python-complete-symbol hangs

python-complete-symbol hangs

[Michael Droettboom]

[-- Attachment #1: Type: text/plain, Size: 331 bytes --]

python-complete-symbol works by calling "complete()" in etc/emacs.py.
If that Python function fails (i.e. raises an exception), emacs will
hang waiting for a response, with no recourse but to kill the emacs
process.

The attached patch fixes this by always outputting a response.

-- 
Michael Droettboom
http://www.droettboom.com/

[-- Attachment #2: emacs.py.patch --]
[-- Type: application/octet-stream, Size: 671 bytes --]

*** emacs.py	2007-09-07 13:38:29.000000000 -0400
--- /home/mdroe/builds/emacs-22.1/etc/emacs.py	2007-05-14 10:56:28.000000000 -0400
***************
*** 176,184 ****
  		    names.add('__class__')
  		    names.union_update (class_members (object))
  	    except: names = all_names (dict)
!     except Exception, e:
!         print '_emacs_out ()'
!         return []
      l = len(name)
      print '_emacs_out (',
      for n in names:
--- 176,182 ----
  		    names.add('__class__')
  		    names.union_update (class_members (object))
  	    except: names = all_names (dict)
!     except: return []
      l = len(name)
      print '_emacs_out (',
      for n in names:

[-- Attachment #3: Type: text/plain, Size: 149 bytes --]

_______________________________________________
bug-gnu-emacs mailing list
bug-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnu-emacs

Re: python-complete-symbol hangs

[Michael Droettboom]

[-- Attachment #1: Type: text/plain, Size: 520 bytes --]

I'm sorry.  The patch in the last e-mail was backwards.

On 9/7/07, Michael Droettboom <mike@droettboom.com> wrote:
> python-complete-symbol works by calling "complete()" in etc/emacs.py.
> If that Python function fails (i.e. raises an exception), emacs will
> hang waiting for a response, with no recourse but to kill the emacs
> process.
>
> The attached patch fixes this by always outputting a response.
>
> --
> Michael Droettboom
> http://www.droettboom.com/
>
>


-- 
Michael Droettboom
http://www.droettboom.com/

[-- Attachment #2: emacs.py.patch --]
[-- Type: application/octet-stream, Size: 671 bytes --]

*** /home/mdroe/builds/emacs-22.1/etc/emacs.py	2007-05-14 10:56:28.000000000 -0400
--- emacs.py	2007-09-07 13:38:29.000000000 -0400
***************
*** 176,182 ****
  		    names.add('__class__')
  		    names.union_update (class_members (object))
  	    except: names = all_names (dict)
!     except: return []
      l = len(name)
      print '_emacs_out (',
      for n in names:
--- 176,184 ----
  		    names.add('__class__')
  		    names.union_update (class_members (object))
  	    except: names = all_names (dict)
!     except Exception, e:
!         print '_emacs_out ()'
!         return []
      l = len(name)
      print '_emacs_out (',
      for n in names:

[-- Attachment #3: Type: text/plain, Size: 149 bytes --]

_______________________________________________
bug-gnu-emacs mailing list
bug-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnu-emacs

Re: python-complete-symbol hangs

[Michael Droettboom]

The emacs version is 22.5.1

On 9/7/07, Michael Droettboom <mike@droettboom.com> wrote:
> I'm sorry.  The patch in the last e-mail was backwards.
>
> On 9/7/07, Michael Droettboom <mike@droettboom.com> wrote:
> > python-complete-symbol works by calling "complete()" in etc/emacs.py.
> > If that Python function fails (i.e. raises an exception), emacs will
> > hang waiting for a response, with no recourse but to kill the emacs
> > process.
> >
> > The attached patch fixes this by always outputting a response.
> >
> > --
> > Michael Droettboom
> > http://www.droettboom.com/
> >
> >
>
>
> --
> Michael Droettboom
> http://www.droettboom.com/
>
>


-- 
Michael Droettboom
http://www.droettboom.com/

Re: python-complete-symbol hangs

[Dave Love]

"Michael Droettboom" <mike@droettboom.com> writes:

> python-complete-symbol works by calling "complete()" in etc/emacs.py.
> If that Python function fails (i.e. raises an exception), emacs will
> hang waiting for a response, with no recourse but to kill the emacs
> process.

Can you supply a test case?  First of all, how does it raise an
exception?

If you have to kill it like that, it's surely a basic Emacs bug,
regardless of any python.el bugs.

> The attached patch fixes this by always outputting a response.

That shouldn't be necessary.  I think there were several things broken
or not fixed in the one shipped with Emacs 22, and that one sounds
familiar.  Sorry I don't have a working version since I didn't keep up
with the incompatibilities introduced by Emacs 22.

http://www.loveshack.ukfsn.org/emacs/python-21.el behaves as I expect
-- providing a null completion list -- if emacs.complete raises an
exception.

Re: python-complete-symbol hangs

[Michael Droettboom]

[-- Attachment #1: Type: text/plain, Size: 1952 bytes --]

On 9/8/07, Dave Love <fx@gnu.org> wrote:
> "Michael Droettboom" <mike@droettboom.com> writes:
>
> > python-complete-symbol works by calling "complete()" in etc/emacs.py.
> > If that Python function fails (i.e. raises an exception), emacs will
> > hang waiting for a response, with no recourse but to kill the emacs
> > process.
>
> Can you supply a test case?

See the attached.  Move cursor to bottom of file, then run
"python-update-imports", then "python-complete-symbol".

This causes a hang for me on

   22.1.5 from source on Linux
   current cvs head from source on Linux
   22.1.1 as CarbonEmacs Package on OS-X

All of these cases are using Python 2.5 (in case that makes the
difference here).

>  First of all, how does it raise an
> exception?

It raises a SyntaxError exception because the imports built by
"python-update-imports" are invalid (see my other bug report).

> If you have to kill it like that, it's surely a basic Emacs bug,
> regardless of any python.el bugs.

The entire emacs process becomes unresponsive to keystrokes and mouse
events.  Is there something else I should try?

> > The attached patch fixes this by always outputting a response.
>
> That shouldn't be necessary.  I think there were several things broken
> or not fixed in the one shipped with Emacs 22, and that one sounds
> familiar.  Sorry I don't have a working version since I didn't keep up
> with the incompatibilities introduced by Emacs 22.
>
> http://www.loveshack.ukfsn.org/emacs/python-21.el behaves as I expect
> -- providing a null completion list -- if emacs.complete raises an
> exception.

Indeed -- that works for me also -- without making any changes to
emacs.py.  That is a workable solution for me for now.

Perhaps some of your changes (the diff between python-21.el and emacs
22's python.el is quite large) should be merged into the main Emacs
CVS.

Thanks for your help.

Cheers,
Mike

-- 
Michael Droettboom
http://www.droettboom.com/

[-- Attachment #2: test.py --]
[-- Type: application/octet-stream, Size: 48 bytes --]

from sys import argv \
    exit
import os

os.p

[-- Attachment #3: Type: text/plain, Size: 149 bytes --]

_______________________________________________
bug-gnu-emacs mailing list
bug-gnu-emacs@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnu-emacs

Re: python-complete-symbol hangs

[Michael Droettboom]

[-- Attachment #1: Type: text/plain, Size: 1110 bytes --]

I have attached a gdb log, with my comments in [[ ]], along with a
simple Python file that triggers this bug.

I have also attached the informational output of "emacs-report-bug",
which I should have provided initially.  (Sorry about that).  The
recent keystrokes are probably irrelevant there, since they are from a
different run (I couldn't wake emacs back up after it hung.)

Thanks for your help with this, and most of all, thanks for emacs!

Cheers,
Mike

On 9/8/07, Richard Stallman <rms@gnu.org> wrote:
>     The entire emacs process becomes unresponsive to keystrokes and mouse
>     events.  Is there something else I should try?
>
> This seems to imply a fundamental Emacs bug.
>
> Please try this: run Emacs under GDB, reproduce the bug, then stop
> Emacs by typing C-z at the terminal where GDB is running.  Then follow
> the advice in etc/DEBUG about what to do when Emacs is looping.
>
> I also asked people to look at Dave's patch and maybe install it.
> But I suspect there are two separate bugs here, and it would be good
> to fix both.
>
>


-- 
Michael Droettboom
http://www.droettboom.com/

[-- Attachment #2: emacs_hang.txt --]
[-- Type: text/plain, Size: 9017 bytes --]

> gdb
GNU gdb Red Hat Linux (6.3.0.0-1.143.el4rh)
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-redhat-linux-gnu".
.gdbinit:22: Error in sourced command file:
No symbol table is loaded.  Use the "file" command.
(gdb) attach 13754
Attaching to process 13754

[[snip]]

(gdb) source .gdbinit
DISPLAY = :0.0
TERM = xterm
Breakpoint 1 at 0x80e8bc2: file emacs.c, line 431.
Breakpoint 2 at 0x81009b9: file sysdep.c, line 1385.
(gdb) c
Continuing.
Detaching after fork from child process 13772.

[[Load attached test.py.  Go to end of last line.  Execute
python-find-imports, then python-complete-symbol.  Emacs stops
responding.  Called 'kill -TSTP 13754']]

Detaching after fork from child process 13781.

Program received signal SIGTSTP, Stopped (user).
[Switching to Thread -1208154432 (LWP 13754)]
0x00aa07a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
(gdb) bt full
#0  0x00aa07a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
No symbol table info available.
#1  0x00b8064d in ___newselect_nocancel () from /lib/tls/libc.so.6
No symbol table info available.
#2  0x0817700e in select_wrapper (n=Variable "n" is not available.
) at process.c:4225
No locals.
#3  0x08178c52 in wait_reading_process_output (time_limit=5, microsecs=0, read_kbd=0,
    do_display=0, wait_for_cell=137377993, wait_proc=0x919aa20, just_wait_proc=0)
    at process.c:4594
        usecs = Variable "usecs" is not available.

Lisp Backtrace:
"accept-process-output" (0x919aa24)
"python-send-receive" (0x913b5bb)
"byte-code" (0x8e1c33b)
"python-symbol-completions" (0x913b5eb)
"python-complete-symbol" (0x83038c9)
"call-interactively" (0x8e443c9)
"execute-extended-command" (0x83038c9)
"call-interactively" (0x830dc01)
(gdb) xbacktrace
"accept-process-output" (0x919aa24)
"python-send-receive" (0x913b5bb)
"byte-code" (0x8e1c33b)
"python-symbol-completions" (0x913b5eb)
"python-complete-symbol" (0x83038c9)
"call-interactively" (0x8e443c9)
"execute-extended-command" (0x83038c9)
"call-interactively" (0x830dc01)
(gdb) bt
#0  0x00aa07a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
#1  0x00b8064d in ___newselect_nocancel () from /lib/tls/libc.so.6
#2  0x0817700e in select_wrapper (n=Variable "n" is not available.
) at process.c:4225
#3  0x08178c52 in wait_reading_process_output (time_limit=5, microsecs=0, read_kbd=0,
    do_display=0, wait_for_cell=137377993, wait_proc=0x919aa20, just_wait_proc=0)
    at process.c:4594
#4  0x0817aef5 in Faccept_process_output (process=152676900, seconds=40, millisec=137377993,
    just_this_one=137377993) at process.c:3966
#5  0x0814a147 in Ffuncall (nargs=3, args=0xbfe69180) at eval.c:3000
#6  0x08172381 in Fbyte_code (bytestr=149233923, vector=149237380, maxdepth=Variable "maxdepth" is not available.
) at bytecode.c:679
#7  0x08149bef in funcall_lambda (fun=149237532, nargs=1, arg_vector=0xbfe692b8) at eval.c:3184
#8  0x0814a01a in Ffuncall (nargs=2, args=0xbfe692b4) at eval.c:3054
#9  0x08172381 in Fbyte_code (bytestr=149013307, vector=148470844, maxdepth=Variable "maxdepth" is not available.
) at bytecode.c:679
#10 0x08149793 in Feval (form=139039869) at eval.c:2328
#11 0x0814b855 in internal_lisp_condition_case (var=137377993, bodyform=139039869,
    handlers=139040085) at eval.c:1426
#12 0x08172776 in Fbyte_code (bytestr=149013803, vector=146456940, maxdepth=Variable "maxdepth" is not available.
) at bytecode.c:869
#13 0x08149bef in funcall_lambda (fun=147814364, nargs=1, arg_vector=0xbfe69654) at eval.c:3184
#14 0x0814a01a in Ffuncall (nargs=2, args=0xbfe69650) at eval.c:3054
#15 0x08172381 in Fbyte_code (bytestr=149013227, vector=148416964, maxdepth=Variable "maxdepth" is not available.
) at bytecode.c:679
#16 0x08149bef in funcall_lambda (fun=148959756, nargs=0, arg_vector=0xbfe697b4) at eval.c:3184
#17 0x0814a01a in Ffuncall (nargs=1, args=0xbfe697b0) at eval.c:3054
#18 0x0814b592 in apply1 (fn=149177289, arg=137377993) at eval.c:2738
#19 0x08147814 in Fcall_interactively (function=149177289, record_flag=137378041, keys=137419308)
    at callint.c:406
#20 0x080efdc5 in Fcommand_execute (cmd=149177289, record_flag=137378041, keys=137377993,
    special=137377993) at keyboard.c:10036
#21 0x080efff5 in Fexecute_extended_command (prefixarg=137377993) at keyboard.c:10152
#22 0x0814a16b in Ffuncall (nargs=2, args=0xbfe69ae0) at eval.c:3000
#23 0x0814767e in Fcall_interactively (function=137419777, record_flag=137377993, keys=137419308)
    at callint.c:860
#24 0x080efdc5 in Fcommand_execute (cmd=137419777, record_flag=137377993, keys=137377993,
    special=137377993) at keyboard.c:10036
#25 0x080f741e in command_loop_1 () at keyboard.c:1873
#26 0x0814852a in internal_condition_case (bfun=0x80f70ac <command_loop_1>, handlers=137423449,
    hfun=0x80f0694 <cmd_error>) at eval.c:1481
#27 0x080eb1b6 in command_loop_2 () at keyboard.c:1329
#28 0x08148209 in internal_catch (tag=137416633, func=0x80eb198 <command_loop_2>, arg=137377993)
    at eval.c:1222
#29 0x080eaff5 in command_loop () at keyboard.c:1308
#30 0x080eb090 in recursive_edit_1 () at keyboard.c:1006
#31 0x080eb15a in Frecursive_edit () at keyboard.c:1067
#32 0x080ea50e in main (argc=1, argv=0xbfe6a2e4) at emacs.c:1762

Lisp Backtrace:
"accept-process-output" (0x919aa24)
"python-send-receive" (0x913b5bb)
"byte-code" (0x8e1c33b)
"python-symbol-completions" (0x913b5eb)
"python-complete-symbol" (0x83038c9)
"call-interactively" (0x8e443c9)
"execute-extended-command" (0x83038c9)
"call-interactively" (0x830dc01)
(gdb) step
Single stepping until exit from function _dl_sysinfo_int80,
which has no line number information.

0x00b8064d in ___newselect_nocancel () from /lib/tls/libc.so.6
(gdb)
Single stepping until exit from function ___newselect_nocancel,
which has no line number information.
select_wrapper (n=Variable "n" is not available.
) at process.c:4226
4226    }

[[Trying to get arguments to system call, but first two are "not available."]]

(gdb) p n
Variable "n" is not available.
(gdb) p rfd
Variable "rfd" is not available.
(gdb) p wfd
$1 = (fd_set *) 0x0
(gdb) p xfd
$2 = (fd_set *) 0x0
(gdb) p tmo
$3 = (struct timeval *) 0xbfe68db0

[[Calling finish to see loop.]]

(gdb) finish
Run till exit from #0  select_wrapper (n=Variable "n" is not available.
) at process.c:4226
0x08178c52 in wait_reading_process_output (time_limit=5, microsecs=0, read_kbd=0, do_display=0,
    wait_for_cell=137377993, wait_proc=0x919aa20, just_wait_proc=0) at process.c:4594
4594              nfds = select (max (max_process_desc, max_keyboard_desc) + 1,
Value returned is $4 = 0
(gdb) finish
Run till exit from #0  0x08178c52 in wait_reading_process_output (time_limit=5, microsecs=0,
    read_kbd=0, do_display=0, wait_for_cell=137377993, wait_proc=0x919aa20, just_wait_proc=0)
    at process.c:4594
0x0817aef5 in Faccept_process_output (process=152676900, seconds=40, millisec=137377993,
    just_this_one=137377993) at process.c:3966
3966      return
Value returned is $5 = 0
(gdb) finish
Run till exit from #0  0x0817aef5 in Faccept_process_output (process=152676900, seconds=40,
    millisec=137377993, just_this_one=137377993) at process.c:3966
0x0814a147 in Ffuncall (nargs=3, args=0xbfe69180) at eval.c:3000
3000              val = (*XSUBR (fun)->function) (internal_args[0], internal_args[1]);
Value returned is $6 = 137377993
(gdb) finish
Run till exit from #0  0x0814a147 in Ffuncall (nargs=3, args=0xbfe69180) at eval.c:3000
Fbyte_code (bytestr=149233923, vector=149237380, maxdepth=Variable "maxdepth" is not available.
) at bytecode.c:681
681                 break;
Value returned is $7 = 137377993
(gdb) finish
Run till exit from #0  Fbyte_code (bytestr=149233923, vector=149237380, maxdepth=Variable "maxdepth" is not available.
)
    at bytecode.c:681

[[gdb stops here.  Call 'kill -TSTP' on emacs again.]]

Program received signal SIGTSTP, Stopped (user).
0x00aa07a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
(gdb) bt full
#0  0x00aa07a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
No symbol table info available.
#1  0x00b8064d in ___newselect_nocancel () from /lib/tls/libc.so.6
No symbol table info available.
#2  0x0817700e in select_wrapper (n=Variable "n" is not available.
) at process.c:4225
No locals.
#3  0x08178c52 in wait_reading_process_output (time_limit=5, microsecs=0, read_kbd=0,
    do_display=0, wait_for_cell=137377993, wait_proc=0x919aa20, just_wait_proc=0)
    at process.c:4594
        usecs = Variable "usecs" is not available.

Lisp Backtrace:
"accept-process-output" (0x919aa24)
"python-send-receive" (0x913b5bb)
"byte-code" (0x8e1c33b)
"python-symbol-completions" (0x913b5eb)
"python-complete-symbol" (0x83038c9)
"call-interactively" (0x8e443c9)
"execute-extended-command" (0x83038c9)
"call-interactively" (0x830dc01)
(gdb)

[-- Attachment #3: test.py --]
[-- Type: application/octet-stream, Size: 47 bytes --]

from sys import argv \
    exit
import os

os.p

[-- Attachment #4: emacs_report.txt --]
[-- Type: text/plain, Size: 1856 bytes --]

In GNU Emacs 22.1.1 (i686-pc-linux-gnu, GTK+ Version 2.10.9)
 of 2007-09-09 on wonkabar.stsci.edu
Windowing system distributor `The X.Org Foundation', version 11.0.60802000
configured using `configure  '--prefix=$USER/usr' '--with-gtk' '--enable-font-backend' '--with-xft' '--with-xpm' '--with-jpeg' '--with-png''

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: en_US.UTF-8
  locale-coding-system: utf-8
  default-enable-multibyte-characters: t

Major mode: Python

Minor modes in effect:
  shell-dirtrack-mode: t
  encoded-kbd-mode: t
  show-paren-mode: t
  cua-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  blink-cursor-mode: t
  unify-8859-on-encoding-mode: t
  utf-translate-cjk-mode: t
  auto-compression-mode: t
  column-number-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
<help-echo> <down-mouse-1> <mouse-1> C-x C-f t e s 
t . p y <return> <down> <down> <down> <down> <down> 
M-x p y t h o n - f i l l - <backspace> <backspace> 
<backspace> n g - <backspace> <backspace> d - i m p 
o <tab> <return> M-x e m a c s <backspace> <backspace> 
<backspace> <backspace> <backspace> r e p o r t - e 
m a c s <tab> <return>

Recent messages:
Loading easy-mmode...done
Loading derived...done
Loading $USER/usr/share/emacs/site-lisp/auctex.el (source)...
Loading $USER/usr/share/emacs/site-lisp/tex-site.el (source)...done
Loading $USER/usr/share/emacs/site-lisp/auctex.el (source)...done
Loading printing...done
Loading server...done
For information about the GNU Project and its goals, type C-h C-p. [2 times]
Loading cl-seq...done
Loading emacsbug...done