From: Gleydson Soares (gsoaresgmail.com)
Date: Fri Aug 08 2008 - 10:58:07 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Fri, Aug 8, 2008 at 5:39 AM, Karl Sjodahl - dunceor
<dunceorgmail.com> wrote:
> On Thu, Aug 7, 2008 at 11:29 PM, Gleydson Soares <gsoaresgmail.com> wrote:
>> this function don't return, void instead of int.
>> Index: pppctl.c
>> ===================================================================
>> RCS file: /cvs/src/usr.sbin/ppp/pppctl/pppctl.c,v
>> retrieving revision 1.19
>> diff -u -r1.19 pppctl.c
>> --- pppctl.c 26 Jul 2007 17:48:41 -0000 1.19
>> +++ pppctl.c 7 Aug 2008 21:16:24 -0000
>> -53,8 +53,8
>> #define LINELEN 2048
>> static char Buffer[LINELEN], Command[LINELEN];
>>
>> -static int
>> -usage()
>> +static void
>> +usage(void)
>> {
>> fprintf(stderr, "usage: pppctl [-v] [-p passwd] [-t n] "
>> "Port|LocalSock [command[;command]...]\n");
>>
>>
>
> Why not mark it as __dead as they usually do in OpenBSD usage()?
>
> BR
> dunceor
>

because yes

Theo a time ago told me about, and he doesn't believe in __dead.

i guess, the code should be totally portable. take o look on google,
there are a lot of compilation problems when the __dead poisoning is
declared on others OS.

the OpenSSH 5.1 code had a compilation problem when compiled on
Interix, the problem was the __dead.

from http://www.openssh.org/txt/release-5.1:

"* Use explicit noreturn attribute instead of __dead, fixing
compilation problems on Interix. (bz#1112)"

fortunally, the openbsd code is used a lot on others OS, and the
portability is the first issue to be considered before to write a
diff.

that is my personal opinion,

//G

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]