra" wrote:
> > 
> > On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> > > 
> > > branch: master
> > > commit b5bbb29634eca42582c01e2fd30c867d22f53298
> > > Merge: a047fb8494c 01fb898420f
> > > Author: Eli Zaretskii <eliz@gnu.org>
> > > Commit: Eli Zaretskii <eliz@gnu.org>
> > > 
> > >     Merge from origin/emacs-29
> > >     
> > >     01fb898420f Simplify after adding internal function to enter a labele...
> > >     b741dc7fcde Add internal function to enter a labeled restriction
> > 
> > Since this commit(?) I have the following issue:
> > 
> > When I open an org file (e.g. simple table) the table and formula is
> > not highlighted. C-u C-x = shows
> > 
> > There are text properties here:
> >   fontified            t
> > 
> > instead of
> > 
> > There are text properties here:
> >   face                 org-table
> >   fontified            t
> > 
> > To reproduce:
> > 
> > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> >   --with-mailutils --with-native-compilation
> > or
> > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> >   --with-mailutils --with-native-compilation=aot
> > % make install
> > % /usr/local/emacs30/bin/emacs -Q --file ~/table.org
> > 
> > The issue does not occur if I
> > 
> > - build without --with-native-compilation(=aot)
> > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> > 
> > table.org is a simple file created with
> > 
> > - C-x C-f ~/table.org
> > - org-table-create RET
> > - C-x C-s
> 
> Well, if I close an (re)open the same or another org file, the file is
> correctly displayed. Any ideas? The problem still occurs in master
> (6ab90effbe5dbda01a934c3d8eaccf4aa835b60d).

This is a very strange issue.  I cannot imagine how the factors you
mention, or the commit to which you point, could possibly affect this
behavior.  Andrea, any ideas?

Maybe you should raise this on the Org mailing list, as the issue
seems to be related to Org.

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Thu, 24 Aug 2023 08:02:29 +0200, Eli Zaretskii wrote:
> 
> > Date: Tue, 22 Aug 2023 14:32:15 +0200
> > From: "Herbert J. Skuhra"
> > 
> > On Mon, 07 Aug 2023 14:07:43 +0200, "Herbert J. Skuhra" wrote:
> > > 
> > > On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> > > > 
> > > > branch: master
> > > > commit b5bbb29634eca42582c01e2fd30c867d22f53298
> > > > Merge: a047fb8494c 01fb898420f
> > > > Author: Eli Zaretskii <eliz@gnu.org>
> > > > Commit: Eli Zaretskii <eliz@gnu.org>
> > > > 
> > > >     Merge from origin/emacs-29
> > > >     
> > > >     01fb898420f Simplify after adding internal function to enter a labele...
> > > >     b741dc7fcde Add internal function to enter a labeled restriction
> > > 
> > > Since this commit(?) I have the following issue:
> > > 
> > > When I open an org file (e.g. simple table) the table and formula is
> > > not highlighted. C-u C-x = shows
> > > 
> > > There are text properties here:
> > >   fontified            t
> > > 
> > > instead of
> > > 
> > > There are text properties here:
> > >   face                 org-table
> > >   fontified            t
> > > 
> > > To reproduce:
> > > 
> > > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> > >   --with-mailutils --with-native-compilation
> > > or
> > > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> > >   --with-mailutils --with-native-compilation=aot
> > > % make install
> > > % /usr/local/emacs30/bin/emacs -Q --file ~/table.org
> > > 
> > > The issue does not occur if I
> > > 
> > > - build without --with-native-compilation(=aot)
> > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> > > 
> > > table.org is a simple file created with
> > > 
> > > - C-x C-f ~/table.org
> > > - org-table-create RET
> > > - C-x C-s
> > 
> > Well, if I close an (re)open the same or another org file, the file is
> > correctly displayed. Any ideas? The problem still occurs in master
> > (6ab90effbe5dbda01a934c3d8eaccf4aa835b60d).
> 
> This is a very strange issue.  I cannot imagine how the factors you
> mention, or the commit to which you point, could possibly affect this
> behavior.  Andrea, any ideas?
> 
> Maybe you should raise this on the Org mailing list, as the issue
> seems to be related to Org.

Thanks.

My git skills are probably limited. A git checkout of

"d09de2f49d708f73e7397273a254a7775c294d05" is OK
"b5bbb29634eca42582c01e2fd30c867d22f53298" is not OK.

I'll try again.

1. Last time I checked issue does not occur on the emacs-29 branch
2. If I replace lisp/org in master with the files from the
emacs-29 branch the issue still occurs.

Two ways to work around the issue:

1.

'M-x load-library <RET> org-loaddefs' before opening an org-file 

2.

# cd /usr/local/emacs30/share/emacs/30.0.50/lisp/org/
# gzip -d org-loaddefs.el.gz
# cd /usr/local/emacs30/share/emacs/30.0.50/lisp/emacs-lisp
# gzip -d cl-loaddefs.el.gz 

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> Date: Thu, 24 Aug 2023 11:02:40 +0200
> From: "Herbert J. Skuhra" <herbert@gojira.at>
> Cc: Andrea Corallo <acorallo@gnu.org>,
> 	emacs-devel@gnu.org
> 
> On Thu, 24 Aug 2023 08:02:29 +0200, Eli Zaretskii wrote:
> > 
> > > Date: Tue, 22 Aug 2023 14:32:15 +0200
> > > From: "Herbert J. Skuhra"
> > > 
> > > On Mon, 07 Aug 2023 14:07:43 +0200, "Herbert J. Skuhra" wrote:
> > > > 
> > > > On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> > > > > 
> > > > > branch: master
> > > > > commit b5bbb29634eca42582c01e2fd30c867d22f53298
> > > > > Merge: a047fb8494c 01fb898420f
> > > > > Author: Eli Zaretskii <eliz@gnu.org>
> > > > > Commit: Eli Zaretskii <eliz@gnu.org>
> > > > > 
> > > > >     Merge from origin/emacs-29
> > > > >     
> > > > >     01fb898420f Simplify after adding internal function to enter a labele...
> > > > >     b741dc7fcde Add internal function to enter a labeled restriction
> > > > 
> > > > Since this commit(?) I have the following issue:
> > > > 
> > > > When I open an org file (e.g. simple table) the table and formula is
> > > > not highlighted. C-u C-x = shows
> > > > 
> > > > There are text properties here:
> > > >   fontified            t
> > > > 
> > > > instead of
> > > > 
> > > > There are text properties here:
> > > >   face                 org-table
> > > >   fontified            t
> > > > 
> > > > To reproduce:
> > > > 
> > > > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> > > >   --with-mailutils --with-native-compilation
> > > > or
> > > > % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
> > > >   --with-mailutils --with-native-compilation=aot
> > > > % make install
> > > > % /usr/local/emacs30/bin/emacs -Q --file ~/table.org
> > > > 
> > > > The issue does not occur if I
> > > > 
> > > > - build without --with-native-compilation(=aot)
> > > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> > > > 
> > > > table.org is a simple file created with
> > > > 
> > > > - C-x C-f ~/table.org
> > > > - org-table-create RET
> > > > - C-x C-s
> > > 
> > > Well, if I close an (re)open the same or another org file, the file is
> > > correctly displayed. Any ideas? The problem still occurs in master
> > > (6ab90effbe5dbda01a934c3d8eaccf4aa835b60d).
> > 
> > This is a very strange issue.  I cannot imagine how the factors you
> > mention, or the commit to which you point, could possibly affect this
> > behavior.  Andrea, any ideas?
> > 
> > Maybe you should raise this on the Org mailing list, as the issue
> > seems to be related to Org.
> 
> Thanks.
> 
> My git skills are probably limited. A git checkout of
> 
> "d09de2f49d708f73e7397273a254a7775c294d05" is OK
> "b5bbb29634eca42582c01e2fd30c867d22f53298" is not OK.
> 
> I'll try again.
> 
> 1. Last time I checked issue does not occur on the emacs-29 branch
> 2. If I replace lisp/org in master with the files from the
> emacs-29 branch the issue still occurs.
> 
> Two ways to work around the issue:
> 
> 1.
> 
> 'M-x load-library <RET> org-loaddefs' before opening an org-file 
> 
> 2.
> 
> # cd /usr/local/emacs30/share/emacs/30.0.50/lisp/org/
> # gzip -d org-loaddefs.el.gz
> # cd /usr/local/emacs30/share/emacs/30.0.50/lisp/emacs-lisp
> # gzip -d cl-loaddefs.el.gz 

Ihor, any suggestions or comments?

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Ihor Radchenko]

Eli Zaretskii <eliz@gnu.org> writes:

>> > > > The issue does not occur if I
>> > > >  ...
>> > > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
>
> Ihor, any suggestions or comments?

The fact that running from ./src/emacs does not trigger the problem
makes me suspect that it might be something to do with .el.gz files. At
least, it is the only problematic thing I heard about when doing make
install.

I'd first try make extraclean; make bootstrap; make install. Just to be
sure.

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Thu, 24 Aug 2023 13:17:49 +0200, Ihor Radchenko wrote:
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> > > > The issue does not occur if I
> >> > > >  ...
> >> > > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> >
> > Ihor, any suggestions or comments?
> 
> The fact that running from ./src/emacs does not trigger the problem
> makes me suspect that it might be something to do with .el.gz files. At
> least, it is the only problematic thing I heard about when doing make
> install.

If I gzip ./lisp/org/org-loaddefs.el and
./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.

> I'd first try make extraclean; make bootstrap; make install. Just to be
> sure.

During my tests I always run 'git clean -xfd' before building Emacs.

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> Date: Thu, 24 Aug 2023 16:10:21 +0200
> From: "Herbert J. Skuhra" <herbert@gojira.at>
> Cc: Eli Zaretskii <eliz@gnu.org>,
> 	acorallo@gnu.org,
> 	emacs-devel@gnu.org
> 
> On Thu, 24 Aug 2023 13:17:49 +0200, Ihor Radchenko wrote:
> > 
> > Eli Zaretskii <eliz@gnu.org> writes:
> > 
> > >> > > > The issue does not occur if I
> > >> > > >  ...
> > >> > > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> > >
> > > Ihor, any suggestions or comments?
> > 
> > The fact that running from ./src/emacs does not trigger the problem
> > makes me suspect that it might be something to do with .el.gz files. At
> > least, it is the only problematic thing I heard about when doing make
> > install.
> 
> If I gzip ./lisp/org/org-loaddefs.el and
> ./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.

Is your build with zlib or without it?  If without, then when you
compress *.el files, Emacs will be unable to use natively-compiled
*.eln files, because it needs to check their consistency with the
corresponding *.el files.  Don't you have warnings about that in the
*Warninngs* buffer?

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Thu, 24 Aug 2023 17:08:24 +0200, Eli Zaretskii wrote:
> 
> > Date: Thu, 24 Aug 2023 16:10:21 +0200
> > From: "Herbert J. Skuhra" <herbert@gojira.at>
> > Cc: Eli Zaretskii <eliz@gnu.org>,
> > 	acorallo@gnu.org,
> > 	emacs-devel@gnu.org
> > 
> > On Thu, 24 Aug 2023 13:17:49 +0200, Ihor Radchenko wrote:
> > > 
> > > Eli Zaretskii <eliz@gnu.org> writes:
> > > 
> > > >> > > > The issue does not occur if I
> > > >> > > >  ...
> > > >> > > > - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> > > >
> > > > Ihor, any suggestions or comments?
> > > 
> > > The fact that running from ./src/emacs does not trigger the problem
> > > makes me suspect that it might be something to do with .el.gz files. At
> > > least, it is the only problematic thing I heard about when doing make
> > > install.
> > 
> > If I gzip ./lisp/org/org-loaddefs.el and
> > ./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.
> 
> Is your build with zlib or without it?

With zlib.

% ldd /usr/local/emacs/bin/emacs |grep libz.so
	libz.so.1 => /usr/lib/libz.so.1 (0x00007fd7cd66b000)

% pacman -Qo /usr/lib/libz.so.1
/usr/lib/libz.so.1 is owned by zlib 1:1.3-1

> If without, then when you compress *.el files, Emacs will be unable
> to use natively-compiled *.eln files, because it needs to check
> their consistency with the corresponding *.el files.  Don't you have
> warnings about that in the *Warninngs* buffer?

When I open an org-file I only see the following lines in
"*Async-native-compile-log*" buffer:

Compiling /usr/local/emacs/share/emacs/30.0.50/lisp/org/org-loaddefs.el.gz...
uncompressing org-loaddefs.el.gz...
uncompressing org-loaddefs.el.gz...done
Compiling /usr/local/emacs/share/emacs/30.0.50/lisp/emacs-lisp/cl-loaddefs.el.gz...
uncompressing cl-loaddefs.el.gz...
uncompressing cl-loaddefs.el.gz...done
Compilation finished.

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> Date: Thu, 24 Aug 2023 18:29:25 +0200
> From: "Herbert J. Skuhra" <herbert@gojira.at>
> Cc: yantar92@posteo.net,
> 	acorallo@gnu.org,
> 	emacs-devel@gnu.org
> 
> On Thu, 24 Aug 2023 17:08:24 +0200, Eli Zaretskii wrote:
> > 
> > > If I gzip ./lisp/org/org-loaddefs.el and
> > > ./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.
> > 
> > Is your build with zlib or without it?
> 
> With zlib.
> 
> % ldd /usr/local/emacs/bin/emacs |grep libz.so
> 	libz.so.1 => /usr/lib/libz.so.1 (0x00007fd7cd66b000)

Then why would compressing a *.el file change the behavior?

Andrea, any ideas?  I'm out of ideas.

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Ihor Radchenko]

Eli Zaretskii <eliz@gnu.org> writes:

> Then why would compressing a *.el file change the behavior?

AFAIR, one of the ways .el.gz files got handled wrongly was an
assumption in the code that file extension must be .elc/.el/.eln. But
.el.gz files have .gz extension, which may be easily missed in the
logic.

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> From: Ihor Radchenko <yantar92@posteo.net>
> Cc: "Herbert J. Skuhra" <herbert@gojira.at>, acorallo@gnu.org,
>  emacs-devel@gnu.org
> Date: Fri, 25 Aug 2023 06:42:14 +0000
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > Then why would compressing a *.el file change the behavior?
> 
> AFAIR, one of the ways .el.gz files got handled wrongly was an
> assumption in the code that file extension must be .elc/.el/.eln. But
> .el.gz files have .gz extension, which may be easily missed in the
> logic.

Where (in what code) did you see this assumption?

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Ihor Radchenko]

Eli Zaretskii <eliz@gnu.org> writes:

>> AFAIR, one of the ways .el.gz files got handled wrongly was an
>> assumption in the code that file extension must be .elc/.el/.eln. But
>> .el.gz files have .gz extension, which may be easily missed in the
>> logic.
>
> Where (in what code) did you see this assumption?

`package--reload-previously-loaded'

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Andrea Corallo]

Eli Zaretskii <eliz@gnu.org> writes:

>> Date: Thu, 24 Aug 2023 18:29:25 +0200
>> From: "Herbert J. Skuhra" <herbert@gojira.at>
>> Cc: yantar92@posteo.net,
>> 	acorallo@gnu.org,
>> 	emacs-devel@gnu.org
>> 
>> On Thu, 24 Aug 2023 17:08:24 +0200, Eli Zaretskii wrote:
>> > 
>> > > If I gzip ./lisp/org/org-loaddefs.el and
>> > > ./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.
>> > 
>> > Is your build with zlib or without it?
>> 
>> With zlib.
>> 
>> % ldd /usr/local/emacs/bin/emacs |grep libz.so
>> 	libz.so.1 => /usr/lib/libz.so.1 (0x00007fd7cd66b000)
>
> Then why would compressing a *.el file change the behavior?
>
> Andrea, any ideas?  I'm out of ideas.

Hi Eli,

not so far must say.  I've hard time linking a native compiler related
issue with what I see in b5bbb29634e.

  Andrea

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> From: Ihor Radchenko <yantar92@posteo.net>
> Cc: herbert@gojira.at, acorallo@gnu.org, emacs-devel@gnu.org
> Date: Fri, 25 Aug 2023 07:42:16 +0000
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> AFAIR, one of the ways .el.gz files got handled wrongly was an
> >> assumption in the code that file extension must be .elc/.el/.eln. But
> >> .el.gz files have .gz extension, which may be easily missed in the
> >> logic.
> >
> > Where (in what code) did you see this assumption?
> 
> `package--reload-previously-loaded'

That probably needs to be fixed, but is package.el relevant to the
issue discussed here?  Org is part of Emacs, so I don't expect
package.el to be involved.

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Ihor Radchenko]

Eli Zaretskii <eliz@gnu.org> writes:

>> > Where (in what code) did you see this assumption?
>> 
>> `package--reload-previously-loaded'
>
> That probably needs to be fixed, but is package.el relevant to the
> issue discussed here?  Org is part of Emacs, so I don't expect
> package.el to be involved.

I do not think that this particular part of the code is responsible.
Just suspect that other places might have similar logic omission.
(Aside: that specific code _was_ a problem in Emacs 28, but I was unable
to create reproducer on master and did not report the problem. So, it
might be perfectly fine, or not. Someone more familiar with package.el
may need to take a look - I can report on emacs-devel or debbugs if you
think that it is worth it)

Also, I doubt that Org has anything to do with handling .el.gz files -
we do not fiddle with package loading, except in `org-assert-version'.
But there is nothing in there that seems relevant to the issue at hand.

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Fri, 25 Aug 2023 11:45:49 +0200, Andrea Corallo wrote:
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> Date: Thu, 24 Aug 2023 18:29:25 +0200
> >> From: "Herbert J. Skuhra" <herbert@gojira.at>
> >> Cc: yantar92@posteo.net,
> >> 	acorallo@gnu.org,
> >> 	emacs-devel@gnu.org
> >> 
> >> On Thu, 24 Aug 2023 17:08:24 +0200, Eli Zaretskii wrote:
> >> > 
> >> > > If I gzip ./lisp/org/org-loaddefs.el and
> >> > > ./lisp/emacs-lisp/cl-loaddefs.el and run ./src/emacs the problem also occurs.
> >> > 
> >> > Is your build with zlib or without it?
> >> 
> >> With zlib.
> >> 
> >> % ldd /usr/local/emacs/bin/emacs |grep libz.so
> >> 	libz.so.1 => /usr/lib/libz.so.1 (0x00007fd7cd66b000)
> >
> > Then why would compressing a *.el file change the behavior?
> >
> > Andrea, any ideas?  I'm out of ideas.
> 
> Hi Eli,
> 
> not so far must say.  I've hard time linking a native compiler related
> issue with what I see in b5bbb29634e.

Sorry, I guess that's the wrong commit. The problem also occurs with
a047fb8494c203bd8f416e7ee9f77ad8dcb03631.

Running 'git log --pretty=oneline' shows:

[...]
b5bbb29634eca42582c01e2fd30c867d22f53298 Merge from origin/emacs-29
d09de2f49d708f73e7397273a254a7775c294d05 Ignore quit while getting interprogram paste in kill-new
afdf54a5313ec898b6045e48eb6dce7d4abfa565 * lisp/progmodes/gdb-mi.el: Fix interactive invocation of 'gud-go'.
a047fb8494c203bd8f416e7ee9f77ad8dcb03631 Fix "Improve Python imports management commands"
[...]

Why do I get Emacs 29.0.92 when I checkout e.g. the commit before
b5bbb29634eca42582c01e2fd30c867d22f53298 (d09de2f49d7 and
afdf54a5313)? The version was bumped (from 29.0.50 to 30.0.50) in
commit 65456668ab1048fe8d77dc76a9e4d84481c218a5 (Mon Nov 28 20:03:07
2022 +0200).

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Eli Zaretskii]

> Date: Sat, 26 Aug 2023 11:05:42 +0200
> From: "Herbert J. Skuhra" <herbert@gojira.at>
> Cc: Eli Zaretskii <eliz@gnu.org>, yantar92@posteo.net, emacs-devel@gnu.org
> 
> > not so far must say.  I've hard time linking a native compiler related
> > issue with what I see in b5bbb29634e.
> 
> Sorry, I guess that's the wrong commit. The problem also occurs with
> a047fb8494c203bd8f416e7ee9f77ad8dcb03631.
> 
> Running 'git log --pretty=oneline' shows:
> 
> [...]
> b5bbb29634eca42582c01e2fd30c867d22f53298 Merge from origin/emacs-29
> d09de2f49d708f73e7397273a254a7775c294d05 Ignore quit while getting interprogram paste in kill-new
> afdf54a5313ec898b6045e48eb6dce7d4abfa565 * lisp/progmodes/gdb-mi.el: Fix interactive invocation of 'gud-go'.
> a047fb8494c203bd8f416e7ee9f77ad8dcb03631 Fix "Improve Python imports management commands"
> [...]
> 
> Why do I get Emacs 29.0.92 when I checkout e.g. the commit before
> b5bbb29634eca42582c01e2fd30c867d22f53298 (d09de2f49d7 and
> afdf54a5313)? The version was bumped (from 29.0.50 to 30.0.50) in
> commit 65456668ab1048fe8d77dc76a9e4d84481c218a5 (Mon Nov 28 20:03:07
> 2022 +0200).

Because you are bisecting inside commits made on the emacs-29 branch,
which were then merged to the master branch.

I suggest to bisect on the emacs-29 branch, not on master.

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Sat, 26 Aug 2023 11:41:08 +0200, Eli Zaretskii wrote:
> 
> > Date: Sat, 26 Aug 2023 11:05:42 +0200
> > From: "Herbert J. Skuhra" <herbert@gojira.at>
> > Cc: Eli Zaretskii <eliz@gnu.org>, yantar92@posteo.net, emacs-devel@gnu.org
> > 
> > > not so far must say.  I've hard time linking a native compiler related
> > > issue with what I see in b5bbb29634e.
> > 
> > Sorry, I guess that's the wrong commit. The problem also occurs with
> > a047fb8494c203bd8f416e7ee9f77ad8dcb03631.
> > 
> > Running 'git log --pretty=oneline' shows:
> > 
> > [...]
> > b5bbb29634eca42582c01e2fd30c867d22f53298 Merge from origin/emacs-29
> > d09de2f49d708f73e7397273a254a7775c294d05 Ignore quit while getting interprogram paste in kill-new
> > afdf54a5313ec898b6045e48eb6dce7d4abfa565 * lisp/progmodes/gdb-mi.el: Fix interactive invocation of 'gud-go'.
> > a047fb8494c203bd8f416e7ee9f77ad8dcb03631 Fix "Improve Python imports management commands"
> > [...]
> > 
> > Why do I get Emacs 29.0.92 when I checkout e.g. the commit before
> > b5bbb29634eca42582c01e2fd30c867d22f53298 (d09de2f49d7 and
> > afdf54a5313)? The version was bumped (from 29.0.50 to 30.0.50) in
> > commit 65456668ab1048fe8d77dc76a9e4d84481c218a5 (Mon Nov 28 20:03:07
> > 2022 +0200).
> 
> Because you are bisecting inside commits made on the emacs-29 branch,
> which were then merged to the master branch.
>
> I suggest to bisect on the emacs-29 branch, not on master.

The emacs-29 branch is OK.

I am now running master (d8b528c8967) with commit 4e8d579f3da reverted
and it seems to resolve the issue.

% git status
On branch master
Your branch is up to date with 'origin/master'.

You are currently reverting commit 4e8d579f3da.
  (all conflicts fixed: run "git revert --continue")
  (use "git revert --skip" to skip this patch)
  (use "git revert --abort" to cancel the revert operation)

Changes to be committed:
  (use "git restore --staged <file>..." to unstage)
	modified:   lisp/emacs-lisp/comp.el

--
Herbert

Re: master b5bbb29634e 1/5: Merge from origin/emacs-29

[Herbert J. Skuhra]

On Mon, 07 Aug 2023 14:07:43 +0200, "Herbert J. Skuhra" wrote:
> 
> On Sat, 15 Jul 2023 11:35:46 +0200, Eli Zaretskii  wrote:
> > 
> > branch: master
> > commit b5bbb29634eca42582c01e2fd30c867d22f53298
> > Merge: a047fb8494c 01fb898420f
> > Author: Eli Zaretskii <eliz@gnu.org>
> > Commit: Eli Zaretskii <eliz@gnu.org>
> > 
> >     Merge from origin/emacs-29
> >     
> >     01fb898420f Simplify after adding internal function to enter a labele...
> >     b741dc7fcde Add internal function to enter a labeled restriction
> 
> Since this commit(?) I have the following issue:
> 
> When I open an org file (e.g. simple table) the table and formula is
> not highlighted. C-u C-x = shows
> 
> There are text properties here:
>   fontified            t
> 
> instead of
> 
> There are text properties here:
>   face                 org-table
>   fontified            t
> 
> To reproduce:
> 
> % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
>   --with-mailutils --with-native-compilation
> or
> % ./configure --prefix=/usr/local/emacs30 --with-x-toolkit=lucid
>   --with-mailutils --with-native-compilation=aot
> % make install
> % /usr/local/emacs30/bin/emacs -Q --file ~/table.org
> 
> The issue does not occur if I
> 
> - build without --with-native-compilation(=aot)
> - run emacs from the source tree: ./src/emacs -Q --file ~/table.org
> 
> table.org is a simple file created with
> 
> - C-x C-f ~/table.org
> - org-table-create RET
> - C-x C-s

Nice, the issue is finally fixed... Bug #69431.

Thanks a lot!