[OpenAFS-devel] [PATCH] flexelint: fun with printf

[Henry B. Hotz]

At 12:01 PM -0500 12/2/03, openafs-devel-request@openafs.org wrote:
>Date: Mon, 1 Dec 2003 18:22:27 +0100
>From: Joerg Sonnenberger <joerg@britannica.bec.de>
>To: openafs-devel@openafs.org
>Subject: Re: [OpenAFS-devel] [PATCH] flexelint: fun with printf
>
>On Mon, Dec 01, 2003 at 10:10:55AM -0500, Ted Anderson wrote:
>>  On 11/21/2003 14:45, Joe Buehler wrote:
>>  >This patch addresses a number of problems with printf-like functions.
>>
>>  I don't know how the gatekeepers feel about this recent spate of
>>  patches, but it seems like a valuable and overdue effort.  Thanks.
>>
>>  >Most of them are %x formats applied to pointers (which could be a
>>  >problem depending on the architecture I would assume).  There is a
>>  >%p format for pointers on many machines but I don't know if it
>>  >applies to all supported OpenAFS platforms so I didn't go that route.
>>
>>  > -	afsi_log("osi_LogCreate log addr %x", afsd_logp);
>>  > +	afsi_log("osi_LogCreate log addr %x", (int)afsd_logp);
>>
>>  I think it would be more correct, and still portable, to use long
>>  instead of int to represent pointers.  I don't know it C mandates it,
>>  but I think the wide-spread convention is that longs are at least as big
>>  as pointers, while the size of ints vary more widely.  So this would
>>  lead to code like this:
>>
>>	afsi_log("osi_LogCreate log addr %lx", (long)afsd_logp);
>
>What about:
>afsi_log("osi_LogCreate log addr %p", (void*)afsd_logp);
>
>The BSD man page doesn't mention %p as non portable, so it may be the
>correct way.
>
>Joerg
>
>>
>  > Ted Anderson

Well someone posted it worked for Windows.  I can confirm it's there 
for OSX and NetBSD, and Solaris.  That's starting to look like good 
coverage.

Any other takers?
-- 
The opinions expressed in this message are mine,
not those of Caltech, JPL, NASA, or the US Government.
Henry.B.Hotz@jpl.nasa.gov, or hbhotz@oxy.edu