Date: Mon, 9 Nov 1998 11:09:55 -0600 From: Dan Nelson <dnelson@emsphone.com> To: Thomas David Rivers <rivers@dignus.com>, freebsd-hackers@freefall.cdrom.com Subject: Re: A stdio question... does fpos_t really need to be 'long long'? Message-ID: <19981109110955.A5248@emsphone.com> In-Reply-To: <199811091627.LAA23614@lakes.dignus.com>; from "Thomas David Rivers" on Mon Nov 9 11:27:10 GMT 1998 References: <199811091627.LAA23614@lakes.dignus.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In the last episode (Nov 09), Thomas David Rivers said: > Ok - here's a question for the stdio internal gurus... > > ftell() is defined to return a 'long' (32 bits). > fpos_t is defined as a 'long long' (64 bits). > fgetpos() accepts an fpos_t as it's second argument, and is > implemented as: > > retval = (*pos = ftell(fp)) == (fpos_t) -1; > return (retval); > > If this is the case... then why is fpos_t a 'long long'? If, in > fact, it can never be set that large? I noticed this back in April and commented on it. I believe the consensus was that fsetpos()/fgetpos()/fseek()/ftell() should be wrappers for the X/Open functions fseeko() and ftello(), which take off_t arguments. Unfortunately, I never submitted patches, and neither did anyone else. -Dan Nelson dnelson@emsphone.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981109110955.A5248>