2.6.11-love2 (usb & genetic-as deplug. oops fix)

[sobers_2002]

this one is much better than the last one it seems

one thing that i wanted to know was how to use the nvidiafb??

as in the syntax and if there were any advantages of it.

[Legoguy]

So yeah, I'm at school now with my laptop... and usually in the morning I unplug my logitech mx1000 from my box, without turning on the monitor, so I can use it at school.. and when I get to school this morning, woah, where's my webserver... ssh server... everything...

Now I know where it is... oopsing at home... sigh.
Does oopsing usually stay at 100% cpu usage, or is it 0%? I don't want to come home to a 66 deg C processor...

[WladyX]

Code: Select all

 pass rootflags=nopseudo to disable metas/ for root reiser4

Do i need to do this if i have reiser4 on / ? Thanks.



Also during make i get this:

Code: Select all

 LD      drivers/usb/input/built-in.o
  CC [M]  drivers/usb/input/hid-core.o
drivers/usb/input/hid-core.c:51: error: parse error before ';' token
make[3]: *** [drivers/usb/input/hid-core.o] Error 1
make[2]: *** [drivers/usb/input] Error 2
make[1]: *** [drivers/usb] Error 2
make: *** [drivers] Error 2

[rmh3093]

the nvidiafb isnt worth your time because if you use nvidiafb, nvidia-kernel wont work because they fight over the card

[Merlin-TC]

I think you need to set the mouse polling rate. At least that's what gave me an error.
You can set it in the kernel menu under the USB HID settings.

[WladyX]

Thanks for the reply i got it to work, are there any "safe" configurations need for reiser4, besides disabling the 4k stacks?

[fallow]

Also during make i get this:

Code: Select all

 LD      drivers/usb/input/built-in.o
  CC [M]  drivers/usb/input/hid-core.o
drivers/usb/input/hid-core.c:51: error: parse error before ';' token
make[3]: *** [drivers/usb/input/hid-core.o] Error 1
make[2]: *** [drivers/usb/input] Error 2
make[1]: *** [drivers/usb] Error 2
make: *** [drivers] Error 2

If You want to complie it , You must set something to USB HID Mouse Interrupt Polling Interval.
Without any setting is wrong.
-->> BAD example

Code: Select all

()    USB HID Mouse Interrupt Polling Interval (NEW)    

-->> GOOD example

Code: Select all

(2)    USB HID Mouse Interrupt Polling Interval (NEW)    

Code: Select all

 pass rootflags=nopseudo to disable metas/ for root reiser4

Do i need to do this if i have reiser4 on / ? Thanks.

in quick answer no .

i have a couple of problems with love-sources
-----------------------------------------------------
i'm not really good at this stuff, but i tried love-sources 2.6.11-r1 and on some heavier compilations (those lasting longer then 25min) my system would go into complete halt. CPU was extremely hot at that moment (i cannot tell you exactly, but the very air inside of the box was at least 40C, which is absolutely rare for me). it happened numerous times on love-sources. gentoo-dev-sources work all right all the time, however.
i'm not an overclocker. i don't do things like that. i have an 1.7Ghz celeron (pentium4-compatible cpu)
i did not use clock modulation features in kernel config, the config was the same as in my gentoo-dev-sources 2.6.10 kernel. i only enabled love-specific features.
so,umm, what did i do wrong?
oh, by the way - i cannot get framebuffer to work with love-sources, but it works on 2.6.10-r6 with the same setup. (i mean links -driver fb complains that device /dev/fb0 does not exist on love-sources, but it works with gentoo-dev-2.6.10-r6 (and older dev-2.6 as well)). (i'm using udev btw.)

hmmm, You can try none-genetic nicksched - maybe the temperature is releated to it .
and feel free to change the sources from love hehe maybe exactly just O(1) based line gentoo-dev-sources is better for U .

about fb dev.

Code: Select all

Enterprise ~ # ls -al /dev/fb0
lr-xr-xr-x  1 root root 4 Mar 10 15:47 /dev/fb0 -> fb/0
Linux Enterprise 2.6.11-love2 #2 Wed Mar 9 22:10:55 CET 2005 i686 AMD Athlon(tm) processor AuthenticAMD GNU/Linux

@DrWoland : BIG thanks for the picture - will be much easier now ...

about this nvidiafb - I added it just as test. I tested if this works and don`t broke vesafb.
compilled fine , alone vesafb working and also nvidiafb - so decided to add it to love for testing
I`m going to expand my knowledge abouit it also - now I can

cheers

[fallow]

I did some io scheduler tests also simple .
1) kernel_src copy
2) parallel copy of kernel_src
3) copy of big video file

results here :

Code: Select all

whole "/" -> reiser4
please note that # of repeats = 1 , more objective results will be with avg of 5 tests for ex.

--test1-----------------------------------------------------------------
singe copy of kernel_src
test1 = time cp -r linux-2.6.11 t1
test2 = time rm -fr t1 
io sched = ---deadline---|-----cfq-----|gen.anticipatory|----no-op-----
test1: 
real         0m42.761s     0m45.699s       0m45.259s       0m40.326s
user         0m0.190s      0m0.204s        0m0.217s        0m0.194s
sys          0m8.317s      0m8.434s        0m8.222s        0m8.162s

test2:
real         0m9.299s      0m9.228s        0m9.327s        0m9.048s
user         0m0.033s      0m0.047s        0m0.049s        0m0.036s
sys          0m5.447s      0m5.373s        0m5.386s        0m5.369s
-----------------------------------------------------------------------
--test2----------------------------------------------------------------
parallel copy of kernel_src
test1 = time cp -r linux-2.6.11 t1 & cp -r linux-2.6.11 t2 
test2 = time rm -fr t1 & rm -fr t2 
io sched = ---deadline---|-----cfq-----|gen.anticipatory|----no-op-----
test1: 
real         1m9.455s       4m13.061s      1m7.417s       1m1.550s
user         0m0.247s       0m0.255s       0m0.230s       0m0.226s
sys          0m7.418s       0m8.004s       0m8.031s       0m7.849s

test2:
real         0m34.322s      0m23.615s      0m25.605s      0m31.493s
user         0m0.063s       0m0.043s       0m0.045s       0m0.064s
sys          0m4.162s       0m4.412s       0m4.344s       0m4.258s
-----------------------------------------------------------------------
--test3----------------------------------------------------------------
bif file copy (400mb.avi)
test1 = time cp trainspotting.avi t1.avi
io sched = ---deadline---|-----cfq-----|gen.anticipatory|----no-op-----
test1: 
real         0m34.294s     0m31.714s      0m30.256s       0m31.008s
user         0m0.018s      0m0.014s       0m0.025s        0m0.015s
sys          0m4.565s      0m4.590s       0m4.825s        0m4.415s

file is here : http://vivid.dat.pl/fallow/pub/mini_io_sched_test.txt

Great job fallow!
I'll just stick with deadline as you have recommended. But I'd like to know, how does
deadline perform with multiple heavy I/O operations at once?

Also, has anyone tried benchmarking genetic schedulers versus normal schedulers (both CPU and I/O)?
They feel more responsive, but I'd be interesting to see some numbers. I lack the time to do proper tests.

this is the of of first answers for the quetion

whats the differences between deadline and cfq? why are ppl using deadline instead of cfq?

http://forums.gentoo.org/viewtopic-p-20 ... ml#2053237 this is IMHO and Bollucks said also imporant thing here.

cheers

[DrWoland]

@DrWoland : BIG thanks for the picture - will be much easier now ...

cheers

No problem, mate, let me know if there's any more guinea pig action you need

[DrWoland]

btw.Propably You can get better genetic results when You turn out the "Rounding" option from the genetic nicksched menu

done. Let's reboot!

Does turn out mean turn on or off? I'm confused Gonna turn it on and give it a shot.

[fallow]

btw.Propably You can get better genetic results when You turn out the "Rounding" option from the genetic nicksched menu

done. Let's reboot!

Does turn out mean turn on or off? I'm confused Gonna turn it on and give it a shot.

sorry - my funny english - I mean is better to DEACTIVATE this extra genetic nicksched options.
so - for final resume - is better to have they "off"

No problem, mate, let me know if there's any more guinea pig action you need

hehe - nice, thx & cheers

cheers.

[WladyX]

I seem to have problems with all the 2.6.11 patchsets, nitro, morph, skunk, and also love. I have a P4 and reiser4 could this be the cause?

[fallow]

I seem to have problems with all the 2.6.11 patchsets, nitro, morph, skunk, and also love. I have a P4 and reiser4 could this be the cause?

but what problem? I also have whole '/' on r4.

cheers.

[WladyX]

But do you also have P4? I heard there are problems with SMP, nitro&morph froze on mounting root fs in read/write mode, skunk randomly, love randomly on boot, or at starting X sometimes, i recompiled it without preempt, i'll let you know if anything changes. Thanks for support

[rmh3093]

could be faulty acpi, I have a server that crashes randomly all the time with ACPI enabled, so I just disabled it and it never crashes

[DrWoland]

btw.Propably You can get better genetic results when You turn out the "Rounding" option from the genetic nicksched menu

done. Let's reboot!

Does turn out mean turn on or off? I'm confused Gonna turn it on and give it a shot.

sorry - my funny english - I mean is better to DEACTIVATE this extra genetic nicksched options.
so - for final resume - is better to have they "off"

No problem, mate, let me know if there's any more guinea pig action you need

hehe - nice, thx & cheers

cheers.

Yeah, I figured it out too when my system started freezing at a random point in bootup

Edit: Nvidiafb is also pretty worthless at this point in time, at least with my GeForce 4 4200 64MB. It screws up the screen majorly (still mildly readable) and prevents nvidia-kernel from loading. I turned it back off. I took a picture of it as well and there appears to be a memory addressing issue, I'll post the pic later.

[ashtophet]

btw.Propably You can get better genetic results when You turn out the "Rounding" option from the genetic nicksched menu

done. Let's reboot!

Does turn out mean turn on or off? I'm confused Gonna turn it on and give it a shot.

well, my english is also a funny thing... so i understood fallow perfectly...


cheers

[WladyX]

could be faulty acpi, I have a server that crashes randomly all the time with ACPI enabled, so I just disabled it and it never crashes

OK, i'll try that too, btw disabling preempt gave me a shock: it booted fine, i loged on, entered mc and when i closed mc someting freeky happend, a text with something like call traceback started scrooling on the tty and it was very fast and it didn't stop, so i had to hard reboot, after that, the kernel wouldn't boot anymore, back on gentoo-dev-sources with reiser4 patch for now, but i'll try disabling the acpi in love, anyway here is my love-config, if anyone could tell me if something is wrong i'd greatly apreciate it. Thanks!

/later, tried without acpi, stoped at a boot script, it does it randomly.

[thoffmeyer]

Yet another great release, thanks fallow

[DrWoland]

Yet another great release, thanks fallow

Sounds like it's time for a sig update, bud

[Naib]

Well thought I would upgrade from rc4-love1 (since that was very good) - compiled fine
Got a "oops" at boot, but it did not lock my system up

Well at teh desktop it started up very jerky but got better ???, things do load faster, but visual are a bit slow (using the nvidiafb - could that be resposnisble)

anyway here is the "oops"

Code: Select all

root@Fluid ~ # cat love-oops.txt
 * Setting NIS domainname to nis.Motion.com ...
 * [ ok ] * Setting DNS domainname to Motion ...
 * [ ok ] * Starting portmap ...
 * [ ok ] * Starting famd ...
 * [ ok ] * Starting FireHOL ...
 * [ ok ] * Starting Hardware Abstraction Layer daemon ...
 * [ ok ] * Starting Automounter ...
 * [ ok ][4294714.250000] Oops: 0000 [#1]
 [4294714.250000] PREEMPT
 [4294714.250000] Modules linked in: tulip usblp usbhid uhci_hcd vfat fat ntfs
 rfcomm l2cap bluetooth snd_emu10k1 snd_rawmidi snd_ac97_codec snd_util_mem
 snd_hwdep nvidia
 [4294714.250000] CPU:    0
 [4294714.250000] EIP:    0060:[<c019723a>]    Tainted: P      VLI
 [4294714.250000] EFLAGS: 00010206   (2.6.11-love2)
 [4294714.250000] EIP is at object_depth+0x6/0x11
 [4294714.250000] eax: 00000001   ebx: c03aea40   ecx: 00000001   edx:
 0007039f
 [4294714.250000] esi: f7b0e274   edi: c19d75dc   ebp: c19d7680   esp:
 f7c2fea8
 [4294714.250000] ds: 007b   es: 007b   ss: 0068
 [4294714.250000] Process hald (pid: 10624, threadinfo=f7c2f000 task=f7c89a80)
 [4294714.250000] Stack: c019743a c19d7680 f7c2ff14 f7c2fec4 c03aea40 f7b0e274
 c19d75dc c19d7680
 [4294714.250000]        c0197568 c19d7680 c19d75dc f7b0f000 f7b0f000 fffffff4
 f7c2ff14 00000100
 [4294714.250000]        c0197644 f7b0e274 f7b0f000 ffffffea f7b0e274 f7c2ff14
 c016d09f f7b0e274
 [4294714.250000] Call Trace:
 [4294714.250000]  [<c019743a>] sysfs_get_target_path+0x13/0x7d
 [4294714.250000]  [<c0197568>] sysfs_getlink+0xc4/0x14d
 [4294714.250000]  [<c0197644>] sysfs_follow_link+0x53/0x74
 [4294714.250000]  [<c016d09f>] generic_readlink+0x32/0x8c
 [4294714.250000]  [<c011fada>] current_fs_time+0x4d/0x5b
 [4294714.250000]  [<c0176e6e>] update_atime+0x48/0xbd
 [4294714.250000]  [<c0165251>] sys_readlink+0x96/0x9a
 [4294714.250000]  [<c010304b>] sysenter_past_esp+0x54/0x75
 [4294714.250000] Code: 70 e8 bf b4 1b 00 e9 fb fd ff ff 8d 42 70 e8 d6 b4 1b
 00 e9 06 fe ff ff 8d 42 70 e8 c9 b4 1b 00 e9
 f7 fe ff ff 8b 54 24 04 31 c0 <8b> 52 24 83 c0 01 85 d2 75 f6 c3 55 bd ff ff
 ff ff 57 56 be 01
 [4294714.250000]   * Mounting network filesystems ...
 [ ok ] * Setting up gdm ...
 [ ok ] * Starting local ...
 [ ok ]

 This is Fluid.Motion (Linux i686 2.6.11-love2) 20:46:56

[Robin79]

I will test this one cause latest nitro isnt stable for me i started with love mabey i end with it

[janiskr]

so, first kernel from 2.6.11 family
if i boots and works as expected - cool
if not - not so cool

till now i have all make errors mentioned here
hope ati-drivers will be fine

Edit:

so results after reboot
anything works, pluging/unplugging usb storage works NICE
ati compiles and loads and works NICE

love ROCKS

[Sith_Happens]

I thought this was the second love patchset for the 2.6.11 kernel (2.6.11-love1 "We Are The Borg")? Anyway about the ati-drivers, make sure you turn of the direct render manager under device drivers -> character devices. I've had this compiled into other kernels and used the ati-drivers with no problem, but when I tried loading the ati-driver modules with love, it wouldn't work till I got rid of it. I've been using this love release for a day now and have nothing but good things to say about it.

EDIT: Oh nevermind, you were talking about your first 2.6.11 kernel, sorry. Good choice.

[janiskr]

this was second try of love sources for me, first try was unsuccessful
after that i worked only on vanilla or gentoo-dev-sources
month ago emerged nitro, but never compiled

these sources works for me "as is"

ofcourse i had to straighten my hands a bit by seting HID device timer

onther reason why love2 - it is based on 2.6.11.2 (as i understood from the post)