On Mon, Jul 22, 2002 at 02:43:00PM -0700, Dan Kegel wrote: > Thomas Dickey wrote: > > that's certainly an understatement (the ifdef's in glibc probably have been > > reasonably well-tested only for the combination that corresponds to > > _GNU_SOURCE - after seeing a number of obvious errors, I decided that it > > must be intentional). > > Really? Do you have a test case? Maybe I can add a regression test to glibc > for it. Here, I use
not offhand - I spent a chunk of time attempting to get it to compile some ANSI+POSIX stuff that works fine on other platforms w/o adding definitions, and didn't succeed (so I added a _GNU_SOURCE test ;-). Occasionally when I'm grep'ing through the headers I come across stray functions that don't appear to belong. But I haven't gone through them systematically (judging by what google tells me, no one has). (Aside from ifdef's, it would be nice to have a set of system header-files that don't generate lots of compiler-warnings for my normal set of gcc options - I complained about that more than once, was told it was fixed, but only saw it get worse ;-) > #define _XOPEN_SOURCE 500 > > to get access to pread() without any trouble. > - Dan -- Thomas E. Dickey <[EMAIL PROTECTED]> http://invisible-island.net ftp://invisible-island.net
