bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Jim Meyering]

When I edit any file and interrupt emacs, I get an abort, in src/eval.c:

  if (gc_in_progress || waiting_for_input)
    abort ();

waiting_for_input is 1.

However, that happens only when using my .emacs file.
When I use -q, e.g., "touch /tmp/x; emacs -q /tmp/x" there is no abort.

Bisecting ~/.emacs, I found the culprit:

  (require 'saveplace)

without that line, there is no problem.

So anyone can reproduce it by running this and then hitting ^C:

  $ touch /tmp/k; emacs -q --eval "(require 'saveplace)" /tmp/k
  ^CFatal error (6)zsh: abort (core dumped)  emacs -q --eval "(require\
    'saveplace)" /tmp/k
  [Exit 134 (ABRT)]

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Paul Eggert]

> So anyone can reproduce it by running this and then hitting ^C:
>  $ touch /tmp/k; emacs -q --eval "(require 'saveplace)" /tmp/k
>  ^CFatal error (6)zsh: abort (core dumped)  emacs -q --eval "(require\
>    'saveplace)" /tmp/k
>  [Exit 134 (ABRT)]

Unfortunately I can't reproduce this problem, with either Fedora 15 Emacs
or with the latest trunk (built with GCC 4.6.2, x86-64).  Which version of
Emacs and which platform are you using?  Can you send a GDB backtrace?
Assuming it's the trunk, it's helpful to compile without optimization
when getting a backtrace.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Jim Meyering]

Paul Eggert wrote:
>> So anyone can reproduce it by running this and then hitting ^C:
>>  $ touch /tmp/k; emacs -q --eval "(require 'saveplace)" /tmp/k
>>  ^CFatal error (6)zsh: abort (core dumped)  emacs -q --eval "(require\
>>    'saveplace)" /tmp/k
>>  [Exit 134 (ABRT)]
>
> Unfortunately I can't reproduce this problem, with either Fedora 15 Emacs
> or with the latest trunk (built with GCC 4.6.2, x86-64).  Which version of
> Emacs and which platform are you using?  Can you send a GDB backtrace?
> Assuming it's the trunk, it's helpful to compile without optimization
> when getting a backtrace.

Hi Paul,
It was with the latest emacs from bzr (some time yesterday),
built on Fedora 16 using gcc-4.7.0 20111124.

If I find time today, I'll narrow it down or post a backtrace.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Glenn Morris]

I can reproduce it on Scientific Linux 6.1, x86_64, with gcc 4.4.5
20110214 (Red Hat 4.4.5-6). Backtrace:

#0  abort () at emacs.c:386
No locals.
#1  0x00000000005f7e51 in Fsignal (error_symbol=12961906, data=19014854)
    at eval.c:1662
        conditions = 15231761
        string = 19738512
        real_error_symbol = 12961906
        clause = 12777890
        h = 0x12224b6
        bp = 0xe86b11
#2  0x00000000005f81ba in xsignal (error_symbol=12961906, data=19014854)
    at eval.c:1758
No locals.
#3  0x00000000005f8289 in xsignal3 (error_symbol=12961906, arg1=15231761, arg2=
    208, arg3=212) at eval.c:1785
No locals.
#4  0x000000000063ad86 in scan_lists (from=15231761, count=1, depth=-2, 
    sexpflag=0) at syntax.c:2607
        comstart_first = 0
        prefix = 0
        syntax = 5
        other_syntax = 0
        val = 12
        stop = 212
        c = 208
        c1 = 208
        stringterm = 34
        quoted = 112
        mathexit = 0
---Type <return> to continue, or q <return> to quit---
        code = Sclose
        temp_code = 208
        min_depth = -1
        comstyle = 0
        comnested = 0
        temp_pos = 3
        last_good = 52
        found = 0
        from_byte = 53
        out_bytepos = 140737488295520
        out_charpos = 47
        temp = 2
        dummy = 0
        multibyte_symbol_p = 0
#5  0x000000000063e831 in Fscan_lists (from=12, count=4, depth=-4)
    at syntax.c:2865
No locals.
#6  0x00000000005fad2a in Ffuncall (nargs=4, args=0x7fffffff19b0)
    at eval.c:2981
        fun = 12160837
        original_fun = 12963634
        funcar = 4
        numargs = 3
        lisp_numargs = 13226320
        val = 12777890
        backtrace = {
          next = 0x7fffffff1e00, 
          function = 0x7fffffff19b0, 
          args = 0x7fffffff19b8, 
---Type <return> to continue, or q <return> to quit---
          nargs = 3, 
          debug_on_exit = 0
        }
        internal_args = 0x7fffffff19b8
        i = 140737488296352
#7  0x0000000000647ae4 in exec_byte_code (bytestr=10875017, vector=10875053, 
    maxdepth=20, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 14
        op = 3
        vectorp = 0xa5f0b8
        stack = {
          pc = 0xb21096 "\206$", 
          byte_string = 10875017, 
          byte_string_start = 0xb21078 "\b\204\006", 
          constants = 10875053, 
          next = 0x7fffffff1ef0
        }
        top = 0x7fffffff19b0
        result = 12777890
#8  0x00000000005fb742 in funcall_lambda (fun=10874957, nargs=1, arg_vector=
    0xa5f0ad) at eval.c:3205
        val = 12777890
        syms_left = 12777890
        next = 13184898
        lexenv = 12777890
        count = 13
        i = 1
        optional = 1
        rest = 0
---Type <return> to continue, or q <return> to quit---
#9  0x00000000005faee6 in Ffuncall (nargs=2, args=0x7fffffff1e98)
    at eval.c:3023
        fun = 10874957
        original_fun = 13178802
        funcar = 1
        numargs = 1
        lisp_numargs = 4294967295
        val = 12777890
        backtrace = {
          next = 0x7fffffff2330, 
          function = 0x7fffffff1e98, 
          args = 0x7fffffff1ea0, 
          nargs = 1, 
          debug_on_exit = 0
        }
        internal_args = 0xfec0f5
        i = 140737488297424
#10 0x0000000000647ae4 in exec_byte_code (bytestr=17578657, vector=16695541, 
    maxdepth=12, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 13
        op = 1
        vectorp = 0xfec100
        stack = {
          pc = 0x12d1f04 "\210\207", 
          byte_string = 17578657, 
          byte_string_start = 0x12d1f00 "ÀÁÂ!\210\207", 
          constants = 16695541, 
          next = 0x7fffffff25d0
        }
---Type <return> to continue, or q <return> to quit---
        top = 0x7fffffff1e98
        result = 12777938
#11 0x0000000000646f63 in Fbyte_code (bytestr=17578657, vector=16695541, 
    maxdepth=12) at bytecode.c:423
No locals.
#12 0x00000000005f9670 in eval_sub (form=18964262) at eval.c:2328
        numargs = 12
        args_left = 12777890
        i = 6582052
        maxargs = 3
        argvals =           {17578657,
          16695541,
          12,
          44,
          0,
          6266547,
          140737488298864,
          1}
        fun = 12161029
        val = 3
        original_fun = 12917730
        original_args = 18964278
        funcar = 3
        backtrace = {
          next = 0x7fffffff29b0, 
          function = 0x7fffffff2360, 
          args = 0x7fffffff2290, 
          nargs = 3, 
          debug_on_exit = 0
---Type <return> to continue, or q <return> to quit---
        }
        gcpro1 = {
          next = 0xd, 
          var = 0xe86ab1, 
          nvars = 140737488299072
        }
        gcpro2 = {
          next = 0x7fffffff2440, 
          var = 0x600bc1, 
          nvars = 48
        }
        gcpro3 = {
          next = 0xd, 
          var = 0x7fffffff2290, 
          nvars = 3
        }
#13 0x00000000005f77c2 in internal_lisp_condition_case (var=16146594, bodyform=
    18964262, handlers=18964358) at eval.c:1453
        val = 12777890
        c = {
          tag = 12777890, 
          val = 12777890, 
          next = 0x7fffffff52f0, 
          gcpro = 0x0, 
          jmp =             {{
              __jmpbuf =                 {448,
                -8826069716758269337,
                0,
                140737488312912,
---Type <return> to continue, or q <return> to quit---
                0,
                0,
                -8826069716840058265,
                8826069120708014695}, 
              __mask_was_saved = 0, 
              __saved_mask = {
                __val =                   {12777890,
                  12777890,
                  5853900,
                  140737488299360,
                  6274425,
                  12777890,
                  55851513923,
                  12777890,
                  19014806,
                  6190615,
                  0,
                  2,
                  2,
                  15226369,
                  15226368,
                  15226368}
              }
            }}, 
          backlist = 0x7fffffff29b0, 
          handlerlist = 0x7fffffff52c0, 
          lisp_eval_depth = 5, 
          pdlcount = 13, 
          poll_suppress_count = 1, 
---Type <return> to continue, or q <return> to quit---
          interrupt_input_blocked = 0, 
          byte_stack = 0x7fffffff25d0
        }
        h = {
          handler = 18964358, 
          var = 16146594, 
          chosen_clause = 140737488299200, 
          tag = 0x7fffffff2440, 
          next = 0x7fffffff52c0
        }
#14 0x0000000000648862 in exec_byte_code (bytestr=17578561, vector=16797397, 
    maxdepth=12, args_template=12777890, nargs=0, args=0x0) at bytecode.c:981
        handlers = 18964358
        body = 18964262
        count = 13
        op = 143
        vectorp = 0x1004ee0
        stack = {
          pc = 0x12d1e7b "\203\061", 
          byte_string = 17578561, 
          byte_string_start = 0x12d1e70 "eb\210m\204X", 
          constants = 16797397, 
          next = 0x7fffffff2aa0
        }
        top = 0x7fffffff2570
        result = 18960278
#15 0x00000000005fb742 in funcall_lambda (fun=16797749, nargs=0, arg_vector=
    0x1004ed5) at eval.c:3205
        val = 12777890
---Type <return> to continue, or q <return> to quit---
        syms_left = 12777890
        next = 480
        lexenv = 12777890
        count = 13
        i = 0
        optional = 0
        rest = 0
#16 0x00000000005faee6 in Ffuncall (nargs=1, args=0x7fffffff2a40)
    at eval.c:3023
        fun = 16797749
        original_fun = 14562050
        funcar = 12777938
        numargs = 0
        lisp_numargs = 12897104
        val = 12777890
        backtrace = {
          next = 0x7fffffff2e80, 
          function = 0x7fffffff2a40, 
          args = 0x7fffffff2a48, 
          nargs = 0, 
          debug_on_exit = 0
        }
        internal_args = 0x1e0
        i = 140737488300592
#17 0x0000000000647ae4 in exec_byte_code (bytestr=17581121, vector=18085877, 
    maxdepth=12, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 11
        op = 0
        vectorp = 0x113f800
---Type <return> to continue, or q <return> to quit---
        stack = {
          pc = 0x12d1c74 "\210Î *\207", 
          byte_string = 17581121, 
          byte_string_start = 
    0x12d1c58 "rÅÆ!q\210Ç\216ÈÉ!\210Ê\b!\210\tË\032\033Ì\fp\"\210*Í \210Î *\207", 
          constants = 18085877, 
          next = 0x7fffffff2f70
        }
        top = 0x7fffffff2a40
        result = 13489890
#18 0x00000000005fb742 in funcall_lambda (fun=17184725, nargs=1, arg_vector=
    0x113f7f5) at eval.c:3205
        val = 42962450898
        syms_left = 12777890
        next = 13006258
        lexenv = 12777890
        count = 10
        i = 1
        optional = 0
        rest = 0
#19 0x00000000005faee6 in Ffuncall (nargs=2, args=0x7fffffff2f18)
    at eval.c:3023
        fun = 17184725
        original_fun = 14562002
        funcar = 12777890
        numargs = 1
        lisp_numargs = 1
        val = 0
---Type <return> to continue, or q <return> to quit---
        backtrace = {
          next = 0x7fffffff3350, 
          function = 0x7fffffff2f18, 
          args = 0x7fffffff2f20, 
          nargs = 1, 
          debug_on_exit = 0
        }
        internal_args = 0x10050b5
        i = 140737488301840
#20 0x0000000000647ae4 in exec_byte_code (bytestr=16796577, vector=16797877, 
    maxdepth=12, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 10
        op = 1
        vectorp = 0x10050c0
        stack = {
          pc = 0x12d2084 "\t\206\t", 
          byte_string = 16796577, 
          byte_string_start = 0x12d2080 "ÃÄ\b!\t\206\t", 
          constants = 16797877, 
          next = 0x7fffffff3440
        }
        top = 0x7fffffff2f18
        result = 6267955
#21 0x00000000005fb742 in funcall_lambda (fun=18065669, nargs=2, arg_vector=
    0x10050b5) at eval.c:3205
        val = 0
        syms_left = 12777890
        next = 16146642
        lexenv = 12777890
---Type <return> to continue, or q <return> to quit---
        count = 8
        i = 2
        optional = 1
        rest = 0
#22 0x00000000005faee6 in Ffuncall (nargs=3, args=0x7fffffff33e0)
    at eval.c:3023
        fun = 18065669
        original_fun = 14562098
        funcar = 12898242
        numargs = 2
        lisp_numargs = 12897008
        val = 14176790
        backtrace = {
          next = 0x7fffffff3820, 
          function = 0x7fffffff33e0, 
          args = 0x7fffffff33e8, 
          nargs = 2, 
          debug_on_exit = 0
        }
        internal_args = 0x10337c5
        i = 140737488303056
#23 0x0000000000647ae4 in exec_byte_code (bytestr=17434609, vector=16988101, 
    maxdepth=16, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 3
        op = 2
        vectorp = 0x10337d0
        stack = {
          pc = 0x1029aab "\210*\016\030\204\066", 
          byte_string = 17434609, 
---Type <return> to continue, or q <return> to quit---
          byte_string_start = 
    0x1029a80 "Æ\b!Ç\031\032rÈÉ!q\210ed|\210\v\203\027", 
          constants = 16988101, 
          next = 0x7fffffff3900
        }
        top = 0x7fffffff33e0
        result = 12777890
#24 0x00000000005fb742 in funcall_lambda (fun=14360165, nargs=0, arg_vector=
    0x10337c5) at eval.c:3205
        val = 12777890
        syms_left = 12777890
        next = 19269826
        lexenv = 12777890
        count = 3
        i = 0
        optional = 0
        rest = 0
#25 0x00000000005faee6 in Ffuncall (nargs=1, args=0x7fffffff38b0)
    at eval.c:3023
        fun = 14360165
        original_fun = 16146258
        funcar = 10783013
        numargs = 0
        lisp_numargs = 12026631
        val = 12777890
        backtrace = {
          next = 0x7fffffff3ce0, 
          function = 0x7fffffff38b0, 
          args = 0x7fffffff38b8, 
---Type <return> to continue, or q <return> to quit---
          nargs = 0, 
          debug_on_exit = 0
        }
        internal_args = 0x102daf5
        i = 12026624
#26 0x0000000000647ae4 in exec_byte_code (bytestr=16765873, vector=16964341, 
    maxdepth=4, args_template=12777890, nargs=0, args=0x0) at bytecode.c:785
        count = 3
        op = 0
        vectorp = 0x102db00
        stack = {
          pc = 0x102a319 "\207", 
          byte_string = 16765873, 
          byte_string_start = 0x102a310 "Á \210\b\205\t", 
          constants = 16964341, 
          next = 0x0
        }
        top = 0x7fffffff38b0
        result = 12777938
#27 0x00000000005fb742 in funcall_lambda (fun=16693349, nargs=0, arg_vector=
    0x102daf5) at eval.c:3205
        val = 140737488305640
        syms_left = 12777890
        next = 140737340148088
        lexenv = 12777890
        count = 3
        i = 0
        optional = 0
        rest = 0
---Type <return> to continue, or q <return> to quit---
#28 0x00000000005faee6 in Ffuncall (nargs=1, args=0x7fffffff3e50)
    at eval.c:3023
        fun = 16693349
        original_fun = 16146450
        funcar = 0
        numargs = 0
        lisp_numargs = 12777890
        val = 20
        backtrace = {
          next = 0x0, 
          function = 0x7fffffff3e50, 
          args = 0x7fffffff3e58, 
          nargs = 0, 
          debug_on_exit = 0
        }
        internal_args = 0x7ffff70a09a0
        i = 10784509
#29 0x00000000005f9ed6 in funcall_nil (nargs=1, args=0x7fffffff3e50)
    at eval.c:2491
No locals.
#30 0x00000000005fa2e7 in run_hook_with_args (nargs=1, args=0x7fffffff3e50, 
    funcall=0x5f9eb3 <funcall_nil>) at eval.c:2680
        global_vals = 12777890
        sym = 12940338
        val = 13947622
        ret = 12777890
        gcpro1 = {
          next = 0x126c852, 
          var = 0x7ffff70a09a0, 
---Type <return> to continue, or q <return> to quit---
          nvars = 7032063
        }
        gcpro2 = {
          next = 0xf, 
          var = 0xc2c9e5, 
          nvars = 820
        }
        gcpro3 = {
          next = 0x7fffffff3e10, 
          var = 0x62aeb8, 
          nvars = 0
        }
#31 0x00000000005f9f22 in Frun_hooks (nargs=1, args=0x7fffffff3e88)
    at eval.c:2518
        hook =           {16146450}
        i = 0
#32 0x00000000005573f1 in Fkill_emacs (arg=12777890) at emacs.c:2005
        gcpro1 = {
          next = 0xa48fd6, 
          var = 0x6b6f2d, 
          nvars = 4294917824
        }
        hook = 12940338
        exit_code = 0
#33 0x000000000056d52a in interrupt_signal (signalnum=2) at keyboard.c:10862
        old_errno = 11
        terminal = 0x0
#34 <signal handler called>
No symbol table info available.
---Type <return> to continue, or q <return> to quit---
#35 0x000000366b8de363 in __select_nocancel () from /lib64/libc.so.6
No symbol table info available.
#36 0x0000000000652a14 in select_wrapper (n=8, rfd=0x7fffffff4660, wfd=
    0x7fffffff45e0, xfd=0x0, tmo=0x7fffffff45c0) at process.c:4245
No locals.
#37 0x0000000000653b1e in wait_reading_process_output (time_limit=0, microsecs=
    0, read_kbd=-1, do_display=1, wait_for_cell=12777890, wait_proc=0x0, 
    just_wait_proc=0) at process.c:4611
        timeout_reduced_for_timers = 1
        channel = -47520
        nfds = 0
        Available = {
          fds_bits =             {128,
            0 <repeats 15 times>}
        }
        Writeok = {
          fds_bits =             {0 <repeats 16 times>}
        }
        check_write = 1
        check_delay = 0
        no_avail = 0
        xerrno = 11
        proc = 0
        timeout = {
          tv_sec = 0, 
          tv_usec = 473112
        }
        end_time = {
          tv_sec = 0, 
---Type <return> to continue, or q <return> to quit---
          tv_usec = 0
        }
        wait_channel = -1
        got_some_input = 1
        count = 2
#38 0x000000000055eac0 in kbd_buffer_get_event (kbp=0x7fffffff4990, 
    used_mouse_menu=0x7fffffff4ec4, end_time=0x0) at keyboard.c:3850
        c = 2
        obj = 4294967296
#39 0x000000000055c67b in read_char (commandflag=1, nmaps=2, maps=
    0x7fffffff4ce0, prev_event=12777890, used_mouse_menu=0x7fffffff4ec4, 
    end_time=0x0) at keyboard.c:2796
        kb = 0x0
        c = 12777890
        jmpcount = 2
        local_getcjmp =           {{
            __jmpbuf =               {2,
              -8826069719702670745,
              4261152,
              140737488312912,
              0,
              0,
              -8826069719851568537,
              8826069075804582503}, 
            __mask_was_saved = 0, 
            __saved_mask = {
              __val =                 {66,
                18446744073709551615,
                4294967294,
---Type <return> to continue, or q <return> to quit---
                0,
                0,
                0,
                0,
                0,
                0,
                46,
                0,
                12777890,
                12777890,
                12777890,
                12777890,
                12777890}
            }
          }}
        save_jump =           {{
            __jmpbuf =               {0,
              0,
              0,
              0,
              0,
              0,
              0,
              0}, 
            __mask_was_saved = 0, 
            __saved_mask = {
              __val =                 {0 <repeats 16 times>}
            }
          }}
---Type <return> to continue, or q <return> to quit---
        key_already_recorded = 0
        tem = 12811970
        save = 140737488309296
        previous_echo_area_message = 12777890
        also_record = 12777890
        reread = 0
        gcpro1 = {
          next = 0x7fffffff4bf0, 
          var = 0x7fffffff4bf8, 
          nvars = 140737488308752
        }
        gcpro2 = {
          next = 0x0, 
          var = 0x7fffffff4a18, 
          nvars = 12777890
        }
        polling_stopped_here = 1
        orig_kboard = 0x12b0150
#40 0x0000000000569c32 in read_key_sequence (keybuf=0x7fffffff5130, bufsize=
    30, prompt=12777890, dont_downcase_last=0, can_return_switch_frame=1, 
    fix_current_buffer=1) at keyboard.c:9290
        interrupted_kboard = 0x12b0150
        interrupted_frame = 0xff3700
        key = 16908133
        used_mouse_menu = 0
        echo_local_start = 0
        last_real_key_start = 0
        keys_local_start = 0
        local_first_binding = 1
---Type <return> to continue, or q <return> to quit---
        from_string = 12777890
        count = 2
        t = 0
        echo_start = 0
        keys_start = 0
        nmaps = 2
        nmaps_allocated = 2
        defs = 0x7fffffff4cc0
        submaps = 0x7fffffff4ce0
        orig_local_map = 12777890
        orig_keymap = 12777890
        localized_local_map = 0
        first_binding = 1
        first_unbound = 31
        mock_input = 0
        fkey = {
          parent = 18294934, 
          map = 18294934, 
          start = 0, 
          end = 0
        }
        keytran = {
          parent = 12757414, 
          map = 12757414, 
          start = 0, 
          end = 0
        }
        indec = {
          parent = 18294950, 
---Type <return> to continue, or q <return> to quit---
          map = 18294950, 
          start = 0, 
          end = 0
        }
        shift_translated = 0
        delayed_switch_frame = 12777890
        original_uppercase = 12777938
        original_uppercase_position = -1
        dummyflag = 0
        starting_buffer = 0xf5f2e0
        fake_prefixed_keys = 12777890
        gcpro1 = {
          next = 0x7fffffff4f00, 
          var = 0x5ddf17, 
          nvars = 0
        }
#41 0x00000000005599d2 in command_loop_1 () at keyboard.c:1447
        cmd = 12899282
        keybuf =           {4261152,
          140737488312912,
          140737488310640,
          6152116,
          2822930839,
          12777890,
          0,
          5,
          140737488310720,
          6154447,
          12777890,
---Type <return> to continue, or q <return> to quit---
          12899282,
          140737488310800,
          12899280,
          7,
          12627840,
          0,
          140737488310752,
          140737488310880,
          6274735,
          13356918,
          8602712482,
          12899282,
          12777890,
          0,
          0,
          4261152,
          140737488312912,
          140737488310880,
          6274155}
        i = 12899280
        prev_modiff = 0
        prev_buffer = 0x0
        already_adjusted = 0
#42 0x00000000005f792a in internal_condition_case (bfun=
    0x5595e9 <command_loop_1>, handlers=12830082, hfun=0x558ed8 <cmd_error>)
    at eval.c:1499
        val = 0
        c = {
          tag = 12777890, 
---Type <return> to continue, or q <return> to quit---
          val = 12777890, 
          next = 0x7fffffff5470, 
          gcpro = 0x0, 
          jmp =             {{
              __jmpbuf =                 {5,
                -8826069720493297049,
                4261152,
                140737488312912,
                0,
                0,
                -8826069720575085977,
                8826069120765948519}, 
              __mask_was_saved = 0, 
              __saved_mask = {
                __val =                   {8826069120765948519,
                  0,
                  4294967295,
                  13307238,
                  1,
                  9389376,
                  0,
                  0,
                  0,
                  0,
                  233723453152,
                  1,
                  0,
                  16110896,
                  233731823104,
---Type <return> to continue, or q <return> to quit---
                  16110896}
              }
            }}, 
          backlist = 0x0, 
          handlerlist = 0x0, 
          lisp_eval_depth = 0, 
          pdlcount = 2, 
          poll_suppress_count = 1, 
          interrupt_input_blocked = 0, 
          byte_stack = 0x0
        }
        h = {
          handler = 12830082, 
          var = 12777890, 
          chosen_clause = 12830082, 
          tag = 0x7fffffff52f0, 
          next = 0x0
        }
#43 0x00000000005592d8 in command_loop_2 (ignore=12777890) at keyboard.c:1158
        val = 5
#44 0x00000000005f72b4 in internal_catch (tag=12825874, func=
    0x5592b2 <command_loop_2>, arg=12777890) at eval.c:1256
        c = {
          tag = 12825874, 
          val = 12777890, 
          next = 0x0, 
          gcpro = 0x0, 
          jmp =             {{
              __jmpbuf =                 {5,
---Type <return> to continue, or q <return> to quit---
                -8826069720545725849,
                4261152,
                140737488312912,
                0,
                0,
                -8826069720484908441,
                8826069120562786919}, 
              __mask_was_saved = 0, 
              __saved_mask = {
                __val =                   {6153416,
                  0,
                  4301645564,
                  0,
                  12777890,
                  13003840,
                  140737488311736,
                  14,
                  12805936,
                  12183296,
                  6151995,
                  140737488311696,
                  12777890,
                  4261152,
                  140737488312912,
                  140737488311712}
              }
            }}, 
          backlist = 0x0, 
          handlerlist = 0x0, 
---Type <return> to continue, or q <return> to quit---
          lisp_eval_depth = 0, 
          pdlcount = 2, 
          poll_suppress_count = 1, 
          interrupt_input_blocked = 0, 
          byte_stack = 0x0
        }
#45 0x000000000055928b in command_loop () at keyboard.c:1137
No locals.
#46 0x0000000000558a1c in recursive_edit_1 () at keyboard.c:757
        count = 1
        val = 5606357
#47 0x0000000000558bbf in Frecursive_edit () at keyboard.c:821
        count = 0
        buffer = 12777890
#48 0x0000000000556d37 in main (argc=5, argv=0x7fffffff5a58) at emacs.c:1707
        dummy = 233731852096
        stack_bottom_variable = 0 '\000'
        do_initial_setlocale = 1
        skip_args = 0
        rlim = {
          rlim_cur = 33554432, 
          rlim_max = 18446744073709551615
        }
        no_loadup = 0
        junk = 0x0
        dname_arg = 0x0
        ch_to_dir = 0x7ffff7526cf0 "@\024\"k6"

Lisp Backtrace:
  "scan-lists" (0xffff19b8)
  "down-list" (0xffff1ea0)
  "byte-code" (0xffff2290)
  "pp-buffer" (0xffff2a48)
  "pp-to-string" (0xffff2f20)
  "pp" (0xffff33e8)
  "save-place-alist-to-file" (0xffff38b8)
  "save-place-kill-emacs-hook" (0xffff3e58)

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Andreas Schwab]

Should be fixed now.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Glenn Morris]

Now interrupts are silently ignored:

bash> ./emacs -Q
ctrl-C    # in the shell

just prints ^C.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Glenn Morris]

Glenn Morris wrote:

> Now interrupts are silently ignored:
>
> bash> ./emacs -Q
> ctrl-C    # in the shell
>
> just prints ^C.

Correction, not silently ignored: Emacs prints "Quit" but is otherwise
unaffected.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Stefan Monnier]

> So anyone can reproduce it by running this and then hitting ^C:

>   $ touch /tmp/k; emacs -q --eval "(require 'saveplace)" /tmp/k
>   ^CFatal error (6)zsh: abort (core dumped)  emacs -q --eval "(require\
>     'saveplace)" /tmp/k
>   [Exit 134 (ABRT)]

I can reproduce it indeed with Debian testing's emacs23 as well as with
the current trunk.

I have to be careful to hit C-c before Emacs actually opens the frame
(i.e. interrupt it during startup), otherwise it exits cleanly.

Can't investigate further right now, so if anyone wants to pick up from
here, he's welcome.


        Stefan

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Dan Nicolaescu]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> So anyone can reproduce it by running this and then hitting ^C:
>
>>   $ touch /tmp/k; emacs -q --eval "(require 'saveplace)" /tmp/k
>>   ^CFatal error (6)zsh: abort (core dumped)  emacs -q --eval "(require\
>>     'saveplace)" /tmp/k
>>   [Exit 134 (ABRT)]
>
> I can reproduce it indeed with Debian testing's emacs23 as well as with
> the current trunk.
>
> I have to be careful to hit C-c before Emacs actually opens the frame
> (i.e. interrupt it during startup), otherwise it exits cleanly.

I cannot reproduce it with emacs-23.3 on Fedora 16... 
I can reproduce it on yesterday's trunk.

I vaguely remember there was some problem with saveplace.el and 
emacs --daemon, something about saveplace.el wanting to chat when
shutting down.  No idea if that's related, and I can't find the
discussion at the moment...

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Dan Nicolaescu]

Glenn Morris <rgm@gnu.org> writes:

> I can reproduce it on Scientific Linux 6.1, x86_64, with gcc 4.4.5
> 20110214 (Red Hat 4.4.5-6). Backtrace:
>
> Lisp Backtrace:
>   "scan-lists" (0xffff19b8)
>   "down-list" (0xffff1ea0)
>   "byte-code" (0xffff2290)
>   "pp-buffer" (0xffff2a48)
>   "pp-to-string" (0xffff2f20)
>   "pp" (0xffff33e8)
>   "save-place-alist-to-file" (0xffff38b8)
>   "save-place-kill-emacs-hook" (0xffff3e58)

A quick hack: put a `when' around the `pp' call in `save-place-alist-to-file'

    (when save-place-alist
      (pp (sort save-place-alist

and it does not crash anymore.  Sounds strange, but unfortunately I
don't have time at the moment to investigate further what's really going
on.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Andreas Schwab]

Glenn Morris <rgm@gnu.org> writes:

> Glenn Morris wrote:
>
>> Now interrupts are silently ignored:
>>
>> bash> ./emacs -Q
>> ctrl-C    # in the shell
>>
>> just prints ^C.
>
> Correction, not silently ignored: Emacs prints "Quit" but is otherwise
> unaffected.

IMHO this is the correct behaviour for an interactive programm.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Glenn Morris]

Andreas Schwab wrote:

> IMHO this is the correct behaviour for an interactive programm.

It seems weird to me, but I'm not qualified to argue with you.
None of xterm, gedit, firefox, librefoffice behave as you describe.

gnome-terminal and gvim "detach" themselves, or whatever the term is
(ie, act like you had used '&' even if you had not; this also seems
weird to me), but Emacs does not. gvim -f behaves like Emacs does now.

I guess this means you fixed your own report, so I will merge it:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=5743

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Dan Nicolaescu]

Andreas Schwab <schwab@linux-m68k.org> writes:

> Glenn Morris <rgm@gnu.org> writes:
>
>> Glenn Morris wrote:
>>
>>> Now interrupts are silently ignored:
>>>
>>> bash> ./emacs -Q
>>> ctrl-C    # in the shell
>>>
>>> just prints ^C.
>>
>> Correction, not silently ignored: Emacs prints "Quit" but is otherwise
>> unaffected.
>
> IMHO this is the correct behaviour for an interactive programm.

Why do you think that and what other widely use programs can you name
that have the same behavior?

Why do you think that pretest is the right time to do such a change to a
long available behavior, and why is it OK to do it with zero discussion?

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Andreas Schwab]

Dan Nicolaescu <dann@gnu.org> writes:

> Why do you think that pretest is the right time to do such a change to a
> long available behavior, and why is it OK to do it with zero discussion?

It fixes a severe bug: it is fatal to run Lisp code in a signal handler.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Dan Nicolaescu]

Andreas Schwab <schwab@linux-m68k.org> writes:

> Dan Nicolaescu <dann@gnu.org> writes:
>
>> Why do you think that pretest is the right time to do such a change to a
>> long available behavior, and why is it OK to do it with zero discussion?
>
> It fixes a severe bug: it is fatal to run Lisp code in a signal handler.

How does that that justify introducing another severe bug: can't kill
emacs using C-c  (as it has always been possible ? 

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Andreas Schwab]

Dan Nicolaescu <dann@gnu.org> writes:

> How does that that justify introducing another severe bug: can't kill
> emacs using C-c  (as it has always been possible ? 

C-c in an interactive shell doesn't kill it either.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Dan Nicolaescu]

Andreas Schwab <schwab@linux-m68k.org> writes:

> Dan Nicolaescu <dann@gnu.org> writes:
>
>> How does that that justify introducing another severe bug: can't kill
>> emacs using C-c  (as it has always been possible ? 
>
> C-c in an interactive shell doesn't kill it either.

What's the similarity between emacs as an X11 application and an
interactive shell?
And again, how do you justify changing a long time behavior?

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Daniel Colascione]

[-- Attachment #1: Type: text/plain, Size: 763 bytes --]

On 12/3/11 2:46 PM, Dan Nicolaescu wrote:
> Andreas Schwab <schwab@linux-m68k.org> writes:
> 
>> Dan Nicolaescu <dann@gnu.org> writes:
>>
>>> How does that that justify introducing another severe bug: can't kill
>>> emacs using C-c  (as it has always been possible ? 
>>
>> C-c in an interactive shell doesn't kill it either.
> 
> What's the similarity between emacs as an X11 application and an
> interactive shell?
> And again, how do you justify changing a long time behavior?

I agree: a SIGINT sent to an Emacs running as a deamon or as a GUI application
should kill it dead, without running any Lisp code at all. It's only when Emacs
is actually interesting with the user over a terminal that Emacs should process
control-c specially.



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 235 bytes --]

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Chong Yidong]

Andreas Schwab <schwab@linux-m68k.org> writes:

> Dan Nicolaescu <dann@gnu.org> writes:
>
>> How does that that justify introducing another severe bug: can't kill
>> emacs using C-c  (as it has always been possible ?
>
> C-c in an interactive shell doesn't kill it either.

The argument is invalid: here, Emacs is not using C-c as keyboard input.
This change is worse than the bug it is trying to fix.  Please revert.

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Andreas Schwab]

Chong Yidong <cyd@gnu.org> writes:

> The argument is invalid: here, Emacs is not using C-c as keyboard input.
> This change is worse than the bug it is trying to fix.  Please revert.

It is an absolute no-go to call Lisp in a signal handler.  So I have
installed a different solution.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

bug#10169: a simple interrupt evokes abort?! (but only with (require 'saveplace))

[Richard Stallman]

    > C-c in an interactive shell doesn't kill it either.

    The argument is invalid: here, Emacs is not using C-c as keyboard input.
    This change is worse than the bug it is trying to fix.  Please revert.

I agree.

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use free telephony http://directory.fsf.org/category/tel/