| View previous topic :: View next topic |
| Author |
Message |
Till Varoquaux
n00b
Joined: 28 Oct 2004
Posts: 24
|
 Posted: Sun Oct 31, 2004 8:25 pm Post subject: |
 |
|
Seppe.
nice patchset really. However:
_why supermount?
It sits in kernel space bringing potentially a lot of instability. And you ejecet or umount a drive supermount is gone...
Con Kolivas doesn't seem to be very found of it either.
The point of file systems like FUSE and LUFS is to put as much as possible in userspace. That's not the way things work with supermount.
I am curious,what is wrong with Ivman? |
|
| Back to top |
|
 |
MagicTom
n00b
Joined: 23 Oct 2003
Posts: 60
Location: Nancy, France
|
| Posted: Mon Nov 01, 2004 1:26 am Post subject: problem on AMD64 |
|
|
Hi,
I can't compile this kernel on my Athlon64 laptop. I get this error :
| Code: | CHK include/linux/version.h
make[1]: `arch/x86_64/kernel/asm-offsets.s' is up to date.
CHK include/linux/compile.h
dnsdomainname: Unknown host
CC kernel/fork.o
kernel/fork.c: In function `do_fork':
kernel/fork.c:1241: error: structure has no member named `xcs'
make[1]: *** [kernel/fork.o] Error 1
make: *** [kernel] Error 2 |
Someone else had the same problem with -nitro1.
I quickly checked : "nitro" do_fork.c is different from "2.6.9-gentoo-r1" do_fork.c, and "2.6.9-gentoo-r1" works perfectly. The problem is in the pt_regs structure, defined in /usr/src/linux-.../include/asm-your-platform-here/ptrace.h. Indeed, there is no xcs member in pt_regs as defined in asm-x86_64/ptrace.h.
I guess there is a patch for the Athlon64 platform that is included in gentoo-dev-sources but not in -nitro2...
Does anyone know where I could download this patch ?
And will it be included in the next nitro release ?
_________________
There's no place like ::1 |
|
| Back to top |
|
|
bollucks
l33t
Joined: 27 Oct 2004
Posts: 606
|
| Posted: Mon Nov 01, 2004 1:49 am Post subject: |
|
|
That doesn't sound like he's not fond of it at all. He has mentioned on his IRC channel that he still uses it himself. |
|
| Back to top |
|
|
devsk
Advocate
Joined: 24 Oct 2003
Posts: 3003
Location: Bay Area, CA
|
| Posted: Mon Nov 01, 2004 2:24 am Post subject: |
|
|
| wonder why patch for software suspend not included in the ebuild itself? I am thinking of going nitro and that's the first thing that struck me. |
|
| Back to top |
|
|
desertstalker
Apprentice
Joined: 18 Sep 2004
Posts: 209
|
| Posted: Mon Nov 01, 2004 6:28 am Post subject: |
|
|
| Apparantly the swsusp2 patch caues some problems with compiling when it was not enabled. |
|
| Back to top |
|
|
Vlad
Apprentice
Joined: 09 Apr 2002
Posts: 264
Location: San Diego, California
|
| Posted: Mon Nov 01, 2004 7:36 am Post subject: |
|
|
| thubble wrote: | | archrival wrote: | | Does anyone else have a random lockup problem with the recent nitro kernels? It normally happens at startup, but not always at the same time. Sometimes it makes it all the way to the login prompt, othertimes it stops at Starting lo, hdparm, calculating module_depencies, etc. It will lock up eventually though, even after a little usage. When I am able to look at dmesg, I normally see something with DMA errors on hda (my linux root is hda4). I have no problem with mm-sources and I was curious if the new libata code might be the culprit. I have a VIA IDE chipset and I use the same options on both kernels. SMP with HT, Reiser4, Pentium4, VIA IDE driver. mm-sources work like a charm. I just like what nitro has to offer. |
It seems to be an incompatibility between Reiser4 and SMP. Try disabling SMP. You may have to disable hyperthreading in your BIOS if that doesn't work. |
I can say for certain that Reiser4 and SMP are not the problem, as I'm running this kernel on two computers, neither of which are SMP, and only one of which is using Reiser4.
I've found that sometimes hitting control+c during the 'freezes' in startup will spit out this error:
/lib/rcscripts/sh/rc-services.sh: line 589: [: : integer expression expected
After doing some looking, I discovered that you can stop this behavor by turning off parallel startup in /etc/conf.d/rc. Unfortunately, I still had major problems booting up, but I suspect that these problems were more my fault than anything elses. In any event, disabling parallel startup will temporarily fix that problem.
Last edited by Vlad on Mon Nov 01, 2004 9:10 am; edited 2 times in total |
|
| Back to top |
|
|
devsk
Advocate
Joined: 24 Oct 2003
Posts: 3003
Location: Bay Area, CA
|
| Posted: Mon Nov 01, 2004 7:47 am Post subject: |
|
|
for the first time ever, I could get any kind of suspend to work on my desktop linux. thanks seppe!!
boot time ~20 seconds into KDE. my X and KDE are not very optimized I think.
seems darn stable to me, probably as stable as any Gentoo self compiled kernel I have used. thanks again for this great set of patches and guide!! |
|
| Back to top |
|
|
bollucks
l33t
Joined: 27 Oct 2004
Posts: 606
|
| Posted: Mon Nov 01, 2004 9:03 am Post subject: |
|
|
| Vlad wrote: | I've found that sometimes hitting control+c during the 'freezes' in startup will spit out this error:
/lib/rcscripts/sh/rc-services.sh: line 589: [: : integer expression expected
After that it will continue booting, sometimes requiring multiple control+c's to (sometimes) make it to the login.
It must be something in the rc-scripts and (guessing here) the name of the kernel. It doesn't seem to do anything to the stability of the platform; sometimes I can ssh in and do a # killall rc and it will spit out a prompt, after which I can login and start any unstarted services. Could this be a naming issue...? |
Parallel boot scripts break with staircase. They are fragile to begin with and staircase speeds up the loading of some things and they end up trying to load out of order. Try disabling parallel boot. |
|
| Back to top |
|
|
bollucks
l33t
Joined: 27 Oct 2004
Posts: 606
|
| Posted: Mon Nov 01, 2004 9:04 am Post subject: |
|
|
| Vlad wrote: | I've found that sometimes hitting control+c during the 'freezes' in startup will spit out this error:
/lib/rcscripts/sh/rc-services.sh: line 589: [: : integer expression expected
After that it will continue booting, sometimes requiring multiple control+c's to (sometimes) make it to the login.
It must be something in the rc-scripts and (guessing here) the name of the kernel. It doesn't seem to do anything to the stability of the platform; sometimes I can ssh in and do a # killall rc and it will spit out a prompt, after which I can login and start any unstarted services. Could this be a naming issue...? |
Parallel boot scripts break with staircase. They are fragile to begin with and staircase speeds up the loading of some things and they end up trying to load out of order. Try disabling parallel boot. |
|
| Back to top |
|
|
Crazor
Tux's lil' helper
Joined: 23 Apr 2003
Posts: 131
|
| Posted: Mon Nov 01, 2004 6:03 pm Post subject: |
|
|
when I compile in Enhanced SpeedStep support (read: frequency scaling) for my Intel Centrino laptop, I get a kernel OOPS at boot which disappears when I put "acpi=off" on the kernel command line:
| Code: | ACPI-0292 *** Error: Looking up [ACST] in namespace, AE_ALREADY_EXISTS
ACPI-1133 *** Error: [NULL NAME], AE_ALREADY_EXISTS
Unable to handle kernel NULL pointer dereference at virtual address 00000010 |
followed by the usual stacktrace etc. (too lazy to note everything and type it in here. will do that if it is useful.)
BTW, is there any way to get the kernel output (like dmesg) when it oopses other than serial console (my laptop aint got no damnn serial port =[ ) |
|
| Back to top |
|
|
racoontje
Veteran
Joined: 19 Jul 2004
Posts: 1290
|
| Posted: Mon Nov 01, 2004 10:17 pm Post subject: |
|
|
| Seppe, are you putting grsecurity/PaX patches in your patchset? |
|
| Back to top |
|
|
timmy334
n00b
Joined: 24 Sep 2003
Posts: 69
Location: Montgomery, AL
|
| Posted: Tue Nov 02, 2004 12:32 am Post subject: reiser4 not showing up |
|
|
When I applied the patches, reiser4 didn't show up in menuconfig at all. What gives? The patching process went without any problems so anybody have any ideas?
_________________
It is by caffeine alone I set my mind in motion,
It is by the beans of Java my thoughts acquire speed,
My hands acquire shaking, the shaking becomes a warning,
It is by caffeine alone I set my mind in motion |
|
| Back to top |
|
|
cptmorgan
Tux's lil' helper
Joined: 28 Jan 2004
Posts: 129
Location: California
|
| Posted: Tue Nov 02, 2004 1:23 am Post subject: Re: reiser4 not showing up |
|
|
| timmy334 wrote: | | When I applied the patches, reiser4 didn't show up in menuconfig at all. What gives? The patching process went without any problems so anybody have any ideas? |
remove 4k kernal stacks from kernel hacking |
|
| Back to top |
|
|
devsk
Advocate
Joined: 24 Oct 2003
Posts: 3003
Location: Bay Area, CA
|
| Posted: Tue Nov 02, 2004 4:53 am Post subject: |
|
|
| with xorg's 'nv' driver, the suspend works like charm from within X, exactly the thing one imagines. no such luck with nvidia binary drivers. anyone? |
|
| Back to top |
|
|
petrjanda
Veteran
Joined: 05 Sep 2003
Posts: 1557
Location: Brno, Czech Republic
|
| Posted: Tue Nov 02, 2004 1:15 pm Post subject: |
|
|
Im getting this error while compiling nitro2.
| Code: |
LD .tmp_vmlinux1
sound/built-in.o(.text+0x33531): In function `snd_seq_device_register_driver':
: undefined reference to `snd_seq_autoload_lock'
sound/built-in.o(.text+0x335b9): In function `snd_seq_device_register_driver':
: undefined reference to `snd_seq_autoload_unlock'
sound/built-in.o(.text+0x335ec): In function `snd_seq_device_register_driver':
: undefined reference to `snd_seq_autoload_unlock'
sound/built-in.o(.text+0x335fb): In function `snd_seq_device_register_driver':
: undefined reference to `snd_seq_autoload_unlock'
make: *** [.tmp_vmlinux1] Error 1
|
_________________
There is, a not-born, a not-become, a not-made, a not-compounded. If that unborn, not-become, not-made, not-compounded were not, there would be no escape from this here that is born, become, made and compounded. - Gautama Siddharta |
|
| Back to top |
|
|
artic
Veteran
Joined: 22 Nov 2003
Posts: 1208
|
| Posted: Tue Nov 02, 2004 2:01 pm Post subject: |
|
|
I get this errors compiling kernel reiser4 modules,it happen with 6 reiser4 modules:
| Code: |
CC fs/reiser4/page_cache.o
fs/reiser4/page_cache.c: In function `set_page_dirty_internal':
fs/reiser4/page_cache.c:536: warning: passing arg 1 of `_read_lock_irq' from incompatible pointer type
fs/reiser4/page_cache.c:552: warning: passing arg 1 of `_read_unlock_irq' from incompatible pointer type
fs/reiser4/page_cache.c: In function `capture_reiser4_inodes':
fs/reiser4/page_cache.c:597: warning: passing arg 1 of `_read_lock_irqsave' from incompatible pointer type
fs/reiser4/page_cache.c:603: warning: passing arg 1 of `_read_unlock_irqrestore' from incompatible pointer type
|
??????????
_________________
Pwd by Gentoo | 2.6.28-gentoo-x64 | kde4.2| ext4 | xorg 7.4 | |
|
| Back to top |
|
|
eikketk
Apprentice
Joined: 03 Jun 2003
Posts: 270
Location: Belgium
|
| Posted: Tue Nov 02, 2004 7:24 pm Post subject: |
|
|
First time I tried nitro (I'm a love user)
2 comments: after a while, my Firefox broke. I could launch new instances, but before I visited 2 pages, it stopped responding. Killall was the solution.
Next, little compile error: didnt want to compile with swsusp2 (yeah I patched) gzip compression enabled (undefined MAX function, easy to track down the problem by watching the compile output)
_________________
Working day and night to enhance the Linux Desktop Experience 
Homepage |
|
| Back to top |
|
|
Headrush
Watchman
Joined: 06 Nov 2003
Posts: 5597
Location: Bizarro World
|
| Posted: Tue Nov 02, 2004 7:26 pm Post subject: |
|
|
Runs great here.
Last edited by Headrush on Wed Nov 03, 2004 12:05 am; edited 1 time in total |
|
| Back to top |
|
|
ryceck
Apprentice
Joined: 13 Jan 2004
Posts: 195
|
| Posted: Tue Nov 02, 2004 11:20 pm Post subject: |
|
|
The nitro1 had me some stability problems but this one is a freaking rock dude 
Fast, responsive and very good and stable.
And the best thing is: it even hibernates correctly (sometimes  ).
Nitro kicks them all |
|
| Back to top |
|
|
deepspace9
Apprentice
Joined: 29 Jan 2003
Posts: 214
Location: Netherlands
|
| Posted: Tue Nov 02, 2004 11:57 pm Post subject: |
|
|
Hi there, I'm just installing this kernel on a fresh system, but it can't seam to autodetect my software raid partitions  Now I can't boot. I know the raid array;s are oke, I installed a complete system on it with a liveCD.
Is it just broken? I realy need reiserfs4 support, ans this kernel is the only one that has it. Please help! |
|
| Back to top |
|
|
Robin79
l33t
Joined: 18 Oct 2003
Posts: 632
Location: /home/valdemarsvik
|
| Posted: Wed Nov 03, 2004 12:22 am Post subject: |
|
|
| Till Varoquaux wrote: | Seppe.
nice patchset really. However:
_why supermount?
It sits in kernel space bringing potentially a lot of instability. And you ejecet or umount a drive supermount is gone...
Con Kolivas doesn't seem to be very found of it either.
The point of file systems like FUSE and LUFS is to put as much as possible in userspace. That's not the way things work with supermount.
I am curious,what is wrong with Ivman? |
What is FUSE and LUFS? Cause supermount messes up for me in Point2Play!
_________________
Gentoo 2005.0
ASUS P5AD2 Deluxe MB
Intel P4 3.2GHz
PCI-E Nvidia Geforce PCX 5750
--------------------
Registered Linux User Nr # 319050 http://counter.li.org |
|
| Back to top |
|
|
deepspace9
Apprentice
Joined: 29 Jan 2003
Posts: 214
Location: Netherlands
|
| Posted: Wed Nov 03, 2004 4:29 pm Post subject: |
|
|
I gave up raid hopes and killed the array and copied al data do a new reiserfs partition: /dev/sda3
after reboot, the nitro kernel complains about a wrong partition table (it works perfectly with the liveCD, which has the 2.6.8-nitro5 kernel), and complains that /dev/sda3 cannot be found An old 2.6.3 kernel (standard)) works just fine, but nu reiserfs4 support)
WTF is wrong here?
My scsi driver is the aic79xx ..should be a piece of cake realy!
And all this to use reiserfs4... I gues it might not be worth the trouble..
If I don't get ansers soon, i'll just use reiser3 and a standard kernel with a new raid array.. has always worked
I'm is pissed of. I thought it would be a pease of cake  and would be up and running within a few houers (as always) |
|
| Back to top |
|
|
peterk0
Apprentice
Joined: 25 Mar 2003
Posts: 188
Location: Prague
|
| Posted: Wed Nov 03, 2004 5:04 pm Post subject: |
|
|
| for me best kernel ever !!! nice job |
|
| Back to top |
|
|
Camp
Tux's lil' helper
Joined: 31 May 2004
Posts: 82
Location: Moscow
|
| Posted: Wed Nov 03, 2004 6:58 pm Post subject: |
|
|
I've just solved nvidia invalid module format under 2.6.9
GREAT WORK SEPPE!
_________________
registered Linux user #357267
GENTOO |
|
| Back to top |
|
|
eean
Guru
Joined: 10 Jun 2002
Posts: 355
Location: Kirksville, MO USA
|
| Posted: Wed Nov 03, 2004 7:39 pm Post subject: Re: Azureus |
|
|
| Avenger902 wrote: | | I like this kernel... its my first time with nitro... Can't get Azureus to stay up though... it keeps crashing.... Anyone able to run Azureus ok with nitro2? or is there a way I can work around it? |
You mean Azureus crashs or the whole computer?
Next time there's a new vanilla kernel released I'll be switching to that. Stablity has noticably lessened with nitro, whereas any speed up is not noticed.
_________________
Amarok - Rediscover your Music |
|
| Back to top |
|
|
|
Unsupported Software
