Possible bug in Emacs 21.3

Possible bug in Emacs 21.3

[Peter Milliken]

I have downloaded the source of 21.3 and built it on a PC running Win2000
and using the Cygwin distribution and following the INSTALL instructions in
emacs-21.3/nt/INSTALL (well, I ran "configure.bat --prefix=/c/emacs" and
then type "make install" as I have done for all other source distributions
:-))

The resulting binary falls over with a "Abort, Retry or Ignore" dialog when
any package referenced from my .emacs that attempts to "require" the cl.el
package.

Just typing M-: (require 'cl) from within the Emacs session will bring up
the problem.

I would imagine this is more likely a problem with my Cygwin setup than an
Emacs bug as I am sure that pretesting would have found this one very
quickly!

I used gcc 3.2.3(?) to build emacs from the source. I then tried using gcc
2.95 but that fails with a source code/compile error part way through the
make, so I rebuilt with gcc 3.2.3 and tried again - same results.

I don't have Microsft C++ so I can't build the binary using the other method
and the pre-built binaries for 20.3 hasn't been produced as yet.

Thanks
Peter

Peter Milliken
Software Engineer
ResMed
Phone: +61 2 9886-5059



Warning:  Copyright ResMed.  Where the contents of this email and/or attachment includes materials prepared by ResMed, the use of those materials is subject exclusively to the conditions of engagement between ResMed and the intended recipient.

This communication is confidential and may contain legally privileged information. By the use of email over the Internet or other communication systems, ResMed is not waiving either confidentiality of, or legal privilege in,the content of the email and of any attachments.
If the recipient of this message is not the intended addressee, please call ResMed immediately on  +61 2 9886 5000 Sydney, Australia.

Re: Possible bug in Emacs 21.3

[maierh]

"Peter Milliken" <PeterM@resmed.com.au> writes:

> I have downloaded the source of 21.3 and built it on a PC running Win2000
> and using the Cygwin distribution and following the INSTALL instructions in
> emacs-21.3/nt/INSTALL (well, I ran "configure.bat --prefix=/c/emacs" and
> then type "make install" as I have done for all other source distributions
> :-))
>
> The resulting binary falls over with a "Abort, Retry or Ignore" dialog when
> any package referenced from my .emacs that attempts to "require" the cl.el
> package.
>
> Just typing M-: (require 'cl) from within the Emacs session will bring up
> the problem.
>
> I would imagine this is more likely a problem with my Cygwin setup than an
> Emacs bug as I am sure that pretesting would have found this one very
> quickly!

> I used gcc 3.2.3(?) to build emacs from the source. I then tried using gcc
> 2.95 but that fails with a source code/compile error part way through the
> make, so I rebuilt with gcc 3.2.3 and tried again - same results.

I veryfied emacs-21.2.95 this with the latest cygwin release and I
could reproduce this on my system too. I was able to compile emacs too
with the latest cygwin 'gcc version 2.95.3-10 (cygwin special)', but
the problem still exists: cl pops up a "Abort, Retry or Ignore"
Dialog.

Compiler                          | State
----------------------------------+-----------------
gcc    2.95.3-10 (cygwin special) | fails with cl.el
gcc 3.2 20020927 (prerelease)     | fails with cl.el
gcc     2.95.3-6 (mingw special)  | okay
msvc                              | okay

With the emacs head version I get by executing the "M-: (require 'cl)"
lisp function the following backtrace:

,----
| Debugger entered--Lisp error: (invalid-read-syntax ". in wrong context")
|   cl-random-time()
|   byte-code("...")
|   require(cl)
|   eval((require (quote cl)))
|   eval-expression((require (quote cl)) nil)
| * call-interactively(eval-expression)
`----

I am cc'ing this too to the cygwin mailing list, maybe someone has an
idea what the problem caused.

Harald

Re: [h-e-w] Re: Possible bug in Emacs 21.3

[Harald.Maier.BW]

Hello David,

I tried your patch below with emacs-21.3 and emacs cvs head and it
works very fine with gcc 3.2 and the latest cygwin release. So
hopefully the patch is soon included into cvs. Thanks for your fast
reply.

Harald

David Ponce <david.ponce@wanadoo.fr> writes:

> Harald.Maier.BW@t-online.de wrote:
>
>> "Peter Milliken" <PeterM@resmed.com.au> writes:
>>
>>>I have downloaded the source of 21.3 and built it on a PC running Win2000
>>>and using the Cygwin distribution and following the INSTALL instructions in
>>>emacs-21.3/nt/INSTALL (well, I ran "configure.bat --prefix=/c/emacs" and
>>>then type "make install" as I have done for all other source distributions
>>>:-))
>>>
>>>The resulting binary falls over with a "Abort, Retry or Ignore" dialog when
>>>any package referenced from my .emacs that attempts to "require" the cl.el
>>>package.
>>>
>>>Just typing M-: (require 'cl) from within the Emacs session will bring up
>>>the problem.

>> ----------------------------------+-----------------
>> gcc    2.95.3-10 (cygwin special) | fails with cl.el
>> gcc 3.2 20020927 (prerelease)     | fails with cl.el
>> gcc     2.95.3-6 (mingw special)  | okay
>> msvc                              | okay

> Maybe are you using the mingw runtime 2.3 or 2.4?
> There is a compatibility problem between that runtime and Emacs,
> because of change in the way it handles binary vs. text IO mode.
>
> Notice that the development version (in CVS) is affected too.
>
> The following patch (discussed in the emacs-devel ML) fixed that
> problem for me.

>
> *** emacs.c.ori	Thu Aug 29 21:27:07 2002
> --- emacs.c	Thu Mar 27 16:29:58 2003
> ***************
> *** 40,45 ****
> --- 40,49 ----
>    #include <sys/ioctl.h>
>    #endif
>
> + #ifdef WINDOWSNT
> + #include <fcntl.h>
> + #endif
> +
>    #include "lisp.h"
>    #include "commands.h"
>    #include "intervals.h"
> ***************
> *** 874,884 ****
>      uninterrupt_malloc ();
>    #endif	/* not SYSTEM_MALLOC */
>
> ! #ifdef MSDOS
>      /* We do all file input/output as binary files.  When we need to translate
>         newlines, we do that manually.  */
>      _fmode = O_BINARY;
>
>    #if __DJGPP__ >= 2
>      if (!isatty (fileno (stdin)))
>        setmode (fileno (stdin), O_BINARY);
> --- 878,890 ----
>      uninterrupt_malloc ();
>    #endif	/* not SYSTEM_MALLOC */
>
> ! #if defined (MSDOS) || defined (WINDOWSNT)
>      /* We do all file input/output as binary files.  When we need to translate
>         newlines, we do that manually.  */
>      _fmode = O_BINARY;
> + #endif /* MSDOS || WINDOWSNT */
>
> + #ifdef MSDOS
>    #if __DJGPP__ >= 2
>      if (!isatty (fileno (stdin)))
>        setmode (fileno (stdin), O_BINARY);