> : Well kind of.. 'make install' stopped to segmentation fault and dumped core,
> : because make itself didn't like new libc ;). So right after the library were
> : at it place the next call to some sub-make dumped core and interrupted
> : installation.
> The 'make' failure is known and there is a way to get it work,
> please read the release notes (you find them in the same directory
> as libc 5.4.44) before you install!!!
Another datapoint - flawless install from sources on an
i486-pc-linux-gnulibc1 system. Even though I've got egcs, I compiled it
with gcc 184.108.40.206 because egcs seems to generate slightly larger binaries
than gcc (yes, even with -fomit-frame-pointer -fno-rtti -fno-exceptions).
I can find no mention in the sources of any make(1) failures and have
noticed no problems since recompilation on Tuesday. System's been up
constantly since then after a shutdown to ensure that everything was using
the new libc; no problems, no surprises, great work.
 if you ignore make(1)'s falling over because a .depend got corrupted
somehow... my fault, I think. Certainly it's got nothing to do with the
sources themselves because doing another `make depend clean; make`
`... I find that if I put certain kinds of meat in a casserole dish, in
water, and add an oxo cube and some salt, and forget about it for a
few hours, it goes soft enough to chew.' - the *real* art of cookery