Date: Wed, 21 Jun 2000 15:33:48 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: pavenis AT lanet DOT lv
cc: djgpp AT delorie DOT com
Subject: Re: Inline asm: lcall & various binutils versions
In-Reply-To: <3950C2BF.16877.34C3C3@localhost>
Message-ID: <Pine.SUN.3.91.1000621152921.13853B-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Wed, 21 Jun 2000 pavenis AT lanet DOT lv wrote:

> > AFAIK, inline assembly doesn't go through cpp; GCC emits it in the
> > form of preprocessed assembly.  So you cannot have any preprocessor 
> > directives inside the asm() block.
> 
> Wrong.
> 
> There are no problems using #ifdef and similar stuff inside inline 
> assembler. Inline assembler is normally recognised by cc1 or cc1plus
> so it's only after cpp have processed file.

This is a misunderstanding: what I meant to say was that the inline 
assembly doesn't go through preprocessor *after* cc1 or cc1plus wrote it 
to the output, i.e. between cc1 and Gas.

It goes without saying that the C/C++ source itself goes through 
preprocessor.  But at that stage, the asm is not interpreted except for 
constraints and clobber lists.

> #define TEST
> #define FOO "inc %%eax"
> 
> int main (void)
> {
>     int foo = 1;
>     asm ( "inc %%eax\t\n"
> 	  "inc %%eax\t\n"
> #ifdef TEST
> 	  "inc %%eax\t\n"
> #endif    
> 	  FOO "\t\n"
> 	  : "=a" (foo)
>           : "a" (foo)
> 	);
>     return foo;
> }

In this example, the preprocessor directives are on the C level, not on 
the assembly level.  To be on assembly level, they would need to be 
inside the quoted assembly code, not outside it.

I'm sorry if my wording was unclear.