summaryrefslogtreecommitdiffstats
path: root/pyanaconda/isys
Commit message (Collapse)AuthorAgeFilesLines
* Don't fail on invalid network --device kickstart specification.Radek Vykydal2013-01-301-6/+5
| | | | | We can get also org.freedesktop.DBus.Error.AccessDenied exception
* Remove the storage module and replace it with blivet.David Lehman2013-01-281-3/+3
|
* Remove unused functions and move storage-specific utils to storage.David Lehman2013-01-286-852/+15
| | | | | | | | | | | | | | | | | | | This is the first pass at setting things up to split storage out into a separate package. It replaces the guts of iutil.execWithRedirect and iutil.execWithCapture with a single backend function, _run_program. The main thing I can think of that might be cause for complaint is that _run_program does not log each output line as the external program outputs it, so you get the whole output buffer at once. Storage gets its own copy of _run_program and all of the calls in storage use the copy from storage.util. The reason I moved all of the iutil arch functions into storage is that they are mostly used in storage and I didn't want there to be two copies of the whole set. The rest is removing of unused things and moving of things only used by storage into storage.
* Add device NM_DEVICE_TYPE_ETHERNET to isys (#893892)Radek Vykydal2013-01-161-0/+1
| | | | Fix of previous patch.
* g_type_init call is no longer neededBrian C. Lane2013-01-141-8/+0
|
* Remove resetResolver function, we don't need it anymore (#868695)Radek Vykydal2012-12-051-14/+0
| | | | | Also removes urlgrabber.grabber.reset_curl_obj() from getDefaultHostname() where all its job were to cause races occasionally.
* Dump default auto connection's ifcfg file instead of writing a new one (#870922)Radek Vykydal2012-11-191-0/+1
| | | | | | | | | | | | | | | | | | | | The case: 1) Dracut does not use network so it does not write any ifcfg file (e.g. DVD install) 2) NM creates default auto connection (DHCP) because it can't find any ifcfg file for a device. 3) Anaconda writes out default ifcfg file (e.g. ifcfg-eth0) 4) When editing a connection in nm-c-e, ifcfg file of edited default auto connection (Wired connection 1) is created (ifcfg-Wired_connection_1). 5) If anaconda tweaks any ifcfg file later (setting ONBOOT, NM_CONTROLLED), it is the ifcfg-eth0. 6) Both ifcfg files are copied to the target system at the end of installation Luckily, the ifcfg-Wired_connection_1 seems to be used after installation, but bug #870922 suggests it might be not so in all cases. Anyway, we shouldn't create anaconda's ifcfg-ethX default file in this case.
* Don't let mount/umount block python threads (#873600).Chris Lumens2012-11-091-0/+6
| | | | | | | | | | Without these lines, the IO blocks everything in python until complete, which means setting up a bad NFS mount in the source spoke gives you an unresponsive hub until the NFS timeout hits. This is only a patch. The better fix would be to use subprocess for mounting instead of our own C blob. The only reason this C blob exists is because loader also required it, and no one's found the time to convert this old code.
* Rewrite isWirelessDevice to Python and DBus calls (#862801)Vratislav Podzimek2012-10-114-56/+17
| | | | | | | | Using libnm-glib from a separate thread results in a hang, probably because libnm-glib has to do some "loops magic" for its internal use of DBus. Signed-off-by: Vratislav Podzimek <vpodzime@redhat.com>
* Use chvt command for tty switchingVratislav Podzimek2012-09-052-21/+0
| | | | | | isys.vtActivate is problematic when using tmux, chvt command works well and we use it so rarely that it should be okay to call an external command.
* Remove the last references to isysLoadKeymap.Chris Lumens2012-08-091-18/+0
|
* Remove our loadKeymap code from isysVratislav Podzimek2012-08-084-133/+0
| | | | | It is used only in the text mode and it should be possible to replace it with calling 'loadkeys' command in the new text mode.
* lower minimum memory requirements by 256M (#708966)Brian C. Lane2012-05-081-5/+6
| | | | | | Now that we are running from the media instead of RAM we can lower our minimum requirements a bit. Tested with a DVD desktop install in 512M and it worked fine.
* Remove misc. references to loaderWill Woods2012-02-152-3/+3
| | | | | | | EVERY LAST REFERENCE TO LOADER MUST BE ELIMINATED RARRRR: - remove references to loader in analog and makeupdates. - change the default progname in isys/log.c from loader to anaconda. - reword comments here and there that refer to loader.
* And stop attempting to load our own fonts, since we no longer ship them.Chris Lumens2011-10-312-42/+0
|
* imount.c: include fcntl.h before ext2fs/ext2fs.h.Ales Kozumplik2011-09-291-1/+1
| | | | | | A glitch in e2fsprogs-devel header files prevents us from building otherwise. Related: rhbz#742147
* Start NM in loader on s390 until we have systemd init here too (#733680)Radek Vykydal2011-09-212-0/+43
| | | | | This is exactly how we used to do it before systemd took over starting NM for anaconda - commit 4acfadedb038744e6b47af74fe584188f82e44aa.
* Tidy warnings.showwarning into anaconda_log.py.Ales Kozumplik2011-08-251-1/+0
| | | | Also remove unused 'warnings' imports.
* Put nolock instead of ,nolock to options if provided options are empty (#727522)Martin Sivak2011-08-171-1/+1
|
* Restart NetworkManager to use anaconda's initial ifcfg config (#727951)Radek Vykydal2011-08-102-24/+28
| | | | | | | NM is started by systemd before anaconda writes out initial ifcfg files which results in creating of automatic default connections (e.g. "Wired connection 1"). The settings of auto connections are not persistent so we are ok just restarting the NM.
* Use unsigned long long type in doTotalMemory()David Cantrell2011-07-251-2/+2
| | | | | | | | Followup to e24592f036ca1d8d0169d215ef45e333c442dacc, this changes the [hopefully] last remaining int used to represent the total system memory to an unsigned long long. Patch from Anton Blanchard <anton AT samba DOT org>
* Handle systems with more than 2147483647 kB of memory (#704593).David Cantrell2011-07-062-56/+55
| | | | | | | | | | | | Our code had a limit of 2TB for the memory check. At least one reporter is trying to install on a system with 3TB of memory and anaconda reports that he does not have enough memory. The log file reports he has -1119974932 kB of memory. Make two changes here. The first is in iutil.py to explicitly use the long() Python type for memory reporting. The second is in the totalMemory() function in loadermisc.c, which is the source of error reported in #704593.
* imount.c: first wait() for mount then close its stdin/stdout.Ales Kozumplik2011-06-011-2/+2
| | | | | | Else mount can get signalled with a SIGPIPE. Resolves: rhbz#693732
* Update the requirements for memory..David Lehman2011-05-053-6/+3
| | | | | | | | New minimum for physical memory is 768MB. New minimum memory for us to not require swap space is 1152MB. Related: rhbz#696805 Related: rhbz#699770
* Use correct interface to obtain HwAddress property (#693614)Radek Vykydal2011-04-181-1/+2
| | | | Resolves: rhbz#693614
* Revert "Don't write HWADDR into ifcfg files (#690589)"Radek Vykydal2011-04-181-2/+1
| | | | | | This reverts commit 6f7ccb89bf9f99f7f0bb4503c9f73a65ce69b841. HWADDR is required for BOOTPROTO=ibft, and more importantly ifcfg-rh/NM needs it when activating devices.
* Changes for NetworkManager API 0.9Radek Vykydal2011-04-113-8/+23
| | | | Related NM bugs: #672282, #604334
* Don't write HWADDR into ifcfg files (#690589)Radek Vykydal2011-04-081-1/+2
| | | | | | Not needed anymore (udev and biosdevname take care of the naming). Resolves: rhbz#690589
* Ensure remount requests go through isys.mount() (#678520)David Cantrell2011-03-011-1/+1
| | | | | | | | If remount=True and the volume is already mounted, the mount() function will bail. We don't want that for remount requests since remount doesn't change the mount count. (cherry picked from commit 517e6d45c736cef1d13f58db55d6f58553368990)
* Use NM for ibft configurationRadek Vykydal2011-03-021-1/+0
| | | | | Port from rhel6-branch Resolves: rhbz#634016
* Add kickstart network --nodefroute optionRadek Vykydal2011-03-022-0/+2
| | | | | | | | | Port from rhel6-branch. Resolves: rhbz#638131 Sets DEFROUTE=no in ifcfg file. Use case: one NIC used to fetch packages, another for iSCSI. Using the option we can prevent the latter to grab default route.
* Wait for activation of specific devices instead of NMRadek Vykydal2011-03-022-0/+31
| | | | | Port from rhel6-branch Resolves: rhbz#638131
* remove unused variablesMichal Schmidt2011-02-032-8/+6
| | | | | | | | | | | There were two kinds of unused but set variables: - the really pointless ones - the ones whose only purpose was to silence the warnings about ignoring the return values from functions declared with __attribute__((__warn_unused_result__)) Since gcc 4.5 the warnings can be disabled using -Wno-unused-result, so the variables are no longer necessary. Then -Wno-unused-but-set-variable is an overkill, so drop it.
* Raise an exception if X*Display functions fail (#663294).Chris Lumens2010-12-151-0/+2
|
* Add losetup to the install image, re-remove it from isys (#662183).Chris Lumens2010-12-101-42/+0
| | | | | Part of the removal never happened in the first place, and part of it got overwritten when I pushed the install image merging patch series.
* Check for ARPHRD_ETHER and ARPHRD_SLIP types in getDevices (#596826)David Cantrell2010-12-061-1/+3
| | | | | | | | | | | | | In the getDevices() function in isys, we check the type of the device during discovery to see if the type is ARPHRD_ETHER. If it isn't, we skip that device when gathering network devices. This filters out lo, sit0, and other devices we don't want in the network device list. But, it does skip over one ethernet device type that is initialized as ARPHRD_SLIP on s390. (This is cherry-picked from rhel6-branch, but since the functionality is already on master, I just made sure the test is using the ARPHRD_* constants rather than 1 and 256.)
* A little too much got deleted from imount.c.Chris Lumens2010-11-111-0/+2
|
* Return mount's actual error codes instead of obfuscating them.David Lehman2010-11-102-42/+10
| | | | | | | | | | | In addition to being unnecessary, this obfuscation does not account for the fact that the codes can be combined using a bitwise or. This is part of the reason we were unable to identify the actual problem with the series of bugs involving the error "SystemError: (2, None)" during F11, F12, and perhaps after that. Related: rhbz#548592
* Remove mkdirChain() from isys, use g_mkdir_with_parents()David Cantrell2010-11-102-51/+2
| | | | | | | Replace uses of mkdirChain() with g_mkdir_with_parents() and remove the existing mkdirChain() code from isys/imount.* Added a wrapper function for g_mkdir_with_parents() usage.
* Remove include lines for stubs.h from isys.David Cantrell2010-11-101-19/+19
| | | | | The stubs.h header file does not exist anymore. Replace gunzip_* macro usage with the real zlib function names.
* Remove isys cpio extraction code.David Cantrell2010-11-107-993/+4
| | | | | | The only place this code is still used is in loader. Loader is now using libarchive to unpack cpio streams and zlib to uncompress them, so we do not need these functions anymore.
* Fix typo in my ctc commit (#648858)Brian C. Lane2010-11-081-1/+1
|
* Fix ctc check logic (#648858)Brian C. Lane2010-11-081-1/+1
| | | | | | | | | | | if ((type != 1) && !((type == 256) && !strncmp(name, "ctc", 3))) #1 #2 #3 type, name 1 !(2&&3) 2 3 Result 1 bob F T F F F skip continue 2 sam T T F F T continue 256 sam T T T F T continue 256 ctc T F T T F skip continue
* Support devices larger than 1.5TB (#649095, rspanton AT zepler DOT net).Chris Lumens2010-11-031-4/+4
|
* Fix test for CTC devices from yesterday.Chris Lumens2010-11-031-1/+1
|
* Remove obsolete networking code.Radek Vykydal2010-11-031-35/+0
|
* Support installation to CTC devices in loader (#648858, karsten).Chris Lumens2010-11-021-1/+2
|
* We no longer need to copy the install.img over and lochangefd to it.Chris Lumens2010-10-152-32/+0
| | | | | Since there is no more install.img, this whole process is pointless and all its support code can come out.
* Remove losetup and unlosetup from isysBrian C. Lane2010-09-031-28/+0
| | | | No longer used. mount can now handle loop devices itself.
* memory: install.img is now >150 MB so count 192 MB extra for it.Ales Kozumplik2010-08-302-2/+2
|
generated by cgit (git 2.34.1) at 2025-09-28 07:37:29 +0000