On 05/09/15 09:25, Jacco Ligthart wrote:
packages that I could not update:
====================
biosdevname-0.3.6-11.fc13.src.rpm
I'm moderately sure this is x86 specific.
firefox-24.2.0-1.el6.0.src.rpm
libvirt-0.10.2-29.el6.2.src.rpm
lldpad-0.9.43-20.el6.src.rpm
microcode_ctl-1.17-4.el6.0.src.rpm
As is this.
mstflint-1.4-6.el6.src.rpm
mvapich2-1.6-3.el6.src.rpm
ocaml-libvirt-0.6.1.0-6.2.el6.src.rpm
openmotif-2.3.3-6.1.el6.src.rpm
parted-2.1-21.el6.src.rpm
thunderbird-24.2.0-1.el6.0.src.rpm
util-linux-ng-2.17.2-12.14.el6.src.rpm
virt-top-1.0.4-3.11.el6.src.rpm
xulrunner-17.0.8-3.el6.0.src.rpm
I'm pretty sure this is deprecated - it has been un-separated from the
firefox package.
I know FF and TB don't need it any more, I did however need it to build
PackageKit, icedtea-web and libproxy.
Won't this build against the firefox package? Or does the last buildable
ff have separate xulruner?
Packages added:
==========
Konsole output Konsole output
google-crosextra-caladea-fonts-1.002-0.3.20130214.el6.1
google-crosextra-carlito-fonts-1.103-0.1.20130920.el6.1
json-c-0.11-12.el6
libgovirt-0.3.2-1.el6
numactl-2.0.9-2.el6.0
I'm pretty sure this is x86 specific. What depends on it? Is there a
configuration option to disable this dependency?
irqbalance depends on it.
I'm not sure it's x86 specific. I found the patch here:
https://wiki.linaro.org/LEG/Engineering/Kernel/NUMA
here it says:
"There aren't any ARM NUMA systems out at the moment, but one can use
the Linux NUMA subsystem to group cores that share an L2 cache together
into the same node. For certain workloads this can then result in a
decrease in the accesses to main memory."
I guess we can argue both ways :)
Fair enough, if it works. :)
packages available upstream but not build:
==========================
PyYAML-3.10-3.1.el6
amtu-1.0.8-11.el6
anaconda-13.21.239-1.el6.centos
busybox-1.15.1-21.el6_6
IIRC this went a little weird from a certain version on, purely due to
the .spec file. I can't remember the gory details off the top of my head.
cfg80211_dup-0.1_rh1-1.el6_6
clufter-0.11.2-1.el6
compat-dapl-1.2.19-2.el6
===
compat-gcc-296-2.96-144.el6
compat-gcc-32-3.2.3-69.el6
compat-gcc-34-3.4.6-19.el6
compat-glibc-2.5-46.2
I'm pretty sure we can live without these.
compat-openmpi-1.4.3-1.2.el6
convmv-1.15-2.el6
cpupowerutils-1.2-7.el6
crash-gcore-command-1.0-5.el6
cryptsetup-reencrypt-1.6.4-1.el6
dapl-2.0.34-2.el6
eclipse-3.6.1-6.13.el6
eclipse-birt-2.6.0-1.1.el6
eclipse-callgraph-0.6.1-1.el6
eclipse-cdt-7.0.1-4.el6
eclipse-changelog-2.7.0-1.el6
eclipse-dtp-1.8.1-1.1.el6
eclipse-emf-2.6.0-1.el6
eclipse-gef-3.6.1-3.el6
eclipse-linuxprofilingframework-0.6.1-1.el6
eclipse-mylyn-3.4.2-9.el6
eclipse-nls-3.6.0.v20120721114722-2.el6
eclipse-oprofile-0.6.1-1.el6
eclipse-rpm-editor-0.5.0-2.el6
eclipse-rse-3.2-1.el6
eclipse-subclipse-1.6.5-6.el6
eclipse-valgrind-0.6.1-1.el6
I never did get to the bottom why eclipse wasn't building. It took so
long to build that I eventually gave up because it meant waiting days
to see if each change I made resulted in a small amount of progress.
:( It may be less painful on an 8-core machine with 4GB of RAM, but on
a SheevaPlug (1.2GHz single core with 512MB of RAM) it wasn't really
workable.
I tried a little. It's difficult, because it depends on itself. I could
not get it to bootstrap after some reasonable attempts, went to google
and found your unanswered questions in various mailinglists :)
I think we can leave it for now - now ARM machine has enough RAM to run
eclipse anyway. :)
foghorn-0.1.2-1.el6
gcc-libraries-5.0.0-1.1.1.el6
gdisk-0.8.10-1.el6
gnu-efi-3.0g-2.el6
gprolog-1.3.1-6.el6
gpxe-0.9.7-6.14.el6
grub-0.97-94.el6
haproxy-1.5.4-2.el6
It surprises me that something as trivial as haproxy didn't build. Do
you have a build log?
No, because I didn't try :)
Until now I only updated packages that were previously in RSEL6. How do
we go from here? cherry picking? or just try and build them all and see
what happens?
A mass build sounds like a good place to start. See what just works.
Cherry pick from the FTBFS if we need to. :)
_______________________________________________
users mailing list
[email protected]
http://lists.redsleeve.org/mailman/listinfo/users
