Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Zen-Kernel Discussion/Support Thread - Part 4
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3 ... 8, 9, 10 ... 21, 22, 23  Next  
Reply to topic    Gentoo Forums Forum Index Unsupported Software
View previous topic :: View next topic  
Author Message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Sun Sep 26, 2010 11:02 pm    Post subject: Reply with quote

use this in addition to the other fixes:

[PATCH] vmscan: raise the bar to PAGEOUT_IO_SYNC stalls

this should really improve things when dealing especially with slow devices and write-stalls

edit:

it should already be applied

that's why it looked so familiar :roll:
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
ponciarello
Apprentice
Apprentice


Joined: 22 Jul 2008
Posts: 223
Location: beach of slack

PostPosted: Mon Sep 27, 2010 12:50 pm    Post subject: Reply with quote

and, again, thanks steven for the superfast update. ;)
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Mon Sep 27, 2010 2:09 pm    Post subject: Reply with quote

ponciarello wrote:
and, again, thanks steven for the superfast update. ;)


yeah, that was fast

thanks ! :)


does tuxonice work for everyone ?

I'm currently trying to get it to work with 2.6.35-zen3 but I got an error-message (from kernel) with worker-something after switching to tuxoniceui_text and starting to copy or freezing stuff
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
ebrsc
n00b
n00b


Joined: 24 Jun 2008
Posts: 31

PostPosted: Mon Sep 27, 2010 3:15 pm    Post subject: Reply with quote

maybe I missed something and very sorry if it's the case but I'm at 1b4f70ca7103 and
Code:
eb64@drama:~/pkg/kernel26zen/src$grep Kolivas include/linux/sched.h
        printk(KERN_INFO"BFS CPU scheduler v0.323 by Con Kolivas.\n");

ck provides incremental patches so that should not be too hard to get latest BFS but is there any reason why I shouldn't do it?
thanks.
Back to top
View user's profile Send private message
ssteinberg
Apprentice
Apprentice


Joined: 09 Jul 2010
Posts: 206
Location: Israel

PostPosted: Mon Sep 27, 2010 3:50 pm    Post subject: Reply with quote

ebrsc wrote:
maybe I missed something and very sorry if it's the case but I'm at 1b4f70ca7103 and
Code:
eb64@drama:~/pkg/kernel26zen/src$grep Kolivas include/linux/sched.h
        printk(KERN_INFO"BFS CPU scheduler v0.323 by Con Kolivas.\n");

ck provides incremental patches so that should not be too hard to get latest BFS but is there any reason why I shouldn't do it?
thanks.


+1 noticed that as well.
TOI does seem to sometimes fail and just keeps writing caches or trying to freeze system with no errors forever.
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Mon Sep 27, 2010 5:43 pm    Post subject: Reply with quote

if the kernel after several minutes or hours starts to stutter - the GUI and mouse becoming jerky for several minutes and the console not responding to input then
switching the clocksource from hpet to tsc or acpi_pm might fix it for you:

put
Code:
clocksource=tsc


or

Code:
clocksource=acpi_pm


to the line where your kernel is called at grub.conf / menu.lst or lilo.conf

that seems to have fixed it for me under 2.6.35+ and 2.6.36-rc*

cheers !
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
Ormaaj
Guru
Guru


Joined: 28 Jan 2008
Posts: 319

PostPosted: Tue Sep 28, 2010 12:55 am    Post subject: Reply with quote

kernelOfTruth wrote:
if the kernel after several minutes or hours starts to stutter - the GUI and mouse becoming jerky for several minutes and the console not responding to input then
switching the clocksource from hpet to tsc or acpi_pm might fix it for you:

put
Code:
clocksource=tsc


or

Code:
clocksource=acpi_pm


to the line where your kernel is called at grub.conf / menu.lst or lilo.conf

that seems to have fixed it for me under 2.6.35+ and 2.6.36-rc*

cheers !
Is one of these preferable under ordinary circumstances? I can't find too much info on these. My box always seems to naturally prefer tsc unless I override it. I think it chooses tsc when dynamic frequency scaling is enabled.
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Tue Sep 28, 2010 12:58 am    Post subject: Reply with quote

Ormaaj wrote:
Is one of these preferable under ordinary circumstances? I can't find too much info on these. My box always seems to naturally prefer tsc unless I override it. I think it chooses tsc when dynamic frequency scaling is enabled.


if in dmesg messages like:

Quote:
Marking TSC unstable due to cpufreq changes


and

Quote:
Clocksource tsc unstable (delta = -291678239 ns)


appear

you should prefer acpi_pm, this should be especially true on laptops
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
bollucks
l33t
l33t


Joined: 27 Oct 2004
Posts: 606

PostPosted: Wed Sep 29, 2010 10:27 am    Post subject: Reply with quote

Ormaaj wrote:
Is one of these preferable under ordinary circumstances? I can't find too much info on these. My box always seems to naturally prefer tsc unless I override it. I think it chooses tsc when dynamic frequency scaling is enabled.

acpi_pm is reliable under all circumstances but is much slower, incurring much greater overhead which is why it's not used by default. However if your other time source is unstable, there is not much you can do but wear the extra overhead.
Back to top
View user's profile Send private message
Ormaaj
Guru
Guru


Joined: 28 Jan 2008
Posts: 319

PostPosted: Wed Sep 29, 2010 3:35 pm    Post subject: Reply with quote

How about the other one? "HPET" sounds "Highly Precise!", which I hear is a good thing to look for in a "Timer" of "Events".

kernelOfTruth wrote:
if in dmesg messages like:

Quote:
Marking TSC unstable due to cpufreq changes



and

Quote:
Clocksource tsc unstable (delta = -291678239 ns)


appear

you should prefer acpi_pm, this should be especially true on laptops


Doesn't look like I'm getting those (even with ondemand cpufreq, linux-PHC patches, intel idle driver, etc etc)

Some relevant looking lines:
Code:
...

[    0.000000] hpet clockevent registered
[    0.000000] Fast TSC calibration using PIT
[    0.001000] Detected 3900.591 MHz processor.
[    0.000003] Calibrating delay loop (skipped), value calculated using timer frequency.. 7801.18 BogoMIPS (lpj=3900591)

...

[    0.164334] APIC calibration not consistent with PM-Timer: 120ms instead of 100ms
[    0.164926] APIC delta adjusted to PM-Timer: 1283345 (1552833)

...

[    0.911857] HPET: 4 timers in total, 0 timers will be used for per-cpu timer
[    0.911888] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0, 0
[    0.911969] hpet0: 4 comparators, 64-bit 14.318180 MHz counter
[    0.917582] Switching to clocksource tsc


No results for "unstable"


Last edited by Ormaaj on Wed Sep 29, 2010 11:02 pm; edited 1 time in total
Back to top
View user's profile Send private message
ppurka
Advocate
Advocate


Joined: 26 Dec 2004
Posts: 3256

PostPosted: Wed Sep 29, 2010 5:09 pm    Post subject: Reply with quote

Looks like it prefers hpet here:
Code:
~> dmesg|grep -C2 clock
[    0.000000] Console: colour VGA+ 80x25
[    0.000000] console [tty1] enabled
[    0.000000] hpet clockevent registered
[    0.000000] Fast TSC calibration using PIT
[    0.000000] Detected 1795.617 MHz processor.
--
[    0.076687] Booting Node   0, Processors  #1 Ok.
[    0.173332] TSC synchronization [CPU#0 -> CPU#1]:
[    0.173332] Measured 965475 cycles TSC warp between CPUs, turning off TSC clock.
[    0.173332] Marking TSC unstable due to check_tsc_sync_source failed
[    0.173346] Brought up 2 CPUs
--
[    0.250345] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0
[    0.250351] hpet0: 3 comparators, 64-bit 14.318180 MHz counter
[    0.253358] Switching to clocksource hpet
[    0.253499] pnp: PnP ACPI init
[    0.253507] ACPI: bus type pnp registered

_________________
emerge --quiet redefined | E17 vids: I, II | Now using kde5 | e is unstable :-/
Back to top
View user's profile Send private message
Ormaaj
Guru
Guru


Joined: 28 Jan 2008
Posts: 319

PostPosted: Wed Sep 29, 2010 11:18 pm    Post subject: Reply with quote

ppurka wrote:
Looks like it prefers hpet here:
Code:
~> dmesg|grep -C2 clock
[    0.000000] Console: colour VGA+ 80x25
[    0.000000] console [tty1] enabled
[    0.000000] hpet clockevent registered
[    0.000000] Fast TSC calibration using PIT
[    0.000000] Detected 1795.617 MHz processor.
--
[    0.076687] Booting Node   0, Processors  #1 Ok.
[    0.173332] TSC synchronization [CPU#0 -> CPU#1]:
[    0.173332] Measured 965475 cycles TSC warp between CPUs, turning off TSC clock.
[    0.173332] Marking TSC unstable due to check_tsc_sync_source failed
[    0.173346] Brought up 2 CPUs
--
[    0.250345] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0
[    0.250351] hpet0: 3 comparators, 64-bit 14.318180 MHz counter
[    0.253358] Switching to clocksource hpet
[    0.253499] pnp: PnP ACPI init
[    0.253507] ACPI: bus type pnp registered
Interesting. What kind of hardware?

Currently used should be in /sys. Try:
Code:
for i in $(find /sys -type f -name '*clocksource'); do echo -e "$(basename ${i}):\n$(cat ${i})"; done
Back to top
View user's profile Send private message
ppurka
Advocate
Advocate


Joined: 26 Dec 2004
Posts: 3256

PostPosted: Wed Sep 29, 2010 11:55 pm    Post subject: Reply with quote

This is a laptop (T61).
Code:
~> for i in $(find /sys -type f -name '*clocksource'); do echo -e "$(basename ${i}):\n$(cat ${i})"; done
current_clocksource:
hpet
available_clocksource:
hpet acpi_pm
Oh, I am not using zen-sources. I was using it until about two days ago. In either case, my system has always used hpet automatically.
_________________
emerge --quiet redefined | E17 vids: I, II | Now using kde5 | e is unstable :-/
Back to top
View user's profile Send private message
Anon-E-moose
Advocate
Advocate


Joined: 23 May 2008
Posts: 4204
Location: Dallas area

PostPosted: Thu Sep 30, 2010 12:09 am    Post subject: Reply with quote

ppurka wrote:
This is a laptop (T61).
Code:
~> for i in $(find /sys -type f -name '*clocksource'); do echo -e "$(basename ${i}):\n$(cat ${i})"; done
current_clocksource:
hpet
available_clocksource:
hpet acpi_pm
Oh, I am not using zen-sources. I was using it until about two days ago. In either case, my system has always used hpet automatically.


Same results, running 2.5.35-zen2
_________________
Asus m5a99fx, FX 8320 - nouveau, oss4, rx550 for qemu passthrough
Acer laptop E5-575, i3-7100u - i965, alsa
---both---
5.0.13 zen kernel, profile 17.1 (no-pie & modified) amd64-no-multilib
gcc 8.2.0, eudev, openrc, openbox, palemoon
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Mon Oct 04, 2010 1:43 pm    Post subject: Reply with quote

important fix for the XFS users among us:

[patch] xfs: properly account for reclaimed inodes

the 3rd hunk failed for me and I can't figure out how to fix (a lot seemingly has changed from the underlying kernel used for this patch and 2.6.35.*) - maybe you zen guys can figure out

I don't even know if it's needed for 2.6.35 ...

edit:

the fix could be:
http://marc.info/?l=linux-kernel&m=128618776007499&w=2
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
bollucks
l33t
l33t


Joined: 27 Oct 2004
Posts: 606

PostPosted: Mon Oct 04, 2010 11:17 pm    Post subject: Reply with quote

You might find this blogpost interesting about possible BFS evolution:

http://ck-hack.blogspot.com/2010/10/interactive-at-any-load.html
Back to top
View user's profile Send private message
Colt45
Tux's lil' helper
Tux's lil' helper


Joined: 05 Sep 2007
Posts: 102
Location: Central Washington

PostPosted: Sun Oct 10, 2010 12:46 am    Post subject: Reply with quote

I pulled current zen-stable master and built a new kernel on my XPS M1330. Low and behold, it actually boots this time. :D
But, no network. Ive tried everything I could think of; this is what I get.
Code:
 
09:00.0 Ethernet controller: Broadcom Corporation NetLink BCM5906M Fast Ethernet PCI Express (rev 02)

        Subsystem: Dell XPS M1330

        Flags: fast devsel, IRQ 17

        Memory at fe5f0000 (64-bit, non-prefetchable) [size=64K]

        Capabilities: [48] Power Management version 3

        Capabilities: [50] Vital Product Data

        Capabilities: [58] Vendor Specific Information: Len=78 <?>

        Capabilities: [e8] MSI: Enable- Count=1/1 Maskable- 64bit+

        Capabilities: [d0] Express Endpoint, MSI 00

        Capabilities: [100] Advanced Error Reporting

        Capabilities: [13c] Virtual Channel

        Kernel modules: tg3

[  196.396429] tg3.c:v3.110 (April 9, 2010)

[  196.396455] tg3 0000:09:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17

[  196.396479] tg3 0000:09:00.0: setting latency timer to 64

[  196.515870] tg3 0000:09:00.0: phy probe failed, err -19

[  196.592731] tg3 0000:09:00.0: Problem fetching invariants of chip, aborting

[  196.592771] tg3 0000:09:00.0: PCI INT A disabled

_________________
"Hopps" - AMD Athlon X2 4450B, 4GB, 128GB Samsung SSD, (Gentoo Hardened amd64; gcc-4.9.3; 4.4.8-hardened-r1)
"Dash" - 2x Intel Xeon 5148, 12GB FB-DIMM, 3x Samsung 1TB (mdadm RAID 5), AMD HD6950, (Gentoo amd64-multilib; gcc-5.3.0; 4.4.0-pf4)
Back to top
View user's profile Send private message
Muso
l33t
l33t


Joined: 22 Oct 2002
Posts: 985
Location: The Holy city of Honolulu

PostPosted: Sun Oct 10, 2010 1:50 am    Post subject: Reply with quote

I switched a laptop and desktop over to Zen this past week. So far so good.

Thanks to those who put this kernel together.
_________________
Time is a great teacher, but unfortunately it kills all its pupils.
Back to top
View user's profile Send private message
Colt45
Tux's lil' helper
Tux's lil' helper


Joined: 05 Sep 2007
Posts: 102
Location: Central Washington

PostPosted: Sun Oct 10, 2010 8:58 am    Post subject: Reply with quote

Turns out, there is more to it than some bad driver for my ethernet. I figured out my wireless was not working either.
Found out the Sonics Silicon Backplane driver was complaining about unknown SPROM version. Some Google-fu provided me with a way to find the sprom version it was getting, which was interesting.
http://pastebin.com/5pHWDKun
You can see when compared to 33-zen3 something looks a bit off. Straight 1's?
Heres boot log for .33-zen3
http://pastebin.com/TztD8CxA
And .35-zen3+
http://pastebin.com/4mh4KRUk
_________________
"Hopps" - AMD Athlon X2 4450B, 4GB, 128GB Samsung SSD, (Gentoo Hardened amd64; gcc-4.9.3; 4.4.8-hardened-r1)
"Dash" - 2x Intel Xeon 5148, 12GB FB-DIMM, 3x Samsung 1TB (mdadm RAID 5), AMD HD6950, (Gentoo amd64-multilib; gcc-5.3.0; 4.4.0-pf4)
Back to top
View user's profile Send private message
AaronPPC
Guru
Guru


Joined: 29 May 2005
Posts: 522
Location: Tucson, AZ

PostPosted: Sun Oct 10, 2010 11:02 am    Post subject: Reply with quote

bollucks wrote:
You might find this blogpost interesting about possible BFS evolution:

It is already in -zen3. I am absolutely shredding a world update at -j16 with no lag at all. And there is this:
Code:
load average: 4.78, 2.69, 1.39

_________________
--Aaron
Back to top
View user's profile Send private message
darklegion
Guru
Guru


Joined: 14 Nov 2004
Posts: 468

PostPosted: Mon Oct 11, 2010 5:51 am    Post subject: Reply with quote

Have all of the 2.6.36 i/o interactivity patches been backported to 2.6.35-zen3 or should I stick with vanilla 2.6.36-rcX for now? I want to try this new version of BFS, but don't want to lose the i/o fixes.
Back to top
View user's profile Send private message
ebrsc
n00b
n00b


Joined: 24 Jun 2008
Posts: 31

PostPosted: Mon Oct 11, 2010 8:45 am    Post subject: Reply with quote

darklegion wrote:
Have all of the 2.6.36 i/o interactivity patches been backported to 2.6.35-zen3 or should I stick with vanilla 2.6.36-rcX for now? I want to try this new version of BFS, but don't want to lose the i/o fixes.

http://ck-hack.blogspot.com/2010/10/2636-rc7-with-ck1-and-bfs-357.html
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Mon Oct 11, 2010 4:50 pm    Post subject: Reply with quote

while we're at it:

the following patch at least doubled or tripled my performance of backup-jobs with ext4 and xfs:

http://marc.info/?l=dm-devel&m=128671249318393&w=2

this applies to 2.6.36*
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
Muso
l33t
l33t


Joined: 22 Oct 2002
Posts: 985
Location: The Holy city of Honolulu

PostPosted: Tue Oct 12, 2010 11:22 pm    Post subject: Reply with quote

AaronPPC wrote:
bollucks wrote:
You might find this blogpost interesting about possible BFS evolution:

It is already in -zen3. I am absolutely shredding a world update at -j16 with no lag at all. And there is this:
Code:
load average: 4.78, 2.69, 1.39


Any indication on when Zen-3 will be in portage?
_________________
Time is a great teacher, but unfortunately it kills all its pupils.
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Tue Oct 12, 2010 11:27 pm    Post subject: Reply with quote

Re: [PATCH try 5] CFS: Add hierarchical tree-based penalty.

http://marc.info/?l=linux-kernel&m=128687654414646&w=2
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Unsupported Software All times are GMT
Goto page Previous  1, 2, 3 ... 8, 9, 10 ... 21, 22, 23  Next
Page 9 of 23

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum