m68k "Invalid character:" build failure look familiar to anyone?

m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

I'm trying to track down a 21.3 build failure that only appears on the
m68k architecture when trying to build the Debian package (all the
other (Debian) architectures build ok). I was hoping to see if this
looked familiar to anyone before I try to track it down the hard way.
I'd be happy to gather/provide more information if it would be
helpful.

During a full boostrap build, the build fails as follows:

for el in /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/bytecomp.el /home/rlb/tmp/emacs21-21.3+1/lisp/subr.el /home/rlb/tmp/emacs21-21.3+1/lisp/progmodes/cc-mode.el /home/rlb/tmp/emacs21-21.3+1/lisp/progmodes/cc-vars.el $els; do \
  echo Compiling $el; \
  EMACSLOADPATH=/home/rlb/tmp/emacs21-21.3+1/lisp ../src/bootstrap-emacs -batch --no-site-file --multibyte -f batch-byte-compile $el || exit 1; \
done
Compiling /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el
  !! error (("Invalid character: 037000000000, -134217728, 0xf8000000"))
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
  !! error (("Invalid character: 037300000000, -83886080, 0xfb000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
  !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
While compiling toplevel forms in file /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el:
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
  !! error (("Invalid character: 037600000011, -33554423, 0xfe000009"))
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
Done
make[3]: *** [compile-files] Error 1

  The full bug is reported here:
    http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=207580
  and the various architecture build statuses are reported here:
    http://buildd.debian.org/build.php?arch=&pkg=emacs21

Thanks
-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Andreas Schwab]

Rob Browning <rlb@defaultvalue.org> writes:

> I'm trying to track down a 21.3 build failure that only appears on the
> m68k architecture when trying to build the Debian package (all the
> other (Debian) architectures build ok). I was hoping to see if this
> looked familiar to anyone before I try to track it down the hard way.
> I'd be happy to gather/provide more information if it would be
> helpful.

Please try -Wall to see whether there is anything suspicious.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Andreas Schwab <schwab@suse.de> writes:

> Rob Browning <rlb@defaultvalue.org> writes:
>
>> I'm trying to track down a 21.3 build failure that only appears on the
>> m68k architecture when trying to build the Debian package (all the
>> other (Debian) architectures build ok). I was hoping to see if this
>> looked familiar to anyone before I try to track it down the hard way.
>> I'd be happy to gather/provide more information if it would be
>> helpful.
>
> Please try -Wall to see whether there is anything suspicious.

OK, according to grep, there were about 382 warnings in the ~100K+
package build log.  I can post the whole thing, or if you only want to
see the output for particular files, I can just post that.

Thanks

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Andreas Schwab]

Rob Browning <rlb@defaultvalue.org> writes:

> Andreas Schwab <schwab@suse.de> writes:
>
>> Rob Browning <rlb@defaultvalue.org> writes:
>>
>>> I'm trying to track down a 21.3 build failure that only appears on the
>>> m68k architecture when trying to build the Debian package (all the
>>> other (Debian) architectures build ok). I was hoping to see if this
>>> looked familiar to anyone before I try to track it down the hard way.
>>> I'd be happy to gather/provide more information if it would be
>>> helpful.
>>
>> Please try -Wall to see whether there is anything suspicious.
>
> OK, according to grep, there were about 382 warnings in the ~100K+
> package build log.  I can post the whole thing, or if you only want to
> see the output for particular files, I can just post that.

The most interesting warnings are about implicit declarations.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Andreas Schwab <schwab@suse.de> writes:

>> OK, according to grep, there were about 382 warnings in the ~100K+
>> package build log.  I can post the whole thing, or if you only want to
>> see the output for particular files, I can just post that.
>
> The most interesting warnings are about implicit declarations.

Below are just the implicit warnings.  I also fixed the "suggest
parentheses around && within ||" warnings in charset.c to match the
code in current CVS.

Also, here's the first bit of a backtrace I just generated in case
that's helpful:

#0  invalid_character (c=-134217728)
    at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:117
#1  0x80033c2a in char_to_string (c=-134217728, 
    str=0x80852ef4 "o88592:\t\t\t\t\tcs_CZ.ISO8859-2")
    at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:270
#2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
    last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
#3  0x800d7a26 in Fconcat (nargs=1, args=0xefffc098)
    at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:434
#4  0x800d326e in Feval (form=1350173948)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:1986
#5  0x800d0d1a in Fprogn (args=274922548)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:431
#6  0x800d1790 in Flet (args=1350170172)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:875
#7  0x800d3068 in Feval (form=1350168404)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:2013

And here are the warnings:

$ grep -B1 implicit emacs21-crest-build.log

/home/rlb/tmp/emacs21-21.3+1/src/charset.c:695: warning: suggest parentheses around && within ||
/home/rlb/tmp/emacs21-21.3+1/src/charset.c:717: warning: implicit declaration of function `Fupdate_coding_systems_internal'
--
/home/rlb/tmp/emacs21-21.3+1/src/xfaces.c: In function `x_update_menu_appearance':
/home/rlb/tmp/emacs21-21.3+1/src/xfaces.c:4475: warning: implicit declaration of function `free_frame_menubar'
--
/home/rlb/tmp/emacs21-21.3+1/src/xterm.c: In function `x_window_to_menu_bar':
/home/rlb/tmp/emacs21-21.3+1/src/xterm.c:8271: warning: implicit declaration of function `xlwmenu_window_p'
/home/rlb/tmp/emacs21-21.3+1/src/xterm.c: In function `XTread_socket':
/home/rlb/tmp/emacs21-21.3+1/src/xterm.c:10247: warning: implicit declaration of function `xlwmenu_redisplay'
--
/home/rlb/tmp/emacs21-21.3+1/src/emacs.c: In function `main':
/home/rlb/tmp/emacs21-21.3+1/src/emacs.c:1473: warning: implicit declaration of function `init_window'
/home/rlb/tmp/emacs21-21.3+1/src/emacs.c: In function `Fdump_emacs':
/home/rlb/tmp/emacs21-21.3+1/src/emacs.c:2027: warning: implicit declaration of function `unexec'
--
/home/rlb/tmp/emacs21-21.3+1/src/keyboard.c: In function `read_char':
/home/rlb/tmp/emacs21-21.3+1/src/keyboard.c:2376: warning: implicit declaration of function `_setjmp'
--
/home/rlb/tmp/emacs21-21.3+1/src/eval.c: In function `call_debugger':
/home/rlb/tmp/emacs21-21.3+1/src/eval.c:270: warning: implicit declaration of function `Ftop_level'
/home/rlb/tmp/emacs21-21.3+1/src/eval.c: In function `internal_catch':
/home/rlb/tmp/emacs21-21.3+1/src/eval.c:1029: warning: implicit declaration of function `_setjmp'
--
/home/rlb/tmp/emacs21-21.3+1/src/sound.c: In function `vox_configure':
/home/rlb/tmp/emacs21-21.3+1/src/sound.c:799: warning: implicit declaration of function `ioctl'
--
/home/rlb/tmp/emacs21-21.3+1/src/prefix-args.c: In function `main':
/home/rlb/tmp/emacs21-21.3+1/src/prefix-args.c:63: warning: implicit declaration of function `exit'


-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Andreas Schwab]

Rob Browning <rlb@defaultvalue.org> writes:

> Also, here's the first bit of a backtrace I just generated in case
> that's helpful:
>
> #0  invalid_character (c=-134217728)
>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:117
> #1  0x80033c2a in char_to_string (c=-134217728, 
>     str=0x80852ef4 "o88592:\t\t\t\t\tcs_CZ.ISO8859-2")
>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:270
> #2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
>     last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
> #3  0x800d7a26 in Fconcat (nargs=1, args=0xefffc098)
>     at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:434
> #4  0x800d326e in Feval (form=1350173948)
>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:1986
> #5  0x800d0d1a in Fprogn (args=274922548)
>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:431

Can you please print out the lisp expression that args points to at this
place?  See etc/DEBUG for more information.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Andreas Schwab <schwab@suse.de> writes:

> Rob Browning <rlb@defaultvalue.org> writes:
>
>> Also, here's the first bit of a backtrace I just generated in case
>> that's helpful:
>>
>> #0  invalid_character (c=-134217728)
>>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:117
>> #1  0x80033c2a in char_to_string (c=-134217728, 
>>     str=0x80852ef4 "o88592:\t\t\t\t\tcs_CZ.ISO8859-2")
>>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:270
>> #2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
>>     last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
>> #3  0x800d7a26 in Fconcat (nargs=1, args=0xefffc098)
>>     at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:434
>> #4  0x800d326e in Feval (form=1350173948)
>>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:1986
>> #5  0x800d0d1a in Fprogn (args=274922548)
>>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:431
>
> Can you please print out the lisp expression that args points to at this
> place?  See etc/DEBUG for more information.

Sure:

(gdb) up
#2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
    last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
822                         toindex_byte += CHAR_STRING (c, p);
(gdb) p *args
$2 = 1350394612
(gdb) pr
(8 131 8 -134217728 194 9 33 136 195 9 64 33 136 196 197 9 65 34 136 198 199 9 65 71 34 135)

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Kevin Rodgers]

Rob Browning wrote:

> Andreas Schwab <schwab@suse.de> writes:
>>Can you please print out the lisp expression that args points to at this
>>place?  See etc/DEBUG for more information.
> 
> Sure:
> 
> (gdb) up
> #2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
>     last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
> 822                         toindex_byte += CHAR_STRING (c, p);
> (gdb) p *args
> $2 = 1350394612
> (gdb) pr
> (8 131 8 -134217728 194 9 33 136 195 9 64 33 136 196 197 9 65 34 136 198 199 9 65 71 34 135)

Hmmm:
(format "%c" -134217728) => "\200"


The other characters are:
(concat
  '(8 131 8 194 9 33 136 195 9 64 33 136 196 197 9 65 34 136 198 199 9 65 71 34 
135)) =>
"\b\203\b\302 
!\210\303 
@!\210\304\305 
A\"\210\306\307 
AG\"\207"

-- 
Kevin Rodgers

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Andreas Schwab]

Rob Browning <rlb@defaultvalue.org> writes:

> Andreas Schwab <schwab@suse.de> writes:
>
>> Rob Browning <rlb@defaultvalue.org> writes:
>>
>>> Also, here's the first bit of a backtrace I just generated in case
>>> that's helpful:
>>>
>>> #0  invalid_character (c=-134217728)
>>>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:117
>>> #1  0x80033c2a in char_to_string (c=-134217728, 
>>>     str=0x80852ef4 "o88592:\t\t\t\t\tcs_CZ.ISO8859-2")
>>>     at /home/rlb/tmp/emacs21-21.3+1/src/charset.c:270
>>> #2  0x800d7fee in concat (nargs=1, args=0xefffc098, target_type=Lisp_String, 
>>>     last_special=0) at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:822
>>> #3  0x800d7a26 in Fconcat (nargs=1, args=0xefffc098)
>>>     at /home/rlb/tmp/emacs21-21.3+1/src/fns.c:434
>>> #4  0x800d326e in Feval (form=1350173948)
>>>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:1986
>>> #5  0x800d0d1a in Fprogn (args=274922548)
>>>     at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:431
>>
>> Can you please print out the lisp expression that args points to at this
>> place?  See etc/DEBUG for more information.
>
> Sure:

I was rather referring to the args in Fprogn.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux AG, Deutschherrnstr. 15-19, D-90429 Nürnberg
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Andreas Schwab <schwab@suse.de> writes:

>>> Can you please print out the lisp expression that args points to at this
>>> place?  See etc/DEBUG for more information.
>>
>> Sure:
>
> I was rather referring to the args in Fprogn.

Ah, ok.  If you mean the next level Fprogn up, then it appears to be
nil at the time of the error.  I also included args from the Flet
above it in case that's helpful.

Thanks again.

(gdb) up
#4  0x800d326e in Feval (form=1350173948)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:1986
1986              val = (*XSUBR (fun)->function) (XINT (numargs), vals);
(gdb) up
#5  0x800d0d1a in Fprogn (args=274922548)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:431
431           val = Feval (Fcar (args_left));
(gdb) p args
$5 = 274922548
(gdb) pr
nil
(gdb) up
#6  0x800d1790 in Flet (args=1350170172)
    at /home/rlb/tmp/emacs21-21.3+1/src/eval.c:875
875       elt = Fprogn (Fcdr (args));
(gdb) p args
$6 = 1350170172
(gdb) pr
(((pc 0) op off (bytes (quote nil)) (patchlist nil) rest rel tmp) (while lap (setq op (car (car lap)) off (cdr (car lap))) (cond ((not (symbolp op)) (error "Non-symbolic opcode `%s'" op)) ((eq op (quote TAG)) (setcar off pc) (setq patchlist (cons off patchlist))) ((memq op byte-goto-ops) (setq pc (+ pc 3)) (setq bytes (cons (cons pc (cdr off)) (cons nil (cons (symbol-value op) bytes)))) (setq patchlist (cons bytes patchlist))) (t (setq bytes (cond ((cond ((consp off) (setq off (cdr off)) (eq op (quote byte-constant)))) (cond ((< off byte-constant-limit) (setq pc (1+ pc)) (cons (+ byte-constant off) bytes)) (t (setq pc (+ 3 pc)) (cons (lsh off -8) (cons (logand off 255) (cons byte-constant2 bytes)))))) ((<= byte-listN (symbol-value op)) (setq pc (+ 2 pc)) (cons off (cons (symbol-value op) bytes))) ((< off 6) (setq pc (1+ pc)) (cons (+ (symbol-value op) off) bytes)) ((< off 256) (setq pc (+ 2 pc)) (cons off (cons (+ (symbol-value op) 6) bytes))) (t (setq pc (+ 3 pc)) (cons (lsh off -8) (cons (logand off 255) (cons (+ (symbol-value op) 7) bytes)))))))) (setq lap (cdr lap))) (let (bytes) (while patchlist (setq bytes (car patchlist)) (cond ((atom (car bytes))) (t (setq pc (car (cdr (car bytes)))) (setcar (cdr bytes) (logand pc 255)) (setcar bytes (lsh pc -8)))) (setq patchlist (cdr patchlist)))) (concat (nreverse bytes)))

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Stefan Monnier]

> During a full boostrap build, the build fails as follows:

> for el in /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/bytecomp.el /home/rlb/tmp/emacs21-21.3+1/lisp/subr.el /home/rlb/tmp/emacs21-21.3+1/lisp/progmodes/cc-mode.el /home/rlb/tmp/emacs21-21.3+1/lisp/progmodes/cc-vars.el $els; do \
>   echo Compiling $el; \
>   EMACSLOADPATH=/home/rlb/tmp/emacs21-21.3+1/lisp ../src/bootstrap-emacs -batch --no-site-file --multibyte -f batch-byte-compile $el || exit 1; \
> done
> Compiling /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el
>   !! error (("Invalid character: 037000000000, -134217728, 0xf8000000"))
>   !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
>   !! error (("Invalid character: 037300000000, -83886080, 0xfb000000"))
>   !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
>   !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
>   !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
> While compiling toplevel forms in file /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el:
>   !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
>   !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
>   !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
>   !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
>   !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
>   !! error (("Invalid character: 037600000011, -33554423, 0xfe000009"))
>   !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
> Done
> make[3]: *** [compile-files] Error 1

Could you use ../src/bootstrap-emacs interactively and byte-compile the
file from there, while setting both debug-on-error and byte-compile-debug
to a non-nil value ?
This should give you a lisp backtrace,


        Stefan

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

> Could you use ../src/bootstrap-emacs interactively and byte-compile
> the file from there, while setting both debug-on-error and
> byte-compile-debug to a non-nil value ?  This should give you a lisp
> backtrace,

Certainly, though it gives me an error:

  $ EMACSLOADPATH=/home/rlb/tmp/emacs21-21.3+1/lisp \
      ../src/bootstrap-emacs --no-site-file --multibyte 
  (setq debug-on-error t)
  t
  (setq byte-compile-debug t)
  t
  (byte-compile-file
   "/home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el")

At this point emacs beeps and prints "Attempt to split fixed-size
window" in the minibuffer.  I get the same result if I just try M-x
debug after firing up bootstrap-emacs.

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Rob Browning <rlb@defaultvalue.org> writes:

> At this point emacs beeps and prints "Attempt to split fixed-size
> window" in the minibuffer.  I get the same result if I just try M-x
> debug after firing up bootstrap-emacs.

The split problem appears to have been an unrelated local issue.

After fixing that, I just get the same output in the *Compile-Log*
buffer that I showed before -- note that there's no lisp error
involved in any of this.  The compile just prints the messages (one at
a time) while compiling:

>From the *Compile-Log*:

Compiling file /home/rlb/tmp/emacs21-21.3+1/lisp/emacs-lisp/byte-opt.el at Sat \
Sep 27 18:43:48 2003
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
  !! error (("Invalid character: 037600000011, -33554423, 0xfe000009"))
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Rob Browning <rlb@defaultvalue.org> writes:

> After fixing that, I just get the same output in the *Compile-Log*
> buffer that I showed before -- note that there's no lisp error
> involved in any of this.  The compile just prints the messages (one at
> a time) while compiling:

(and then exits with a non-zero value, which is what causes the make
 bootstrap to fail.)

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Stefan Monnier]

> After fixing that, I just get the same output in the *Compile-Log*
> buffer that I showed before -- note that there's no lisp error
> involved in any of this.  The compile just prints the messages (one at
> a time) while compiling:

Are you sure you've set both debug-on-error and byte-compile-debug
to non-nil values ?


        Stefan

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

> Are you sure you've set both debug-on-error and byte-compile-debug
> to non-nil values ?

I thought so.  I know I did the first time above (when I had the split
windown problem), but I'll try again when the testing machine is
available again (apparently it's down for repair).

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4

Re: m68k "Invalid character:" build failure look familiar to anyone?

[Rob Browning]

Rob Browning <rlb@defaultvalue.org> writes:

> Stefan Monnier <monnier@IRO.UMontreal.CA> writes:
>
>> Are you sure you've set both debug-on-error and byte-compile-debug
>> to non-nil values ?
>
> I thought so.  I know I did the first time above (when I had the split
> windown problem), but I'll try again when the testing machine is
> available again (apparently it's down for repair).

(One possibly relevant point -- Debian is using gcc 3.3 by default in
 unstable now (where this is being compiled), and I know there have
 been some issues with the new gcc.)

OK, the machine's back.  Here's *exactly* what I've done this time:

  $ tar xzf ../emacs21_21.3+1.orig.tar.gz
  # this is just the emacs 21.3 tar + the leim tar - the .elc files.

  $ cd emacs21-21.3
  $ ./configure \
      --with-x=yes \
      --with-x-toolkit=athena \
      --without-toolkit-scroll-bars \
        && make

make stops with the errors I reproduced previously

  ...
    !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
  Done
  make[2]: *** [compile-files] Error 1
  make[2]: Leaving directory `/home/rlb/non-deb/emacs-21.3/lisp'
  make[1]: *** [bootstrap-lisp] Error 2
  make[1]: Leaving directory `/home/rlb/non-deb/emacs-21.3'
  make: *** [maybe_bootstrap] Error 2

and then

  rlb@crest:~/non-deb/emacs-21.3$ cd lisp
  rlb@crest:~/non-deb/emacs-21.3/lisp$ \
    EMACSLOADPATH=`pwd` ../src/bootstrap-emacs --no-site-file --multibyte

and then from within the *scratch* buffer

  (setq debug-on-error t)
  t
  (setq byte-compile-debug t)
  t
  (byte-compile-file "/home/rlb/non-deb/emacs-21.3/lisp/emacs-lisp/byte-opt.el")

which results in the window splitting into the *scratch* buffer and a
*Compile-Log* buffer, which looks like this:

File Edit Options Buffers Tools Help
(setq debug-on-error t)
t
(setq byte-compile-debug t)
t
(byte-compile-file "/home/rlb/non-deb/emacs-21.3/lisp/emacs-lisp/byte-opt.el")
nil

----:**-F1  *scratch*         (Lisp Interaction)--L7--All-----------------------
Compiling file /home/rlb/non-deb/emacs-21.3/lisp/emacs-lisp/byte-opt.el at Fri \
Oct  3 23:17:14 2003
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037200000000, -100663296, 0xfa000000"))
  !! error (("Invalid character: 037500000000, -50331648, 0xfd000000"))
  !! error (("Invalid character: 037100000000, -117440512, 0xf9000000"))
  !! error (("Invalid character: 037600000011, -33554423, 0xfe000009"))
  !! error (("Invalid character: 037600000000, -33554432, 0xfe000000"))

----:**-F1  *Compile-Log*      (Fundamental)--L9--Bot---------------------------
Compiling /home/rlb/non-deb/emacs-21.3/lisp/emacs-lisp/byte-opt.el... (byte-com\
pile-log-lap-1)

-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4