emacs-26 build breaking on RHEL 6.6 GTK2

emacs-26 build breaking on RHEL 6.6 GTK2

[Kaushal Modi]

[-- Attachment #1: Type: text/plain, Size: 2889 bytes --]

Hello,

emacs-26 build is breaking after the last 2 commits ([1], [2]) related to
GTK. This is after "make bootstrap".

First I get this warning:

xfns.c: In function ‘Fx_display_monitor_attributes_list’:
xfns.c:4993:15: warning: implicit declaration of function
‘gdk_screen_get_monitor_scale_factor’ [-Wimplicit-function-declaration]
       scale = gdk_screen_get_monitor_scale_factor (gscreen, i);
               ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
xfns.c:4993:7: warning: nested extern declaration of
‘gdk_screen_get_monitor_scale_factor’ [-Wnested-externs]
       scale = gdk_screen_get_monitor_scale_factor (gscreen, i);
       ^~~~~
  CC       xselect.o

And then this error:

  CC       lastfile.o
make -C ../admin/charsets all
make[2]: Entering directory '/home/kmodi/downloads/git/emacs/admin/charsets'
make[2]: Nothing to be done for 'all'.
make[2]: Leaving directory '/home/kmodi/downloads/git/emacs/admin/charsets'
make -C ../admin/unidata charscript.el
make[2]: Entering directory '/home/kmodi/downloads/git/emacs/admin/unidata'
make[2]: Nothing to be done for 'charscript.el'.
make[2]: Leaving directory '/home/kmodi/downloads/git/emacs/admin/unidata'
  CCLD     temacs
xfns.o: In function `Fx_display_monitor_attributes_list':
xfns.c:(.text+0x9ab3): undefined reference to
`gdk_screen_get_monitor_scale_factor'
collect2: error: ld returned 1 exit status
make[1]: *** [Makefile:600: temacs] Error 1

Here is my Emacs build info from the last successful build (yesterday):

Emacs version: GNU Emacs 26.0.91 (build 1, x86_64-pc-linux-gnu, GTK+
Version 2.24.23)
 of 2018-01-23, built using commit 1fc98ed073a2d26ad941748ecb13aa6f47dd7b01.

./configure options:
  --with-modules --prefix=/home/kmodi/usr_local/apps/6/emacs/emacs-26
'--program-transform-name=s/^ctags$/ctags_emacs/'
--enable-checking=yes,glyphs --enable-check-lisp-object-type
'CPPFLAGS=-I/home/kmodi/usr_local/6/include -I/usr/include/freetype2
-I/usr/include' 'CFLAGS=-ggdb3 -O0' 'CXXFLAGS=-ggdb3 -O0'
'LDFLAGS=-L/home/kmodi/usr_local/6/lib -L/home/kmodi/usr_local/6/lib64
-ggdb3'
PKG_CONFIG_PATH=/home/kmodi/usr_local/6/lib/pkgconfig:/home/kmodi/usr_local/6/lib64/pkgconfig:/cad/adi/apps/gnu/linux/x86_64/6/lib/pkgconfig:/cad/adi/apps/gnu/linux/x86_64/6/lib64/pkgconfig:/home/kmodi/stowed/lib/pkgconfig:/usr/lib/pkgconfig:/usr/lib64/pkgconfig:/usr/share/pkgconfig:/lib/pkgconfig:/lib64/pkgconfig

Features:
  XPM JPEG TIFF GIF PNG RSVG IMAGEMAGICK SOUND GPM DBUS GSETTINGS NOTIFY
ACL LIBSELINUX GNUTLS LIBXML2 FREETYPE LIBOTF XFT ZLIB TOOLKIT_SCROLL_BARS
GTK2 X11 MODULES THREADS

Thanks.

[1]:
http://git.savannah.gnu.org/cgit/emacs.git/commit/?h=emacs-26&id=2892f05792e1f52b0966f92c5ed1aa75dcdd66a3
[2]:
http://git.savannah.gnu.org/cgit/emacs.git/commit/?h=emacs-26&id=59db8dca030ba6a34d143c3cc6715f02beba1068

-- 

Kaushal Modi

[-- Attachment #2: Type: text/html, Size: 3580 bytes --]

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Kaushal Modi]

[-- Attachment #1: Type: text/plain, Size: 408 bytes --]

On Wed, Jan 24, 2018 at 10:48 AM Kaushal Modi <kaushal.modi@gmail.com>
wrote:

> Hello,
>
> emacs-26 build is breaking after the last 2 commits ([1], [2]) related to
> GTK. This is after "make bootstrap".
>
> Here is my Emacs build info from the last successful build (yesterday):
>

This addition info might be useful:

"pkg-config --modversion gtk+-2.0" in the terminal prints "2.24.23".
-- 

Kaushal Modi

[-- Attachment #2: Type: text/html, Size: 887 bytes --]

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

Kaushal Modi <kaushal.modi@gmail.com> writes:

> On Wed, Jan 24, 2018 at 10:48 AM Kaushal Modi <kaushal.modi@gmail.com>
> wrote:
>
>> Hello,
>>
>> emacs-26 build is breaking after the last 2 commits ([1], [2]) related to
>> GTK. This is after "make bootstrap".
>>
>> Here is my Emacs build info from the last successful build (yesterday):
>>
>
> This addition info might be useful:
>
> "pkg-config --modversion gtk+-2.0" in the terminal prints "2.24.23".

This should be fixed in the emacs-26 branch already. Could you update
and try again?

Thanks

Robert

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Kaushal Modi]

[-- Attachment #1: Type: text/plain, Size: 341 bytes --]

On Tue, Feb 20, 2018 at 6:05 AM Robert Pluim <rpluim@gmail.com> wrote:

>
> This should be fixed in the emacs-26 branch already. Could you update
> and try again?
>

Hello,

I am not sure why I got this reply now; I already confirmed[1] the fix back
in Jan :)

[1]: http://lists.gnu.org/r/emacs-devel/2018-01/msg00691.html
-- 

Kaushal Modi

[-- Attachment #2: Type: text/html, Size: 803 bytes --]

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

Kaushal Modi <kaushal.modi@gmail.com> writes:

> On Tue, Feb 20, 2018 at 6:05 AM Robert Pluim <rpluim@gmail.com> wrote:
>
>>
>> This should be fixed in the emacs-26 branch already. Could you update
>> and try again?
>>
>
> Hello,
>
> I am not sure why I got this reply now; I already confirmed[1] the fix back
> in Jan :)

My apologies. Looks like the email I replied to had been stuck
somewhere in the depths of gmane for 3 weeks, and my memory is
terrible :-)

Thanks for (re-)confirming.

Robert

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

[-- Attachment #1: Type: text/plain, Size: 214 bytes --]

Kaushal Modi <kaushal.modi@gmail.com> writes:

> Hello,
>
> emacs-26 build is breaking after the last 2 commits ([1], [2]) related to
> GTK. This is after "make bootstrap".
>

Eek, GTK2. How about the following:



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Only-check-for-scaling-on-GTK3.patch --]
[-- Type: text/x-diff, Size: 1201 bytes --]

From 3d4e551f3481558680a57351c6abed8321578cbd Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Wed, 24 Jan 2018 17:02:50 +0100
Subject: [PATCH] Only check for scaling on GTK3

* src/xfns.c (Fx_display_monitor_attributes_list): Only check
for scaling on GTK3
---
 src/xfns.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/xfns.c b/src/xfns.c
index 4ea5113265..3dbd669edc 100644
--- a/src/xfns.c
+++ b/src/xfns.c
@@ -4940,7 +4940,7 @@ Internal use only, use `display-monitor-attributes-list' instead.  */)
       gint width_mm = -1, height_mm = -1;
       GdkRectangle rec, work;
       struct MonitorInfo *mi = &monitors[i];
-      int scale;
+      int scale = 1;
 
 #if GTK_CHECK_VERSION (3, 22, 0)
       GdkMonitor *monitor = gdk_display_get_monitor (gdpy, i);
@@ -4989,7 +4989,7 @@ Internal use only, use `display-monitor-attributes-list' instead.  */)
       /* GTK returns scaled sizes for the workareas.  */
 #if GTK_CHECK_VERSION (3, 22, 0)
       scale = gdk_monitor_get_scale_factor (monitor);
-#else
+#elif defined HAVE_GTK3
       scale = gdk_screen_get_monitor_scale_factor (gscreen, i);
 #endif
       rec.width *= scale;
-- 
2.16.0.rc1

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Kaushal Modi]

[-- Attachment #1: Type: text/plain, Size: 161 bytes --]

On Wed, Jan 24, 2018 at 11:12 AM Robert Pluim <rpluim@gmail.com> wrote:

> Eek, GTK2. How about the following:
>

That patch works, thank you.
-- 

Kaushal Modi

[-- Attachment #2: Type: text/html, Size: 615 bytes --]

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Glenn Morris]

And --with-x-toolkit=lucid (on rhel7):

  xterm.o: In function `x_set_offset':
  xterm.c:10313: undefined reference to `xg_get_scale'

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

[-- Attachment #1: Type: text/plain, Size: 298 bytes --]

Glenn Morris <rgm@gnu.org> writes:

> And --with-x-toolkit=lucid (on rhel7):
>
>   xterm.o: In function `x_set_offset':
>   xterm.c:10313: undefined reference to `xg_get_scale'

*sigh* too many combinations to test. [1]

Robert


Footnotes: 
[1]  Did we ever get anywhere on a CI setup for emacs?


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Fix-gtk-scaling-changes.patch --]
[-- Type: text/x-diff, Size: 692 bytes --]

From 3f0866efb84902b89b56e3897471e646fe11b8d9 Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Wed, 24 Jan 2018 19:01:52 +0100
Subject: [PATCH] Fix gtk scaling changes

* src/xterm.c (x_set_offset): Only call xg_get_scale if using GTK.
---
 src/xterm.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/xterm.c b/src/xterm.c
index f05ac6147e..364a8a8db0 100644
--- a/src/xterm.c
+++ b/src/xterm.c
@@ -10310,7 +10310,9 @@ void
 x_set_offset (struct frame *f, register int xoff, register int yoff, int change_gravity)
 {
   int modified_top, modified_left;
+#ifdef USE_GTK
   int scale = xg_get_scale (f);
+#endif
 
   if (change_gravity > 0)
     {
-- 
2.16.0.rc1

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Glenn Morris]

Robert Pluim wrote:

> [1]  Did we ever get anywhere on a CI setup for emacs?

"We" already have one. I find it irritating that it is ignored,
given the effort that has gone into it.

But yes, there is apparently a plan to make a new setup that will be
different, and perhaps it will (somehow) test all possible build
combinations.

I still think the onus is on people changing the display code to at
least check that it compiles with the common toolkit options,
particularly on the release branch at a time like this.

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

Glenn Morris <rgm@gnu.org> writes:

> Robert Pluim wrote:
>
>> [1]  Did we ever get anywhere on a CI setup for emacs?
>
> "We" already have one. I find it irritating that it is ignored,
> given the effort that has gone into it.
>

I didn't ignore it. I thought there wasn't one. I'll search for it.

> But yes, there is apparently a plan to make a new setup that will be
> different, and perhaps it will (somehow) test all possible build
> combinations.
>
> I still think the onus is on people changing the display code to at
> least check that it compiles with the common toolkit options,
> particularly on the release branch at a time like this.

Yes. Mea culpa.

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Noam Postavsky]

On Wed, Jan 24, 2018 at 1:39 PM, Robert Pluim <rpluim@gmail.com> wrote:
> Glenn Morris <rgm@gnu.org> writes:
>
>> Robert Pluim wrote:
>>
>>> [1]  Did we ever get anywhere on a CI setup for emacs?
>>
>> "We" already have one. I find it irritating that it is ignored,
>> given the effort that has gone into it.
>>
>
> I didn't ignore it. I thought there wasn't one. I'll search for it.

There are some links at https://www.emacswiki.org/emacs/Hydra (the
second section)

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[martin rudalics]

 > *sigh* too many combinations to test. [1]

One more:

../../src/xfns.c: In function ‘Fx_display_monitor_attributes_list’:
../../src/xfns.c:4993:7: warning: implicit declaration of function ‘gdk_screen_get_monitor_scale_factor’ [-Wimplicit-function-declaration]
../../src/xfns.c:4993:7: warning: nested extern declaration of ‘gdk_screen_get_monitor_scale_factor’ [-Wnested-externs]

Here with GTK+ Version 3.4.2.

Sorry.  I must have compiled the wrong branch when testing this.

martin

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

martin rudalics <rudalics@gmx.at> writes:

>> *sigh* too many combinations to test. [1]
>
> One more:
>
> ../../src/xfns.c: In function ‘Fx_display_monitor_attributes_list’:
> ../../src/xfns.c:4993:7: warning: implicit declaration of function ‘gdk_screen_get_monitor_scale_factor’ [-Wimplicit-function-declaration]
> ../../src/xfns.c:4993:7: warning: nested extern declaration of ‘gdk_screen_get_monitor_scale_factor’ [-Wnested-externs]
>
> Here with GTK+ Version 3.4.2.
>
> Sorry.  I must have compiled the wrong branch when testing this.

3.4.2? That's bizarre. Eli, I apologize for screwing this up. Probably
best to revert until I've improved my testing coverage, and then try
again for emacs-27.

Robert

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[martin rudalics]

> 3.4.2? That's bizarre.

It's defined only since GDK 3.10.

martin

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

martin rudalics <rudalics@gmx.at> writes:

>> 3.4.2? That's bizarre.
>
> It's defined only since GDK 3.10.

Which I didn't find in the GTK doc. Definitely back to the drawing
board on these changes.

Robert

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[martin rudalics]

 > Definitely back to the drawing
 > board on these changes.

Don't worry.  They build all now here.

It would be nice if someone tried with a GTK version 3.10..3.21.

Sorry for the inconveniences, martin

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Robert Pluim]

martin rudalics <rudalics@gmx.at> writes:

>> Definitely back to the drawing
>> board on these changes.
>
> Don't worry.  They build all now here.
>
> It would be nice if someone tried with a GTK version 3.10..3.21.
>

I'd tried on 3.22, I don't think I have a 3.10 lying around.

Robert

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Date: Wed, 24 Jan 2018 19:52:25 +0100
> Cc: emacs-devel@gnu.org
> 
> martin rudalics <rudalics@gmx.at> writes:
> 
> >> 3.4.2? That's bizarre.
> >
> > It's defined only since GDK 3.10.
> 
> Which I didn't find in the GTK doc. Definitely back to the drawing
> board on these changes.

Why not simply change the prerequisite for the GTK version number, and
leave the rest of the patch alone?  IOW, instead of checking HAVE_GTK3
test GTK_CHECK_VERSION (3, 10, 0).  Would that work?

Re: emacs-26 build breaking on RHEL 6.6 GTK2

[martin rudalics]

> Why not simply change the prerequisite for the GTK version number, and
> leave the rest of the patch alone?  IOW, instead of checking HAVE_GTK3
> test GTK_CHECK_VERSION (3, 10, 0).  Would that work?

Hopefully, that's what I've pushed.

martin