Re: trunk r117534: Unbreak compilation of derived cc-mode modes

Re: trunk r117534: Unbreak compilation of derived cc-mode modes

[Glenn Morris]

Daniel Colascione wrote:

> --- a/lisp/progmodes/cc-defs.el	2014-06-29 11:26:47 +0000
> +++ b/lisp/progmodes/cc-defs.el	2014-07-14 23:58:52 +0000
[...]
> @@ -1946,7 +1943,7 @@
>  	;; reason, but we also use this expansion handle
>  	;; `c-lang-defconst-eval-immediately' and to register
>  	;; dependencies on the `c-lang-const's in VAL.)
> -	(setq val (cl-macroexpand-all val))
> +	(setq val (macroexpand-all val))
>  
>  	(setq bindings (cons (cons assigned-mode val) bindings)
>  	      args (cdr args))))

Why is this needed, when cc-defs requires cl at compile time,
which defines cl-macroexpand-all as an alias for macroexpand-all?

It might need backporting to emacs-24 - if so, could you do that please?
TIA.

Re: trunk r117534: Unbreak compilation of derived cc-mode modes

[Daniel Colascione]

[-- Attachment #1: Type: text/plain, Size: 1031 bytes --]

On 07/14/2014 08:47 PM, Glenn Morris wrote:
> Daniel Colascione wrote:
> 
>> --- a/lisp/progmodes/cc-defs.el	2014-06-29 11:26:47 +0000
>> +++ b/lisp/progmodes/cc-defs.el	2014-07-14 23:58:52 +0000
> [...]
>> @@ -1946,7 +1943,7 @@
>>  	;; reason, but we also use this expansion handle
>>  	;; `c-lang-defconst-eval-immediately' and to register
>>  	;; dependencies on the `c-lang-const's in VAL.)
>> -	(setq val (cl-macroexpand-all val))
>> +	(setq val (macroexpand-all val))
>>  
>>  	(setq bindings (cons (cons assigned-mode val) bindings)
>>  	      args (cdr args))))
> 
> Why is this needed, when cc-defs requires cl at compile time,
> which defines cl-macroexpand-all as an alias for macroexpand-all?

I was trying to fix some code that broke when requiring cc-langs and
cc-fonts only. Requiring these features doesn't bring the needed alias,
but apparently used to:

ELISP> (require 'cc-langs)
cc-langs
ELISP> (require 'cc-fonts)
cc-fonts
ELISP> (featurep 'cl)
nil

Maybe that's the real bug.


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]

Re: trunk r117534: Unbreak compilation of derived cc-mode modes

[Glenn Morris]

Daniel Colascione wrote:

>> Why is this needed, when cc-defs requires cl at compile time,
>> which defines cl-macroexpand-all as an alias for macroexpand-all?
>
> I was trying to fix some code that broke when requiring cc-langs and
> cc-fonts only. Requiring these features doesn't bring the needed alias,
> but apparently used to:
>
> ELISP> (require 'cc-langs)
> cc-langs
> ELISP> (require 'cc-fonts)
> cc-fonts
> ELISP> (featurep 'cl)
> nil
>
> Maybe that's the real bug.

Doesn't whether that's "right" or not depend on whether you are loading
the compiled or uncompiled versions? cc-mode was certainly not
_supposed_ to be loading cl (or cc-langs) at runtime when run compiled,
though for a time it did:

http://debbugs.gnu.org/cgi/bugreport.cgi?bug=17463

I can't say I understand all this cc-mode stuff, but I made the change
from cl to cl-lib in emacs-24, so if something needs fixing it
probably needs fixing there too.

Re: trunk r117534: Unbreak compilation of derived cc-mode modes

[Glenn Morris]

PS I can't see any way that changing from cl-macroexpand-all to
macroexpand-all could do any harm in emacs-24, so if in doubt just do
it.