@@ -1119,7 +1119,7 @@ def setscene_depvalid(task, taskdependees, notneeded, d, log=None):
# This is due to the [depends] in useradd.bbclass complicating matters
# The logic *is* reversed here due to the way hard setscene dependencies are injected
- if (taskdependees[task][1] == 'do_package' or taskdependees[task][1] == 'do_populate_sysroot') and taskdependees[dep][0].endswith(('shadow-native', 'shadow-sysroot', 'base-passwd', 'pseudo-native')) and taskdependees[dep][1] == 'do_populate_sysroot':
+ if (taskdependees[task][1] == 'do_package' or taskdependees[task][1] == 'do_populate_sysroot') and taskdependees[dep][0].endswith(('shadow-sysroot', 'base-passwd', 'pseudo-native')) and taskdependees[dep][1] == 'do_populate_sysroot':
continue
# Consider sysroot depending on sysroot tasks
This results in executables from shadow-native, such as useradd, falling through to host libraries instead of using native ones that they were linked with, because the native libraries would not be installed because of this dependency exclusion. Until now that could be coincidentally fulfilled with host libattr (which is present on all hosts), but latest useradd also wants libbsd and libmd, and those are not typically installed. The original reason to add this exclusion was: https://git.yoctoproject.org/poky/commit/?id=0d2a301bca4eb9d43243f1a03c1cd1ab3f27f30a While I can't be sure the reason no longer applies, sstate and sysroot handling has changed massively since early 2013, testing on a-full revealed no issues, and avoiding incorrect fall-through to host libraries is better than pulling unneeded dependencies. Signed-off-by: Alexander Kanavin <alex@linutronix.de> --- meta/classes-global/sstate.bbclass | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)