Bruno Haible <[EMAIL PROTECTED]> wrote: > Jim Meyering wrote: >> Do you feel like adding a few coreutils tests that currently fail >> on common platforms but that would succeed if coreutils switched? >> E.g., here: >> coreutils/tests/seq/basic >> coreutils/tests/misc/printf > > Yes, sure. Which tests exactly do you want to be included in gnulib? > (I don't see any of the coreutils tests being commented out.)
I was asking if you could construct printf and seq tests _in coreutils_ (in those tests/* files I mentioned) that would currently fail. ... > On a glibc 2.5 system, we should have: ... > $ ls -l printf.o vfprintf.o vasnprintf.o > /bin/ls: printf.o: No such file or directory > /bin/ls: vfprintf.o: No such file or directory > -rw-r--r-- 1 bruno user 20092 May 20 15:28 vasnprintf.o > $ size vasnprintf.o > text data bss dec hex filename > 3746 0 0 3746 ea2 vasnprintf.o > > So, on a glibc 2.5 system, the size increase of executables will be 0 bytes, > since vasnprintf.o will not be linked in (coreutils only needs printf(), > not vasnprintf()). Good! I hope it stays that way.
