Reiser4 enabled LiveCDv2 - Release Notes

thechris · Veteran Joined: 12 Oct 2003 Posts: 1203

i got the ebuild for r3 and it fails.

i'll try the redeeman4 patch now.

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

what about Redeeman5? ;P

Hal9k · n00b Joined: 06 Aug 2003 Posts: 6

Very nice CD! Was able to convert to Reiser4 without any issues. The new filesystem is doing completely fine, but it hasn't really been "tested" (power outage) yet. Reiser3 was good, but couldn't always handle the hard crashes; one time it actually corrupted the filesystem (though in a very minor way where I couldn't remove a directory).

Possible things of interest for the future...

1) Hopefully the BadRAM (http://rick.vanrein.org/linux/badram/download.html) patch can be included in the kernel for those of us who need to bypass a bad address. Did note the LiveCD included the new Memtest+ -- very nice!

2) Inclusion of the dar utility (emerge dar). It permits one to make and restore multi-volume archives. For the LiveCD, only /usr/bin/dar_static is needed to keep bloat under control. (When doing CD work from a LiveCD, it is necessary to pass the "docache" option to the kernel so that the LiveCD can be unmounted and ejected. However, it means everything goes to memory.)

3) Related to #2 and quite minor: inclusion of the CD utilities (cdrecord/mkisofs). Helps with archiving, but a user can always adjust the library path and run the programs from the hard drive. (#2 is more important because we might be restoring an entire hard drive from CDs or DVDs.)

4) Noticed the framebuffer kernel didn't work here with an NForce2 motherboard (Asus A7N8X) and an ATI 9700 video card. Fortunately the non-fb kernel was available.

Thanks again for the very up-to-date LiveCD!

thechris · Veteran Joined: 12 Oct 2003 Posts: 1203

tried redeeman5 with genkernel -- i get issues with i think it was alloc.c with things like __small_alloc or __small_malloc. i'll try a manual compile next.

this is with gcc3.4, so i'm not ruling that out, but i would assume since you can compile the kernel and its listed that you use gcc3.4 that the issue is resolvable. i'm going to try to jerryrig this now -- move /tmp and /var to a different hard disk and try using vanilla or gentoo dev sources without reiser4.

tinyau · n00b Joined: 15 Apr 2004 Posts: 10

redeeman, you have mentioned this version of LiveCD has iteraid patched. However, I use this LiveCD to boot in my PC with ITE Raid, ITE Raid is not detected.

wrc1944 · Posted: Wed Jun 30, 2004 4:23 am Post subject:

Redeeman,
Thought you might like to know- I've also been unable to reach your "sources" site for 2 days now. I could reach the lamda site, and got the new live-cd iso in a marathon dialup download session.

An error occurred while loading http://kaspersandberg.com/~redeeman/Redeeman-Sources/:

Timeout on server
Connection was to kaspersandberg.com at port 80

This happens at all times, day or night.

wrc1944
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.36-r7, gcc-13.2.1_p20230304
kernel-6.9.1 USE=experimental python3_11

thechris · Veteran Joined: 12 Oct 2003 Posts: 1203

ok, a manual compile almost boots -- ithas issues with kernel RAID1. i'll doublecheck the config. it got the RAID0 for the root fs just fine. what is ironic is that the raid1 partition was supposed to be the "always there" thing...

edited fstab and raidtab rebooted.

DHCP still takes 2 minutes to fail... i need to figure out what part of the script fails. i'm thinking of making my own initscript to _ACTUALLY_ bring up eth0 with dhcp. but this happens on the liveCD and every gentoo install AND with other distros. so i guess i'm fine there... from my understanding it is a IRQ issue as all other "netmount" solutions have failed.

and now it does boot, which is good. and tulip loads. so far so good. i guess this is the first time genkernel has ever left me down!

edit--update i compiled raid1 as a module. so that issue is resolved. i changed /etc/init.d/net.eth0 and that should fix that issue as well. at least i hope it does. right now i'm compiling xfree becuase last time i used xorg with 2.6.7 and nvidia 5336 i had issues with hardlocking. not sure if it was xorg's fault or 2.6.7 or 5336, but since i've heard the switch from xfree to xorg isn't hard, i figure i could just go with what worked in the past.

wrc1944 · Posted: Wed Jun 30, 2004 5:10 am Post subject:

Wedge_,
Sorry, I still have some questions:

Regarding the 2nd file, you say "Put this in /etc/ppp/peers (remember to give it the same name as used in the previous file)."

Since the first file is placed in /etc/ppp/peers, and named "yourISP," how can I place another file there, and name it exactly the same thing?

Also, shouldn't:

OK "ATDTphonenumberhere"
"ogin:" "username"
"word:" "password"
"otocol:" ppp"

actually be:

OK "ATDTphonenumberhere"
"login:" "username"
"password:" "password"
"protocol:" ppp"

Was this a typo?
---------------------------------------------
All that I have in /etc/ppp/peers are 2 files:

wvdial, consisting only of:

noauth
name wvdial
usepeerdns
--------------------------------------
And, wvdial-pipe, consisting of:

noauth
name wvdial

plugin passwordfd.so

defaultroute
replacedefaultroute
----------------------------------------------------
And of course I also have an /etc/wvdial.conf file:

[Dialer Defaults]
Modem = /dev/ttyS0
Baud = 115200
Init1 = ATZ
Init2 = ATQ0 V1 E1 S0=0 &C1 &D2 +FCLASS=0
ISDN = 0
Modem Type = Analog Modem
Phone = 3840323
Username = flacycads
Password = xxxxxx
Stupid Mode = 2

Which has some items you have in your 2nd file.

What about the ip-up, ip-down, chap/pap secrets, chat-default files in /etc/ppp?
-------------------------------------------------------------
My chat-defaults file looks like this:

# /etc/ppp/chat-default:
# $Header: /home/cvsroot/gentoo-x86/net-dialup/ppp/files/2.4.2/chat-default,v 1.1 2004/02/12 20:25:27 lanius Exp $

'ABORT' 'BUSY'
'ABORT' 'ERROR'
'ABORT' 'NO ANSWER'
'ABORT' 'NO CARRIER'
'ABORT' 'NO DIALTONE'
'ABORT' 'Invalid Login'
'ABORT' 'Login incorrect'
'' 'ATZ'
'OK' 'ATDT$NUMBER'
'CONNECT' ''
'TIMEOUT' '5'
'~--' ''
----------------------------------------------------------
Needless to say, all this is a bit confusing, and I don't see how it's going to work, if I follow your instructions exactly.

BTW, I had to edit ip-up, adding the chmod 644 line, because the new ppp rewrites the /etc/resolv.conf file permissions at every reboot, so that you can't connect to any website as user. Like this:

one
cp -a $REALRESOLVCONF $REALRESOLVCONF.pppd-backup
mv $REALRESOLVCONF.tmp $REALRESOLVCONF
chmod 644 $REALRESOLVCONF

fi

I guess I still don't understand how ppp is going to connect, without all the info in these other files in /etc/ppp that you didn't mention.

Thanks,
wrc1944
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.36-r7, gcc-13.2.1_p20230304
kernel-6.9.1 USE=experimental python3_11

Wedge_ · Posted: Wed Jun 30, 2004 10:23 am Post subject:

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

ITEraid:
i patched the kernel.

reach sources:
its viewable from ~12:00 - 02:30 GMT+1 most likely...
dammit, i will upload to mirror soon

thechris:
dont dont dont use genkernel... its evil, and besides, it disallows udev with the evil initrd

anyway, i always compile my kernel with 3.4, and it works perfect..

---

btw, dont panic if the ebuild does not work, theres very high chance my ebuilds does not work as i do not use them myself.. but check for a type in the fetch url if it does not

wrc1944 · Posted: Wed Jun 30, 2004 1:08 pm Post subject:

Thanks again, Wedge_!

That really clarifies it for me- much appreciated.

So- let me summarize, to be absolutely sure I understand.

I create these files with nano, NOT in the chrooted /mnt/gentoo environment, but in the Redeeman live-cd environment, which is loaded into RAM.

In other words, the Redeeman cd, like Knoppix, has it's own /etc directory loaded into ram, and I can edit/add the needed files there, and connect just like I do with a knoppix live-cd with kppp (which is loaded to ram), but instead of the gui kppp, I'm using the pure command line pppd method, as you've described.

Then, once connected, I can chroot over to the Gentoo environment, and proceed with downloading/emerging into my reiser4 Gentoo system.

Have I finally got it correct?

wrc1944
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.36-r7, gcc-13.2.1_p20230304
kernel-6.9.1 USE=experimental python3_11

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

Wedge_ · Posted: Wed Jun 30, 2004 1:49 pm Post subject:

Isaiah · Guru Joined: 25 Feb 2003 Posts: 359

thechris · Veteran Joined: 12 Oct 2003 Posts: 1203

i got the kernel working. i can boot, but Xfree hardlocks when i use nvidia. i can get 320x240x8bit on the vga module. i had similar issues with gcc3.3+xorg+gnome+nvidia. a fuller explanation of the issue is in the desktop enviroment section.

StringCheesian · l33t Joined: 21 Oct 2003 Posts: 887

I have found a reproducable error, which I think is reiser4 related.

After extracting a stage tarball and emerge syncing, the next thing I do is exit the chroot and copy over the distfiles (1.3 gigs) from my other Gentoo installation. xterm, or whatever console I'm using freezes when I execute the cp command, and Ctrl+C doesn't work.

It's happened all three times I've tried reiser4, and never when I use reiserfs or xfs.

Has anybody been able to copy more than a gig of files to reiser4 from another filesystem all at once? And/or more than 700 files at once? Can somebody try it, I hope it's not just me (what could I be doing wrong?)

EDIT: If I cd to the directory I was trying to copy to on the reiser4 partition, and type ls, it freezes the same way. It also happens if I try to rm -R the directory. umount won't unmount it, instead it complains that the filesystem is busy (and yet it isn't because I've closed everything that could be accessing it and cd-ed out of the mount point).

EDIT #2: This time I was using libaal-0.5.2, reiser4progs-0.5.5, 2.6.7-mm4 patched with reiser4-2004.06.30-19.37-linux-2.6.7-mm4.diff from http://www.namesys.com/auto-snapshots/. I was using an older kernel (about 2.6.5, I think it was redeeman's kernel) the other two times this happened.

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

yes... reiser4 related.. this bug was fixed, seems it came again

ett_gramse_nap · Posted: Thu Jul 01, 2004 8:09 am Post subject:

When I'm booting with this livecd on my old Dell Latitude laptop my kernel hangs at "Freeing unused kernel memory: xxx".

I've searched around about it but haven't found anything that seems to do with it. I have also tried booting with most of the kernel parameters, such as "nodetect". Any suggestions of what can be done about it?
_________________
Don't bother!

StringCheesian · l33t Joined: 21 Oct 2003 Posts: 887

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

i have latest snapshot in redeeman6

StringCheesian · l33t Joined: 21 Oct 2003 Posts: 887

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

download it off http://kaspersandberg.com/~redeeman/Redeeman-Sources/2.6.7/2.6.7-Redeeman6

wrc1944 · Posted: Thu Jul 01, 2004 6:10 pm Post subject:

Does redeeman6 apply to 2.6.7-mm4, or mm5, and still use -p3? There's nothing in the notes about this.

wrc1944
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.36-r7, gcc-13.2.1_p20230304
kernel-6.9.1 USE=experimental python3_11

Redeeman · l33t Joined: 25 Sep 2003 Posts: 958 Location: Portugal

take redeeman7

against 2.6.7-mm5

tinyau · n00b Joined: 15 Apr 2004 Posts: 10