| View previous topic :: View next topic |
| Author |
Message |
danone
Guru
Joined: 18 Jan 2004
Posts: 398
Location: Germany
|
 Posted: Sat Jun 05, 2004 7:53 am Post subject: |
 |
|
stahlsau..i only put -03 into the flag..since I got errors when i used processor spec options
_________________
[:: Processor: Intel Core 2 Duo E6300 ]::[ Mainboard: ASUS P5B Deluxe ]::[ GPU: nVidia 7900GTO ::]
[:: RAM: HyperX DDR2 800 ]::[ Samsung SH-183A SATA:: CREATiVE X-Fi XtremeMusic :: ] |
|
| Back to top |
|
 |
evermind
Guru
Joined: 10 Jan 2004
Posts: 322
|
| Posted: Sat Jun 05, 2004 7:59 am Post subject: |
|
|
| PrakashKC wrote: | @OneOfOne
Look closely at evermind's patch. It is NOT useless. The 0x4 is missing, therefore it is not identical. |
The patch (from reiserfs-mailinglist) was just for reiser4-2004.06.02-19.39-linux-2.6.7-rc1-mm1.diff. In any newer snapshot it´s already in
@OneOfOne thx for the new reiser4patch I´ll test it. |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sat Jun 05, 2004 8:03 am Post subject: |
|
|
| evermind wrote: | | PrakashKC wrote: | @OneOfOne
Look closely at evermind's patch. It is NOT useless. The 0x4 is missing, therefore it is not identical. |
The patch (from reiserfs-mailinglist) was just for reiser4-2004.06.02-19.39-linux-2.6.7-rc1-mm1.diff. In any newer snapshot its already in
@OneOfOne thx for the new reiser4patch Ill test it. |
3:12hrs up time now, mysql doesnt go corrupting the database anymore and generally it feels good 
mysql survived emerge regen, emerge -s python and emerge -euDp world.
peace |
|
| Back to top |
|
|
ed0n
l33t
Joined: 23 Apr 2003
Posts: 638
Location: Prishtine/Kosove
|
| Posted: Sat Jun 05, 2004 8:04 am Post subject: |
|
|
| Working like a charm here + reiser4 patch. |
|
| Back to top |
|
|
eikketk
Apprentice
Joined: 03 Jun 2003
Posts: 270
Location: Belgium
|
| Posted: Sat Jun 05, 2004 10:51 am Post subject: |
|
|
| OneOfOne wrote: | @eikke hey, you gotta give me a share of the profit you get!! |
What profit?  
_________________
Working day and night to enhance the Linux Desktop Experience
Homepage |
|
| Back to top |
|
|
evermind
Guru
Joined: 10 Jan 2004
Posts: 322
|
| Posted: Sat Jun 05, 2004 10:53 am Post subject: |
|
|
I haven't fsck.my reiser4 partition before (they had errors) but this shouldn't occur either
after some time I got this and all is freezing
| Code: |
Unable to handle kernel paging request at virtual address 000104dc
printing eip:
c01b15c5
*pde = 00000000
Oops: 0000 [#1]
PREEMPT
Modules linked in: sg radeon md5 ipv6 pppoe pppox lp snd_seq_midi snd_via82xx snd_ac97_codec snd_mpu401_uart snd_rawmidi snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_pcm snd_page_alloc snd_timer snd_mixer_oss snd soundcore usbhid ide_scsi usblp uhci_hcd usbcore 3c59x sr_mod cdrom
CPU: 0
EIP: 0060:[<c01b15c5>] Not tainted VLI
EFLAGS: 00010202 (2.6.7-rc2-love2-reiser4)
EIP is at unhash_unformatted_node_nolock+0x55/0xc0
eax: 00010460 ebx: d3780440 ecx: 00010504 edx: 00000000
esi: c135a000 edi: 00000000 ebp: d382f814 esp: c135bc84
ds: 007b es: 007b ss: 0068
Process pdflush (pid: 30, threadinfo=c135a000 task=c139f160)
Stack: c135a000 d3780440 c01b24b4 00000000 d3780470 c135a000 d3780440 d30f1dc0
c01be75a 00000000 d382f800 c135a000 d3fc5e00 c01c595e c010494a 00000082
d3871460 d30f1d60 c135bd60 d3871460 d3871460 c01ca4fe d387146c d30f1d60
Call Trace:
[<c01b24b4>] jnode_try_drop+0x164/0x1b0
[<c01be75a>] invalidate_list+0x4a/0x60
[<c01c595e>] reiser4_write_logs+0x6e/0x2e0
[<c010494a>] apic_timer_interrupt+0x1a/0x20
[<c01ca4fe>] release_prepped_list+0xfe/0x120
[<c01ca098>] finish_fq+0x38/0x50
[<c01ca121>] finish_all_fq+0x71/0xd0
[<c01be000>] commit_current_atom+0x130/0x220
[<c01be949>] try_commit_txnh+0xe9/0x1b0
[<c01bea3f>] commit_txnh+0x2f/0xb0
[<c01bd700>] txn_end+0x30/0x40
[<c01be648>] flush_some_atom+0x1e8/0x2b0
[<c01ce66a>] writeout+0x4a/0xd0
[<c01ce74b>] reiser4_sync_inodes+0x5b/0xa0
[<c01ce6f0>] reiser4_sync_inodes+0x0/0xa0
[<c0170229>] sync_sb_inodes+0x19/0x20
[<c0170266>] writeback_inodes+0x36/0xc0
[<c013898c>] background_writeout+0x9c/0xf0
[<c0139488>] __pdflush+0xf8/0x200
[<c011713f>] set_user_nice+0xaf/0xd0
[<c0139590>] pdflush+0x0/0x20
[<c01395aa>] pdflush+0x1a/0x20
[<c01388f0>] background_writeout+0x0/0xf0
[<c0139590>] pdflush+0x0/0x20
[<c012b6e5>] kthread+0x95/0xd0
[<c012b650>] kthread+0x0/0xd0
[<c010228d>] kernel_thread_helper+0x5/0x18
Code: 89 c2 8d 42 1c 8d 44 20 00 39 da 75 ea 8b 53 1c 89 11 8d 74 26 00 be 00 e0 ff ff 21 e6 ff 46 14 8b 43 18 8b 08 8d 81 5c ff ff ff <8b> 50 7c 4a 85 d2 89 50 7c 75 0a 81 a1 08 01 00 00 ff fd ff ff
<6>note: pdflush[30] exited with preempt_count 2
|
now I fscked all partition and try it again
EDIT: after fscking I get the same error after some time so I went back to my previous patch |
|
| Back to top |
|
|
primero.gentoo
Guru
Joined: 23 Dec 2003
Posts: 402
|
| Posted: Sat Jun 05, 2004 1:40 pm Post subject: |
|
|
I've not yet understood wich process scheduler is the "default" for love sources.
I'm using right now 2.6.7-rc2-love1 because i want "Staircase" and because in love2 it has changed to "Nick"... is this right?
Could be possible to make a separate patch to switch from one to the other? i don't really know how to do it but if someone could explain me the basis of Kernel patching i could try it ... so i could also include Systrace patch to this wonderfull patchset
bye
_________________
"Linux, the choice of a GNU generation"
==Micro$oft - just say NO==
(L#USER 353039) |
|
| Back to top |
|
|
garlicbread
Apprentice
Joined: 06 Mar 2004
Posts: 182
|
| Posted: Sat Jun 05, 2004 2:04 pm Post subject: |
|
|
I've noticed that USB memory (MTD) sticks still don't work
it doesn't generate an OOPS this time around which is better, but it doesn't appear to register the device within the SCSI system
(which also interferes with a reboot when the system trys to take USB down) |
|
| Back to top |
|
|
alinv
Guru
Joined: 19 Nov 2002
Posts: 395
Location: Bucharest
|
| Posted: Sat Jun 05, 2004 2:29 pm Post subject: |
|
|
| BNoise wrote: | [code]
How do i fix this?
[code]
LD vmlinux
fs/built-in.o(.init.text+0x109f): In function `proc_misc_init':
: undefined reference to `proc_schedstat_operations'
make: *** [vmlinux] Error 1
[/code]
|
Disable "Collect scheduler statistics".
_________________
Ever tried. Ever failed. No matter. Try Again. Fail again. Fail better.
S.B. |
|
| Back to top |
|
|
spb
Retired Dev
Joined: 02 Jan 2004
Posts: 2135
Location: Cambridge, UK
|
| Posted: Sat Jun 05, 2004 2:32 pm Post subject: |
|
|
| ACX100 is broken in this release. The module name changed with 0.2.0_pre8, so you need to rename drivers/net/wireless/acx100 to drivers/net/wireless/acx, and update drivers/net/wireless/Makefile accordingly. Having done that it compiles, thought I haven't tried using it yet. |
|
| Back to top |
|
|
Pink
Veteran
Joined: 24 Jul 2003
Posts: 1062
|
| Posted: Sat Jun 05, 2004 3:24 pm Post subject: |
|
|
| garlicbread wrote: | I've noticed that USB memory (MTD) sticks still don't work
it doesn't generate an OOPS this time around which is better, but it doesn't appear to register the device within the SCSI system
(which also interferes with a reboot when the system trys to take USB down) |
It's an mm-patch problem, not a love-sources problem.
It messes up usb at the moment - but remember, mm isn't classed as stable, never mind love-sources |
|
| Back to top |
|
|
Cerement
Guru
Joined: 14 Jun 2003
Posts: 404
|
| Posted: Sat Jun 05, 2004 4:45 pm Post subject: |
|
|
| primero.gentoo wrote: | I've not yet understood wich process scheduler is the "default" for love sources.
I'm using right now 2.6.7-rc2-love1 because i want "Staircase" and because in love2 it has changed to "Nick"... is this right?
Could be possible to make a separate patch to switch from one to the other? |
love-sources uses (and traditionally prefers) Nick Piggin's nicksched CPU scheduler. The switch to Con Kolivas' staircase CPU scheduler was a one time deal (and I'm guessing it was in response to some benchmarks I posted earlier: too many people thought that an artificial benchmark and real world results were equivalent, and OneOfOne was more than happy to prove them wrong )
Currently it is only possible to switch back and forth on IO schedulers, CPU schedulers are an all-or-nothing affair, but after a comment by Con Kolivas on KernelTrap, it sounds very probable that, due to the recent proliferation of CPU schedulers, sometime early in kernel 2.7 it will be possible to switch CPU schedulers back and forth.
[editorial] -- if you want the staircase scheduler, stick with the CK patchsets (2.6.7-*-ck1 should be coming soon) -- if you want nicksched, stick with love-sources -- and after seeing the last round of patches, it is pretty clear that, while they do not touch common code, staircase and nickvm "do not play well together"  |
|
| Back to top |
|
|
Cerement
Guru
Joined: 14 Jun 2003
Posts: 404
|
| Posted: Sat Jun 05, 2004 4:47 pm Post subject: |
|
|
| PickledOnion wrote: | | but remember, mm isn't classed as stable, never mind love-sources |
I still like steel300's quote from #gentoo a whiles back when someone asked about love-sources:
| steel300 wrote: | | If mm-sources is bleeding edge, then love-sources is open heart surgery. |
|
|
| Back to top |
|
|
primero.gentoo
Guru
Joined: 23 Dec 2003
Posts: 402
|
| Posted: Sat Jun 05, 2004 5:45 pm Post subject: |
|
|
| Cerement wrote: |
[editorial] -- if you want the staircase scheduler, stick with the CK patchsets (2.6.7-*-ck1 should be coming soon) -- if you want nicksched, stick with love-sources -- and after seeing the last round of patches, it is pretty clear that, while they do not touch common code, staircase and nickvm "do not play well together" |
Ok, is it possible to apply 2.6.7-rc2-ck1 on love sources? maybe i'm mad, but i've read "staircase scheduler" in 2.6.7-rc2-love1 notes ... is this real or not?
bye
_________________
"Linux, the choice of a GNU generation"
==Micro$oft - just say NO==
(L#USER 353039) |
|
| Back to top |
|
|
Cerement
Guru
Joined: 14 Jun 2003
Posts: 404
|
| Posted: Sat Jun 05, 2004 6:00 pm Post subject: |
|
|
| primero.gentoo wrote: | | Ok, is it possible to apply 2.6.7-rc2-ck1 on love sources? maybe i'm mad, but i've read "staircase scheduler" in 2.6.7-rc2-love1 notes ... is this real or not? |
You cannot apply staircase to love-sources, because love-sources already has nicksched in place. Yes, it was in one release of love-sources, that was what I was explaining above. |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sat Jun 05, 2004 8:19 pm Post subject: |
|
|
| primero.gentoo wrote: | | Cerement wrote: |
[editorial] -- if you want the staircase scheduler, stick with the CK patchsets (2.6.7-*-ck1 should be coming soon) -- if you want nicksched, stick with love-sources -- and after seeing the last round of patches, it is pretty clear that, while they do not touch common code, staircase and nickvm "do not play well together" |
Ok, is it possible to apply 2.6.7-rc2-ck1 on love sources? maybe I'm mad, but I've read "staircase scheduler" in 2.6.7-rc2-love1 notes ... is this real or not?
bye |
well, staircase proved to be a bad idea in last love, didnt it? 
maybe in the future I'd make sperate patches for both nicksech/staircase, but so far staircase doesnt like the love!
@evermind i didnt have that, but when i fsck my / it seg faults (which is for some reason beyond me is normal) and i found few corruptions on another reiser4 partition which got fixed with --build-fs.
redeeman says there's a bug that will make it freeze the computer under heavy load.
peace
peace |
|
| Back to top |
|
|
Suicidal
l33t
Joined: 30 Jul 2003
Posts: 959
Location: /dev/null
|
| Posted: Sat Jun 05, 2004 11:10 pm Post subject: |
|
|
| Any way to get the NX (no execute) patch into love? |
|
| Back to top |
|
|
evermind
Guru
Joined: 10 Jan 2004
Posts: 322
|
| Posted: Sun Jun 06, 2004 1:31 am Post subject: |
|
|
| OneOfOne wrote: |
@evermind i didnt have that, but when i fsck my / it seg faults (which is for some reason beyond me is normal) and i found few corruptions on another reiser4 partition which got fixed with --build-fs.
redeeman says there's a bug that will make it freeze the computer under heavy load.
peace
peace |
yeah I had the same expirience with your latest patch - but when i use my personal patch it works for me (but only for me - on a friend machine it also crashes)
I tested your patch under heavy load and then I experienced a freeze
but this is the greatest community I ever found 
hf
save the penguins |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sun Jun 06, 2004 3:29 am Post subject: |
|
|
| evermind wrote: | | OneOfOne wrote: |
@evermind i didnt have that, but when i fsck my / it seg faults (which is for some reason beyond me is normal) and i found few corruptions on another reiser4 partition which got fixed with --build-fs.
redeeman says there's a bug that will make it freeze the computer under heavy load.
peace
peace |
yeah I had the same experience with your latest patch - but when i use my personal patch it works for me (but only for me - on a friend machine it also crashes)
I tested your patch under heavy load and then I experienced a freeze
but this is the greatest community I ever found
hf
save the penguins |
okie dokie, i backed up my stuff and reformatted, copied everything back (8.7GB) and ran fsck and it found 2 corruptions, fixed with --build-fs and so far so good.
I haven't had any of those "heavy load" freezes, glibc froze the comp during ./configure but now its compiling fine.
last night php also froze the comp while doing "pear config" stuff, but i haven't tried after the reformat.
so far mysql works perfectly and finally i'm able to enjoy my mysql portage module.
I'll report back in a few.
peace |
|
| Back to top |
|
|
Lord Angelbane
n00b
Joined: 11 Apr 2004
Posts: 8
|
| Posted: Sun Jun 06, 2004 3:38 am Post subject: |
|
|
Not sure Suicidal, but it looks like the NX patch you pointed to is already in this love. I tried patching it myself and all of the hunks were rejected; and each hunk already appears in the files. So, looks like you got your wish.
Jeff
_________________
ACX100 project: acx100.sourceforge.net |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sun Jun 06, 2004 4:21 am Post subject: |
|
|
| Lord Angelbane wrote: | Not sure Suicidal, but it looks like the NX patch you pointed to is already in this love. I tried patching it myself and all of the hunks were rejected; and each hunk already appears in the files. So, looks like you got your wish.
Jeff |
?! noone told me that i included it in love!
lol probably got broken by nicksched or something but i didn't add it.
i'll try to get it in (or atleast as an optional patch).
peace |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sun Jun 06, 2004 4:24 am Post subject: |
|
|
ok i made a hack and disabled the metas/ support by default (reversed the nopseudo option), that (in theory) should fix the weird permission problems with glibc and apache.
It's compiling now, I'll report back asap.
peace |
|
| Back to top |
|
|
OneOfOne
Guru
Joined: 28 May 2003
Posts: 368
|
| Posted: Sun Jun 06, 2004 6:22 am Post subject: |
|
|
| OneOfOne wrote: | ok i made a hack and disabled the metas/ support by default (reversed the nopseudo option), that (in theory) should fix the weird permission problems with glibc and apache.
It's compiling now, I'll report back asap.
peace |
glibc compiles and apache's permissions problems are gone but apache still goes randomaly zombie and uses 99% cpu.
[poll] should the reiser4 metas/ feature be disabled by default in next love?
if you dont use it, it does more harm than good ATM.
peace |
|
| Back to top |
|
|
stahlsau
Guru
Joined: 09 Jan 2004
Posts: 584
Location: WildWestwoods
|
| Posted: Sun Jun 06, 2004 7:56 am Post subject: |
|
|
| Quote: | | [poll] should the reiser4 metas/ feature be disabled by default in next love? |
YES |
|
| Back to top |
|
|
metalh34d
Tux's lil' helper
Joined: 04 Feb 2004
Posts: 126
|
| Posted: Sun Jun 06, 2004 8:33 am Post subject: |
|
|
emerge nvidia-kernel looks like it works smoothly but when I try to use the module it won't work. Says error inserting, invalid format when I do a modprobe nvidia.
Here's a snippet of me trying to startx:
| Code: | (II) Setting vga for screen 0.
(II) Loading sub module "vgahw"
(II) LoadModule: "vgahw"
(II) Loading /usr/X11R6/lib/modules/libvgahw.a
(II) Module vgahw: vendor="X.Org Foundation"
compiled for 6.7.0, module version = 0.1.0
ABI class: X.Org Video Driver, version 0.7
(**) NVIDIA(0): Depth 24, (--) framebuffer bpp 32
(==) NVIDIA(0): RGB weight 888
(==) NVIDIA(0): Default visual is TrueColor
(==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
(--) NVIDIA(0): Linear framebuffer at 0xF0000000
(--) NVIDIA(0): MMIO registers at 0xF8000000
(EE) NVIDIA(0): Failed to initialize the NVIDIA kernel module!
(EE) NVIDIA(0): *** Aborting ***
(II) UnloadModule: "nvidia"
(II) UnloadModule: "vgahw"
(II) Unloading /usr/X11R6/lib/modules/libvgahw.a
(EE) Screen(s) found, but none have a usable configuration.
Fatal server error:
no screens found
Please consult the The X.Org Foundation support
at http://wiki.X.Org
for help.
Please also check the log file at "/var/log/Xorg.0.log" for additional information.
|
|
|
| Back to top |
|
|
|
Unsupported Software
