On Tue, May 29, 2012 at 04:43:43PM -0500, Bruce Dubbs wrote: > Ken Moffat wrote: > > > You seem to be making good progress here. I've only managed to do > > a DESTDIR install for the accelerometer ... v4l_id batch of programs > > and for libudev.* : Haven't managed to build udevadm or > > systemd-udevd without DBUS at the moment. I'll take a look at your > > details later, to see if I can steal any ideas :) > > Use whatever you want. > > -- Bruce
I'm Trying to get around the libtoolisms, and failing badly. Can't compile udev-builtin-firmware.c [ none of those -builtin files were being picked up ], it needs FIRMWARE_PATH which is only defined in the standard Makefile in libudev_core_la_CPPFLAGS. Trying a new udevadmlfs target, with those modules in the udevadmlfs_OBJECTS. Tried setting udevadmlfs_CPPFLAGS to include that define, but it isn't being used. Needless to say, this is still trying to work with the shipped configure and the resulting Makefile. Giving up for the night. Forking udev starts to seem more attractive. Well, at least until I looked at the amount of commits in the systemd git tree. I've gone back about 141 commits but I'm still only back to 14th May. The last commit in udev.git at the kernel was on 29th March. Of those 141 commits, 25 are since systemd-183 was released, of which 4 look useful, 1 looks possibly deleterious, and 2 more are probably useful. Not sure I've got the time or motivation to get back to when the merge to systemd took place. Anyone for makedev.sh ? ;-) ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
