Re: [Fedora-livecd-list] [PATCH] Implement %post --erroronfail
On Sat, Dec 19, 2009 at 3:27 PM, Daniel Drake d...@laptop.org wrote: I was a bit surprised to find that this flag is documented and parsed, but not acted upon. Am I missing anything? No one ever asked for it :-) Patch looks correct enough, although if we're going to check, it's probably worth at least logging a warning on scriptlet errors if they haven't asked for fail on error. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Re: Unnecessary SELinux Failure Condition?
On Wed, Dec 16, 2009 at 12:03 AM, Alan Pevec ape...@gmail.com wrote: I've just found one issue with the patch: @@ -734,6 +725,9 @@ class ImageCreator(object): self.__run_post_scripts() + # selinux should always come last + kickstart.SelinuxConfig(self._instroot).apply(ksh.selinux) + def launch_shell(self): Launch a shell in the install root. NAK. Nothing *EVER* can come after running %post scripts. It breaks one of the very few things that have to always be held true for kickstart scripts that came about after a few years of flip-flopping things around for various reasons that seemed good at the time. Other issue is that pykickstart selinux command doesn't have --type option to specify the policy and lokkit. assumes targeted if --selinuxtype=type is not specified. SelinuxConfig should not touch SELINUXTYPE already set in /etc/selinux/config by %post script. To avoid lokkit side-effects, I'll amend the patch to use Augeas to modify /etc/selinux/config I'd rather not have a dependency on augeas here. If it's important that people be able to specify the policy type in the kickstart config, then we should push that support in via pykickstart and lokkit, not change to a whole new toolset. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] lzma + squashfs should be easy to use in livecd-tools
On Tue, Dec 8, 2009 at 8:11 PM, Bruno Wolff III br...@wolff.to wrote: Phillip Lougher posted lmza+squashfs patches to lkml yesterday and today as a followup he posted a message indicating how to build / use the new sqaushfs-tools to build or read squashfs file systems. (http://marc.info/?l=linux-embeddedm=126029628723357w=2) [snip] If it is desired to have a single version of livecd-creator that works regardless of whether or not mksquashfs supports lzma, it is possible to run mksquashfs without any arguments and examine the output to see if lzma is supported. It may also be necessary to check for a minimum kernel being used for livecd image if livecd-creator is supposed to be able to compose older versions of Fedora. While one version is better, we continue to end up being a little hosed by squashfs changes here so I probably wouldn't end up being too concerned by it :/ Checking kernel versions is kind of crummy as an answer given the propensity towards things like patching kernels, etc. The biggest question as far as whether or not it's feasible for F13 in my mind really is whether or not the code gets accepted upstream. The progress of squashfs patches into the kernel has historically not been very fast there. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] launch anaconda kickstart from running livecd
On Fri, Oct 16, 2009 at 10:49 AM, Patrice Guay patrice.g...@nanotechnologies.qc.ca wrote: Peter Scheie wrote: Thanks for the clarification. Is the 'liveinst' utility needed in order to install the image from the livecd to the hard drive? Or is it needed because anaconda isn't built to start an installation from an already running system (such as a running livecd session)? I assume it is the former, as the webpage says as much, but I wanted verify that I wasn't misunderstanding the requirements. (This *does* start drifting toward being an anaconda question, but it's all within the context of the livecd session.) This is definitely a question related to anaconda. The liveinst utility was added to anaconda in order to support installation from a LiveCD. You should ask the package maintainers what are the limitations of their utility and its normal usage. pedantic liveinst is purely a shell script to kick off anaconda with the various arguments for install off of a live system mode /pedantic Installing a live image using anaconda requires the anaconda changes that went into Fedora 8 to do so -- you're not going to be able to get away with easily getting those running on a CentOS 5 system. You could hack up something to do the basic bits, but any of the, eg, help with partitioning you get or otherwise isn't going to be there. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] How to Pass Parameters to isolinux.cfg for console redirect
On Sat, Oct 10, 2009 at 12:03 AM, Jon Jaroker fedoraliv...@jaroker.com wrote: I would like to add “console=ttyS0” to the append line in the isolinux.cfg file in order to capture boot messages when testing liveCDs. I notice that the live.py script includes an $(extra) parameter. Is it possible to add “console=ttyS0” to this $(extra) parameter? If so, where should it be set? Is there a better way to pass such parameters directly through the kickstart file, such as with the “bootloader” parameter? That's exactly how -- as the kickstart docs say bootloader --append=console=ttyS0 should do the trick with livecds and a real system install - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] RFC: Stop creating /etc/modprobe.conf
On Thu, Oct 1, 2009 at 2:12 PM, Warren Togami wtog...@redhat.com wrote: mclasen noticed this is one of the last things creating modprobe.conf, which in turn causes many annoying deprecation warnings. I'm committing this to livecd git and tagging a new build for F-12 later today unless there are any objections. That looks fine. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Buyer Beware: A Major Change in NFS is about to happen
On Tue, Sep 29, 2009 at 8:15 PM, Steve Dickson ste...@redhat.com wrote: My main concern is with installer, installing from NFS shares from older servers, say RHEL5. How will anaconda handle mounting? Will there be odd errors that are difficult to figure out? Has this been tested in the anaconda environment at all? This issue is this... when the the F12 does a mount to a linux server and that linux server is *not* configured with a / *(ro,fsid=0) export, the mount will fail with ENOENT (or No such file or directory). If the server does have that export, things will work as expected... So my advice is to added that one line to your rhel5 server and every thing should as expected... or may even better... ;-) Another workaround is to added the '-o v3' mount options... would that be hard? Why not just see the error and fall back and try v3 programatically rather than forcing that upon unsuspecting users? If someone explicitly specifies v4, then sure, if that fails, it should fail. But if they don't, we should be forgiving in what we do rather than giving cryptic error messages. How will anaconda handle mounting? I don't know since I believe anaconda has its own mounting code so it does its own thing... and in the end that will mostly likely work, since that code didn't change... anaconda hasn't been using its own mounting code for several releases now. The change to force v3 could be done, but that then makes things different from the real system which is something that's been trying to be avoided. - Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: [Fedora-livecd-list] Re: Unnecessary SELinux Failure Condition?
On Fri, Sep 18, 2009 at 10:26 AM, Jay Greguske jgreg...@redhat.com wrote: Daniel J Walsh wrote: On 09/11/2009 04:47 PM, Jay Greguske wrote: While using livecd-creator and poking around the code, I found a check that I don't understand the reason for. livecd-creator will bail out if the host has SELinux disabled and the kickstart file requests it be enabled. Why is that? I would think that if SELinux was disabled but you still had the policy available, that would be all you need to build a properly labeled image. [snip] Perhaps the failure condition is no longer necessary? Yes I think that is no longer necessary. And it should definitely be supported. Attached is a cleaner patch that removes the check and some other unnecessary code (thanks Dan). With this users should be able to build livecd images that have SELinux enabled on an SELinux-disabled host. I've tested this on an F10 system with an F10 and a RHEL 5 kernel. Both kernels I was able to build images with the SELinux enabled and disabled on the host (but always enabled in the kickstart file). There have been some problems more recently with the booleans stuff if SELinux isn't enabled. Does that all end up working correctly still? I'm not fundamentally opposed to the patch; it's just historically been something which didn't work. - Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: issues with livecd-creator on F-11?
On Tue, Aug 25, 2009 at 2:49 AM, Peter Robinson pbrobin...@gmail.com wrote: Is anyone else having issues with livecd-creator? I'm seeing the issue below on a 2 F-11 boxes, and possible a rawhide one as well. [snip] IOError: [Errno 2] No such file or directory: '/var/tmp/imgcreate-qZYEPV/install_root/boot/initrd-2.6.31-0.167.rc6.git6.fc12.i686.img' The move to dracut required some changes in livecd-creator which have been in git for a week or so. Those are in livecd-tools-026 which should show up in today's rawhide Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
[Fedora-livecd-list] 2 commits - imgcreate/live.py tools/livecd-iso-to-disk.sh
imgcreate/live.py |1 + tools/livecd-iso-to-disk.sh |2 ++ 2 files changed, 3 insertions(+) New commits: commit 59aa5640bf2216d1823ce7d552ae8530d44bdabb Author: Jeremy Katz ka...@redhat.com Date: Tue Aug 25 10:59:42 2009 -0400 Fix traceback when implantisomd5 isn't available (Ron Yorston) diff --git a/imgcreate/live.py b/imgcreate/live.py index 528840d..c49b358 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -264,6 +264,7 @@ class LiveImageCreatorBase(LoopImageCreator): implantisomd5 = /usr/lib/anaconda-runtime/implantisomd5 else: logging.warn(isomd5sum not installed; not setting up mediacheck) +return subprocess.call([implantisomd5, iso]) commit 89720ad085e3996bded0e9570878e56a158232ff Author: Jeremy Katz ka...@redhat.com Date: Thu Aug 20 14:13:41 2009 -0400 Sleep to avoid races (#517477) Try to sleep a little to avoid racing with system stuff poking at new filesystems diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 0efef58..cd75153 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -28,6 +28,7 @@ usage() { } cleanup() { +sleep 2 [ -d $CDMNT ] umount $CDMNT rmdir $CDMNT [ -d $USBMNT ] umount $USBMNT rmdir $USBMNT } @@ -598,6 +599,7 @@ if [ $homesizemb -gt 0 ]; then done mke2fs -j /dev/mapper/EncHomeFoo tune2fs -c0 -i0 -ouser_xattr,acl /dev/mapper/EncHomeFoo + sleep 2 cryptsetup luksClose EncHomeFoo losetup -d $loop else -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] Makefile
Makefile |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 3553dafeccf8461b966ab7d5c327d9e22e9955ad Author: Jeremy Katz ka...@redhat.com Date: Mon Aug 24 16:41:52 2009 -0400 Bump version diff --git a/Makefile b/Makefile index 68cb635..f208c41 100644 --- a/Makefile +++ b/Makefile @@ -1,5 +1,5 @@ -VERSION = 025 +VERSION = 026 INSTALL = /usr/bin/install -c INSTALL_PROGRAM = ${INSTALL} -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] resize2fs: No space left while trying to resize ... ext3fs.img
On Thursday, August 20 2009, Warren Togami said: Trying livecd-creator as of git 1d215ebcf584b2bc19dece88b7f75944de2ad3f0 with August 20th rawhide + ntfsprogs to fix the broken dep. [snip] Resizing the filesystem on /opt/tmp/imgcreate-3h7N84/tmp-1Tw3yD/ext3fs.img to 534815 (4k) blocks. /sbin/resize2fs: No space left on device while trying to resize /opt/tmp/imgcreate-3h7N84/tmp-1Tw3yD/ext3fs.img resize2fs here is saying you don't have enough space on /opt for the resize and then failing. If you _do_ have lots of disk space there, file a bug against e2fsprogs and get sandeen to look at why it's failing Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] imgcreate/live.py
imgcreate/live.py | 67 +++--- 1 file changed, 49 insertions(+), 18 deletions(-) New commits: commit c6b31fa922b322fc5b0b9f48bca3a9ff812f8bdb Author: Jeremy Katz ka...@redhat.com Date: Tue Aug 18 14:07:58 2009 -0400 Support dracut based initrds With dracut, the initrd is named initrd-generic-*.img, so support looking for that. We also end up wanting to use a different root= line which also lets the hybrid images Just Work (tm) diff --git a/imgcreate/live.py b/imgcreate/live.py index f44e173..756097b 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -364,8 +364,14 @@ class x86LiveImageCreator(LiveImageCreatorBase): shutil.copyfile(bootdir + /vmlinuz- + version, isodir + /isolinux/vmlinuz + index) -shutil.copyfile(bootdir + /initrd- + version + .img, -isodir + /isolinux/initrd + index + .img) +isDracut = False +if os.path.exists(bootdir + /initrd-generic- + version + .img): +shutil.copyfile(bootdir + /initrd-generic- + version + .img, +isodir + /isolinux/initrd + index + .img) +isDracut = True +else: +shutil.copyfile(bootdir + /initrd- + version + .img, +isodir + /isolinux/initrd + index + .img) is_xen = False if os.path.exists(bootdir + /xen.gz- + version[:-3]): @@ -373,7 +379,7 @@ class x86LiveImageCreator(LiveImageCreatorBase): isodir + /isolinux/xen + index + .gz) is_xen = True -return is_xen +return (is_xen, isDracut) def __is_default_kernel(self, kernel, kernels): if len(kernels) == 1: @@ -408,18 +414,23 @@ menu hidden menu hiddenrow 5 % args -def __get_image_stanza(self, is_xen, **args): +def __get_image_stanza(self, is_xen, isDracut, **args): +if isDracut: +args[rootlabel] = live:LABEL=%(fslabel)s % args +else: +args[rootlabel] = CDLABEL=%(fslabel)s % args + if not is_xen: template = label %(short)s menu label %(long)s kernel vmlinuz%(index)s - append initrd=initrd%(index)s.img root=CDLABEL=%(fslabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s + append initrd=initrd%(index)s.img root=%(rootlabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s else: template = label %(short)s menu label %(long)s kernel mboot.c32 - append xen%(index)s.gz --- vmlinuz%(index)s root=CDLABEL=%(fslabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s --- initrd%(index)s.img + append xen%(index)s.gz --- vmlinuz%(index)s root=%(rootlabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s --- initrd%(index)s.img return template % args @@ -438,7 +449,9 @@ menu hiddenrow 5 index = 0 for version in versions: -is_xen = self.__copy_kernel_and_initramfs(isodir, version, index) +(is_xen, isDracut) = self.__copy_kernel_and_initramfs(isodir, version, index) +if index == 0: +self._isDracut = isDracut default = self.__is_default_kernel(kernel, kernels) @@ -449,7 +462,7 @@ menu hiddenrow 5 else: long = Boot %s(%s) % (self.name, kernel) -cfg += self.__get_image_stanza(is_xen, +cfg += self.__get_image_stanza(is_xen, isDracut, fslabel = self.fslabel, isofstype = auto, liveargs = kernel_options, @@ -462,7 +475,7 @@ menu hiddenrow 5 cfg += menu default\n if checkisomd5: -cfg += self.__get_image_stanza(is_xen, +cfg += self.__get_image_stanza(is_xen, isDracut, fslabel = self.fslabel, isofstype = auto, liveargs = kernel_options, @@ -540,8 +553,12 @@ hiddenmenu %args def __get_efi_image_stanza(self, **args): +if self._isDracut: +args[rootlabel] = live:LABEL=%(fslabel)s % args +else: +args[rootlabel] = CDLABEL=%(fslabel)s % args return title %(long)s - kernel /EFI/boot/vmlinuz%(index)s root=CDLABEL=%(fslabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s + kernel /EFI/boot/vmlinuz%(index)s root=%(rootlabel)s rootfstype=%(isofstype)s %(liveargs)s %(extra)s initrd /EFI/boot/initrd%(index)s.img %args @@ -649,6 +666,7 @@ class ppcLiveImageCreator(LiveImageCreatorBase): return { 32 : False, 64 : True } def __copy_kernel_and_initramfs(self, destdir, version): +isDracut = False bootdir = self._instroot + /boot makedirs(destdir) @@ -656,8 +674,15 @@ class
Re: Thoughts on NOPASSWD and disabling agent forwarding on publictest machines?
On Sunday, August 16 2009, Mike McGrath said: I'm conflicted on this, there's valid points here but also the risks are fairly low. As far as disabling agent forwarding, that's trivial to re-enable if the box gets rooted. We could add something to the security doc suggesting something like the following in ~/.ssh/config Host publictest*.fedoraproject.org ForwardAgent no Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
[Fedora-livecd-list] 2 commits - imgcreate/fs.py imgcreate/live.py
imgcreate/fs.py |3 +++ imgcreate/live.py |5 - 2 files changed, 7 insertions(+), 1 deletion(-) New commits: commit 2315f1a3cfbd38257e86b9f785f11e9a99784d67 Author: Jeremy Katz ka...@redhat.com Date: Mon Aug 17 09:50:08 2009 -0400 Apparently the python traceback isn't clear enough (#517841) diff --git a/imgcreate/live.py b/imgcreate/live.py index b98f7b3..773d2c2 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -171,7 +171,10 @@ class LiveImageCreatorBase(LoopImageCreator): LiveOS/ext3fs.img nor os.img exist % base_on) -shutil.copyfile(os_image, self._image) +try: +shutil.copyfile(os_image, self._image) +except IOError, e +raise CreatorError(Failed to copy base live image to %s for modification: %s %(self._image, e)) finally: squashloop.cleanup() isoloop.cleanup() commit 7a62ff24c4f1f7e46326210d0514e4b71de5772b Author: Jeremy Katz ka...@redhat.com Date: Mon Aug 17 09:47:14 2009 -0400 Add a sleep to try to avoid tracebacks with snapshot removal (#506644) diff --git a/imgcreate/fs.py b/imgcreate/fs.py index a9512f2..31e5ca3 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -26,6 +26,7 @@ import random import string import logging import tempfile +import time from imgcreate.errors import * @@ -516,6 +517,8 @@ class DeviceMapperSnapshot(object): if not self.__created: return +# sleep to try to avoid any dm shenanigans +time.sleep(2) rc = subprocess.call([/sbin/dmsetup, remove, self.__name]) if not ignore_errors and rc != 0: raise SnapshotError(Could not remove snapshot device) -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] imgcreate/live.py
imgcreate/live.py |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 832c78e2d6f54e130b845dd7d63fa457b0f513ea Author: Jeremy Katz ka...@redhat.com Date: Mon Aug 17 10:50:01 2009 -0400 Fix my stupid typo. *sigh* diff --git a/imgcreate/live.py b/imgcreate/live.py index 773d2c2..f44e173 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -173,7 +173,7 @@ class LiveImageCreatorBase(LoopImageCreator): try: shutil.copyfile(os_image, self._image) -except IOError, e +except IOError, e: raise CreatorError(Failed to copy base live image to %s for modification: %s %(self._image, e)) finally: squashloop.cleanup() -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] xen livecd creation
On Sunday, August 16 2009, Patrice Guay said: I'm trying to create a LiveCD with a xen enabled kernel (kernel-xen). I'm using livecd-tools 013 (the last compatible version with CentOS5). Creating a regular LiveCD works fine but I'm having trouble with the Xen LiveCD. [snip] Anyone has a clue about what is going wrong with this Xen LiveCD? Poke around with the shell and make sure the CD device is showing up and has the right label (/lib/udev/vol_id against the cd device) Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Why resize back ext3fs.img to bigger size.
On Sunday, August 16 2009, Mads Kiilerich said: Alexander Todorov wrote, On 08/15/2009 10:39 PM: Hi, in fs.py in ExtDiskMount.resparse() function we resize and truncate the image to minimum possible size (i.e. ext3 becomes 100% full) and then resize it back to the size specified in kickstart. In most cases this results in ext3fs.img which when mounted has lots of free space on it. Why is that ? Perhaps it is because unused space might contain old data which probably compresses badly. By doing it this way the extended space will be initialized with zeroes which can be compressed to (almost) nothing, and the compressed image of the specified size thus doesn't take up more space than if it had the minimal size. Correct. And by having the free space, we are able to let you write things to the filesystem when you're running the live image rather than having every write operation return -ENOSPC. We also keep around a snapshot of the minimal image so that we use the minimal image as the basis to copy over after an install from the live image (and then we resize that minimal image on the disk to the size of the partition you created) Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: An error while using livecd-creator
On Tuesday, August 11 2009, Kushal Das said: I am using livecd-creator on a F-11 box. I have 27GB free on my / partition. The error I am getting is given below: [snip] Error creating Live CD : Unable to install: [('installing package bug-buddy-1:2.26.0-2.fc11.i586 needs 684KB on the /var/tmp/imgcreate-rScTyr/install_root filesystem', (9, [snip] Any pointer on how to fix this ? As the message says, there's not enough space in the root fs of the livecd (/var/tmp/imgcreate-X/install_root) to install the packages. If you were overriding the size of the / partition, be sure you have the latest pykickstart package installed as there was a bug there that broke the overriding Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
[Fedora-livecd-list] 3 commits - imgcreate/fs.py imgcreate/kickstart.py Makefile
Makefile |2 +- imgcreate/fs.py|2 +- imgcreate/kickstart.py | 12 +--- 3 files changed, 7 insertions(+), 9 deletions(-) New commits: commit ff31b963edb642ae210455cf38a88392feb84a9f Author: Jeremy Katz ka...@redhat.com Date: Fri Aug 14 10:50:59 2009 -0400 If resize2fs fails, raise an error also (related to #517222) diff --git a/imgcreate/fs.py b/imgcreate/fs.py index 563b7a9..a9512f2 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -68,7 +68,7 @@ def resize2fs(fs, size = None, minimal = False): args.append(%sK %(size / 1024,)) ret = subprocess.call(args) if ret != 0: -return ret +raise ResizeError(resize2fs returned an error (%d)! image to debug at %s %(ret, saved_image)) if e2fsck(fs) != 0: raise ResizeError(fsck after resize returned an error! image to debug at %s %(saved_image,)) commit b1496d9f5ee11c42ca4446eb5e4da9de025500cd Author: Jeremy Katz ka...@redhat.com Date: Thu Jul 30 16:43:42 2009 -0400 Bumpv ersion diff --git a/Makefile b/Makefile index a3b2aa9..68cb635 100644 --- a/Makefile +++ b/Makefile @@ -1,5 +1,5 @@ -VERSION = 024 +VERSION = 025 INSTALL = /usr/bin/install -c INSTALL_PROGRAM = ${INSTALL} commit 8d1a422ca611e7f05da8f8f3b8e46c7d3bcb4e86 Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 27 09:26:58 2009 -0400 Use system-config-keyboard's bits in preference to rhpl.keyboard system-config-keyboard = 1.3.0 has support for the keyboard pieces that used to be in rhpl. Use them if we can but fall back gracefully as people running from git could be on older releases diff --git a/imgcreate/kickstart.py b/imgcreate/kickstart.py index 83a62e4..98db856 100644 --- a/imgcreate/kickstart.py +++ b/imgcreate/kickstart.py @@ -24,7 +24,10 @@ import time import logging import urlgrabber -import rhpl.keyboard +try: +import system_config_keyboard.keyboard as keyboard +except ImportError: +import rhpl.keyboard as keyboard import pykickstart.commands as kscommands import pykickstart.constants as ksconstants @@ -132,12 +135,7 @@ class LanguageConfig(KickstartConfig): class KeyboardConfig(KickstartConfig): A class to apply a kickstart keyboard configuration to a system. def apply(self, kskeyboard): -# -# FIXME: -# should this impact the X keyboard config too? -# or do we want to make X be able to do this mapping? -# -k = rhpl.keyboard.Keyboard() +k = keyboard.Keyboard() if kskeyboard.keyboard: k.set(kskeyboard.keyboard) k.write(self.instroot) -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
The death of rhpl
We're hoping to be able to not ship rhpl for Fedora 12 -- this isn't something which is really a feature, but it might be something that a heads up is useful for. Anyone who has a package with a dependency on rhpl should have a bug filed, 98% of them with patches, to switch to using something else. If you are using rhpl and *don't* have a Requires: rhpl in your spec file (or are using it for something not in Fedora), here's the replacements for the things which were commonly used * rhpl.translate - Use python's gettext module. It's a lot more functional these days. rhpl.translate was a good idea 7 years ago, but not anymore :) * rhpl.ethtool - python-ethtool provides a module with similar/the same API * rhpl.exception - Chris Lumens has split out a new package python-meh which includes the exception handling capabilities from rhpl with the added support for filing to bugzilla like anaconda does * rhpl.keyboard* - these now live in system-config-keyboard Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: [Fedora-livecd-list] unbootable rawhide images
On Tuesday, July 28 2009, Sebastian Vahl said: Are there some issues known that current live images with livecd-creator from git are unbootable atm? With my images I have the problem that the booting stops after the squashfs message of the kernel (eg. here from VirtualBox: http://img525.imageshack.us/img525/2016/livecdcreatorrawhidenob.png) There was something which broke things last week, but I fixed it up last Monday or Tuesday. Haven't heard/noticed anything new. Does your rawhide mirror have mkinitrd-6.0.92? Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Rawhide error
On Tuesday, July 28 2009, David Huff said: Retrieving http://download.fedora.devel.redhat.com/pub/fedora/linux/development/x86_64/os/repodata/repomd.xml ...OK /usr/lib/python2.6/site-packages/imgcreate/errors.py:45: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6 return unicode(self.message) Error creating Live CD : Dependency check failed! Make sure you have a rawhide rpm (rpm 4.7.1) -- otherwise, you're probably hitting that the version of rpm that you have installed can't install packages with the new payload format Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Rawhide error
On Tuesday, July 28 2009, Adam Miller said: On Tue, Jul 28, 2009 at 10:07 AM, Jeremy Katzka...@redhat.com wrote: Make sure you have a rawhide rpm (rpm 4.7.1) -- otherwise, you're probably hitting that the version of rpm that you have installed can't install packages with the new payload format snip Is it safe to assume that version of rpm is destined to F10 and F11 at some point? It's already in at least F11 updates-testing since I picked it up on one of my boxes recently. Not sure what the intent is for f10 Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Updated Anaconda packages
On Monday, July 27 2009, Jeff Garzik said: Honestly, I always thought Fedora install images should be regenerated far more frequently. I think back to my days as a Solaris sysadmin in the late 90's, where ordering the latest media kit (CD-ROM) from Sun meant I got a fresh installer, fresh kernel, and all recommended patches. And Sun was doing those media kits roughly every six to twelve months from what I remember. And our (major) release frequency is every six months, so ... Even in the face of known Linux kernel bugs, people always seemed reluctant to regenerate the Fedora install images. I think Fedora would better serve its users by being much more willing to update install images after initial release. Regenerating the images is expensive -- it requires effort on the part of the developers doing fixes, release engineering doing builds with the fixes, QA testing the fixes, infrastructure (mirrors) carrying a significant amount more bits[1], ... When our releases are at most six months apart, how much effort are we willing to divert from the next release to make this happen? It's not going to happen for free. Jeremy [1] What gets respun? All the images (DVD, live, etc)? Some of them? If some, which ones? How do we differentiate between them? -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: Updated Anaconda packages
On Tuesday, July 28 2009, Ralf Corsepius said: On 07/27/2009 10:21 PM, Jeremy Katz wrote: Regenerating the images is expensive -- it requires effort on the part of the developers doing fixes, release engineering doing builds with the fixes, QA testing the fixes, infrastructure (mirrors) carrying a significant amount more bits[1], ... Not quite true. Instead of building all images, you could build a minimalist network install image, which installs from Everything+updates. Except that the majority of our users don't do an install in that fashion, so you're going to be helping the 1% case that least needs help. I don't know how you build images, but it's hard to image building such an image frequently (e.g. whenever any package it contains has changed) is such kind of expensive. The build time, test time[1], mirror time/space is all still there... those are the expensive parts. An alternative would be, to ship a script to let people build such an image themselves. It would not help everybody in all situations, but it at least help people who have some (other) version of Fedora running somewhere. As it turns out, we ship all the tools to build the distribution the exact way we do! And as David said, he's been working with Jeroen for occasional updated anaconda packages. Jeremy [1] And associated find bug/fix bug/respin cycle times -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
[Fedora-livecd-list] 2 commits - imgcreate/errors.py imgcreate/live.py
imgcreate/errors.py |2 ++ imgcreate/live.py |3 +++ 2 files changed, 5 insertions(+) New commits: commit 2ef4083a8c1adea520dc22bc43ae4ceb65dc85de Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 20 21:26:23 2009 -0400 Make live image a hybrid Use isohybrid if available so that images can be dd'd to a usb stick. Note that there's still some initrd changes required for this to be fully functional diff --git a/imgcreate/live.py b/imgcreate/live.py index a4914e0..b98f7b3 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -248,6 +248,9 @@ class LiveImageCreatorBase(LoopImageCreator): if subprocess.call(args) != 0: raise CreatorError(ISO creation failed!) +if os.path.exists(/usr/bin/isohybrid): +subprocess.call([/usr/bin/isohybrid, iso]) + self.__implant_md5sum(iso) def __implant_md5sum(self, iso): commit 733f9ce7b96ed492e4b974b089fa647db569e957 Author: Jeremy Katz ka...@redhat.com Date: Thu Jul 9 13:44:12 2009 -0400 Include new error definition which I apparently didn't commit diff --git a/imgcreate/errors.py b/imgcreate/errors.py index a29b841..800dc3b 100644 --- a/imgcreate/errors.py +++ b/imgcreate/errors.py @@ -52,3 +52,5 @@ class SnapshotError(CreatorError): pass class SquashfsError(CreatorError): pass +class ResizeError(CreatorError): +pass -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Fit and Finish test day: batteries and suspend
On Tuesday, July 21 2009, Matthias Clasen said: On Tue, 2009-07-21 at 13:10 +1000, Rodd Clarkson wrote: It would be great if the test day (and others) could link to a iso for rawhide that fits on a CD to make this part of the process simple. hint, hint ;-] Yeah, I'm working on it. However, todays (and yesterdays) spins so far had the unfortunate tendency to not boot at all, at least in qemu. It should be fixed with today's rawhide (mkinitrd-6.0.92) -- Jens mentioned that something was broken last night on IRC and I tracked it down to set -e in the /init of the initramfs breaking something again. Also need to figure out why we end up losing all of the (useful) output when plymouthd is running during the live /init, but at this point, I'm going to see if it continues happening once I get live images booting with dracut and just fix it there if so Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: creating minimal boot image for network installations
On Thursday, July 16 2009, Dan Hor?k said: I don't know about a simple way to start a network installation of Fedora from USB stick. Creating of boot.img, that server that purpose, was stopped by rel-engs some releases ago. I have found a blog post [1] that makes possible to transform a boot.iso image into a bootable USB stick. But it still requires to download the 200+ MB large boot.iso file and do it by hand, while in theory only initrd.img and vmlinuz files should be needed and the whole process could be automated. So I have created mkbootimg.sh script [2] that automates all required steps and has a bootable disk image as a result. It can be transferred to real USB disk by the dd command. I find it useful when testing rawhide installations, but can be used for releases too. There is naturally no warranty for the script, it can eat whole your system :-) With Fedora 12, you'll be able to just dd the boot.iso onto a USB stick since we're making isohybrid'd images. Note that while kernel + initrd is all you need, not having the stage2 means we have to ask where to find it rather than being able to automate the use of the mirror list for everything Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: [Fedora-livecd-list] 3 more megs if we get rid of redundant copy of kernel?
On Wednesday, July 15 2009, Douglas McClendon said: In my livecd generation tools I go to a bit of extra hassle to play symlink games with /boot/vmlinuz-$(uname -r), such that it doesn't get put into the squashfs, since a copy of it is in the container iso/usb filesystem. Maybe we want to do that for f12? It is a bit messy, but I think it's worth it for 3/700M. Nope, it breaks the ability to easily install by copying over the filesystem. And when you consider that someone may be booting via pxe and then trying to install (see livecd-iso-to-pxeboot), there's no guarantee that we can get that file back. Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] livecd-creator creates ISO without /home
On Wednesday, July 15 2009, Martin Dengler said: On Tue, Jul 14, 2009 at 09:37:33PM -0400, Jeremy Katz wrote: On Tuesday, July 14 2009, Martin Dengler said: On Tue, Jul 14, 2009 at 11:02:25AM -0400, Jeremy Katz wrote: On Tuesday, July 14 2009, Martin Dengler said: I'm getting a .ISO without a /home despite the filesystem rpm being installed, and I'd really appreciate any tips on debugging this or hints as to how this might happen. [...] The fact that groups aren't existing early in the install after the setup package has been installed is the more concerning part of the log to me. The machine in question has an LDAP user database, from what I understand. I don't know how that might influence the users/groups in the created filesystem. Ahh, that is useful and could be related. Are you also using nscd by chance? If so, try turning it off Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Orphaning garmin-sync and pyusb
Since I upgraded from the Garmin Edge 305 to the 705, I'm no longer using either of garmin-sync or its dependency pyusb on any sort of regular basis. This is probably less than ideal for adequately maintaining them. If you *have* one of the older Garmin Edge or Forerunner devices, they're easy enough packages to maintain -- upstream is responsive to the one or two things I threw at him and the userbase is not that large. Orphaned in pkgdb for devel -- I'll keep responding to things for F10/F11 unless someone else picks them up to avoid leaving users of existing releases out in the cold Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: [Fedora-livecd-list] livecd-creator creates ISO without /home
On Tuesday, July 14 2009, Martin Dengler said: I'm getting a .ISO without a /home despite the filesystem rpm being installed, and I'd really appreciate any tips on debugging this or hints as to how this might happen. Any SELinux errors? Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] livecd-creator creates ISO without /home
On Tuesday, July 14 2009, Martin Dengler said: On Tue, Jul 14, 2009 at 11:02:25AM -0400, Jeremy Katz wrote: On Tuesday, July 14 2009, Martin Dengler said: I'm getting a .ISO without a /home despite the filesystem rpm being installed, and I'd really appreciate any tips on debugging this or hints as to how this might happen. Any SELinux errors? These don't look like SELinux errors but they look like something security-related: Those messages are far later (and not abnormal, although authconfig makes them look a little scary). The fact that groups aren't existing early in the install after the setup package has been installed is the more concerning part of the log to me. Which has usually only happened if SELinux is denying things. So checking for AVCs with ausearch -m avc would be very helfpul Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 4 commits - imgcreate/creator.py imgcreate/fs.py
imgcreate/creator.py |2 - imgcreate/fs.py | 62 +++ 2 files changed, 35 insertions(+), 29 deletions(-) New commits: commit 6a3849d072b142f6c5c5bba07dd6521a1ef0c03a Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 6 14:29:30 2009 -0400 Store a copy of the filesystem image before resize Save a copy of an e2image before we do the resize so that if it fails, people can have something to provide via bugzilla diff --git a/imgcreate/fs.py b/imgcreate/fs.py index c19e8a0..c536de3 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -25,6 +25,7 @@ import subprocess import random import string import logging +import tempfile from imgcreate.errors import * @@ -56,6 +57,9 @@ def resize2fs(fs, size = None, minimal = False): raise RuntimeError(Must specify either a size or minimal for resize!) e2fsck(fs) +(fd, saved_image) = tempfile.mkstemp(, resize-image-, /tmp) +os.close(fd) +subprocess.call([/sbin/e2image, -r, fs, saved_image]) dev_null = os.open(/dev/null, os.O_WRONLY) args = [/sbin/resize2fs, fs] @@ -72,7 +76,8 @@ def resize2fs(fs, size = None, minimal = False): return ret if e2fsck(fs) != 0: -raise CreatorError(fsck after resize returned an error!) +raise CreatorError(fsck after resize returned an error! image to debug at %s %(saved_image,)) +os.unlink(saved_image) return 0 def e2fsck(fs): commit 44c17e8655e2787d3bb2fdc5ab32be18526ce284 Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 6 14:26:43 2009 -0400 Consolidate calls to fsck We always want to fsck before/after resizing, so just do it in the resize method rather than having it scattered everywhere. diff --git a/imgcreate/fs.py b/imgcreate/fs.py index 567376a..c19e8a0 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -54,6 +54,9 @@ def resize2fs(fs, size = None, minimal = False): raise RuntimeError(Can't specify both minimal and a size for resize!) if not minimal and size is None: raise RuntimeError(Must specify either a size or minimal for resize!) + +e2fsck(fs) + dev_null = os.open(/dev/null, os.O_WRONLY) args = [/sbin/resize2fs, fs] if minimal: @@ -61,9 +64,21 @@ def resize2fs(fs, size = None, minimal = False): else: args.append(%sK %(size / 1024,)) try: -return subprocess.call(args, stdout = dev_null, stderr = dev_null) +ret = subprocess.call(args, stdout = dev_null, stderr = dev_null) finally: os.close(dev_null) + +if ret != 0: +return ret + +if e2fsck(fs) != 0: +raise CreatorError(fsck after resize returned an error!) +return 0 + +def e2fsck(fs): +logging.debug(Checking filesystem %s % fs) +rc = subprocess.call([/sbin/e2fsck, -f, -y, fs]) +return rc class BindChrootMount: Represents a bind mount of a directory into a chroot. @@ -402,11 +417,7 @@ class ExtDiskMount(DiskMount): if size current_size: self.disk.expand(size) -self.__fsck() - resize2fs(self.disk.lofile, size) - -self.__fsck() return size def __create(self): @@ -426,8 +437,7 @@ class ExtDiskMount(DiskMount): DiskMount.mount(self) def __fsck(self): -logging.debug(Checking filesystem %s % self.disk.lofile) -rc = subprocess.call([/sbin/e2fsck, -f, -y, self.disk.lofile]) +return e2fsck(self.disk.lofile) return rc def __get_size_from_filesystem(self): @@ -449,12 +459,8 @@ class ExtDiskMount(DiskMount): return int(parse_field(out, Block count)) * self.blocksize def __resize_to_minimal(self): -self.__fsck() resize2fs(self.disk.lofile, minimal = True) -min = self.__get_size_from_filesystem() -if self.__fsck() != 0: -raise CreatorError(fsck returned an error!) -return min +return self.__get_size_from_filesystem() def resparse(self, size = None): self.cleanup() commit 1e5e6019a058c6acde0b4cb85a21cac4d7b66a60 Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 6 13:49:14 2009 -0400 Use resize2fs -M instead of binary search esandeen got a resize to minimal size option added to resize2fs, so we should use it rather than our bisecting down to a minimal size. And maybe then we'll hit fewer resize2fs bugs. diff --git a/imgcreate/fs.py b/imgcreate/fs.py index 65f16e4..567376a 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -49,11 +49,19 @@ def mksquashfs(in_img, out_img): raise SquashfsError('%s' exited with error (%d) % (string.join(args, ), ret)) -def resize2fs(fs, size): +def resize2fs(fs, size = None, minimal = False): +if minimal and size is not None: +raise RuntimeError(Can't specify both minimal and a size for resize!) +if not minimal and size is None
[Fedora-livecd-list] imgcreate/fs.py
imgcreate/fs.py |7 +-- 1 file changed, 1 insertion(+), 6 deletions(-) New commits: commit 94d23f008f8f42e0cf93cd8b6cc0a4381d408dfb Author: Jeremy Katz ka...@redhat.com Date: Mon Jul 6 14:51:06 2009 -0400 Also, don't lose resize2fs output (per request of sandeen) diff --git a/imgcreate/fs.py b/imgcreate/fs.py index ae79b1c..563b7a9 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -61,17 +61,12 @@ def resize2fs(fs, size = None, minimal = False): os.close(fd) subprocess.call([/sbin/e2image, -r, fs, saved_image]) -dev_null = os.open(/dev/null, os.O_WRONLY) args = [/sbin/resize2fs, fs] if minimal: args.append(-M) else: args.append(%sK %(size / 1024,)) -try: -ret = subprocess.call(args, stdout = dev_null, stderr = dev_null) -finally: -os.close(dev_null) - +ret = subprocess.call(args) if ret != 0: return ret -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: FESCo meeting summary for 2009-06-26
On Saturday, June 27 2009, Kevin Kofler said: * fixing comps so task-oriented groups like SoundVideo aren't biased towards GNOME apps (this most likely requires extending the comps format or having separate comps-kde and comps-gnome - I think extending the format to handle conditionals based on the desktop selected via radiobutton would be the more maintainable solution in the long run): where I'd like to get to is that if I select SoundVideo after having selected GNOME as the desktop, I get Totem and Rhythmbox by default and kdemultimedia, Amarok and Kaffeine as optional checkboxes, but if I select SoundVideo after having selected KDE as the desktop, I get kdemultimedia, Amarok and Kaffeine by default and Totem and Rhythmbox as optional checkboxes, fwiw, concrete discussions on how to actually improve comps so that it scales to the order of magnitude more packages + many many many more use cases we have are welcome. We had a long discussion about it at the FUDCon in January, but no one was really happy with the idea we came up with (as it introduced new problems and didn't solve all the existing ones) and so we continue to live with where we are. Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: fedora-git-commit-mail-hook: Fix command for finding previous tag
On Tuesday, June 23 2009, Todd Zullinger said: Jeremy Katz wrote: That sounds sensible. I have no problem with us changing, although I won't have the time to sit down and make it happen any time in the next couple of weeks. If someone can get to it before, I'm more than happy to look over it. I'm happy to work on this. I mostly wanted to check that the idea was agreeable first. I'll try not to come up with too many suggestions that cause other people work -- other than the work of reading my long emails... :) Awesome. Reading and replying to emails is what I do best at this point ;) I an ideal world, we would use the upstream post-receive-email unmodified. I'm not sure if that will be possible or not, as we'd want to pipe mail to the send-unicode-email.py script, rather than just through sendmail. A patch to the upstream script to allow setting the path to 'sendmail' might be in order. And at the least, a simple puppet exec could fix that one part and leave us otherwise using the upstream hook. That sounds sensible My biggest concern is that the format of the notifications has changed a good bit since the current fedora-git-commit-mail-hook was added. [snip] The upstream hook should produce more useful results for various operations, adding tags, branches, and such. But as with any change, some people are bound to not like it. (I admit that I'm not all that fond of the output for the simple case of pushing a new commit, as the summary is added after the diff.) Hmmm, most of the changes I think are harmless aside from people getting thrown off the first time. I do agree that the summary being at the bottom is a little bit more bothersome than most of the changes. Also, it looks like we'd lose the X-Git-Module header which is sort of important for at least some people. BTW, Jeremy, do you happen to recall just what the origin of the current script is? I was trying to find a common ancestor that I could use to cherry-pick patches from git.git into. The current script was originally running on git.freedesktop.org and written by Carl Worth, then hijacked and tweaked more than I would have liked by me to fit our needs Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |3 +++ 1 file changed, 3 insertions(+) New commits: commit d2de65a9ee73af0d001660ee8b2939d88f4f83fe Author: Martin Dengler mar...@martindengler.com Date: Tue Jun 23 21:59:16 2009 +0100 don't need to checkMBR() for loop devices livecd-iso-to-disk.sh doesn't do the right thing w.r.t. loop devices and MBR detection. I don't think this is intentional, but it currently requires loop devices to start with an MBR, by virtue of getdisk() being a NOOP for loop devices and checkMBR assuming that the disk must need an MBR. checkMBR() should just ignore loop devices like other key functions do. diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 672e6db..0efef58 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -93,6 +93,9 @@ resetMBR() { } checkMBR() { +if [[ $DEV =~ /dev/loop* ]]; then + return 0 +fi getdisk $1 bs=$(mktemp /tmp/bs.XX) -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] don't need to checkMBR() for loop devices
On Tuesday, June 23 2009, Martin Dengler said: livecd-iso-to-disk.sh doesn't do the right thing w.r.t. loop devices and MBR detection. I don't think this is intentional, but it currently requires loop devices to start with an MBR, by virtue of getdisk() being a NOOP for loop devices and checkMBR assuming that the disk must need an MBR. checkMBR() should just ignore loop devices like other key functions do. Thanks, applied Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: fedora-git-commit-mail-hook: Fix command for finding previous tag
On Monday, June 22 2009, Todd Zullinger said: [snip] I think we may want to look at converting from our use of the update hook to the post-receive hook for sending commit notifications. Upstream moved the mail notification bits from the update hook to a new post-receive hook only a few weeks after Jeremy added the initial fedora-git-commit-mail-hook. So what we're running now isn't something which will receive any attention or bug fixes from upstream, unfortunately. The reasoning for upstream moving mail notification is stated in the the commit message for 46d409d: That sounds sensible. I have no problem with us changing, although I won't have the time to sit down and make it happen any time in the next couple of weeks. If someone can get to it before, I'm more than happy to look over it. Alternately, I can stick on my queue for July :) Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
[Fedora-livecd-list] 2 commits - tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh | 21 + 1 file changed, 13 insertions(+), 8 deletions(-) New commits: commit a3227625a1317026fbbd331fcbca560716b6cfa8 Author: Jason Farrell farre...@gmail.com Date: Wed Jun 17 19:53:15 2009 -0400 small fix for --multi mode One other tiny patch for the mostly unused --multi mode. The initrd= in isolinux.cfg wasn't getting replaced, and, it's easier if the $LIVEOS/$SYSLINUXPATH paths are absolute when editing the main /syslinux/syslinux.cfg or /syslinux/extlinux.conf menu entries to add them for multibooting. diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 01dadb3..672e6db 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -696,8 +696,8 @@ if [ -z $multi ]; then fi else # we need to do some more config file tweaks for multi-image mode - sed -i -e s;kernel vm;kernel $LIVEOS/syslinux/vm; $USBMNT/$SYSLINUXPATH/isolinux.cfg - sed -i -e s;initrd i;initrd $LIVEOS/syslinux/i; $USBMNT/$SYSLINUXPATH/isolinux.cfg + sed -i -e s;kernel vm;kernel /$LIVEOS/syslinux/vm; $USBMNT/$SYSLINUXPATH/isolinux.cfg + sed -i -e s;initrd=i;initrd=/$LIVEOS/syslinux/i; $USBMNT/$SYSLINUXPATH/isolinux.cfg mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/syslinux.cfg cleanup fi commit dd2f7eccd287932b1d84287877997a9cf94a0d33 Author: Jason Farrell farre...@gmail.com Date: Wed Jun 17 19:28:54 2009 -0400 Make overlay space requirements clearer My biggest annoyance with the livecd-iso-to-disk script has simply been having to do a little math in my head when trying to fit the largest possible overlay combo onto my liveusb. The script should be doing this work - it should tell you by how much you're over so you can free it and/or quickly adjust your sizes. diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index de439d5..01dadb3 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -476,13 +476,18 @@ if [ -n $skipcompress ]; then fi free=$(df -B1M $USBDEV |tail -n 1 |awk {'print $4;'}) -if [ $(($overlaysizemb + $homesizemb + $livesize + $swapsizemb)) -gt $(($free + $tbd)) ]; then + tba=$(($overlaysizemb + $homesizemb + $livesize + $swapsizemb)) +if [ $tba -gt $(($free + $tbd)) ]; then echo Unable to fit live image + overlay on available space on USB stick - echo Size of live image: $livesize - [ $overlaysizemb -gt 0 ] echo Overlay size: $overlaysizemb - [ $homesizemb -gt 0 ] echo Home overlay size: $homesizemb - [ $swapsizemb -gt 0 ] echo Home overlay size: $swapsizemb - echo Available space: $(($free + $tbd)) + echo + Size of live image: $livesize + [ $overlaysizemb -gt 0 ] echo + Overlay size: $overlaysizemb + [ $homesizemb -gt 0 ] echo + Home overlay size: $homesizemb + [ $swapsizemb -gt 0 ] echo + Swap overlay size: $swapsizemb + echo --- + echo = Requested: $tba + echo - Available: $(($free + $tbd)) + echo --- + echo = To fit, free or decrease requested size total by: $(($tba - $free + $tbd)) exitclean fi -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Do we need split media CDs for F12?
On Sunday, June 14 2009, Chris Adams said: Once upon a time, Jeremy Katz ka...@redhat.com said: See the livecd-iso-to-pxeboot script, although it does place some (somewhat) different requirements on things. AFAIK livecd-iso-to-pxeboot is useless for 32 bit, at least for the standard Fedora LiveCD images. I think the kernel will only use an initrd that is less than half the size of lowmem, or 448M. To be honest, I'm not sure. I've never actually used the script -- it was written by some of the guys working on the project now known as ovirt and they use it pretty heavily It would be useful to be able to export the root FS from a LiveCD via NFS, or maybe have an alternate initrd for PXE booting that could NFS mount the ISO image (and then the LiveCD root), or fetch the ISO into RAM via HTTP, or something along those lines. With the way we build the initrd today, there's no way[1] to do so. But as I said in my previous mail, with dracut maybe we can start to look at other interesting things like this Jeremy [1] Well, there's a way... but not anything sustainable. Having multiple completely separate initrd paths is all kinds of pain -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
[Fedora-livecd-list] docs/livecd-iso-to-disk.pod
docs/livecd-iso-to-disk.pod | 18 +- 1 file changed, 17 insertions(+), 1 deletion(-) New commits: commit d4c5e39d9898ced0ef8fd0ebe41938d1460a78cf Author: Jeremy Katz ka...@redhat.com Date: Mon Jun 15 15:38:50 2009 -0400 Update for current options (#505742) Update for current options based on a patch from Michel Duquaine (michelduquaine AT gmail DOT com) diff --git a/docs/livecd-iso-to-disk.pod b/docs/livecd-iso-to-disk.pod index 7449c17..9274ab8 100644 --- a/docs/livecd-iso-to-disk.pod +++ b/docs/livecd-iso-to-disk.pod @@ -6,7 +6,7 @@ livecd-iso-to-disk - installs bootable CD images on a USB storage devices. =head1 SYNOPSIS -Blivecd-iso-to-disk [--reset-mbr] [--noverify] [--overlay-size-mb size] path-to-iso usb storage device +Blivecd-iso-to-disk [--format] [--reset-mbr] [--noverify] [--overlay-size-mb size] [\-\-home\-size\-mb size] [\-\-unencrypted\-home] [\-\-skipcopy] path-to-iso usb storage device =head1 DESCRIPTION @@ -18,6 +18,10 @@ Blivecd-iso-to-disk is not a destructive process; any data you currently have =over 4 +=item --format + +Formats the USB stick and creates a GPT partition table. + =item --reset-mbr Sets the Master Boot Record(MBR) of the USB storage device to the mbr.bin file from the image's syslinux directory. @@ -30,6 +34,18 @@ Disables the image validation process which occurs before the image is installed This option sets the overlay size in megabytes. The overlay is additional storage available to the live operating system if the operating system supports it. The USB storage device must have enough free space for the image and the overlay. +=item --home-size-mb + +Sets the home directory size in megabytes. + +=item --unencrypted-home + +Disables the encryption of the home directory. + +=item --skipcopy + +Skips the copy of the live image to the USB stick. + =back =head1 CONTRIBUTORS -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Do we need split media CDs for F12?
On Saturday, June 13 2009, Jussi Lehtola said: On Sat, 2009-06-13 at 11:12 -0500, Matt Domsch wrote: On Sat, Jun 13, 2009 at 07:04:12PM +0300, Jussi Lehtola wrote: Hmm, I'd want netboot.img back, since I normally use a USB stick to start the network install (OK, there is the possibility of using livecd-iso-to-disk, but that's a lot more hassle than downloading a minimalistic img and running dd). We have it, it's now called netinst.iso Yes but not netboot.img that could be dd'd straight away to a USB drive or whatnot; the iso needs livecd-iso-to-disk which a) is extra work and b) is only available on Fedora and Windows. [Also, the livecd tools need an own homepage so that users of other distros can get them.] We really need to finish the push in F12 to get liveusb-creator working for all cases (including command line) so that we can kick the silly shell script to the curb as liveusb-creator has its own homepage, etc. Also, I want to look a bit more at isohybrid to see if we can build iso images that can just be dd'd, at least for the case of boot.iso/netinst.iso Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: Do we need split media CDs for F12?
On Sunday, June 14 2009, King InuYasha said: Also, maybe we should support PXE/network booting the Live version from mirrors or whatever with the advent of netbooks and other computers without an optical drive. While doing it via USB is preferable, it is not always possible. For example I have a laptop with a completely damaged drive bay where the CD drive is and it does not support booting from USB devices. Being able to boot the Live distro from a network would be a great alternative. See the livecd-iso-to-pxeboot script, although it does place some (somewhat) different requirements on things. If we can get dracut in for F12, we might be able to be more clever with netboot + live images Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: GDM Language list...
On Sunday, June 14 2009, Jens Petersen said: - Bill Nottingham nott...@redhat.com wrote: https://fedoraproject.org/wiki/Features/YumLangpackPlugin My one concern with this is that the conditional stuff is also used on the compose side when making LiveCDs, etc. We need to make sure that still works somehow. Right. (Though since F11 we have dropped all the lang-support groups from the default spins.) Thanks for bringing it up - I see the potential weakness: so would it be sufficient to ship the plugin by default then or would livecd-tools need to pull it in? livecd-tools doesn't use any plugins at present, so there'd be some work needed to ensure the right thing happened Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: GRUB 2 in Ubuntu 9.10
On Thursday, June 11 2009, Michael Schwendt said: On Wed, 10 Jun 2009 17:17:07 -0400, Jeremy wrote: [1] So, the gory history for those who might be interested. Eight years ago (!), we decided that the advantage of not having to rerun lilo after changing the config file as you can just read the config file off the filesystem with grub was worthwhile. Why is GRUB reinstalled with every kernel update then? It's not. grubby is run, but that's just a tool that modifies your grub.conf to reflect the new kernel that's been installed Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: GRUB 2 in Ubuntu 9.10
On Wednesday, June 10 2009, King InuYasha said: On Wed, Jun 10, 2009 at 11:36 AM, Adam Jackson a...@redhat.com wrote: On Wed, 2009-06-10 at 11:07 -0500, King InuYasha wrote: Well, the existing GRUB used in distros was declared Legacy a long time ago. GRUB 2 is a rewrite that is supposed to include all the features the various vendors have been patching into GRUB Legacy, as well as being able to support EFI and basically supporting what the Chameleon bootloader does in addition to the GRUB Legacy's support. Though I doubt fake-EFI would be implemented in GRUB 2 The grub we're already shipping has EFI support. I have yet to hear of a problem we're actually having that would be solved with grub2. EFI support is not the same as fake-EFI. Erm, the EFI support in our grub today isn't fake-EFI. Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
Re: anaconda loses six versions from fc11 to fc12 ?
On Tuesday, June 09 2009, John Reiser said: It looks like the versions from .54 through .59 were dropped in the changeover from fc11 to fc12. Version .54 was in: [snip] so I hope that the work of two weeks can be recovered soon. (The changes were not in Monday's [20090608] rawhide report for fc12, either.) It's just because after things branched, builds were done from F-11/ instead of devel/. Regular builds from git master of anaconda to devel/ (and thus f12 rawhide) will probably pick up shortly. Jeremy -- fedora-devel-list mailing list fedora-devel-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-devel-list
[Fedora-livecd-list] Changes to 'refs/tags/livecd-tools-0.17.3'
Tag 'livecd-tools-0.17.3' created by Jeremy Katz ka...@redhat.com at 2009-05-27 14:11 + livecd-tools-0.17.3 Changes since livecd-tools-017.1-6: --- 0 files changed --- -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Change Request - email alias handling
On Thursday, May 21 2009, Toshio Kuratomi said: Fixing these requires updating the fas server code. Attaching a patch to hotfix our servers with to do this. The patch has been tested on fas1.stg successfully. Looks good, +1 Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
Re: [Fedora-livecd-list] should a livecd be ext3 or ext4
On Wednesday, May 20 2009, Rahul Sundaram said: On 05/20/2009 12:43 PM, Williamson Grant wrote: Should a livecd be created ext3 or ext4. The reason I ask, is we tend offer the livecd's also as usb sticks, when I last tried the usb stick does not boot if the filesystem is ext4. Comments anyone? Fedora 11 Live CD's will use Ext4 by default but retain a small /boot partition formatted as Ext3. Since GRUB doesn't support LVM and Fedora uses LVM by default, a separate /boot is needed anyway. You can create a customized live cd with a different filesystem by using --fstype in kickstart. This isn't entirely accurate. The live images themselves are entirely ext4. When you install them on a hard drive, we make an ext3 /boot and do some shuffling in anaconda. The ext4 live image is compressed inside of a squashfs which is then placed on your USB stick which should be formatted either as vfat, ext2, or ext3. I don't believe that extlinux (which we use for booting ext* USB sticks with live images on them) supports ext4 yet Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] livecd-tools 024-1
On Monday, May 11 2009, Williamson Grant said: Anyone else having issues with the latest livecd-tools 024-1. Almost every custom livecd I create it fails on fsck. Update your e2fsprogs. If you have a perfect reproducer, esandeen would love to hear it in bug 499452 I worked around this, however, could it be changed so it just throws a warning and not an exception? Having it be fatal was by the explicit request of the ext4 maintainer as there really shouldn't be errors being found. And having errors found means that there are critical bugs which need fixing Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] Changes to 'refs/tags/livecd-tools-024'
Changes since livecd-tools-023-10: --- 0 files changed --- -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 2 commits - config/livecd-fedora-minimal.ks imgcreate/live.py Makefile
Makefile|2 +- config/livecd-fedora-minimal.ks |1 + imgcreate/live.py |4 ++-- 3 files changed, 4 insertions(+), 3 deletions(-) New commits: commit 8ddeb5380c41e9a65218bbd4e92cec8210255acf Author: Jeremy Katz ka...@redhat.com Date: Wed May 6 14:02:46 2009 -0400 Bump version diff --git a/Makefile b/Makefile index c597ffc..a3b2aa9 100644 --- a/Makefile +++ b/Makefile @@ -1,5 +1,5 @@ -VERSION = 023 +VERSION = 024 INSTALL = /usr/bin/install -c INSTALL_PROGRAM = ${INSTALL} commit e0328e752e0fc5f53e8643bf02fe1aba23543ace Author: Josh Boyer jwbo...@gmail.com Date: Tue Apr 28 21:10:35 2009 -0400 Fix ppc creation The updated patch below fixes iso creation for ppc. With this I can now build an iso to completion. diff --git a/config/livecd-fedora-minimal.ks b/config/livecd-fedora-minimal.ks index 59fced6..147f4ea 100644 --- a/config/livecd-fedora-minimal.ks +++ b/config/livecd-fedora-minimal.ks @@ -11,6 +11,7 @@ repo --name=development --mirrorlist=http://mirrors.fedoraproject.org/mirrorlist %packages @core +anaconda-runtime bash kernel passwd diff --git a/imgcreate/live.py b/imgcreate/live.py index f91a5d2..a4914e0 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -606,7 +606,7 @@ hiddenmenu class ppcLiveImageCreator(LiveImageCreatorBase): def _get_mkisofs_options(self, isodir): -return [ -hfs, -nodesktop, -part +return [ -hfs, -no-desktop, -part, -map, isodir + /ppc/mapping, -hfs-bless, isodir + /ppc/mac, -hfs-volid, self.fslabel ] @@ -697,7 +697,7 @@ image=/ppc/ppc%(bit)s/vmlinuz f.write(cfg) f.close() -def __write_not_supported(isodir, bit): +def __write_not_supported(self, isodir, bit): makedirs(isodir + /ppc/ppc + bit) message = Sorry, this LiveCD does not support your hardware -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Live.py question
On Wednesday, April 29 2009, David Huff said: What is the version parameter in __copy_kernel_and_initramfs used for??? The version is to handle multiple kernels, eg, PAE vs regular. I ask because livecd-to-pxeboot looks statically for version 0, if this ever changes it looks like it would cause the script to fail. There will always be an index of 0 as the default kernel. Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] Make USB disk into full DVD installer
On Tuesday, April 28 2009, Warren Togami said: This patch allows livecd-iso-to-disk.sh to detect if the specified .iso image is an Anaconda DVD installer. If so, it copies it to the USB disk and configures syslinux to boot it. Anaconda will automatically find the DVD iso on the USB stick and install from it. Requires Fedora 11 Preview DVD iso and later. So I've been thinking about this all day and I really don't know that it makes sense. The script is explicitly for *live* images (it's in the name! :-) and making it support more than that is just going to complicate things. What happens when someone then wants it to support Ubuntu images or something else? Also, I'm hoping to stop adding functionality to the shell script at some point as liveusb-creator really should be preferred for most cases, although it's command line mode needs some work Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] Make USB disk into full DVD installer
On Wednesday, April 29 2009, Warren Togami said: On 04/29/2009 10:15 PM, Jeremy Katz wrote: On Tuesday, April 28 2009, Warren Togami said: This patch allows livecd-iso-to-disk.sh to detect if the specified .iso image is an Anaconda DVD installer. If so, it copies it to the USB disk and configures syslinux to boot it. Anaconda will automatically find the DVD iso on the USB stick and install from it. Requires Fedora 11 Preview DVD iso and later. So I've been thinking about this all day and I really don't know that it makes sense. The script is explicitly for *live* images (it's in the name! :-) and making it support more than that is just going to complicate things. What happens when someone then wants it to support Ubuntu images or something else? Also, I'm hoping to stop adding functionality to the shell script at some point as liveusb-creator really should be preferred for most cases, although it's command line mode needs some work liveusb-creator has live in its name too! Yes, but at least it's written in python so that things can be abstracted nicely rather than the horrors of shell :-) At which point, maybe it should have a rename. But I'll leave that for Luke ;-) Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |1 + 1 file changed, 1 insertion(+) New commits: commit e94db68943a0ed78cbbd72b152a2aed0c16a5e03 Author: Warren Togami wtog...@redhat.com Date: Tue Apr 28 21:32:08 2009 -0400 chattr -i extlinux.sys chattr -i extlinux.sys after it is installed, because it isn't useful to be protected, and it is annoying as it prevents easy removal. diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 4effb3e..4ed16c1 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -686,6 +686,7 @@ if [ -z $multi ]; then # and has to be run with the file system mounted mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/extlinux.conf extlinux -i $USBMNT/$SYSLINUXPATH +chattr -i $USBMNT/$SYSLINUXPATH/extlinux.sys cleanup fi else -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] Branch 'f10-branch' - livecd-tools.spec Makefile
Makefile |2 +- livecd-tools.spec | 10 +- 2 files changed, 10 insertions(+), 2 deletions(-) New commits: commit b9f5f1335a6605ef21e1be1c59cfdaeaea11575b Author: Jeremy Katz ka...@redhat.com Date: Tue Apr 21 14:09:10 2009 -0400 Bump version diff --git a/Makefile b/Makefile index 48c89d5..5c65cb8 100644 --- a/Makefile +++ b/Makefile @@ -1,5 +1,5 @@ -VERSION = 020 +VERSION = 020.1 INSTALL = /usr/bin/install -c INSTALL_PROGRAM = ${INSTALL} diff --git a/livecd-tools.spec b/livecd-tools.spec index 7ab1468..5956add 100644 --- a/livecd-tools.spec +++ b/livecd-tools.spec @@ -4,7 +4,7 @@ Summary: Tools for building live CD's Name: livecd-tools -Version: 020 +Version: 020.1 Release: 1%{?dist} License: GPLv2 Group: System Environment/Base @@ -64,6 +64,14 @@ rm -rf $RPM_BUILD_ROOT %{python_sitelib}/imgcreate/*.pyc %changelog +* Tue Apr 21 2009 Jeremy Katz ka...@redhat.com - 020.1-1 +- Fix up --base-on (#471656) +- Fix macro name for excludedocs (bkearney) +- Fix device command (apevec) +- Fix unicode errors (Felix Schwarz) +- Support syslinux in /usr/share +- Use our syslinux's menu com32 modules in livecd-iso-to-disk (#492370) + * Wed Nov 12 2008 Jeremy Katz ka...@redhat.com - 020-1 - Support setting up a swap file - Verify integer args in livecd-iso-to-disk (#467257) -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 5 commits - imgcreate/fs.py imgcreate/live.py
imgcreate/fs.py |8 +++- imgcreate/live.py |8 2 files changed, 11 insertions(+), 5 deletions(-) New commits: commit 149e1d01eef4da1064038792e46c8765d8bb0080 Author: Jeremy Katz ka...@redhat.com Date: Mon Apr 27 15:31:15 2009 -0400 More ppc syntax fixing (#497193) diff --git a/imgcreate/live.py b/imgcreate/live.py index 4b5286e..f91a5d2 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -670,7 +670,7 @@ image=/ppc/ppc%(bit)s/vmlinuz % args -def __write_yaboot_config(isodir, bit): +def __write_yaboot_config(self, isodir, bit): cfg = self.__get_basic_yaboot_config(name = self.name, timeout = self._timeout * 100) commit d218b52f16389d510c74f989e24f471f1ab56075 Author: Jeremy Katz ka...@redhat.com Date: Thu Apr 23 12:51:54 2009 -0400 Raise an error if the fsck finds problems diff --git a/imgcreate/fs.py b/imgcreate/fs.py index 5dfacec..65f16e4 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -457,7 +457,8 @@ class ExtDiskMount(DiskMount): else: bot = t -self.__fsck() +if self.__fsck() != 0: +raise CreatorError(fsck returned an error!) return top def resparse(self, size = None): commit 3141a0ca6d0588605619aded646e638305691226 Author: Jeremy Katz ka...@redhat.com Date: Thu Apr 23 12:36:58 2009 -0400 Fix up ppc live image creation (#497193) diff --git a/imgcreate/live.py b/imgcreate/live.py index 7b07d22..4b5286e 100644 --- a/imgcreate/live.py +++ b/imgcreate/live.py @@ -733,9 +733,9 @@ image=/ppc/ppc32/vmlinuz configure the boot loader havekernel = { 32: False, 64: False } -self.__copy_boot_file(mapping, isodir + /ppc) -self.__copy_boot_file(bootinfo.txt, isodir + /ppc) -self.__copy_boot_file(ofboot.b, isodir + /ppc/mac) +self.__copy_boot_file(isodir + /ppc, mapping) +self.__copy_boot_file(isodir + /ppc, bootinfo.txt) +self.__copy_boot_file(isodir + /ppc/mac, ofboot.b) shutil.copyfile(self._instroot + /usr/lib/yaboot/yaboot, isodir + /ppc/mac/yaboot) commit 786dbdd9edf41866544f6810353ed6c5208d618d Author: Jeremy Katz ka...@redhat.com Date: Thu Apr 23 12:36:44 2009 -0400 Return return code from fsck in case we want to do something with it diff --git a/imgcreate/fs.py b/imgcreate/fs.py index cc4d503..5dfacec 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -419,7 +419,8 @@ class ExtDiskMount(DiskMount): def __fsck(self): logging.debug(Checking filesystem %s % self.disk.lofile) -subprocess.call([/sbin/e2fsck, -f, -y, self.disk.lofile]) +rc = subprocess.call([/sbin/e2fsck, -f, -y, self.disk.lofile]) +return rc def __get_size_from_filesystem(self): def parse_field(output, field): commit f2719f41b9e6be0b15b533d166620a31d0730d16 Author: Jeremy Katz ka...@redhat.com Date: Thu Apr 23 12:33:45 2009 -0400 Fsck after resize to catch/fix problems (#497377) diff --git a/imgcreate/fs.py b/imgcreate/fs.py index 4a698e8..cc4d503 100644 --- a/imgcreate/fs.py +++ b/imgcreate/fs.py @@ -397,6 +397,8 @@ class ExtDiskMount(DiskMount): self.__fsck() resize2fs(self.disk.lofile, size) + +self.__fsck() return size def __create(self): @@ -453,6 +455,8 @@ class ExtDiskMount(DiskMount): top = t else: bot = t + +self.__fsck() return top def resparse(self, size = None): -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] SELinux and livecd-creator
On Friday, April 24 2009, Marc Herbert said: http://fedoraproject.org/wiki/How_to_create_and_use_a_Live_CD#Set_SELinux_to_permissive_mode SELinux should be in permissive mode for livecd-creator to work. Is this information obsolete? Or did all my past livecd-creator runs fail in a subtle and hard to notice way? We ensure that contexts are all set correctly at the end with a restorecon. The biggest problem with being in enforcing mode vs permissive is you can get a flood of AVCs SELinux being disabled only works if your live image config has SELinux disabled Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] liveinst, root and ext3
On Friday, April 24 2009, Sebastian Vahl said: Isn't it possible to install the F11 live images to an ext3 root filesystem (and only one partition)? AFAIR ext2 worked with F9 and F10 images. Nope -- ext4 has more substantial changes to the on-disk format than the ext2-ext3 move did Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] imgcreate/creator.py tools/image-creator tools/livecd-creator
On Wednesday, April 22 2009, Sebastian Vahl said: After this patch livecd-creator is broken: [snip] optparse.OptionConflictError: option -d/--debug: conflicting option string(s): -d, --debug That'll teach me to apply obvious patches without trying them Martin -- can you please send an actual *tested* version if you want the functionality here? Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] tools/livecd-iso-to-pxeboot.sh
tools/livecd-iso-to-pxeboot.sh |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 1acd7f772f1d6b4dfeb84e0aed5ee41bcd7fb908 Author: Marc Herbert marc.herb...@gmail.com Date: Mon Apr 20 12:56:41 2009 +0100 cpio -L to dereference any symbolic link to the actual ISO This is safe because the cpio archive contains only the ISO image. (Such a symbolic link is helpful to track the date-based, constantly changing ISO name). diff --git a/tools/livecd-iso-to-pxeboot.sh b/tools/livecd-iso-to-pxeboot.sh index 22810d9..45a1ffe 100755 --- a/tools/livecd-iso-to-pxeboot.sh +++ b/tools/livecd-iso-to-pxeboot.sh @@ -87,7 +87,7 @@ fi # filesystem. ISOBASENAME=`basename $ISO` ISODIRNAME=`dirname $ISO` -( cd $ISODIRNAME echo $ISOBASENAME | cpio -H newc --quiet -o ) | +( cd $ISODIRNAME echo $ISOBASENAME | cpio -H newc --quiet -L -o ) | gzip -9 | cat $CDMNT/isolinux/initrd0.img - tftpboot/initrd0.img -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] imgcreate/creator.py tools/image-creator tools/livecd-creator
imgcreate/creator.py |3 +++ tools/image-creator |5 + tools/livecd-creator |6 ++ 3 files changed, 14 insertions(+) New commits: commit 456c1682e5352a1b2277195d5c261d530a014d30 Author: Martin Langhoff mar...@laptop.org Date: Mon Apr 20 18:50:03 2009 +0200 creator: -d opt and matching setdebug() method that gets rpm in debug mode Useful to diagnose problems with %post scripts during the build. This patch adds the method to the ImageCreator class, and the corresponding options to image-creator and livecd-creator diff --git a/imgcreate/creator.py b/imgcreate/creator.py index 823afc5..a7c79c7 100644 --- a/imgcreate/creator.py +++ b/imgcreate/creator.py @@ -760,6 +760,9 @@ class ImageCreator(object): shutil.move(os.path.join(self._outdir, f), os.path.join(destdir, f)) +def setdebug(self): +rpm.setVerbosity(rpm.RPMLOG_DEBUG) + def create(self): Install, configure and package an image. diff --git a/tools/image-creator b/tools/image-creator index 6f2604c..51b2558 100755 --- a/tools/image-creator +++ b/tools/image-creator @@ -30,6 +30,8 @@ def parse_options(args): parser.add_option(-n, --name, type=string, dest=name, help=Image name and filesystem label) +parser.add_option(-d, --debug, action=store_true, dest=debug, + help=Enable debug mode) imgcreate.setup_logging(parser) @@ -61,6 +63,9 @@ def main(): creator = imgcreate.LoopImageCreator(ks, name) +if options.debug: +creator.setdebug() + try: creator.create() except imgcreate.CreatorError, e: diff --git a/tools/livecd-creator b/tools/livecd-creator index 39f7478..c534900 100755 --- a/tools/livecd-creator +++ b/tools/livecd-creator @@ -60,6 +60,9 @@ def parse_options(args): # Start a shell in the chroot for post-configuration. parser.add_option(-l, --shell, action=store_true, dest=give_shell, help=optparse.SUPPRESS_HELP) +parser.add_option(-d, --debug, action=store_true, dest=debug, + help=Enable debugging during the build) + # Don't compress the image. parser.add_option(-s, --skip-compression, action=store_true, dest=skip_compression, help=optparse.SUPPRESS_HELP) @@ -118,6 +121,9 @@ def main(): if options.cachedir: options.cachedir = os.path.abspath(options.cachedir) +if options.debug: +creator.setdebug() + try: creator.mount(options.base_on, options.cachedir) creator.install() -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] imgcreate/creator.py
imgcreate/creator.py |2 ++ 1 file changed, 2 insertions(+) New commits: commit fd252dc757a086b104decb2c1bd88345450efcf5 Author: Jeremy Katz ka...@redhat.com Date: Mon Apr 13 15:44:56 2009 -0400 Fix traceback with selinux disabled (#495269) diff --git a/imgcreate/creator.py b/imgcreate/creator.py index 498f7e6..823afc5 100644 --- a/imgcreate/creator.py +++ b/imgcreate/creator.py @@ -433,6 +433,8 @@ class ImageCreator(object): def __getbooleans(self): booleans = [] +if not kickstart.selinux_enabled(self.ks): +return booleans for i in selinux.security_get_boolean_names()[1]: on = selinux.security_get_boolean_active(i) booleans.append((/booleans/%s % i, %d %d % (on, on))) -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] className - PatitionTypeWindow
On Thursday, April 09 2009, Williamson Grant said: Did a spin of the fedora live desktop en from rawhide. After iso is created, trying to install fails to hard drive with. className - PatitionTypeWindow No module named partedUtils Not a livecd issue but, thought I would mention it here Yes, not a livecd issue and really not the right list Did the partedUtils module get removed from anaconda? But yes, and I pushed the fix to anaconda.git yesterday Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 2 commits - imgcreate/kickstart.py imgcreate/yuminst.py
imgcreate/kickstart.py |2 +- imgcreate/yuminst.py |2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) New commits: commit e08034c7738a22edaddbdddf3f8d63991cafb2b6 Author: Marc Herbert marc.herb...@gmail.com Date: Thu Apr 9 18:19:03 2009 +0100 quiet restorecon In my configuration, the attached patch is reducing the output of livecd-creator from 2000 lines to 700 lines. I suspect that these 1300 lines of output from restorecon are not very useful to anyone, are they? diff --git a/imgcreate/kickstart.py b/imgcreate/kickstart.py index 5afe8f9..83a62e4 100644 --- a/imgcreate/kickstart.py +++ b/imgcreate/kickstart.py @@ -419,7 +419,7 @@ class SelinuxConfig(KickstartConfig): if not os.path.exists(self.path(/sbin/restorecon)): return -self.call([/sbin/restorecon, -l, -v, -r, -F, -e, /proc, -e, /sys, -e, /dev, -e, /selinux, /]) +self.call([/sbin/restorecon, -l, -r, -F, -e, /proc, -e, /sys, -e, /dev, -e, /selinux, /]) def apply(self, ksselinux): if os.path.exists(self.path(/usr/sbin/lokkit)): commit 727e6a818b8e542ee484021febea6cb9db8adaff Author: Jesse Keating jkeat...@redhat.com Date: Thu Apr 9 09:22:23 2009 -0700 Add options to force fetching of repomd.xml every run. This will make repomd.xml timeout every run and force yum to fetch a new one. However if it matches the rest of the repodata no other files will be downloaded. This helps when using the same cachedir for 32bit and 64bit runs. Also expire the mirrorlist every run to ensure we get proper mirrors. diff --git a/imgcreate/yuminst.py b/imgcreate/yuminst.py index 080034b..d3f36ac 100644 --- a/imgcreate/yuminst.py +++ b/imgcreate/yuminst.py @@ -144,6 +144,8 @@ class LiveCDYum(yum.YumBase): repo.basecachedir = self.conf.cachedir repo.failovermethod = priority repo.metadata_expire = 0 +repo.mirrorlist_expire = 0 +repo.timestamp_check = 0 # disable gpg check??? repo.gpgcheck = 0 repo.enable() -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] Add options to force fetching of repomd.xml every run.
On Thursday, April 09 2009, Jesse Keating said: This will make repomd.xml timeout every run and force yum to fetch a new one. However if it matches the rest of the repodata no other files will be downloaded. This helps when using the same cachedir for 32bit and 64bit runs. Also expire the mirrorlist every run to ensure we get proper mirrors. Thanks, applied Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] quiet restorecon
On Thursday, April 09 2009, Marc Herbert said: In my configuration, the attached patch is reducing the output of livecd-creator from 2000 lines to 700 lines. I suspect that these 1300 lines of output from restorecon are not very useful to anyone, are they? Seems reasonable enough, applied Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Report On F-11 Beta With Orca
On Tuesday, April 07 2009, Janina Sajka said: Jeremy Katz writes: On Monday, April 06 2009, Janina Sajka said: * My attempt to install F-11 from the Live desktop icon failed. * All I could glean is that Orca had the installation screen * listed as inaccessible. This is on an x86_64 box using x86_64 * Live. It may not be there in i386. I wqould like to help debug this, but not sure how to proceed. The problem here (unfortunately) is that we have to run the install as root. But the session isn't running as root and hence you don't get any of the a11y stuff. While an argument could be made that the work could be done to separate out the frontend and backend of the install and do everything over dbus or some other ipc, the root of the matter is that there are too many other things in the installer pile that are likely to always end up taking priority. So the question is how can you get an app running as root from consolehelper to work with the a11y technologies and then I can try to push those tweaks as needed OK. I'll try and help chase down the latest and greatest advice on this. For starters: * I believe Ubuntu Live is doing it. I'll get their image and * check. * Fedora's GUI Virt-Manager (System Tools) is doing it * beautifully. While I'm still having problems learning to set up * virtual machines, it's not for a11y reasons as this tool is * working as it should--perhaps 95-98% of the screens. So, * whatever virt-manager is doing in the gui, is the way to go. I, * as the ordinary user, am prompted once (and accessibly) for * root's passwd, and that's that. virt-manager runs all of the UI as the user and then just communicates with the local libvirt daemon for privileged bits. That's the sort of split that I don't really see making the priority list of the anaconda team * There's a page, not recently updated, on the Orca wiki on this * topic: http://live.gnome.org/Orca/SysAdmin This looks like it might do the trick although some testing is almost certainly needed Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Report On F-11 Beta With Orca
On Tuesday, April 07 2009, Janina Sajka said: Jeremy Katz writes: On Monday, April 06 2009, Janina Sajka said: Problem Areas * It would be preferable to use gnome-speech-espeak as the default * Orca driver for several reasons: -- Espeak is more responsive and that's a big deal with a screen reader. Low latency, quick shutup, etc., are far more important to screen reading than high quality speech synthesis. -- Festival is a particularly bad choice in the shutup department. It will keep speaking until it has finished the string it was given to say, whatever the user does. This makes for very sluggish handling. The user must wait for the computer to finish speaking, with no ability to stop speech and move on. This has things the wrong way round as computers should wait on people, not people on computers. How do they compare size-wise? In any case, file an RFE against the gnome-speech package as it looks like that's where the change would need to be. I'm so glad you asked! big grin Looking just at the rpm sizes, espeak is about 1.3M and gnome-speech-espeak is about 16K. This compares to about 13.8M for the three Festival packages--so a significantly smaller footprint. Is Festival present only for Orca? Or for other applications in the Live image as well? I believe just Orca Also, are you suggesting filing the RFE with the upstream GNOME project? Or for the Fedora Live product itself? File it in Fedora bugzilla against gnome-speech. The maintainer may then want something upstream, but it's the right place to start Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Report On F-11 Beta With Orca
On Monday, April 06 2009, Janina Sajka said: Problem Areas * It would be preferable to use gnome-speech-espeak as the default * Orca driver for several reasons: -- Espeak is more responsive and that's a big deal with a screen reader. Low latency, quick shutup, etc., are far more important to screen reading than high quality speech synthesis. -- Festival is a particularly bad choice in the shutup department. It will keep speaking until it has finished the string it was given to say, whatever the user does. This makes for very sluggish handling. The user must wait for the computer to finish speaking, with no ability to stop speech and move on. This has things the wrong way round as computers should wait on people, not people on computers. How do they compare size-wise? In any case, file an RFE against the gnome-speech package as it looks like that's where the change would need to be. * My attempt to install F-11 from the Live desktop icon failed. * All I could glean is that Orca had the installation screen * listed as inaccessible. This is on an x86_64 box using x86_64 * Live. It may not be there in i386. I wqould like to help debug this, but not sure how to proceed. The problem here (unfortunately) is that we have to run the install as root. But the session isn't running as root and hence you don't get any of the a11y stuff. While an argument could be made that the work could be done to separate out the frontend and backend of the install and do everything over dbus or some other ipc, the root of the matter is that there are too many other things in the installer pile that are likely to always end up taking priority. So the question is how can you get an app running as root from consolehelper to work with the a11y technologies and then I can try to push those tweaks as needed Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Using fedora 10 to spin fedora 11 live
On Friday, April 03 2009, Williamson Grant said: is possible to use a fedora 10 based machine to spin a fedora 11 beta livecd. I rebuilt the livecd tools from fedora 11 on fedora 10, so that /usr/ share/syslinux is handled, and it appears to generate an iso, however when booting, it complains there is a squashfs mismatch. If you update to rawhide squashfs-tools as well, that will help. I'm not sure if the F10 kernel has all of the ext4 fixes needed not to deadlock, though. At least for the problems of, eg, squashfs-tools, there's some work underway to get livecd-creator running under mock which will long-term help Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] vmware fusion and f11 live
On Friday, April 03 2009, Williamson Grant said: Anyone have any ideas why livecds created in Fedora 11 beta just will not boot in vmware. The same iso boots in sun virtual box and on a real machine. This is as far as it gets during the boot process. I'm guessing kernel problem -- try removing quiet and rhgb from the kernel command line and you might see a more useful error. Also, try turning on (or off if it's on) VMI support for your guest Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 2 commits - imgcreate/kickstart.py tools/livecd-iso-to-disk.sh
imgcreate/kickstart.py | 18 +++--- tools/livecd-iso-to-disk.sh | 21 + 2 files changed, 28 insertions(+), 11 deletions(-) New commits: commit d9880bf3823b21421f4916d64648a990a7392ecb Author: Jeremy Katz ka...@redhat.com Date: Tue Mar 31 14:12:00 2009 -0400 Support the rest of the firewall args for kickstart configs Support trusteddevs, services and ports in the firewall line diff --git a/imgcreate/kickstart.py b/imgcreate/kickstart.py index 253d349..5afe8f9 100644 --- a/imgcreate/kickstart.py +++ b/imgcreate/kickstart.py @@ -172,18 +172,22 @@ class AuthConfig(KickstartConfig): class FirewallConfig(KickstartConfig): A class to apply a kickstart firewall configuration to a system. def apply(self, ksfirewall): -# -# FIXME: should handle the rest of the options -# if not os.path.exists(self.path(/usr/sbin/lokkit)): return +args = [/usr/sbin/lokkit, -f, --quiet, --nostart] if ksfirewall.enabled: -status = --enabled +args.append(--enabled) + +for port in ksfirewall.ports: +args.append(--port=%s %(port,)) +for svc in ksfirewall.services: +args.append(--service=%s %(svc,)) +for dev in ksfirewall.trusts: +args.append(--trust=%s %(dev,)) else: -status = --disabled +args.append(--disabled) -self.call([/usr/sbin/lokkit, - -f, --quiet, --nostart, status]) +self.call(args) class RootPasswordConfig(KickstartConfig): A class to apply a kickstart root password configuration to a system. commit b9e8cc6bf67ed6e4202453ffb6ae6a7149a100ab Author: Alexander Boström a...@stacken.kth.se Date: Fri Mar 27 22:00:20 2009 +0100 Handle the case when the kernel can't mount squashfs.img (probably a mksquashfs/running kernel mismatch). diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 96869c7..daadc3d 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -191,6 +191,11 @@ checkGPT() { partinfo=$(/sbin/parted --script -m $device print |grep ^$partnum:) volname=$(echo $partinfo |cut -d : -f 6) flags=$(echo $partinfo |cut -d : -f 7) +if [ $volname != EFI System Partition ]; then + echo Partition name must be 'EFI System Partition' + echo This can be set in parted or you can run with --reset-mbr + exitclean +fi if [ $(echo $flags |grep -c boot) = 0 ]; then echo Partition isn't marked bootable! echo You can mark the partition as bootable with @@ -457,9 +462,17 @@ else fi livesize=$(du -s -B 1M $check | awk {'print $1;'}) if [ -n $skipcompress ]; then -mount -o loop $CDMNT/LiveOS/squashfs.img $CDMNT -livesize=$(du -s -B 1M $CDMNT/LiveOS/ext3fs.img | awk {'print $1;'}) -umount $CDMNT +if [ -e $CDMNT/LiveOS/squashfs.img ]; then + if mount -o loop $CDMNT/LiveOS/squashfs.img $CDMNT; then + livesize=$(du -s -B 1M $CDMNT/LiveOS/ext3fs.img | awk {'print $1;'}) + umount $CDMNT + else + echo WARNING: --skipcompress or --xo was specified but the currently + echo running kernel can not mount the squashfs from the ISO file to extract + echo it. The compressed squashfs will be copied to the USB stick. + skipcompress= + fi +fi fi free=$(df -B1M $USBDEV |tail -n 1 |awk {'print $4;'}) @@ -500,7 +513,7 @@ if [ -z $skipcopy ];then [ ! -d $USBMNT/$LIVEOS ] mkdir $USBMNT/$LIVEOS [ -n $keephome -a -f $USBMNT/$HOMEFILE ] mv $USBMNT/$HOMEFILE $USBMNT/$LIVEOS/$HOMEFILE if [ -n $skipcompress -a -f $CDMNT/LiveOS/squashfs.img ]; then - mount -o loop $CDMNT/LiveOS/squashfs.img $CDMNT + mount -o loop $CDMNT/LiveOS/squashfs.img $CDMNT || exitclean cp $CDMNT/LiveOS/ext3fs.img $USBMNT/$LIVEOS/ext3fs.img || (umount $CDMNT ; exitclean) umount $CDMNT elif [ -f $CDMNT/LiveOS/squashfs.img ]; then -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 80dddee3c642848e36a4ae19a27f0e1734e16489 Author: Jeremy Katz ka...@redhat.com Date: Mon Mar 30 11:44:12 2009 -0400 Fix unterminated sed command (#492376) diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 8702936..96869c7 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -596,7 +596,7 @@ if [ -n $xo ]; then args=$args persistenthome=mtd0 fi args=$args reset_overlay -xosyspath=$(echo $SYSLINUXPATH | sed -e 's;/;\;') +xosyspath=$(echo $SYSLINUXPATH | sed -e 's;/;\\;') if [ ! -d $USBMNT/boot ]; then mkdir -p $USBMNT/boot ; fi cat $USBMNT/boot/olpc.fth EOF \ Boot script for USB boot -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: bind-mounting files in mock
On Friday, March 27 2009, Jesse Keating said: On Fri, 2009-03-27 at 10:37 -0500, Clark Williams wrote: It's not described in the man page, so I'm a bit leery about depending on that behavior. That being said, it's not a normal use-case for mock builds, so it'll probably only bite the livecd creation stuff it the bind-mount behavior changes. Ok, I'll pull it in for the next release. I need that functionality too when building install images. Rather than see more and different code paths when chroot generating, I think there would be some value in using the same code path regardless of how the chroot is used. I think the bind mounting of dev/ is only there for the loop entries, I don't know of any other mock consumers that require a real /dev/ tree. We might consider just always file bind mounting a few loop entries in and only having one code path. You can't really just do a few as you may need however many there are on the real system. If the idea is to make them by default it's probably better to just make n of them based on the host. But then the chroot does have a way of affecting outside of it :/ Jeremy -- Fedora-buildsys-list mailing list Fedora-buildsys-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-buildsys-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |5 - 1 file changed, 5 deletions(-) New commits: commit 3c14c6c41b8f541bbfa3095d6b23e39ed7f91a4f Author: Jim Radford radf...@blackbean.org Date: Thu Mar 26 09:51:00 2009 -0700 Only the UUID of the EFI System Partition is important, not its name EFI systems should boot fine regardless of the name of the EFI System Partition as it's defined by its UUID, not its label, which the boot flag check already takes care of. I verfied this on a MacBook. The check for the boot flag should probably check for :boot; to not just boot to keep from accidentally picking up the label. -Jim diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 05fe080..8702936 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -191,11 +191,6 @@ checkGPT() { partinfo=$(/sbin/parted --script -m $device print |grep ^$partnum:) volname=$(echo $partinfo |cut -d : -f 6) flags=$(echo $partinfo |cut -d : -f 7) -if [ $volname != EFI System Partition ]; then - echo Partition name must be 'EFI System Partition' - echo This can be set in parted or you can run with --reset-mbr - exitclean -fi if [ $(echo $flags |grep -c boot) = 0 ]; then echo Partition isn't marked bootable! echo You can mark the partition as bootable with -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] Re: [PATCH] Only the UUID of the EFI System Partition is important, not its name
On Thursday, March 26 2009, Jim Radford said: EFI systems should boot fine regardless of the name of the EFI System Partition as it's defined by its UUID, not its label, which the boot flag check already takes care of. I verfied this on a MacBook. Thanks, applied Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] The variable xosyspath escaped into olpc.fth. Expand it to the proper path instead.
Thanks for the patch; applied Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) New commits: commit a54b23dd751bd4b52154a3df011f4aa1f75a4c09 Author: Alexander Boström a...@stacken.kth.se Date: Tue Mar 24 23:50:36 2009 +0100 The variable xosyspath escaped into olpc.fth. Expand it to the proper path instead. diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index f0c7331..05fe080 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -635,8 +635,8 @@ hex rom-pa fffc7 + 4 \$number drop h# 2e19 [if] set-bootpath-dev $args to boot-file - \${BOOTPATHDEV}\$xosyspath\initrd0.img expand$ to ramdisk - \${BOOTPATHDEV}\$xosyspath\vmlinuz0 expand$ to boot-device + \${BOOTPATHDEV}$xosyspath\initrd0.img expand$ to ramdisk + \${BOOTPATHDEV}$xosyspath\vmlinuz0 expand$ to boot-device unfreeze boot EOF -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 2 commits - tools/livecd-iso-to-disk.sh tools/livecd-iso-to-pxeboot.sh
tools/livecd-iso-to-disk.sh|2 ++ tools/livecd-iso-to-pxeboot.sh |8 +--- 2 files changed, 7 insertions(+), 3 deletions(-) New commits: commit b6c69841c1e372173b8633cbd9ede0baa7847aba Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 20 10:46:49 2009 -0400 Handle changed syslinux paths for livecd-iso-to-pxeboot as well diff --git a/tools/livecd-iso-to-pxeboot.sh b/tools/livecd-iso-to-pxeboot.sh index 280d354..22810d9 100755 --- a/tools/livecd-iso-to-pxeboot.sh +++ b/tools/livecd-iso-to-pxeboot.sh @@ -42,8 +42,8 @@ if [ $(id -u) != 0 ]; then fi # Check pxelinux.0 exists. -if [ ! -f /usr/lib/syslinux/pxelinux.0 ]; then -echo Warning: /usr/lib/syslinux/pxelinux.0 not found. +if [ ! -f /usr/share/syslinux/pxelinux.0 -a ! -f /usr/lib/syslinux/pxelinux.0 ]; then +echo Warning: pxelinux.0 not found. echo Make sure syslinux or pxelinux is installed on this system. fi @@ -95,7 +95,9 @@ ISODIRNAME=`dirname $ISO` cp $CDMNT/isolinux/vmlinuz0 tftpboot/vmlinuz0 # pxelinux bootloader. -if [ -f /usr/lib/syslinux/pxelinux.0 ]; then +if [ -f /usr/share/syslinux/pxelinux.0 ]; then +cp /usr/share/syslinux/pxelinux.0 tftpboot +elif [ -f /usr/lib/syslinux/pxelinux.0 ]; then cp /usr/lib/syslinux/pxelinux.0 tftpboot else echo Warning: You need to add pxelinux.0 to tftpboot/ subdirectory commit d2c2a0c911090cb55e56e207fba0fa472913e5dd Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 20 10:44:09 2009 -0400 Don't prompt about overwriting files from mtools (#491234) New syslinux uses mtools for the installer and then mmove prompts about overwriting the existing ldlinux.sys. So let's just remove it as there's not a good way to force mmove to not ask diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index f840c6e..f0c7331 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -654,6 +654,8 @@ if [ -z $multi ]; then # syslinux expects the config to be named syslinux.cfg # and has to run with the file system unmounted mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/syslinux.cfg +# deal with mtools complaining about ldlinux.sys +if [ -f $USBMNT/$SYSLINUXPATH/ldlinux.sys ] ; then rm -f $USBMNT/$SYSLINUXPATH/ldlinux.sys ; fi cleanup if [ -n $SYSLINUXPATH ]; then syslinux -d $SYSLINUXPATH $USBDEV -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Using livecd-tools for recovery disc
On Wednesday, March 18 2009, Rhonda Kern said: #1 - I would like for the install to complete *without* prompting with the reboot message. I do not have a reboot instruction in my kickstart file. Still, it prompts with a reboot message. I press enter and the script continues. But I'd rather it just continue without the prompting. Does anyone know how to make that happen? reboot in your kickstart config will exit the installer automatically; otherwise, it prompts #2 - I need to be able to remove the live DVD and place another disc in the drive to un-tar some files. If I try to unmount the DVD right after the machine is booted, before the script is run, it works fine. However, if I try to unmount the DVD after the script has run, I get errors like: Ext3-fs error (device dm-0): ext3_journal_start_sb: detected aborted journal Remounting filesystem read-only Also, I am un-able to run any commands. It says it can't find /bin/find, or whatever command I'm trying to run. Does anyone know if it's possible to remove the live disc? If so, how is it done? Or does anyone know if it's possible in F10 or F11? You can't (generally) do this as the system is running off of the livecd. So removing it is removing the disc of your rootfs. If you have enough ram, you can add 'live_ram' to your kernel command line and instead the image will be copied and run out of ram Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] LiveUSB overlay size is smaller than expected
On Tuesday, March 17 2009, S'orlok Reaves said: I'm using the LiveUSB feature of the LiveCDs, with the overlay feature to persist data. I have to say, I'm really enjoying the portability. However, I'm unable to install all the packages I would like to, due to a mysterious fact: the overlay file is way smaller than I created it to be. [snip] Running df -h on a clean boot of the usb shows me: /dev/mapper/live-rw 741M (size) 715M(used) 20M(avail) (mounted as /) and several tmpfs's I'm using for /var/run and yum's cache While the overlay is larger than this, the filesystem size may be smaller. If you look at the kickstart config you used to create your image, you probably have a rootfs size specified to be on the smaller size Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] a boot option to remove the persistent layer and create a new one
On Friday, March 13 2009, said: if one used more than the size of the persistent layer, system becomes corrupted and unusable one then can remove the layer to restore it to factory defaults but telling people to create a new layer is a but difficult for end users If you boot with 'reset_overlay', it resets the overlay. [snip] resetPersistentHome and resetOverlay should simply check the size of the img file we want to rest then remove it then create a new one with the same size Resetting the persistent home seems a little more questionable. Since it's just a filesystem image and not the weird dm-snapshot, it's a less straight-forward call of how/when it could get corrupted and then need any sort of resetting Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Low overlay/home dir space
On Monday, March 16 2009, Williamson Grant said: Has anyone found a workaround for low disk space on the livecd. I know you can repair the luks home directory but I want to try and prevent this. Has anyone looked in to gnome-volume-manager or quota(or something else) to prevent the root and/or home dir from becoming full? There's really not a good way to do it with the overlay -- since we're using dm-snapshot, the changes are all block based and so things like quota or even shrinking the filesystem won't work. The only real solution is to be able to use something more robust than dm-snapshot, but then we get into the land of not-upstreamed-kernel-bits which are basically a non-starter. There's a new effort underway to hopefully make an upstreamable unionfs-like thing which could be interesting in the future. For persistent /home, though, it's a real filesystem and thus you get the usual filesystem limits to writing to avoid low disk space problems (including the default 5% of reserved space for root) Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: Change request - Using git send-email
On Thursday, March 12 2009, Jesse Keating said: Here is a message from git send-email. To get here, I used: $ git commit -a Created commit cb85f54: Disable rawhide. 1 files changed, 2 insertions(+), 1 deletions(-) $ git format-patch HEAD^ 0001-Disable-rawhide.patch $ git send-email --compose --to Fedora-infrastructure-list@redhat.com 0001-Disable-rawhide.patch Note that if it's a single patch, --compose is probably overkill as it makes two messages rather than one. Your commit message is self-explanatory, no? (... and if not, edit the 0001 file before running git send-email) Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
Re: change request
On Tuesday, March 10 2009, Mike McGrath said: I need to install transifex-extras on app1 (same deal as before) Low risk. It does pull in bzr, bzrtools pysvn, python-paramiko. 2 +1's? +1 Jeremy ___ Fedora-infrastructure-list mailing list Fedora-infrastructure-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-infrastructure-list
[Fedora-livecd-list] Changes to 'refs/tags/livecd-tools-022'
Changes since livecd-tools-020-40: --- 0 files changed --- -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] Error creating livecd: Cannot unmount, device is busy
On Sunday, March 08 2009, Joshua C. said: My last 3 attempts to create a livecd from rawhide resulted in errror (see the attached file). There is some problem with the umount that cannot unmount the device. The installation of the packages and relabeling with selinux works fine but then come the errors like umount: /mnt/disk/tmp/imgcreate-WLYOXG/install_root: device is busy. (In some cases useful info about processes that use the device is found by lsof(8) or fuser(1)) Before this everything was fine. The only difference between this live image and the others is that I added the java RE. without the java packages everything is fine. It sounds like the java packages start a process in their %post that isn't killed, which would be a bug in those packages. Which packages in particular? Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] livecd-creator from git failure, Could not remove snapshot device
On Thursday, March 05 2009, Warren Togami said: Jeremy Katz wrote: On Wednesday, March 04 2009, Warren Togami said: Warren Togami wrote: Pass 5: Checking group summary information fedora-livecd--i: 78136/139264 files (1.7% non-contiguous), 542199/542199 blocks device-mapper: remove ioctl failed: Device or resource busy Command failed /usr/lib/python2.6/site-packages/imgcreate/errors.py:45: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6 return unicode(self.message) Error creating Live CD : Could not remove snapshot device I figured it out. A week or two ago, GNOME began mounting things that it shouldn't (a separate bug). In this case GNOME is mounting the device mapper device, interfering with LiveCD creator when it tries to destroy the device. What GNOME package would this bug be? Probably DeviceKit-disks DavidZ says we need to run livecd-creator wrapped from devkit-disks: devkit-disks --inhibit -- /bin/foo As I just noted in the bug, this then introduces new sets of problems instead 1) Running a livecd build breaks mounting usb sticks or other things that you plug in 2) Running livecd-creator under some other command means that we lose some of the care that we have around keyboard interrupting and having that sanely clean up So I really don't think that's the right solution to the problem as instead it's just papering over the fact that devkit-disks thinks it owns the world :/ Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] livecd-iso-to-disk giving wrong directory path to extlinux
On Wednesday, March 04 2009, Warren Togami said: [PATCH] livecd-iso-to-disk giving wrong directory path to extlinux This is actually a result of something I was trying out for multiple images on a key that I didn't realize I had pushed. It didn't work out, so I'm just going to revert the commit in question instead (and push the revert :) Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
Re: [Fedora-livecd-list] [PATCH] livecd-iso-to-disk giving wrong directory path to extlinux
On Thursday, March 05 2009, Warren Togami said: Warren Togami wrote: There is a similar problem with vfat filesystems, where syslinux is a path that does not exist, and mmove subsequently segfaults because it doesn't know how to handle it. I'm working on that issue separately. The previous patch to make the extlinux code-path work is confirmed to work. The syslinux code path however is proving to be trickier. It appears that livecd-iso-to-disk is doing something wrong, then mmove via syslinux crashes. mtools itself is buggy in crashing here, but it appears to be livecd-iso-to-disk's fault. I am still figuring this... With the path changes reverted (and thus, basically exactly what shipped in F10), I still hit this. Going back to F10's mtools works fine. And trying a newer version from upstream (4.0.9) seems to fix things also, so I've bounced this over to the mtools maintainer to do an update Jeremy -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 0d1394b0f6134d7b476e6f39a5d24afcd31d28c1 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 11:39:01 2009 -0500 Revert Move syslinux path to be under LiveOS to prep for multiple livedirs This reverts commit 75201f2180bb2e35cc439f9219aaf333d9577daa as it seems to introduce some odd cases diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index 7d5ba1d..c310936 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -246,7 +246,7 @@ checkSyslinuxVersion() { if ! syslinux 21 | grep -qe -d; then SYSLINUXPATH= else - SYSLINUXPATH=LiveOS/syslinux + SYSLINUXPATH=syslinux fi } -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 3 commits - tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh | 12 1 file changed, 4 insertions(+), 8 deletions(-) New commits: commit 33c94391bdc8cfcc38ec65e5c9d78f06b5e1fa81 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 14:38:52 2009 -0500 Handle varying syslinux dirs for --xo mode also diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index ea6cb58..7ab1d86 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -612,6 +612,7 @@ if [ -n $xo ]; then args=$args persistenthome=mtd0 fi args=$args reset_overlay +xosyspath=$(echo $SYSLINUXPATH | sed -e 's;/;\;') if [ ! -d $USBMNT/boot ]; then mkdir -p $USBMNT/boot ; fi cat $USBMNT/boot/olpc.fth EOF \ Boot script for USB boot @@ -645,8 +646,8 @@ hex rom-pa fffc7 + 4 \$number drop h# 2e19 [if] set-bootpath-dev $args to boot-file - \${BOOTPATHDEV}\syslinux\initrd0.img expand$ to ramdisk - \${BOOTPATHDEV}\syslinux\vmlinuz0 expand$ to boot-device + \${BOOTPATHDEV}\$xosyspath\initrd0.img expand$ to ramdisk + \${BOOTPATHDEV}\$xosyspath\vmlinuz0 expand$ to boot-device unfreeze boot EOF commit 0066fc2223ba4990d673e621914bc4647c77ecf3 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 14:21:42 2009 -0500 Fix up syslinuxpath for extlinux Inspired by something Warren sent, but we should always be listening to $SYSLINUXPATH for extlinux too diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index d7aa536..ea6cb58 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -673,7 +673,7 @@ elif [ $USBFS = ext2 -o $USBFS = ext3 ]; then # extlinux expects the config to be named extlinux.conf # and has to be run with the file system mounted mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/extlinux.conf -extlinux -i $USBMNT/syslinux +extlinux -i $USBMNT/$SYSLINUXPATH cleanup fi commit b2ee33d0d9b7a1ef17a8ca30eb137c8867e2ba34 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 14:21:11 2009 -0500 Remove legacy cases without LiveOS dir diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index c310936..d7aa536 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -488,19 +488,14 @@ if [ -z $skipcopy ];then echo Copying live image to USB stick [ ! -d $USBMNT/LiveOS ] mkdir $USBMNT/LiveOS [ -n $keephome -a -f $USBMNT/$HOMEFILE ] mv $USBMNT/$HOMEFILE $USBMNT/LiveOS/$HOMEFILE - # cases without /LiveOS are legacy detection, remove for F10 if [ -n $skipcompress -a -f $CDMNT/LiveOS/squashfs.img ]; then mount -o loop $CDMNT/LiveOS/squashfs.img $CDMNT cp $CDMNT/LiveOS/ext3fs.img $USBMNT/LiveOS/ext3fs.img || (umount $CDMNT ; exitclean) umount $CDMNT elif [ -f $CDMNT/LiveOS/squashfs.img ]; then cp $CDMNT/LiveOS/squashfs.img $USBMNT/LiveOS/squashfs.img || exitclean - elif [ -f $CDMNT/squashfs.img ]; then - cp $CDMNT/squashfs.img $USBMNT/LiveOS/squashfs.img || exitclean elif [ -f $CDMNT/LiveOS/ext3fs.img ]; then cp $CDMNT/LiveOS/ext3fs.img $USBMNT/LiveOS/ext3fs.img || exitclean - elif [ -f $CDMNT/ext3fs.img ]; then - cp $CDMNT/ext3fs.img $USBMNT/LiveOS/ext3fs.img || exitclean fi if [ -f $CDMNT/LiveOS/osmin.img ]; then cp $CDMNT/LiveOS/osmin.img $USBMNT/LiveOS/osmin.img || exitclean -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list
[Fedora-livecd-list] 4 commits - tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh | 132 +--- 1 file changed, 63 insertions(+), 69 deletions(-) New commits: commit 3db0a1cc2cbd218673cb9174e83fcfb059c1a59f Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 15:58:27 2009 -0500 Basic support for multi-image mode Multi-image mode doesn't install a boot loader, so you have to set it up by hand for now which is left as an exercise for the reader diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index def083b..c803bc7 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -245,6 +245,8 @@ checkSyslinuxVersion() { fi if ! syslinux 21 | grep -qe -d; then SYSLINUXPATH= +elif [ -n $multi ]; then + SYSLINUXPATH=$LIVEOS/syslinux else SYSLINUXPATH=syslinux fi @@ -346,7 +348,11 @@ while [ $# -gt 2 ]; do LIVEOS=$2 shift ;; + --multi) + multi=1 + ;; *) + echo invalid arg -- $1 usage ;; esac @@ -637,28 +643,33 @@ EOF fi -echo Installing boot loader -if [ -n $efi ]; then +if [ -z $multi ]; then + echo Installing boot loader + if [ -n $efi ]; then # replace the ia32 hack if [ -f $USBMNT/EFI/boot/boot.conf ]; then cp -f $USBMNT/EFI/boot/bootia32.conf $USBMNT/EFI/boot/boot.conf ; fi -fi + fi -if [ $USBFS = vfat -o $USBFS = msdos ]; then + if [ $USBFS = vfat -o $USBFS = msdos ]; then # syslinux expects the config to be named syslinux.cfg # and has to run with the file system unmounted mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/syslinux.cfg cleanup if [ -n $SYSLINUXPATH ]; then - syslinux -d $SYSLINUXPATH $USBDEV + syslinux -d $SYSLINUXPATH $USBDEV else - syslinux $USBDEV + syslinux $USBDEV fi -elif [ $USBFS = ext2 -o $USBFS = ext3 ]; then + elif [ $USBFS = ext2 -o $USBFS = ext3 ]; then # extlinux expects the config to be named extlinux.conf # and has to be run with the file system mounted mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/extlinux.conf extlinux -i $USBMNT/$SYSLINUXPATH cleanup + fi +else + cleanup + echo Multi-image mode selected; not setting up bootloader fi echo USB stick set up as live image! commit 96e874e2893f2f73348d19037a08b51ee72651d2 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 14:56:12 2009 -0500 More generally, this is EFI support, not just Intel Macs diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index b7677f0..def083b 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -68,8 +68,8 @@ resetMBR() { return fi getdisk $1 -# if mactel, we need to use the hybrid MBR -if [ -n $mactel ];then +# if efi, we need to use the hybrid MBR +if [ -n $efi ];then if [ -f /usr/lib/syslinux/gptmbr.bin ]; then gptmbr='/usr/lib/syslinux/gptmbr.bin' elif [ -f /usr/share/syslinux/gptmbr.bin ]; then @@ -313,8 +313,8 @@ while [ $# -gt 2 ]; do --reset-mbr|--resetmbr) resetmbr=1 ;; - --mactel) - mactel=1 + --efi|--mactel) + efi=1 ;; --format) format=1 @@ -384,19 +384,19 @@ fi checkMounted $USBDEV if [ -n $format ];then # checks for a valid filesystem - if [ -n $mactel ];then + if [ -n $efi ];then createGPTLayout $USBDEV else createMSDOSLayout $USBDEV fi fi checkFilesystem $USBDEV -if [ -n $mactel ]; then +if [ -n $efi ]; then checkGPT $USBDEV fi checkSyslinuxVersion # Because we can't set boot flag for EFI Protective on msdos partition tables -[ -z $mactel ] checkPartActive $USBDEV +[ -z $efi ] checkPartActive $USBDEV [ -n $resetmbr ] resetMBR $USBDEV checkMBR $USBDEV @@ -436,7 +436,7 @@ if [ -f $USBMNT/$LIVEOS/$HOMEFILE -a -n $keephome -a $homesizemb -gt 0 ]; exitclean fi -if [ -n $mactel -a ! -d $CDMNT/EFI/boot ]; then +if [ -n $efi -a ! -d $CDMNT/EFI/boot ]; then echo ERROR: This live image does not support EFI booting exitclean fi @@ -492,7 +492,7 @@ fi # Bootloader is always reconfigured, so keep these out of the if skipcopy stuff. [ ! -d $USBMNT/$SYSLINUXPATH ] mkdir -p $USBMNT/$SYSLINUXPATH -[ -n $mactel -a ! -d $USBMNT/EFI/boot ] mkdir -p $USBMNT/EFI/boot +[ -n $efi -a ! -d $USBMNT/EFI/boot ] mkdir -p $USBMNT/EFI/boot if [ -z $skipcopy ];then echo Copying live image to USB stick @@ -516,7 +516,7 @@ cp $CDMNT/isolinux/* $USBMNT/$SYSLINUXPATH BOOTCONFIG=$USBMNT/$SYSLINUXPATH/isolinux.cfg # Set this to nothing so sed doesn't care BOOTCONFIG_EFI= -if [ -n $mactel ];then +if [ -n $efi ];then cp $CDMNT/EFI/boot/* $USBMNT/EFI/boot # this is a little ugly, but it gets the interesting named config file @@ -638,7 +638,7 @@ EOF fi echo Installing boot loader -if [ -n $mactel
[Fedora-livecd-list] tools/livecd-iso-to-disk.sh
tools/livecd-iso-to-disk.sh |5 - 1 file changed, 4 insertions(+), 1 deletion(-) New commits: commit 3ea5e21f5d49ce967d6dc71e1d543e42a0eff606 Author: Jeremy Katz ka...@redhat.com Date: Fri Mar 6 17:01:42 2009 -0500 Tweak kernel/initrd paths for the multi-image case When you load the extra syslinux.cfg, we need to be sure we reference the kernel and initrd with full paths so the right ones are loaded. Go ahead and set that up. For those who want to try, add a section like the following to your top-level syslinux/syslinux.cfg label otherimage menu label Other Image kernel vesamenu.c32 APPEND /livedir/syslinux/syslinux.cfg diff --git a/tools/livecd-iso-to-disk.sh b/tools/livecd-iso-to-disk.sh index c803bc7..f840c6e 100755 --- a/tools/livecd-iso-to-disk.sh +++ b/tools/livecd-iso-to-disk.sh @@ -668,8 +668,11 @@ if [ -z $multi ]; then cleanup fi else + # we need to do some more config file tweaks for multi-image mode + sed -i -e s;kernel vm;kernel $LIVEOS/syslinux/vm; $USBMNT/$SYSLINUXPATH/isolinux.cfg + sed -i -e s;initrd i;initrd $LIVEOS/syslinux/i; $USBMNT/$SYSLINUXPATH/isolinux.cfg + mv $USBMNT/$SYSLINUXPATH/isolinux.cfg $USBMNT/$SYSLINUXPATH/syslinux.cfg cleanup - echo Multi-image mode selected; not setting up bootloader fi echo USB stick set up as live image! -- Fedora-livecd-list mailing list Fedora-livecd-list@redhat.com https://www.redhat.com/mailman/listinfo/fedora-livecd-list