src/pypungi/__init__.py | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/pypungi/__init__.py b/src/pypungi/__init__.py
index 72759a3..4769427 100644
@@ -238,9 +238,9 @@ class Pungi(pypungi.PungiBase):
self.ayum.compatarch = yumarch
# Filter out all the multilib arches, anaconda won't use them.
- # only makes sense on x86_64
+ # only makes sense on x86_64 and s390x
full_archlist = set(yum.rpmUtils.arch.getArchList(yumarch))
- if arch == 'x86_64':
+ if arch in ['x86_64', 's390x']:
compat_archinfo = yum.rpmUtils.arch.getMultiArchInfo(yumarch)
compat_archlist = set(yum.rpmUtils.arch.getArchList(compat_archinfo))
newarchlist = list(full_archlist.difference(compat_archlist))
So I'm not sure the best forum to ask this in, here seems reasonable,
though not directly related to koji/mash/etc (but could be :) ).
Here's the situation:
We've got 3 or 4 buckets of content that we care about as an OS
engineering group. First, is the stuff that is on the path to
production for the current build that we're working on. Second, stuff
that's released from Red Hat - i.e. whatever exists in RHN channels
right now. Third, is beta stuff that we get from Red Hat - some of
which may never see the light of day in prod. We also have our own
content that we put into each of these.
The problem lies in how to maintain these three separate repos,
considering that when we're working on a build, we continue to get
errata from Red Hat. For some period of time, we'll just blindly take
the updates into the next release. But after some certain "freeze"
point, we cherry pick stuff from the current Red Hat bits. Then
there's the beta bits that we need to be able to build in the lab, and
in a worst case scenario (hasn't happened, but I can see where it
might) we cherry pick from the beta bits to go to prod.
How do people handle this? I don't believe that I'm in a particularly