I've made some progress tracking down the issues with kernel 4.18 booting
on i686, which is causing all of the i686 images to fail QA.
The quick summary is something introduced in kernel 4.18 (after rc0.git2.1)
is causing systemd to coredump, dropping the user to a dracut prompt. I was
able to reproduce this on a Fedora 28 install with the default systemd in
I've filed the following BZ:
I just tried a rebuild of a package in Rawhide, it worked on most
arches, but failed on armv7hl and i686 (our only 32-bit arches). On
both arches, cmake seems to have segfaulted:
+ /usr/bin/cmake -DCMAKE_C_FLAGS_RELEASE:STRING=-DNDEBUG
/var/tmp/rpm-tmp.0imLnb: line 47: 15870 Segmentation fault (core
dumped) /usr/bin/cmake -DCMAKE_C_FLAGS_RELEASE:STRING="-DNDEBUG"
Just wanted to flag this up. I haven't yet tried doing the build in a
mock to try and get the core out, or anything like that.
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
It looks like the 32-bit PTI patches are queued up and are expected to be pulled for the next kernel merge widow. This means they will be coming into rawhide sometime in the middle of August. Everything _should_ be tested by now but as we all know bugs can come up. I wanted to make sure the SIG was aware of this to keep an eye out for unexpected problems. The patches are at https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=x86/pti if anyone wants to test on their own.
FESCo met today to discuss an i686 build failure in which it seems that the x86 SIG did not have a timely response, which resulted in wider-than-expected problems in the distribution due to causing build failures in other dependent packages. FESCo hereby notifies the x86 SIG that future events like this one may result in dropping i686 support from Fedora.
The kernel team has proposed splitting out the kernel-headers package to alleviate similar problems in the future.
I'm going to take a whack at triaging x86-related bugs. I see a few listed
on the x86 tracker bug and the x86 exclude arch bug, which I'll take a
closer look at.
Before I start searching for anything i686-related, are there any issues
you'd like us to take a look at first?
My apologies, there were a group of emails that had been sent to the
list by non-subscribers which were held up. I had missed them in the
daily flood of "RE: offer knitting fabrics to Disney" emails we get
Stephen J Smoogen.