Anaconda-devel May 2010

anaconda-devel@lists.fedoraproject.org

25 participants
180 discussions

[PATCH master] nm-c-e integration: fix some leftovers from patch porting.
by Radek Vykydal 26 May '10

26 May '10

We don't have InstallData on master (patch was ported from rhel). --- gui.py | 5 ++--- 1 files changed, 2 insertions(+), 3 deletions(-) diff --git a/gui.py b/gui.py index d3423d1..efb510c 100755 --- a/gui.py +++ b/gui.py @@ -954,7 +954,6 @@ class InstallInterface(InstallInterfaceBase): pass - # TODORV: handle one device case - don't ask # just_setup is used for [Configure Network] button def enableNetwork(self, just_setup=False): @@ -987,7 +986,7 @@ class InstallInterface(InstallInterfaceBase): if network.hasWirelessDev(): # NOTE: For wireless, we need supplicant to go to ready state, # that means to get the wireless device managed by NM - self.anaconda.id.network.writeIfcfgFiles() + self.anaconda.network.writeIfcfgFiles() w = self.anaconda.intf.waitWindow(_("Wireless setup"), _("Scanning access points for wireless devices")) # get available wireless APs @@ -999,7 +998,7 @@ class InstallInterface(InstallInterfaceBase): dev_ssids = selectSSIDsDialog(dev_all_ssids) or dev_all_ssids self.anaconda.network.updateIfcfgsSSID(dev_ssids) - self.anaconda.id.network.writeIfcfgFiles() + self.anaconda.network.writeIfcfgFiles() network.logIfcfgFiles(header="========== before nm-c-e run\n") runNMCE(self.anaconda) network.logIfcfgFiles(header="========== after nm-c-e run\n") -- 1.6.0.6

2 1

[PATCH] Fix 2 variable references in formats/fs.py
by Hans de Goede 26 May '10

26 May '10

--- storage/formats/fs.py | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/storage/formats/fs.py b/storage/formats/fs.py index 7c0b19b..7a2f1a9 100644 --- a/storage/formats/fs.py +++ b/storage/formats/fs.py @@ -369,7 +369,7 @@ class FS(DeviceFormat): w.pop() if ret.rc: - raise FormatCreateError("format failed: %s" % rc, self.device) + raise FormatCreateError("format failed: %s" % ret.rc, self.device) self.exists = True self.notifyKernel() @@ -478,7 +478,7 @@ class FS(DeviceFormat): w.pop() if ret.rc: - raise FSResizeError("resize failed: %s" % rc, self.device) + raise FSResizeError("resize failed: %s" % ret.rc, self.device) self.doCheck(intf=intf) -- 1.7.0.1

2 1

[rhel6-branch] Fixed the setting of LD_LIBRARY_PATH in rescue (#594034)
by Martin Gracik 26 May '10

26 May '10

The hardcoded setting of LD_LIBRARY_PATH was not taking into consideration if there are lib or lib64 directories, and was also overwriting everything loader put there before. Changed to just prepend "/mnt/sysimage" to the directories we already have in the path. --- rescue.py | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/rescue.py b/rescue.py index 632d6ad..f360cc7 100644 --- a/rescue.py +++ b/rescue.py @@ -411,7 +411,9 @@ def runRescue(anaconda, instClass): fd.close() # set a library path to use mounted fs - os.environ["LD_LIBRARY_PATH"] = "/lib:/usr/lib:/usr/X11R6/lib:/lib:/mnt/usr/lib:/mnt/sysimage/lib:/mnt/sysimage/usr/lib:/mnt/sysimage/usr/X11R6/lib" + libdirs = os.environ["LD_LIBRARY_PATH"].split(":") + mounted = map(lambda dir: "/mnt/sysimage%s" % dir, libdirs) + os.environ["LD_LIBRARY_PATH"] = ":".join(libdirs + mounted) # find groff data dir try: -- 1.6.6.1

2 3

[PATCH rhel6-branch] Add nm-c-e locales to stage 2 (#594982)
by Radek Vykydal 26 May '10

26 May '10

--- scripts/upd-instroot | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/scripts/upd-instroot b/scripts/upd-instroot index 22b7192..bbacf03 100755 --- a/scripts/upd-instroot +++ b/scripts/upd-instroot @@ -680,6 +680,7 @@ usr/share/locale/*/LC_MESSAGES/anaconda.mo usr/share/locale/*/LC_MESSAGES/cracklib.mo usr/share/locale/*/LC_MESSAGES/gtk20.mo usr/share/locale/*/LC_MESSAGES/libc.mo +usr/share/locale/*/LC_MESSAGES/nm-applet.mo usr/share/locale/*/LC_MESSAGES/parted.mo usr/share/locale/*/LC_MESSAGES/redhat-dist.mo usr/share/locale/*/LC_MESSAGES/system-config-date.mo -- 1.6.0.6

2 1

[master rhel6-branch] zFCP component for cio_ignore functionality
by David Cantrell 26 May '10

26 May '10

The what component for what functionality? Yeah, exactly. So we have this capability on s390x to prevent scanning of all devices on boot up and then you just enable the devices you need during installation. We have had that for network and dasd, this patch rounds that out by adding in zFCP support. I've done test installs passing FCP_x lines at boot time as well as using the Add Advanced Storage dialog and passing the parameters that way. The nice thing is that most of the hard work has been offloaded to the /sbin/zfcp_cio_free command. Patch developed on rhel6-branch, but intended for master too. Don't want to leave those Fedora s390x users hanging. -- David Cantrell <dcantrell(a)redhat.com> Red Hat / Honolulu, HI

3 3

[PATCH mkinitrd] Use location of bindings file from multipath.conf (#593050)
by Brian C. Lane 25 May '10

25 May '10

Copy the bindings file from the location in the multipath.conf file, or copy from /var/lib/multipath/bindings Resolves: rhbz#593050 --- mkinitrd | 12 ++++++++++-- 1 files changed, 10 insertions(+), 2 deletions(-) diff --git a/mkinitrd b/mkinitrd index 5a4ad2e..cd98ba0 100755 --- a/mkinitrd +++ b/mkinitrd @@ -1606,9 +1606,17 @@ if [ "$use_multipath" == "1" ]; then if [ -f /etc/multipath.conf ]; then inst /etc/multipath.conf $MNTIMAGE/etc/multipath.conf fi - if [ -f /var/lib/multipath/bindings ]; then - inst /var/lib/multipath/bindings $MNTIMAGE/var/lib/multipath/bindings + + if [ -f /etc/multipath.conf ]; then + if [ "$(egrep '^[[:space:]]*user_friendly_names[[:space:]]+yes' /etc/multipath.conf)" ] ; then + bindings=$(awk '/^[[:blank:]]*bindings_file[[:blank:]]/ { gsub(/"/,""); print $2 }' /etc/multipath.conf) + fi fi + bindings=${bindings:-/var/lib/multipath/bindings} + if [ -f $bindings ]; then + inst $bindings $MNTIMAGE/$bindings + fi + inst /sbin/scsi_id $MNTIMAGE/bin/scsi_id inst /etc/scsi_id.config $MNTIMAGE/etc/scsi_id.config # This includes all mpath_prio checkers in the image (~2.5M) -- 1.6.6.1

3 2

[PATCH rhel6-branch] Skip the bootloader placement window if we're on UEFI (#582143)
by Peter Jones 25 May '10

25 May '10

Since we don't install a bootloader physically to a disk on UEFI, and it doesn't use an MBR, there's no real point in selecting which MBR to stick the bootloader on. --- installclass.py | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/installclass.py b/installclass.py index deb7e53..236f778 100644 --- a/installclass.py +++ b/installclass.py @@ -129,7 +129,7 @@ class BaseInstallClass(object): if not BETANAG: dispatch.skipStep("betanag", permanent=1) - if not iutil.isX86(): + if iutil.isEfi() or not iutil.isX86(): dispatch.skipStep("bootloader", permanent=1) # allow backends to disable interactive package selection -- 1.7.0.1

2 1

[master rhel6-branch] s390 patch and nfs ks patch
by David Cantrell 25 May '10

25 May '10

1) Fix utils/addrsize.c to actually write out the size. 2) Add a missing newline to certain "nfs" lines in the ks file we write out. Both for master and rhel6-branch. -- David Cantrell <dcantrell(a)redhat.com> Red Hat / Honolulu, HI

3 5

[rhel6-branch] Add fipscheck to the install.img (#592186)
by Martin Gracik 25 May '10

25 May '10

--- scripts/upd-instroot | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/scripts/upd-instroot b/scripts/upd-instroot index 22b7192..c16eade 100755 --- a/scripts/upd-instroot +++ b/scripts/upd-instroot @@ -165,7 +165,7 @@ PACKAGES="GConf2 NetworkManager NetworkManager-gnome ORBit2 acl anaconda device-mapper-multipath device-mapper-multipath-libs dhclient dmraid dmraid-libs dogtail dosfstools e2fsprogs e2fsprogs-libs echo-icon-theme eject ethtool - elfutils-libelf expat fcoe-utils + elfutils-libelf expat fcoe-utils fipscheck firstboot fontconfig fonts-ISO8859-2 freetype gail gawk gdb-gdbserver gdk-pixbuf gfs2-utils glib2 glibc-common gnome-python2-canvas gnome-python2-gconf gnome-themes gpm grep gtk2 gtk2-engines hal @@ -538,6 +538,7 @@ usr/bin/dbus-launch usr/bin/du usr/bin/eject usr/bin/expr +usr/bin/fipscheck usr/bin/gdb-gdbserver usr/bin/gdialog usr/bin/gdk-pixbuf-query-loaders* -- 1.6.6.1

3 3

Restarting anaconda.
by Ales Kozumplik 25 May '10

25 May '10

Hi, While grooming rawhide for work on the threaded UI, I made a patch providing a shell script to restart the stage2 anaconda. This is for our internal development and debugging so we can avoid some reboots every day. There had to be a change in the loader too because the script needs to communicate with it that anaconda should be restarted. The mechanism this works is by writing an empty /tmp/restart-anaconda file and then killing anaconda and all its descendant processes. Even though I think this is fine, I welcome suggestions about a nicer/better way to do this. I tested restarting anaconda early after its start and it works pretty reliably, even after storageinit. I think we will see problems when one does the restart after e.g. /mnt/sysimage is mounted or similar, but I'll try to fix those too so the restart method becomes pretty universal. The second patch re-downloads updates.img with each restart so changes to the code can be done that way. (another way is modifying an uncompressed stage2 image directly, described here https://www.redhat.com/archives/anaconda-devel-list/2010-April/msg00359.html. That's more general, one can even add libs etc. to the image without rebooting.) Hope you'll find this useful. Ales

4 11

← Newer
1
2
3
4
5
6
7
...
18
Older →

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel May 2010