2.6.13.2-nitro1 "Down with latency!'

Message

1U · Post by 1U » Mon Sep 26, 2005 8:30 am

You have a good point about installing a new kernel, but it's really not that hard. I think if enough fixes and updates were made to this one it would justify a new realease. After all, 13 had nitro1-5 so just a second version couldn't hurt.

Also, the 2.6.13.2 thing is a bit redundant. It's almost like how some developers have "program 5.1.6.19.823.62 build 1592496010234234". You and tiger should agree on a standard for naming the releases, I think kernel.version-nitroX is good enough, no need for extra -rcx or .x and etc.

Zyne · Post by **Zyne** » Mon Sep 26, 2005 8:43 am

tagwar wrote:just installed on my home machine, which has two sata discs. Now I want to do SMART-readings of my discs from time to time to ensure that everything is OK.

I applied that patch, but it doesn't seem to work for me...

Code: Select all

# hddtemp /dev/sd?
/dev/sda: ATA Maxtor 6Y120M0:  known drive, but it doesn't have a temperature sensor.
/dev/sdb: ATA Maxtor 6Y160M0:  known drive, but it doesn't have a temperature sensor.
/dev/sdc: ATA WDC WD740GD-00FL:  known drive, but it doesn't have a temperature sensor.

Could it be that I missed something in the kernel config? I just applied the patch, did a make clean, and recompiled the kernel... I assumed that was enough...

as for the logo problem, that one is fixed perfectly by manually editing the files mentioned in a previous thread!

EDIT: should be solved now... apparently the file I downloaded was 0 bytes, so the patch really didn't do anything at all...
more info here: http://forums.gentoo.org/viewtopic-t-31 ... ature.html

EDIT2: :p

Code: Select all

# hddtemp /dev/sd?
/dev/sda: Maxtor 6Y120M0: 37Â°C
/dev/sdb: Maxtor 6Y160M0: 39Â°C
/dev/sdc: WDC WD740GD-00FLA2:  36Â°C

miedo95 · Post by **miedo95** » Mon Sep 26, 2005 2:37 pm

|cub| wrote:

miedo95 wrote:...

i was following instrucions and get

Code: Select all

  CC      drivers/video/logo/logo.o
  LOGO  drivers/video/logo/logo_linux_mono.c
  CC      drivers/video/logo/logo_linux_mono.o
  LOGO  drivers/video/logo/logo_linux_vga16.c
  CC      drivers/video/logo/logo_linux_vga16.o
  LOGO  drivers/video/logo/logo_linux_clut224.c
  CC      drivers/video/logo/logo_linux_clut224.o
  LOGO  drivers/video/logo/logo_nitro_clut224.c
drivers/video/logo/logo_nitro_clut224.ppm: Binary PNM is not supported
Use pnmnoraw(1) to convert it to ASCII PNM
make[3]: *** [drivers/video/logo/logo_nitro_clut224.c] Error 1
make[2]: *** [drivers/video/logo] Error 2
make[1]: *** [drivers/video] Error 2
make: *** [drivers] Error 2

can't jump this step

backup
mv logo_nitro_clut224.ppm logo_nitro_clut224.ppm_old

write
pnmnoraw logo_nitro_clut224.ppm_old > logo_nitro_clut224.ppm

a13x · Post by **a13x** » Mon Sep 26, 2005 3:21 pm

The logo is broken in this release. Just disable it and it will work. Or ... wait for the next relase.

Anthaus · Post by **Anthaus** » Mon Sep 26, 2005 3:37 pm

1U wrote:Also, the 2.6.13.2 thing is a bit redundant. It's almost like how some developers have "program 5.1.6.19.823.62 build 1592496010234234". You and tiger should agree on a standard for naming the releases, I think kernel.version-nitroX is good enough, no need for extra -rcx or .x and etc.

Glad to hear I'm NOT the only one who thinks this way.

satanskin · Post by **satanskin** » Mon Sep 26, 2005 4:27 pm

Wow. I've noticed this release is a bit sluggish. Especially when switching between open windows. I don't know if this is the problem with the swapping you are talking about, but it is rather annoying. I still have half a gig of ram free and a 64-bit processor so it shouldn't be hardware limitations.

Robin79 · Post by **Robin79** » Mon Sep 26, 2005 5:33 pm

hmm i tried to compile and gets this anyone?

Code: Select all

CC      init/main.o
init/main.c: I funktion `setup_per_cpu_areas':
init/main.c:329: error: `__per_cpu_end' undeclared (first use in this function)
init/main.c:329: error: (Each undeclared identifier is reported only once
init/main.c:329: error: for each function it appears in.)
init/main.c:329: error: `__per_cpu_start' undeclared (first use in this function)
make[1]: *** [init/main.o] Fel 1
make: *** [init] Fel 2

jurrie · Post by **jurrie** » Mon Sep 26, 2005 6:00 pm

1U wrote:Also, the 2.6.13.2 thing is a bit redundant. It's almost like how some developers have "program 5.1.6.19.823.62 build 1592496010234234". You and tiger should agree on a standard for naming the releases, I think kernel.version-nitroX is good enough, no need for extra -rcx or .x and etc.

well, since this nitro release is the first one based on kernel 2.6.13.2, it's kind of obvious it is named 2.6.13.2-nitro1, no?

Robin79 · Post by **Robin79** » Mon Sep 26, 2005 6:20 pm

dam Juyrrie

i tought you had a answer to my problem so i fast looked back to this thread ;;)

tagwar · Post by **tagwar** » Mon Sep 26, 2005 7:48 pm

sorry, if you're already running this kernel, you probably already noticed, so this is for all those who're wondering if it is worth the trouble to install this wonderful thing...

I just got back to my laptop (1.3ghz pentium m), started up firefox, checked my mail, all the "normal" stuff one does... then I did a "ps ax" cause I wondered about the pid of a process I just started... well, what I saw realy interesting... emerge was running and compiling the new firefox... I mean... normaly my system got quite a bit unresponsible when doing big compiles.. now I didn't even notice...

funny... and yes, I'm absolutely sure that it is still the same laptop...

Legoguy · Post by **Legoguy** » Mon Sep 26, 2005 10:57 pm

satanskin wrote:Wow. I've noticed this release is a bit sluggish. Especially when switching between open windows. I don't know if this is the problem with the swapping you are talking about, but it is rather annoying. I still have half a gig of ram free and a 64-bit processor so it shouldn't be hardware limitations.

I've noticed this too, 3.2GHz P4 (Prescott), NVIDIA GeForce FX 5900 128mb, 1gb mem -- it wasn't like this on 2.6.13.1-nitro1 :-/

Hauser · Post by **Hauser** » Tue Sep 27, 2005 12:27 am

iphitus wrote:
Fleta wrote:
iphitus wrote:
Fleta wrote:ipw2200 failed to compile:...
I released patches to fix this on the archck webpage as the same problem occurs on archck. The patches apply to the module not the kernel.

http://iphitus.loudas.com/archck.php
How do I apply the patch?
I have no idea how to do it the gentoo way, but if you wish to do it manually...

download the driver's source (http://unc.dl.sourceforge.net/sourcefor ... -1.0.6.tgz)
extract it. cd into the driver's source directory,
patch -p1 < path/to/patch/ipw2200-fix.diff
then 'make', and as root, 'make install'

will patch and install the driver.

Code: Select all

# ebuild /usr/portage/net-wireless/ipw2200/ipw2200-1.0.6.ebuild unpack
# cd /var/portage/ipw2200-1.0.6/work/ipw2200-1.0.6
# patch -p1 < /path/to/ipw2200-fix.diff
# ebuild /usr/portage/net-wireless/ipw2200/ipw2200-1.0.6.ebuild merge

bbkr · Post by **bbkr** » Tue Sep 27, 2005 1:31 am

works fine.

Code: Select all

localhost ~ # uname -a
Linux localhost 2.6.13.2-nitro1 #2 Mon Sep 26 23:40:20 CEST 2005 x86_64 AMD Athlon(tm) 64 Processor 3000+ AuthenticAMD GNU/Linux
localhost ~ # gcc --version
gcc (GCC) 3.4.4 (Gentoo Hardened 3.4.4-r1, ssp-3.4.4-1.0, pie-8.7.8)

thanks for great job!!

alinv · Post by **alinv** » Tue Sep 27, 2005 6:35 am

satanskin wrote:Wow. I've noticed this release is a bit sluggish. Especially when switching between open windows. I don't know if this is the problem with the swapping you are talking about, but it is rather annoying. I still have half a gig of ram free and a 64-bit processor so it shouldn't be hardware limitations.

Same here

I also noticed that my free memory doesn't go below 300Mb, while I have ~1Gb in swap. I never had more than 2-300Mb in swap (for 1Gb of RAM).

massimo · Post by **massimo** » Tue Sep 27, 2005 6:43 am

I upgraded my kernel on my home desktop pc (Athlon XP 3200, 1GB, GF6600) from 2.6.12-nitro5 to this one as well as at my office's desktop (Dell Optiplex GX-280 - P4 3GHz ht and smp enabled, 512MB, intel onboardgfx). At home I've one GB of RAM and at my office "only" 512 MB of RAM, which since I've upgraded to this kernel was leading to out of memory errors at the desktop pc at my office -> applications are killed automatically (randomly?) too free up some mem. I'm running a few memory intensive Java applications, so the 512MB RAM are used up very fast. There are just a few architectural and network .config differences, so I really didn't try to change any settings there. This happens although there is enough swapable memory (total ~ 800MB). So at my office I'm back on 2.6.12-nitro5. I'm just curious if anybody else did experience something like that. If anybody knows what to do to resolve this issue with 2.6.13.2-nitro1 I'd appreciate your suggestions.

bollucks · Post by **bollucks** » Tue Sep 27, 2005 7:07 am

Disable swap prefetch. It was clearly marked as experimental and was under heavy development. The version in this nitro kernel is known broken.

massimo · Post by **massimo** » Tue Sep 27, 2005 7:14 am

bollucks wrote:Disable swap prefetch. It was clearly marked as experimental and was under heavy development. The version in this nitro kernel is known broken.

Thank you, I didn't know that.

alinv · Post by **alinv** » Tue Sep 27, 2005 7:37 am

bollucks wrote:Disable swap prefetch. It was clearly marked as experimental and was under heavy development. The version in this nitro kernel is known broken.

I did that when I switched from 2.6.13.1 to 2.6.13.2, but I didn't update my grub.conf. Stupid me

Thanks!

XioXouS · Post by **XioXouS** » Tue Sep 27, 2005 7:52 am

well i was gonna wait for nitro2, but i had so homework to put off, so seemed like the perfect time to upgrade kernels

aside from one issue, looking good so far

i applied the logo fix and libata patch mentioned earlier, the ck5 patch to see if there was a difference at all with swap prefetching - think i'm gonna turn it back off completely though, just doesn't seem to be much need, and i also got l7-filters working (2.0.0_beta)

vmware, fglrx, emu10k1 (external alsa), all work normally so far, although fglrx seemed a little slow just in comparison to previous glxgear runs, but the important stuff like ut run fine

alright problem is this, I try to put my laptop to sleep, which has worked great for over a year now, but still one of the first things I check with a new kernel, and I get this nasty message. It's got LVM2 with several Reiser4 partitions, one of which seem to be the failure(s). Can anyone enlighten me at all about these messages? Thanks.

Code: Select all

...
ACPI: PCI interrupt for device 0000:00:1f.5 disabled
ACPI: PCI interrupt for device 0000:02:08.0 disabled
Stopping tasks failed.
Tasks that refused to be refrigerated and haven't since exited:
 - ktxnmgrd:hda3:w (#758) signalled and todo list empty.
 - ent:hda3. (#759) signalled and todo list empty.
 - ktxnmgrd:dm-0:t (#1602) signalled and todo list empty.
 - ent:dm-0. (#1603) signalled and todo list empty.
 - ktxnmgrd:dm-1:w (#1617) signalled and todo list empty.
 - ent:dm-1. (#1618) signalled and todo list empty.
 - ktxnmgrd:dm-2:w (#1619) signalled and todo list empty.
 - ent:dm-2. (#1620) signalled and todo list empty.
 - ktxnmgrd:dm-3:w (#1621) signalled and todo list empty.
 - ent:dm-3. (#1622) signalled and todo list empty.
 - ktxnmgrd:dm-4:w (#1623) signalled and todo list empty.
 - ent:dm-4. (#1624) signalled and todo list empty.
agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
...

aenertia · Post by **aenertia** » Tue Sep 27, 2005 10:10 am

Hi there I have been using .12-nitro5 in both my Opteron Server/Workstation and in A machine upstairs which is p4 based. I have never been able to emerge pcmcia-cs (the cardmgr tools) with nitro5.

I am just about to image the p4 machine to send to my partner who has a p4 laptop. And I need to get those pesky pcmcia-cs tools workin with nitro-5. Can someone let me know if this new e-build works. And anyone who has been using the cardbus succesfully with nitro5.

Kind regards

Joel W (Ænertia)

aenertia · Post by **aenertia** » Tue Sep 27, 2005 10:14 am

I also noticed that my free memory doesn't go below 300Mb, while I have ~1Gb in swap. I never had more than 2-300Mb in swap (for 1Gb of RAM).[/quote]

You should never have more than 5% of memory free if the kernel is doing a good job. Basically all good kernels will ramp up to 100% useage after a while, then drop down to around 97%. The key is looking at how much of that 97% is cached... it should be around 30% odd. This is good.

More free mem != a good thing, and is a sign that somthing is not working right.

XioXouS · Post by **XioXouS** » Tue Sep 27, 2005 7:36 pm

XioXouS wrote:alright problem is this, I try to put my laptop to sleep, which has worked great for over a year now, but still one of the first things I check with a new kernel, and I get this nasty message. It's got LVM2 with several Reiser4 partitions, one of which seem to be the failure(s). Can anyone enlighten me at all about these messages? Thanks.

Code: Select all

...
ACPI: PCI interrupt for device 0000:00:1f.5 disabled
ACPI: PCI interrupt for device 0000:02:08.0 disabled
Stopping tasks failed.
Tasks that refused to be refrigerated and haven't since exited:
 - ktxnmgrd:hda3:w (#758) signalled and todo list empty.
 - ent:hda3. (#759) signalled and todo list empty.
 - ktxnmgrd:dm-0:t (#1602) signalled and todo list empty.
 - ent:dm-0. (#1603) signalled and todo list empty.
 - ktxnmgrd:dm-1:w (#1617) signalled and todo list empty.
 - ent:dm-1. (#1618) signalled and todo list empty.
 - ktxnmgrd:dm-2:w (#1619) signalled and todo list empty.
 - ent:dm-2. (#1620) signalled and todo list empty.
 - ktxnmgrd:dm-3:w (#1621) signalled and todo list empty.
 - ent:dm-3. (#1622) signalled and todo list empty.
 - ktxnmgrd:dm-4:w (#1623) signalled and todo list empty.
 - ent:dm-4. (#1624) signalled and todo list empty.
agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
...

I've since confirmed this on three machines (some without lvm), with various simplified kernel configs for both mm-sources (just for reiser4) and nitro-sources - both were the 2.6.13 variant. If the machine is put to sleep via

Code: Select all

echo mem > /sys/power/state

those reiser4 parts running (ktxnmgrd*, ent*) get angry, display the above messages, and then return to the system hogging all the CPU resources aka 100% utilization.
Help? Should I be submiting this to someone else as a bug? Thanks.

Fleta · Post by **Fleta** » Wed Sep 28, 2005 12:23 am

Hauser wrote:
iphitus wrote:
Fleta wrote:
iphitus wrote:
Fleta wrote:ipw2200 failed to compile:...
I released patches to fix this on the archck webpage as the same problem occurs on archck. The patches apply to the module not the kernel.

http://iphitus.loudas.com/archck.php
How do I apply the patch?
I have no idea how to do it the gentoo way, but if you wish to do it manually...

download the driver's source (http://unc.dl.sourceforge.net/sourcefor ... -1.0.6.tgz)
extract it. cd into the driver's source directory,
patch -p1 < path/to/patch/ipw2200-fix.diff
then 'make', and as root, 'make install'

will patch and install the driver.
Code: Select all
# ebuild /usr/portage/net-wireless/ipw2200/ipw2200-1.0.6.ebuild unpack
# cd /var/portage/ipw2200-1.0.6/work/ipw2200-1.0.6
# patch -p1 < /path/to/ipw2200-fix.diff
# ebuild /usr/portage/net-wireless/ipw2200/ipw2200-1.0.6.ebuild merge

Thanks a lot, now it works!

Cinder6 · Post by **Cinder6** » Wed Sep 28, 2005 1:21 am

alinv wrote:
bollucks wrote:Disable swap prefetch. It was clearly marked as experimental and was under heavy development. The version in this nitro kernel is known broken.
I did that when I switched from 2.6.13.1 to 2.6.13.2, but I didn't update my grub.conf. Stupid me

Thanks!

Wait...you have to enable swap prefetching via GRUB?

Gnufsh · Post by **Gnufsh** » Wed Sep 28, 2005 3:37 am

Cinder6 wrote:
alinv wrote:
bollucks wrote:Disable swap prefetch. It was clearly marked as experimental and was under heavy development. The version in this nitro kernel is known broken.
I did that when I switched from 2.6.13.1 to 2.6.13.2, but I didn't update my grub.conf. Stupid me

Thanks!
Wait...you have to enable swap prefetching via GRUB?

I think he means he didn't update it to point to the new kernel.

2.6.13.2-nitro1 "Down with latency!'

Re: sata and smart

Re: Problem resolved!!!

daaaamnn... this is soooo cool..

PCMCIA-CS Fixed in this version?