[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20130506074051.GB17240@openwall.com>
Date: Mon, 6 May 2013 11:40:51 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: Core: build warning
On Mon, May 06, 2013 at 10:53:57AM +0400, Alexander Cherepanov wrote:
> While you are at it you could also think about fixing undefined behavior
> for atoi:
>
> http://openwall.com/lists/john-dev/2013/04/23/5
I think this issue does not affect the core tree.
> >>trip_fmt.c: In function ?cmp_exact?:
> >>trip_fmt.c:520: warning: ISO C90 forbids mixed declarations and code
> >
> >I think these are spurious (gcc bug?),
>
> Sorry, forgot to describe the cause for these warnings. After checking
> the result of preprocessing it becomes clear that the problem is in
> extra semicolon in "MAYBE_T0;". It leads to an empty statement after
> macro substitution.
Oh, this makes everything clear indeed. Thank you!
> >compiler.c provides fallback code for non-gcc, with no dependency on
> >computed gotos. It runs 2-3 times slower, though.
>
> Thanks, I see it now.
>
> Except for this file everything now compiles cleanly with the following
> options:
>
> gcc -c -std=c89 -pedantic -Wall -Wextra -Wno-missing-field-initializers
> -Wno-unused-parameter -Wno-sign-compare -Wno-overlength-strings
You may add -U__GNUC__ to force compilation of non-gcc versions of the
code, where those are available. I've been using this approach for
testing (years ago).
Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
