Is this location (crypto/.libs) deliberate or has something gone haywire? I should expect the .so in atom/.libs along with the remainder of the loadable modules.
I suspect that make was done in the crypto directory rather than in atom; that is considered a bad idea by makefile aficionados. Instead, include a makefile from crypto that adds the source modules with the source directory.
HHC01603I modpath /usr/data/src/hercules/atom/crypto/.libs
HHC01603I ldmod dyncrypt
HHC01526I HDL: loading module dyncrypt...
HHC00150I Crypto module loaded (c) Copyright 2003-2015 by Bernard van der Helm
HHC00151I Activated facility: Message Security Assist
HHC00151I Activated facility: Message Security Assist Extension 1, 2, 3 and 4
HHC01527I HDL: module dyncrypt loaded
I can do this circumvention in the configuration file:
# This seems to be a bug in crypto build:
modpath crypto/.libs
ldmod dyncrypt
# And now where device managers live
modpath .libs
However that will fail if I specify -p .libs on the hercules command.
atom is the out-of-source build directory.
Is this location (
crypto/.libs
) deliberate or has something gone haywire? I should expect the .so in atom/.libs along with the remainder of the loadable modules.I suspect that make was done in the crypto directory rather than in atom; that is considered a bad idea by makefile aficionados. Instead, include a makefile from crypto that adds the source modules with the source directory.
I can do this circumvention in the configuration file:
However that will fail if I specify
-p .libs
on the hercules command.