On Wed, Feb 17, 2010 at 9:23 AM, Barry Scott <[email protected]> wrote: > On 16/02/10 20:30, Jeff Trawick wrote: >> >> On Tue, Feb 16, 2010 at 3:24 PM,<[email protected]> wrote: >> >>> >>> Author: poirier >>> Date: Tue Feb 16 20:24:33 2010 >>> New Revision: 910673 >>> >>> URL: http://svn.apache.org/viewvc?rev=910673&view=rev >>> Log: >>> Fix compile warning (discarding constness of fname) >>> >>> Modified: >>> httpd/httpd/trunk/server/config.c >>> >>> Modified: httpd/httpd/trunk/server/config.c >>> URL: >>> http://svn.apache.org/viewvc/httpd/httpd/trunk/server/config.c?rev=910673&r1=910672&r2=910673&view=diff >>> >>> ============================================================================== >>> --- httpd/httpd/trunk/server/config.c (original) >>> +++ httpd/httpd/trunk/server/config.c Tue Feb 16 20:24:33 2010 >>> @@ -1670,7 +1670,7 @@ >>> int current; >>> >>> /* find the first part of the filename */ >>> - rest = ap_strchr(fname, '/'); >>> + rest = ap_strchr((char*)fname, '/'); >>> >> >> Casting isn't the right fix, which I guess is to make rest const char >> * and then use ap_strchr_c() instead of ap_strchr() (hopefully that >> doesn't snowball). >> >> > > This is a common problem that the C RTL solves using the > idiom of accepting a const char * as input and returning > a char * as output: > > char *strchr(const char *s, int c); > > You may wish to fix the ap_strchr signature to follow the > C standard solution rather then invite snow balling.
The C standard signatures for strchr() and strstr() are busted and don't allow some simple program errors to be caught at compile time, so httpd provides different flavors which take char * or const char * as input depending on what operations on the returned pointer are desired. (You have to build in maintainer mode to trigger the compile warnings for misuse of the httpd APIs.) The fix was simply to use the httpd APIs properly.
