bug#11083: guile-2.0.5 bug report

[Andy Wingo]

Closing as it's been 3 years without any update.  Please feel free to
test the latest 2.0.11 release from the 2.0 series and if you still see
the bug, let us know.

Cheers,

Andy

On Wed 13 Mar 2013 11:55, Andy Wingo  writes:

> Hi Kiyoshi,
>
> We're following up on some bugs in anticipation of a 2.0.8 release.  Can
> you try the following tarball:
>
>   http://wingolog.org/priv/guile-2.0.7.194-dfd1d.tar.gz
>
> Git from stable-2.0 is fine as well.
>
> Thanks!
>
> Andy
>
> On Thu 05 Jul 2012 23:24, Andy Wingo  writes:
>
>> On Fri 13 Apr 2012 18:23, Kiyoshi KANAZAWA  
>> writes:
>>
>>> I downloaded the latest snapshot "guile-2.0.5.100-0f6dd" and tryed making.
>>> The result is:
>>> (1) with -m64
>>> make passed, and "make check " also passed except for 1 FAIL:
>>> FAIL: time.test: strftime: C99 %z format: EST+5
>>
>> Do you have any more details on this test?
>>
>>> (2) with -m32
>>> Failed with Segmentation Fault as:
>>> make[3]: Entering directory `/tmp/guile-2.0.5.100-0f6dd'
>>> Testing /tmp/guile-2.0.5.100-0f6dd/meta/guile ...
>>> with GUILE_LOAD_PATH=/tmp/guile-2.0.5.100-0f6dd/test-suite
>>> /bin/bash: line 5: 14661 Segmentation Fault  (core dumped) 
>>> CHARSETALIASDIR="/tmp/guile-2.0.5.100-0f6dd/lib" ${dir}$tst
>>> FAIL: check-guile
>>>
>>> This is less inportant because it occurs in "make check" phase, not in 
>>> "make" phase.
>>>
>>>   [1] __cexp(0x0, 0x0, 0xf01b866e, 0x400921f9, 0x6, 0x89b1f20), at 
>>> 0xfe801928 
>>> =>[2] scm_exp(z = (nil)), line 167 in "numbers.c"
>>
>> What can the meaning of this be?
>>
>> Regards,
>>
>> Andy

bug#11083: guile-2.0.5 bug report

[Andy Wingo]

Hi Kiyoshi,

We're following up on some bugs in anticipation of a 2.0.8 release.  Can
you try the following tarball:

  http://wingolog.org/priv/guile-2.0.7.194-dfd1d.tar.gz

Git from stable-2.0 is fine as well.

Thanks!

Andy

On Thu 05 Jul 2012 23:24, Andy Wingo wi...@pobox.com writes:

 On Fri 13 Apr 2012 18:23, Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp 
 writes:

 I downloaded the latest snapshot guile-2.0.5.100-0f6dd and tryed making.
 The result is:
 (1) with -m64
 make passed, and make check  also passed except for 1 FAIL:
 FAIL: time.test: strftime: C99 %z format: EST+5

 Do you have any more details on this test?

 (2) with -m32
 Failed with Segmentation Fault as:
 make[3]: Entering directory `/tmp/guile-2.0.5.100-0f6dd'
 Testing /tmp/guile-2.0.5.100-0f6dd/meta/guile ...
 with GUILE_LOAD_PATH=/tmp/guile-2.0.5.100-0f6dd/test-suite
 /bin/bash: line 5: 14661 Segmentation Fault  (core dumped) 
 CHARSETALIASDIR=/tmp/guile-2.0.5.100-0f6dd/lib ${dir}$tst
 FAIL: check-guile

 This is less inportant because it occurs in make check phase, not in 
 make phase.

   [1] __cexp(0x0, 0x0, 0xf01b866e, 0x400921f9, 0x6, 0x89b1f20), at 
 0xfe801928 
 =[2] scm_exp(z = (nil)), line 167 in numbers.c

 What can the meaning of this be?

 Regards,

 Andy

-- 
http://wingolog.org/

bug#11083: guile-2.0.5 bug report

[Andy Wingo]

Hi Kiyoshi,

Did you make any progress here?

On Fri 13 Apr 2012 18:23, Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:

 I downloaded the latest snapshot guile-2.0.5.100-0f6dd and tryed making.
 The result is:
 (1) with -m64
 make passed, and make check  also passed except for 1 FAIL:
 FAIL: time.test: strftime: C99 %z format: EST+5

Do you have any more details on this test?

 (2) with -m32
 Failed with Segmentation Fault as:
 make[3]: Entering directory `/tmp/guile-2.0.5.100-0f6dd'
 Testing /tmp/guile-2.0.5.100-0f6dd/meta/guile ...
 with GUILE_LOAD_PATH=/tmp/guile-2.0.5.100-0f6dd/test-suite
 /bin/bash: line 5: 14661 Segmentation Fault  (core dumped) 
 CHARSETALIASDIR=/tmp/guile-2.0.5.100-0f6dd/lib ${dir}$tst
 FAIL: check-guile

 This is less inportant because it occurs in make check phase, not in make 
 phase.

   [1] __cexp(0x0, 0x0, 0xf01b866e, 0x400921f9, 0x6, 0x89b1f20), at 0xfe801928 
 =[2] scm_exp(z = (nil)), line 167 in numbers.c

What can the meaning of this be?

Regards,

Andy
-- 
http://wingolog.org/

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Ludovic,

mem_iconveh () returns ENOSYS, in striconveh.c line 1139.
This is out of #if HAVE_ICONV but in #else.

Maybe configure did not find iconv.h.

I built guile again with configure options
  --with-libiconv-prefix=
  --with-libunistring-prefix=
  --with-libltdl-prefix=
  --with-libgmp-prefix=
  --with-libintl-prefix=
and set all the include pathes to CPATH environment for gcc.
With these, make passed without any problem.

I downloaded the latest snapshot guile-2.0.5.100-0f6dd and tryed making.
The result is:
(1) with -m64
make passed, and make check  also passed except for 1 FAIL:
FAIL: time.test: strftime: C99 %z format: EST+5

(2) with -m32
Failed with Segmentation Fault as:
make[3]: Entering directory `/tmp/guile-2.0.5.100-0f6dd'
Testing /tmp/guile-2.0.5.100-0f6dd/meta/guile ...
with GUILE_LOAD_PATH=/tmp/guile-2.0.5.100-0f6dd/test-suite
/bin/bash: line 5: 14661 Segmentation Fault  (core dumped) 
CHARSETALIASDIR=/tmp/guile-2.0.5.100-0f6dd/lib ${dir}$tst
FAIL: check-guile

This is less inportant because it occurs in make check phase, not in make 
phase.
But if you do not mind helping me, I'll try to find what causes this.
Backtrace is attached as backtrace.txt.

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Fri, 2012/4/13, Ludovic Courtès l...@gnu.org wrote:

 Hi,
 
 Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp skribis:
 
  % grep ac_cv_libunistring_with_iconv_support config.log
  ac_cv_libunistring_with_iconv_support=yes
 
 In that case, could you step through ‘scm_to_stringn’ and its
 ‘mem_iconveh’ call to see where that error value comes from?
 
 To that end, run Guile from the build tree with
 ‘./meta/gdb-uninstalled-guile’ (substitute “gdb” in that script with the
 actual debugger command name, if needed.)
 
 Thanks a lot for your help and perseverance!
 
 Ludo’.
  [1] __cexp(0x0, 0x0, 0xf01b866e, 0x400921f9, 0x6, 0x89b1f20), at 0xfe801928 
=[2] scm_exp(z = (nil)), line 167 in numbers.c
  [3] vm_debug_engine(vm = 0x80fb690, program = 0x80dcad0, argv = 0x8046f14, 
nargs = 1), line 892 in vm-i-system.c
  [4] scm_c_vm_run(vm = (nil), program = (nil), argv = (nil), nargs = 0), line 
586 in vm.c
  [5] scm_primitive_eval(exp = 0x9a93378), line 685 in eval.c
  [6] scm_primitive_load(filename = (nil)), line 132 in load.c
  [7] vm_debug_engine(vm = 0x80fb690, program = 0xfef67930, argv = 0x8047124, 
nargs = 1), line 892 in vm-i-system.c
  [8] scm_c_vm_run(vm = (nil), program = (nil), argv = (nil), nargs = 0), line 
586 in vm.c
  [9] scm_primitive_eval(exp = 0x81e9458), line 685 in eval.c
  [10] scm_eval(exp = (nil), module_or_state = (nil)), line 719 in eval.c
  [11] scm_shell(argc = 11, argv = (nil)), line 441 in script.c
  [12] invoke_main_func(body_data = (nil)), line 338 in init.c
  [13] c_body(d = (nil)), line 511 in continuations.c
  [14] apply_catch_closure(clo = (nil), args = 0x304), line 146 in throw.c
  [15] apply_1(smob = (nil), a = (nil)), line 141 in smob.c
  [16] vm_regular_engine(vm = 0x80fb690, program = 0x80f50a0, argv = 0x80473c0, 
nargs = 2), line 895 in vm-i-system.c
  [17] scm_c_vm_run(vm = (nil), program = (nil), argv = (nil), nargs = 0), line 
586 in vm.c
  [18] scm_call_4(proc = (nil), arg1 = (nil), arg2 = (nil), arg3 = (nil), arg4 
= (nil)), line 507 in eval.c
  [19] scm_catch_with_pre_unwind_handler(key = 0x404, thunk = 0x81d6ea0, 
handler = 0x81d6e90, pre_unwind_handler = (nil)), line 86 in throw.c
r_data = 0x80fb650), line 213 in throw.c
  [21] scm_i_with_continuation_barrier(body = (nil), body_data = (nil), handler 
= (nil), handler_data = (nil), pre_unwind_handler = (nil), 
pre_unwind_handler_data = (nil)), line 449 in continuations.c
  [22] scm_c_with_continuation_barrier(func = (nil), data = (nil)), line 545 in 
continuations.c
  [23] with_guile_and_parent(base = 0x80474f0, data = (nil)), line 905 in 
threads.c
  [24] GC_call_with_stack_base(fn = 0xfeef2310 = 
`libguile-2.0.so.22.4.0`threads.c`with_guile_and_parent(struct GC_stack_base 
*base, register void *data), arg = (nil)), line 1530 in misc.c
  [25] scm_i_with_guile_and_parent(func = (nil), data = (nil), parent = (nil)), 
line 948 in threads.c
  [26] scm_boot_guile(argc = 0, argv = (nil), main_func = (nil), closure = 
(nil)), line 321 in init.c
  [27] main(argc = 0, argv = (nil)), line 71 in guile.c

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi,

--- On Thu, 2012/4/12, Ludovic Courtès l...@gnu.org wrote:

 Hi,
 
 Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp skribis:
 
  Whole the backtrace is:
    [1] __lwp_kill(0x1, 0x6, 0x89b048a0, 0x7fc5338e, 
 0x6f00312e6f732e63, 0x612f6e00312e), at 0x7fc52a9a 
    [2] _thr_kill(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fc4b393 
    [3] raise(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fbf7d59 
    [4] abort(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fbd68c0 
  =[5] pre_init_throw(k = 0x45b4c0, args = 0x4e2220), line 526 in throw.c
    [6] vm_regular_engine(vm = 0x4db270, program = 0x4d6320, argv = 0x1, 
 nargs = 2), line 895 in vm-i-system.c
    [7] scm_c_vm_run(vm = 0x4db270, program = 0x4d6320, argv = 
 0xfd7fffdfe130, nargs = 6), line 567 in vm.c
    [8] scm_call_with_vm(vm = 0x4db270, proc = 0x4d6320, args = 0x4db270), 
 line 859 in vm.c
    [9] scm_to_stringn(str = 0x4e3620, lenp = (nil), encoding = 
 0x7ff05411 ISO-8859-1, handler = SCM_FAILED_CONVERSION_QUESTION_MARK), 
 line 1966 in strings.c
 
 This seems to indicate a failure to convert STR from ISO-8859-1, with a
 failed attempt to raise an exception.
 
 Can you try running Guile under the “C” locale?

Last one was also ran with environment LANG = C
 
   $ LANG=C ./meta/guile -c 1

$  $ LANG=C ./meta/guile -c 1
Throw without catch before boot:
Throw to key encoding-error with args (scm_to_stringn cannot convert narrow 
string to output locale 89 #f #f)Aborting.
Abort


I found /usr/lib/locale/ja/amd64 in my LD_LIBRARY_PATH_64 and removed it, but 
the result is same.

Regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

bug#11083: guile-2.0.5 bug report

[Ludovic Courtès]

Hi,

Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp skribis:

 $  $ LANG=C ./meta/guile -c 1
 Throw without catch before boot:
 Throw to key encoding-error with args (scm_to_stringn cannot convert 
 narrow string to output locale 89 #f #f)Aborting.

On a Solaris box, I found this:

   grep 89 /usr/include/*/errno*  
  /usr/include/sys/errno.h:/* Copyright (c) 1984, 1986, 1987, 1988, 1989 
ATT */
  /usr/include/sys/errno.h: * (c) 1983,1984,1985,1986,1987,1988,1989  ATT.
  /usr/include/sys/errno.h:#defineENOSYS  89  /* Unsupported file 
system operation*/

This errors comes from the ‘mem_iconveh’ call, which, according to
striconveh.c, means that libunistring was compiled without iconv
support.  Can you check whether this is the case?

This problem is normally caught by Guile’s configure.  Could you check
the value of ‘ac_cv_libunistring_with_iconv_support’ in Guile’s
‘config.log’?

Thanks,
Ludo’.

bug#11083: guile-2.0.5 bug report

[Ludovic Courtès]

Hi,

Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp skribis:

 % grep ac_cv_libunistring_with_iconv_support config.log
 ac_cv_libunistring_with_iconv_support=yes

In that case, could you step through ‘scm_to_stringn’ and its
‘mem_iconveh’ call to see where that error value comes from?

To that end, run Guile from the build tree with
‘./meta/gdb-uninstalled-guile’ (substitute “gdb” in that script with the
actual debugger command name, if needed.)

Thanks a lot for your help and perseverance!

Ludo’.

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi,

% grep ac_cv_libunistring_with_iconv_support config.log
ac_cv_libunistring_with_iconv_support=yes

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Thu, 2012/4/12, Ludovic Courtès l...@gnu.org wrote:

 Hi,
 
 Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp skribis:
 
  $  $ LANG=C ./meta/guile -c 1
  Throw without catch before boot:
  Throw to key encoding-error with args (scm_to_stringn cannot convert 
  narrow string to output locale 89 #f #f)Aborting.
 
 On a Solaris box, I found this:
 
    grep 89 /usr/include/*/errno*                                             
  
   /usr/include/sys/errno.h:/*     Copyright (c) 1984, 1986, 1987, 1988, 1989 
 ATT */
   /usr/include/sys/errno.h: *     (c) 1983,1984,1985,1986,1987,1988,1989  
 ATT.
   /usr/include/sys/errno.h:#define        ENOSYS  89      /* Unsupported file 
 system operation    */
 
 This errors comes from the ‘mem_iconveh’ call, which, according to
 striconveh.c, means that libunistring was compiled without iconv
 support.  Can you check whether this is the case?
 
 This problem is normally caught by Guile’s configure.  Could you check
 the value of ‘ac_cv_libunistring_with_iconv_support’ in Guile’s
 ‘config.log’?
 
 Thanks,
 Ludo’.

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Mark,

I was reconstructing GNU software building environment.
I put 32 bit code in /usr/local/GNU and 64bit ones in /usr/local/GNU/amd64.
I sometimes had to specify -I/usr/local/GNU/amd64/include 
-L/usr/local/GNU/amd64/lib,
but it was harmful.
Now I do not have to use them in almost softwares.

As the result of this, backtrace of guile has got easy to understand.
Segmentation Fault happens in line 895 of vm-i-system.c,
  ret = subr (sp[-1], sp[0]);

Whole the backtrace is:
  [1] __lwp_kill(0x1, 0x6, 0x89b048a0, 0x7fc5338e, 
0x6f00312e6f732e63, 0x612f6e00312e), at 0x7fc52a9a 
  [2] _thr_kill(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fc4b393 
  [3] raise(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fbf7d59 
  [4] abort(0x0, 0x0, 0x0, 0x0, 0x0, 0x0), at 0x7fbd68c0 
=[5] pre_init_throw(k = 0x45b4c0, args = 0x4e2220), line 526 in throw.c
  [6] vm_regular_engine(vm = 0x4db270, program = 0x4d6320, argv = 0x1, nargs = 
2), line 895 in vm-i-system.c
  [7] scm_c_vm_run(vm = 0x4db270, program = 0x4d6320, argv = 
0xfd7fffdfe130, nargs = 6), line 567 in vm.c
  [8] scm_call_with_vm(vm = 0x4db270, proc = 0x4d6320, args = 0x4db270), line 
859 in vm.c
  [9] scm_to_stringn(str = 0x4e3620, lenp = (nil), encoding = 0x7ff05411 
ISO-8859-1, handler = SCM_FAILED_CONVERSION_QUESTION_MARK), line 1966 in 
strings.c
  [10] search_path(path = 0x4e2360, filename = 0x4e3620, extensions = 0x304, 
require_exts = 0x4, stat_buf = 0xfd7fffdfe330), line 444 in load.c
  [11] scm_init_eval_in_scheme(), line 971 in load.c
  [12] scm_i_init_guile(base = 0x7ff04bb4), line 499 in init.c
  [13] scm_i_init_thread_for_guile(base = 0xfd7fffdfe460, parent = (nil)), 
line 829 in threads.c
  [14] with_guile_and_parent(base = 0xfd7fffdfe460, data = (nil)), line 895 
in threads.c
  [15] GC_call_with_stack_base(fn = (nil), arg = (nil)), line 1530 in misc.c
  [16] scm_i_with_guile_and_parent(func = (nil), data = (nil), parent = (nil)), 
line 945 in threads.c
  [17] scm_boot_guile(argc = 0, argv = (nil), main_func = (nil), closure = 
(nil)), line 319 in init.c
  [18] main(argc = 0, argv = (nil)), line 71 in guile.c

Regard,

-- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Wed, 2012/4/11, Mark H Weaver m...@netris.org wrote:

 Hi Kiyoshi,
 
 I wrote:
  It appears that 'environ_locale_charset' returns a bad address on
  Solaris 10, but only when Guile is compiled with '-m64'.
 
 Kiyoshi: would you be willing to help investigate further?  I would
 recommend running the 64-bit version of Guile in a debugger, setting a
 breakpoint on 'environ_locale_charset' and stepping through that
 function to find out more precisely where the bad pointer originates.
 
     Regards,
       Mark
 
 
  Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
  =[1] libunistring_c_strcasecmp(s1 = value unavailable, s2 =
  value unavailable) (optimized), at 0x7fa102a8 (line ~39) in
  c-strcasecmp.c
    [2] u8_conv_from_encoding(fromcode = 0xfff56ea0 bad
  address 0xfff56ea0, handler = iconveh_escape_sequence, src
  = 0xfd7fffdff6ea snarf-check-and-output-texi, srclen = 27U,
  offsets = (nil), resultbuf = (nil), lengthp = 0xfd7fffdfecf8)
  (optimized), at 0x7fa1ce80 (line ~39) in u8-conv-from-enc.c
    [3] u32_conv_from_encoding(fromcode = value unavailable, handler
  = value unavailable, src = value unavailable, srclen = 27U,
  offsets = (nil), resultbuf = (nil), lengthp = 0xfd7fffdfed60)
  (optimized), at 0x7fa1c730 (line ~52) in u-conv-from-enc.h
    [4] scm_from_stringn(str = 0xfd7fffdff6ea
  snarf-check-and-output-texi, len = 27U, encoding =
  0xfff56ea0 bad address 0xfff56ea0, handler =
  SCM_FAILED_CONVERSION_ESCAPE_SEQUENCE), line 1515 in strings.c
 
  Note the bad 'encoding' parameter above.
 
    [5] locale_arguments_to_string_list(argc = 5, argv = 
 0xfd7fffdff478), line 394 in script.c
    [6] scm_i_set_boot_program_arguments(argc = 0, argv = (nil)), line 405 
 in script.c
    [7] invoke_main_func(body_data = 0xfd7fffdff420), line 335 in 
 init.c
    [8] c_body(d = 0xfd7fffdff370), line 512 in continuations.c
    [9] vm_regular_engine(vm = 0x4da270, program = 0x3679880, argv = 
 0x7fec7210, nargs = 1), line 960 in vm-i-system.c
    [10] scm_c_vm_run(vm = 0x4da270, program = 0x3679a80, argv = 
 0xfd7fffdfef70, nargs = 0), line 567 in vm.c
    [11] eval(x = 0x589570, env = 0x3681820), line 341 in eval.c
    [12] eval(x = 0x5893b0, env = 0x3681820), line 297 in eval.c
    [13] eval(x = 0x4f0920, env = 0x36817f0), line 462 in eval.c
    [14] boot_closure_apply(closure = (nil), args = (nil)), line 956 in 
 eval.c
    [15] vm_regular_engine(vm = 0x4da270, program = 0x36799e0, argv = 
 0x7fe53260, nargs = 1), line 960 in vm-i-system.c
    [16] scm_c_vm_run(vm = 0x4da270, program = 0x5011c0, argv = 
 0xfd7fffdff2c0, nargs = 4), line 567 in vm.c
    [17] scm_call_4(proc = 0x5011c0, arg1 = (nil), arg2 = (nil), arg3 = 
 (nil), arg4 = (nil)), line 507 in 

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Mark

I installed the latest gdb-7.4 to run 'meta/gdb-uninstalled-guile', but failed.
% meta/gdb-uninstalled-guile
GNU gdb (GDB) 7.4
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type show copying
and show warranty for details.
This GDB was configured as i386-pc-solaris2.10.
For bug reporting instructions, please see:
http://www.gnu.org/software/gdb/bugs/...
Reading symbols from /tmp/guile-2.0.5/libguile/.libs/guile...done.
(gdb) run
Starting program: /tmp/guile-2.0.5/libguile/.libs/guile 
procfs:4200 -- process not stopped.
procfs: ...giving up...

We still have the same problem as 
http://sources.redhat.com/ml/gdb-prs/2011-q4/msg00050.html in the latest gdb.


I used dbx debugger instead.
At first, back traced libguile/core (See BackTrace1.txt attached).
After that, executed meta/uninstalled-env guild snarf-check-and-output-texi 
from the build directory wirh ENVIRONMENT GUILE_AUTO_COMPILE = 0, and  found 
core dumped again.
Back traced this new core again (See BackTrace2.txt) to compare.
The function scm_from_stringn () is called with bad address.

Keep in mind that patch 5de0053178b4acc793ae62838175e5f3ab56c603 is appyed.

Regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Thu, 2012/4/5, Mark H Weaver m...@netris.org wrote:

 Presumably something is broken for 64-bit builds on Solaris 10.  I
 suggest using a debugger to find out where the segfault is happening.
 Can you please run 'meta/gdb-uninstalled-guile' within the build
 directory and show us the backtrace when it segfaults?
 
     Regards,
       Mark
$ pwd
/tmp/guile-2.0.5 
$ ldd libguile/.libs/guile
libguile-2.0.so.22 =
/tmp/guile-2.0.5/libguile/.libs/libguile-2.0.so.22
libgc.so.1 =/usr/local/GNU/amd64/lib/libgc.so.1
libpthread.so.1 =   /usr/lib/amd64/libpthread.so.1
libdl.so.1 =/usr/lib/amd64/libdl.so.1
libffi.so.5 =   /usr/local/GNU/amd64/lib/libffi.so.5
libintl.so.8 =  /usr/local/GNU/amd64/lib/libintl.so.8
libc.so.1 = /usr/lib/amd64/libc.so.1
libunistring.so.0 = /usr/local/GNU/amd64/lib/libunistring.so.0
libiconv.so.2 = /usr/local/GNU/amd64/lib/libiconv.so.2
libgmp.so.10 =  /usr/local/GNU/amd64/lib/libgmp.so.10
libltdl.so.7 =  /usr/local/GNU/amd64/lib/libltdl.so.7
librt.so.1 =/usr/lib/amd64/librt.so.1
libsocket.so.1 =/usr/lib/amd64/libsocket.so.1
libnsl.so.1 =   /usr/lib/amd64/libnsl.so.1
libm.so.2 = /usr/lib/amd64/libm.so.2
libgcc_s.so.1 = /usr/local/gcc/lib/amd64/libgcc_s.so.1
libaio.so.1 =   /usr/lib/amd64/libaio.so.1
libmd.so.1 =/usr/lib/amd64/libmd.so.1
libmp.so.2 =/usr/lib/amd64/libmp.so.2
libscf.so.1 =   /usr/lib/amd64/libscf.so.1
libdoor.so.1 =  /usr/lib/amd64/libdoor.so.1
libuutil.so.1 = /usr/lib/amd64/libuutil.so.1
libgen.so.1 =   /usr/lib/amd64/libgen.so.1
$ dbx libguile/.libs/guile libguile/core
For information about new features see `help changes'
To remove this message, put `dbxenv suppress_startup_message 7.9' in your .dbxrc
Reading guile
core file header read successfully
Reading ld.so.1
Reading libguile-2.0.so.22.4.0
Reading libgc.so.1.0.3
Reading libpthread.so.1
Reading libdl.so.1
Reading libffi.so.5.0.10
Reading libintl.so.8.1.1
Reading libc.so.1
Reading libunistring.so.0.1.2
Reading libiconv.so.2.5.1
Reading libgmp.so.10.0.4
dbx: internal warning: (null)::srcpath(): srcpath abs.c being overriden to 
abs.c
dbx: internal warning: (null)::srcpath(): srcpath add.c being overriden to 
add.c
dbx: internal warning: (null)::srcpath(): srcpath add_ui.c being overriden 
to add_ui.c
dbx: internal warning: (null)::srcpath(): srcpath clear.c being overriden 
to clear.c
dbx: internal warning: (null)::srcpath(): srcpath clears.c being overriden 
to clears.c
dbx: internal warning: (null)::srcpath(): srcpath cmp.c being overriden to 
cmp.c
dbx: internal warning: (null)::srcpath(): srcpath cmp_d.c being overriden 
to cmp_d.c
dbx: internal warning: (null)::srcpath(): srcpath cmp_si.c being overriden 
to cmp_si.c
dbx: internal warning: (null)::srcpath(): srcpath cmp_ui.c being overriden 
to cmp_ui.c
dbx: internal warning: (null)::srcpath(): srcpath dump.c being overriden to 
dump.c
dbx: internal warning: (null)::srcpath(): srcpath fits_sint.c being 
overriden to fits_sint.c
dbx: internal warning: (null)::srcpath(): srcpath fits_slong.c being 
overriden to fits_slong.c
dbx: internal warning: (null)::srcpath(): srcpath fits_sshort.c being 
overriden to fits_sshort.c
dbx: internal warning: (null)::srcpath(): srcpath fits_uint.c being 
overriden to fits_uint.c
dbx: internal warning: (null)::srcpath(): srcpath fits_ulong.c being 
overriden to fits_ulong.c
dbx: internal warning: 

bug#11083: guile-2.0.5 bug report

[Mark H Weaver]

Hi Ludovic,

It appears that 'environ_locale_charset' returns a bad address on
Solaris 10, but only when Guile is compiled with '-m64'.  Kiyoshi says
that 5de0053178b4acc793ae62838175e5f3ab56c603 is applied in his tree.

Do you have time to look into this?

 Regards,
   Mark


Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
 =[1] libunistring_c_strcasecmp(s1 = value unavailable, s2 = value 
 unavailable) (optimized), at 0x7fa102a8 (line ~39) in c-strcasecmp.c
   [2] u8_conv_from_encoding(fromcode = 0xfff56ea0 bad address 
 0xfff56ea0, handler = iconveh_escape_sequence, src = 
 0xfd7fffdff6ea snarf-check-and-output-texi, srclen = 27U, offsets = 
 (nil), resultbuf = (nil), lengthp = 0xfd7fffdfecf8) (optimized), at 
 0x7fa1ce80 (line ~39) in u8-conv-from-enc.c
   [3] u32_conv_from_encoding(fromcode = value unavailable, handler = value 
 unavailable, src = value unavailable, srclen = 27U, offsets = (nil), 
 resultbuf = (nil), lengthp = 0xfd7fffdfed60) (optimized), at 
 0x7fa1c730 (line ~52) in u-conv-from-enc.h
   [4] scm_from_stringn(str = 0xfd7fffdff6ea 
 snarf-check-and-output-texi, len = 27U, encoding = 0xfff56ea0 bad 
 address 0xfff56ea0, handler = 
 SCM_FAILED_CONVERSION_ESCAPE_SEQUENCE), line 1515 in strings.c

Note the bad 'encoding' parameter above.

   [5] locale_arguments_to_string_list(argc = 5, argv = 0xfd7fffdff478), 
 line 394 in script.c
   [6] scm_i_set_boot_program_arguments(argc = 0, argv = (nil)), line 405 in 
 script.c
   [7] invoke_main_func(body_data = 0xfd7fffdff420), line 335 in init.c
   [8] c_body(d = 0xfd7fffdff370), line 512 in continuations.c
   [9] vm_regular_engine(vm = 0x4da270, program = 0x3679880, argv = 
 0x7fec7210, nargs = 1), line 960 in vm-i-system.c
   [10] scm_c_vm_run(vm = 0x4da270, program = 0x3679a80, argv = 
 0xfd7fffdfef70, nargs = 0), line 567 in vm.c
   [11] eval(x = 0x589570, env = 0x3681820), line 341 in eval.c
   [12] eval(x = 0x5893b0, env = 0x3681820), line 297 in eval.c
   [13] eval(x = 0x4f0920, env = 0x36817f0), line 462 in eval.c
   [14] boot_closure_apply(closure = (nil), args = (nil)), line 956 in eval.c
   [15] vm_regular_engine(vm = 0x4da270, program = 0x36799e0, argv = 
 0x7fe53260, nargs = 1), line 960 in vm-i-system.c
   [16] scm_c_vm_run(vm = 0x4da270, program = 0x5011c0, argv = 
 0xfd7fffdff2c0, nargs = 4), line 567 in vm.c
   [17] scm_call_4(proc = 0x5011c0, arg1 = (nil), arg2 = (nil), arg3 = (nil), 
 arg4 = (nil)), line 507 in eval.c
   [18] scm_i_with_continuation_barrier(body = 0x7fe49e70 = 
 `libguile-2.0.so.22.4.0`continuations.c`c_body(register void *d), body_data 
 = 0xfd7fffdff370, handler = 0x7fe49ea0 = 
 `libguile-2.0.so.22.4.0`continuations.c`c_handler(register void *d, register 
 SCM tag, register SCM args), handler_data = 0xfd7fffdff370, 
 pre_unwind_handler = 0x7fe49f00 = 
 `libguile-2.0.so.22.4.0`continuations.c`pre_unwind_handler(register void 
 *error_port, register SCM tag, register SCM args), pre_unwind_handler_data = 
 0x4da1a0), line 450 in continuations.c
   [19] scm_c_with_continuation_barrier(func = (nil), data = (nil)), line 546 
 in continuations.c
   [20] with_guile_and_parent(base = 0xfd7fffdff3d0, data = (nil)), line 
 902 in threads.c
   [21] GC_call_with_stack_base(fn = (nil), arg = (nil)), line 1530 in misc.c
   [22] scm_i_with_guile_and_parent(func = (nil), data = (nil), parent = 
 (nil)), line 945 in threads.c
   [23] scm_boot_guile(argc = 0, argv = (nil), main_func = (nil), closure = 
 (nil)), line 319 in init.c
   [24] main(argc = 0, argv = (nil)), line 71 in guile.c
 (dbx)

bug#11083: guile-2.0.5 bug report

[Mark H Weaver]

Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
 I analysed the problems.
 (1) Why can not link libguile-2.0.so.22
 LD_LIBRARY_PATH_64 is not set correctly.
 I changed LD_LIBRARY_PATH to LD_LIBRARY_PATH_64 in all the files
 in guile-2.0.5, and found libguile-2.0.so.22 can be linked.

This is a libtool issue.  See:

  
http://news.gmane.org/find-root.php?group=gmane.comp.gnu.libtool.bugsarticle=7988

 (2) About Segmentation Fault
 This seems to be  same to
 http://lists.gnu.org/archive/html/bug-guile/2011-03/msg00111.html.

This is the first time Guile is run during the build process, therefore
there is no evidence that these two problems are related.

 I do not know why it occurs only with -m64 option.
 (I have no problem with -m32).

Presumably something is broken for 64-bit builds on Solaris 10.  I
suggest using a debugger to find out where the segfault is happening.
Can you please run 'meta/gdb-uninstalled-guile' within the build
directory and show us the backtrace when it segfaults?

Regards,
  Mark

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Mark,

I analysed the problems.
(1) Why can not link libguile-2.0.so.22
LD_LIBRARY_PATH_64 is not set correctly.
I changed  LD_LIBRARY_PATH to LD_LIBRARY_PATH_64 in all the files in 
guile-2.0.5, and found libguile-2.0.so.22 can be linked.

(2) About Segmentation Fault
This seems to be  same to
http://lists.gnu.org/archive/html/bug-guile/2011-03/msg00111.html.
I do not know why it occurs only with -m64 option.
(I have no problem with -m32).

Regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Fri, 2012/3/30, Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp wrote:

 Hi, Mark,
 
 Thank you for your kindness.
 
 I'm discussing with a member of gc project these 5 days.
 He also told me to configure with --enable-threads=posix instead of 
 -DGC_THREADS.
 
 And I'm also testing the latest gc-7.2 candidate and gc-7.3 (developement) 
 candidate.
 
 
 Now, I'm retrying to build guile-2.0.5 again.
 Make  make check pass with gcc-3.4.3 -m32, but make fail with -m64.
 Message of make check is:as follows.
 (libguile-2.0.so.22 is ELF 64-bit LSB executable, so that it can be linked.)
 
 make  all-am
 make[3]: Entering directory `/tmp/guile-2.0.5/libguile'
   GEN    guile-procedures.texi
 ld.so.1: guile: fatal: libguile-2.0.so.22: open failed: No such file or 
 directory
 /bin/bash: line 1: 23126 Broken Pipe             cat alist.doc arbiters.doc 
 array-handle.doc array-map.doc arrays.doc async.doc backtrace.doc boolean.doc 
 bitvectors.doc bytevectors.doc chars.doc control.doc continuations.doc 
 debug.doc deprecated.doc deprecation.doc dynl.doc dynwind.doc eq.doc 
 error.doc eval.doc evalext.doc expand.doc extensions.doc feature.doc 
 filesys.doc fluids.doc foreign.doc fports.doc gc-malloc.doc gc.doc 
 gettext.doc generalized-arrays.doc generalized-vectors.doc goops.doc 
 gsubr.doc guardians.doc hash.doc hashtab.doc hooks.doc i18n.doc init.doc 
 ioext.doc keywords.doc list.doc load.doc macros.doc mallocs.doc memoize.doc 
 modules.doc numbers.doc objprop.doc options.doc pairs.doc ports.doc print.doc 
 procprop.doc procs.doc promises.doc r6rs-ports.doc random.doc rdelim.doc 
 read.doc root.doc rw.doc scmsigs.doc script.doc simpos.doc smob.doc sort.doc 
 srcprop.doc srfi-1.doc srfi-4.doc srfi-13.doc srfi-14.doc srfi-60.doc 
 stackchk.doc
  stacks.doc stime.doc strings.doc strorder.doc strports.doc struct.doc 
 symbols.doc threads.doc throw.doc trees.doc uniform.doc values.doc 
 variable.doc vectors.doc version.doc vports.doc weaks.doc dynl.doc posix.doc 
 net_db.doc socket.doc regex-posix.doc
      23127 Killed                  | GUILE_AUTO_COMPILE=0 
 ../meta/uninstalled-env guild snarf-check-and-output-texi  
 guile-procedures.texi
 make[3]: *** [guile-procedures.texi] Error 1
 
 Regards,
 
 --- Kiyoshi yoi_no_myou...@yahoo.co.jp
 
 --- On Thu, 2012/3/29, Mark H Weaver m...@netris.org wrote:
 
  Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
   Thank you Mark,
   Sending this again with Cc:
  
   (1) Now, make succeeds with threads
   Version of libgc is 7.1, but I found some macro define is required for 
   threads on Solaris.
   After re-install gc-7.1 with -DGC_SOLARIS_THREADS or -DGC_THREADS,
   make succeeds.
  
  This is not my area of expertise, but for lack of other suggestions, I
  have two ideas:
  
  1. You could try recompiling libgc using POSIX threads instead of
     Solaris threads, and then rebuilding Guile from scratch, starting
     with configure, and linking against the new libgc.
  
     Guile supports only POSIX threads, and I wouldn't be surprised if
     there are problems associated with linking Guile against a libgc that
     uses a different threading library with slightly different semantics.
  
  2. You could try a newer version of libgc: 7.2alpha6, from
     http://www.hpl.hp.com/personal/Hans_Boehm/gc/gc_source/gc-7.2alpha6.tar.gz
  
     Despite the scary version number, Hans Boehm himself now recommends
     this version because 7.1 is so old (see Where to get the collector
     at http://www.hpl.hp.com/personal/Hans_Boehm/gc/ )
  
      Thanks,
        Mark
  
  
   (2) Segmentation Fault occur in make check
   Still encounterd FAILS as follows:
   make  check-TESTS
   make[5]: Entering directory `/tmp/guile-2.0.5/test-suite/standalone'
   Assertion failed: SCM_I_IS_THREAD (data.thread), file threads.c, line 1135
   /bin/bash: line 5: 11634 Abort                   srcdir=. builddir=. 
   CHARSETALIASDIR
   =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
   ${dir}$tst
   FAIL: test-system-cmds
     :
   PASS: test-extensions
   /bin/bash: line 5: 12409 Segmentation Fault      srcdir=. builddir=. 
   CHARSETALIASDIR
   =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
   ${dir}$tst
   FAIL: test-with-guile-module
   PASS: test-scm-with-guile
   /bin/bash: line 5: 12455 Segmentation Fault      srcdir=. builddir=. 
   CHARSETALIASDIR
   =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
   ${dir}$tst
   FAIL: 

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Mark,

I'd better mention:
this Segmentation Fault occurs even if patch 
5de0053178b4acc793ae62838175e5f3ab56c603 is applied.

Regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp


--- On Sun, 2012/4/1, Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp wrote:

 Hi, Mark,
 
 I found scm_from_stringn () is called with bad address.
 Following is the back trace of core which I encountered during making 
 guile-2.0.5 with gcc -m64. (gcc version is 3.4.3)
 
 % pwd
 /tmp/guile-2.0.5/libguile/.libs 
 % dbx guile ../core
 For information about new features see `help changes'
 To remove this message, put `dbxenv suppress_startup_message 7.9' in your 
 .dbxrc
 Reading guile
 core file header read successfully
 Reading ld.so.1
 Reading libguile-2.0.so.22.4.0
 Reading libgc.so.1.0.3
 Reading libpthread.so.1
 Reading libdl.so.1
 Reading libffi.so.5.0.10
 Reading libintl.so.8.1.1
 Reading libc.so.1
 Reading libunistring.so.0.1.2
 Reading libiconv.so.2.5.1
 Reading libgmp.so.10.0.4
 dbx: internal warning: (null)::srcpath(): srcpath abs.c being overriden 
 to abs.c
 dbx: internal warning: (null)::srcpath(): srcpath add.c being overriden 
 to add.c
 dbx: internal warning: (null)::srcpath(): srcpath add_ui.c being 
 overriden to add_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath clear.c being overriden 
 to clear.c
 dbx: internal warning: (null)::srcpath(): srcpath clears.c being 
 overriden to clears.c
 dbx: internal warning: (null)::srcpath(): srcpath cmp.c being overriden 
 to cmp.c
 dbx: internal warning: (null)::srcpath(): srcpath cmp_d.c being overriden 
 to cmp_d.c
 dbx: internal warning: (null)::srcpath(): srcpath cmp_si.c being 
 overriden to cmp_si.c
 dbx: internal warning: (null)::srcpath(): srcpath cmp_ui.c being 
 overriden to cmp_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath dump.c being overriden 
 to dump.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_sint.c being 
 overriden to fits_sint.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_slong.c being 
 overriden to fits_slong.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_sshort.c being 
 overriden to fits_sshort.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_uint.c being 
 overriden to fits_uint.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_ulong.c being 
 overriden to fits_ulong.c
 dbx: internal warning: (null)::srcpath(): srcpath fits_ushort.c being 
 overriden to fits_ushort.c
 dbx: internal warning: (null)::srcpath(): srcpath get_d.c being overriden 
 to get_d.c
 dbx: internal warning: (null)::srcpath(): srcpath get_d_2exp.c being 
 overriden to get_d_2exp.c
 dbx: internal warning: (null)::srcpath(): srcpath get_si.c being 
 overriden to get_si.c
 dbx: internal warning: (null)::srcpath(): srcpath get_str.c being 
 overriden to get_str.c
 dbx: internal warning: (null)::srcpath(): srcpath get_ui.c being 
 overriden to get_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath init.c being overriden 
 to init.c
 dbx: internal warning: (null)::srcpath(): srcpath init2.c being overriden 
 to init2.c
 dbx: internal warning: (null)::srcpath(): srcpath inits.c being overriden 
 to inits.c
 dbx: internal warning: (null)::srcpath(): srcpath inp_str.c being 
 overriden to inp_str.c
 dbx: internal warning: (null)::srcpath(): srcpath iset.c being overriden 
 to iset.c
 dbx: internal warning: (null)::srcpath(): srcpath iset_d.c being 
 overriden to iset_d.c
 dbx: internal warning: (null)::srcpath(): srcpath iset_si.c being 
 overriden to iset_si.c
 dbx: internal warning: (null)::srcpath(): srcpath iset_str.c being 
 overriden to iset_str.c
 dbx: internal warning: (null)::srcpath(): srcpath iset_ui.c being 
 overriden to iset_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath mul.c being overriden 
 to mul.c
 dbx: internal warning: (null)::srcpath(): srcpath mul_2exp.c being 
 overriden to mul_2exp.c
 dbx: internal warning: (null)::srcpath(): srcpath mul_ui.c being 
 overriden to mul_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath neg.c being overriden 
 to neg.c
 dbx: internal warning: (null)::srcpath(): srcpath nextprime.c being 
 overriden to nextprime.c
 dbx: internal warning: (null)::srcpath(): srcpath out_str.c being 
 overriden to out_str.c
 dbx: internal warning: (null)::srcpath(): srcpath pow_ui.c being 
 overriden to pow_ui.c
 dbx: internal warning: (null)::srcpath(): srcpath random2.c being 
 overriden to random2.c
 dbx: internal warning: (null)::srcpath(): srcpath set.c being overriden 
 to set.c
 dbx: internal warning: (null)::srcpath(): srcpath set_d.c being overriden 
 to set_d.c
 dbx: internal warning: (null)::srcpath(): srcpath set_q.c being overriden 
 to set_q.c
 dbx: internal warning: (null)::srcpath(): srcpath set_si.c being 
 overriden to set_si.c
 dbx: internal warning: (null)::srcpath(): srcpath set_str.c being 
 overriden to set_str.c
 dbx: internal warning: (null)::srcpath(): srcpath set_ui.c being 
 overriden to set_ui.c
 dbx: internal warning: 

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hi, Mark,

Thank you for your kindness.

I'm discussing with a member of gc project these 5 days.
He also told me to configure with --enable-threads=posix instead of 
-DGC_THREADS.

And I'm also testing the latest gc-7.2 candidate and gc-7.3 (developement) 
candidate.


Now, I'm retrying to build guile-2.0.5 again.
Make  make check pass with gcc-3.4.3 -m32, but make fail with -m64.
Message of make check is:as follows.
(libguile-2.0.so.22 is ELF 64-bit LSB executable, so that it can be linked.)

make  all-am
make[3]: Entering directory `/tmp/guile-2.0.5/libguile'
  GENguile-procedures.texi
ld.so.1: guile: fatal: libguile-2.0.so.22: open failed: No such file or 
directory
/bin/bash: line 1: 23126 Broken Pipe cat alist.doc arbiters.doc 
array-handle.doc array-map.doc arrays.doc async.doc backtrace.doc boolean.doc 
bitvectors.doc bytevectors.doc chars.doc control.doc continuations.doc 
debug.doc deprecated.doc deprecation.doc dynl.doc dynwind.doc eq.doc error.doc 
eval.doc evalext.doc expand.doc extensions.doc feature.doc filesys.doc 
fluids.doc foreign.doc fports.doc gc-malloc.doc gc.doc gettext.doc 
generalized-arrays.doc generalized-vectors.doc goops.doc gsubr.doc 
guardians.doc hash.doc hashtab.doc hooks.doc i18n.doc init.doc ioext.doc 
keywords.doc list.doc load.doc macros.doc mallocs.doc memoize.doc modules.doc 
numbers.doc objprop.doc options.doc pairs.doc ports.doc print.doc procprop.doc 
procs.doc promises.doc r6rs-ports.doc random.doc rdelim.doc read.doc root.doc 
rw.doc scmsigs.doc script.doc simpos.doc smob.doc sort.doc srcprop.doc 
srfi-1.doc srfi-4.doc srfi-13.doc srfi-14.doc srfi-60.doc stackchk.doc
 stacks.doc stime.doc strings.doc strorder.doc strports.doc struct.doc 
symbols.doc threads.doc throw.doc trees.doc uniform.doc values.doc variable.doc 
vectors.doc version.doc vports.doc weaks.doc dynl.doc posix.doc net_db.doc 
socket.doc regex-posix.doc
 23127 Killed  | GUILE_AUTO_COMPILE=0 
../meta/uninstalled-env guild snarf-check-and-output-texi  
guile-procedures.texi
make[3]: *** [guile-procedures.texi] Error 1

Regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

--- On Thu, 2012/3/29, Mark H Weaver m...@netris.org wrote:

 Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
  Thank you Mark,
  Sending this again with Cc:
 
  (1) Now, make succeeds with threads
  Version of libgc is 7.1, but I found some macro define is required for 
  threads on Solaris.
  After re-install gc-7.1 with -DGC_SOLARIS_THREADS or -DGC_THREADS,
  make succeeds.
 
 This is not my area of expertise, but for lack of other suggestions, I
 have two ideas:
 
 1. You could try recompiling libgc using POSIX threads instead of
    Solaris threads, and then rebuilding Guile from scratch, starting
    with configure, and linking against the new libgc.
 
    Guile supports only POSIX threads, and I wouldn't be surprised if
    there are problems associated with linking Guile against a libgc that
    uses a different threading library with slightly different semantics.
 
 2. You could try a newer version of libgc: 7.2alpha6, from
    http://www.hpl.hp.com/personal/Hans_Boehm/gc/gc_source/gc-7.2alpha6.tar.gz
 
    Despite the scary version number, Hans Boehm himself now recommends
    this version because 7.1 is so old (see Where to get the collector
    at http://www.hpl.hp.com/personal/Hans_Boehm/gc/ )
 
     Thanks,
       Mark
 
 
  (2) Segmentation Fault occur in make check
  Still encounterd FAILS as follows:
  make  check-TESTS
  make[5]: Entering directory `/tmp/guile-2.0.5/test-suite/standalone'
  Assertion failed: SCM_I_IS_THREAD (data.thread), file threads.c, line 1135
  /bin/bash: line 5: 11634 Abort                   srcdir=. builddir=. 
  CHARSETALIASDIR
  =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
  ${dir}$tst
  FAIL: test-system-cmds
    :
  PASS: test-extensions
  /bin/bash: line 5: 12409 Segmentation Fault      srcdir=. builddir=. 
  CHARSETALIASDIR
  =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
  ${dir}$tst
  FAIL: test-with-guile-module
  PASS: test-scm-with-guile
  /bin/bash: line 5: 12455 Segmentation Fault      srcdir=. builddir=. 
  CHARSETALIASDIR
  =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
  ${dir}$tst
  FAIL: test-scm-spawn-thread
  /bin/bash: line 5: 12479 Segmentation Fault      srcdir=. builddir=. 
  CHARSETALIASDIR
  =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
  ${dir}$tst
  FAIL: test-pthread-create
  SKIP: test-pthread-create-secondary
  ==
  4 of 27 tests failed
  (1 test was not run)
  Please report to bug-guile@gnu.org
  ==

bug#11083: guile-2.0.5 bug report

[Mark H Weaver]

Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
 Thank you Mark,
 Sending this again with Cc:

 (1) Now, make succeeds with threads
 Version of libgc is 7.1, but I found some macro define is required for 
 threads on Solaris.
 After re-install gc-7.1 with -DGC_SOLARIS_THREADS or -DGC_THREADS,
 make succeeds.

This is not my area of expertise, but for lack of other suggestions, I
have two ideas:

1. You could try recompiling libgc using POSIX threads instead of
   Solaris threads, and then rebuilding Guile from scratch, starting
   with configure, and linking against the new libgc.

   Guile supports only POSIX threads, and I wouldn't be surprised if
   there are problems associated with linking Guile against a libgc that
   uses a different threading library with slightly different semantics.

2. You could try a newer version of libgc: 7.2alpha6, from
   http://www.hpl.hp.com/personal/Hans_Boehm/gc/gc_source/gc-7.2alpha6.tar.gz

   Despite the scary version number, Hans Boehm himself now recommends
   this version because 7.1 is so old (see Where to get the collector
   at http://www.hpl.hp.com/personal/Hans_Boehm/gc/ )

Thanks,
  Mark


 (2) Segmentation Fault occur in make check
 Still encounterd FAILS as follows:
 make  check-TESTS
 make[5]: Entering directory `/tmp/guile-2.0.5/test-suite/standalone'
 Assertion failed: SCM_I_IS_THREAD (data.thread), file threads.c, line 1135
 /bin/bash: line 5: 11634 Abort   srcdir=. builddir=. 
 CHARSETALIASDIR
 =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
 ${dir}$tst
 FAIL: test-system-cmds
   :
 PASS: test-extensions
 /bin/bash: line 5: 12409 Segmentation Fault  srcdir=. builddir=. 
 CHARSETALIASDIR
 =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
 ${dir}$tst
 FAIL: test-with-guile-module
 PASS: test-scm-with-guile
 /bin/bash: line 5: 12455 Segmentation Fault  srcdir=. builddir=. 
 CHARSETALIASDIR
 =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
 ${dir}$tst
 FAIL: test-scm-spawn-thread
 /bin/bash: line 5: 12479 Segmentation Fault  srcdir=. builddir=. 
 CHARSETALIASDIR
 =/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
 ${dir}$tst
 FAIL: test-pthread-create
 SKIP: test-pthread-create-secondary
 ==
 4 of 27 tests failed
 (1 test was not run)
 Please report to bug-guile@gnu.org
 ==

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Thank you Mark,
Sending this again with Cc:

(1) Now, make succeeds with threads
Version of libgc is 7.1, but I found some macro define is required for threads 
on Solaris.
After re-install gc-7.1 with -DGC_SOLARIS_THREADS or -DGC_THREADS,
make succeeds.

(2) Segmentation Fault occur in make check
Still encounterd FAILS as follows:
make  check-TESTS
make[5]: Entering directory `/tmp/guile-2.0.5/test-suite/standalone'
Assertion failed: SCM_I_IS_THREAD (data.thread), file threads.c, line 1135
/bin/bash: line 5: 11634 Abort   srcdir=. builddir=. 
CHARSETALIASDIR
=/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
${dir}$tst
FAIL: test-system-cmds
  :
PASS: test-extensions
/bin/bash: line 5: 12409 Segmentation Fault  srcdir=. builddir=. 
CHARSETALIASDIR
=/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
${dir}$tst
FAIL: test-with-guile-module
PASS: test-scm-with-guile
/bin/bash: line 5: 12455 Segmentation Fault  srcdir=. builddir=. 
CHARSETALIASDIR
=/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
${dir}$tst
FAIL: test-scm-spawn-thread
/bin/bash: line 5: 12479 Segmentation Fault  srcdir=. builddir=. 
CHARSETALIASDIR
=/tmp/guile-2.0.5/lib GUILE_AUTO_COMPILE=0 ../../meta/uninstalled-env 
${dir}$tst
FAIL: test-pthread-create
SKIP: test-pthread-create-secondary
==
4 of 27 tests failed
(1 test was not run)
Please report to bug-guile@gnu.org
==

First one seems to be related with thread, and the rest may be related whth (3).

(3) LANG  environment
Yes, % guile --version goes well, with LANG=ja_JP.eucJP


I'll wait for Guile 2.0.6 release, thank you.

--- Kiyoshi yoi_no_myou...@yahoo.co.jp


--- On Sun, 2012/3/25, Mark H Weaver m...@netris.org wrote:

 Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:
 
  I'm trying to install guile-2.0.5 on Solaris 10 x86-64 system.
  I have some problems.
 
  (1) make fails with threads
  make fails with configure CC=gcc -I/usr/local/GNU/include 
  -L/usr/local/GNU/lib --prefix=/usr/local/GNU
  make stops with messages as:
  make[3]: Entering directory `/tmp/guile-2.0.5/libguile'
    CCLD   guile
  Undefined                       first referenced
   symbol                             in file
  GC_unregister_my_thread             ./.libs/libguile-2.0.so
  GC_register_my_thread               ./.libs/libguile-2.0.so
  GC_pthread_create                   ./.libs/libguile-2.0.so
  GC_pthread_detach                   ./.libs/libguile-2.0.so
  ld: fatal: symbol referencing errors. No output written to .libs/guile
  collect2: ld returned 1 exit status
 
  libgc.a, libgc.la, libgc.so, libgc.so.1, libgc.so.1.0.3 can be found on 
  /usr/local/GNU/lib.
 
 What version of libgc is installed?  Was it built with POSIX thread
 support?  On my GNU/Linux system, libgc's configure script output the
 following lines:
 
   checking for thread model used by GCC... posix
   checking for pthread_self in -lpthread... yes
 
 Do you see anything like that in your libgc configure output?
 
  (2) make passes with --without-threads but Segmentation Fault occur in make 
  check
  make succeeds with configure --without-threads CC=gcc 
  -I/usr/local/GNU/include -L/usr/local/GNU/lib --prefix=/usr/local/GNU
  But, make check failes as:
  Testing /tmp/guile-2.0.5/meta/guile ...
  with GUILE_LOAD_PATH=/tmp/guile-2.0.5/test-suite
  /bin/bash: line 5: 29264 Segmentation Fault      
  CHARSETALIASDIR=/tmp/guile-2.0.5/lib 
  ${dir}$tst
  FAIL: check-guile
 
 Hmm, I'm not sure what's happening here.  I wonder if it's related to
 the locale bug (see below).
 
  (3) guile does not go correctly if LANG is not C
  Installed guile ignoring the result of make check, and ran guile --version 
  with some LANG environment.
 
 This is a known bug, fixed in 5de0053178b4acc793ae62838175e5f3ab56c603.
 The fix will be in Guile 2.0.6.
 
 To work around this problem, use only locale identifiers with an
 explicit character set, such as ja_JP.eucJP.  The bug only affects
 locale identifiers without the character set suffix.
 
    Thanks,
      Mark

bug#11083: guile-2.0.5 bug report

[Kiyoshi KANAZAWA]

Hello,

I'm trying to install guile-2.0.5 on Solaris 10 x86-64 system.
I have some problems.

(1) make fails with threads
make fails with configure CC=gcc -I/usr/local/GNU/include 
-L/usr/local/GNU/lib --prefix=/usr/local/GNU
make stops with messages as:
make[3]: Entering directory `/tmp/guile-2.0.5/libguile'
  CCLD   guile
Undefined   first referenced
 symbol in file
GC_unregister_my_thread ./.libs/libguile-2.0.so
GC_register_my_thread   ./.libs/libguile-2.0.so
GC_pthread_create   ./.libs/libguile-2.0.so
GC_pthread_detach   ./.libs/libguile-2.0.so
ld: fatal: symbol referencing errors. No output written to .libs/guile
collect2: ld returned 1 exit status

libgc.a, libgc.la, libgc.so, libgc.so.1, libgc.so.1.0.3 can be found on 
/usr/local/GNU/lib.

(2) make passes with --without-threads but Segmentation Fault occur in make 
check
make succeeds with configure --without-threads CC=gcc -I/usr/local/GNU/include 
-L/usr/local/GNU/lib --prefix=/usr/local/GNU
But, make check failes as:
Testing /tmp/guile-2.0.5/meta/guile ...
with GUILE_LOAD_PATH=/tmp/guile-2.0.5/test-suite
/bin/bash: line 5: 29264 Segmentation Fault  
CHARSETALIASDIR=/tmp/guile-2.0.5/lib 
${dir}$tst
FAIL: check-guile
==
1 of 1 test failed
Please report to bug-guile@gnu.org
==

(3) guile does not go correctly if LANG is not C
Installed guile ignoring the result of make check, and ran guile --version with 
some LANG environment.
(3-1) LANG=C
% guile --version
guile (GNU Guile) 2.0.5
Copyright (C) 2011 Free Software Foundation, Inc.

License LGPLv3+: GNU LGPL 3 or later http://gnu.org/licenses/lgpl.html.
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
(3-2) LANG=ja
guile --version
Backtrace:
In ice-9/boot-9.scm:
 149: 2 [catch #t #catch-closure 81ce670 ...]
 157: 1 [#procedure 81938c0 ()]
In unknown file:
   ?: 0 [catch-closure]

ERROR: In procedure catch-closure:
ERROR: Throw to key `decoding-error' with args `(scm_from_stringn input 
locale conversion error 22 #vu8(45 45 118 101 114 115 105 111 110))'.
(3-3) LANG=es
guile --version
Backtrace:
In ice-9/boot-9.scm:
 149: 2 [catch #t #catch-closure 81ce670 ...]
 157: 1 [#procedure 81938c0 ()]
In unknown file:
   ?: 0 [catch-closure]

ERROR: In procedure catch-closure:
ERROR: Throw to key `decoding-error' with args `(scm_from_stringn input 
locale conversion error 22 #vu8(45 45 118 101 114 115 105 111 110))'.


Best regards,

--- Kiyoshi yoi_no_myou...@yahoo.co.jp

bug#11083: guile-2.0.5 bug report

[Mark H Weaver]

Kiyoshi KANAZAWA yoi_no_myou...@yahoo.co.jp writes:

 I'm trying to install guile-2.0.5 on Solaris 10 x86-64 system.
 I have some problems.

 (1) make fails with threads
 make fails with configure CC=gcc -I/usr/local/GNU/include 
 -L/usr/local/GNU/lib --prefix=/usr/local/GNU
 make stops with messages as:
 make[3]: Entering directory `/tmp/guile-2.0.5/libguile'
   CCLD   guile
 Undefined   first referenced
  symbol in file
 GC_unregister_my_thread ./.libs/libguile-2.0.so
 GC_register_my_thread   ./.libs/libguile-2.0.so
 GC_pthread_create   ./.libs/libguile-2.0.so
 GC_pthread_detach   ./.libs/libguile-2.0.so
 ld: fatal: symbol referencing errors. No output written to .libs/guile
 collect2: ld returned 1 exit status

 libgc.a, libgc.la, libgc.so, libgc.so.1, libgc.so.1.0.3 can be found on 
 /usr/local/GNU/lib.

What version of libgc is installed?  Was it built with POSIX thread
support?  On my GNU/Linux system, libgc's configure script output the
following lines:

  checking for thread model used by GCC... posix
  checking for pthread_self in -lpthread... yes

Do you see anything like that in your libgc configure output?

 (2) make passes with --without-threads but Segmentation Fault occur in make 
 check
 make succeeds with configure --without-threads CC=gcc 
 -I/usr/local/GNU/include -L/usr/local/GNU/lib --prefix=/usr/local/GNU
 But, make check failes as:
 Testing /tmp/guile-2.0.5/meta/guile ...
 with GUILE_LOAD_PATH=/tmp/guile-2.0.5/test-suite
 /bin/bash: line 5: 29264 Segmentation Fault  
 CHARSETALIASDIR=/tmp/guile-2.0.5/lib 
 ${dir}$tst
 FAIL: check-guile

Hmm, I'm not sure what's happening here.  I wonder if it's related to
the locale bug (see below).

 (3) guile does not go correctly if LANG is not C
 Installed guile ignoring the result of make check, and ran guile --version 
 with some LANG environment.

This is a known bug, fixed in 5de0053178b4acc793ae62838175e5f3ab56c603.
The fix will be in Guile 2.0.6.

To work around this problem, use only locale identifiers with an
explicit character set, such as ja_JP.eucJP.  The bug only affects
locale identifiers without the character set suffix.

   Thanks,
 Mark