On Thu, May 16, 2019 at 07:48:30AM -0700, Richard Henderson wrote: > On 5/15/19 9:53 AM, Daniel P. Berrangé wrote: > > So I think we should just use $(crypto-obj-y) unconditionally in > > the user emulators, and get rid of crypto-aes-obj-y too. > > That results in > > LINK arm-linux-user/qemu-arm > ../crypto/tlssession.o: In function `qcrypto_tls_session_check_certificate': > /home/rth/qemu/qemu/crypto/tlssession.c:356: undefined reference to > `qauthz_is_allowed_by_id' > collect2: error: ld returned 1 exit status > > which means all of the authz objects need to come in as well. > I suppose they're not big, but still... > > I'm leaning toward reviving crypto-user-obj-y, with just the > crypto random and aes objects in. > > Thoughts?
Coincidentally I think I need to split the $authz-obj-y variable into two parts authz-obj-y containing only base.o authz-obj-impl-y containing everything else Most things will thus be satisfied by just $authz-obj-y. Only the system emulators and qemu-nbd then need to have authz-obj-impl-y This will avoid qemu-img, qemu-io, and other tools and linux-user from having to link to the pam library. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|