aboutsummaryrefslogtreecommitdiffstats
path: root/sysconfig/network-scripts/network-functions
Commit message (Collapse)AuthorAgeFilesLines
* Add infiniband mapping (#648524, <monis@voltaire.com>)Bill Nottingham2010-11-011-0/+3
|
* Fix ifdown calling nmcli for NM-managed devices; others are handled by ↵Jiří Klimeš2010-11-011-0/+5
| | | | ifdown-${DEVICETYPE} (#612934).
* Cleanup.Bill Nottingham2010-06-241-2/+2
|
* Handle quoted SUBCHANNELS (#607481)Bill Nottingham2010-06-241-2/+2
|
* Check NM_CONTROLLED before deciding to use NM for a connection. (#599707, ↵Bill Nottingham2010-06-041-1/+1
| | | | probably others.)
* Behave sanely if we're not passed a device at all. (#589521, #592108)Bill Nottingham2010-05-131-0/+2
|
* Get rid of a silly error message for loopback devices.Bill Nottingham2010-05-131-1/+1
|
* Update for new nmcli syntax.initscripts-9.12-1Bill Nottingham2010-05-051-2/+2
|
* Redirect error messages to /dev/null to catch if nmcli is not installed. ↵Bill Nottingham2010-03-051-2/+2
| | | | (#570828, <zing@fastmail.fm>)
* Only use UUID to bring connections up/down; ensure we have a UUID by asking ↵Bill Nottingham2010-03-011-3/+7
| | | | NM for it.
* Quote 'name' when passing to get_config_by_name.Bill Nottingham2010-03-011-1/+1
|
* Fix is_nm_activeBill Nottingham2010-03-011-1/+1
| | | | Some versions of nmcli have trailing whitespace.
* When bringing devices down, disconnect the device, don't down the connection.Bill Nottingham2010-03-011-0/+5
|
* Remove quotes from uuid.Bill Nottingham2010-03-011-1/+1
|
* Use NM if it's running and we have a configuration for the device.Bill Nottingham2010-03-011-0/+4
|
* Add a function that determines the status of NetworkManager; use it.Bill Nottingham2010-03-011-0/+6
|
* Add get_config_by_name, use it.Bill Nottingham2010-03-011-0/+12
|
* Add get_config_by_device, adjust caller.Bill Nottingham2010-03-011-3/+9
|
* Make sure we default to 'ok' if link checking fails.Bill Nottingham2010-02-151-1/+1
|
* Don't use ethtool to check for device w/bonding driver; just check that the ↵Bill Nottingham2010-02-151-1/+1
| | | | | | | passed device exists. If they have a device that already exists but isn't a bond, the later add will fail anyway.
* Don't use ethtool to check link ; operate on the carrier sysfs attribute.Bill Nottingham2010-02-151-20/+1
|
* Fix various device checks to not misidentify devices.Bill Nottingham2010-02-121-14/+4
| | | | | Also, don't use ifconfig for checking link state on alias devices; /sbin/ip does fine with that.
* If HWADDR is set and DEVICE is not, use HWADDR to determine DEVICE. (#545597)Bill Nottingham2010-01-221-6/+7
|
* Eliminate bunch of unnecessary stat calls and command invocations.Ville Skyttä2010-01-121-1/+1
|
* Get rid of uses of deprecated egrep and fgrep.Ville Skyttä2010-01-121-1/+1
|
* Fix some bashisms.Ville Skyttä2010-01-071-4/+4
|
* Use sysfs rather than iwconfig for wireless checkAdel Gadllah2010-01-041-3/+1
|
* Silence error. (#516569)Bill Nottingham2009-12-151-1/+1
|
* Remove IUCV support. (#507217)Bill Nottingham2009-10-261-3/+0
|
* Load bonding driver if BONDING_OPTS is defined. (#516569)Bill Nottingham2009-10-201-0/+2
|
* Use $(( )) for arithmetic statements instead of let or $[ ].Victor Lowther2009-07-301-3/+3
|
* Use $() instead of `` for process expansion.Bill Nottingham2009-07-301-19/+19
| | | | Adapted from <victor.lowther@gmail.com>.
* Don't use mii-tool for link status; only use ethtool.Bill Nottingham2009-07-021-13/+3
| | | | Essentially a consequence of bug #491358.
* Explicitly source from the proper directory. (#496233)Bill Nottingham2009-04-211-2/+2
|
* Handle full paths in need_config/source_config by normalizing them (#483257)Bill Nottingham2009-03-161-2/+3
| | | | | | This broke when we fixed the source commands for bash 4. It's not as having configurations outside of /etc/sysconfig/network-scripts was supported, so just normalize any full path passed in to look there.
* Remove support for 'ifcfg-foo-bar' configurations that automatically inherit ↵Bill Nottingham2009-03-161-10/+0
| | | | | | | | from 'ifcfg-foo'. 1) I'm not sure of anything at all that ever supported this except for manual configuration. 2) '-' is a valid character in ethernet device names. So you can't have 'ifcfg-eth0-2' files that work. 3) Having it work for every '-' separated combination, except those that end in 'g' is just strange.
* Use a leading path marker when sourcing files. (#482826)Bill Nottingham2009-01-281-4/+4
| | | | bash 4 breaks the prior code.
* fixed typo in get_config_by_subchannel()Harald Hoyer2008-11-251-1/+1
|
* fixed SUBCHANNEL grep for s390Harald Hoyer2008-11-121-1/+3
|
* fixed bug #459044Harald Hoyer2008-11-111-1/+1
| | | | refined the grep statement
* Remove reference to rename_device.Bill Nottingham2008-10-031-1/+0
| | | | rename_device no longer exists.
* https://bugzilla.redhat.com/show_bug.cgi?id=447928Harald Hoyer2008-09-161-1/+1
| | | | | | | | All networking scripts from /etc/sysconfig/network-scripts call logger without full path. This can prevent system from booting if let's say dcc is installed when compiled with wrong parameters and installs /usr/sbin/logger (see bellow) It's not really a bug but this can be easily prevented. All /etc/init.d scripts use full path for logger network-scripts should use it also
* Use 0.0.0.0/0, not 0/0. (#460580)Bill Nottingham2008-09-031-1/+1
| | | | Updated iproute doesn't like the old syntax.
* Don't try to rename devices - udev rules are the way to goBill Nottingham2008-05-231-33/+0
|
* add sit tunnel type (#441635)Bill Nottingham2008-04-091-0/+3
|
* restore file context on /etc/resolv.conf (#230776, <dwalsh@redhat.com>)Bill Nottingham2007-04-161-0/+1
|
* fix logicBill Nottingham2007-02-221-1/+1
|
* simplify bonding test (#215887, <herbert.xu@redhat.com>)Bill Nottingham2007-01-231-4/+1
|
* don't run ethtool if there's no device. Prevents reloading the moduleBill Nottingham2006-11-061-0/+1
| | | | on ifdown called by udev on device unregistration.
* use sysfs interface for bonding (#202443, <agospoda@redhat.com>)Bill Nottingham2006-10-181-0/+13
|