GNU bug report logs - #19634
counting MANY function args more reliably

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 19634 in the body.
You can then email your comments to 19634 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Emacs bug reports and feature requests <bug-gnu-emacs <at> gnu.org>
Subject: counting MANY function args more reliably
Date: Tue, 20 Jan 2015 01:26:25 -0800

[Message part 1 (text/plain, inline)]

Commit ef5a526f1b51b76b0f753e0936c80743a7f4463d fixed a bug in Emacs where the C 
code passed 10 arguments to a function using the MANY calling convention, but 
the caller mistakenly passed 8 as the argument count, and this mistakenly 
discarded the last two arguments; see <http://bugs.gnu.org/3228#63>.

To help prevent this sort of mistake in the future, I would like to install 
something like the attached patch, which uses a new C macro CALLN to count these 
arguments automatically.  The key lines in the patch are the following additions 
to lisp.h:

#define CALLMANY(f, array) (f) (ARRAYELTS (array), array)

#define CALLN(f, ...) CALLMANY (f, ((Lisp_Object []) {__VA_ARGS__}))

This lets code use 'return CALLN (foo, a, b, c, d);' instead of the current 
error-prone usages which are like '{ Lisp_Object args[4]; args[0] = a; args[1] = 
b; args[2] = c; args[3] = d; return foo (3, args); }'.  (Oops, that last '3' 
should have been a '4'....)

This patch still needs a ChangeLog entry and some more testing on my part, but I 
thought I'd get it out for review now.

[calln.diff (text/x-patch, attachment)]

Message #10 received at 19634 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Paul Eggert <eggert <at> cs.ucla.edu>
Cc: 19634 <at> debbugs.gnu.org
Subject: Re: bug#19634: counting MANY function args more reliably
Date: Tue, 20 Jan 2015 11:37:39 -0500

> #define CALLN(f, ...) CALLMANY (f, ((Lisp_Object []) {__VA_ARGS__}))

I like that,


        Stefan

Message #13 received at 19634 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 19634 <at> debbugs.gnu.org, eggert <at> cs.ucla.edu
Subject: Re: bug#19634: counting MANY function args more reliably
Date: Thu, 22 Jan 2015 18:31:21 +0200

> From: Stefan Monnier <monnier <at> iro.umontreal.ca>
> Date: Tue, 20 Jan 2015 11:37:39 -0500
> Cc: 19634 <at> debbugs.gnu.org
> 
> > #define CALLN(f, ...) CALLMANY (f, ((Lisp_Object []) {__VA_ARGS__}))
> 
> I like that,

Do we want to use this for every call to a function that accepts MANY
args, or just those where the argument is an array?

I'm asking because the proposed patch left out this part of w32fns.c:

  static Lisp_Object
  x_create_tip_frame (struct w32_display_info *dpyinfo,
		      Lisp_Object parms, Lisp_Object text)
  {
    [...]
    Ferase_buffer ();
    Finsert (1, &text);  <<<<<<<<<<<<<<<<<<<<<<

and I'm not sure whether this was just an omission or this kind of
calls doesn't need to use CALLN.

Message #16 received at 19634 <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Eli Zaretskii <eliz <at> gnu.org>, 
 Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 19634 <at> debbugs.gnu.org
Subject: Re: bug#19634: counting MANY function args more reliably
Date: Thu, 22 Jan 2015 09:42:10 -0800

On 01/22/2015 08:31 AM, Eli Zaretskii wrote:
> Do we want to use this for every call to a function that accepts MANY
> args, or just those where the argument is an array?

We should use it for calls where counting the number of arguments is 
likely to be error-prone.

> I'm asking because the proposed patch left out this part of w32fns.c:
>
>    static Lisp_Object
>    x_create_tip_frame (struct w32_display_info *dpyinfo,
> 		      Lisp_Object parms, Lisp_Object text)
>    {
>      [...]
>      Ferase_buffer ();
>      Finsert (1, &text);  <<<<<<<<<<<<<<<<<<<<<<

The patch does not affect calls like that, as they're unlikely to give 
rise to the sort of typographical error that motivated the patch.

Message #19 received at 19634 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Paul Eggert <eggert <at> cs.ucla.edu>
Cc: 19634 <at> debbugs.gnu.org, monnier <at> iro.umontreal.ca
Subject: Re: bug#19634: counting MANY function args more reliably
Date: Thu, 22 Jan 2015 20:16:05 +0200

> Date: Thu, 22 Jan 2015 09:42:10 -0800
> From: Paul Eggert <eggert <at> cs.ucla.edu>
> CC: 19634 <at> debbugs.gnu.org
> 
> On 01/22/2015 08:31 AM, Eli Zaretskii wrote:
> > Do we want to use this for every call to a function that accepts MANY
> > args, or just those where the argument is an array?
> 
> We should use it for calls where counting the number of arguments is 
> likely to be error-prone.
> 
> > I'm asking because the proposed patch left out this part of w32fns.c:
> >
> >    static Lisp_Object
> >    x_create_tip_frame (struct w32_display_info *dpyinfo,
> > 		      Lisp_Object parms, Lisp_Object text)
> >    {
> >      [...]
> >      Ferase_buffer ();
> >      Finsert (1, &text);  <<<<<<<<<<<<<<<<<<<<<<
> 
> The patch does not affect calls like that, as they're unlikely to give 
> rise to the sort of typographical error that motivated the patch.

OK, thanks.

Message #24 received at 19634-done <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 19634-done <at> debbugs.gnu.org
Subject: Re: bug#19634: counting MANY function args more reliably
Date: Sun, 25 Jan 2015 09:15:58 -0800

Stefan Monnier wrote:
> I like that,

Thanks, I tested the patch on a few more platforms, rebased it to the current 
master, added some commentary in response to the remarks in this bug report, and 
installed the patch into the master as commit 
a3689d3c661fe36df971c875760f8d500b5ae994.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.