[Rpm-maint] RPM 4.13.0 rc2 released
pmatilai at laiskiainen.org
Sat Oct 22 10:18:46 UTC 2016
On 10/21/2016 02:42 PM, Panu Matilainen wrote:
> On 10/21/2016 02:21 PM, Thierry Vignaud wrote:
>> Might also be a different fakechroot version...
>> Nope, both are using 2.18.
> I'm actually using 2.16 because on my work laptop, because with
> fakechroot 2.18 from Fedora practically everything fails. Haven't had a
> chance to investigate, not really looking forward to it either.
>> Though Mageia has this rpm related patch:
>> "Workaround for glob* vs rpm glob*
>> glob/glob64 is also defined by rpm in librpmio, which breaks fakechroot.
>> This patch tries to find which library is defining the symbol
>> (using dladdr, which is gnu specific) and if found, check if the symbol
>> comes from libc. If it's not from libc, call it directly."
> Right, that seems to ring some faint, distant bell. Sounds about right
> for the failing tests too since they're all about "file not found" mumble.
The mystery is deeper than that.
Like said, on my work laptop, with fakechroot 2.18 the testsuite barfs
big time, 144 cases fail. Downgrading to (self-built) fakechroot 2.10
However on my home laptop, the testsuite passes with flying colors with
They both are running Fedora 24, the difference is that the work laptop
was installed something like five years ago and since then
dist-upgraded, home laptop started its new life as Fedora 23 half a year
Wife and kid are away for the weekend and here I am, somehow ended up
investigating rpm testsuite failures. Doh :)
Testing with a different, "clean" user account revealed that it's
something in my accounts environment. On further investigation, it's a
symlink in the git repository path (~/repos is a symlink into /srv
partition) that somehow throws fakechroot 2.18 off.
Oh well, mystery solved, to some extent at least. And now off to
something completely different, really...
- Panu -
More information about the Rpm-maint