bug#34826: Possible bug on "define-charset" of Emacs 27.0.50

bug#34826: Possible bug on "define-charset" of Emacs 27.0.50

[Ikki Tachikawa]

[-- Attachment #1: Type: text/plain, Size: 7878 bytes --]

 Dear developers,

Perhaps at the following code, Emacs 27.0.50 (newest HEAD Windows version)
fails.
It works fine with Emacs 26.1.  Perhaps it happens independent of platform.

    (define-charset 'adobe-japan1 "Adobe-Japan1"
      :map (file-name-sans-extension aj16map)
      :invalid-code 128)))

Thread 1 received signal SIGSEGV, Segmentation fault.
0x00000004001c3042 in assq_no_quit (key=XIL(0xfffffffc04270ab0),
    list=XIL(0x100000000ff)) at fns.c:1607
1607        if (CONSP (XCAR (list)) && EQ (XCAR (XCAR (list)), key))
(gdb) bt
#0  0x00000004001c3042 in assq_no_quit (key=XIL(0xfffffffc04270ab0),
    list=XIL(0x100000000ff)) at fns.c:1607
#1  0x000000040019a32c in swap_in_symval_forwarding (symbol=0x49647b0,
    blv=0x509fb60) at data.c:1222
#2  0x000000040019a479 in find_symbol_value (symbol=XIL(0xfffffffc04270ab0))
    at data.c:1257
#3  0x000000040019a4dc in Fsymbol_value (symbol=XIL(0xfffffffc04270ab0))
    at data.c:1274
#4  0x000000040012f80d in set_buffer_internal_2 (b=0x4b4fbd0) at
buffer.c:2136
#5  0x000000040012f6f7 in set_buffer_internal_1 (b=0x469d648) at
buffer.c:2085
#6  0x0000000400102752 in set_buffer_internal (b=0x469d648) at buffer.h:1137
#7  0x00000004000c4ec2 in make_conversion_work_buffer (multibyte=true)
    at coding.c:7832
#8  0x00000004000c4f7f in code_conversion_save (with_work_buf=true,
    multibyte=true) at coding.c:7860
#9  0x00000004000c5cd2 in decode_coding_object (coding=0xbfb010,
    src_object=XIL(0x64c6764), from=0, from_byte=0, to=20, to_byte=20,
    dst_object=XIL(0xd200)) at coding.c:8106
#10 0x00000004000cae51 in code_convert_string (string=XIL(0x64c6764),
    coding_system=XIL(0xe760), dst_object=XIL(0xd200), encodep=false,
    nocopy=false, norecord=true) at coding.c:9468
#11 0x00000004000caf13 in code_convert_string_norecord (
    string=XIL(0x64c6764), coding_system=XIL(0xe760), encodep=false)
    at coding.c:9488
#12 0x00000004000caf6c in decode_file_name (fname=XIL(0x64c6764))
    at coding.c:9501
#13 0x000000040014c901 in Fexpand_file_name (name=XIL(0x64c6744),
    default_directory=XIL(0x508d1d4)) at fileio.c:1105
#14 0x00000004001e8885 in openp (path=XIL(0x50899b3), str=XIL(0x64c6744),
    suffixes=XIL(0xbfb9b3), storeptr=0x0, predicate=XIL(0), newer=false)
    at lread.c:1613
#15 0x000000040009dd63 in load_charset_map_from_file (charset=0xbfbaf0,
    mapfile=XIL(0x64c6744), control_flag=0) at charset.c:489
#16 0x000000040009e498 in load_charset (charset=0xbfbaf0, control_flag=0)
    at charset.c:643
#17 0x00000004000a0ed4 in Fdefine_charset_internal (nargs=17, args=0xbfbea8)
    at charset.c:1169
#18 0x00000004001ba8e9 in funcall_subr (
    subr=0x400621300 <Sdefine_charset_internal>, numargs=17, args=0xbfbea8)
    at eval.c:2902
#19 0x00000004001ba5fb in Ffuncall (nargs=18, args=0xbfbea0) at eval.c:2847
#20 0x00000004001b9a1e in Fapply (nargs=3, args=0xbfc100) at eval.c:2467
#21 0x00000004001ba8e9 in funcall_subr (subr=0x40062b340 <Sapply>,
numargs=3,
    args=0xbfc100) at eval.c:2902
#22 0x00000004001ba5fb in Ffuncall (nargs=4, args=0xbfc0f8) at eval.c:2847
#23 0x00000004002089ff in exec_byte_code (bytestr=XIL(0x4b67164),
    vector=XIL(0x4b666ad), maxdepth=make_number(7), args_template=XIL(0),
    nargs=0, args=0x0) at bytecode.c:633
#24 0x00000004001bb33a in funcall_lambda (fun=XIL(0x4b6667d), nargs=6,
    arg_vector=0x4b666ad) at eval.c:3123
#25 0x00000004001bad71 in apply_lambda (fun=XIL(0x4b6667d),
    args=XIL(0x64c3983), count=40) at eval.c:2981
#26 0x00000004001b9391 in eval_sub (form=XIL(0x64c3953)) at eval.c:2360
#27 0x00000004001b514f in Fprogn (body=XIL(0)) at eval.c:478
#28 0x00000004001b8e12 in eval_sub (form=XIL(0x64c2673)) at eval.c:2267
#29 0x00000004001b5020 in Fif (args=XIL(0x64c2693)) at eval.c:433
#30 0x00000004001b8e12 in eval_sub (form=XIL(0x64c26a3)) at eval.c:2267
#31 0x00000004001b514f in Fprogn (body=XIL(0)) at eval.c:478
#32 0x00000004001b64c9 in Flet (args=XIL(0x64c2773)) at eval.c:1005
#33 0x00000004001b8e12 in eval_sub (form=XIL(0x64c2783)) at eval.c:2267
#34 0x00000004001e943b in readevalloop_eager_expand_eval
(val=XIL(0x64c4263),
    macroexpand=XIL(0xfffffffc04305a58)) at lread.c:1904
#35 0x00000004001e9daf in readevalloop (readcharfun=XIL(0x50e8885),
    infile0=0x0, sourcename=XIL(0x5dd98c4), printflag=false, unibyte=XIL(0),
    readfun=XIL(0), start=XIL(0), end=XIL(0)) at lread.c:2087
#36 0x00000004001ea1f1 in Feval_buffer (buffer=XIL(0x50e8885),
    printflag=XIL(0), filename=XIL(0x50cd424), unibyte=XIL(0),
    do_allow_print=XIL(0xd200)) at lread.c:2154
#37 0x00000004001baa88 in funcall_subr (subr=0x40062de80 <Seval_buffer>,
    numargs=5, args=0xbfd040) at eval.c:2934
#38 0x00000004001ba5fb in Ffuncall (nargs=6, args=0xbfd038) at eval.c:2847
#39 0x00000004002089ff in exec_byte_code (bytestr=XIL(0x481f3b4),
    vector=XIL(0x481ec35), maxdepth=make_number(6), args_template=XIL(0),
    nargs=0, args=0x0) at bytecode.c:633
#40 0x00000004001bb33a in funcall_lambda (fun=XIL(0x481ec05), nargs=4,
    arg_vector=0x481ec35) at eval.c:3123
#41 0x00000004001ba63f in Ffuncall (nargs=5, args=0xbfd560) at eval.c:2849
#42 0x00000004001ba0ca in call4 (fn=XIL(0xfffffffc0412aed0),
    arg1=XIL(0x50cd424), arg2=XIL(0x50cd424), arg3=XIL(0xd200),
    arg4=XIL(0xd200)) at eval.c:2723
#43 0x00000004001e7f45 in Fload (file=XIL(0x50cd5c4),
    noerror=XIL(0xfffffffc0412b590), nomessage=XIL(0xfffffffc0412b3c0),
    nosuffix=XIL(0), must_suffix=XIL(0)) at lread.c:1381
#44 0x00000004001baa88 in funcall_subr (subr=0x40062de00 <Sload>, numargs=3,
    args=0xbfdd50) at eval.c:2934
#45 0x00000004001ba5fb in Ffuncall (nargs=4, args=0xbfdd48) at eval.c:2847
#46 0x00000004002089ff in exec_byte_code (bytestr=XIL(0x4bc47ec),
    vector=XIL(0x4bc4555), maxdepth=make_number(18),
    args_template=make_number(769), nargs=3, args=0xbfe2f8) at
bytecode.c:633
#47 0x00000004001bafde in funcall_lambda (fun=XIL(0x4bc4525), nargs=3,
    arg_vector=0xbfe2e0) at eval.c:3045
#48 0x00000004001ba63f in Ffuncall (nargs=4, args=0xbfe2d8) at eval.c:2849
#49 0x00000004002089ff in exec_byte_code (bytestr=XIL(0x4bc517c),
    vector=XIL(0x4bc19f5), maxdepth=make_number(12),
    args_template=make_number(0), nargs=0, args=0xbfedf8) at bytecode.c:633
#50 0x00000004001bafde in funcall_lambda (fun=XIL(0x4bc19c5), nargs=0,
    arg_vector=0xbfedf8) at eval.c:3045
#51 0x00000004001ba63f in Ffuncall (nargs=1, args=0xbfedf0) at eval.c:2849
#52 0x00000004002089ff in exec_byte_code (bytestr=XIL(0x4bc5cec),
    vector=XIL(0x4bc534d), maxdepth=make_number(12),
    args_template=make_number(0), nargs=0, args=0xbff3e0) at bytecode.c:633
#53 0x00000004001bafde in funcall_lambda (fun=XIL(0x4bc531d), nargs=0,
    arg_vector=0xbff3e0) at eval.c:3045
#54 0x00000004001bad71 in apply_lambda (fun=XIL(0x4bc531d), args=XIL(0),
    count=4) at eval.c:2981
#55 0x00000004001b9391 in eval_sub (form=XIL(0x4d1bab3)) at eval.c:2360
#56 0x00000004001b8814 in Feval (form=XIL(0x4d1bab3), lexical=XIL(0))
    at eval.c:2134
#57 0x00000004001090c0 in top_level_2 () at keyboard.c:1099
#58 0x00000004001b6e9e in internal_condition_case (
    bfun=0x400109095 <top_level_2>, handlers=XIL(0x52e0),
    hfun=0x4001089df <cmd_error>) at eval.c:1369
#59 0x0000000400109118 in top_level_1 (ignore=XIL(0)) at keyboard.c:1107
#60 0x00000004001b6769 in internal_catch (tag=XIL(0xda70),
    func=0x4001090c6 <top_level_1>, arg=XIL(0)) at eval.c:1132
#61 0x0000000400108f94 in command_loop () at keyboard.c:1068
#62 0x0000000000000000 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)

Lisp Backtrace:
"define-charset-internal" (0xbfbea8)
"apply" (0xbfc100)
"define-charset" (0xbfc620)
"progn" (0xbfc8b0)
"if" (0xbfca30)
"let" (0xbfcc80)
"eval-buffer" (0xbfd040)
"load-with-code-conversion" (0xbfd568)
"load" (0xbfdd50)
"startup--load-user-init-file" (0xbfe2e0)
"command-line" (0xbfedf8)
"normal-top-level" (0xbff3e0)

Regards,

[-- Attachment #2: Type: text/html, Size: 65549 bytes --]

bug#34826: Possible bug on "define-charset" of Emacs 27.0.50

[Eli Zaretskii]

> From: Ikki Tachikawa <tckwik@gmail.com>
> Date: Tue, 12 Mar 2019 19:36:29 +0900
> 
> Perhaps at the following code, Emacs 27.0.50 (newest HEAD Windows version) fails.
> It works fine with Emacs 26.1.  Perhaps it happens independent of platform.
> 
>     (define-charset 'adobe-japan1 "Adobe-Japan1"
>       :map (file-name-sans-extension aj16map)
>       :invalid-code 128)))

What is aj16map?  I guess it's a variable that holds the name of a map
file?  Please show that file as well, as otherwise it's impossible to
reproduce and investigate this issue.

Thanks.

bug#34826: Possible bug on "define-charset" of Emacs 27.0.50

[Eli Zaretskii]

> From: Ikki Tachikawa <tckwik@gmail.com>
> Date: Tue, 12 Mar 2019 19:36:29 +0900
> 
> Perhaps at the following code, Emacs 27.0.50 (newest HEAD Windows version) fails.
> It works fine with Emacs 26.1.  Perhaps it happens independent of platform.
> 
>     (define-charset 'adobe-japan1 "Adobe-Japan1"
>       :map (file-name-sans-extension aj16map)
>       :invalid-code 128)))

This was a bug introduced when we switched to using the portable
dumping: an attempt to create any additional charset after restoring
from a pdump file would crash.

Should be fixed now.

Thanks.