summaryrefslogtreecommitdiffstats
path: root/mdk-stage1
Commit message (Collapse)AuthorAgeFilesLines
* pcmcia works on x86_64 too, update stage1 build for 10.0Gwenolé Beauchesne2004-09-211-0/+9
|
* clean-up, merge, fix pcmcia subsystem to make it work on x86_64Gwenolé Beauchesne2004-09-2110-117/+16
|
* define io ports on x86_64 too for pcmcia stuffGwenolé Beauchesne2004-09-211-1/+1
|
* revert previous commit, that won't work after pivot_root (proc has toOlivier Blin2004-07-301-6/+3
| | | | | | | be mounted, insmod would have to be able to find the modules in the new root), we'll have to use hotplug to load eagle-usb drivers in Move (in next release maybe)
* s/MandrakeMove/Move/Pascal Rigaux2004-07-292-3/+4
|
* do not probe usb interface too soon, wait for the Move images to be mountedOlivier Blin2004-07-281-3/+6
| | | | | (or else usb drivers won't be automatically loaded at boot with Move)
* don't redefine buf for /proc/scsi/scsi (size 5000 instead of 512)Nicolas Planel2004-06-241-1/+0
|
* increase the number of loopbacks (needed for mandrakemove where the default ↵Pascal Rigaux2004-06-151-4/+4
| | | | (8) is much too low :)
* in move, use stg1_info_message() instead of stg1_error_message() not to ↵Olivier Blin2004-06-091-1/+1
| | | | disable MODE_AUTOMATIC if not enough memory
* This commit was manufactured by cvs2svn to create branchMystery Man2004-06-091-0/+669
| | | | 'MDK-10-update'.
* use the same technique as ldetect for detecting usb and firewire controllers ↵Pascal Rigaux2004-06-031-31/+61
| | | | (based on the pci class)
* usb cdrom has a weird size in /proc/partitions, the result is that it is ↵Pascal Rigaux2004-05-131-1/+1
| | | | detected as floppy
* This commit was manufactured by cvs2svn to create branchMystery Man2004-05-111-671/+0
| | | | 'MDK-10-update'.
* don't use msgboot-graphicallogo.img.bz2 anymore, build it (easy and simpler)Pascal Rigaux2004-05-032-0/+0
|
* stage1-disk and stage1-medias-usb are dead alreadyPascal Rigaux2004-05-031-32/+4
|
* obsolete stuffPascal Rigaux2004-05-032-0/+0
|
* blank.img has been removed, so remove special code handling itPascal Rigaux2004-05-033-13/+1
|
* rename MandrakeMove to MandrakemovePascal Rigaux2004-04-262-3/+3
|
* - no need for chloop device anymore, my kernel patch applies directly on the ↵Pascal Rigaux2004-04-081-11/+5
| | | | | | | default loop module :) - gzloop needs cryptoloop and zlib_inflate (why must i handle this by hand? is my gzloop ugly?...)
* clean init-move when cleaningPascal Rigaux2004-04-021-1/+1
|
* enable nfs install on old i586 machines (or maybe poor network cards) ↵Pascal Rigaux2004-04-021-3/+5
| | | | (bugzilla #9322) (thanks to Michael Riss)
* fix eth[67] detection (gc)Thierry Vignaud2004-03-111-1/+1
|
* (get_net_devices) increase max number of detectable net cards from 6 to 10Thierry Vignaud2004-03-101-1/+1
|
* unmounting /sys is not needed (useful to have one less line displayed after ↵Pascal Rigaux2004-03-041-0/+1
| | | | "unmounting filesystems...")
* when in automatic=disk mode but no partition is given, fall backGuillaume Cottenceau2004-02-271-3/+1
| | | | | on non automatic mode evidently
* give us more chance to get bootloader write actually commited toGuillaume Cottenceau2004-02-251-0/+3
| | | | | disk before reboot
* Patch from gc to allow booting a MandrakeMove on a Harddrive using theErwan Velu2004-02-251-10/+30
| | | | | automatic settings.
* 10.0Guillaume Cottenceau2004-02-201-1/+1
|
* support automatic installs even when needing the additional drivers floppyGuillaume Cottenceau2004-02-181-2/+12
|
* support 2.6 insmoding for expert third party modules and loading named modulesGuillaume Cottenceau2004-02-173-18/+28
|
* when available, always load usb interface, as soon as possible, should help ↵Guillaume Cottenceau2004-02-174-21/+3
| | | | usb mouse detection in stage2 which otherwise timeouts
* fix buggy redhat code exhausted only in -Os which corruptedGuillaume Cottenceau2004-02-161-1/+1
| | | | | | memory and gave wrong results for download file size in FTP/HTTP installs
* sleep 10 only when debuggingPascal Rigaux2004-02-131-0/+2
|
* linux/fd.h is too broken to compile with -W :(Guillaume Cottenceau2004-02-121-1/+1
|
* try to detect regular floppy drivesGuillaume Cottenceau2004-02-123-5/+22
|
* be sure to load usb-storage after SCSI adapters, so that they are inGuillaume Cottenceau2004-02-111-8/+10
| | | | | same order than reboot, so that naming is the same
* don't insmod usb-storage the soonest, we can end up mixing scsi devices ↵Guillaume Cottenceau2004-02-111-5/+11
| | | | orders at reboot time, do it only when needed
* floppy_device returns first floppy device, look for medias first in case of ↵Guillaume Cottenceau2004-02-102-1/+14
| | | | usb floppies (maybe ide floppies also? never tested) then regular floppy drive
* in network mode we need to create nodes as well because of usb floppies supportGuillaume Cottenceau2004-02-101-2/+0
|
* use floppy_device rather than /dev/fd0 because we now support usb floppies ↵Guillaume Cottenceau2004-02-102-3/+3
| | | | as well
* fake usb floppies detection by looking at blocks size of scsi disk devices, ↵Guillaume Cottenceau2004-02-101-2/+17
| | | | 1048575 is for 2.4 and 1440 for 2.6
* log server http responseGuillaume Cottenceau2004-02-061-0/+1
|
* differentiate no floppy disk in driver or ext2 fsGuillaume Cottenceau2004-02-031-2/+5
|
* document when netauto is not provided and bootfile DHCP server param is givenGuillaume Cottenceau2004-02-031-2/+6
|
* more doc on DHCP server response confGuillaume Cottenceau2004-02-031-0/+10
|
* allow having no DNS by only using the ip callback once, see DNS == IP as a ↵Guillaume Cottenceau2004-01-282-2/+22
| | | | special case, don't do DNS calls when no DNS is configured to avoid timeouts
* don't forget to umount additional drivers floppy after successful copy of ↵Guillaume Cottenceau2004-01-261-0/+1
| | | | modules.mar (gc sux)
* since dietlibc always reports null strings for RPC errors, at least provide ↵Guillaume Cottenceau2004-01-231-3/+2
| | | | something useful instead
* moving forward our story with nice kernel guys, now they decided to remove ↵Guillaume Cottenceau2004-01-221-1/+1
| | | | the trailing space at the end of "Attached devices:" of /proc/scsi/scsi, probably that they had nothing more interesting to do this day
* better do 2.4->2.6 compat module alias as soon as possible to display ↵Guillaume Cottenceau2004-01-221-37/+40
| | | | correct stuff in logs and simplify code
generated by cgit v1.2.1 (git 2.21.0) at 2024-11-05 21:36:10 +0000