| View previous topic :: View next topic |
| Author |
Message |
kernelOfTruth Watchman


Joined: 20 Dec 2005 Posts: 5345 Location: Vienna, Austria; Germany; hello world :)
|
Posted: Tue May 01, 2007 9:50 pm Post subject: |
|
|
just got a mail from Edward from Namesys:
snip:
| Quote: | It may be because of races intruduced by this unsuccessful patch:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc2/2.6.21-rc2-mm1/broken-out/reiser4-drop-unused-semaphores.patch
make sure, that the following fix is applied:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc7/2.6.21-rc7-mm2/broken-out/reiser4-fix-for-drop-unused-semaphorespatch.patch
If this fix is applied, but the problem still exists, then, please, let me know.
and I had no problems so far (except the 'wrong bytes' error, which get's
fixed fine with fsck.reiser4 <-- those started with later 2.6.20 kernels, don't
know what change in reiser4 or kernel caused it)
http://vipernicus.org/files/patches/2.6.20/viper4/broken-out/reiser4-2.6.20.patch
<== this reiser4-patch works fine with 2.6.20 kernels, elder kernels seemed to
cause 'wrong bytes' errors
Note, that "wrong bytes" reported by fsck for cryptcompress partition is expected
thing (some cosmetic changes is required). Use the option -n for such partitions for a while.
Thanks for report,
Edward. |
snap
so those using reiser4 or willing to try, please check out the following patch (fix) in addition to reiser4 patch:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc7/2.6.21-rc7-mm2/broken-out/reiser4-fix-for-drop-unused-semaphorespatch.patch
according to his mail this should fix the problems, 'wrong bytes' error are only cosmetical: thank god I also didn't experience any data corruptions _________________ Unofficial minimal livecd x86/amd64 w/reiser4+truecrypt (by Neo2)
2.6.37.2_plus_v1: BFS, CFS,THP,compaction, zcache or TOI
Hardcore Linux user since 2004  |
|
| Back to top |
|
 |
Pse Apprentice

Joined: 01 Mar 2005 Posts: 188 Location: by the plate river
|
Posted: Wed May 02, 2007 2:55 am Post subject: |
|
|
| kernelOfTruth wrote: | just got a mail from Edward from Namesys:
snip:
| Quote: | It may be because of races intruduced by this unsuccessful patch:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc2/2.6.21-rc2-mm1/broken-out/reiser4-drop-unused-semaphores.patch
make sure, that the following fix is applied:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc7/2.6.21-rc7-mm2/broken-out/reiser4-fix-for-drop-unused-semaphorespatch.patch
If this fix is applied, but the problem still exists, then, please, let me know.
and I had no problems so far (except the 'wrong bytes' error, which get's
fixed fine with fsck.reiser4 <-- those started with later 2.6.20 kernels, don't
know what change in reiser4 or kernel caused it)
http://vipernicus.org/files/patches/2.6.20/viper4/broken-out/reiser4-2.6.20.patch
<== this reiser4-patch works fine with 2.6.20 kernels, elder kernels seemed to
cause 'wrong bytes' errors
Note, that "wrong bytes" reported by fsck for cryptcompress partition is expected
thing (some cosmetic changes is required). Use the option -n for such partitions for a while.
Thanks for report,
Edward. |
snap
so those using reiser4 or willing to try, please check out the following patch (fix) in addition to reiser4 patch:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21-rc7/2.6.21-rc7-mm2/broken-out/reiser4-fix-for-drop-unused-semaphorespatch.patch
according to his mail this should fix the problems, 'wrong bytes' error are only cosmetical: thank god I also didn't experience any data corruptions |
Great stuff! Thanks for posting, I hope this fix gets into the next Viper release =) In the mean time, 'patch -p1' FTW! |
|
| Back to top |
|
 |
dR0PS Tux's lil' helper


Joined: 27 Feb 2005 Posts: 116 Location: DE/Hamburg
|
Posted: Wed May 02, 2007 8:53 pm Post subject: |
|
|
patched over viper-sources-2.6.21-r1, compiled, works
back in business
thanx for posting
dR0PS _________________
~x86 gentoo / zen-sources 2.6.36-zen1
|
|
| Back to top |
|
 |
vipernicus Veteran


Joined: 17 Jan 2005 Posts: 1462 Location: Your College IT Dept.
|
|
| Back to top |
|
 |
devsk Advocate


Joined: 24 Oct 2003 Posts: 2632 Location: Bay Area, CA
|
Posted: Mon May 07, 2007 8:38 pm Post subject: |
|
|
Longtime R4 user. Never a single corruption.
R4 is highly recommended.
Always apply patch yourself on either vanilla or gentoo-sources.
Do not apply any other patch on top.
Do not use mm-sources.
Others non-official patches can screw up things and everyone gets to blame R4.
Do not use compression plugin with very old cpu. A recent dual core is recommended.
Do not use compression plugin on partitions with mostly audio/video files...
Use the default compression plugin.
The compression plugin gave me 50% space back (a fully loaded install went from 5.2G to 2.5G) and tar cpf (with drop_caches to avoid kernel cache effect) on kernel sources became 50% faster. See this thread for exact numbers:
http://forums.gentoo.org/viewtopic.php?p=4045301
Latest official patch for 2.6.21.
ftp://ftp.namesys.com/pub/reiser4-for-2.6/2.6.21/
Now you are free to enjoy the best file system ever... |
|
| Back to top |
|
 |
vipernicus Veteran


Joined: 17 Jan 2005 Posts: 1462 Location: Your College IT Dept.
|
|
| Back to top |
|
 |
vipernicus Veteran


Joined: 17 Jan 2005 Posts: 1462 Location: Your College IT Dept.
|
|
| Back to top |
|
 |
lucke n00b

Joined: 11 May 2005 Posts: 35
|
Posted: Tue May 08, 2007 11:24 am Post subject: |
|
|
| RIP has reiser4progs 1.0.6. |
|
| Back to top |
|
 |
mbar Veteran


Joined: 19 Jan 2005 Posts: 1902 Location: Poland
|
Posted: Tue May 08, 2007 11:25 am Post subject: |
|
|
| I have my 64bit livecd with reiser4 1.0.6 and your kernel (2.6.21-viper1), just tell me where to upload it (approx. 158 MB). |
|
| Back to top |
|
 |
yngwin Developer


Joined: 19 Dec 2002 Posts: 4219 Location: Suzhou, China
|
|
| Back to top |
|
 |
Elv13 Guru


Joined: 13 Nov 2005 Posts: 377 Location: Socialist land of North America
|
Posted: Tue May 08, 2007 12:35 pm Post subject: |
|
|
| Reiser4 killed my gentoo install and almost my computer. I had to close it using reset and some time it gived me "disk boot failure". I also lost some data and rebuilted the fs from live-cd many time because it was not booting anymore. Now gentoo just don't boot and i installed sabayon 3.3 for replace it until i get some time. DON'T USE REISER4 IT'S BAD. No it is not stable. |
|
| Back to top |
|
 |
MrVahn Apprentice


Joined: 30 Dec 2006 Posts: 227 Location: Makati
|
Posted: Tue May 15, 2007 2:35 am Post subject: |
|
|
| That's why it is still experimental. I put my portage into a 500Mb reiser4 partition. The rest is exi3 to be safe. |
|
| Back to top |
|
 |
mbar Veteran


Joined: 19 Jan 2005 Posts: 1902 Location: Poland
|
Posted: Tue May 15, 2007 4:32 am Post subject: |
|
|
| bollocks |
|
| Back to top |
|
 |
devsk Advocate


Joined: 24 Oct 2003 Posts: 2632 Location: Bay Area, CA
|
Posted: Thu May 17, 2007 7:38 pm Post subject: |
|
|
OK, I am hitting a very strange problem. This is with official reiser4 patch on top of gentoo-sources-2.6.21.
Some times smbd, firefox or vmware processes end up in 'D' state and they block everything else. IO wait goes to 100%. I can login from my other machine most of the time. Some times /bin/bash goes into 'D' state disallowing me that login. If I am lucky and I have a root window accessible, and I do 'sync', everything starts working again, until the next time it happens. Otherwise the problem resolves itself after 10-15 minutes. This has become annoying.
Did anyone else experience this situation using the compress plugin? Does anyone know what causes this, and how to fix it? This does not happen if '/' partition doesn't use crypto-compress plugin. Also, I notice that the compress plugin incorrectly concludes that .iso files are compressible. A simple comparison between livecd.iso on a compress partition and a non-compress partition shows a huge speed loss in reading/writing an ISO on a crypto-compress partition. |
|
| Back to top |
|
 |
Vegan Tux's lil' helper

Joined: 29 Mar 2007 Posts: 133
|
Posted: Thu May 17, 2007 8:13 pm Post subject: |
|
|
compress plugin is quite experimental , besides copmression on reiser4 especially with gzip requiere strong cpu and from my experience 100%CPU usage is normal with some operations , Use it only if you got really strong cpu and you really ned it ( you got small disk ) _________________ Go Vegan! For The Sake of Animals
//Gentoo 2007.0/desktop, portage- 2.1.3_rc8, kernel @ 2.6.22-kamikaze1 (CFSv19) i686, "/" REISER4, ~x86, gcc-4.1.2, glibc-2.5-r3
//P4 2.4Ghz, 768 RAM , ATI Radeon X1600pro AGP, P4P800, SB Live ! 5.1, WDC 80GB |
|
| Back to top |
|
 |
devsk Advocate


Joined: 24 Oct 2003 Posts: 2632 Location: Bay Area, CA
|
Posted: Thu May 17, 2007 8:31 pm Post subject: |
|
|
| I am using default plugin, not the gzip. Also, I have a dual core (3800 X2), so cpu usage is fine. The problem is 100% IO wait which freezes the system for a while. A quick 'sync' (if I can issue one) resolves the issue. So, it seems like some processes aren't getting what they want from reiser4 when it is running in compress mode. A 'sync' forces the issue and gets processes out of the 'D' state. |
|
| Back to top |
|
 |
vipernicus Veteran


Joined: 17 Jan 2005 Posts: 1462 Location: Your College IT Dept.
|
Posted: Fri May 18, 2007 1:13 am Post subject: |
|
|
| devsk wrote: | | I am using default plugin, not the gzip. Also, I have a dual core (3800 X2), so cpu usage is fine. The problem is 100% IO wait which freezes the system for a while. A quick 'sync' (if I can issue one) resolves the issue. So, it seems like some processes aren't getting what they want from reiser4 when it is running in compress mode. A 'sync' forces the issue and gets processes out of the 'D' state. |
What I/O scheduler are you using? _________________ Viper-Sources Maintainer || nesl247 Projects || vipernicus.org blog |
|
| Back to top |
|
 |
devsk Advocate


Joined: 24 Oct 2003 Posts: 2632 Location: Bay Area, CA
|
Posted: Fri May 18, 2007 1:44 am Post subject: |
|
|
| vipernicus wrote: | | devsk wrote: | | I am using default plugin, not the gzip. Also, I have a dual core (3800 X2), so cpu usage is fine. The problem is 100% IO wait which freezes the system for a while. A quick 'sync' (if I can issue one) resolves the issue. So, it seems like some processes aren't getting what they want from reiser4 when it is running in compress mode. A 'sync' forces the issue and gets processes out of the 'D' state. |
What I/O scheduler are you using? | I tried deadline and CFQ. Currently using deadline. |
|
| Back to top |
|
 |
Gentree Watchman


Joined: 01 Jul 2003 Posts: 5077 Location: France, Old Europe
|
Posted: Sun May 20, 2007 8:47 pm Post subject: |
|
|
hi, this has been tempting me for a while. What are you calling "old" cpu here. I have Barton core XP clocking at 2.4GHz , is that likely to be upto the job?
@vipers , I've been using 2.6.20-nicus1 for a while which has been very reliable. Many thx. Which of your offerings would you recommend to test compression?
TIA.  _________________ Linux, because I'd rather own a free OS than steal one that's not worth paying for.
KX7-333 , AthlonXP1800+ @2.3GHz
AthlonXP-M on A7N8X @ 2.6/2.4GHz (winter/summer)
2.6.32-hh1 : portage ~x86 |
|
| Back to top |
|
 |
devsk Advocate


Joined: 24 Oct 2003 Posts: 2632 Location: Bay Area, CA
|
Posted: Mon May 21, 2007 6:07 pm Post subject: |
|
|
| Gentree wrote: | hi, this has been tempting me for a while. What are you calling "old" cpu here. I have Barton core XP clocking at 2.4GHz , is that likely to be upto the job?
| that's old... well, not really. I think you should be fine. Don't use compression plugin with vmware. I found too many problems with the compression plugin, with processes going into 'D' state and system requiring reboot. |
|
| Back to top |
|
 |
enderandrew l33t


Joined: 25 Oct 2005 Posts: 731
|
Posted: Mon Sep 17, 2007 8:14 pm Post subject: |
|
|
I'm swearing off ext4 and I've been considering giving JFS a try, or going back to Reiser4.
http://linuxhelp.150m.com/resources/fs-benchmarks.htm
I know for a while the Reiser4 crowd has been saying, just wait until we get the compression code in.
The above site has REALLY impressive numbers (if you can trust them). The site also throws in a political jab about Hans that I'm ignoring, but if I could have a reasonably fast FS that did transparent compression, that would be amazing. My main concern is that it will be very CPU intensive.
My next concern is if I wipe my root system (already corrupted by ext4 anyway) and do a fresh new install, then I need a 64-bit live CD/DVD with a recent enough kernel to create Reiser4 with compression.
Anyone know of one, or a quick/easy way to make a live CD myself? _________________ Nihilism makes me smile. |
|
| Back to top |
|
 |
kernelOfTruth Watchman


Joined: 20 Dec 2005 Posts: 5345 Location: Vienna, Austria; Germany; hello world :)
|
|
| Back to top |
|
 |
Mantaar Apprentice


Joined: 17 May 2007 Posts: 219
|
Posted: Tue Sep 18, 2007 1:43 am Post subject: |
|
|
The one time I had reiser4 on my system I went through a complicated bootstrapping procedure that involved installing a base system on one partition that was to become swap and then compiling a patched kernel from there.
Not that complicated actually, if you know what you're doing. Just one more reboot. Wasn't really that hard and rolling your own liveCD surely is likely to cause more blood sweat and tears... _________________ Error compiling committee.c: too many arguments to function. |
|
| Back to top |
|
 |
enderandrew l33t


Joined: 25 Oct 2005 Posts: 731
|
Posted: Tue Sep 18, 2007 2:52 am Post subject: |
|
|
You're a good man, Charlie Brown. _________________ Nihilism makes me smile. |
|
| Back to top |
|
 |
tokugawa n00b

Joined: 11 Oct 2002 Posts: 40 Location: Olesnica, Poland
|
Posted: Fri Dec 28, 2007 1:29 am Post subject: No! just No |
|
|
Hi,
I'll vote No for Reiser4. Maybe on non changing kernel works ok (not my case - I like to upgrade kernel). Maybe for 99.99% of time it works.
I have used it for very long time (years) and it worked... I was amazed by it's speed (emerge works great on top of it).
...but yesterday I saw 'reiser4 panic' on my fresh kernel - it happens randomly on half-year or so.
Yes, i know that my system is .. well.. experimental but I have switched to ext3 on / and problems went off (I remind once more - reiser4 failed on me only a few times during that years). Now I'm switching to ext3 on /home. And please remember "switched" means that I worked on ext3 for one year on one partition.
In my opinion ? Reiser4 is STILL experimental. |
|
| Back to top |
|
 |
|