Hi Fabrice, On Tue, Sep 15, 2009 at 07:21:24AM +0200, fabrice wrote: > Sebastian Harl escribió: > > Uhm … I've misread that on the first try. It's not that that function is > > called with a non-const argument, but rather the argument is passed > > along to another function. Line 198 of pfspanoramic.cpp reads: > > > > if(opts = strchr(name, '/')) > > > > … where 'name' is the argument of the surrounding function: > > > > static Projection *getProjection(const char *name) > > > > According to POSIX (and that's how it's defined on Debian), strchr() is > > defined as: > > > > char *strchr(const char *s, int c); > > In gcc 4.4 headers (string.h), strchr is defined in 2 ways:
Hrm … talking about headers is a good point. Are you talking about /usr/include/string.h? That'd be provided by libc (rather than GCC). As you might know, Debian has recently switched to eglibc. Did Ubuntu do so as well? Which version of libc do you use? > when __OPTIMIZE__ is set > char *strchr (char *__s, int __c) > __const char *strchr (__const char *__s, int __c) *Both* definitions are active at the same time? Unfortunately, neither of those is POSIX-compliant :-/ > otherwise: > char *strchr (__const char *__s, int __c) This one is, though. > > So, there's nothing wrong with that. If this causes a build-failure on > > Ubuntu, something's wrong with the build environment. > > Also, it doesn't seems to be a problem with the build environment, as > it's failing to build in the buildd of Ubuntu > (https://edge.launchpad.net/ubuntu/+source/pfstools/1.8.1-1build1/+build/1207611/+files/buildlog_ubuntu-karmic-i386.pfstools_1.8.1-1build1_FAILEDTOBUILD.txt.gz). Sorry, I was a bit unclear on that. I was talking about the environment in general, i.e. including the tool chain, etc. > So, rather a problem with gcc 4.4? So, yes, this does look like a problem in GCC / libc to me (or some missing defines to ask for POSIX-compliant features). > Anyway, as upstream proposed the fix (if I am correct), I'll patch the > source in Ubuntu to be able to build pfstools. Yes, the "fix" was proposed by upstream. I'd rather call that a work- around for strange compiler behavior though, but I might be a bit "extreme" in that respect (i.e. standard-compliance) ;-) Anyway, I guess, I'll include the fix in the next upload as well (if that happens before the next upstream release). I'm a bit busy at the moment, though, so that won't happen within the next few days. Thanks for the follow-up! Cheers, Sebastian -- Sebastian "tokkee" Harl +++ GnuPG-ID: 0x8501C7FC +++ http://tokkee.org/ Those who would give up Essential Liberty to purchase a little Temporary Safety, deserve neither Liberty nor Safety. -- Benjamin Franklin
signature.asc
Description: Digital signature
