bug#22932: 25.0.92; X protocol error: BadRequest (invalid request code or no such operation) on protocol request 149

[Ken Raeburn]

In sync mode, there isn't a lot going on that the error could be caused by. By any chance could there have been an old display server in use that doesn't support all the modern RandR extensions?

John, could you try "xrandr -v" against the same display where you were running Emacs? If it says the server supports version 1.3, then the RRGetScreenResourcesCurrent request (the one in use at xfns.c:4266) should be supported, as I understand it.

If there are servers out there still supporting only 1.2, then we may need to check the server version before issuing that request.

Ken

On Tue, Mar 8, 2016 at 11:27 AM, Eli Zaretskii <eliz@gnu.org> wrote:

> From: John Eismeier <john.eismeier@emc.com>
> Cc: Eli Zaretskii <eliz@gnu.org>, 22932@debbugs.gnu.org
> Date: Tue, 08 Mar 2016 09:36:13 -0500
>
>
> I have recompiled with the arguments below, what this correct?
>
> CFLAGS='-O0 -g3' ./configure --prefix=/jenkins/emacs-25/build --with-x-toolkit=lucid --with-sound=no  --enable-checking='yes,glyphs' --enable-check-lisp-object-type
>
> launched with, was this correct?

Yes, thanks.

> /jenkins/emacs-25/build/bin/emacs -Q -xrm "emacs.synchronous: true"
> X protocol error: BadRequest (invalid request code or no such operation) on protocol request 149
> Fatal error 6: Aborted
>
>
> does this help ?

It does.

> Program received signal SIGABRT, Aborted.
> 0x00007f7fea31820b in raise (sig=6)
>     at ../nptl/sysdeps/unix/sysv/linux/pt-raise.c:37
> 37    ../nptl/sysdeps/unix/sysv/linux/pt-raise.c: No such file or directory.
> (gdb)
> #0  0x00007f7fea31820b in raise (sig=6)
>     at ../nptl/sysdeps/unix/sysv/linux/pt-raise.c:37
> #1  0x0000000000588c38 in terminate_due_to_signal (sig=6, backtrace_limit=40)
>     at emacs.c:380
> #2  0x00000000005b5efc in emacs_abort () at sysdep.c:2247
> #3  0x000000000046833f in redisplay_internal () at xdisp.c:13458
> #4  0x000000000046aad6 in redisplay_preserve_echo_area (from_where=13)
>     at xdisp.c:14250
> #5  0x0000000000695009 in Fdelete_process (process=...) at process.c:841
> #6  0x00000000006a41de in kill_buffer_processes (buffer=...) at process.c:7231
> #7  0x000000000058af10 in shut_down_emacs (sig=6, stuff=...) at emacs.c:1978
> #8  0x0000000000588be8 in terminate_due_to_signal (sig=6, backtrace_limit=40)
>     at emacs.c:364
> #9  0x00000000005b5efc in emacs_abort () at sysdep.c:2247
> #10 0x000000000045f65c in message3_nolog (m=...) at xdisp.c:10280
> #11 0x000000000045f46d in message3 (m=...) at xdisp.c:10240
> #12 0x00000000006329af in Fmessage (nargs=2, args=0x7fffa51e3680)
>     at editfns.c:3686
> #13 0x00000000006419ec in Ffuncall (nargs=3, args=0x7fffa51e3678)
>     at eval.c:2673
> #14 0x000000000068fc5c in exec_byte_code (bytestr=..., vector=...,
>     maxdepth=..., args_template=..., nargs=0, args=0x7fffa51e3e00)
>     at bytecode.c:880
> #15 0x000000000064253a in funcall_lambda (fun=..., nargs=0,
>     arg_vector=0x7fffa51e3e00) at eval.c:2855
> #16 0x0000000000641db8 in Ffuncall (nargs=1, args=0x7fffa51e3df8)
>     at eval.c:2742
> #17 0x000000000068fc5c in exec_byte_code (bytestr=..., vector=...,
>     maxdepth=..., args_template=..., nargs=2, args=0x7fffa51e4590)
>     at bytecode.c:880
> #18 0x000000000064253a in funcall_lambda (fun=..., nargs=2,
>     arg_vector=0x7fffa51e4580) at eval.c:2855
> #19 0x0000000000641db8 in Ffuncall (nargs=3, args=0x7fffa51e4578)
>     at eval.c:2742
> #20 0x000000000068fc5c in exec_byte_code (bytestr=..., vector=...,
>     maxdepth=..., args_template=..., nargs=0, args=0x0) at bytecode.c:880
> #21 0x00000000006429ab in funcall_lambda (fun=..., nargs=0,
>     arg_vector=0x32a78dd) at eval.c:2921
> #22 0x0000000000641db8 in Ffuncall (nargs=1, args=0x7fffa51e4ce8)
>     at eval.c:2742
> #23 0x000000000068fc5c in exec_byte_code (bytestr=..., vector=...,
>     maxdepth=..., args_template=..., nargs=0, args=0x0) at bytecode.c:880
> #24 0x00000000006429ab in funcall_lambda (fun=..., nargs=0,
>     arg_vector=0x32a2bbd) at eval.c:2921
> #25 0x0000000000641db8 in Ffuncall (nargs=1, args=0x7fffa51e5500)
>     at eval.c:2742
> #26 0x0000000000640e0f in funcall_nil (nargs=1, args=0x7fffa51e5500)
>     at eval.c:2332
> #27 0x000000000064130c in run_hook_with_args (nargs=1, args=0x7fffa51e5500,
>     funcall=0x640dec <funcall_nil>) at eval.c:2509
> #28 0x0000000000640e93 in Frun_hook_with_args (nargs=1, args=0x7fffa51e5500)
>     at eval.c:2374
> #29 0x0000000000641399 in run_hook (hook=...) at eval.c:2522
> #30 0x000000000058ad41 in Fkill_emacs (arg=...) at emacs.c:1898

This Emacs was killed 4 times over, oh boy...

> #31 0x0000000000557d31 in x_connection_closed (dpy=0xe6d880,
>     error_message=0x7fffa51e5760 "X protocol error: BadRequest (invalid request code or no such operation) on protocol request 149", ioerror=false)
>     at xterm.c:9484
> #32 0x0000000000557e67 in x_error_quitter (display=0xe6d880,
>     event=0x7fffa51e5910) at xterm.c:9553
> #33 0x0000000000557db4 in x_error_handler (display=0xe6d880,
>     event=0x7fffa51e5910) at xterm.c:9523
> #34 0x00007f7fee9a854b in _XError () from /usr/lib/x86_64-linux-gnu/libX11.so.6
> #35 0x00007f7fee9a55e7 in ?? () from /usr/lib/x86_64-linux-gnu/libX11.so.6
> #36 0x00007f7fee9a6687 in _XReply () from /usr/lib/x86_64-linux-gnu/libX11.so.6
> #37 0x00007f7fec2fd6f9 in ?? () from /usr/lib/x86_64-linux-gnu/libXrandr.so.2
> #38 0x0000000000568c78 in x_get_monitor_attributes_xrandr (dpyinfo=0xe79020)
>     at xfns.c:4266
> #39 0x000000000056905e in x_get_monitor_attributes (dpyinfo=0xe79020)
>     at xfns.c:4369
> #40 0x0000000000569127 in Fx_display_monitor_attributes_list (terminal=...)
>     at xfns.c:4515
> #41 0x000000000056c65f in compute_tip_xy (f=0x61b5260, parms=..., dx=...,
>     dy=..., width=463, height=22, root_x=0x7fffa51e5e80, root_y=0x7fffa51e5e84)
>     at xfns.c:5707
> #42 0x000000000056da31 in Fx_show_tip (string=..., frame=..., parms=...,
>     timeout=..., dx=..., dy=...) at xfns.c:6055

OK, the error happens when Emacs attempts to display a tooltip.  Is it
correct that you cannot show a tooltip at all without crashing Emacs?

As to why, we need help from X experts.  The immediate reason is the
call to x_get_monitor_attributes_xrandr.  Ken, can you take a look?
Why would that call trigger an X error?