| View previous topic :: View next topic |
| Author |
Message |
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
 Posted: Tue Dec 13, 2005 3:38 am Post subject: [realtime-sources] 2.6.14 (-rt22 based + ck6) [UPDATED] |
 |
|
Use it for realtime tasks like audio/video, for solve JACK issues and Gaming
1. NO reiser4
2. Tested only on amd64 and x86
3. Patches from Gentoo genpatches
4. Patches from Debian SID kernel
5. Patches from Fedora kernel
6. Patches from CK patchset (except staircase scheduler)
7. NO git unstable patches
8. NO staircase scheduler
The ebuild's home is:
https://fantoo.ru/svn/sys-kernel/realtime-sources/
NOTE: use only 2.6.14 kernel! the ebuild for 2.6.15 isn't usable at all (bugs in RT patchset)
You may try setup FULL overlay using command:
mkdir /usr/local/fantoo
svn co https://fantoo.ru/svn/ /usr/local/fantoo
The russian documentation is here: http://wiki.fantoo.ru/index.php/Realtime_Sources_HOWTO
-= Good Luck with REALTIME! =-
P.S. My probjems with JACK has resolved with CONFIG_HZ=250
>>> Unpacking linux-2.6.14.tar.bz2 to /var/tmp/portage/realtime-sources-2.6.14_p22-r4/work
* genpatches-2.6.14-5.base.tar.bz2 unpacked
* genpatches-2.6.14-5.extras.tar.bz2 unpacked
* Excluding Patch #1004_2_thread-sendqueue-race.patch ...
* Excluding Patch #1021_3_alsa-rtc-soft-lockup.patch ...
* Excluding Patch #4500_fbsplash-0.9.2-r5.patch ...
* Excluding Patch #4305_dm-bbr.patch ...
* Excluding Patch #4900_speakup-20051028.patch ...
* Excluding Patch #*alpha-* ...
* Excluding Patch #*sparc-* ...
* Excluding Patch #*ppc-* ...
* Excluding Patch #*ppc64-* ...
* Applying 1001_patch-2.6.14-rt22.patch (-p0+) ...
* Applying 1002_1g_lowmem1_i386.patch (-p0+) ...
* Applying 1003_defaultcfq.patch (-p0+) ...
* Applying 1004_pdflush-tweaks.patch (-p0+) ...
* Applying 1005_jiffies-max.patch (-p0+) ...
* Applying 1006_kconfig-hz-integer.patch (-p0+) ...
* Applying 1010_mm-swap_prefetch-18.patch (-p0+) ...
* Applying 1011_vm-mapped.patch (-p0+) ...
* Applying 1012_vm-lots_watermark.patch (-p0+) ...
* Applying 1013_vm-background_scan-1.patch (-p0+) ...
* Applying 1014_mm-kswapd_inherit_prio.patch (-p0+) ...
* Applying 1015_mm-prio_dependant_scan.patch (-p0+) ...
* Applying 1020_fix-swap_prefetch-lock.patch (-p0+) ...
* Applying 1021_fix-vgettimeofday-amd64.patch (-p0+) ...
* Applying 1022_alsa-rtc-soft-lockup.patch (-p0+) ...
* Applying 1023_fbsplash-0.9.2-r5.patch (-p0+) ...
* Applying amd64-int3-fix.patch (-p0+) ...
* Applying buslogic-pci-id-table.patch (-p0+) ...
* Applying fbdev-radeon-noaccel.patch (-p0+) ...
* Applying linux-2.6-debug-taint-proprietary-helpers.patch (-p0+) ...
* Applying linux-2.6-input-kill-stupid-messages.patch (-p0+) ...
* Applying linux-2.6-max-symlinks.patch (-p0+) ...
* Applying linux-2.6-net-sundance-ip100A.patch (-p0+) ...
* Applying linux-2.6-obsolete-idescsi-warning.patch (-p0+) ...
* Applying linux-2.6-obsolete-oss-warning.patch (-p0+) ...
* Applying linux-2.6-procfs-i_nlink-miscalculate.patch (-p0+) ...
* Applying linux-2.6-sata-enable-atapi-by-default.patch (-p0+) ...
* Applying linux-2.6-swsusp-nofreeze.patch (-p0+) ...
* Applying linux-2.6-valid-ether-addr.patch (-p0+) ...
* Applying linux-2.6-vm-oomkiller-debugging.patch (-p0+) ...
* Applying linux-2.6-vm-silence-atomic-alloc-failures.patch (-p0+) ...
* Applying linux-2.6-x86-vga-vidfail.patch (-p0+) ...
* Applying linux-2.6-x86_64-silence-up-apic-errors.patch (-p0+) ...
* Applying linux-2.6.14-kauditd-suspend.patch (-p0+) ...
* Applying net-bonding-consolidation-fix.patch (-p0+) ...
* Applying net-ipconntrack-nat-fix.patch (-p0+) ...
* Applying qla2xxx-removed.patch (-p0+) ...
* Applying setkeys-needs-root-1.patch (-p0+) ...
* Applying tty-locking-fixes9.patch (-p0+) ...
* Applying 1000_1_sysctl-unregistration.patch (-p0+) ...
* Applying 1001_2_airo_cs-correct-prototypes.patch (-p0+) ...
* Applying 1002_2_cfq-io-sched-fix.patch (-p0+) ...
* Applying 1005_2_ptrace-self-attach.patch (-p0+) ...
* Applying 1006_2_signal-live-leak.patch (-p0+) ...
* Applying 1007_2_zero-size-datagram.patch (-p0+) ...
* Applying 1008_2_ipvs-connection-leak.patch (-p0+) ...
* Applying 1009_2_prism54-frame-size.patch (-p0+) ...
* Applying 1010_2_tcp-bic-max-increment.patch (-p0+) ...
* Applying 1011_2_usb-interface-modalias.patch (-p0+) ...
* Applying 1012_2_xfs-modular-quota.patch (-p0+) ...
* Applying 1013_3_4GB-memory-intel-dual-core.patch (-p0+) ...
* Applying 1014_3_ctnetlink-protoinfo.patch (-p0+) ...
* Applying 1015_3_ctnetlink-no-icmp-oops.patch (-p0+) ...
* Applying 1016_3_eicon-signature.patch (-p0+) ...
* Applying 1017_3_ipv6-ah-length.patch (-p0+) ...
* Applying 1018_3_file-lease-leak.patch (-p0+) ...
* Applying 1019_3_ipv6-advapi-memory.patch (-p0+) ...
* Applying 1020_3_ipv6-extension-headers.patch (-p0+) ...
* Applying 1022_3_hdlc-netif-carrier.patch (-p0+) ...
* Applying 1023_3_hwmon-it87-fan-speed-limits.patch (-p0+) ...
* Applying 1024_3_hwmon-lm78-vid.patch (-p0+) ...
* Applying 1025_3_hwmon-w83627hf-limits.patch (-p0+) ...
* Applying 1026_3_conntrack-high-ports.patch (-p0+) ...
* Applying 1027_3_conntrack-syn-push.patch (-p0+) ...
* Applying 1028_3_nat-refcount.patch (-p0+) ...
* Applying 1029_3_netfilter-queue-oops.patch (-p0+) ...
* Applying 1030_3_packet-writing-oops-fix.patch (-p0+) ...
* Applying 1032_3_pptp-endianness.patch (-p0+) ...
* Applying 1033_3_pptp-call-id.patch (-p0+) ...
* Applying 1034_3_trace-auto-reap.patch (-p0+) ...
* Applying 1035_3_netfilter-refcount-leak.patch (-p0+) ...
* Applying 1100_setkeys-needs-root.patch (-p0+) ...
* Applying 1101_setkeys-read-ok.patch (-p0+) ...
* Applying 1105_invalidate-inodes-overflow.patch (-p0+) ...
* Applying 1910_kconfig-no-nls.patch (-p0+) ...
* Applying 2500_dvb-budget-ci-linking.patch (-p0+) ...
* Applying 2505_cx88-quiet.patch (-p0+) ...
* Applying 2700_acpi-driver-binding-hp-boot-hang.patch (-p0+) ...
* Applying 2705_ht-slowdown.patch (-p0+) ...
* Applying 2710_ht-dlowdown-2.patch (-p0+) ...
* Applying 2900_microtek-fix.patch (-p0+) ...
* Applying 4101_deprecate-sk98lin.patch (-p0+) ...
* Applying 4300_via-vt6410.patch (-p0+) ...
* Applying 4320_promise-pdc2037x.patch (-p0+) ...
* Applying 4351_megaraid-compatibility.patch (-p0+) ...
* Applying 4505_vesafb-tng-1.0-rc1-r2.patch (-p0+) ...
* Applying 4705_squashfs-2.2.patch (-p0+) ...
CHK include/linux/version.h
UPD include/linux/version.h
>>> Source unpacked.
Last edited by amax on Wed Dec 14, 2005 9:32 am; edited 7 times in total |
|
| Back to top |
|
 |
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 3:44 am Post subject: |
|
|
Updated to the realtime-sources-2.6.14_p22-r4.ebuild
Now it uses fbslpash 
Waiting for feedback... |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 3:48 am Post subject: |
|
|
what should I add else?
* mppe-mppc
* iptables layer7
Help me please.. |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 3:54 am Post subject: |
|
|
| NOTE: it is positioned like STABLE kernel, so no extra-experimental patches has applied, ONLY stabilization and bug fixes. |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 4:06 am Post subject: |
|
|
Addtionaly, you will need:
sys-process/schedutils >= 1.4.0
sys-apps/realtime-lsm >= 0.8.5-r1
sys-libs/pam >= 0.78-r4
(you can find pre-patched for "rtprio" PAM in the overlay at https://fantoo.ru/svn/sys-libs/pam/ , in the official portage PAM w/o realtime support)
edit /etc/security/limits.conf
and append
@audio - rtprio 80
@audio - nice -10
@audio - memlock 250000
also use this settings for the kernel:
=============================================================
[*] Prompt for development and/or incomplete code/drivers
[*] Generic x86 support
[*] HPET Timer Support
[*] Ktimers 64bit scalar representation
[*] High Resolution Timer Support
(10000) High Resolution Timer resolution (nanoseconds)
Preemption Mode (Complete Preemption (Real-Time)) --->
[*] Local APIC support on uniprocessors
[*] IO-APIC support on uniprocessors
[*] enhanced IO-APIC support
High Memory Support (off) --->
[*] 1Gb Low Memory Support
Timer frequency (250 HZ) --->
<*> CFQ I/O scheduler
[ ] Sharing PCI IDE interrupts support
[ ] Paraniod Timekeeping Checks
[ ] Wakeup latency timing (NEW)
[ ] Automatic spinlock/rwlock/mutex/rwsem deadlock detection (NEW)
[*] Enable different security models
<M> Default Linux Capabilities
=============================================================
use "modprobe realtime" with options in /etc/modules.d/realtime
options realtime gid=18
You may contact to me - Rusnet IRC Network (irc.rinet.ru, utf8.rusnet.org.ru) - #fantoo channel |
|
| Back to top |
|
|
Ma3oxuct
Guru
Joined: 18 Apr 2003
Posts: 523
|
| Posted: Tue Dec 13, 2005 5:39 am Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) |
|
|
| amax wrote: |
6. Patches from CK patchset (except staircase scheduler)
|
Was not that the very good and improved staircase? If it was, then which one is in your patchset, and is it better than the new and improved one? |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 6:09 am Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) |
|
|
| Ma3oxuct wrote: | | amax wrote: |
6. Patches from CK patchset (except staircase scheduler)
|
Was not that the very good and improved staircase? If it was, then which one is in your patchset, and is it better than the new and improved one? |
The staircase is still incompatible with RT patchset, talking to the Con Kolivas for getting new one.. The staircase is great. I expect to get it sometime (after Con will back from vacation) |
|
| Back to top |
|
|
Ma3oxuct
Guru
Joined: 18 Apr 2003
Posts: 523
|
| Posted: Tue Dec 13, 2005 6:25 am Post subject: |
|
|
| I'd love to help a fellow ruski out but, I'm too happy with archck5 and the staircase. Good luck with your patchset tovarish (I hate being stuck with an english keyboard!). |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Tue Dec 13, 2005 9:23 am Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) |
|
|
| amax wrote: |
The staircase is still incompatible with RT patchset, talking to the Con Kolivas for getting new one.. The staircase is great. I expect to get it sometime (after Con will back from vacation) |
For help from Con on this one, you can wait until hell freezes over 
Con loathes preemption at all, he himself advices no preemption at all for staircase. Earlier, as i included RT patches in an experimental nitro, Con was laughing at us, removed the link to nitro website from his one, started hating me and ceized every (even so inofficial) support for nitro sources. We had to apologize hime REALLY hard and long.....
So, well, good luch with this one 
_________________
Retired gentoo user |
|
| Back to top |
|
|
enderandrew
l33t
Joined: 25 Oct 2005
Posts: 731
|
| Posted: Tue Dec 13, 2005 9:29 am Post subject: |
|
|
I asked about help getting ck patches to work with mm patches, and Con basically described the mm patchset as a bunch of garbage. He made it quite clear he doesn't want the two patches to be be compatible with each other.
_________________
Nihilism makes me smile. |
|
| Back to top |
|
|
gimpel
Advocate
Joined: 15 Oct 2004
Posts: 2720
Location: Munich, Bavaria
|
| Posted: Tue Dec 13, 2005 11:17 am Post subject: |
|
|
very interesting patchset - gentoo, deb and fedora fixes too.. weird stuff.
If you manage to get reiser4 to compile, i'll be able to do some big tests toghether with some audio guys from jacklab, and if this runs stable we'll make it the kernel for the audio distribution of opensuse. (SuSE has built-in reiser4 support since version 9.3 and it would be cool not to loose it in jad)
so i'm getting impatient on next release!
also many thanks for patched pam overlay! good stuff!
Ah, and before someone starts flaming... I KNOW using reiser4 on DAW is pretty munch pointless, but half of my system runs on it because it's the only fs i know that stands tons of hard resets, and i'm not doing audio fun all the time - so i'm good atm with -ck's SCHED_ISO stuff.
_________________
http://proaudio.tuxfamily.org/wiki - pro-audio software overlay
|
|
| Back to top |
|
|
enderandrew
l33t
Joined: 25 Oct 2005
Posts: 731
|
| Posted: Tue Dec 13, 2005 3:33 pm Post subject: |
|
|
I'm putting together my first patchset merely for educational purposes.
So far it is going fairly well.
I'm curious where you found broken-out patches for the Debian and Fedora kernel-sources. And if you find a way to make staircase work with the realtime patchset, please let me know!
_________________
Nihilism makes me smile. |
|
| Back to top |
|
|
bollucks
l33t
Joined: 27 Oct 2004
Posts: 606
|
| Posted: Tue Dec 13, 2005 4:04 pm Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) |
|
|
| Tiger683 wrote: | | amax wrote: |
The staircase is still incompatible with RT patchset, talking to the Con Kolivas for getting new one.. The staircase is great. I expect to get it sometime (after Con will back from vacation) |
For help from Con on this one, you can wait until hell freezes over
Con loathes preemption at all, he himself advices no preemption at all for staircase. Earlier, as i included RT patches in an experimental nitro, Con was laughing at us, removed the link to nitro website from his one, started hating me and ceized every (even so inofficial) support for nitro sources. We had to apologize hime REALLY hard and long..... |
If I recall correctly that was the "licensed you can't get the source patchset" era that we should all forget that caused that backlash. I doubt it had anything to do with preemption... |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Tue Dec 13, 2005 7:24 pm Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) |
|
|
| bollucks wrote: |
If I recall correctly that was the "licensed you can't get the source patchset" era that we should all forget that caused that backlash. I doubt it had anything to do with preemption... |
The license was the second story, about preemption, con said it into my face, additionally you can ask predatorfreak, he will tell you enough stories about Con and his *cough* love *cough* towards preemption, especially ingo's patch.
PS: don't waste your time trying to let staircase code work with RTP, unless you are as skilled as ingo and con together i know, cause i've been there :S
_________________
Retired gentoo user |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 10:42 pm Post subject: |
|
|
Talk with Con Kolivas
==========================================================================
15:34:42 <ssb> con, what do you think about -rt ?
15:35:07 <con> it serves a useful purpose, but of no use to ordinary desktop users
15:35:12 <con> gug
15:35:19 <ssb> why not?
15:35:40 <con> huh? it does nothing of any use to desktop users, thats why
15:35:59 <ssb> it looks like jack works better with -rt
15:36:07 <con> well it does
15:36:22 <con> but desktop users dont need jack
15:36:22 <obiwan> Realtime is something for people to write papers and ask for grants about 
15:36:30 <obiwan> con : nice play on words.
15:36:36 <con>
15:36:50 <predatorfreak> con: I heard you where on vacation.
15:36:56 <con> jack is overqualified for playing audio on a desktop
15:36:58 <con> predatorfreak I was
15:37:00 <con> I got home today
15:37:09 <ssb> con, is it possible to merge staircase in -rt ?
15:37:17 <con> anything's possible
15:37:19 <con> but what's the point?
15:37:23 <predatorfreak> con: welcome back than.
15:37:34 <ssb> jack will work even better
15:37:43 <con> if you want ultra high resolution low latency performance with real time
scheduling then staircase doesn't help you any more
15:37:53 <con> predatorfreak ty
15:38:30 <con> ssb staircase only has significant effect on non-rt scheduled things
15:39:01 <con> so if you are after studio quality realtime performance with all its extra
overhead and bad throughput then there is no point adding staircase to it
15:39:43 <predatorfreak> con: I always loved the way you shoot things down.
15:39:44 <neenee> glug
15:40:09 * con shrugs
15:40:12 <con> predatorfreak I just tell it like it is
15:40:32 <con> the -rt tree is for professional audio and video work; it has no place on the
deskotp
15:40:35 <con> *desktop
15:40:45 <con> it costs you in so many other ways
15:40:57 * ssb still thinks that with staircase -rt will perform even better
15:41:10 <predatorfreak> ssb: than you don't listen very well
15:42:41 <ssb> no, i listen, i just compared -rt, -ck and vanilla and it was obvious that
staircase rocks
15:43:25 <con> heh
15:43:33 <con> well for ordinary desktop work -ck is what you want
15:48:22 * ssb wants SCHED_ISO on -rt :]
15:48:46 <ssb> i tried to merge but without serial console i failed to debug it :]
15:50:28 <con> that is a non trivial thing to do
15:52:52 <predatorfreak> uhh.
15:53:12 <predatorfreak> why bother with SCHED_ISO on -rt, since as con pointed out, it's
clearly not for desktop use.
15:53:30 <predatorfreak> I'd recommend realtime-lsm or rlimits, as those are more suited to
the task of -rt.
15:58:31 <ssb> well i've never mentioned desktop. my point here is that I think that staircase
is better than ingo's scheduler for any workload.
16:00:45 <predatorfreak> not entirely.
==========================================================================
So.. let's use original scheduler.
Also, general note: THE REISER4 FILESYSTEM IS FULLY INCOMPATIBLE WITH REALTIME PATCHSET!!! WE ARE SORRY.
Also, it is well known issue, the reiser4 is bad filesystem for audio/video realtime tasks. |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 10:56 pm Post subject: |
|
|
THE GOOD NEWS
* New ebuild realtime-sources-2.6.14_p22-r5.ebuild
* added some patches from archck and nitro patchset
* Published single-file patch over 2.6.14 vanilla
* now fixes for ppc and sparc applied too
=====================================
Patches added:
8001_shfs-0.35.patch
8002_shfs-0.35-fixes.patch
8003_shfs_Makefile.patch
8004_acpi-dsdt-initramfs-fix-2.6.10-cleanup.patch
8005_acpi-dsdt-initrd-v0.7d-2.6.12.patch
8006_invis-sata-pixx.patch
8007_ipw2x00-hz-fix.patch
8008_reiser4-2.6.14-rc5-mm1-fix.patch
8009_scsi-low_hz_fix.patch
8010_unionfs-1.1.1-fix.patch
8011_reiser4-rt-fix.patch
8012-realtime-lsm.patch
and
9001_ide-probe-delay.patch
9002_increase-MAX_MP_BUSSES.patch
9003_increase-firmware-loading-timeout.patch
9004-daconfig-2.3.0.patch
9005-linux-2.6.13-e100-badeeprom.patch
9006-mppe_mppc-iptables_layer7-2.6.14.patch
9007_kmalloc-max.patch
9008_parport-mutex.patch
9009_posix-locks-match.patch
9010-svgalib_helper-1.9.21.patch
9011-usbhid-readd-kconfig.patch
9012_readahead-tune.patch
=====================================
Some new log about used patches:
>>> Unpacking source...
>>> Unpacking linux-2.6.14.tar.bz2 to /var/tmp/portage/realtime-sources-2.6.14_p22-r5/work
* genpatches-2.6.14-5.base.tar.bz2 unpacked
* genpatches-2.6.14-5.extras.tar.bz2 unpacked
* Excluding Patch #1004_2_thread-sendqueue-race.patch ... [ ok ]
* Excluding Patch #1021_3_alsa-rtc-soft-lockup.patch ... [ ok ]
* Excluding Patch #4500_fbsplash-0.9.2-r5.patch ... [ ok ]
* Excluding Patch #4305_dm-bbr.patch ... [ ok ]
* Excluding Patch #8008_reiser4-2.6.14-rc5-mm1-fix.patch ... [ ok ]
* Excluding Patch #8011_reiser4-rt-fix.patch ... [ ok ]
* Applying 1001_patch-2.6.14-rt22.patch (-p0+) ... [ ok ]
* Applying 1002_1g_lowmem1_i386.patch (-p0+) ... [ ok ]
* Applying 1003_defaultcfq.patch (-p0+) ... [ ok ]
* Applying 1004_pdflush-tweaks.patch (-p0+) ... [ ok ]
* Applying 1005_jiffies-max.patch (-p0+) ... [ ok ]
* Applying 1006_kconfig-hz-integer.patch (-p0+) ... [ ok ]
* Applying 1010_mm-swap_prefetch-18.patch (-p0+) ... [ ok ]
* Applying 1011_vm-mapped.patch (-p0+) ... [ ok ]
* Applying 1012_vm-lots_watermark.patch (-p0+) ... [ ok ]
* Applying 1013_vm-background_scan-1.patch (-p0+) ... [ ok ]
* Applying 1014_mm-kswapd_inherit_prio.patch (-p0+) ... [ ok ]
* Applying 1015_mm-prio_dependant_scan.patch (-p0+) ... [ ok ]
* Applying 1020_fix-swap_prefetch-lock.patch (-p0+) ... [ ok ]
* Applying 1021_fix-vgettimeofday-amd64.patch (-p0+) ... [ ok ]
* Applying 1022_alsa-rtc-soft-lockup.patch (-p0+) ... [ ok ]
* Applying 1023_fbsplash-0.9.2-r5.patch (-p0+) ... [ ok ]
* Applying 8001_shfs-0.35.patch (-p0+) ... [ ok ]
* Applying 8002_shfs-0.35-fixes.patch (-p0+) ... [ ok ]
* Applying 8003_shfs_Makefile.patch (-p0+) ... [ ok ]
* Applying 8004_acpi-dsdt-initramfs-fix-2.6.10-cleanup.patch (-p0+) ... [ ok ]
* Applying 8005_acpi-dsdt-initrd-v0.7d-2.6.12.patch (-p0+) ... [ ok ]
* Applying 8006_invis-sata-pixx.patch (-p0+) ... [ ok ]
* Applying 8007_ipw2x00-hz-fix.patch (-p0+) ... [ ok ]
* Applying 8009_scsi-low_hz_fix.patch (-p0+) ... [ ok ]
* Applying 8010_unionfs-1.1.1-fix.patch (-p0+) ... [ ok ]
* Applying 8012-realtime-lsm.patch (-p0+) ... [ ok ]
* Applying 9001_ide-probe-delay.patch (-p0+) ... [ ok ]
* Applying 9002_increase-MAX_MP_BUSSES.patch (-p0+) ... [ ok ]
* Applying 9003_increase-firmware-loading-timeout.patch (-p0+) ... [ ok ]
* Applying 9004-daconfig-2.3.0.patch (-p0+) ... [ ok ]
* Applying 9005-linux-2.6.13-e100-badeeprom.patch (-p0+) ... [ ok ]
* Applying 9006-mppe_mppc-iptables_layer7-2.6.14.patch (-p0+) ... [ ok ]
* Applying 9007_kmalloc-max.patch (-p0+) ... [ ok ]
* Applying 9008_parport-mutex.patch (-p0+) ... [ ok ]
* Applying 9009_posix-locks-match.patch (-p0+) ... [ ok ]
* Applying 9010-svgalib_helper-1.9.21.patch (-p0+) ... [ ok ]
* Applying 9011-usbhid-readd-kconfig.patch (-p0+) ... [ ok ]
* Applying 9012_readahead-tune.patch (-p0+) ... [ ok ]
* Applying amd64-int3-fix.patch (-p0+) ... [ ok ]
* Applying buslogic-pci-id-table.patch (-p0+) ... [ ok ]
* Applying fbdev-radeon-noaccel.patch (-p0+) ... [ ok ]
* Applying linux-2.6-debug-taint-proprietary-helpers.patch (-p0+) ... [ ok ]
* Applying linux-2.6-input-kill-stupid-messages.patch (-p0+) ... [ ok ]
* Applying linux-2.6-max-symlinks.patch (-p0+) ... [ ok ]
* Applying linux-2.6-net-sundance-ip100A.patch (-p0+) ... [ ok ]
* Applying linux-2.6-obsolete-idescsi-warning.patch (-p0+) ... [ ok ]
* Applying linux-2.6-obsolete-oss-warning.patch (-p0+) ... [ ok ]
* Applying linux-2.6-procfs-i_nlink-miscalculate.patch (-p0+) ... [ ok ]
* Applying linux-2.6-sata-enable-atapi-by-default.patch (-p0+) ... [ ok ]
* Applying linux-2.6-swsusp-nofreeze.patch (-p0+) ... [ ok ]
* Applying linux-2.6-valid-ether-addr.patch (-p0+) ... [ ok ]
* Applying linux-2.6-vm-oomkiller-debugging.patch (-p0+) ... [ ok ]
* Applying linux-2.6-vm-silence-atomic-alloc-failures.patch (-p0+) ... [ ok ]
* Applying linux-2.6-x86-vga-vidfail.patch (-p0+) ... [ ok ]
* Applying linux-2.6-x86_64-silence-up-apic-errors.patch (-p0+) ... [ ok ]
* Applying linux-2.6.14-kauditd-suspend.patch (-p0+) ... [ ok ]
* Applying net-bonding-consolidation-fix.patch (-p0+) ... [ ok ]
* Applying net-ipconntrack-nat-fix.patch (-p0+) ... [ ok ]
* Applying qla2xxx-removed.patch (-p0+) ... [ ok ]
* Applying setkeys-needs-root-1.patch (-p0+) ... [ ok ]
* Applying tty-locking-fixes9.patch (-p0+) ... [ ok ]
* Applying 1000_1_sysctl-unregistration.patch (-p0+) ... [ ok ]
* Applying 1001_2_airo_cs-correct-prototypes.patch (-p0+) ... [ ok ]
* Applying 1002_2_cfq-io-sched-fix.patch (-p0+) ... [ ok ]
* Applying 1003_2_fix-alpha-breakage.patch (-p0+) ... [ ok ]
* Applying 1005_2_ptrace-self-attach.patch (-p0+) ... [ ok ]
* Applying 1006_2_signal-live-leak.patch (-p0+) ... [ ok ]
* Applying 1007_2_zero-size-datagram.patch (-p0+) ... [ ok ]
* Applying 1008_2_ipvs-connection-leak.patch (-p0+) ... [ ok ]
* Applying 1009_2_prism54-frame-size.patch (-p0+) ... [ ok ]
* Applying 1010_2_tcp-bic-max-increment.patch (-p0+) ... [ ok ]
* Applying 1011_2_usb-interface-modalias.patch (-p0+) ... [ ok ]
* Applying 1012_2_xfs-modular-quota.patch (-p0+) ... [ ok ]
* Applying 1013_3_4GB-memory-intel-dual-core.patch (-p0+) ... [ ok ]
* Applying 1014_3_ctnetlink-protoinfo.patch (-p0+) ... [ ok ]
* Applying 1015_3_ctnetlink-no-icmp-oops.patch (-p0+) ... [ ok ]
* Applying 1016_3_eicon-signature.patch (-p0+) ... [ ok ]
* Applying 1017_3_ipv6-ah-length.patch (-p0+) ... [ ok ]
* Applying 1018_3_file-lease-leak.patch (-p0+) ... [ ok ]
* Applying 1019_3_ipv6-advapi-memory.patch (-p0+) ... [ ok ]
* Applying 1020_3_ipv6-extension-headers.patch (-p0+) ... [ ok ]
* Applying 1022_3_hdlc-netif-carrier.patch (-p0+) ... [ ok ]
* Applying 1023_3_hwmon-it87-fan-speed-limits.patch (-p0+) ... [ ok ]
* Applying 1024_3_hwmon-lm78-vid.patch (-p0+) ... [ ok ]
* Applying 1025_3_hwmon-w83627hf-limits.patch (-p0+) ... [ ok ]
* Applying 1026_3_conntrack-high-ports.patch (-p0+) ... [ ok ]
* Applying 1027_3_conntrack-syn-push.patch (-p0+) ... [ ok ]
* Applying 1028_3_nat-refcount.patch (-p0+) ... [ ok ]
* Applying 1029_3_netfilter-queue-oops.patch (-p0+) ... [ ok ]
* Applying 1030_3_packet-writing-oops-fix.patch (-p0+) ... [ ok ]
* Applying 1031_3_ppc64-numa.patch (-p0+) ... [ ok ]
* Applying 1032_3_pptp-endianness.patch (-p0+) ... [ ok ]
* Applying 1033_3_pptp-call-id.patch (-p0+) ... [ ok ]
* Applying 1034_3_trace-auto-reap.patch (-p0+) ... [ ok ]
* Applying 1035_3_netfilter-refcount-leak.patch (-p0+) ... [ ok ]
* Applying 1100_setkeys-needs-root.patch (-p0+) ... [ ok ]
* Applying 1101_setkeys-read-ok.patch (-p0+) ... [ ok ]
* Applying 1105_invalidate-inodes-overflow.patch (-p0+) ... [ ok ]
* Applying 1300_ppc64-multilib.patch (-p0+) ... [ ok ]
* Applying 1305_ppc-powerbook-5-8.patch (-p0+) ... [ ok ]
* Applying 1340_sparc-obp64-naming.patch (-p0+) ... [ ok ]
* Applying 1360_sparc-kconfig-depends.patch (-p0+) ... [ ok ]
* Applying 1362_sparc-kconfig-fb.patch (-p0+) ... [ ok ]
* Applying 1399_sparc-U1-hme-lockup.patch (-p0+) ... [ ok ]
* Applying 1910_kconfig-no-nls.patch (-p0+) ... [ ok ]
* Applying 2500_dvb-budget-ci-linking.patch (-p0+) ... [ ok ]
* Applying 2505_cx88-quiet.patch (-p0+) ... [ ok ]
* Applying 2700_acpi-driver-binding-hp-boot-hang.patch (-p0+) ... [ ok ]
* Applying 2705_ht-slowdown.patch (-p0+) ... [ ok ]
* Applying 2710_ht-dlowdown-2.patch (-p0+) ... [ ok ]
* Applying 2900_microtek-fix.patch (-p0+) ... [ ok ]
* Applying 4101_deprecate-sk98lin.patch (-p0+) ... [ ok ]
* Applying 4300_via-vt6410.patch (-p0+) ... [ ok ]
* Applying 4320_promise-pdc2037x.patch (-p0+) ... [ ok ]
* Applying 4351_megaraid-compatibility.patch (-p0+) ... [ ok ]
* Applying 4505_vesafb-tng-1.0-rc1-r2.patch (-p0+) ... [ ok ]
* Applying 4705_squashfs-2.2.patch (-p0+) ... [ ok ]
* Applying 4900_speakup-20051028.patch (-p0+) ... [ ok ]
* Applying 4905_alpha-sysctl-uac.patch (-p0+) ... [ ok ]
CHK include/linux/version.h
UPD include/linux/version.h
>>> Source unpacked.
=====================================
try cd /usr/src/linux
and
cp 2.6.14_p22-realtime-r5.config .config
You will use default config for realtime-sources
=====================================
Single file patchset you can download HERE http://chronos.nsu.ru/fantoo/realtime/linux-2.6.14_p22-realtime-r5.patch.bz2
Now we have only one problem with ati-drivers, but I expected to fix this problem today.
Waiting for feedback, for fixes or any ideas/improvements AND benchmark results! (use something like interbench)
My personal notes:
1. Use reiserfs instead reiser4, b/c reiserfs works well with evms/lvm2, and can be resized online.
2. For me the staircase is BAD scheduler, with it I have 2 unresloved issues:
- The quake3 game have lags with sound events and quick movements
- If I compite anything, and I try start other compile process, it starts VERY long (long python/emerge work, etc)
- If I have LA > 10 - the start any new program at first time like OOo or FF or Opera is suicidal for time (with realtime kernel all ok with LA ~20)
Last edited by amax on Tue Dec 13, 2005 11:10 pm; edited 1 time in total |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Tue Dec 13, 2005 11:03 pm Post subject: |
|
|
Second note:
Now we have 2 disabled patches in ebuild
* 8008_reiser4-2.6.14-rc5-mm1-fix.patch
* 8011_reiser4-rt-fix.patch
Probably anyone can rewrite parts of reiser4 code for compatibility with realtime. |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
|
| Back to top |
|
|
Earthwings
Bodhisattva
Joined: 14 Apr 2003
Posts: 7753
Location: Germany
|
| Posted: Tue Dec 13, 2005 11:22 pm Post subject: |
|
|
Please use the edit function to add content when your post is the last in a thread. Rule of thumb is waiting at least a day before bumping a thread. Thank you.
PS: Gaming and realtime patches are really a bad idea. I don't want to offend you, but maybe you should do some reading on that thing "realtime".
_________________
KDE |
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
|
| Back to top |
|
|
amax
Tux's lil' helper
Joined: 22 Feb 2004
Posts: 76
Location: Russia, Novosibirsk (Silicon Taiga)
|
| Posted: Wed Dec 14, 2005 12:00 am Post subject: interbench results |
|
|
| Earthwings wrote: | Please use the edit function to add content when your post is the last in a thread. Rule of thumb is waiting at least a day before bumping a thread. Thank you.
PS: Gaming and realtime patches are really a bad idea. I don't want to offend you, but maybe you should do some reading on that thing "realtime". |
the results from interbench (see Read and None tests):
VANILLA:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 1.06 +/- 3.27 10.8 99
Video 69.2 +/- 69.8 90.3 59.1
X 96.7 +/- 214 1500 50.8
Burn 318 +/- 416 811 23.9
Write 9.37 +/- 53.7 757 91.4
Read 3.85 +/- 5.3 15.9 96.3
Compile 346 +/- 405 1149 22.4
Memload 3.08 +/- 5.75 31.6 97
CK6:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 0.551 +/- 2.38 13.5 99.5
Video 66.8 +/- 67.1 83.8 60
X 102 +/- 200 1050 49.6
Burn 392 +/- 436 488 20.3
Write 9.76 +/- 28.4 164 91.1
Read 1.48 +/- 2.96 14 98.5
Compile 436 +/- 487 634 18.6
Memload 2.1 +/- 5.09 39.5 97.9
RT:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 0.184 +/- 0.713 5.85 99.8
Video 66.2 +/- 66.5 74.1 60.2
X 101 +/- 202 1221 49.7
Burn 312 +/- 354 403 24.3
Write 10.5 +/- 65.1 810 90.5
Read 2.76 +/- 3.04 5.1 97.3
Compile 342 +/- 418 1321 22.6
Memload 4.81 +/- 7.53 53.6 95.4 |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Wed Dec 14, 2005 2:39 pm Post subject: Re: interbench results |
|
|
| amax wrote: | | Earthwings wrote: | Please use the edit function to add content when your post is the last in a thread. Rule of thumb is waiting at least a day before bumping a thread. Thank you.
PS: Gaming and realtime patches are really a bad idea. I don't want to offend you, but maybe you should do some reading on that thing "realtime". |
the results from interbench (see Read and None tests):
VANILLA:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 1.06 +/- 3.27 10.8 99
Video 69.2 +/- 69.8 90.3 59.1
X 96.7 +/- 214 1500 50.8
Burn 318 +/- 416 811 23.9
Write 9.37 +/- 53.7 757 91.4
Read 3.85 +/- 5.3 15.9 96.3
Compile 346 +/- 405 1149 22.4
Memload 3.08 +/- 5.75 31.6 97
CK6:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 0.551 +/- 2.38 13.5 99.5
Video 66.8 +/- 67.1 83.8 60
X 102 +/- 200 1050 49.6
Burn 392 +/- 436 488 20.3
Write 9.76 +/- 28.4 164 91.1
Read 1.48 +/- 2.96 14 98.5
Compile 436 +/- 487 634 18.6
Memload 2.1 +/- 5.09 39.5 97.9
RT:
--- Benchmarking simulated cpu of Gaming in the presence of simulated ---
Load Latency +/- SD (ms) Max Latency % Desired CPU
None 0.184 +/- 0.713 5.85 99.8
Video 66.2 +/- 66.5 74.1 60.2
X 101 +/- 202 1221 49.7
Burn 312 +/- 354 403 24.3
Write 10.5 +/- 65.1 810 90.5
Read 2.76 +/- 3.04 5.1 97.3
Compile 342 +/- 418 1321 22.6
Memload 4.81 +/- 7.53 53.6 95.4 |
Yeah, i did that too once, and guess what, interbench says nothing about RTP patches.
Read up on preemption a little and then come back.
I used to be a big fan of Ingo's patch until i got sick and tired of real life applications
spasmatically chopping around on my desktop just for those few msecs of startup time .
Hard Preemption is ONLY usable for applications with strict demand for meeting all critical
deadlines of their SHORT living, critical tasks/threads.
If an applications involves threads running longer than a few msecs which are cpu/throughput intensive
you have a problem, they will get cpu preempted away from underneeth their feet,
such applications are for example games or computational applications, X also seems
to get affected by this misbehaviour under certain circumstances, in particular
when you have more than just one or two applications running in it.
If you are trying to emulate a whole professional music studio on your box, and that with demand
for digitally high quality sound and no latencies/lag between tracks etc. then you are best suited with this one.
Also, realtime, time-critical embedded systems might profit from this too
rgds.
T
_________________
Retired gentoo user |
|
| Back to top |
|
|
mpalencia
n00b
Joined: 16 Feb 2005
Posts: 51
Location: Colombia
|
| Posted: Wed Dec 14, 2005 4:51 pm Post subject: |
|
|
Tiger: Is realtime good for digital video capture? firewire and/or analog video trough v4l?.
I do a lot of these captures, but dont use realtime patchset, just wanted to know if this colud be useful for this sort of task.
_________________
Lucky Livecd 32 and 64 bit versions (reiser4 suppport)
https://forums.gentoo.org/viewtopic-p-2986139.html#2986139
There are 2 things infinite: the universe and human stupidity, but I am not sure about the first one. (Einstein) |
|
| Back to top |
|
|
Dominique_71
Veteran
Joined: 17 Aug 2005
Posts: 1944
Location: Switzerland (Romandie)
|
| Posted: Sun Feb 12, 2006 10:06 pm Post subject: |
|
|
I have done my own realtime kernel with a vanilla kernel, Ingo Molnar's patch and the realtime-lsm module. With an Audigy2, I get a sound latency with jack of 1.33msec at 48kHz. I want to know which latency with a such soundcard you are able to archive with this RT kernel.
_________________
"Confirm You are a robot." - the singularity |
|
| Back to top |
|
|
indianiec
Tux's lil' helper
Joined: 01 Feb 2005
Posts: 97
Location: Krakow
|
| Posted: Thu Mar 09, 2006 10:54 pm Post subject: Re: [realtime-sources] 2.6.14 (-rt22 based + ck6) [UPDATED] |
|
|
And english here |
|
| Back to top |
|
|
|
Unsupported Software
