[python-meh] Last missing pieces for Python3 compatibility
by Vratislav Podzimek
These four patches are all related to making python-meh compatible with both
Python2 and Python3 (reported as bug #985294) and they build on patches provided
by mhroncok (in the bug) for Makefile and python-meh.spec that add building and
packaging code that results in both python-meh and python3-meh packages being
created.
Vratislav Podzimek (4):
Make sure we work with strings when we think we do
Add one more flag to test (for Python3)
Use dict.items() instead of dict.iteritems()
Encode str/unicode object before hashing it
meh/dump.py | 21 ++++++++++++---------
meh/handler.py | 2 +-
2 files changed, 13 insertions(+), 10 deletions(-)
--
1.9.3
9 years, 4 months
[anaconda:master 0/3] Using a pointless overrides checker (v2)
by Anne Mulhern
The first two patches are the same as the equivalent from previous version.
The last patch substitutes for the previous "Factor computation of paths..."
patch.
mulhern (3):
Add a simple pointless-override checker to pylint checkers.
Remove pointless overrides identified by the pointless override
checker.
Remove UserInterface.basepath and UserInterface.basemask definitions.
pyanaconda/bootloader.py | 3 -
pyanaconda/installclasses/fedora.py | 2 -
pyanaconda/installclasses/rhel.py | 1 -
pyanaconda/packaging/dnfpayload.py | 3 -
pyanaconda/ui/__init__.py | 3 -
pyanaconda/ui/tui/tuiobject.py | 9 --
tests/pylint/intl.py | 2 -
tests/pylint/pointless-override.py | 252 ++++++++++++++++++++++++++++++++++++
tests/pylint/pylint-one.sh | 2 +-
9 files changed, 253 insertions(+), 24 deletions(-)
create mode 100644 tests/pylint/pointless-override.py
--
1.9.3
9 years, 4 months
[master 0/2] stylesheet changes
by David Shea
Give the fedora-productimg-* packages the tools they need to apply a
stylesheet, and remove the broken logo handling code because the logos aren't
broken anymore.
David Shea (2):
Add an additional stylesheet path.
Remove _assureLogoImage
pyanaconda/ui/gui/__init__.py | 66 +++++++++----------------------------------
1 file changed, 14 insertions(+), 52 deletions(-)
--
2.1.0
9 years, 4 months
[PATCH] Print an error when the kickstart file is missing
by Brian C. Lane
If --kickstart has been passed we should make sure it actually exists.
If we don't the installation tries to continue but gets confused because
automatedInstall has been set.
---
anaconda | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/anaconda b/anaconda
index 6707e33..70ff3fc 100755
--- a/anaconda
+++ b/anaconda
@@ -1049,6 +1049,11 @@ if __name__ == "__main__":
# shipped with the installation media.
ksdata = None
if opts.ksfile and not opts.liveinst:
+ if not os.path.exists(opts.ksfile):
+ stdoutLog.error("Kickstart file %s is missing.", opts.ksfile)
+ iutil.ipmi_report(constants.IPMI_ABORTED)
+ sys.exit(1)
+
flags.automatedInstall = True
flags.eject = False
ksFiles = [opts.ksfile]
--
1.9.3
9 years, 5 months
[master/rhel7-branch] Fix EOF error that occurs if user input required in x3270. (#1171135)
by Samantha N. Bueno
Prior to fixing rhbz#1040933, anaconda was always launched from a tmux
session. Now however, if you install via text/cmdline kickstart, sshd
is suspended and everything is dumped to the x3270 console.
If it happened that user input was required (e.g. if you supplied an
incomplete ks file), this EOF error was triggered when raw_input() was
called from pyanaconda.ui.tui.simpleline.base, because there was no std
input device for it to open. Well, standard input on x3270 is a tty, so
define that here to prevent the traceback and allow users to provide
input if it is needed.
Resolves: rhbz#1171135
---
data/systemd/anaconda-direct.service | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/data/systemd/anaconda-direct.service b/data/systemd/anaconda-direct.service
index 9bf172a..91a88b3 100644
--- a/data/systemd/anaconda-direct.service
+++ b/data/systemd/anaconda-direct.service
@@ -11,7 +11,7 @@ Environment=HOME=/root MALLOC_CHECK_=2 MALLOC_PERTURB_=204 PATH=/usr/bin:/bin:/s
Type=oneshot
WorkingDirectory=/root
ExecStart=/usr/sbin/anaconda
-StandardInput=null
+StandardInput=tty
StandardOutput=journal+console
StandardError=journal+console
TimeoutSec=0
--
1.9.3
9 years, 5 months
[blivet:master 0/6] Patches about setting NTFS._formattable to True
by Anne Mulhern
mulhern (6):
Make NTFS._formattable True.
Don't make overridden values actual properties.
Change the generic badly formatted label to one that's bad for all.
Make unnecessarily verbose properties into simple class attributes.
Make it possible for NTFS to recognize the label it reads.
Tests NTFS labeling.
blivet/formats/fs.py | 1 +
blivet/formats/fslabel.py | 34 +++++++++++++++++-----------------
tests/formats_test/fs_test.py | 1 -
tests/formats_test/fslabeling.py | 2 +-
tests/formats_test/labeling_test.py | 32 ++++++++++++++++++--------------
5 files changed, 37 insertions(+), 33 deletions(-)
--
1.9.3
9 years, 5 months
[anaconda][rhel7-branch][PATCH 0/2] Fix help subpackage version breakage
by Martin Kolman
As it looks like the build process does not like the help subpackage having
the same version across multiple builds, the first patch makes it use the
ever changing Anaconda version.
To be still easily able to find what help content version is used in a given
Anaconda package and also to be able to easily check in the installation
environment, the second patch passes the help content version to the
help content generation script, so that it can write the version to a version
file. The script does not actually do that just yet, but the script in the
next version of the help content tarball will. :)
Martin Kolman (2):
Use Anaconda version for the help subpackage (#1072033)
Pass the help content version to the script that generates it
(#1072033)
anaconda.spec.in | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
--
1.9.3
9 years, 5 months
[anaconda:master 0/3] Using a pointless overrides checker
by Anne Mulhern
The first patch is motivated by the results of the pointless overrides checker.
The checker identifies somethings as pointless overrides which aren't really
meant to be class attributes at all. It seemed sensible to just refactor all
that path calculation into a staticmethod.
The other two patches just add the pylint checker and then remove pointless
overrides identified. The checker has to be used with some judgement in
when there is multiple inheritance, which occurs with TUIObject, so I did
have to check a few by hand. The restrictions of the checker are all
documented in the comments.
mulhern (3):
Factor computation of paths attribute into UserInterface class.
Add a simple pointless-override checker to pylint checkers.
Remove pointless overrides identified by the pointless override
checker.
pyanaconda/bootloader.py | 3 -
pyanaconda/installclasses/fedora.py | 2 -
pyanaconda/installclasses/rhel.py | 1 -
pyanaconda/packaging/dnfpayload.py | 3 -
pyanaconda/ui/__init__.py | 32 ++++-
pyanaconda/ui/gui/__init__.py | 21 +--
pyanaconda/ui/tui/__init__.py | 20 +--
pyanaconda/ui/tui/tuiobject.py | 9 --
tests/pylint/intl.py | 2 -
tests/pylint/pointless-override.py | 252 ++++++++++++++++++++++++++++++++++++
tests/pylint/pylint-one.sh | 2 +-
11 files changed, 285 insertions(+), 62 deletions(-)
create mode 100644 tests/pylint/pointless-override.py
--
1.9.3
9 years, 5 months
[lorax/master] aarch64 no longer needs explicit console setting (#1170412)
by Brian C. Lane
As of fedora kernel commit 831ed82ff581629c aarch64 can sort out what to
use for the console on its own, so drop the console= from the aarch64
grub2-efi.cfg template.
---
share/config_files/aarch64/grub2-efi.cfg | 10 +++-------
1 file changed, 3 insertions(+), 7 deletions(-)
diff --git a/share/config_files/aarch64/grub2-efi.cfg b/share/config_files/aarch64/grub2-efi.cfg
index 3cb8465..1cc389d 100644
--- a/share/config_files/aarch64/grub2-efi.cfg
+++ b/share/config_files/aarch64/grub2-efi.cfg
@@ -14,10 +14,6 @@ function load_video {
fi
}
-serial --speed=115200
-terminal_input serial console
-terminal_output serial console
-
load_video
set gfxpayload=keep
insmod gzio
@@ -31,11 +27,11 @@ search --no-floppy --set=root -l '@ISOLABEL@'
### BEGIN /etc/grub.d/10_linux ###
menuentry 'Install @PRODUCT@ @VERSION@' --class red --class gnu-linux --class gnu --class os {
- linux @KERNELPATH@ @ROOT@ ro console=ttyS0,115200n8
+ linux @KERNELPATH@ @ROOT@ ro
initrd @INITRDPATH@
}
menuentry 'Test this media & install @PRODUCT@ @VERSION@' --class red --class gnu-linux --class gnu --class os {
- linux @KERNELPATH@ @ROOT@ rd.live.check console=ttyS0,115200n8
+ linux @KERNELPATH@ @ROOT@ rd.live.check
initrd @INITRDPATH@
}
submenu 'Troubleshooting -->' {
@@ -44,7 +40,7 @@ submenu 'Troubleshooting -->' {
initrd @INITRDPATH@
}
menuentry 'Rescue a @PRODUCT@ system' --class red --class gnu-linux --class gnu --class os {
- linux @KERNELPATH@ @ROOT@ rescue console=ttyS0,115200n8
+ linux @KERNELPATH@ @ROOT@ rescue
initrd @INITRDPATH@
}
}
--
1.9.3
9 years, 5 months
[blivet:rhel7:master 0/3] #1090009 related patches
by Anne Mulhern
The first is just a name change, trivial.
The second is really simple.
The third bears a look from someone who has a better idea of the actual
impact of that lvm_cc_filter stuff.
mulhern (3):
Change allow_degraded_mdraid flag to allow_imperfect_devices
(#1090009)
Be even more accepting when allow_imperfect_devices is True (#1090009)
Be more accepting for dm when allow_imperfect_devices is True
(#1090009)
blivet/devicetree.py | 20 ++++++++++++++++----
blivet/flags.py | 4 ++--
2 files changed, 18 insertions(+), 6 deletions(-)
--
1.9.3
9 years, 5 months