Status of libtool 2.2.X?

[Kevin Kofler]

Till Maas wrote:
> Also does %cmake also make sure that -O3 / -s is not passed to gcc? Also
> it seems imposible to see, whether cmake uses "install -s" to install
> binaries or not.

CMake doesn't use the install binary at all, it creates a CMake script with
the installations to perform and runs cmake -P with that script. So there
are no command invocations to show. I'm not sure whether that even supports
stripping, I've yet to see a CMake-using program which tries to strip on
install.

But to be sure, just look at the contents of the -debuginfo package, you'll
see immediately when it is invalid (it will be empty or almost empty). If
there's useful debuginfo, the files haven't been stripped.

As for -O3, that's not necessarily a problem, I'd be more worried about
things like -fno-stack-protector (i.e. "Listen crackers, this code is so
full of buffer overflows that it doesn't work at all with buffer overflow
protection, and so I helpfully disabled it for you, have fun!" ;-) ).

> Can %make maybe modified in a way, that it is possible to use something
> like "rpmbuild --without cmake_verbose" or add something to ones
> .rpmmacros file to enable/disable this for local/mock rebuilds?

That (but in %cmake and %cmake_kde4, not %make) sounds like a good idea.

        Kevin Kofler