On 12/13/17 9:15 PM, Tom Rini wrote: > On Wed, Dec 13, 2017 at 09:45:13PM -0500, Tom Rini wrote: >> On Wed, Dec 13, 2017 at 02:40:34PM +0000, Burton, Ross wrote: >> >>> I finally made it break locally, with MACHINE=qemux86 and using sysvinit >>> instead of systemd. Basically, a pristine poky local.conf. I've not dug >>> into it beyond that, or why it passes with systemd. >> >> I found it, and, erm, ugh? Here's what appears to be a systemd vs >> sysvinit difference. With systemd, if you have an fstab entry and the >> mount point doesn't exist, it will get made. So an entry of /uuid (a >> very non-HFS and just made up directory) the mount happens. With >> sysvinit, no, the mount fails. That's why the test fails. >> >> My gut reaction is that no, it's not wic's job to ensure that the >> underlying filesystems have all mount points and I should just re-do the >> test to use /mnt instead of /uuid, for the UUID test mount point. > > And in fact, systemd just can't be used for oe-selftest and have > everything expect to pass today. The existing wic.Wic.test_qemu will > fail on /dev/root vs /dev/sda2. So I'm going to post a v4 that puts the > UUID one into /media, and I've run the tests and they pass atm. Thanks! >
/media is usually automounted when removable media is inserted. you might want to use something else. -- _______________________________________________ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core