summaryrefslogtreecommitdiffstats
path: root/loader
Commit message (Collapse)AuthorAgeFilesLines
...
* Convert urlinstall.c to using the new urlinstTransfer function.Chris Lumens2009-09-301-190/+51
|
* Add proxy support to urlinstTransfer by setting more curl options.Chris Lumens2009-09-303-3/+25
|
* Add the urlinstTransfer function, which replaces urlinst*Transfer.Chris Lumens2009-09-302-81/+63
| | | | | | | These older functions operate by initiating a connection, returning a file descriptor, the caller performing some random operation, and then closing the connection. This new function does everything for you. If the downloaded file doesn't end up where you want, you'll need to do the move yourself.
* Add a function to construct an array of HTTP headers and cache the result.Chris Lumens2009-09-301-0/+85
|
* Add a CURL instance to the loader data.Chris Lumens2009-09-302-0/+8
|
* Add checks for libcurl into the makefile process.Chris Lumens2009-09-301-0/+2
|
* Improve loader messages in parseCmdLineFlags when passing vnc (#526350).Steffen Maier2009-09-291-2/+3
| | | | | | | | | | | | | | Instead of the misleading 21:55:23 INFO : vnc forced cmdline mode from cmdline 21:55:23 INFO : early networking required for syslog it now prints correctly 21:55:23 INFO : vnc forced graphical mode from cmdline 21:55:23 INFO : early networking required for vnc The early networking message was introduced solely for syslog in 620bdb978bfe46f96f04865e776edd15728b85d8 but later reused without adaptation also for vnc in b27dcdc6a37b5429b50579dbf0650e880a95988c.
* Whitespace cleanup in loader/net.c.David Cantrell2009-09-291-56/+70
| | | | | Clean up whitespace in chooseNetworkInterface() and make the whole function use consistent whitespace.
* Clean up getHostandPath() debugging messages for host & file.David Cantrell2009-09-291-3/+4
|
* We can't prompt for new network info in cmdline mode (#526262).Chris Lumens2009-09-291-0/+8
|
* Add 'install' user to start anaconda on s390x.David Cantrell2009-09-211-3/+3
| | | | | | | | | | This is a usability improvment. Once you've booted up the kernel and initrd on s390x, you are told to ssh in to the host to start the installation. In the past, we had rigged the root account to launch loader on the first login and a shell on subsequent logins. In this patch, I've added an 'install' user that has loader for the shell. The root user still has bash for the shell. Also change the recommendation to run 'ssh -X' rather than 'ssh -x'.
* Do not assume we found a module in addOption() in loader/modules.cDavid Cantrell2009-09-211-6/+10
| | | | | Avoid SIGSEGV in addOption() in loader/modules.c. Just because found does not equal 0 does not mean we found the module requested.
* Do not try to load floppy, edd, pcspkr, or iscsi_ibft on s390x.David Cantrell2009-09-211-1/+5
| | | | | Gets rid of the 'FATAL: Module _______ not found' errors in the log files.
* Fix going back from hd install UI when stage2 is given as boot param (#519206).Radek Vykydal2009-09-171-7/+4
| | | | | | | Prevents SIGSEGV. We need to reinitialize stage2Data here too (as for other methods). Also when going back from Driver Disk Source dialog - go back to hd install dialog instead of going straight to install method dialog.
* load_policy has moved from /usr/sbin to /sbin (#523506).Chris Lumens2009-09-151-2/+2
|
* Remove duplicate search_cu() in linuxrc.s390David Cantrell2009-09-152-83/+49
| | | | | | Remove duplicate search_cu() function in linuxrc.s390, move CU_ data structures to controlunits.sh and have linuxrc.s390 source /sbin/controlunits when it runs.
* It's controlunits, not controlunits.sh.Chris Lumens2009-09-101-1/+1
|
* Fix askmethod + stage2= (#516973, #519288, #518194)Radek Vykydal2009-09-101-0/+6
| | | | Adds to Jeremy's 1fbf3c9d8d550fc01ae34528533377adc19d1611.
* Make anaconda know its version number (#520061)David Cantrell2009-09-041-1/+0
| | | | | | | Remove the ANACONDAVERSION environment variable and embed the version number at build time. The isys.getAnacondaVersion() function returns the version number to Python code. If need be, it can be extended to libisys.
* We convert cmdline args to longs in several places, so reduce to a function.Chris Lumens2009-09-021-44/+23
|
* Add some silent make support for sed, mkctype, and other commands.David Cantrell2009-09-021-5/+25
| | | | Match up to the silent make thing we have by default now.
* Fix going back in "Inst. Method" and "Configure TCP/IP" screens in stage 1 ↵Radek Vykydal2009-08-262-4/+15
| | | | | | | | | | (#515450) Going back from method screen was broken with commit 245d1951c40ecb61cee11f3ecc00e483bc8dd142. This patch implements Retry flow differently - by staying on the Configure TCP/IP screen instead of going there through hacked (skipped) method screen. The interface can still be changed by going Back from Configure TCP/IP which is fixed in this patch too.
* Fix going back from stage1 nfs/url setup dialog.Radek Vykydal2009-08-262-2/+6
| | | | | | | | | | | | | | We need to clear stage2Data when going back so that * we don't sigsegv on it later if we change nfs method to url (and also if we change from url to nfs in case of location specified via ks or boot opts) * we don't try to fetch stage2 from the location in stage2Data right after reselecting the method. Also fixes [C] and [A] from https://www.redhat.com/archives/anaconda-devel-list/2009-August/msg00043.html (sigsegv after going Back when incorrect stage2 url is specified via ks or boot params) which were partly fixed by Jeremy's ad1e95df0e277442a19f7989afe3bd758cfccb47.
* Allocate memory for login and password and do not meddle with host pointer ↵Martin Sivak2009-08-192-5/+10
| | | | so we can correctly free it (#483818)
* Remove unused reipl code in linuxrc.s390Steffen Maier2009-08-181-24/+2
| | | | reipl is done correctly for DASD and zFCP in iutil.
* 64 bit sparc linux does not define __sparc64__ we need to use ↵Dennis Gilmore2009-08-181-1/+1
| | | | "(defined(__sparc__) && defined(__arch64__))" fixes building 64 bit sparc
* sparc no longer needs and special keyboard handling. it uses the standard ↵Dennis Gilmore2009-08-181-24/+1
| | | | api's interfaces
* setup termcap for sparcDennis Gilmore2009-08-181-1/+9
|
* If asked, put the system SN (as given by dmidecode) into an HTTP header.Chris Lumens2009-08-171-0/+32
|
* Add the kssendsn parameter and corresponding flag.Chris Lumens2009-08-172-0/+4
|
* Don't keep testing if we're doing URL_METHOD_HTTP.Chris Lumens2009-08-171-26/+26
|
* Fix askmethod to work with stage= being specified (#516973)Jeremy Katz2009-08-141-1/+4
| | | | | | We now are specifying stage2= but if they add askmethod, we should still end up doing the asking. So add the check for stage2 implying skipping the method dialog under the !askmethod case
* Correctly inform the user once about obsolete parm/conf file options on s390Steffen Maier2009-08-121-1/+22
|
* Handle activation of DASDs in linuxrc.s390 since loader no longer worksSteffen Maier2009-08-121-28/+111
| | | | | | | | | | | | | | Since we automatically load device drivers in linuxrc.s390 by means of udev, loader can no longer explicitly load the DASD device driver module and pass the module option DASD=... in order to activate DASDs based on the user specification in parm/conf file or interactively in linuxrc.s390. This patch does the necessary work in linuxrc.s390 including full support for usability features such as restarting configuration without reboot. In a next step we could remove the DASD module stuff from loader. It is recommended to use /dev/disk/by-* instead of /dev/dasdXY.
* make IPv4 configuration in linuxrc.s390 compatible with NM in loaderSteffen Maier2009-08-121-9/+18
| | | | | | | | | | | Fix new step 2 of the following 3 steps of network configuration on s390x: 1) linuxrc.s390 completely configures initial network (ifconfig, route, /etc/hosts, /etc/resolv.conf) 2) loader starts NM, which in turn reconfigures network based on /etc/sysconfig/network, /etc/sysconfig/network-scripts/ifcfg-XYZ 3) loader reconfigures network based on /tmp/install.cfg using NM, which rewrites parts of /etc/sysconfig/network, /etc/sysconfig/network-scripts/ifcfg-XYZ
* suggest disabled X11-forwarding for ssh login in linuxrc.s390Steffen Maier2009-08-121-1/+1
| | | | | | | The GTK used for the GUI in anaconda requires some recent X extensions, which most probably many clients don't have, so do not suggest using ssh logins with X11-forwarding but rather suggest explicitly disabled X11-forwarding so the users may then choose VNC, which works fine.
* Fix an erroneous "!" in the test for doKill, and make reboot explicit.Peter Jones2009-08-124-8/+18
| | | | | | This test causes it not to reboot if you don't specify anything about rebooting in kickstart or on the command line. It also treats /all/ installs with "nokill" specified as "halt" installs.
* Honor network config boot params for CD-booted ks installs (#433214)David Cantrell2009-08-102-5/+9
| | | | | | | | | | | | | | | | | Users booting from CD or DVD media, performing a kickstart install, and passing network configuration boot parameters were hitting an issue where loader was insisting on trying to get a DHCP lease to fetch the kickstart file. Even if you provided 'asknetwork' or 'ip=' settings, loader was still forcing DHCP at the point where the kickstart file is downloaded. This patch makes the 'asknetwork' parameter work if you are booting from CD or DVD and performing a kickstart install. By default, we would try DHCP, but asknetwork forces the configureTCPIP() code path to run. Likewise, if the user boots from media and passes ip=, netmask=, gateway=, dns=, and ksdevice=, the network is configured using that information and the kickstart file is downloaded.
* Fix going back from "NFS Setup" screen in stage 1 (#507064)Radek Vykydal2009-08-071-1/+1
|
* Add missing 'i' in loader/loader.c for non-s390 arches.David Cantrell2009-08-061-0/+2
| | | | | We actually need 'i' on everything but s390. The world is not all mainframes.
* Call shutDown() correctly for s390Karsten Hopp2009-08-061-1/+1
| | | | | Syntax of shutDown() function changed recently, so make sure we call it correctly.
* Remove unused variable from loader/loader.cKarsten Hopp2009-08-061-1/+1
|
* Complete udev setup in linuxrc.s390 for automatic module loadingroot2009-08-061-6/+37
| | | | Patch from Steffen Maier.
* Fix ctrl-alt-deleter behavior /before/ end of install.Peter Jones2009-08-043-3/+8
| | | | | | The new reboot code didn't take into account the fact that there's a different sigint handler while anaconda is running vs when we're done installing. This patch fixes that bad assumption.
* Rework shutDown() to better accomidate "nokill" better.Peter Jones2009-07-303-45/+97
| | | | | | Rework shutDown() so that we don't do unmounts until we're going to reboot or power off the machine. Also some minor cleanups like taking out all the negative-phrasing of "noKill" in the code itself.
* Make sure controlunits.sh is installed to initrd on s390David Cantrell2009-07-282-2/+2
| | | | | | Put controlunits.sh in /usr/lib/anaconda-runtime/loader on s390, install as /sbin/controlunits in the initrd.img, modify linuxrc.s390 to source /sbin/controlunits.
* Skip writeDisabledNetInfo() when loader starts on s390David Cantrell2009-07-281-0/+2
| | | | This causes NetworkManager to hang on s390.
* Update linuxrc.s390 and friends to reflect review comments.Steffen Maier2009-07-272-343/+288
| | | | | | | | | | | | | | | | | | | | | | This brings linuxrc.s390 and friends to the latest version as posted here https://www.redhat.com/archives/anaconda-devel-list/2009-February/msg00392.html plus some updates after that. Changes: - update usage of udev to rawhide - no more telnetd for network login during install - adapt readcmsfile to automatic loading of device driver modules - refactor main part into function init_main - consider renaming of network device driver module from ctc to ctcm - refactor control unit lookup tables into separate file - no more rely on kernel versions to detect device driver features - upstream qeth driver supports IPv6 on HiperSockets in layer3 mode - no more manual (un)loading of device driver modules - correctly handle default MAC address for layer2 qeth devices - no more kill udevd at end of linuxrc.s390 - write network device configuration to correct new place (ifcfg) Splitting up into single patches is impossible without merge conflicts.
* Try to auto-find the CD even if stage2= is specifiedJeremy Katz2009-07-211-2/+2
| | | | | | | | | Try to auto-find the CD even if stage2= is specified; this lets us add stage2= and then the isohybrid'd images can be dd'd to a usb stick and stage2 gets find there as well as on the CD. Next step is to be able to get rid of the auto CD finding by always explicitly specifying stage2=
* Add hipersockets to NETTYPE description (bhinson, #511962).Chris Lumens2009-07-171-1/+1
|
generated by cgit (git 2.34.1) at 2025-08-31 18:58:01 +0000