On 2013-04-29 21:44, Gabriel L. Somlo wrote: > On Mon, Apr 29, 2013 at 09:06:37PM +0200, Jan Kiszka wrote: >>> Bisect says that commit 4fb60de74f93a601775dbda053b3237634d7a417 >> >> That commit alone is broken as it depends on the refactorings being >> selected by the submodule update. >> >> Does the problem persist with current master checked out into a clean >> folder? > > OK, after further digging, following your canonical steps > (i.e. "git submodule update --init") seems to work without problems. > > Instead, i was replacing the "linux" directory from kvm-kmod with a > symlink to ../kvm (itself obtained via "git clone"), in a (potentially > misguided) attempt to stay on the master branch of both projects. > > That *used* to work until the recent update. Now that I know I'm > "off the reservation", I haven't yet figured out how "reasonable" vs. > "bone-headed" of a thing that is... :)
Linking should work as well, though is unnecessary: make LINUX=/path/to/your/repo sync That's what I'm doing all the time to test my kvm hacks on my not-so-recent host kernel. Jan
signature.asc
Description: OpenPGP digital signature