as infelicity

Tony van der Hoff tony.linux at mk-net.demon.co.uk
Sun Jul 13 10:13:54 PDT 2003


Peter Naulls <peter at chocky.org> wrote:

> In message <gemini.3f11896f021e2510%tony.linux at mk-net.demon.co.uk>
>           Tony van der Hoff <tony.linux at mk-net.demon.co.uk> wrote:
> 
> > as has a global variable "gcc_backend", which is currently
> > unconditionally true. Its primary effect appears to be to cause as to
> > treat the character "@" as a comment flag, equivalent to ";".
> > 
> > In the OBJASM syntax, "@" is the value of the current storage location
> > counter, and is therefore incompatible with its use as a comment.
> > 
> > So, what is the intended use for gcc_backend; can it be safely set to
> > false in the distribution, of should it be set via a command-line
> > switch?
> 
> The current behaviour doesn't seem entirely useful.  Having said that,
> having @ as a comment is helpful for some inline assembler situations, as
> long it's unambigious.  Perhaps you can advise if this is possible.
> 

Is it just a matter of having something other than ; to introduce a comment?
If so, maybe it would be better to define something like "//", rather than
@. But maybe there's already too much code using this feature.

I guess it is unlikely that anyone will want to perform storage layout using
in-line assembler; so alternatively, to enable @ to introduce a comment, I
would be inclined to create a new command line switch, say -b(ackend) which
can be used to override the default (OBJASM-like) behaviour.

However, when this discussion has been long forgotten, someone will discover
that storage layout doesn't work inline, and start the bug-hunt all over
again. My feeling is that it's a bodge.

-- 
Tony van der Hoff           | mailto: tony.linux at mk-net.demon.co.uk
Buckinghamshire, England    | http: www.mk-net.demon.co.uk



More information about the gcc mailing list