July 2009 - Anaconda-devel - Fedora mailing-lists

[PATCH 1/3] Fix lvm on top of dmraid
by Hans de Goede 15 Jul '09

15 Jul '09

The lvm rules were failing to identify PV's on mdraid because blkid/vol_id has not been run on md raidsets by previous rules. --- modules.d/90lvm/64-lvm-vol_id.rules | 1 + modules.d/90lvm/64-lvm.rules | 1 + 2 files changed, 2 insertions(+), 0 deletions(-) diff --git a/modules.d/90lvm/64-lvm-vol_id.rules b/modules.d/90lvm/64-lvm-vol_id.rules index d3a07c1..f5f3a6f 100644 --- a/modules.d/90lvm/64-lvm-vol_id.rules +++ b/modules.d/90lvm/64-lvm-vol_id.rules @@ -8,6 +8,7 @@ SUBSYSTEM!="block", GOTO="lvm_end" ACTION!="add|change", GOTO="lvm_end" KERNEL=="dm-*", IMPORT{program}="vol_id --export $tempnode" +KERNEL=="md*", IMPORT{program}="/sbin/blkid -o udev -p $tempnode" ENV{ID_FS_TYPE}=="LVM2_member", \ TEST!="/tmp/.lvm_scan-%k", \ diff --git a/modules.d/90lvm/64-lvm.rules b/modules.d/90lvm/64-lvm.rules index 3403020..a1c2d1f 100644 --- a/modules.d/90lvm/64-lvm.rules +++ b/modules.d/90lvm/64-lvm.rules @@ -8,6 +8,7 @@ SUBSYSTEM!="block", GOTO="lvm_end" ACTION!="add|change", GOTO="lvm_end" KERNEL=="dm-*", IMPORT{program}="/sbin/blkid -o udev -p $tempnode" +KERNEL=="md*", IMPORT{program}="/sbin/blkid -o udev -p $tempnode" ENV{ID_FS_TYPE}=="LVM2_member", \ TEST!="/tmp/.lvm_scan-%k", \ -- 1.6.2.2

2 4

[PATCH] Run firstaidkit-qs script instead of the shell (shows rescue menu) (#508512)
by Martin Sivak 14 Jul '09

14 Jul '09

--- rescue.py | 25 ++++++++++++++++++++++++- scripts/upd-instroot | 4 +++- 2 files changed, 27 insertions(+), 2 deletions(-) diff --git a/rescue.py b/rescue.py index 665c56c..ba8455c 100644 --- a/rescue.py +++ b/rescue.py @@ -34,6 +34,7 @@ import iutil import shutil import time import network +import subprocess import gettext _ = lambda x: gettext.ldgettext("anaconda", x) @@ -128,6 +129,22 @@ def makeMtab(instPath, fsset): finally: f.close() +def makeFStab(instPath = ""): + if os.access("/etc/mtab", os.R_OK): + f = open("/etc/mtab" %(instPath,), "r") + buf = f.read() + f.close() + else: + buf = "" + + try: + f = open(instPath + "/etc/fstab", "w+") + if buf: + f.write(buf) + f.close() + except IOError, e: + log.info("failed to write /etc/fstab: %s" % e) + # make sure they have a resolv.conf in the chroot def makeResolvConf(instPath): if not os.access("/etc/resolv.conf", os.R_OK): @@ -183,7 +200,10 @@ def runShell(screen = None, msg=""): "system will reboot.")) print - if os.path.exists("/bin/bash"): + if os.path.exists("/usr/bin/firstaidkit-qs") and os.path.exists("/usr/bin/dialog"): + proc = subprocess.Popen(["/bin/firstaidkit-qs"]) + proc.wait() + elif os.path.exists("/bin/bash"): iutil.execConsole() else: print(_("Unable to find /bin/sh to execute! Not starting shell")) @@ -436,6 +456,9 @@ def runRescue(anaconda, instClass): log.error("error making a resolv.conf: %s" %(e,)) msgStr = _("Your system is mounted under the %s directory.") % (anaconda.rootPath,) + #create /etc/fstab in ramdisk, so it is easier to work with RO mounted filesystems + makeFStab() + # run %post if we've mounted everything if anaconda.isKickstart: from kickstart import runPostScripts diff --git a/scripts/upd-instroot b/scripts/upd-instroot index 8a6ddca..31e5d89 100755 --- a/scripts/upd-instroot +++ b/scripts/upd-instroot @@ -235,7 +235,7 @@ PACKAGESRESCUE="bzip2 bzip2-libs dump libidn modutils mtools mt-st mtr net-tools open smartmontools openssh openssh-clients pciutils rsh traceroute tar rsync device-mapper device-mapper-libs dmraid ntfsprogs samba-client - firstaidkit-plugin-all" + firstaidkit firstaidkit-plugin-all dialog" if [ $ARCH = i386 -o $ARCH = i586 -o $ARCH = x86_64 ]; then PACKAGESRESCUE="$PACKAGESRESCUE gpart grub" @@ -815,6 +815,8 @@ usr/sbin/smartctl usr/sbin/traceroute usr/$LIBDIR/firstaidkit/plugins/ usr/lib/python?.?/site-packages/pyfirstaidkit/ +usr/bin/dialog +usr/bin/firstaidkit-qs EOF if [ $ARCH = i386 -o $ARCH = i586 -o $ARCH = x86_64 ]; then -- 1.5.4.3

3 5

breaking exn handling out of rhpl & anaconda
by Chris Lumens 13 Jul '09

13 Jul '09

I've spent the last few days trying to break the exception handling out of both rhpl and anaconda, combining them into their own project. The benefits for this are obvious: helping to get rid of rhpl, and letting tons of other programs make use of save-to-bugzilla. Anyway, my current work in progress is here: http://clumens.fedorapeople.org/exn/ Making this generic enough is a little annoying but I think I've managed it. The UI is probably the most complicated part. My reasoning for doing it like you see is so we can still support text and graphical interfaces, and so that anaconda can provide its own wacky UI classes that override what I've done. That's why the abstract interfaces look awful familiar. test.py is obviously not a part of this module, but it does show how to use the thing from a programmer's point of view. This is likely to get a little more complicated when I work on bugzilla support but that's just the nature of working with bugzilla. Things still left to do: - test save to bugzilla - it should "compile", but probably doesn't do anything - work up a patch for anaconda, rhpl, firstboot, system-config-*, etc. - do all the typical project stuff (translations, spec file, etc.) - add wait dialogs as well - there's a bug about this somewhere - write docs describing how to make this easier to use - make sure everything's robust - your pet requirement here Any questions? - Chris

1 0

ADP Anaconda Patches, 7/9/2009 Edition
by Alford, Seth 13 Jul '09

13 Jul '09

The ADP Anaconda Patches, 7/9/2009 Edition $Id: README /main/10 2009/07/10 00:12:59 alfords Exp $ Copyright (c) 2001-2009 Automatic Data Processing Inc. NOTICE: This is a README which describes a series of anaconda patches which ADP is redistributing. We are publishing this software to fulfill the requirements of the GPL. This software is offered with NO WARRANTY and NO SUPPORT of any kind. If you are reading this and you are not an ADP employee, see subsection 1.2, "Licensing, Intended Audience, and a Note for Non-ADP Readers", below. CHANGELOG: 6/30/2009 Completely rewrote this README from a previous version. I deleted a lot of extraneous information, including syslinux changes, and some suggestions for how to debug anaconda. The syslinux changes are described in a separate posting I made to the syslinux mailing list. Other features or debugging techniques are described in documentation from RedHat. I also included information specific to individual patches with those specific patches, rather than having all that documentation here. That makes this README much shorter. Since it's shorter, people might actually read it. Since this is a complete rewrite, I restarted the CHANGELOG. TABLE OF CONTENTS: 1. INTRODUCTION 1.1 What is this? 1.2 Licensing, Intended Audience, and a Note for Non-ADP Readers 1.3 What ADP Needed versus Anaconda's Capabilities 2. HELPFUL RESOURCES 2.1 What to Read 2.2 Preparing for Anaconda Development 1. INTRODUCTION 1.1 What is this? Some years ago, ADP modified the anaconda which came with RedHat 7.0, and, later, RedHat 8.0. To comply with the GPL, we redistributed changes that we made to those versions of anaconda. To explain what we did, we wrote a large README which we posted with the modifications. ADP has now modified the anaconda which came with CentOS 5.3. The newer version of anaconda is more powerful than the older versions. The newer anaconda is much closer to what ADP needs. But, we still had to modify anaconda itself to get it to do what we wanted. We're now sharing those modifications. That's so that: 1. We fulfill the requirements of the GPL 2. Our modifications might be picked up and included in the base code. That way we don't have to make the same modifications again, later. Also, rather than produce a single large README, I split up the information between this smaller README and each patch. 1.2 Licensing, Intended Audience, and a Note for Non-ADP Readers An intended audience for this README are developers at ADP, who might have to further modify anaconda in the future. Note that I said "might". ADP makes no promise that ADP will make any further modifications to this software. As I said above, this software is offered with NO WARRANTY and NO SUPPORT of any kind. A secondary intended audience for this README are people who are not at ADP. We could have written two versions of this README file: one for ADP readers, and one for non-ADP readers. But, we did not want to have to maintain internal and external versions of this README and/or the accompanying software. Hence these patches may contain some references to non-GPL'd software, or ADP's build environment, which are "independent and separate works" under section 2 of the GPL, and therefore not covered by the GPL. THEREFORE, WE ARE NOT PUBLISHING ANYTHING ELSE UNDER THE GPL, INCLUDING BUT NOT LIMITED TO: * files which do not appear in the patches, * ADP's proprietary backup and restore software, * our build environment, * other ADP proprietary software referenced or included by this README, including but not limited to DSDA. That means: * this README may contain some references which won't make sense to non-ADP readers. These include, but are not limited to, our build environment, and build process. We don't provide explanation for these references, either. * other parts of anaconda, as modified, may also be non-functional. * some readers may find themselves saying, "yes, I know about that already." Other readers may say, "I don't understand that reference, it must be an ADP-ism." 1.3 What ADP Needed versus Anaconda's Capabilities For ADP's purposes, anaconda, even with a customized kickstart file, lacked several features, including: * simplified messages for the user * completely automatic partitioning on a variety of disk configurations, and support for some specific-to-ADP disk partitioning requirements * support for tape restore * --userinput flag to pre and post install scripts * -s flag passed to pre and post install scripts when running on a serial console * support for a restore install type 2. HELPFUL RESOURCES 2.1 What to Read Here is an incomplete list of resources that I think will help you in learning how to modify anaconda and kickstart: * Beazley's _Python Essential Reference_. Remember, anaconda is written in Python. Check which version of Python comes with your version of anaconda. Make sure that your version of Beazley addresses a matching version of Python. Otherwise, you may find yourself trying to take advantage of a new Python feature that doesn't exist in your environment. Or, you might find you are puzzled by a Python feature in one of the anaconda sources that Beazley does not explain. * The RedHat Enterprise Linux Installation Guide, available at http://www.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/Installati… * The kickstart-list and anaconda-dev mailing lists. You can join at www.redhat.com. The list archives are located there, too. * Most importantly, read Chris Lumens' pykickstart documentation. You should be able to get this by installing the pykickstart rpm then looking in /usr/share/doc/pykickstart*/programmers-guide . 2.2 Preparing for Anaconda Development Install at least the python, python-devel, pykickstart, anaconda, and anaconda-runtime packages and their dependencies on your Linux development system. This message and any attachments are intended only for the use of the addressee and may contain information that is privileged and confidential. If the reader of the message is not the intended recipient or an authorized representative of the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by e-mail and delete the message and any attachments from your system.

2 15

[PATCH][RHEL5.4] update support for ext4 in anaconda
by Jeffrey Bastian 13 Jul '09

13 Jul '09

ext4 support is being updated in RHEL 5.4 from an early development release to the stable code from Fedora 10. See https://bugzilla.redhat.com/show_bug.cgi?id=485315 As a result, the kernel module has changed names from ext4dev.ko to just ext4.ko, and anaconda no longer lists ext4/ext4dev as an available filesystem (if the 'ext4' command line option is used). See https://bugzilla.redhat.com/show_bug.cgi?id=510634 The attached patch updates anaconda to use the new module name. More work is probably needed to drop ext4dev. Jeff

2 1

[PATCH] Support for bootloader on md raid devices.
by Radek Vykydal 13 Jul '09

13 Jul '09

In f10, for /boot on md raid 1 we used to have only option to install bootloader on partition but in fact, we installed on MBR. In f11 we added option to instal to MBR in UI, and we are still installing on MBR for both options. In rhel 5 we are always installing bootloader on MBR too. Also grub-install /dev/md0 installs on MBR, so maybe this is not something we want to do at all, but in that case, we should probably remove the option to install bootloader on /dev/md0 raid device from UI. Two issues: * The patch assumes that 'mdX' bootloader target is md raid 1 device with members being partitions. Intel bios raid using mdraid (or whole-disk md devices in general?) need to be handled/considered. * If we allow (start to do) installing on md raid partitions, upgrade should be patched to count with this possibility. A little problem is that up to f11 boot=/dev/md0 we write into /etc/sysconfig/grub or grub.conf means that bootloader was installed on mbr, and after patching it would mean that it is on raid device (partitions). --- booty/x86.py | 17 ++++------------- 1 files changed, 4 insertions(+), 13 deletions(-) diff --git a/booty/x86.py b/booty/x86.py index 3890e69..8527442 100644 --- a/booty/x86.py +++ b/booty/x86.py @@ -112,14 +112,13 @@ class x86BootloaderInfo(efiBootloaderInfo): for bootDev in bootDevs: cmds = [] - gtPart = self.getMatchingPart(bootDev, grubTarget) - gtDisk = self.grubbyPartitionName(getDiskPart(gtPart, self.storage)[0]) + gtDisk = self.grubbyPartitionName(getDiskPart(bootDev, self.storage)[0]) bPart = self.grubbyPartitionName(bootDev) cmd = "root %s\n" % (bPart,) stage1Target = gtDisk if target == "partition": - stage1Target = self.grubbyPartitionName(gtPart) + stage1Target = bPart cmd += "install %s%s/stage1 d %s %s/stage2 p %s%s/grub.conf" % \ (args, grubPath, stage1Target, grubPath, bPart, grubPath) @@ -155,7 +154,8 @@ class x86BootloaderInfo(efiBootloaderInfo): if grubTarget[-2] == 'p' or \ (grubTarget[-2].isdigit() and grubTarget[-3] == 'p'): target = "partition" - elif grubTarget[-1].isdigit() and not path.startswith('md'): + # NOTE: this includes also mdX grub targets + elif grubTarget[-1].isdigit(): target = "partition" f = open(cf, "w+") @@ -362,15 +362,6 @@ class x86BootloaderInfo(efiBootloaderInfo): return 0 - def getMatchingPart(self, bootDev, target): - bootName, bootPartNum = getDiskPart(bootDev, self.storage) - devices = self.getPhysicalDevices(target) - for device in devices: - name, partNum = getDiskPart(device, self.storage) - if name == bootName: - return device - return devices[0] - def grubbyDiskName(self, name): return "hd%d" % self.drivelist.index(name) -- 1.6.0.6

2 3

[PATCH] Use tail process to ensure line buffered output to screen and log file at the same moment... (#506664)
by Martin Sivak 13 Jul '09

13 Jul '09

--- iutil.py | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++------------ 1 files changed, 68 insertions(+), 16 deletions(-) diff --git a/iutil.py b/iutil.py index ba1d0b4..0a4adfa 100644 --- a/iutil.py +++ b/iutil.py @@ -33,6 +33,7 @@ from flags import flags from constants import * from rhpl.translate import _ import re +import select import logging log = logging.getLogger("anaconda") @@ -81,37 +82,88 @@ def execWithRedirect(command, argv, stdin = None, stdout = None, runningLog = open("/tmp/program.log", "a") runningLog.write("Running... %s\n" % ([command] + argv,)) + + #prepare os pipes for feeding tee proceses + pstdout, pstdin = os.pipe() + perrout, perrin = os.pipe() env = os.environ.copy() env.update({"LC_ALL": "C"}) + #prepare "empty" tee proceses so we can properly detect errors + proc_std = None + proc_err = None + try: - proc = subprocess.Popen([command] + argv, stdin=stdin, - stdout=subprocess.PIPE, + #create output pipes with logging (it needs to be different process to avoid deadlock and EOF detection..) + proc_std = subprocess.Popen(["tee", "-a", "/tmp/program.log"], stdin=pstdout, + stdout=stdout, stderr=subprocess.PIPE, + cwd=root, + env=env, bufsize = 1) + + proc_err = subprocess.Popen(["tee", "-a", "/tmp/program.log"], stdin=perrout, + stdout=stderr, + stderr=subprocess.PIPE, + cwd=root, + env=env, bufsize = 1) + + #run the command + proc = subprocess.Popen([command] + argv, stdin=stdin, + stdout=pstdin, + stderr=perrin, preexec_fn=chroot, cwd=root, - env=env) + env=env, bufsize = 1) + + proc.wait() + + #close the output pipes + if proc_std: + proc_std.terminate() + proc_std.wait() + proc_std = None + + if proc_err: + proc_err.terminate() + proc_err.wait() + proc_err = None + + ret = proc.returncode - while True: - (outStr, errStr) = proc.communicate() - if outStr: - os.write(stdout, outStr) - runningLog.write(outStr) - if errStr: - os.write(stderr, errStr) - runningLog.write(errStr) - if proc.returncode is not None: - ret = proc.returncode - break except OSError as e: errstr = "Error running %s: %s" % (command, e.strerror) log.error(errstr) + + #free descriptors + os.close(pstdout) + os.close(perrout) + os.close(pstdin) + os.close(perrin) + + #close the output pipes + if proc_std: + proc_std.terminate() + proc_std.wait() + proc_std = None + + if proc_err: + proc_err.terminate() + proc_err.wait() + proc_err = None + runningLog.write(errstr) runningLog.close() raise RuntimeError, errstr runningLog.close() + + #free descriptors + os.close(pstdout) + os.close(perrout) + os.close(pstdin) + os.close(perrin) + return ret ## Run an external program and capture standard out. -- 1.5.4.3

1 1

[PATCH] rhel5 include graphical libs.
by Joel Granados Moreno 13 Jul '09

13 Jul '09

Hello List The libdrm is needed to make ATI Technologies Inc M24 work on rhel5. make sure we include the sofile in the image. Review appreciated. Regards. Joel Andres Granados Brno, Czech Republic, Red Hat.

2 2

[master] expand info in initialize disk dialog (#460697)
by David Cantrell 11 Jul '09

11 Jul '09

Originally filed for RHEL-5. I posted a patch earlier and Steffen Maier commented on it saying it would be useful to have more information for the drive in question, especially if you have a large number of the same type in the system. The dialog box could probably be improved, but I'm mostly just going for functionality with this patch. In addition to displaying the device model string and the size in megabytes (as originally requested in the BZ), I added a device details string if we can find it in /dev/disk/by-path on the system. Example output on my system: Error processing drive sda 305243MB (ATA ST3320620AS). Maybe it needs to be reinitialized. YOU WILL LOSE ALL DATA ON THIS DRIVE IN YOUR REINITIALIZE IT! Device details: pci-0000:00:1f.2-scsi-0:0:0:0 For DASD and zFCP, you'll get a slightly different clown barf string for the device details. I think the dialog wording could be improved, but since Peter and Andy are working on error string cleanups, I would like to leave that to them. -- David Cantrell <dcantrell(a)redhat.com> Red Hat / Honolulu, HI

2 3

[PATCH] Don't write removable devices to /etc/fstab (#505697).
by Chris Lumens 10 Jul '09

10 Jul '09

--- storage/__init__.py | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/storage/__init__.py b/storage/__init__.py index 03b9876..7996d00 100644 --- a/storage/__init__.py +++ b/storage/__init__.py @@ -1851,6 +1851,10 @@ class FSSet(object): if not device.format.mountable and device.format.type != "swap": continue + # Don't write removable devices, either. Let the desktop do it. + if device.removable: + continue + fstype = getattr(device.format, "mountType", device.format.type) if fstype == "swap": mountpoint = "swap" -- 1.6.1.3

4 4

[master] include comps.dtd (patch from RHEL 5.4)
by David Cantrell 10 Jul '09

10 Jul '09

This is another 5.4 patch of mine, updated for master. -- David Cantrell <dcantrell(a)redhat.com> Red Hat / Honolulu, HI

3 8

[PATCH] Clean yum caches following preupgrade, too (#503096).
by Chris Lumens 10 Jul '09

10 Jul '09

--- yuminstall.py | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/yuminstall.py b/yuminstall.py index 0491fad..31ed4ea 100644 --- a/yuminstall.py +++ b/yuminstall.py @@ -1584,7 +1584,7 @@ reposdir=/etc/anaconda.repos.d,/tmp/updates/anaconda.repos.d,/tmp/product/anacon repo.dirCleanup() # expire yum caches on upgrade - if anaconda.id.getUpgrade() and os.path.exists("%s/var/cache/yum" %(anaconda.rootPath,)): + if (flags.cmdline.has_key("preupgrade") or anaconda.id.getUpgrade()) and os.path.exists("%s/var/cache/yum" %(anaconda.rootPath,)): log.info("Expiring yum caches") try: iutil.execWithRedirect("yum", ["clean", "all"], @@ -1595,7 +1595,7 @@ reposdir=/etc/anaconda.repos.d,/tmp/updates/anaconda.repos.d,/tmp/product/anacon pass # nuke preupgrade - if flags.cmdline.has_key("preupgrade") and anaconda.id.getUpgrade() and os.path.exists("%s/var/cache/yum/anaconda-upgrade" %(anaconda.rootPath,)): + if flags.cmdline.has_key("preupgrade") and os.path.exists("%s/var/cache/yum/anaconda-upgrade" %(anaconda.rootPath,)): try: shutil.rmtree("%s/var/cache/yum/anaconda-upgrade" %(anaconda.rootPath,)) except: -- 1.6.1.3

2 1

[PATCH] All other teardown methods take a "recursive" argument (#506166).
by Chris Lumens 10 Jul '09

10 Jul '09

--- storage/devices.py | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/storage/devices.py b/storage/devices.py index b46479b..e236710 100644 --- a/storage/devices.py +++ b/storage/devices.py @@ -2835,7 +2835,7 @@ class FileDevice(StorageDevice): for parent in self.parents: parent.format.setup() - def teardown(self): + def teardown(self, recursive=None): StorageDevice.teardown(self) if self.format and self.format.exists and not self.format.status: self.format.device = self.path -- 1.6.1.3

2 1

[PATCH] You must be this tall to ride.
by Bill Nottingham 10 Jul '09

10 Jul '09

Adjust memory requirements. Signed-off-by: Bill Nottingham <notting(a)redhat.com> --- isys/isys.h | 10 +++------- 1 files changed, 3 insertions(+), 7 deletions(-) diff --git a/isys/isys.h b/isys/isys.h index 43355ca..15e461e 100644 --- a/isys/isys.h +++ b/isys/isys.h @@ -20,13 +20,9 @@ #ifndef H_ISYS #define H_ISYS -#define MIN_RAM 64000 -#define MIN_GUI_RAM 192000 -#if defined(__x86_64__) || defined(__ia64__) || defined(__s390x__) || defined(__ppc64__) -#define EARLY_SWAP_RAM 400000 -#else -#define EARLY_SWAP_RAM 270000 -#endif +#define MIN_RAM 131072 +#define MIN_GUI_RAM 393216 +#define EARLY_SWAP_RAM 524288 #define OUTPUT_TERMINAL "/dev/tty5" -- 1.6.2.5

2 1

Installing livecd from USB stick - Bug #502771 - Investigation and patch
by Eric Kerin 09 Jul '09

09 Jul '09

I put this in bugzilla in the closed bug - it was marked as a duplicate of another, it was suggested that I post it here. - Not subscribed please CC. Quick summary - mke2fs will finish it's work quickly, and anaconda will spend quite sometime reading in the output, and updating the progress dialog in certain cases - the attached patch adds a check to the read/progress loop to check to see if the process has finished, and jumps out early if that's the case. Long story - I'm working on a netbook livecd install for Fedora. With the screen size of the machine I have to test on, anaconda goes into full-screen mode, the progress dialog goes full-screen as well (and remains hidden too, have not checked into this yet), it looks like because of this, it takes a long time for the progress pulses to return. (FYI - my machine also happens to be running maximus, but anaconda is excluded from it's maximizing effects) I ran into the same problem as others have, where you can hit next early, and applied the patch to make the progress dialog modal, but still ran into long periods of time where it was "formatting" with no disk IO. A quick check, and mke2fs was defunct, and anaconda was using up CPU nicely. I did some digging, and found that it was still reading the stdout data from mke2fs (it was printed very slowly, one character at a time to /dev/tty5) What I found, was that since iutil.py's execWithPulseProgress method reads one character at a time, in cases where the progress.pulse() method takes a long time to return, it slows up the reading of data, and the eventual reaping of the process. Which manifests as a defunct mke2fs process. If you give it long enough, it will continue the install. Long enough being 10 minutes or so, depending on your system. There's a loop in gui.py that smooths the pulses if it took a long time between calls, that's where a lot of the delay is spent, doing processEvents(), and because of the loop and smoothing it can start to take even longer - the first call to pulse that waits is is for <1s, next one smoothing a delta of 5 seconds, the next call is for 20 s delta, then 80 s delta...Eventually, it reads every character, after this exponentially increasing wait state. The attached patch does two things, only the change to iutil.py is really necessary - it just checks to see if the child process has exited already, and breaks if it is, since we don't need the output of the child process anyway. The patch for gui.py just keeps the delay from getting exponentially longer. Thanks, Eric Kerin

1 0

[PATCH] mdmon missing in install.img
by Jacek Danecki 09 Jul '09

09 Jul '09

mdmon added to install.img mdmon application is needed to switch imsm md array to read/write mode. diff --git a/scripts/upd-instroot b/scripts/upd-instroot index 1ee3084..a560733 100755 --- a/scripts/upd-instroot +++ b/scripts/upd-instroot @@ -405,6 +405,7 @@ sbin/ldconfig sbin/lspci sbin/lvm* sbin/mdadm +sbin/mdmon sbin/mkdosfs sbin/mke2fs sbin/mkfs.btrfs

2 1

[PATCH] rhel5 include graphical libs.
by Joel Granados Moreno 09 Jul '09

09 Jul '09

Hello List The libdrm is needed to make ATI Technologies Inc M24 work on rhel5. Review appreciated. Regards. Joel Andres Granados Brno, Czech Republic, Red Hat.

4 4

PATCH [grub]: add support for partitionable md
by Hans de Goede 09 Jul '09

09 Jul '09

2 1

[PATCH 1/2] Write out NM_CONTROLLED=no for NICs used for FCoE
by Hans de Goede 09 Jul '09

09 Jul '09

Write out NM_CONTROLLED=no for NICs used for FCoE, note that unlike with iSCSI we do not blindly write out NM_CONTROLLED=no for all NICs, but just for the NIC which is used for FCoE. The iSCSI behaviour is undesirable, but the whole writing of NM_CONTROLLED=no for iSCSI will go away as soon as NetworkManager is fixed to not down devices when it takes over control, which should be fixed soon. --- network.py | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/network.py b/network.py index e3ce2c1..b4a8c89 100644 --- a/network.py +++ b/network.py @@ -587,8 +587,9 @@ class Network: # FIXME: use device.host_address to only add "NM_CONTROLLED=no" # for interfaces actually used enroute to the device for d in anaconda.id.storage.devices: - if rootdev.dependsOn(d) and isinstance(d, - storage.devices.NetworkStorageDevice): + if rootdev.dependsOn(d) and \ + isinstance(d, storage.devices.NetworkStorageDevice) and\ + (not d.nic or d.nic == device): f.write("NM_CONTROLLED=no\n") break -- 1.6.2.2

3 7

[PATCH 1/4] 70-anaconda.rules: always import blkid output
by Hans de Goede 09 Jul '09

09 Jul '09

We are currently not calling blkid when a block device is not an mdraid set or a device mapper device, resulting in us not identifying mdraid members as such. This probably worked when still using vol_id because some other rules file was doing an import vol_id before our rules. --- 70-anaconda.rules | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/70-anaconda.rules b/70-anaconda.rules index 0efd536..0fd7647 100644 --- a/70-anaconda.rules +++ b/70-anaconda.rules @@ -4,6 +4,8 @@ SUBSYSTEM!="block", GOTO="anaconda_end" ENV{ANACBIN}="/sbin" TEST!="$env{ANACBIN}/dmsetup", ENV{ANACBIN}="/usr/sbin" +IMPORT{program}="$env{ANACBIN}/blkid -o udev -p $tempnode" + KERNEL!="dm-*", GOTO="anaconda_mdraid" IMPORT{program}="$env{ANACBIN}/dmsetup info -c --nameprefixes --unquoted --rows --noheadings -o name,uuid,suspended,readonly,major,minor,open,tables_loaded -j%M -m%m" @@ -14,7 +16,6 @@ ENV{DM_UUID}=="?*", SYMLINK+="disk/by-id/dm-uuid-$env{DM_UUID}" ENV{DM_STATE}=="SUSPENDED", GOTO="anaconda_end" -IMPORT{program}="blkid -o udev -p $tempnode" OPTIONS="link_priority=-100" ENV{ID_FS_USAGE}=="filesystem|other|crypto", ENV{ID_FS_UUID_ENC}=="?*", SYMLINK+="disk/by-uuid/$env{ID_FS_UUID_ENC}" @@ -37,7 +38,6 @@ ENV{DEVTYPE}=="partition", ENV{MD_UUID}=="?*", SYMLINK+="disk/by-id/md-uuid-$env ENV{DEVTYPE}=="partition", ENV{MD_DEVNAME}=="*[^0-9]", SYMLINK+="md/$env{MD_DEVNAME}%n" ENV{DEVTYPE}=="partition", ENV{MD_DEVNAME}=="*[0-9]", SYMLINK+="md/$env{MD_DEVNAME}p%n" -IMPORT{program}="blkid -o udev -p $tempnode" OPTIONS+="link_priority=100" OPTIONS+="watch" ENV{ID_FS_USAGE}=="filesystem|other|crypto", ENV{ID_FS_UUID_ENC}=="?*", SYMLINK+="disk/by-uuid/$env{ID_FS_UUID_ENC}" -- 1.6.2.2

2 8

[PATCH] Remove _catchallCategory. yum handles this for us now.
by Chris Lumens 09 Jul '09

09 Jul '09

--- yuminstall.py | 25 ------------------------- 1 files changed, 0 insertions(+), 25 deletions(-) diff --git a/yuminstall.py b/yuminstall.py index d1cb626..623911d 100644 --- a/yuminstall.py +++ b/yuminstall.py @@ -1097,8 +1097,6 @@ reposdir=/etc/anaconda.repos.d,/tmp/updates/anaconda.repos.d,/tmp/product/anacon self.ayum._setGroups(None) continue - self._catchallCategory() - def doRepoSetup(self, anaconda, thisrepo = None, fatalerrors = True): self.__withFuncDo(anaconda, lambda r: self.ayum.doRepoSetup(thisrepo=r.id), thisrepo=thisrepo, fatalerrors=fatalerrors) @@ -1186,29 +1184,6 @@ reposdir=/etc/anaconda.repos.d,/tmp/updates/anaconda.repos.d,/tmp/product/anacon self.ayum.repos.callback = None - def _catchallCategory(self): - # FIXME: this is a bad hack, but catch groups which aren't in - # a category yet are supposed to be user-visible somehow. - # conceivably should be handled by yum - grps = {} - for g in self.ayum.comps.groups: - if g.user_visible: - grps[g.groupid] = g - - for cat in self.ayum.comps.categories: - for g in cat.groups: - if grps.has_key(g): - del grps[g] - - if len(grps.keys()) == 0: - log.info("no groups missing") - return - c = yum.comps.Category() - c.name = _("Uncategorized") - c._groups = grps - c.categoryid = "uncategorized" - self.ayum.comps._categories[c.categoryid] = c - def getDefaultGroups(self, anaconda): langs = anaconda.id.instLanguage.getCurrentLangSearchList() rc = map(lambda x: x.groupid, -- 1.6.1.3

2 1

[PATCH] Use a method yum provides, rather than inventing our own.
by Chris Lumens 08 Jul '09

08 Jul '09

--- yuminstall.py | 17 ++++++----------- 1 files changed, 6 insertions(+), 11 deletions(-) diff --git a/yuminstall.py b/yuminstall.py index 623911d..478b22b 100644 --- a/yuminstall.py +++ b/yuminstall.py @@ -743,17 +743,12 @@ class AnacondaYum(YumSorter): downloadpkgs = [] totalSize = 0 totalFiles = 0 - for txmbr in self.tsInfo.getMembers(): - if txmbr.ts_state in ['i', 'u']: - po = txmbr.po - else: - continue - - if po: - totalSize += int(po.returnSimple("installedsize")) / 1024 - for filetype in po.returnFileTypes(): - totalFiles += len(po.returnFileEntries(ftype=filetype)) - downloadpkgs.append(po) + for txmbr in self.tsInfo.getMembersWithState(output_states=TS_INSTALL_STATES): + if txmbr.po: + totalSize += int(txmbr.po.returnSimple("installedsize")) / 1024 + for filetype in txmbr.po.returnFileTypes(): + totalFiles += len(txmbr.po.returnFileEntries(ftype=filetype)) + downloadpkgs.append(txmbr.po) return (downloadpkgs, totalSize, totalFiles) -- 1.6.1.3

2 1

[PATCH 1/2] pykickstart: Add fcoe command (take 2)
by Hans de Goede 08 Jul '09

08 Jul '09

Add an command to activate FCoE attached SAN's from kickstart --- pykickstart/commands/__init__.py | 4 +- pykickstart/commands/fcoe.py | 73 ++++++++++++++++++++++++++++++++++++++ pykickstart/handlers/control.py | 2 + 3 files changed, 77 insertions(+), 2 deletions(-) create mode 100644 pykickstart/commands/fcoe.py diff --git a/pykickstart/commands/__init__.py b/pykickstart/commands/__init__.py index 7ce21f7..0952bf8 100644 --- a/pykickstart/commands/__init__.py +++ b/pykickstart/commands/__init__.py @@ -18,8 +18,8 @@ # with the express permission of Red Hat, Inc. # import authconfig, autopart, autostep, bootloader, clearpart, device -import deviceprobe, displaymode, dmraid, driverdisk, firewall, firstboot, group -import ignoredisk, interactive, iscsi, iscsiname, key, keyboard, lang +import deviceprobe, displaymode, dmraid, driverdisk, fcoe, firewall, firstboot +import group, ignoredisk, interactive, iscsi, iscsiname, key, keyboard, lang import langsupport, lilocheck, logging, logvol, mediacheck, method, monitor import mouse, multipath, network, partition, raid, reboot, repo, rescue, rootpw import selinux, services, skipx, timezone, updates, upgrade, user, vnc diff --git a/pykickstart/commands/fcoe.py b/pykickstart/commands/fcoe.py new file mode 100644 index 0000000..e3ffce8 --- /dev/null +++ b/pykickstart/commands/fcoe.py @@ -0,0 +1,73 @@ +# +# Hans de Goede <hdegoede(a)redhat.com> +# +# Copyright 2009 Red Hat, Inc. +# +# This copyrighted material is made available to anyone wishing to use, modify, +# copy, or redistribute it subject to the terms and conditions of the GNU +# General Public License v.2. This program is distributed in the hope that it +# will be useful, but WITHOUT ANY WARRANTY expressed or implied, including the +# implied warranties of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +# See the GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License along with +# this program; if not, write to the Free Software Foundation, Inc., 51 +# Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. Any Red Hat +# trademarks that are incorporated in the source code or documentation are not +# subject to the GNU General Public License and may only be used or replicated +# with the express permission of Red Hat, Inc. +# +from pykickstart.base import * +from pykickstart.options import * + +class F12_FcoeData(BaseData): + removedKeywords = BaseData.removedKeywords + removedAttrs = BaseData.removedAttrs + + def __init__(self, *args, **kwargs): + BaseData.__init__(self, *args, **kwargs) + self.nic = kwargs.get("nic", None) + + def __str__(self): + retval = BaseData.__str__(self) + retval += "fcoe" + + if self.nic: + retval += " --nic=%s" % self.nic + + return retval + "\n" + + +class F12_Fcoe(KickstartCommand): + removedKeywords = KickstartCommand.removedKeywords + removedAttrs = KickstartCommand.removedAttrs + + def __init__(self, writePriority=71, *args, **kwargs): + KickstartCommand.__init__(self, writePriority, *args, **kwargs) + self.op = self._getParser() + self.fcoe = kwargs.get("fcoe", []) + + def __str__(self): + retval = "" + for fcoe in self.fcoe: + retval += fcoe.__str__() + + return retval + + def _getParser(self): + op = KSOptionParser() + op.add_option("--nic", dest="nic", required=1) + return op + + def parse(self, args): + zd = self.handler.FcoeData() + (opts, extra) = self.op.parse_args(args=args, lineno=self.lineno) + if len(extra) > 0: + mapping = {"command": "fcoe", "options": extra} + raise KickstartValueError, formatErrorMsg(self.lineno, msg=_("Unexpected arguments to %(command)s command: %(options)s") % mapping) + + self._setToObj(self.op, opts, zd) + return zd + + def dataList(self): + return self.fcoe diff --git a/pykickstart/handlers/control.py b/pykickstart/handlers/control.py index 8e1ab66..8f78c3e 100644 --- a/pykickstart/handlers/control.py +++ b/pykickstart/handlers/control.py @@ -541,6 +541,7 @@ commandMap = { "deviceprobe": deviceprobe.FC3_DeviceProbe, "dmraid": dmraid.FC6_DmRaid, "driverdisk": driverdisk.FC3_DriverDisk, + "fcoe": fcoe.F12_Fcoe, "firewall": firewall.F10_Firewall, "firstboot": firstboot.FC3_Firstboot, "graphical": displaymode.FC3_DisplayMode, @@ -873,6 +874,7 @@ dataMap = { "DriverDiskData": driverdisk.F12_DriverDiskData, "DeviceData": device.F8_DeviceData, "DmRaidData": dmraid.FC6_DmRaidData, + "FcoeData": fcoe.F12_FcoeData, "GroupData": group.F12_GroupData, "IscsiData": iscsi.F10_IscsiData, "LogVolData": logvol.F9_LogVolData, -- 1.6.2.2

3 3

[PATCH] Use the yum preconf object to do $releasever substitution.
by Chris Lumens 08 Jul '09

08 Jul '09

--- yuminstall.py | 16 ++++++---------- 1 files changed, 6 insertions(+), 10 deletions(-) diff --git a/yuminstall.py b/yuminstall.py index 5fdca80..5170e94 100644 --- a/yuminstall.py +++ b/yuminstall.py @@ -555,11 +555,11 @@ class AnacondaYum(YumSorter): # We need to make sure $releasever gets set up before .repo files are # read. Since there's no redhat-release package in /mnt/sysimage (and # won't be for quite a while), we need to do our own substutition. - def getReposFromConfig(self): - def _getReleasever(): - from ConfigParser import ConfigParser - c = ConfigParser() + def _getReleasever(): + from ConfigParser import ConfigParser + c = ConfigParser() + try: if os.access("%s/.treeinfo" % self.anaconda.methodstr, os.R_OK): ConfigParser.read(c, "%s/.treeinfo" % self.anaconda.methodstr) else: @@ -569,13 +569,8 @@ class AnacondaYum(YumSorter): ConfigParser.read(c, "/tmp/.treeinfo") return c.get("general", "version") - - try: - self.yumvar["releasever"] = _getReleasever() except: - self.yumvar["releasever"] = productVersion - - YumSorter.getReposFromConfig(self) + return productVersion # Override this method so yum doesn't nuke our existing logging config. def doLoggingSetup(self, *args, **kwargs): @@ -611,6 +606,7 @@ class AnacondaYum(YumSorter): if hasattr(self, "preconf"): self.preconf.fn = fn self.preconf.root = root + self.preconf.releasever = self._getReleasever() self.preconf.enabled_plugins = ["whiteout", "blacklist"] YumSorter._getConfig(self) else: -- 1.6.1.3

2 1

More arch fixes.
by Bill Nottingham 08 Jul '09

08 Jul '09

iutil.py | 16 +++++++++++++--- storage/devicelibs/lvm.py | 2 +- 2 files changed, 14 insertions(+), 4 deletions(-)

2 4

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel July 2009