Re: EXPORT_SYMTAB and 2.3.50

From: Keith Owens (kaos@ocs.com.au)
Date: Fri Mar 10 2000 - 02:04:34 EST


On Thu, 9 Mar 2000 13:17:55 +0100,
f5ibh <f5ibh@db0bm.ampr.org> wrote:
>Keith Owens <kaos@ocs.com.au> wrote :
>> WRONG! module.h is designed to map EXPORT_SYMTAB to
>> "EXPORT_SYMTAB_not_defined" when the gcc command is incorrect. In this
>> case the problem is the gcc line does include "-DEXPORT_SYMTAB", the
>
>The fact that EXPORT_SYMTAB should be in the Makefile or in the code appears
>as a comment in some files.
>
>Here is an example in sb_common.c
>...
>/* FIXME: *grr* why can't the f**in Makefile do this for me ? */
>#define EXPORT_SYMTAB
>...
>
>I think that 'somebody' has to choose how the things have to be done.

'somebody' chose the correct way years ago when the symbol export
mechanism was changed from X() to EXPORT_SYMTAB(). All modules that
export symbols must be defined in LX_OBJS, OX_OBJS, MX_OBJS or
MIX_OBJS. If, and only if, a module is defined in *X_OBJS, the
makefile rules in Rules.make will include -DEXPORT_SYMTAB in the gcc
command. Create a bad Makefile and you get a bad compile - GIGO.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Wed Mar 15 2000 - 21:00:17 EST