Builder requires libgcc_s.so.1

[awokd]

Qubes OS version:

R4.0

Affected component(s):

builder

---

Steps to reproduce the behavior:

make clean
make clean-chroot
Attempt to make qubes with standard selections

Expected behavior:

Shiny new ISO

Actual behavior:

Error: package gcc-6.4.1-1.qubes1.fc25.x86_64 requires libgcc_s.so.1, but none of the providers can be installed

General notes:

---

Related issues:

#3650 ?

[awokd]

Only log is gcc-dom0-fc25.log, attaching.
gcc-dom0-fc25.log

[awokd]

Only log is gcc-dom0-fc25.log, attaching.
gcc-dom0-fc25.log

[awokd]

Inside the chroot:

bash-4.3# dnf install libgcc_s.so.1
Qubes OS Builder Repository                     137 MB/s | 522 kB     00:00    
Error: cannot install both libgcc-6.2.1-2.fc25.x86_64 and libgcc-6.4.1-1.qubes1.fc25.x86_64
(try to add '--allowerasing' to command line to replace conflicting packages)

Trying again with a full make dist-clean...

[awokd]

Inside the chroot:

bash-4.3# dnf install libgcc_s.so.1
Qubes OS Builder Repository                     137 MB/s | 522 kB     00:00    
Error: cannot install both libgcc-6.2.1-2.fc25.x86_64 and libgcc-6.4.1-1.qubes1.fc25.x86_64
(try to add '--allowerasing' to command line to replace conflicting packages)

Trying again with a full make dist-clean...

[awokd]

Looks like marmarek/qubes-vmm-xen@b79f692 is intended to address it.

[awokd]

Looks like marmarek/qubes-vmm-xen@b79f692 is intended to address it.

[marmarek]

There is gcc-6.4.1-1.qubes2.fc25 which have this problem resolved.
Also, this is a duplicate of #3650

[marmarek]

There is gcc-6.4.1-1.qubes2.fc25 which have this problem resolved.
Also, this is a duplicate of #3650

[awokd]

Had that linked as a related issue but the problem myself and a couple others on the mailing list are running into is that the standard make clean is not sufficient to get qubes-builder back in working order.
Will report over on the linked issue instead of this one once the gcc build finishes...

[awokd]

Had that linked as a related issue but the problem myself and a couple others on the mailing list are running into is that the standard make clean is not sufficient to get qubes-builder back in working order.
Will report over on the linked issue instead of this one once the gcc build finishes...