April 2015 - anaconda-patches - Fedora mailing-lists

New: [rhinstaller/lorax/pulls/16 master] Make 64-bit kernel on 32-bit firmware work for x86 efi machines
by dashea 14 Jan '16

14 Jan '16

This enables Baytrail and similar atom CPUs that typically ship with a 32-bit firmware, but have a 64-bit capable CPU. (This patch requires a newer shim package that is not ready yet, so don't pull this yet.) Signed-off-by: Peter Jones <pjones(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/16

1 4

[PATCH rhel6-branch] Increment MD container child counter even if its volumes are ignored (#1120640)
by Artur Paszkiewicz 11 Jan '16

11 Jan '16

Volumes which are not selected in Specialized Storage Devices are ignored in addUdevDevice. This causes that their parent container won't have its child counter incremented, which can later cause an incorrect unusedRaidMembersWarning. Resolves: rhbz#1120640 Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz(a)intel.com> --- storage/devicetree.py | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/storage/devicetree.py b/storage/devicetree.py index 4045cdf..b9ea399 100644 --- a/storage/devicetree.py +++ b/storage/devicetree.py @@ -1328,6 +1328,13 @@ class DeviceTree(object): # slice off the "/dev/" part, lvm filter cares only about the rest partitions_paths = [p[5:] for p in partitions_paths] map(lvm.lvm_cc_addFilterRejectRegexp, partitions_paths) + + if udev_device_get_md_container(info): + parentName = devicePathToName(udev_device_get_md_container(info)) + container = self.getDeviceByName(parentName) + if container: + container.addChild() + return log.debug("scanning %s (%s)..." % (name, sysfs_path)) -- 1.8.4.5

4 14

New: [rhinstaller/anaconda/pulls/92 rhel6-branch] Add zanata.xml to the files included in the archive (#1205286)
by dashea 14 Dec '15

14 Dec '15

Related: rhbz#1205286 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/92

1 7

New: [rhinstaller/anaconda/pulls/70 rhel6-branch] Write --cdl clearpart option to /root/anaconda-ks.cfg (#1211559)
by dashea 01 Dec '15

01 Dec '15

Resolves: rhbz#1211559 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/70

1 8

Updated: [rhinstaller/anaconda/pulls/6] network: Overwrite resolv.conf generated by systemd
by dashea 22 Oct '15

22 Oct '15

[Note: against F22-branch, but installing rawhide content, that's how I hit the bug] systemd changed to generate a symlink resolv.conf -> /run/networkd/resolv.conf which doesn't exist at installation time. (What we should do at some point is tell NetworkManager to serialize its state into the installation root and get out of the business of writing these files directly) https://bugzilla.redhat.com/show_bug.cgi?id=1116651 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/6

1 5

Updated: [rhinstaller/anaconda/pulls/49 master] Fix issues with fetching pull requests.
by dashea 22 Oct '15

22 Oct '15

The "base" SHA is not actually where the pull request branched from, though that would be pretty handy. Instead of trying to find a point to branch from, just fetch the PR branch into a new local branch and rebase that onto the destination branch. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/49

1 6

Updated: [rhinstaller/anaconda/pulls/33 f22-branch] master/f22-branch - Set the default fs label when reformatting (#1141981)
by dashea 21 Oct '15

21 Oct '15

If the new format has a default label, and a label hasn't been set by the user, use the default. This allows the ESP to be reformatted and used without removing it and recreating it. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/33

1 18

New: [rhinstaller/blivet/pulls/59 master] Implement the __dir__ method of the LazyImportObject class
by dashea 10 Jul '15

10 Jul '15

This makes the 'TAB-TAB' completion in ipython and dir() calls work as if the lazy-imported object was imported in a normal way. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/59

1 4

Updated: [rhinstaller/anaconda/pulls/32 master] Check if gnome-icon-theme is installed
by dashea 25 Jun '15

25 Jun '15

Just minimize the false negatives when this is installed. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/32

1 5

New: [rhinstaller/blivet/pulls/44 master] Set format options for mkfs
by dashea 18 Jun '15

18 Jun '15

Currently format options can be specified only for doFormat function. This patch provides options to set the format options. Usage ex:- format = blivet.formats.getFormat("xfs", device=lv.path) format.formatOptions = ["-f", "-K", "-i", "size=256", "-d", ...] blivetEnv.formatDevice(lv, format) blivetEnv.doIt() bug-url: https://bugzilla.redhat.com/show_bug.cgi?id=1180221 Signed-off-by: Timothy Asir Jeyasingh <tjeyasin(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/44

1 7

New: [rhinstaller/anaconda/pulls/17] Backport pyanaconda_tests/keyboard_test.py updates from master
by dashea 03 Jun '15

03 Jun '15

Seems OK (nosetests passes) but there's one thing on master that worries me: commit 45a05d0a16014e9a46a625bc2f418888d3bdeb7d Author: Vratislav Podzimek Date: Thu Jul 17 16:12:14 2014 +0200 Prevent crashes due to accessing X server from multiple threads I don't know how that applies to rhel7-branch. Please comment. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/17

1 9

New: [rhinstaller/anaconda/pulls/19] Backport grub_raid_test.py from master
by dashea 03 Jun '15

03 Jun '15

A few notes here: As is this test fails in grub_mbr_raid1_test(). See the lines: ``` expected_targets = set([(self.sda, self.boot_md), (self.sdb, self.boot_md)]) # expected_targets = set([(self.sda, self.sda2), (self.sdb, self.sdb2)]) ``` If the second one is used instead of the first one then the result is PASS. This is due to different support for GRUB stage2 between master and rhel7-branch. On RHEL7: ``` for stage2dev in self.stage2_device.parents: # if target disk contains any of /boot array's member # partitions, set up stage1 on each member's disk # and stage2 on each member partition stage1dev = stage2dev.disk targets.append((stage1dev, stage2dev)) ``` While on master: ``` for stage2dev in stage2_parents: # if target disk contains any of /boot array's member # partitions, set up stage1 on each member's disk stage1dev = stage2dev.disk targets.append((stage1dev, self.stage2_device)) ``` Notice the self.stage2_device vs. stage2dev. The first one is the md1 partition while the second one is either sda2 or sdb2 - the mdraid member partitions. I don't know which one is the expected behavior for RHEL 7. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/19

1 7

New: [rhinstaller/anaconda/pulls/12] Remove W0123 which is not defined in older pylint and causes traceback
by dashea 03 Jun '15

03 Jun '15

Running the test suite on RHEL7 yields: ``` Traceback (most recent call last): File "/usr/bin/pylint", line 9, in <module> load_entry_point('pylint==1.0.0', 'console_scripts', 'pylint')() File "/usr/lib/python2.7/site-packages/pylint/__init__.py", line 21, in run_pylint Run(sys.argv[1:]) File "/usr/lib/python2.7/site-packages/pylint/lint.py", line 990, in __init__ args = linter.load_command_line_configuration(args) File "/usr/lib/python2.7/site-packages/logilab/common/configuration.py", line 685, in load_command_line_configuration (options, args) = self.cmdline_parser.parse_args(args=args) File "/usr/lib64/python2.7/optparse.py", line 1399, in parse_args stop = self._process_args(largs, rargs, values) File "/usr/lib64/python2.7/optparse.py", line 1439, in _process_args self._process_long_opt(rargs, values) File "/usr/lib64/python2.7/optparse.py", line 1514, in _process_long_opt option.process(opt, value, values, self) File "/usr/lib/python2.7/site-packages/logilab/common/optik_ext.py", line 237, in process self.action, self.dest, opt, value, values, parser) File "/usr/lib64/python2.7/optparse.py", line 808, in take_action self.callback(self, opt, value, parser, *args, **kwargs) File "/usr/lib/python2.7/site-packages/logilab/common/configuration.py", line 540, in cb_set_provider_option self.global_set_option(opt, value) File "/usr/lib/python2.7/site-packages/logilab/common/configuration.py", line 544, in global_set_option self._all_options[opt].set_option(opt, value) File "/usr/lib/python2.7/site-packages/pylint/lint.py", line 383, in set_option meth(_id) File "/usr/lib/python2.7/site-packages/pylint/utils.py", line 251, in disable msg = self.check_message_id(msgid) File "/usr/lib/python2.7/site-packages/pylint/utils.py", line 315, in check_message_id raise UnknownMessage('No such message id %s' % msgid) pylint.utils.UnknownMessage: No such message id W0123 ``` With this small fix it displays correct results from pylint: ``` FAIL: pylint/runpylint ====================== pylint reports the following issues: ************* Module ../pyanaconda/ui/gui/spokes/lib/resize.py E0611: 25,0: : No name 'Gdk' in module 'gi.repository' E0611: 25,0: : No name 'Gtk' in module 'gi.repository' ************* Module ../pyanaconda/anaconda.py F0401:161,8: Anaconda.dumpState: Unable to import 'meh' F0401:162,8: Anaconda.dumpState: Unable to import 'meh.dump' ``` -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/12

1 9

New: [rhinstaller/anaconda/pulls/18] backport iutil_test.py updates from master
by dashea 02 Jun '15

02 Jun '15

Please comment on this PR. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/18

1 6

New: [rhinstaller/anaconda/pulls/22 rhel7-branch] Backport glade tests from master
by dashea 02 Jun '15

02 Jun '15

I've cherry picked these two commits from David and fixed a small conflict while applying the second one. The tests also discover some issues. Will open a bug for them. Please comment on the PR state? Do you want me to squash the two commits together ? -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/22

1 8

New: [rhinstaller/anaconda/pulls/13] Backport pylint updates from master
by dashea 02 Jun '15

02 Jun '15

Notes: - this PR also contains 2206654b6cedd2dc48618435db7a5c55ff294afa which was submitted previously as PR #12. - I've copied the add_metaclass() into pointless-override.py b/c RHEL7 has an older version of six, which doesn't define it Please comment! -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/13

1 7

Updated: [rhinstaller/anaconda/pulls/10] backported gettext updates from master
by dashea 02 Jun '15

02 Jun '15

Hi guys, I'm not sure about removal of this line: potcheckfile = checkfile + ".h" but as far as I can tell it should be the correct thing to do. There are both .glade and .glade.h files and I think the .h files need to go. Everything else appears to be OK. Please comment. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/10

1 7

New: [rhinstaller/anaconda/pulls/16] Move the pykickstart version test into a different subdirectory.
by dashea 02 Jun '15

02 Jun '15

I've just cherry-picked this from master. Doesn't seem to break anything although it doesn't add new value except making rhel7-branch closer to master. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/16

1 6

Updated: [rhinstaller/blivet/pulls/46 f22-branch] Fix XFS' and BTRFS' max size limits (#1200812)
by dashea 01 Jun '15

01 Jun '15

These should be max 64-bit unsigned integer value which is 16 EiB - 1. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/46

1 13

New: [rhinstaller/anaconda/pulls/15] Backport: Fix # handling in SimpleConfigFile (#1045687)
by dashea 19 May '15

19 May '15

This PR includes the original commit by Brian just modified to use uppercase_ASCII_string() instead of upperASCII(). Please comment. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/15

1 3

New: [rhinstaller/blivet/pulls/98 rhel7-branch] Include LUKSDevice information in kickstart data (#1139222)
by dashea 12 May '15

12 May '15

Resolves: rhbz#1139222 Previously, if a device had a child that was a LUKS device that child was ignored when generating ksdata. What this meant was that that device's format, which was luks, was used in generating the kickstart data. Now, the LUKSDevice, which contains the format that should appear in the ks output is used. This means that the LUKS format is used to set some ks data, but the mountpoint and so forth are overwritten with the correct values when the format on the LUKSDevice is used to populate those values. The --passphrase option with passphrase continues to not appear in the output anaconda-ks.cfg file. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/98

1 6

New: [rhinstaller/blivet/pulls/60 master] Master tasks
by dashea 12 May '15

12 May '15

The purpose of this patch is to make blivet able to operate in a context in which it is aware of what external dependencies are available. What this means is a couple of things: * If a DeviceAction that requires an external resource is constructed and that resource is unavailable then the constructor will raise an exception. Currently these exceptions are ValueErrors, which are the same kind of exception raised when the device or some other aspect of the environment has a problem. * All methods on formats like formattable, supported, are environment aware. Previously they usually returned True if the functionality was implemented in blivet...now they only return True if the functionality is implemented _and_ also the external tool required is available. So, where anaconda uses methods to determine whether to display a particular filesystem type in a combo, those should all return False if the operations they check can not be run. * Ultimately, if by some chance a method is called requiring a task with an unavailable external dependency, an exception will be raised indicating that the problem is the unavailability of the external application. This is already implemented in the filesystem format tasks and will be implemented in the libblockdev plugins. There are two external apps in devices, multipath and kpartx for which perhaps it should be implemented as well. Availability of resources has its own separate simple packages. There are only two kinds of external resources so far, libblockdev plugins and executables. libblockdev plugin availability is checked by loadedness, which is always going to be correct on an installed system. executable are available if their names are in $PATH. There are three different categories of things to track availability of, in order of difficulty are: * non-filesystem formats * filesystem formats * devices Non-filesystem formats are easy, because if they depend on anything external they depend on exactly one libblockdev plugin. SwapSpace is an example and all others are handled similarly. All methods like formattable, supportable, typically just check whether the plugin is available. There's just not much to look at for this. Filesystem formats are bit harder. Each filesystem has a lot of tasks to do, create, mount, fsck, info getting, size calculating, etc, and these can be handled by different applications that get installed by different packages. To handle this, these tasks are abstracted out into subclasses of Task and each is set up with an unavailable and an unimplemented method. It is the availability of these tasks that is checked for the various filesystem format properties. Abstracting these tasks into separate tasks forms a significant part of these patches but these probably don't require a lot of scrutiny either. There are no additional tests for these tasks, as they are like friend classes of the filesystem formats and don't really make sense being tested outside the formats classes. They all pass a bunch of regression tests on filesystem formats. Devices is of course the hardest. Besides device classes just generally being more intricate there are the following problems: * Calculations must be done recursively on ancestor devices. * Dependencies seem to accumulate as you travel down the class hierarchy. * Abstracting into tasks like for filesystem formats is not going to happen, but devices aren't really like non-filesystems formats in having just one simple dependency, either. Under these circumstances it is impossible to be correct about all dependencies so I have aimed for being conservative, i.e., sometimes raising an exception when constructing a DeviceAction that actually could go through by some special circumstances. There are only two mechanisms: * an ability to find unavailable external resources for a device and all its ancestors * an ability to find unavailable external resources required for setting up formats for the formats on all the ancestors of a device I've chosen these because it seems, from study of the code, that for almost every DeviceAction a setup task must be performed that requires doing things to the device and to the formats on that device. Generally there will be some small check on the specifics of the action as well; for example, the ActionDestroyFormat class has a check whether the format to be destroyed is actually destroyable. When creating a format, I check whether the format is formattable, DeviceFormat's are not formattable so an exception is raised in that case. Currently, the search for unavailable external resources for a device and all its ancestors has only two modes, Default and Destroy. I chose to add the Destroy mode, because a PartitionDevice has a dependency on the dm plugin but only in the _postDestroy method and AFAICT, it would be a little harsh to require the dm plugin for every combination of DeviceAction, PartitionDevice. The mode mechanism is ready to be extended as necessary to refine the search to only relevant dependencies in case some other situation comes up where the DeviceAction constructor would raise an exception even though the action could be succesfully executed (and it is considered worthwhile to distinguish). For checking whether formats are setupable I wrote two functions that mirror the structure of Device.setup and Device.setupParents, so that these function do not check the setupability of formats that setup does not actually try to setup. In summary, it's probably OK to skip the code about formats, although you might want to look at the Tasks abstractions, and to focus on how I deal with devices, and how I insert calls into DeviceAction classes. That's the tricky part. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/60

1 84

Updated: [rhinstaller/blivet/pulls/90 master] Master tasks 20150422
by dashea 12 May '15

12 May '15

Now with fewer commits. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/90

1 48

[Patch for Review] Introduce delay for slow USB devices to detect OEMDRV on time.
by B_B_Singh＠DELL.com 06 May '15

06 May '15

Hi , During the installation of RHEL-7.X the driver pack drivers or OEMDRV is NOT getting detected on time, due to which the latest drivers will not get installed from OEMDRV. The problem is due to slow usb. Upon introducing the delay for slow usb (OEMDRV) during the insertion of modules after "udevadm trigger" everything works fine. Here is the patch. --- anaconda-19.31.79/dracut/driver-updates.sh 2015-04-20 23:03:45.160081886 +0530 +++ anaconda-19.31.79_new/dracut/driver-updates.sh 2015-04-20 23:07:38.719101243 +0530 @@ -4,6 +4,7 @@ # load all modules udevadm trigger +sleep 2 udevadm settle # Look for devices with the OEMDRV label Regards Balaji Singh

3 5

New: [rhinstaller/anaconda/pulls/87 master] Filter spoke fixes and improvements
by dashea 05 May '15

05 May '15

I'd also like to port at least the first 6 to rhel7-branch -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/87

1 17

Re: [Patch for Review] iSCSI Name Validation using regex.
by Sujith Pandel 04 May '15

04 May '15

Hi Vratislav, Thank you for your feedback on this. One of the changes which you mentioned in this: > > --- a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py > > +++ b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py > > @@ -18,7 +18,7 @@ > > # > > # Red Hat Author(s): Chris Lumens <clumens(a)redhat.com> > > # > > - > > +import re > > +from pyanaconda.regexes import ISCSI_IQN_NAME_REGEX, ISCSI_EUI_NAME_REGEX > >.... > > + #iSCSI Name validation using regex. Name should either match IQN format or EUI format. > > + if re.match(ISCSI_IQN_NAME_REGEX,stripped) or re.match(ISCSI_EUI_NAME_REGEX,stripped): > We prefer calling the match() method of the compiled regexes instead of > using re.match() on them. That would also remove allow you to get rid of > the commas that would otherwise need a space right after them. :) > > > + return True > > + else: > > + return False > This if statement should be reduced to just: > > return ISCSI_ION_NAME_REGEX.match(stripped) or ISCSI_EUI_NAME_REGEX.match(stripped) > > Please note that this results in a run time exception when return value becomes 'None'. Verified this condition. (Boolean values are expected at the receiving end.) Based on your suggestions above, I am thinking of changing it to: if ISCSI_IQN_NAME_REGEX.match(stripped) or ISCSI_EUI_NAME_REGEX.match(stripped): return True else: return False This will also prevent the need to "import re" in iscsi.py Let me know if any changes are required to it. > > +import unittest > > +import re > Seems like the 're' module is not needed anywhere below. > Yes. I'll remove it. I'll send the modified patch incorporating the changes once you confirm on these changes. Thanks and Regards, Sujith Pandel

2 1

New: [rhinstaller/blivet/pulls/101 rhel7-branch] Include LUKSDevice information in kickstart data (#1139222)
by dashea 01 May '15

01 May '15

Resolves: rhbz#1139222 Previously, if a device had a child that was a LUKS device that child was ignored when generating ksdata. What this meant was that that device's format, which was luks, was used in generating the kickstart data. Now, the LUKSDevice, which contains the format that should appear in the ks output is used. This means that the LUKS format is used to set some ks data, but the mountpoint and so forth are overwritten with the correct values when the format on the LUKSDevice is used to populate those values. The --passphrase option with passphrase continues to not appear in the output anaconda-ks.cfg file. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/101

1 5

[PATCH] Adapt to the new argument list for save_servers_to_config.
by Chris Lumens 30 Apr '15

30 Apr '15

This was caught by pylint. --- anaconda | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/anaconda b/anaconda index b124b52..565a263 100755 --- a/anaconda +++ b/anaconda @@ -1362,7 +1362,7 @@ if __name__ == "__main__": # setup ntp servers and start NTP daemon if not requested otherwise if can_touch_runtime_system("start chronyd"): if anaconda.ksdata.timezone.ntpservers: - ntp.save_servers_to_config(anaconda.ksdata.timezone.ntpservers) + ntp.save_servers_to_config([], anaconda.ksdata.timezone.ntpservers) if not anaconda.ksdata.timezone.nontp: iutil.start_service("chronyd") -- 2.2.2

3 2

[anaconda][f22-blocker] Catch libblockdev's CryptoError when trying to unlock LUKS
by Vratislav Podzimek 30 Apr '15

30 Apr '15

Otherwise we get a traceback instead of warning and another chance for the user. Signed-off-by: Vratislav Podzimek <vpodzime(a)redhat.com> --- pyanaconda/rescue.py | 5 +++-- pyanaconda/ui/gui/spokes/custom.py | 3 ++- 2 files changed, 5 insertions(+), 3 deletions(-) diff --git a/pyanaconda/rescue.py b/pyanaconda/rescue.py index b3c96b3..6fba78d 100644 --- a/pyanaconda/rescue.py +++ b/pyanaconda/rescue.py @@ -42,6 +42,7 @@ from blivet.devices import LUKSDevice from blivet.osinstall import storageInitialize, mountExistingSystem from pykickstart.constants import KS_REBOOT, KS_SHUTDOWN +from gi.repository import BlockDev as blockdev import meh.ui.text @@ -245,8 +246,8 @@ def _unlock_devices(intf, storage): unlocked = True # try to use the same passhprase for other devices try_passphrase = passphrase - except StorageError as serr: - log.error("Failed to unlock %s: %s", device.name, serr) + except (StorageError, blockdev.CryptoError) as err: + log.error("Failed to unlock %s: %s", device.name, err) device.teardown(recursive=True) device.format.passphrase = None try_passphrase = None diff --git a/pyanaconda/ui/gui/spokes/custom.py b/pyanaconda/ui/gui/spokes/custom.py index 19b376a..95e53a0 100644 --- a/pyanaconda/ui/gui/spokes/custom.py +++ b/pyanaconda/ui/gui/spokes/custom.py @@ -88,6 +88,7 @@ from pyanaconda.ui.gui.utils import really_hide, really_show, timed_action from pyanaconda.ui.categories.system import SystemCategory from gi.repository import Gdk, Gtk +from gi.repository import BlockDev as blockdev from gi.repository.AnacondaWidgets import MountpointSelector from functools import wraps @@ -2653,7 +2654,7 @@ class CustomPartitioningSpoke(NormalSpoke, StorageChecker): try: device.setup() device.format.setup() - except StorageError as e: + except (StorageError, blockdev.CryptoError) as e: log.error("failed to unlock %s: %s", device.name, e) device.teardown(recursive=True) self._error = e -- 2.1.0

2 1

New: [rhinstaller/anaconda/pulls/83 master] Move some blivet Requires to anaconda.
by dashea 29 Apr '15

29 Apr '15

When blivet becomes environment aware, it will not need to require these packages, since it should not fail in the event that these packages are not installed. Presumably anaconda still wants the capabilities these packages provide, since it is used to having them through blivet, so add these Requires to anaconda. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/83

1 8

New: [rhinstaller/anaconda/pulls/77 master] Replace GtkSpinner (#1204242)
by dashea 29 Apr '15

29 Apr '15

Evolution had the same problem, so steal their fix. The spinning image comes out a good bit lighter than GtkSpinner. This may have something to do with how gtk processes symbolic icons? I don't know. I put the steps I took to create the png into a comment in the widget code. Something weird and dumb is going on with what gtk thinks a color is and if anyone can fix that that would be great. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/77

1 11

New: [rhinstaller/anaconda/pulls/90 master] Remove the restriction that /boot be below 2TB for grub (#1082331)
by dashea 29 Apr '15

29 Apr '15

If the disk is over 2TiB, it's going to be a gpt disklabel, so the stage1 bootloader will either be on a biosboot or EFI partition. Letting stage2 be wherever is probably ok. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/90

1 4

New: [rhinstaller/anaconda/pulls/89 master] Set the GUI-selected environment in the ksdata (#1192100)
by dashea 29 Apr '15

29 Apr '15

This way an environment selected in the GUI and an environment selected in kickstart are treated in the same way. This fixes the problem where changing payloads forgets the previous environment selection, and it changes the generated %packages section of anaconda-ks.cfg to contain the environment information instead of a list of the groups in the environment. This commit restores the dnf _select_environment method removed in 602f47dcd5d7543f66dc6833685d0e810595acb5, but modifies it so that the environment is both selected in dnf and in the kickstart data. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/89

1 10

Updated: [rhinstaller/anaconda/pulls/75 master] Add support for chrony pool directive
by dashea 29 Apr '15

29 Apr '15

With chrony-2.0 it is possible to specify a pool of NTP servers instead of individual servers. It selects four servers from the addresses that the pool name is resolved to and replaces them automatically when they are unreachable or falsetickers. This is now used in the default Fedora chrony configuration. To avoid extending the ntp API and the date time spoke, pools are simply presented as four separate servers with the same name. Resolves: rhbz#1193389 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/75

1 5

[lorax] Add product.img support for s390 templates
by Dan Horák 29 Apr '15

29 Apr '15

--- share/live/s390.tmpl | 10 ++++++++++ share/s390.tmpl | 17 +++++++++++++++++ 2 files changed, 27 insertions(+) diff --git a/share/live/s390.tmpl b/share/live/s390.tmpl index f02963d..86e020d 100644 --- a/share/live/s390.tmpl +++ b/share/live/s390.tmpl @@ -4,6 +4,7 @@ configdir="tmp/config_files/s390" BOOTDIR="images" KERNELDIR=BOOTDIR INITRD_ADDRESS="0x02000000" +LORAXDIR="usr/share/lorax/" # The assumption seems to be that there is only one s390 kernel, ever kernel = kernels[0] %> @@ -31,3 +32,12 @@ createaddrsize ${INITRD_ADDRESS} ${outroot}/${BOOTDIR}/initrd.img ${outroot}/${B treeinfo images-${basearch} initrd.addrsize ${BOOTDIR}/initrd.addrsize treeinfo images-${basearch} generic.prm ${BOOTDIR}/generic.prm treeinfo images-${basearch} generic.ins generic.ins + +# Create optional product.img and updates.img +<% imggraft=""; images=["product", "updates"] %> +%for img in images: + %if exists("%s/%s/" % (LORAXDIR, img)): + installimg ${LORAXDIR}/${img}/ images/${img}.img + <% imggraft += " images/{0}.img={1}/images/{0}.img".format(img, outroot) %> + %endif +%endfor diff --git a/share/s390.tmpl b/share/s390.tmpl index fa712fd..9d7eb00 100644 --- a/share/s390.tmpl +++ b/share/s390.tmpl @@ -4,6 +4,7 @@ configdir="tmp/config_files/s390" BOOTDIR="images" KERNELDIR=BOOTDIR INITRD_ADDRESS="0x02000000" +LORAXDIR="usr/share/lorax/" # The assumption seems to be that there is only one s390 kernel, ever kernel = kernels[0] %> @@ -38,3 +39,19 @@ treeinfo images-${basearch} initrd.addrsize ${BOOTDIR}/initrd.addrsize treeinfo images-${basearch} generic.prm ${BOOTDIR}/generic.prm treeinfo images-${basearch} genericdvd.prm ${BOOTDIR}/genericdvd.prm treeinfo images-${basearch} generic.ins generic.ins + +# Create optional product.img and updates.img +<% imggraft=""; images=["product", "updates"] %> +%for img in images: + %if exists("%s/%s/" % (LORAXDIR, img)): + installimg ${LORAXDIR}/${img}/ images/${img}.img + <% imggraft += " images/{0}.img={1}/images/{0}.img".format(img, outroot) %> + %endif +%endfor + +# Inherit iso-graft/ if it exists from external templates +<% + import os + if os.path.exists(workdir + "/iso-graft"): + imggraft += " " + workdir + "/iso-graft" +%> -- 1.9.3

2 2

New: [rhinstaller/blivet/pulls/99 master] Do not update size information on LVM snapshots
by dashea 29 Apr '15

29 Apr '15

Current updateSize method implementation checks the size of devices in /sys where the size of LVM snapshots is incorrect (it reports the size of the LV instead of the actual size of the snapshot). My fix actually prevents blivet from updating size of the snapshot. The question is if we want to update the size because it would require to run the devicetree.lvInfo() function again (because only "lvs" reports the correct size for snapshots). -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/99

1 4

New: [rhinstaller/blivet/pulls/97 master] If the parent volume has a label, use it in subvol's kickstart (#1072060...
by dashea 29 Apr '15

29 Apr '15

...) Resolves: rhbz#1072060 The LABEL= entry identifies the parent of the specified subvolume by its label. If the parent volume does not have a label, then no entry is made that identifies the parent volume and the generated kickstart file will be rejected by ksvalidator. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/97

1 4

Re: [Patch for Review] iSCSI Name Validation using regex.
by Sujith Pandel 29 Apr '15

29 Apr '15

Hi Chris, I did a few changes to the previously sent patch. Also added the test script for regexes' validation. Please have a look and let me know your thoughts on them. Regarding the lookaheads and the dots, I feel regexes might need major changes to accomodate those changes. I have added few comments for the regexes also. Feel free to add/modify them, if required. Thanks and Regards, Sujith Pandel When adding iSCSI target disk during installation, the iSCSI Initiator name and the Target IP is validated to enable/disable the "Start Discovery" button on the page. Currently, the validation of iSCSI Initiator Name is based on the existence of dots only (".") and this can be more efficiently done by using a regular expression. The conditions for iSCSI initiator name used in this regex are (https://tools.ietf.org/html/rfc3720#section-3.2.6.3.1 , https://tools.ietf.org/html/rfc3721#page-5 and http://standards.ieee.org/regauth/oui/tutorials/EUI64.html): 1. For iqn format: a. Starts with string 'iqn.' b. A date code specifying the year and month in which the organization registered the domain or sub-domain name used as the naming authority string. "yyyy-mm" c. A dot (".") d. The organizational naming authority string, which consists of a valid, reversed domain or subdomain name. e. Optionally, a colon (":"), followed by a string of the assigning organization's choosing, which must make each assigned iSCSI name unique. With the exception of the colon prefix, the owner of the domain name can assign everything after the reversed domain name as desired. 2. For eui format: The format is "eui." followed by an EUI-64 identifier (16 ASCII-encoded hexadecimal digits). Signed-off-by: Sujith Pandel <sujithpshankar(a)gmail.com> --- pyanaconda/regexes.py | 19 +++++++ pyanaconda/ui/gui/spokes/advstorage/iscsi.py | 9 +++- tests/regex_tests/iscsi_name_test.py | 78 ++++++++++++++++++++++++++++ 3 files changed, 104 insertions(+), 2 deletions(-) create mode 100644 tests/regex_tests/iscsi_name_test.py diff --git a/pyanaconda/regexes.py b/pyanaconda/regexes.py index b4e52a3..a30b451 100644 --- a/pyanaconda/regexes.py +++ b/pyanaconda/regexes.py @@ -135,3 +135,22 @@ REPO_NAME_VALID = re.compile(r'^[a-zA-Z0-9_.:-]+$') # Product Version string, just the starting numbers like 21 or 21.1 VERSION_DIGITS = r'([\d.]+)' + + +#Regexes to validate iSCSI Names according to RFC 3720 and RFC 3721 +#The conditions for iSCSI name used in the following regexes are +#(https://tools.ietf.org/html/rfc3720#section-3.2.6.3.1 , https://tools.ietf.org/html/rfc3721#page-5 and http://standards.ieee.org/regauth/oui/tutorials/EUI64.html): +#1. For iqn format: +# a. Starts with string 'iqn.' +# b. A date code specifying the year and month in which the organization registered the domain or sub-domain name used as the naming authority string. "yyyy-mm" +# c. A dot (".") +# d. The organizational naming authority string, which consists of a valid, reversed domain or subdomain name. +# e. Optionally, a colon (":"), followed by a string of the assigning organization's choosing, which must make each assigned iSCSI name unique. With the exception of the colon prefix, the owner of the domain name can assign everything after the reversed domain name as desired. + +#2. For eui format: +# a. The format is "eui." followed by an EUI-64 identifier (16 ASCII-encoded hexadecimal digits). +ISCSI_IQN_NAME_REGEX = re.compile(r'^iqn\.\d{4}-\d{2}((?<!-)\.(?!-)[a-zA-Z0-9\-]+){1,63}(?<!-)(?<!\.)(:[^:]+)?$') + +ISCSI_EUI_NAME_REGEX = re.compile(r'^eui\.[a-fA-F0-9]{16}$') + + diff --git a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py index 22e11c3..26c34c3 100644 --- a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py +++ b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py @@ -18,7 +18,7 @@ # # Red Hat Author(s): Chris Lumens <clumens(a)redhat.com> # - +import re from IPy import IP from collections import namedtuple from gi.repository import GLib @@ -29,6 +29,7 @@ from pyanaconda.ui.gui import GUIObject from pyanaconda.ui.gui.utils import escape_markup from pyanaconda.i18n import _ from pyanaconda import nm +from pyanaconda.regexes import ISCSI_IQN_NAME_REGEX, ISCSI_EUI_NAME_REGEX __all__ = ["ISCSIDialog"] @@ -303,7 +304,11 @@ class ISCSIDialog(GUIObject): stripped = text.strip() #iSCSI Naming Standards: RFC 3720 and RFC 3721 - return "." in stripped + #iSCSI Name validation using regex. Name should either match IQN format or EUI format. + if re.match(ISCSI_IQN_NAME_REGEX,stripped) or re.match(ISCSI_EUI_NAME_REGEX,stripped): + return True + else: + return False def on_discover_field_changed(self, *args): # Make up a credentials object so we can test if it's valid. diff --git a/tests/regex_tests/iscsi_name_test.py b/tests/regex_tests/iscsi_name_test.py new file mode 100644 index 0000000..8571396 --- /dev/null +++ b/tests/regex_tests/iscsi_name_test.py @@ -0,0 +1,78 @@ +#!/usr/bin/python +# vim:set fileencoding=utf-8 +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see <http://www.gnu.org/licenses/>. + +# Author(s): Sujith Pandel <sujithpshankar(a)gmail.com> +# Ajith Pandel <ajithpandel(a)gmail.com> +# +import unittest +import re + +from regexcheck import regex_match +from pyanaconda.regexes import ISCSI_IQN_NAME_REGEX, ISCSI_EUI_NAME_REGEX + +class iSCSIiqnnameRegexTestCase(unittest.TestCase): + def iqnname_test(self): + good_tests = [ + 'iqn.2014-15.com.example', + 'iqn.2014-15.com.example:iscsi', + 'iqn.2014-15.c-om.example:iscsi', + 'iqn.2014-15.c.om.example:iscsi', + 'iqn.2014-15.com.example:...', + 'iqn.2014-15.com.example:iscsi_@nything_except_colon_after_colon!' + ] + + bad_tests = [ + 'iqn', + 'iqn.', + 'iqn.2014-15', + 'iqn.2014-15.', + 'iqn.2014-15..', + 'iqn.2014-15.com.example.', + 'iqn.2014-15.com.example...', + 'iqn.2014-15.com.example:', + 'iqn.2014-15.-com.example', + 'iqn.2014-15.com-.example', + 'iqn.2014-15.-.example', + 'iqn.2014-15.com.example-:iscsi', + 'abciqn.2014-15.com.example:iscsi', + 'iqn.2014-15.-.example:iscsi', + 'iqn.2014-15.com&com.example:iscsi', + 'iqn.2014-15.com.example:iscsi:doublecolon', + 'iqn.2014-15..om.example:iscsi', + ] + + if not regex_match(ISCSI_IQN_NAME_REGEX, good_tests, bad_tests): + self.fail() + +class iSCSIeuinameRegexTestCase(unittest.TestCase): + def euiname_test(self): + good_tests = [ + 'eui.ABCDEF0123456789', + 'eui.abcdef0123456789', + 'eui.0123456789ABCDEF' + ] + + bad_tests = [ + 'eui', + 'eui.', + 'eui.2014-', + 'eui.exampleeui789abc' + 'eui.AAAABBBBCCC2345', + 'eui.AAAABBBBCCCCD4567' + ] + + if not regex_match(ISCSI_EUI_NAME_REGEX, good_tests, bad_tests): + self.fail() -- 2.1.0

2 1

[PATCH 01/11] Move blivet to using pocketlint.
by Chris Lumens 29 Apr '15

29 Apr '15

There's a fair number of errors right now, some of which could be false positives, but I am going to take care of those in subsequent commits. --- Makefile | 3 + python-blivet.spec | 2 + tests/pylint/intl.py | 63 --------- tests/pylint/pointless-override.py | 273 ------------------------------------ tests/pylint/pylint-false-positives | 28 ---- tests/pylint/pylint-one.sh | 33 ----- tests/pylint/runpylint.py | 34 +++++ tests/pylint/runpylint.sh | 125 ----------------- 8 files changed, 39 insertions(+), 522 deletions(-) delete mode 100644 tests/pylint/intl.py delete mode 100644 tests/pylint/pointless-override.py delete mode 100644 tests/pylint/pylint-false-positives delete mode 100755 tests/pylint/pylint-one.sh create mode 100755 tests/pylint/runpylint.py delete mode 100755 tests/pylint/runpylint.sh diff --git a/Makefile b/Makefile index 9f60551..a55f500 100644 --- a/Makefile +++ b/Makefile @@ -36,6 +36,9 @@ coverage: PYTHONPATH=.:tests/ coverage run --branch -m unittest discover -v -s tests/ -p '*_test.py' coverage report --include="blivet/*" +check: + PYTHONPATH=. tests/pylint/runpylint.py + clean: -rm *.tar.gz blivet/*.pyc blivet/*/*.pyc ChangeLog $(MAKE) -C po clean diff --git a/python-blivet.spec b/python-blivet.spec index eeaaba8..ef4bacb 100644 --- a/python-blivet.spec +++ b/python-blivet.spec @@ -13,6 +13,7 @@ Source0: http://github.com/dwlehman/blivet/archive/%{realname}-%{version}.tar.gz # Versions of required components (done so we make sure the buildrequires # match the requires versions of things). %define pykickstartver 1.99.22 +%define pocketlintver 0.4 %define partedver 1.8.1 %define pypartedver 3.10.3 %define e2fsver 1.41.0 @@ -22,6 +23,7 @@ Source0: http://github.com/dwlehman/blivet/archive/%{realname}-%{version}.tar.gz BuildArch: noarch BuildRequires: gettext BuildRequires: python-setuptools +BuildRequires: python3-pocketlint >= %{pocketlintver} %if 0%{with_python3} BuildRequires: python3-devel python3-setuptools diff --git a/tests/pylint/intl.py b/tests/pylint/intl.py deleted file mode 100644 index 3052680..0000000 --- a/tests/pylint/intl.py +++ /dev/null @@ -1,63 +0,0 @@ -# I18N-related pylint module -# -# Copyright (C) 2013 Red Hat, Inc. -# -# This copyrighted material is made available to anyone wishing to use, -# modify, copy, or redistribute it subject to the terms and conditions of -# the GNU General Public License v.2, or (at your option) any later version. -# This program is distributed in the hope that it will be useful, but WITHOUT -# ANY WARRANTY expressed or implied, including the implied warranties of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General -# Public License for more details. You should have received a copy of the -# GNU General Public License along with this program; if not, write to the -# Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA -# 02110-1301, USA. Any Red Hat trademarks that are incorporated in the -# source code or documentation are not subject to the GNU General Public -# License and may only be used or replicated with the express permission of -# Red Hat, Inc. -# -# Red Hat Author(s): Chris Lumens <clumens(a)redhat.com> -# - -import astroid - -from pylint.checkers import BaseChecker -from pylint.checkers.utils import check_messages -from pylint.interfaces import IAstroidChecker - -translationMethods = ["_", "N_", "P_", "C_", "CN_", "CP_"] - -class IntlChecker(BaseChecker): - __implements__ = (IAstroidChecker, ) - name = "internationalization" - msgs = {"W9901": ("Found % in a call to a _() method", - "found-percent-in-_", - "% in a call to one of the _() methods results in incorrect translations"), - "W9902": ("Found _ call at module/class level", - "found-_-in-module-class", - "Calling _ at the module or class level results in translations to the wrong language") - } - - @check_messages("W9901") - def visit_binop(self, node): - if node.op != "%": - return - - curr = node - while curr.parent: - if isinstance(curr.parent, astroid.CallFunc) and getattr(curr.parent.func, "name", "") in translationMethods: - self.add_message("W9901", node=node) - break - - curr = curr.parent - - @check_messages("W9902") - def visit_callfunc(self, node): - # The first test skips internal functions like getattr. - if isinstance(node.func, astroid.Name) and node.func.name == "_": - if isinstance(node.scope(), astroid.Module) or isinstance(node.scope(), astroid.Class): - self.add_message("W9902", node=node) - -def register(linter): - """required method to auto register this checker """ - linter.register_checker(IntlChecker(linter)) diff --git a/tests/pylint/pointless-override.py b/tests/pylint/pointless-override.py deleted file mode 100644 index e5f0fd8..0000000 --- a/tests/pylint/pointless-override.py +++ /dev/null @@ -1,273 +0,0 @@ -# Pylint checker for pointless class attributes overrides. -# -# Copyright (C) 2014 Red Hat, Inc. -# -# This copyrighted material is made available to anyone wishing to use, -# modify, copy, or redistribute it subject to the terms and conditions of -# the GNU General Public License v.2, or (at your option) any later version. -# This program is distributed in the hope that it will be useful, but WITHOUT -# ANY WARRANTY expressed or implied, including the implied warranties of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General -# Public License for more details. You should have received a copy of the -# GNU General Public License along with this program; if not, write to the -# Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA -# 02110-1301, USA. Any Red Hat trademarks that are incorporated in the -# source code or documentation are not subject to the GNU General Public -# License and may only be used or replicated with the express permission of -# Red Hat, Inc. -# -# Red Hat Author(s): Anne Mulhern <amulhern(a)redhat.com> -# - -import abc - -from six import add_metaclass - -import astroid - -from pylint.checkers import BaseChecker -from pylint.checkers.utils import check_messages -from pylint.interfaces import IAstroidChecker - -(a)add_metaclass(abc.ABCMeta) -class PointlessData(object): - - _DEF_CLASS = abc.abstractproperty(doc="Class of interesting definitions.") - message_id = abc.abstractproperty(doc="Pylint message identifier.") - - @classmethod - @abc.abstractmethod - def _retain_node(cls, node, restrict=True): - """ Determines whether to retain a node for the analysis. - - :param node: an AST node - :type node: astroid.Class - :param restrict bool: True if results returned should be restricted - :returns: True if the node should be kept, otherwise False - :rtype: bool - - Restricted nodes are candidates for being marked as overridden. - Only restricted nodes are put into the initial pool of candidates. - """ - raise NotImplementedError() - - @staticmethod - @abc.abstractmethod - def _extract_value(node): - """ Return the node that contains the assignment's value. - - :param node: an AST node - :type node: astroid.Class - :returns: the node corresponding to the value - :rtype: bool - """ - raise NotImplementedError() - - @staticmethod - @abc.abstractmethod - def _extract_targets(node): - """ Generates the names being assigned to. - - :param node: an AST node - :type node: astroid.Class - :returns: a list of assignment target names - :rtype: generator of str - """ - raise NotImplementedError() - - @classmethod - def get_data(cls, node, restrict=True): - """ Find relevant nodes for this analysis. - - :param node: an AST node - :type node: astroid.Class - :param restrict bool: True if results returned should be restricted - - :rtype: generator of astroid.Class - :returns: a generator of interesting nodes. - - Note that all nodes returned are guaranteed to be instances of - some class in self._DEF_CLASS. - """ - nodes = (n for n in node.body if isinstance(n, cls._DEF_CLASS)) - for n in nodes: - if cls._retain_node(n, restrict): - for name in cls._extract_targets(n): - yield (name, cls._extract_value(n)) - - @classmethod - @abc.abstractmethod - def check_equal(cls, node, other): - """ Check whether the two nodes are considered equal. - - :param node: some ast node - :param other: some ast node - - :rtype: bool - :returns: True if the nodes are considered equal, otherwise False - - If the method returns True, the nodes are actually equal, but it - may return False when the nodes are equal. - """ - raise NotImplementedError() - -class PointlessFunctionDefinition(PointlessData): - """ Looking for pointless function definitions. """ - - _DEF_CLASS = astroid.Function - message_id = "W9952" - - @classmethod - def _retain_node(cls, node, restrict=True): - return not restrict or \ - (len(node.body) == 1 and isinstance(node.body[0], astroid.Pass)) - - @classmethod - def check_equal(cls, node, other): - return len(node.body) == 1 and isinstance(node.body[0], astroid.Pass) and \ - len(other.body) == 1 and isinstance(other.body[0], astroid.Pass) - - @staticmethod - def _extract_value(node): - return node - - @staticmethod - def _extract_targets(node): - yield node.name - -class PointlessAssignment(PointlessData): - - _DEF_CLASS = astroid.Assign - message_id = "W9951" - - _VALUE_CLASSES = ( - astroid.CallFunc, - astroid.Const, - astroid.Dict, - astroid.List, - astroid.Tuple - ) - - @classmethod - def _retain_node(cls, node, restrict=True): - return not restrict or isinstance(node.value, cls._VALUE_CLASSES) - - @classmethod - def check_equal(cls, node, other): - if type(node) != type(other): - return False - if isinstance(node, astroid.Const): - return node.value == other.value - if isinstance(node, (astroid.List, astroid.Tuple)): - return len(node.elts) == len(other.elts) and \ - all(cls.check_equal(n, o) for (n, o) in zip(node.elts, other.elts)) - if isinstance(node, astroid.Dict): - return len(node.items) == len(other.items) and \ - all(cls.check_equal(n, o) for (n, o) in zip(sorted(node.items), sorted(other.items))) - if isinstance(node, astroid.CallFunc): - if type(node.func) != type(other.func): - return False - - if isinstance(node.func, astroid.Getattr): - if node.func.attrname != other.func.attrname or \ - type(node.func.expr) != type(other.func.expr) or \ - node.func.expr.name != other.func.expr.name: - return False - elif isinstance(node.func, astroid.Name): - if node.func.name != other.func.name: - return False - else: - return False - - return node.starargs == other.starargs and \ - node.kwargs == other.kwargs and \ - node.args == other.args - - return False - - @staticmethod - def _extract_value(node): - return node.value - - @staticmethod - def _extract_targets(node): - for target in node.targets: - yield target.name - -class PointlessClassAttributeOverrideChecker(BaseChecker): - """ If the nearest definition of the class attribute in the MRO assigns - it the same value, then the overriding definition is said to be - pointless. - - The algorithm for detecting a pointless attribute override is the following. - - * For each class, C: - - For each attribute assignment, - name_1 = name_2 ... name_n = l (where l is a literal): - * For each n in (n_1, n_2): - - Traverse the linearization of the MRO until the first - matching assignment n = l' is identified. If l is equal to l', - then consider that the assignment to l in C is a - pointless override. - - The algorithm for detecting a pointless method override has the same - general structure, and the same defects discussed below. - - Note that this analysis is neither sound nor complete. It is unsound - under multiple inheritance. Consider the following class hierarchy:: - - class A(object): - _attrib = False - - class B(A): - _attrib = False - - class C(A): - _attrib = True - - class D(B,C): - pass - - In this case, starting from B, B._attrib = False would be considered - pointless. However, for D the MRO is B, C, A, and removing the assignment - B._attrib = False would change the inherited value of D._attrib from - False to True. - - The analysis is incomplete because it will find some values unequal when - actually they are equal. - - The analysis is both incomplete and unsound because it expects that - assignments will always be made by means of the same syntax. - """ - - __implements__ = (IAstroidChecker,) - - name = "pointless class attribute override checker" - msgs = { - "W9951": - ( - "Assignment to class attribute %s overrides identical assignment in ancestor %s.", - "pointless-class-attribute-override", - "Assignment to class attribute that overrides assignment in ancestor that assigns identical value has no effect." - ), - "W9952": - ( - "Definition of %s method overrides identical method definition in ancestor %s.", - "pointless-method-definition-override", - "Overriding empty method definition with another empty method definition has no effect." - ) - } - - @check_messages("W9951", "W9952") - def visit_class(self, node): - for checker in (PointlessAssignment, PointlessFunctionDefinition): - for (name, value) in checker.get_data(node): - for a in node.ancestors(): - match = next((v for (n, v) in checker.get_data(a, False) if n == name), None) - if match is not None: - if checker.check_equal(value, match): - self.add_message(checker.message_id, node=value, args=(name,a.name)) - break - -def register(linter): - linter.register_checker(PointlessClassAttributeOverrideChecker(linter)) diff --git a/tests/pylint/pylint-false-positives b/tests/pylint/pylint-false-positives deleted file mode 100644 index 9037510..0000000 --- a/tests/pylint/pylint-false-positives +++ /dev/null @@ -1,28 +0,0 @@ -^(blivet|tests)/.*: \[E1101$no-member$.*\] Instance of 'int' has no .* member -^blivet/.*: \[E0611$no\-name\-in\-module$.*\] No name 'blockdev' in module 'gi.repository'$ -^blivet/.*: \[E0712$catching\-non\-exception$.*\] Catching an exception which doesn't inherit from BaseException: (BlockDev|DM|Crypto|Swap|LVM|Btrfs|MDRaid|)Error$ -^blivet/.*: \[E0712$catching\-non\-exception$.*\] Catching an exception which doesn't inherit from BaseException: GError$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'BootLoaderError' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'ERROR_RAISE' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'ROOT_PATH' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'errorHandler' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'get_bootloader' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'isys' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'iutil' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'productName' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0602$global\-variable\-not\-assigned$, enable_installer_mode\] Using global for 'shortProductName' but no assignment is done$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'ERROR_RAISE'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'ROOT_PATH'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'errorHandler'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'get_bootloader'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'isys'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'productName'$ -^blivet/__init__.py:[[:digit:]]+: \[W0612$unused\-variable$, enable_installer_mode\] Unused variable 'shortProductName'$ -^blivet/arch.py:[[:digit:]]+: \[F0401$import\-error$, numBits\] Unable to import 'platform'$ -^blivet/blivet.py:[[:digit:]]+: \[E1101$no\-member$, Blivet.savePassphrase\] Instance of 'DeviceTree' has no '_DeviceTree__luksDevs' member$ -^blivet/blivet.py:[[:digit:]]+: \[E1101$no\-member$, Blivet.savePassphrase\] Instance of 'DeviceTree' has no '_DeviceTree__passphrases' member$ -^blivet/formats/__init__.py:[[:digit:]]+: \[W0201$attribute\-defined\-outside\-init$, DeviceFormat._setDevice\] Attribute '_device' defined outside __init__$ -^dm.c: [[:digit:]]+: not running as root returning empty list$ -^tests/devicelibs_test/mdraid_test.py:[[:digit:]]+: \[E1003$bad\-super\-call$, ([[:alnum:]\.]+).__init__\] Bad first argument YES given to super$ -^tests/devicelibs_test/raid_test.py:[[:digit:]]+: \[E1120$no\-value\-for\-parameter$, [[:alnum:]\.]+\] No value [[:alnum:] ]+ 'member_count' in [[:alnum:] ]+ call$ -^tests/devicelibs_test/raid_test.py:[[:digit:]]+: \[E1120$no\-value\-for\-parameter$, [[:alnum:]\.]+\] No value [[:alnum:] ]+ 'smallest_member_size' in [[:alnum:] ]+ call$ diff --git a/tests/pylint/pylint-one.sh b/tests/pylint/pylint-one.sh deleted file mode 100755 index 9b1e79b..0000000 --- a/tests/pylint/pylint-one.sh +++ /dev/null @@ -1,33 +0,0 @@ -#!/bin/bash -# -# $1 -- python source to run pylint on -# - -if [ $# -lt 1 ]; then - # no source, just exit - exit 1 -fi - -file_suffix="$(eval echo \$$#|sed s?/?_?g)" - -pylint_output="$(pylint \ - --msg-template='{path}:{line}: [{msg_id}({symbol}), {obj}] {msg}' \ - -r n --disable=C,R --rcfile=/dev/null \ - --dummy-variables-rgx=_ \ - --ignored-classes=Popen,TransactionSet \ - --defining-attr-methods=__init__,_grabObjects,initialize,reset,start,setUp \ - --load-plugins=intl,pointless-override \ - $DISABLED_WARN_OPTIONS \ - $DISABLED_ERR_OPTIONS \ - $NON_STRICT_OPTIONS "$@" 2>&1 | \ - egrep -v -f "$FALSE_POSITIVES" \ - )" - -# I0011 is the informational "Locally disabling ...." message -if [ -n "$(echo "$pylint_output" | fgrep -v '************* Module ' |\ - grep -v '^I0011:')" ]; then - # Replace the Module line with the actual filename - pylint_output="$(echo "$pylint_output" | sed "s|\* Module .*|* Module $(eval echo \$$#)|")" - echo "$pylint_output" > pylint-out_$file_suffix - touch "pylint-$file_suffix-failed" -fi diff --git a/tests/pylint/runpylint.py b/tests/pylint/runpylint.py new file mode 100755 index 0000000..859c3ea --- /dev/null +++ b/tests/pylint/runpylint.py @@ -0,0 +1,34 @@ +#!/usr/bin/python3 + +import os + +from pocketlint import FalsePositive, PocketLintConfig, PocketLinter + +class BlivetLintConfig(PocketLintConfig): + def __init__(self): + PocketLintConfig.__init__(self) + + self.falsePositives = [ FalsePositive(r"Catching an exception which doesn't inherit from BaseException: (BlockDev|DM|Crypto|Swap|LVM|Btrfs|MDRaid|)Error$"), + FalsePositive(r"Instance of 'int' has no .* member$"), + FalsePositive(r"No value for argument 'member_count' in unbound method call$"), + FalsePositive(r"No value for argument 'smallest_member_size' in unbound method call$"), + ] + + @property + def disabledOptions(self): + return [ "W0105", # String statement has no effect + "W0110", # map/filter on lambda could be replaced by comprehension + "W0141", # Used builtin function %r + "W0142", # Used * or ** magic + "W0212", # Access to a protected member of a client class + "W0511", # Used when a warning note as FIXME or XXX is detected. + "W0603", # Using the global statement + "W0614", # Unused import %s from wildcard import + "I0011", # Locally disabling %s + ] + +if __name__ == "__main__": + conf = BlivetLintConfig() + linter = PocketLinter(conf) + rc = linter.run() + os._exit(rc) diff --git a/tests/pylint/runpylint.sh b/tests/pylint/runpylint.sh deleted file mode 100755 index 18200bf..0000000 --- a/tests/pylint/runpylint.sh +++ /dev/null @@ -1,125 +0,0 @@ -#!/bin/bash - -# This script will check for any pylint warning and errors using a set -# of options minimizing false positives, in combination with filtering of any -# warning regularexpressions listed in pylint-false-positives. -# -# If any warnings are found they will be stored in pylint-log and printed -# to stdout and this script will exit with a status of 1, if no (non filtered) -# warnings are found it exits with a status of 0 - -# If $top_srcdir is set, assume this is being run from automake and we don't -# need to keep a separate log -export pylint_log=0 -if [ -z "$top_srcdir" ]; then - export pylint_log=1 -fi - -# Unset TERM so that things that use readline don't output terminal garbage -unset TERM - -# Don't try to connect to the accessibility socket -export NO_AT_BRIDGE=1 - -# If $top_srcdir has not been set by automake, import the test environment -if [ -z "$top_srcdir" ]; then - top_srcdir="$(dirname "$0")/../.." - . ${top_srcdir}/tests/testenv.sh -fi - -. ${top_srcdir}/tests/lib/testlib.sh - -srcdir="${top_srcdir}/tests/pylint" -builddir="${top_builddir}/tests/pylint" - -# Need to add the pylint module directory to PYTHONPATH as well. -export PYTHONPATH="${PYTHONPATH}:${srcdir}" - -# Save analysis data in the pylint directory -export PYLINTHOME="${builddir}/.pylint.d" -[ -d "$PYLINTHOME" ] || mkdir "$PYLINTHOME" - -export FALSE_POSITIVES="${srcdir}"/pylint-false-positives - -# W0212 - Access to a protected member %s of a client class -export NON_STRICT_OPTIONS="--disable=W0212" - -# E1103 - %s %r has no %r member (but some types could not be inferred) -export DISABLED_ERR_OPTIONS="--disable=E1103" - -# W0105 - String statement has no effect -# W0110 - map/filter on lambda could be replaced by comprehension -# W0141 - Used builtin function %r -# W0142 - Used * or ** magic -# W0511 - Used when a warning note as FIXME or XXX is detected. -# W0603 - Using the global statement -# W0614 - Unused import %s from wildcard import -# I0011 - Locally disabling %s (i.e., pylint: disable) -export DISABLED_WARN_OPTIONS="--disable=W0105,W0110,W0141,W0142,W0511,W0603,W0614,I0011" - -usage () { - echo "usage: `basename $0` [--strict] [--help] [files...]" - exit $1 -} - -# Separate the module parameters from the files list -ARGS= -FILES= -while [ $# -gt 0 ]; do - case $1 in - --strict) - export NON_STRICT_OPTIONS= - ;; - --help) - usage 0 - ;; - -*) - ARGS="$ARGS $1" - ;; - *) - FILES=$@ - break - esac - shift -done - -exit_status=0 - -if [ -s pylint-log ]; then - rm pylint-log -fi - -# run pylint one file / module at a time, otherwise it sometimes gets -# confused -if [ -z "$FILES" ]; then - # Test any file that either ends in .py or contains #!/usr/bin/python in - # the first line. Scan everything except old_tests - FILES=$(findtestfiles $ -name '*.py' -o \ - -exec /bin/sh -c "head -1 {} | grep -q '#!/usr/bin/python'" \; $ -print | \ - egrep -v '(|/)doc/conf.py') -fi - -num_cpus=$(getconf _NPROCESSORS_ONLN) -# run pylint in paralel -echo $FILES | xargs --max-procs=$num_cpus -n 1 "$srcdir"/pylint-one.sh $ARGS || exit 1 - -for file in $(find -name 'pylint-out*'); do - cat "$file" >> pylint-log - rm "$file" -done - -fails=$(find -name 'pylint*failed' -print -exec rm '{}' \;) -if [ -z "$fails" ]; then - exit_status=0 -else - exit_status=1 -fi - -if [ -s pylint-log ]; then - echo "pylint reports the following issues:" - cat pylint-log -elif [ -e pylint-log ]; then - rm pylint-log -fi - -exit "$exit_status" -- 2.2.2

2 21

Updated: [rhinstaller/blivet/pulls/93 master] Fix adding new partitions on disks with active devices
by dashea 29 Apr '15

29 Apr '15

See https://bugzilla.redhat.com/show_bug.cgi?id=1212841 Currently is not possible to add new partitions to disk with an existing LVM VG. 'findActiveDevicesOnActionDisks' should disallow only actions involving changes on the disklabel not adding new partitions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/93

1 5

New: [rhinstaller/blivet/pulls/96 master] Master lvm
by dashea 29 Apr '15

29 Apr '15

Correct a bug where only lvm logical volumes and physical volumes where looked up correctly in devicetree. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/96

1 6

New: [rhinstaller/blivet/pulls/100 master] Un-escape '-'s in names or paths for _all_ lvm lv or vgs.
by dashea 28 Apr '15

28 Apr '15

Assume that whenever lvm presents blivet w/ a name with two consecutive dashes this represents an escaped single dash. Previously only doing this for lvmvg's and lvmlv's, which turned out to be a problem when looking up subclass instances, like lvm thin lvs. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/100

1 4

Updated: [rhinstaller/anaconda/pulls/86 master] Master docs
by dashea 28 Apr '15

28 Apr '15

I've been trying to figure out how to create good docs for pyanaconda using Read the Docs, and have hit a dead end. RTD depends on the autodoc feature in Sphinx, but can't install anything other than python projects with a setup.py. You can mock various objects like I did in the previous pass at this, but that's fragile and still doesn't result in complete docs. So, I propose that we generate the docs on a system with all the build deps installed or available via repos so that they are complete and nice looking, and that we carry the html output in the repo. This allows us to either server it up ourselves from someplace, or use gh-pages to serve it from github (which is what I suggest). My understanding is you create a gh-pages branch for the project, then copy the html into the root of that branch (so it really isn't a branch, just a place to store the html for github). I'd suggest doing this manually so that the output can be checked for import problems. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/86

1 6

New: [rhinstaller/anaconda/pulls/88 f22-branch] Configure proxy settings for dnf payload (#1211122)
by dashea 28 Apr '15

28 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/88

1 7

New: [rhinstaller/blivet/pulls/82 rhel7-branch] Include LUKSDevice information in kickstart data (#1139222)
by dashea 28 Apr '15

28 Apr '15

Resolves: rhbz#1139222 Previously, if a device had a child that was a LUKS device that child was ignored when generating ksdata. What this meant was that that device's format, which was luks, was used in generating the kickstart data. Now, the LUKSDevice, which contains the format that should appear in the ks output is used. This means that the LUKS format is used to set some ks data, but the mountpoint and so forth are overwritten with the correct values when the format on the LUKSDevice is used to populate those values. The --passphrase option with passphrase continues to not appear in the output anaconda-ks.cfg file. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/82

1 8

New: [rhinstaller/blivet/pulls/88 rhel7-branch] If the parent volume has a label, use it in subvol's kickstart (#1072060...
by dashea 28 Apr '15

28 Apr '15

...) Resolves: rhbz#1072060 The LABEL= entry identifies the parent of the specified subvolume by its label. If the parent volume does not have a label, then no entry is made that identifies the parent volume and the generated kickstart file will be rejected by ksvalidator. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/88

1 4

New: [rhinstaller/blivet/pulls/94 master] Move definition of _level attribute above super-constructor call.
by dashea 28 Apr '15

28 Apr '15

The basic problem is that level is a very important concept for md so a lot of log messages log this information. Some of these log messages are executed as a result of running the parent constructor. The particular problem I saw is that self.updateSize() is called by StorageDevice.__init__, which understandably logs the size, and the size() methods logs its result, along with the level, leading to an AttributeError. Now there will be no AttributeError, but the RAID level will be displayed as None when this happens. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/94

1 4

[Patch for Review] iSCSI Name Validation using regex.
by Sujith Pandel 27 Apr '15

27 Apr '15

Hi All, The following patch is to make use of a regular expression to validate the user given iSCSI Initiator name in Anaconda, while selecting iSCSI disks during OS installation. I have tested it for cases which I could think of at this time. Requesting review/feedback and inclusion. Thanks and Regards, Sujith Pandel When adding iSCSI target disk during installation, the iSCSI Initiator name and the Target IP is validated to enable/disable the "Start Discovery" button on the page. Currently, the validation of iSCSI Initiator Name is based on the existence of dots only (".") and this can be more efficiently done by using a regular expression. The conditions for iSCSI initiator name used in this regex are (https://tools.ietf.org/html/rfc3720#section-3.2.6.3.1 , https://tools.ietf.org/html/rfc3721#page-5 and http://standards.ieee.org/regauth/oui/tutorials/EUI64.html): 1. For iqn format: a. Starts with string 'iqn.' b. A date code specifying the year and month in which the organization registered the domain or sub-domain name used as the naming authority string. “yyyy-mm” c. A dot (".") d. The organizational naming authority string, which consists of a valid, reversed domain or subdomain name. e. Optionally, a colon (":"), followed by a string of the assigning organization's choosing, which must make each assigned iSCSI name unique. With the exception of the colon prefix, the owner of the domain name can assign everything after the reversed domain name as desired. 2. For eui format: The format is "eui." followed by an EUI-64 identifier (16 ASCII-encoded hexadecimal digits). Signed-off-by: Sujith Pandel <sujithpshankar(a)gmail.com> --- pyanaconda/regexes.py | 8 ++++++++ pyanaconda/ui/gui/spokes/advstorage/iscsi.py | 18 ++++++++++++++++-- 2 files changed, 24 insertions(+), 2 deletions(-) diff --git a/pyanaconda/regexes.py b/pyanaconda/regexes.py index b4e52a3..f153915 100644 --- a/pyanaconda/regexes.py +++ b/pyanaconda/regexes.py @@ -135,3 +135,11 @@ REPO_NAME_VALID = re.compile(r'^[a-zA-Z0-9_.:-]+$') # Product Version string, just the starting numbers like 21 or 21.1 VERSION_DIGITS = r'([\d.]+)' + + +#iSCSI Naming Standards: RFC 3720 and RFC 3721 +ISCSI_IQN_NAMING_FORMAT = re.compile(r'^iqn\.\d{4}-\d{2}((?<!-)\.(?!-)[a-zA-Z0-9\-]+){1,63}(?<!-)(?<!\.)(:[^:]+)?') + +ISCSI_EUI_NAMING_FORMAT = re.compile(r'^eui\.[a-fA-F0-9]{16}') + + diff --git a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py index 22e11c3..764e81f 100644 --- a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py +++ b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py @@ -18,7 +18,7 @@ # # Red Hat Author(s): Chris Lumens <clumens(a)redhat.com> # - +import re from IPy import IP from collections import namedtuple from gi.repository import GLib @@ -29,6 +29,9 @@ from pyanaconda.ui.gui import GUIObject from pyanaconda.ui.gui.utils import escape_markup from pyanaconda.i18n import _ from pyanaconda import nm +from pyanaconda.regexes import ISCSI_IQN_NAMING_FORMAT +from pyanaconda.regexes import ISCSI_EUI_NAMING_FORMAT + __all__ = ["ISCSIDialog"] @@ -303,7 +306,18 @@ class ISCSIDialog(GUIObject): stripped = text.strip() #iSCSI Naming Standards: RFC 3720 and RFC 3721 - return "." in stripped + if stripped.startswith('iqn.'): + initiator_name_match = re.match(ISCSI_IQN_NAMING_FORMAT,stripped) + elif stripped.startswith('eui.'): + initiator_name_match = re.match(ISCSI_EUI_NAMING_FORMAT,stripped) + else: + return False + + if initiator_name_match and (len(initiator_name_match.group(0)) == len(stripped)): + return True + else: + return False + def on_discover_field_changed(self, *args): # Make up a credentials object so we can test if it's valid. -- 2.1.0

2 1

[blivet][master] Python3-ready pyparted needed
by Vratislav Podzimek 27 Apr '15

27 Apr '15

Blivet needs newer pyparted to provide real Python 3 compatibility. The pyparted package is not built yet, so I'll keep holding this patch off until that changes. Vratislav Podzimek (1): Require new version of pyparted with Python 3 related fixes python-blivet.spec | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.1.0

2 2

Updated: [rhinstaller/blivet/pulls/87 rhel7-branch] Fix "anaconda hangs while trying to discover iscsi..." (#1166652)
by dashea 27 Apr '15

27 Apr '15

This was 2 errors in one. First problem: anaconda freeze for some time. Second problem: libiscsi was run on thread and libiscsi the library is not thread safe. Anaconda freeze is because of GIL is taken on library thread which is trying to connect to server. The thread safe problem is because libiscsi is using signals internally which were delivered to another thread (this is hard to solve in python threads...). Solution is to start discover method on new process. So new process is now created to call libiscsi discover_sendtargets method. This should fix both problems. Note: Will be applied to master too. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/87

1 7

Updated: [rhinstaller/blivet/pulls/93 master] Fix adding new partitions on disks with active devices
by dashea 26 Apr '15

26 Apr '15

See https://bugzilla.redhat.com/show_bug.cgi?id=1212841 Currently is not possible to add new partitions to disk with an existing LVM VG. 'findActiveDevicesOnActionDisks' should disallow only actions involving changes on the disklabel not adding new partitions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/93

1 6

[python-meh][master/f22-branch/rhel7-branch][PATCH] Make sure the date in RPM changelog is always in English
by Martin Kolman 24 Apr '15

24 Apr '15

Make sure that the date in the RPM changelog is always in English regardless of system locale as various RPM tools don't really handle localized dates. Signed-off-by: Martin Kolman <mkolman(a)redhat.com> --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index cb5675e..910d850 100644 --- a/Makefile +++ b/Makefile @@ -73,7 +73,7 @@ bumpver: potfile zanata push $(ZANATA_PUSH_ARGS) || ( echo "zanata push failed"; exit 1 ) @NEWSUBVER=$$((`echo $(VERSION) |cut -d . -f 2` + 1)) ; \ NEWVERSION=`echo $(VERSION).$$NEWSUBVER |cut -d . -f 1,3` ; \ - DATELINE="* `date "+%a %b %d %Y"` `git config user.name` <`git config user.email`> - $$NEWVERSION-1" ; \ + DATELINE="* `LANG=c date "+%a %b %d %Y"` `git config user.name` <`git config user.email`> - $$NEWVERSION-1" ; \ cl=`grep -n %changelog python-meh.spec |cut -d : -f 1` ; \ tail --lines=+$$(($$cl + 1)) python-meh.spec > speclog ; \ (head -n $$cl python-meh.spec ; echo "$$DATELINE" ; make --quiet rpmlog 2>/dev/null ; echo ""; cat speclog) > python-meh.spec.new ; \ -- 2.3.5

2 1

New: [rhinstaller/anaconda/pulls/74 f22-branch] Fix issues with disk selection changes (#1166598)
by dashea 24 Apr '15

24 Apr '15

**Please ignore the first two commits, those have already been ACKed on the mailing list and just were not pushed to f22-branch yet (I'm waiting for Beta to be declared GO)** The first of the latter four commits make the ``on_back_clicked`` method of the *StorageSpoke* less complicated because it has been bloating with if's and duplicated chunks of code for quite some time into the shape that was really hard to understand not even modify. To review this commit I highly recommend to check the result instead of/before going through the changes. The second commit is just a minor change simplifying the logic a bit. The last two commits implement and use a snapshot of on-disk storage as we get it from scanning disks before we do any modifications so that we can revert everything to it in case user changes their mind about disk selection. The previous "limbo" with hiding/unhiding disks and relying on the result of cancelling actions on the way being the original state of everything was brave, but not working. However, it still is the right thing to do and that's why this PR is only for the f22-branch with quite a bit of hope that action cancelling will have been fixed in blivet by the time of F23. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/74

3 27

New: [rhinstaller/blivet/pulls/93 master] Fix adding new partitions on disks with active devices
by dashea 24 Apr '15

24 Apr '15

See https://bugzilla.redhat.com/show_bug.cgi?id=1212841 Currently is not possible to add new partitions to disk with an existing LVM VG. 'findActiveDevicesOnActionDisks' should disallow only actions involving changes on the disklabel not adding new partitions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/93

1 7

[Patch for Review] iSCSI Name Validation using regex.
by Sujith Pandel 23 Apr '15

23 Apr '15

Hi All, The following patch is to make use of a regular expression to validate the user given iSCSI Initiator name in Anaconda, while selecting iSCSI disks during OS installation. I have tested it for cases which I could think of at this time. Requesting review/feedback and inclusion. Thanks and Regards, Sujith Pandel When adding iSCSI target disk during installation, the iSCSI Initiator name and the Target IP is validated to enable/disable the "Start Discovery" button on the page. Currently, the validation of iSCSI Initiator Name is based on the existence of dots only (".") and this can be more efficiently done by using a regular expression. The conditions for iSCSI initiator name used in this regex are (https://tools.ietf.org/html/rfc3720#section-3.2.6.3.1 , https://tools.ietf.org/html/rfc3721#page-5 and http://standards.ieee.org/regauth/oui/tutorials/EUI64.html): 1. For iqn format: a. Starts with string 'iqn.' b. A date code specifying the year and month in which the organization registered the domain or sub-domain name used as the naming authority string. “yyyy-mm” c. A dot (".") d. The organizational naming authority string, which consists of a valid, reversed domain or subdomain name. e. Optionally, a colon (":"), followed by a string of the assigning organization's choosing, which must make each assigned iSCSI name unique. With the exception of the colon prefix, the owner of the domain name can assign everything after the reversed domain name as desired. 2. For eui format: The format is "eui." followed by an EUI-64 identifier (16 ASCII-encoded hexadecimal digits). Signed-off-by: Sujith Pandel <sujithpshankar(a)gmail.com> --- pyanaconda/regexes.py | 8 ++++++++ pyanaconda/ui/gui/spokes/advstorage/iscsi.py | 18 ++++++++++++++++-- 2 files changed, 24 insertions(+), 2 deletions(-) diff --git a/pyanaconda/regexes.py b/pyanaconda/regexes.py index b4e52a3..f153915 100644 --- a/pyanaconda/regexes.py +++ b/pyanaconda/regexes.py @@ -135,3 +135,11 @@ REPO_NAME_VALID = re.compile(r'^[a-zA-Z0-9_.:-]+$') # Product Version string, just the starting numbers like 21 or 21.1 VERSION_DIGITS = r'([\d.]+)' + + +#iSCSI Naming Standards: RFC 3720 and RFC 3721 +ISCSI_IQN_NAMING_FORMAT = re.compile(r'^iqn\.\d{4}-\d{2}((?<!-)\.(?!-)[a-zA-Z0-9\-]+){1,63}(?<!-)(?<!\.)(:[^:]+)?') + +ISCSI_EUI_NAMING_FORMAT = re.compile(r'^eui\.[a-fA-F0-9]{16}') + + diff --git a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py index 22e11c3..764e81f 100644 --- a/pyanaconda/ui/gui/spokes/advstorage/iscsi.py +++ b/pyanaconda/ui/gui/spokes/advstorage/iscsi.py @@ -18,7 +18,7 @@ # # Red Hat Author(s): Chris Lumens <clumens(a)redhat.com> # - +import re from IPy import IP from collections import namedtuple from gi.repository import GLib @@ -29,6 +29,9 @@ from pyanaconda.ui.gui import GUIObject from pyanaconda.ui.gui.utils import escape_markup from pyanaconda.i18n import _ from pyanaconda import nm +from pyanaconda.regexes import ISCSI_IQN_NAMING_FORMAT +from pyanaconda.regexes import ISCSI_EUI_NAMING_FORMAT + __all__ = ["ISCSIDialog"] @@ -303,7 +306,18 @@ class ISCSIDialog(GUIObject): stripped = text.strip() #iSCSI Naming Standards: RFC 3720 and RFC 3721 - return "." in stripped + if stripped.startswith('iqn.'): + initiator_name_match = re.match(ISCSI_IQN_NAMING_FORMAT,stripped) + elif stripped.startswith('eui.'): + initiator_name_match = re.match(ISCSI_EUI_NAMING_FORMAT,stripped) + else: + return False + + if initiator_name_match and (len(initiator_name_match.group(0)) == len(stripped)): + return True + else: + return False + def on_discover_field_changed(self, *args): # Make up a credentials object so we can test if it's valid. -- 2.1.0

1 0

Updated: [rhinstaller/anaconda/pulls/85 master] Ignore %anaconda section in parse-kickstart
by dashea 23 Apr '15

23 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/85

1 6

New: [rhinstaller/anaconda/pulls/84 master] Master pxe stage2
by dashea 23 Apr '15

23 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/84

1 6

New: [rhinstaller/anaconda/pulls/75 master] Add support for chrony pool directive
by dashea 23 Apr '15

23 Apr '15

With chrony-2.0 it is possible to specify a pool of NTP servers instead of individual servers. It selects four servers from the addresses that the pool name is resolved to and replaces them automatically when they are unreachable or falsetickers. This is now used in the default Fedora chrony configuration. To avoid extending the ntp API and the date time spoke, pools are simply presented as four separate servers with the same name. Resolves: rhbz#1193389 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/75

1 17

New: [rhinstaller/anaconda/pulls/86 master] Master docs
by dashea 23 Apr '15

23 Apr '15

I've been trying to figure out how to create good docs for pyanaconda using Read the Docs, and have hit a dead end. RTD depends on the autodoc feature in Sphinx, but can't install anything other than python projects with a setup.py. You can mock various objects like I did in the previous pass at this, but that's fragile and still doesn't result in complete docs. So, I propose that we generate the docs on a system with all the build deps installed or available via repos so that they are complete and nice looking, and that we carry the html output in the repo. This allows us to either server it up ourselves from someplace, or use gh-pages to serve it from github (which is what I suggest). My understanding is you create a gh-pages branch for the project, then copy the html into the root of that branch (so it really isn't a branch, just a place to store the html for github). I'd suggest doing this manually so that the output can be checked for import problems. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/86

1 2

[initial-setup][master/f22-branch][PATCH 0/2] Makefile cleanup and improvements
by Martin Kolman 23 Apr '15

23 Apr '15

The first patch removes some obsolete stuff from the Initial Setup Makefile, the second patch improves the Makefile with code mostly based on what the python-meh Makefile does. Martin Kolman (2): Remove old GUI testing code from the Makefile Improve the Makefile Makefile | 115 +++++++++++++++++++++++++++++------------------------ initial-setup.spec | 15 ++++--- po/Makefile | 2 +- 3 files changed, 73 insertions(+), 59 deletions(-) -- 2.3.5

2 3

New: [rhinstaller/anaconda/pulls/81 rhel7-branch] Change of label in iscsi storage spoke
by dashea 23 Apr '15

23 Apr '15

It may takes some time to discover targets when "Bind target to network interfaces" is checked. User is now warn about this possibility. It's related to (#1166652) which is fixed in blivet. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/81

1 8

New: [rhinstaller/anaconda/pulls/85 master] Ignore %anaconda section in parse-kickstart
by dashea 22 Apr '15

22 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/85

1 4

New: [rhinstaller/anaconda/pulls/82 master] Fix doReqPartition import from autopart
by dashea 22 Apr '15

22 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/82

1 3

New: [rhinstaller/blivet/pulls/89 master] Fix BlockDev import in populator.py
by dashea 22 Apr '15

22 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/89

1 4

New: [rhinstaller/blivet/pulls/87 rhel7-branch] Fix "anaconda hangs while trying to discover iscsi..." (#1166652)
by dashea 22 Apr '15

22 Apr '15

This was 2 errors in one. First problem: anaconda freeze for some time. Second problem: libiscsi was run on thread and libiscsi the library is not thread safe. Anaconda freeze is because of GIL is taken on library thread which is trying to connect to server. The thread safe problem is because libiscsi is using signals internally which were delivered to another thread (this is hard to solve in python threads...). Solution is to start discover method on new process. So new process is now created to call libiscsi discover_sendtargets method. This should fix both problems. Note: Will be applied to master too. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/87

1 10

New: [rhinstaller/blivet/pulls/86 master] Master tasks 20140417 a
by dashea 22 Apr '15

22 Apr '15

Changes from previous: removed unsetupable* methods and resizeSupported method since it seems to fall into the same category of unnecessary methods. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/86

1 31

New: [rhinstaller/blivet/pulls/79 master] Master tasks 20140414
by dashea 22 Apr '15

22 Apr '15

Another revision of PR for issue #12. I didn't rebase against master because it's in a tricky state right now. Changes are: * Remove Task classmethod available(), since nobody seems to want it (3d668a6) * Use defined constants for any external resource objects that are instantiated more than once (667ff92) * Change the return type for un*, and impossible methods as well as their names (fc61f7a) -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/79

1 31

New: [rhinstaller/blivet/pulls/83 master] Master tasks 20140416
by dashea 22 Apr '15

22 Apr '15

The only significant changes are the major effects of the rebase and the final commit. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/83

1 32

New: [rhinstaller/blivet/pulls/90 master] Master tasks 20150422
by dashea 22 Apr '15

22 Apr '15

Now with fewer commits. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/90

1 13

[rhel6-branch] Check for FBA DASDs so we don't try to format them with dasdfmt. (#1144979)
by Samantha N. Bueno 21 Apr '15

21 Apr '15

dasdfmt fails if you try to run it against an FBA DASD. Resolves: rhbz#1144979 --- isys/dasd.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/isys/dasd.c b/isys/dasd.c index 6d59742..259d9f8 100644 --- a/isys/dasd.c +++ b/isys/dasd.c @@ -92,6 +92,9 @@ int isLdlDasd(char *device) { if (ret == 2) return 0; + if (dasd_info.FBA_layout) + return 0; + if (dasd_info.format == DASD_FORMAT_CDL) /* VOL1, CDL */ return 0; else -- 1.9.3

2 3

New: [rhinstaller/anaconda/pulls/76 master] Master kexec reboot
by dashea 21 Apr '15

21 Apr '15

This implements rebooting into the new system using kexec to speed things up. Also depends on new pykickstart with reboot --kexec -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/76

1 12

New: [rhinstaller/anaconda/pulls/80 master] Add kickstart %pre-install section support
by dashea 21 Apr '15

21 Apr '15

The %pre-install section will run scripts after all the filesystems have been setup and mounted on /mnt/sysimage, but before any packages have been installed. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/80

1 14

New: [rhinstaller/anaconda/pulls/79 master] Remove the custom help button from the toolbar
by dashea 21 Apr '15

21 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/79

1 6

New: [rhinstaller/blivet/pulls/75 master] Master tasks 20150410
by dashea 21 Apr '15

21 Apr '15

New version for issue #12 . Lots of minor changes, also: * Added ability to check package version for some applications where the spec requires a version but it has to be obtained from the package. Remembered to remove version info from spec file where appropriate. Added requires of "rpm" to specfile. * Caught exception when failing to load plugin, since init methods raise one if fail to load a requested plugin. * Got rid of checking for dependencies by means of modes, thus got rid of external.py file and some other stuff. * Call StorageDevice.resizeSupported from StorageDevice.resizable instead of from ActionResizeDevice constructor. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/75

1 31

New: [rhinstaller/anaconda/pulls/65 f22-branch] translations and unicode
by dashea 21 Apr '15

21 Apr '15

Fix live not being translated and also crashing when it is translated. Also needs https://github.com/rhinstaller/blivet/pull/73 I managed to get through a live install in Russian with this but I'm sure to have missed something. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/65

3 15

New: [rhinstaller/blivet/pulls/73 f22-branch] F22 unicode
by dashea 21 Apr '15

21 Apr '15

The second commit might need to be ported to master in some form -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/73

1 8

[initial-setup][master/f22-branch/rhel7-branch][PATCH] Update upstream URL
by Martin Kolman 21 Apr '15

21 Apr '15

The python-meh upstream source code repository has moved to the rhinstaller organisation on GitHub. Signed-off-by: Martin Kolman <mkolman(a)redhat.com> --- python-meh.spec | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/python-meh.spec b/python-meh.spec index 0b7facb..205cfb9 100644 --- a/python-meh.spec +++ b/python-meh.spec @@ -5,14 +5,14 @@ Summary: A python library for handling exceptions Name: python-meh -Url: http://git.fedorahosted.org/git/?p=python-meh.git +Url: https://github.com/rhinstaller/python-meh Version: 0.39 Release: 1%{?dist} # This is a Red Hat maintained package which is specific to # our distribution. Thus the source is only available from # within this srpm. # This tarball was created from upstream git: -# git clone git://git.fedoraproject.org/git/python-meh.git +# git clone https://github.com/rhinstaller/python-meh # cd python-meh && make archive Source0: %{name}-%{version}.tar.gz -- 2.1.0

3 10

Updated: [rhinstaller/anaconda/pulls/46 master] Use proxy when configured for the base repo (#1196953)
by dashea 20 Apr '15

20 Apr '15

Signed-off-by: Seth Jennings <sjenning(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/46

1 4

New: [rhinstaller/anaconda/pulls/78 master] Use multiple streams for zRAM instead of multiple devices
by dashea 20 Apr '15

20 Apr '15

This simplifies our scripts for setting up/tearing down zRAM swap and the result should be the same as multiple compression streams easily allow concurrency. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/78

1 4

[libblockdev][PATCH] Resolve md symlinks in swap_status (#1212073)
by David Lehman 20 Apr '15

20 Apr '15

This isn't the root cause of the referenced bug, but it does prevent us from deactivating swap on md devices by making them appear to always be active. This should go on master and f22-branch.

3 4

[anaconda][master/f22-branch/rhel7-branch]PATCH] Remove old help from custom spoke
by Martin Kolman 20 Apr '15

20 Apr '15

This patch has been been already posted during the built-in-help system patch review but but did not actually make it to master after the review for some reason. So I've rebased it and I'm ready to make sure it ends in the correct branches this time. :) Martin Kolman (1): Remove the old custom partitioning help dialog pyanaconda/ui/gui/spokes/custom.py | 7 +- .../ui/gui/spokes/lib/custom_storage_helpers.glade | 71 -------------------- .../ui/gui/spokes/lib/custom_storage_helpers.py | 78 +--------------------- 3 files changed, 2 insertions(+), 154 deletions(-) -- 2.1.0

3 3

New: [rhinstaller/blivet/pulls/85 f22-branch] Make use of the new libblockdev error reporting
by dashea 20 Apr '15

20 Apr '15

As of version 0.10 libblockdev provides better error reporting if its ErrorProxy instances are used (e.g. blockdev.lvm.lvcreate instead of lvm.lvm_lvcreate where blockdev.lvm is an ErrorProxy instance). Let's make use of this feature and catch native python exceptions defined in libblockdev's python GI overrides instead of GLib.GError. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/85

1 6

[PATCH] Update upstream URL (#1213101)
by Martin Kolman 20 Apr '15

20 Apr '15

There is now an Initial Setup page on the Fedora wiki, so use it as the upstream URL. Signed-off-by: Martin Kolman <mkolman(a)redhat.com> --- initial-setup.spec | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/initial-setup.spec b/initial-setup.spec index 6363716..697b9b7 100644 --- a/initial-setup.spec +++ b/initial-setup.spec @@ -1,6 +1,6 @@ Summary: Initial system configuration utility Name: initial-setup -URL: http://fedoraproject.org/wiki/FirstBoot +URL: https://fedoraproject.org/wiki/InitialSetup Version: 0.3.32 Release: 1%{?dist} -- 2.1.0

2 1

New: [rhinstaller/blivet/pulls/80 master] Add libselinux-python to package dependencies (#1211834)
by dashea 20 Apr '15

20 Apr '15

Blivet imports selinux module but doen't depend on libselinux-python (only anaconda-core depends on that). Resolves https://bugzilla.redhat.com/show_bug.cgi?id=1211834 -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/80

1 16

[initial-setup][master/f22-branch/rhel7-branch][PATCH] Update upstream Git repository URL
by Martin Kolman 20 Apr '15

20 Apr '15

The Initial Setup upstream source code repository has moved to the rhinstaller organisation on GitHub. --- initial-setup.spec | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/initial-setup.spec b/initial-setup.spec index 47e4fa2..6363716 100644 --- a/initial-setup.spec +++ b/initial-setup.spec @@ -9,7 +9,7 @@ Release: 1%{?dist} # # The source is thus available only from within this SRPM # or via direct git checkout: -# git clone git://git.fedorahosted.org/initial-setup.git +# git clone https://github.com/rhinstaller/initial-setup Source0: %{name}-%{version}.tar.gz %define debug_package %{nil} -- 2.1.0

2 1

Updated: [rhinstaller/blivet/pulls/81 master] In FS._postSetup() check the mountpoint options that were actually used.
by dashea 16 Apr '15

16 Apr '15

Should have gone in with commit 10bc41e6f73796298a3a78faaf69dd363429ea53. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/81

1 2

Updated: [rhinstaller/blivet/pulls/66 master] Fix some Python 3 compatibility issues and add the python3 subpackage
by dashea 16 Apr '15

16 Apr '15

Fixes for Python 3 compatibility issues discovered while running tests with Python 3 including hundreds of installations with Python 3 Anaconda. With all that passing in the end, I decided to add the pieces that make sure the python3 subpackage is built too (tested in [Copr builds] (https://copr.fedoraproject.org/coprs/bkabrda/py3anaconda/builds/)). **Disclaimer:** These patches don't try to fix anything else than Python 3 incompatibilities in the most straightforward way. Many changes are subject for future improvements and nicer solutions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/66

1 1

[lorax][master/f22-branch][PATCH] Don't crash on error messages containing Unicode characters
by Martin Kolman 16 Apr '15

16 Apr '15

Signed-off-by: Martin Kolman <mkolman(a)redhat.com> --- src/pylorax/ltmpl.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/pylorax/ltmpl.py b/src/pylorax/ltmpl.py index 5056ed3..28947ce 100644 --- a/src/pylorax/ltmpl.py +++ b/src/pylorax/ltmpl.py @@ -538,7 +538,7 @@ class LoraxTemplateRunner(object): base.do_transaction(display=display) except BaseException as e: logger.error("The transaction process has ended abruptly: %s", e) - queue.put(('quit', str(e))) + queue.put(('quit', unicode(e))) def report_progress(queue): (token, msg) = queue.get() -- 2.3.4

2 1

[rhel6-branch] Include hyperv modules in mk-images (#1132770)
by Samantha N. Bueno 15 Apr '15

15 Apr '15

Necessary for hypver-V installations to be successful. This also fixes a typo from hid-hyperv to hid_hyperv. Resolves: rhbz#1132770 --- scripts/mk-images | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/mk-images b/scripts/mk-images index ad90b70..ce3f23b 100755 --- a/scripts/mk-images +++ b/scripts/mk-images @@ -113,7 +113,7 @@ CRYPTOMODS="sha256_generic cbc xts lrw aes_generic crypto_blkcipher crc32c ecb a PCMCIASOCKMODS="yenta_socket i82365 tcic pcmcia" INFINIBANDMODS="ib_qib mlx4_ib" VMPVMODS="vmxnet3 vmware_balloon vmw_pvsci" -MSPVMODS="hv_storvsc hid-hyperv hv_netvsc hv_vmbus hv_utils ata_piix" +MSPVMODS="hv_storvsc hid_hyperv hv_netvsc hv_vmbus hv_utils ata_piix hyperv_keyboard hyperv_fb" INITRDMODS="$USBMODS $FIREWIREMODS $IDEMODS $SCSIMODS $FSMODS $LVMMODS $RAIDMODS $CRYPTOMODS $COMMONMODS \ $PCMCIASOCKMODS $INFINIBANDMODS $SDMODS $VMPVMODS $MSPVMODS =scsi =net =drm" -- 1.9.3

2 1

New: [rhinstaller/blivet/pulls/81 master] In FS._postSetup() check the mountpoint options that were actually used.
by dashea 15 Apr '15

15 Apr '15

Should have gone in with commit 10bc41e6f73796298a3a78faaf69dd363429ea53. Signed-off-by: mulhern <amulhern(a)redhat.com> -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/81

1 4

Updated: [rhinstaller/blivet/pulls/77 master] Master fixups
by dashea 15 Apr '15

15 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/77

1 4

Updated: [rhinstaller/blivet/pulls/66 master] Fix some Python 3 compatibility issues and add the python3 subpackage
by dashea 15 Apr '15

15 Apr '15

Fixes for Python 3 compatibility issues discovered while running tests with Python 3 including hundreds of installations with Python 3 Anaconda. With all that passing in the end, I decided to add the pieces that make sure the python3 subpackage is built too (tested in [Copr builds] (https://copr.fedoraproject.org/coprs/bkabrda/py3anaconda/builds/)). **Disclaimer:** These patches don't try to fix anything else than Python 3 incompatibilities in the most straightforward way. Many changes are subject for future improvements and nicer solutions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/66

1 18

New: [rhinstaller/blivet/pulls/78 rhel7-branch] Fix "anaconda hangs while trying to discover iscsi..." (#1166652)
by dashea 15 Apr '15

15 Apr '15

This was 2 errors in one. First problem: anaconda freeze for some time. Second problem: libiscsi was run on thread and libiscsi the library is not thread safe. Anaconda freeze is because of GIL is taken on library thread which is trying to connect to server. The thread safe problem is because libiscsi is using signals internally which were delivered to another thread (this is hard to solve in python threads...). Solution is to start discover method on new process. So new process is now created to call libiscsi discover_sendtargets method. This should fix both problems. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/78

1 9

[blivet][master] Fixes for master changes
by Vratislav Podzimek 14 Apr '15

14 Apr '15

The first two patches are fixes for recent changes on blivet's master branch, the third one is something I've been keeping hitting for long enough to make me patch it. Vratislav Podzimek (3): Do not try to stat FileDevice's path if it doesn't exist Add back DeviceTree's support for saving LUKS passphrases Do not try importing hidden/backup files as formats blivet/blivet.py | 3 +-- blivet/devices/file.py | 2 +- blivet/devicetree.py | 5 +++++ blivet/formats/__init__.py | 2 +- blivet/populator.py | 7 +++++++ 5 files changed, 15 insertions(+), 4 deletions(-) -- 2.1.0

3 5

New: [rhinstaller/anaconda/pulls/69 master] Check for Gtk before importing escape_markup
by dashea 14 Apr '15

14 Apr '15

Early errors can cause errors.py to be imported before any of the GUI are setup. Trying to do a full import triggers various side-effects that cause it to crash in _isys instead of just using the lambda. This method for checking for Gtk was lifted from exception.py -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/69

1 5

New: [rhinstaller/blivet/pulls/77 master] Master fixups
by dashea 14 Apr '15

14 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/77

1 5

New: [rhinstaller/anaconda/pulls/68 master] If the network is disabled, also disable the network part of the source ...
by dashea 14 Apr '15

14 Apr '15

...spoke. (#1192104) Cherry pick a commit from rhel7 to disable the network parts of the source spoke if networking is not enabled. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/68

1 5

[PATCH rhel7-branch] Implement the new reqpart command.
by Chris Lumens 14 Apr '15

14 Apr '15

Resolves: rhbz#1164660 --- pyanaconda/kickstart.py | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/pyanaconda/kickstart.py b/pyanaconda/kickstart.py index 0d722d1..0b72ada 100644 --- a/pyanaconda/kickstart.py +++ b/pyanaconda/kickstart.py @@ -27,6 +27,7 @@ from blivet.devicelibs import swap as swap_lib from blivet.formats import getFormat from blivet.partitioning import doPartitioning from blivet.partitioning import growLVM +from blivet.platform import platform from blivet.errors import StorageError, BTRFSValueError from blivet.size import Size from blivet import udev @@ -1465,6 +1466,28 @@ class RepoData(commands.repo.RHEL7_RepoData): commands.repo.RHEL7_RepoData.__init__(self, *args, **kwargs) +class ReqPart(commands.reqpart.RHEL7_ReqPart): + def execute(self, storage, ksdata, instClass): + from blivet.partitioning import doReqPartition + + if not self.reqpart: + return + + reqs = platform.setPlatformBootloaderReqs() + if self.addBoot: + bootPartition = platform.setPlatformBootPartition() + + # blivet doesn't know this - anaconda sets up the default boot fstype + # in various places in this file, as well as in setDefaultPartitioning + # in the install classes. We need to duplicate that here. + for part in bootPartition: + if part.mountpoint == "/boot": + part.fstype = storage.defaultBootFSType + + reqs += bootPartition + + doReqPartition(storage, reqs) + class RootPw(commands.rootpw.F18_RootPw): def execute(self, storage, ksdata, instClass, users): if not self.password and not flags.automatedInstall: @@ -1725,6 +1748,7 @@ commandMap = { "partition": Partition, "raid": Raid, "realm": Realm, + "reqpart": ReqPart, "rootpw": RootPw, "selinux": SELinux, "services": Services, @@ -1937,6 +1961,7 @@ def doKickstartStorage(storage, ksdata, instClass): ksdata.bootloader.execute(storage, ksdata, instClass) ksdata.autopart.execute(storage, ksdata, instClass) + ksdata.reqpart.execute(storage, ksdata, instClass) ksdata.partition.execute(storage, ksdata, instClass) ksdata.raid.execute(storage, ksdata, instClass) ksdata.volgroup.execute(storage, ksdata, instClass) -- 2.2.2

2 7

[anaconda][master][PATCH] Sort disks by name when checking disk selection
by Vratislav Podzimek 14 Apr '15

14 Apr '15

Not sure about what Python 2 makes up to sort the disks, but Python 3 complains about DiskDevice objects being unsortable. Making the use of name as the key explicit is however good no matter which python we are using. Signed-off-by: Vratislav Podzimek <vpodzime(a)redhat.com> --- pyanaconda/ui/lib/disks.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pyanaconda/ui/lib/disks.py b/pyanaconda/ui/lib/disks.py index 0febf85..9264360 100644 --- a/pyanaconda/ui/lib/disks.py +++ b/pyanaconda/ui/lib/disks.py @@ -106,7 +106,7 @@ def checkDiskSelection(storage, selected_disks): errors = [] for name in selected_disks: selected = storage.devicetree.getDeviceByName(name, hidden=True) - related = sorted(storage.devicetree.getRelatedDisks(selected)) + related = sorted(storage.devicetree.getRelatedDisks(selected), key=lambda d: d.name) missing = [r.name for r in related if r.name not in selected_disks] if missing: errors.append(P_("You selected disk %(selected)s, which contains " -- 2.1.0

2 1

Updated: [rhinstaller/anaconda/pulls/66 master] Give users a chance to fix unusable initial storage layouts
by dashea 13 Apr '15

13 Apr '15

If there is an issue with the initial storage configuration that must be resolved before installation can continue, give the user an opportunity to resolve it. If they do, they can retry the storage scan and proceed with the installation. If not, they have to exit the installer. Limits on input quality are not unreasonable. This relies on rhinstaller/blivet#76 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/66

1 5

Updated: [rhinstaller/blivet/pulls/76 master] Do not continue when presented with unusable storage configurations
by dashea 13 Apr '15

13 Apr '15

There are some issues that must be resolved before blivet can do anything useful. The two that this PR addresses are multiple LVM VGs with the same name and disklabel corruption that prevents parted from acknowledging a disklabel when partitions are present on the disk. Until the lvm tools accept UUID in place of VG name for at least lvremove, vgchange, and vgremove we have no way to work with multiple VGs with the same name. This related bug has been sitting since 2008: https://bugzilla.redhat.com/show_bug.cgi?id=449832 The first patch enables parted to proceed with a partially-corrupt GPT disklabel, but I suspect it will not cover all such cases, hence the exception handling. There will soon be an accompanying anaconda PR that handles these exceptions by showing the error and the suggested course of action to the user. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/76

1 4

Updated: [rhinstaller/blivet/pulls/71 master] Consolidate common code in DeviceFormat class methods.
by dashea 13 Apr '15

13 Apr '15

This is the same model used within the device classes, which has worked pretty well. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/71

1 1

New: [rhinstaller/anaconda/pulls/61 master] Master multiple mounts
by dashea 13 Apr '15

13 Apr '15

Use changes in https://github.com/rhinstaller/blivet/pull/69 to handle cases where we have a device mounted multiple places. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/61

1 6

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 13 Apr '15

13 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 8

New: [rhinstaller/anaconda/pulls/66 master] Give users a chance to fix unusable initial storage layouts
by dashea 13 Apr '15

13 Apr '15

If there is an issue with the initial storage configuration that must be resolved before installation can continue, give the user an opportunity to resolve it. If they do, they can retry the storage scan and proceed with the installation. If not, they have to exit the installer. Limits on input quality are not unreasonable. This relies on rhinstaller/blivet#76 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/66

1 10

[anaconda][master/f22-branch] Make reuse of LUKS devices work
by Vratislav Podzimek 12 Apr '15

12 Apr '15

These two patches fix issues with reuse of LUKS devices for a new installation reported in bug #1210254 and discovered by me after fixing the first part. Vratislav Podzimek (2): Make the Encrypt checkbox insensitive for encrypted non-BTRFS devices (#1210254) Set both .format's and .originalFormat's passphrase on unlock pyanaconda/ui/gui/spokes/custom.py | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) -- 2.1.0

2 4

New: [rhinstaller/blivet/pulls/76 master] Do not continue when presented with unusable storage configurations
by dashea 12 Apr '15

12 Apr '15

There are some issues that must be resolved before blivet can do anything useful. The two that this PR addresses are multiple LVM VGs with the same name and disklabel corruption that prevents parted from acknowledging a disklabel when partitions are present on the disk. Until the lvm tools accept UUID in place of VG name for at least lvremove, vgchange, and vgremove we have no way to work with multiple VGs with the same name. This related bug has been sitting since 2008: https://bugzilla.redhat.com/show_bug.cgi?id=449832 The first patch enables parted to proceed with a partially-corrupt GPT disklabel, but I suspect it will not cover all such cases, hence the exception handling. There will soon be an accompanying anaconda PR that handles these exceptions by showing the error and the suggested course of action to the user. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/76

1 9

New: [rhinstaller/blivet/pulls/71 master] Consolidate common code in DeviceFormat class methods.
by dashea 12 Apr '15

12 Apr '15

This is the same model used within the device classes, which has worked pretty well. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/71

1 17

Updated: [rhinstaller/blivet/pulls/72 master] Pickling support for Size
by dashea 12 Apr '15

12 Apr '15

I need to pickle size instances in blivet-gui for multiprocess communication and \__reduce__ function inherited from decimal causes precision loss and other problems. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/72

1 15

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 10 Apr '15

10 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 0

New: [rhinstaller/anaconda/pulls/64 rhel6-branch] Fix a regression caused by adding thinp support
by dashea 10 Apr '15

10 Apr '15

Resolves: rhbz#1200832 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/64

2 8

New: [rhinstaller/blivet/pulls/74 master] Trivial Test Fixes
by dashea 10 Apr '15

10 Apr '15

Easy ack. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/74

1 5

[PATCH lorax rhel7-branch] Port of lmc pxe to live support to
by Radek Vykydal 10 Apr '15

10 Apr '15

The patches are already on master and also on rhel 7 atomic dist-git branch. I needed to pull two patches from master.

2 8

New: [rhinstaller/lorax/pulls/15 master] Added Sphinx Documentation
by dashea 09 Apr '15

09 Apr '15

Run 'make docs' to update the apidocs. -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/15

1 7

Updated: [rhinstaller/blivet/pulls/66 master] Fix some Python 3 compatibility issues and add the python3 subpackage
by dashea 09 Apr '15

09 Apr '15

Fixes for Python 3 compatibility issues discovered while running tests with Python 3 including hundreds of installations with Python 3 Anaconda. With all that passing in the end, I decided to add the pieces that make sure the python3 subpackage is built too (tested in [Copr builds] (https://copr.fedoraproject.org/coprs/bkabrda/py3anaconda/builds/)). **Disclaimer:** These patches don't try to fix anything else than Python 3 incompatibilities in the most straightforward way. Many changes are subject for future improvements and nicer solutions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/66

1 16

[master/f22-branch] Fix for a possible F22 final blocker
by Vratislav Podzimek 09 Apr '15

09 Apr '15

PATCH 1/2 is just a minor "style" change, PATCH 2/2 actually fixes the bug #1210003. Vratislav Podzimek (2): Mark the back_clicked attribute of the Storage spoke as private Reset the back_clicked flag if we stay on the Storage spoke (#1210003) pyanaconda/ui/gui/spokes/storage.py | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) -- 2.1.0

2 3

[PATCH] Preserve the order of boot args added by kickstart.
by Chris Lumens 08 Apr '15

08 Apr '15

This adds OrderedSet to anaconda, just like it is in pykickstart. I have added additional functionality that we need to the Arguments class in bootloader.py so if OrderedSet is needed elsewhere, it won't inherit the special behavior. This also attempts to keep "rhgb quiet" at the end of the line by moving where passed-in command line arguments are processed to the very end. Note that the storage args get popped off much earlier. Resolves: rhbz#1188948 --- anaconda.spec.in | 2 +- pyanaconda/bootloader.py | 47 ++++++++++++++++++------------------ pyanaconda/orderedset.py | 63 ++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 88 insertions(+), 24 deletions(-) create mode 100644 pyanaconda/orderedset.py diff --git a/anaconda.spec.in b/anaconda.spec.in index 92a54d2..b97f7d6 100644 --- a/anaconda.spec.in +++ b/anaconda.spec.in @@ -4,7 +4,7 @@ Summary: Graphical system installer Name: anaconda Version: @PACKAGE_VERSION@ Release: @PACKAGE_RELEASE@%{?dist} -License: GPLv2+ +License: GPLv2+ and MIT Group: Applications/System URL: http://fedoraproject.org/wiki/Anaconda diff --git a/pyanaconda/bootloader.py b/pyanaconda/bootloader.py index cdd596f..568a61d 100644 --- a/pyanaconda/bootloader.py +++ b/pyanaconda/bootloader.py @@ -43,6 +43,7 @@ from pyanaconda.nm import nm_device_hwaddress from blivet import platform from blivet.size import Size from pyanaconda.i18n import _, N_ +from pyanaconda.orderedset import OrderedSet import logging log = logging.getLogger("anaconda") @@ -124,12 +125,7 @@ def _is_on_iscsi(device): class BootLoaderError(Exception): pass -class Arguments(set): - ordering_dict = { - "rhgb" : 99, - "quiet" : 100 - } - +class Arguments(OrderedSet): def _merge_ip(self): """ Find ip= arguments targetting the same interface and merge them. @@ -140,7 +136,7 @@ class Arguments(set): # automatic network setup: return arg.startswith("ip=") and arg.count(":") == 1 ip_params = filter(partition_p, self) - rest = set(filter(lambda p: not partition_p(p), self)) + rest = OrderedSet(filter(lambda p: not partition_p(p), self)) # split at the colon: ip_params = map(lambda p: p.split(":"), ip_params) @@ -162,13 +158,18 @@ class Arguments(set): def __str__(self): self._merge_ip() - # sort the elements according to their values in ordering_dict. The - # higher the number the closer to the final string the argument - # gets. The default is 50. - lst = sorted(self, key=lambda s: self.ordering_dict.get(s, 50)) - + lst = list(self) return " ".join(lst) + def add(self, key): + self.discard(key) + OrderedSet.add(self, key) + + def update(self, other): + for key in other: + self.discard(key) + self.add(key) + class BootLoaderImage(object): """ Base class for bootloader images. Suitable for non-linux OS images. """ def __init__(self, device=None, label=None, short=None): @@ -845,17 +846,6 @@ class BootLoader(object): self.boot_args.update(setup_args) self.dracut_args.update(setup_args) - # passed-in objects - for cfg_obj in list(args) + kwargs.values(): - if hasattr(cfg_obj, "dracutSetupArgs"): - setup_args = cfg_obj.dracutSetupArgs() - self.boot_args.update(setup_args) - self.dracut_args.update(setup_args) - else: - setup_string = cfg_obj.dracutSetupString() - self.boot_args.add(setup_string) - self.dracut_args.add(setup_string) - # This is needed for FCoE, bug #743784. The case: # We discover LUN on an iface which is part of multipath setup. # If the iface is disconnected after discovery anaconda doesn't @@ -890,6 +880,17 @@ class BootLoader(object): self.boot_args.add(new_arg) + # passed-in objects + for cfg_obj in list(args) + kwargs.values(): + if hasattr(cfg_obj, "dracutSetupArgs"): + setup_args = cfg_obj.dracutSetupArgs() + self.boot_args.update(setup_args) + self.dracut_args.update(setup_args) + else: + setup_string = cfg_obj.dracutSetupString() + self.boot_args.add(setup_string) + self.dracut_args.add(setup_string) + # # configuration # diff --git a/pyanaconda/orderedset.py b/pyanaconda/orderedset.py new file mode 100644 index 0000000..125d57c --- /dev/null +++ b/pyanaconda/orderedset.py @@ -0,0 +1,63 @@ +# Copyright 2009 Raymond Hettinger +# Distributed under the MIT license +# Obtained at http://code.activestate.com/recipes/576694/ + +import collections + +class OrderedSet(collections.MutableSet): + + def __init__(self, iterable=None): + self.end = end = [] + end += [None, end, end] # sentinel node for doubly linked list + self.map = {} # key --> [key, prev, next] + if iterable is not None: + self |= iterable + + def __len__(self): + return len(self.map) + + def __contains__(self, key): + return key in self.map + + def add(self, key): + if key not in self.map: + end = self.end + curr = end[1] + curr[2] = end[1] = self.map[key] = [key, curr, end] + + def discard(self, key): + if key in self.map: + key, prev, next = self.map.pop(key) # pylint: disable=redefined-builtin + prev[2] = next + next[1] = prev + + def __iter__(self): + end = self.end + curr = end[2] + while curr is not end: + yield curr[0] + curr = curr[2] + + def __reversed__(self): + end = self.end + curr = end[1] + while curr is not end: + yield curr[0] + curr = curr[1] + + def pop(self, last=True): # pylint: disable=arguments-differ + if not self: + raise KeyError('set is empty') + key = self.end[1][0] if last else self.end[2][0] + self.discard(key) + return key + + def __repr__(self): + if not self: + return '%s()' % (self.__class__.__name__,) + return '%s(%r)' % (self.__class__.__name__, list(self)) + + def __eq__(self, other): + if isinstance(other, OrderedSet): + return len(self) == len(other) and list(self) == list(other) + return set(self) == set(other) -- 2.2.2

2 1

New: [rhinstaller/blivet/pulls/72 master] Pickling support for Size
by dashea 08 Apr '15

08 Apr '15

I need to pickle size instances in blivet-gui for multiprocess communication and \__reduce__ function inherited from decimal causes precision loss and other problems. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/72

1 8

[PATCH] Don't unconditionally set ksdata.lang.seen to True (#1209927)
by Martin Kolman 08 Apr '15

08 Apr '15

Only set it to True if language is set by the user. Otherwise Geolocation would not work. Signed-off-by: Martin Kolman <mkolman(a)redhat.com> --- anaconda | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/anaconda b/anaconda index ad0a69c..d9ef103 100755 --- a/anaconda +++ b/anaconda @@ -1227,6 +1227,7 @@ if __name__ == "__main__": # If command line options or kickstart set LANG, override the environment if opts.lang: os.environ["LANG"] = opts.lang # pylint: disable=environment-modify + ksdata.lang.seen = True elif ksdata.lang.lang: os.environ["LANG"] = ksdata.lang.lang # pylint: disable=environment-modify @@ -1245,7 +1246,6 @@ if __name__ == "__main__": os.environ["LANG"] = constants.DEFAULT_LANG # pylint: disable=environment-modify localization.setup_locale(os.environ["LANG"], ksdata.lang) - ksdata.lang.seen = True import blivet blivet.enable_installer_mode() -- 2.1.0

2 2

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 08 Apr '15

08 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 7

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 07 Apr '15

07 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 4

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 07 Apr '15

07 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 4

Updated: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 07 Apr '15

07 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 5

Updated: [rhinstaller/blivet/pulls/68 master] Use /sys to obtain sizes of existing devices
by dashea 07 Apr '15

07 Apr '15

parted.Device is quite heavy for just tracking device size. Due to its intended purpose being quite different from how blivet uses it, it becomes problematic in certain situations related to uevents, udev rules, and things like md arrays. Specifically, deleting a parted.Device involves closing a read/write file descriptor, which generates a "change" uevent, which triggers all sorts of udev rules. If the device whose parted.Device you just deleted happened to be a member of an md array, the result would be that the array would get activated. One way this can happen is by simply calling Blivet.reset on a system with an inactive md array. Obviously, this is unacceptable on the basis that storage libraries (or modules) must not activate system devices without having been instructed to do so. This business of closing rw fds gets even less convenient when blivet begins listening for and reacting to uevents, so this is a prerequisite for that. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/68

1 6

New: [rhinstaller/anaconda/pulls/62 master] TUI pwpolicy setup was supposed to be in __init__ not refresh (#1208607)
by dashea 07 Apr '15

07 Apr '15

-- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/62

1 5

New: [rhinstaller/blivet/pulls/69 master] Master multiple mounts
by dashea 07 Apr '15

07 Apr '15

Fix dynamic mountpoint code to allow multiple mounts when the user specifies the mountpoint. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/69

1 11

New: [rhinstaller/anaconda/pulls/63 f22-branch] Revert "allow /boot on btrfs subvol or filesystem"
by dashea 07 Apr '15

07 Apr '15

This reverts commit 6dd9f042b607a6dc4e8d299009c62c5f98e6d5d4. We really don't have support for /boot on btrfs working, turn this off to prevent bugs. Related: rhbz#1200539 Related: rhbz#864198 Related: rhbz#1094489 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/63

1 6

New: [rhinstaller/blivet/pulls/70 master] Master 1208536
by dashea 07 Apr '15

07 Apr '15

Resolves: fed#1208536 -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/70

1 6

Updated: [rhinstaller/blivet/pulls/67 f22-branch] F22 1208536
by dashea 07 Apr '15

07 Apr '15

Also for master. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/67

1 9

New: [rhinstaller/blivet/pulls/68 master] Use /sys to obtain sizes of existing devices
by dashea 07 Apr '15

07 Apr '15

parted.Device is quite heavy for just tracking device size. Due to its intended purpose being quite different from how blivet uses it, it becomes problematic in certain situations related to uevents, udev rules, and things like md arrays. Specifically, deleting a parted.Device involves closing a read/write file descriptor, which generates a "change" uevent, which triggers all sorts of udev rules. If the device whose parted.Device you just deleted happened to be a member of an md array, the result would be that the array would get activated. One way this can happen is by simply calling Blivet.reset on a system with an inactive md array. Obviously, this is unacceptable on the basis that storage libraries (or modules) must not activate system devices without having been instructed to do so. This business of closing rw fds gets even less convenient when blivet begins listening for and reacting to uevents, so this is a prerequisite for that. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/68

1 10

[rhel6-branch] Make sure authconfig and firewalld packages are installed. (#1161682)
by Samantha N. Bueno 06 Apr '15

06 Apr '15

Unless a user specifies --nocore. Based on a patch by Masahiro Matsuya (mmatsuya(a)redhat.com) Resolves: rhbz#1161682 --- yuminstall.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/yuminstall.py b/yuminstall.py index da35677..ba9d048 100644 --- a/yuminstall.py +++ b/yuminstall.py @@ -1600,6 +1600,10 @@ debuglevel=6 anaconda.id.ksdata.services.enabled) or \ anaconda.id.storage.services or anaconda.id.network.hasActiveIPoIBDevice(): self.selectPackage("chkconfig") + else: + if not anaconda.id.ksdata.packages.nocore: + for pkg in ['authconfig', 'system-config-firewall-base']: + self.selectPackage(pkg) def doPostSelection(self, anaconda): # Only solve dependencies on the way through the installer, not the way back. -- 1.8.3.1

2 1

New: [rhinstaller/anaconda/pulls/58 master] Connect scroll adjustments in the right class (#1206472)
by dashea 06 Apr '15

06 Apr '15

Sure are a lot of classes in that file -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/58

1 6

Updated: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 02 Apr '15

02 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. This requires rhinstaller/blivet#63 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 3

Updated: [rhinstaller/blivet/pulls/63 master] Bits needed to enforce sane disk selection in anaconda
by dashea 02 Apr '15

02 Apr '15

This goes with rhinstaller/anaconda#60 - move getRelatedDisks into blivet as suggested by both @vpodzime and @mulkieran - tweak getDependentDevices to (optionally) work with hidden devices -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/63

1 3

New: [rhinstaller/lorax/pulls/14 f22-branch] Include cryptsetup in the image (#1208214)
by dashea 02 Apr '15

02 Apr '15

blivet dropped its requirement on cryptsetup, but fedup needs it in the upgrade.img and it is useful for rescue mode so add it back. -- To view this pull request on github, visit https://github.com/rhinstaller/lorax/pull/14

1 5

Updated: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 02 Apr '15

02 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. This requires rhinstaller/blivet#63 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 1

Updated: [rhinstaller/blivet/pulls/67 f22-branch] F22 1208536
by dashea 02 Apr '15

02 Apr '15

Also for master. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/67

1 3

New: [rhinstaller/blivet/pulls/67 f22-branch] F22 1208536
by dashea 02 Apr '15

02 Apr '15

Also for master. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/67

1 5

Updated: [rhinstaller/anaconda/pulls/59 f22-branch] Allow overriding password policy check
by dashea 02 Apr '15

02 Apr '15

The Fedora Engineering Steering Committee decided that Fedora 22 will restore the double-click of "done" to override the password policy check. This is a temporary measure while FESCo works on a distro-wide security policy to be implemented in Fedora 23. Details: https://fedorahosted.org/fesco/ticket/1412#comment:40 https://fedoraproject.org/wiki/User:Kevin/Draft_Passwordpolicy -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/59

1 4

Updated: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 02 Apr '15

02 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. This requires rhinstaller/blivet#63 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 2

Updated: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 02 Apr '15

02 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. This requires rhinstaller/blivet#63 -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 1

Updated: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 02 Apr '15

02 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 4

New: [rhinstaller/blivet/pulls/66 master] Fix some Python 3 compatibility issues and add the python3 subpackage
by dashea 02 Apr '15

02 Apr '15

Fixes for Python 3 compatibility issues discovered while running tests with Python 3 including hundreds of installations with Python 3 Anaconda. With all that passing in the end, I decided to add the pieces that make sure the python3 subpackage is built too (tested in [Copr builds] (https://copr.fedoraproject.org/coprs/bkabrda/py3anaconda/builds/)). **Disclaimer:** These patches don't try to fix anything else than Python 3 incompatibilities in the most straightforward way. Many changes are subject for future improvements and nicer solutions. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/66

1 15

Updated: [rhinstaller/blivet/pulls/61 master] Use LVname not VG-LVname when adding LVs
by dashea 02 Apr '15

02 Apr '15

The VG-LVname string is not a key in the dictionary constructed from the 'lvm lvs' output and if by some chance (really high on Python 3) e.g. a thin pool is requested to be added by a thin LV, using VG-LVname as the key raises a KeyError while the info about the pool LV is in the dictionary as expected. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/61

1 4

New: [rhinstaller/blivet/pulls/64 master] ]Wrap keys() with a list so that the dictionary can be changed
by dashea 02 Apr '15

02 Apr '15

In Python 3 the keys() dictionary method returns and iterator, which we can't use if we want to modify the dictionary while iterating over it (it raises a RuntimeError). So wrap keys() with a list to prevent this from happening. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/64

1 5

New: [rhinstaller/blivet/pulls/62 master] Add back DeviceTree's methods and properties used from the outside
by dashea 02 Apr '15

02 Apr '15

These methods/properties were moved to Populator and ActionList classes, but to keep the API we need to provide them in the DeviceTree class too. Also make the 'dasd' property settable because that's what Anaconda does. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/62

1 4

New: [rhinstaller/blivet/pulls/63 master] Bits needed to enforce sane disk selection in anaconda
by dashea 02 Apr '15

02 Apr '15

This goes with rhinstaller/anaconda#60 - move getRelatedDisks into blivet as suggested by both @vpodzime and @mulkieran - tweak getDependentDevices to (optionally) work with hidden devices -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/63

1 10

New: [rhinstaller/anaconda/pulls/60 master] Enforce sane disk selections.
by dashea 01 Apr '15

01 Apr '15

There is no value (and a great deal of trouble) in allowing users to split up an existing md array, lvm vg, or btrfs volume by selecting a subset of the disks it uses. This practice has led to the filing of several bugs, mainly by individuals engaged in testing the installer. The only bug id I have handy is 1198367, which I'll edit into the commit message for the rhel7-branch commit. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/60

1 16

New: [rhinstaller/anaconda/pulls/59 f22-branch] Allow overriding password policy check
by dashea 01 Apr '15

01 Apr '15

The Fedora Engineering Steering Committee decided that Fedora 22 will restore the double-click of "done" to override the password policy check. This is a temporary measure while FESCo works on a distro-wide security policy to be implemented in Fedora 23. Details: https://fedorahosted.org/fesco/ticket/1412#comment:40 https://fedoraproject.org/wiki/User:Kevin/Draft_Passwordpolicy -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/59

1 5

New: [rhinstaller/blivet/pulls/61 master] Use LVname not VG-LVname when adding LVs
by dashea 01 Apr '15

01 Apr '15

The VG-LVname string is not a key in the dictionary constructed from the 'lvm lvs' output and if by some chance (really high on Python 3) e.g. a thin pool is requested to be added by a thin LV, using VG-LVname as the key raises a KeyError while the info about the pool LV is in the dictionary as expected. -- To view this pull request on github, visit https://github.com/rhinstaller/blivet/pull/61

1 5

New: [rhinstaller/anaconda/pulls/57 master] Master anaconda dracut tests
by dashea 31 Mar '15

31 Mar '15

Do some testing on parse-kickstart so that we can be more confident that a python3 version of it is working correctly. -- To view this pull request on github, visit https://github.com/rhinstaller/anaconda/pull/57

1 10

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

anaconda-patches April 2015