View previous topic :: View next topic |
Author |
Message |
Dottout l33t
Joined: 07 Mar 2006 Posts: 882
|
Posted: Sat Oct 13, 2007 5:28 pm Post subject: |
|
|
yes, only performance as default has no sense, this was one of my favourite patches from -mm |
|
Back to top |
|
|
termite Guru
Joined: 06 May 2007 Posts: 466
|
Posted: Sat Oct 13, 2007 5:29 pm Post subject: |
|
|
Can anyone help with my TOI issues? As I said before, I can suspend, but when resuming, I get either a blank screen (text mode) or a black screen with a flickering cursor (X)... |
|
Back to top |
|
|
Dottout l33t
Joined: 07 Mar 2006 Posts: 882
|
Posted: Sat Oct 13, 2007 5:33 pm Post subject: |
|
|
I dont use TOI so I cannot help you. but I suggest you to give more info about your hardware |
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
Posted: Sat Oct 13, 2007 5:46 pm Post subject: |
|
|
termite wrote: | Can anyone help with my TOI issues? As I said before, I can suspend, but when resuming, I get either a blank screen (text mode) or a black screen with a flickering cursor (X)... |
Videocard, TOI config...? Also check TOI bugzilla: http://bugzilla.tuxonice.net/ |
|
Back to top |
|
|
SlashBeast Retired Dev
Joined: 23 May 2006 Posts: 2922
|
Posted: Sat Oct 13, 2007 5:55 pm Post subject: |
|
|
dodo1122 wrote: | prestige wrote: | massimo wrote: | Tagx wrote: | It seems that whenever my cpu is under load, input starts to lag causing letters to repeat (and keyboard shortcuts).
Everything was working fine with 2.6.22-kamikazex and prior.
If i can't fix it, itll prob force me to downgrade to .22.
|
I'm experiencing something similar: sometimes the mouse movement seems not that smooth at all even if the cpu is running idle. I'll try turning off new features like air group CPU scheduler since I do not know what might causing this... |
I have experienced that too. When I compile something and the cpu is very high load the mouse is... jumping. After remove the Fair group CPU scheduler the problem is gone. |
Confirmed, it happens to me is well.
ygwin wrote: | I had a weird issue with python not being able to merge, due to some vague IO error. I suspect the reiser4 patch isn't 100% guaranteed. |
Have the same problem, with python and net-tools :/
dodo |
&& www servers have problems on reiser4 with this patch. For example, i try to download txt, html, png and others files i have error Code: | 19:22:54 (0.00 B/s) - Connection closed at byte 0. Ponawianie próby. | but php on mod_fastcgi work fine.
Its reiser4 error or this reiser4 patch from kamikaze ? I use gentoo-sources 2.6.23 with reiser4patch from kamikaze.
Sorry about my english.
ADDED:http://gentoo-wiki.com/TIP_reiser4emergetip
Last edited by SlashBeast on Sat Oct 13, 2007 6:04 pm; edited 1 time in total |
|
Back to top |
|
|
Dottout l33t
Joined: 07 Mar 2006 Posts: 882
|
Posted: Sat Oct 13, 2007 6:01 pm Post subject: |
|
|
only reiser4 users? anyone who run a working ext4 -o mballoc root?I can't, it looks broken. however, data=writeback + delalloc increased a lot my root's performances |
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
Posted: Sat Oct 13, 2007 6:36 pm Post subject: |
|
|
SlashBeast wrote: | Its reiser4 error or this reiser4 patch from kamikaze ? I use gentoo-sources 2.6.23 with reiser4patch from kamikaze. |
There have been some changes in 2.6.23 (sendfile() function) that affect reiser4.
Quote: | 2.13. Use splice in the sendfile() implementation
Splice is a innovative I/O method which was added in Linux 2.6.17, based in a in-kernel buffer that the user has control over, where "splice()" moves data to/from the buffer from/to an arbitrary file descriptor with splice(), while "tee()" copies the data in one buffer to another, ie: it "duplicates" it, or vmsplice() to splice the data from/to user memory. Because the in-kernel buffer isn't really copied from one address space to another, it allows to move data from/to a fd without an extra copy (ie, "zero-copy").
For the particular case of sending the data from a file descriptor to a fd socket, there's been always the sendfile() syscall. splice() however is a generic mechanism, not just limited to what sendfile(). In other words, sendfile() is just a small subset of what splice can do, splice obsoletes it. In Linux 2.6.23, the sendfile() mechanism implementation is killed, but the API and its functionality is not removed, it's instead implemented internally with the splice() mechanisms.
Because sendfile() is critical for many programs, specially for static web servers and FTPs, performance regressions could happen (and performance improvements!) and the kernel hackers would really like to hear about them both in linux-kernel@vger.kernel.org and/or other usual communication channels. |
|
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
Posted: Sat Oct 13, 2007 9:00 pm Post subject: |
|
|
Testing preview version of kamikaze3 with "acpi-release-20070116", "sata-an" and "sata-alpm" patches. Working fine atm.
And the best of all: only ~50 wakeups per second (NEW RECORD!) |
|
Back to top |
|
|
prestige Tux's lil' helper
Joined: 24 Sep 2005 Posts: 95
|
Posted: Sat Oct 13, 2007 11:34 pm Post subject: |
|
|
I have one general question about the patches. How can I exclude some patch from the big patchset. For example, how can I exclude the three shed-cfs patches from the kamikaze2? Or how can I exclude the ck patch from the kamikaze2? What is the procedure?
I am asking this because I have faced another problem with kamikaze2 and I want to try some "custom" kamikaze2 patchset. The problem is with the .avi files. They skips for a second every about a minute.
Waninkoko wrote: | Testing preview version of kamikaze3 with "acpi-release-20070116", "sata-an" and "sata-alpm" patches. Working fine atm.
And the best of all: only ~50 wakeups per second (NEW RECORD!) |
Sorry, but this is another lame question: what are this "wakeups"? |
|
Back to top |
|
|
dodo1122 Guru
Joined: 02 Sep 2006 Posts: 347 Location: York, England
|
Posted: Sat Oct 13, 2007 11:54 pm Post subject: |
|
|
prestige wrote: | I have one general question about the patches. How can I exclude some patch from the big patchset. For example, how can I exclude the three shed-cfs patches from the kamikaze2? Or how can I exclude the ck patch from the kamikaze2? What is the procedure?
I am asking this because I have faced another problem with kamikaze2 and I want to try some "custom" kamikaze2 patchset. The problem is with the .avi files. They skips for a second every about a minute.
Waninkoko wrote: | Testing preview version of kamikaze3 with "acpi-release-20070116", "sata-an" and "sata-alpm" patches. Working fine atm.
And the best of all: only ~50 wakeups per second (NEW RECORD!) |
Sorry, but this is another lame question: what are this "wakeups"? |
You can download broken-out patches, and apply only the ones you want to use, or apply all of them, and then
Code: | patch -p1 -R < patch-you-don't-want.patch |
You need to download the patches you don't want though.
dodo _________________ #zen-sources on irc.rizon.net |
|
Back to top |
|
|
prestige Tux's lil' helper
Joined: 24 Sep 2005 Posts: 95
|
Posted: Sun Oct 14, 2007 12:10 am Post subject: |
|
|
dodo1122 wrote: |
You can download broken-out patches, and apply only the ones you want to use, or apply all of them, and then
Code: | patch -p1 -R < patch-you-don't-want.patch |
You need to download the patches you don't want though.
dodo |
This was my first guess, but did not work. Then I decided to ask because obviously this is a wrong approach.
Code: |
# tar jxf linux-2.6.23.tar.bz2
# cp linux-2.6.23-kamikaze2 linux-2.6.23/
# cp ckpatches-2.6.23.patch linux-2.6.23/
# patch -p1 < linux-2.6.23-kamikaze2
...(cut many lines here. All goes fine)...
# patch -p1 -R < ckpatches-2.6.23.patch
patching file include/linux/sched.h
Hunk #1 succeeded at 40 (offset 3 lines).
patching file kernel/sched.c
Hunk #1 succeeded at 4113 (offset -252 lines).
Hunk #2 succeeded at 4148 with fuzz 2 (offset -253 lines).
Hunk #3 succeeded at 4209 (offset -243 lines).
patching file block/cfq-iosched.c
Hunk #1 succeeded at 1278 (offset 2 lines).
patching file include/linux/ioprio.h
Hunk #2 succeeded at 61 with fuzz 1 (offset 10 lines).
patching file Documentation/sysctl/vm.txt
Hunk #2 succeeded at 33 with fuzz 1 (offset 2 lines).
Hunk #3 succeeded at 89 (offset 3 lines).
Hunk #4 succeeded at 277 (offset 61 lines).
patching file include/linux/swap.h
Hunk #2 FAILED at 188.
Hunk #3 succeeded at 239 (offset 2 lines).
1 out of 3 hunks FAILED -- saving rejects to file include/linux/swap.h.rej
patching file init/Kconfig
Hunk #1 succeeded at 104 (offset -1 lines).
patching file kernel/sysctl.c
Hunk #2 succeeded at 69 (offset -1 lines).
Hunk #3 succeeded at 797 (offset 170 lines).
Hunk #4 succeeded at 875 (offset 170 lines).
Hunk #5 succeeded at 1246 with fuzz 1 (offset 364 lines).
patching file mm/Makefile
Hunk #1 succeeded at 15 with fuzz 2 (offset -2 lines).
patching file mm/swap.c
patching file mm/swap_prefetch.c
patching file mm/swap_state.c
patching file mm/vmscan.c
Hunk #3 succeeded at 63 with fuzz 1.
Hunk #4 succeeded at 101 (offset -9 lines).
Hunk #5 succeeded at 968 (offset 165 lines).
Hunk #6 succeeded at 1120 (offset 165 lines).
Hunk #7 succeeded at 1176 (offset 165 lines).
Hunk #8 succeeded at 1183 (offset 164 lines).
Hunk #9 succeeded at 1205 (offset 165 lines).
Hunk #10 succeeded at 1235 (offset 165 lines).
Hunk #11 succeeded at 1285 (offset 165 lines).
Hunk #12 succeeded at 1295 with fuzz 1 (offset 165 lines).
Hunk #13 succeeded at 1304 (offset 166 lines).
Hunk #14 succeeded at 1311 (offset 166 lines).
Hunk #15 succeeded at 1329 (offset 166 lines).
Hunk #16 succeeded at 1364 (offset 166 lines).
Hunk #17 succeeded at 1404 (offset 166 lines).
Hunk #18 FAILED at 1438.
Hunk #19 succeeded at 1552 (offset 233 lines).
Hunk #20 succeeded at 1565 (offset 233 lines).
Hunk #21 succeeded at 1582 (offset 233 lines).
Hunk #22 succeeded at 1599 (offset 236 lines).
Hunk #23 succeeded at 1719 (offset 337 lines).
Hunk #24 succeeded at 1778 (offset 337 lines).
Hunk #25 succeeded at 1813 (offset 337 lines).
Hunk #26 succeeded at 1877 (offset 337 lines).
Hunk #27 succeeded at 1961 (offset 337 lines).
1 out of 27 hunks FAILED -- saving rejects to file mm/vmscan.c.rej
patching file include/linux/mm_inline.h
patching file include/linux/swap-prefetch.h
patching file mm/page_io.c
patching file include/linux/sysctl.h
Hunk #1 succeeded at 194 (offset 4 lines).
patching file include/linux/mmzone.h
Hunk #2 succeeded at 190 (offset 9 lines).
Hunk #3 succeeded at 479 (offset 27 lines).
Hunk #4 succeeded at 495 (offset 27 lines).
patching file mm/page_alloc.c
Hunk #1 succeeded at 1275 (offset 25 lines).
Hunk #2 succeeded at 1339 (offset 25 lines).
Hunk #3 succeeded at 1601 (offset 31 lines).
Hunk #4 succeeded at 1612 (offset 31 lines).
Hunk #5 succeeded at 3673 (offset 531 lines).
patching file fs/buffer.c
Hunk #1 succeeded at 440 (offset 84 lines).
patching file mm/filemap.c
Hunk #1 succeeded at 469 (offset 3 lines).
Hunk #2 succeeded at 844 with fuzz 2 (offset 5 lines).
Hunk #3 succeeded at 1045 (offset -6 lines).
patching file Documentation/filesystems/proc.txt
Hunk #1 succeeded at 1407 (offset 74 lines).
patching file arch/i386/Kconfig
Hunk #1 succeeded at 567 (offset 17 lines).
Hunk #2 succeeded at 586 (offset 17 lines).
patching file kernel/Kconfig.hz
Hunk #4 FAILED at 45.
1 out of 4 hunks FAILED -- saving rejects to file kernel/Kconfig.hz.rej
patching file arch/i386/defconfig
Hunk #1 succeeded at 221 (offset -5 lines).
patching file arch/x86_64/defconfig
Hunk #1 succeeded at 183 (offset -2 lines).
patching file include/linux/jiffies.h
patching file include/net/inet_timewait_sock.h
patching file init/calibrate.c
patching file arch/i386/kernel/cpu/proc.c
Hunk #1 succeeded at 166 (offset 9 lines).
patching file arch/i386/kernel/smpboot.c
Hunk #1 succeeded at 1081 (offset -13 lines).
patching file include/linux/nfsd/stats.h
patching file arch/x86_64/kernel/setup.c
Hunk #1 succeeded at 1098 (offset 51 lines).
|
|
|
Back to top |
|
|
dodo1122 Guru
Joined: 02 Sep 2006 Posts: 347 Location: York, England
|
Posted: Sun Oct 14, 2007 12:15 am Post subject: |
|
|
just try to compile it, maybe the fails are just empty lines or unneeded stuff
dodo _________________ #zen-sources on irc.rizon.net |
|
Back to top |
|
|
SlashBeast Retired Dev
Joined: 23 May 2006 Posts: 2922
|
Posted: Sun Oct 14, 2007 8:12 am Post subject: |
|
|
I have another question: ck have SD scheduler and kamikaze use ckpatches, 2.6.23 have CFS scheduler - how it work together ?
btw thanks for this patchset. |
|
Back to top |
|
|
mbar Veteran
Joined: 19 Jan 2005 Posts: 1990 Location: Poland
|
Posted: Sun Oct 14, 2007 9:06 am Post subject: |
|
|
they don't, read the thread ffs. |
|
Back to top |
|
|
r2dtu Tux's lil' helper
Joined: 23 Aug 2002 Posts: 87
|
|
Back to top |
|
|
buddabrod Apprentice
Joined: 15 Oct 2006 Posts: 241 Location: Germany
|
Posted: Sun Oct 14, 2007 9:23 am Post subject: |
|
|
Waninkoko wrote: | Testing preview version of kamikaze3 with "acpi-release-20070116", "sata-an" and "sata-alpm" patches. Working fine atm.
And the best of all: only ~50 wakeups per second (NEW RECORD!) | Yay, i'm at home again^^ I got 10-20wakepus per second with 2.6.22-kamikaze9
I was about to rewrite my buggy alpm patch, but it seems like you were faster |
|
Back to top |
|
|
treffer Apprentice
Joined: 14 Dec 2004 Posts: 150
|
Posted: Sun Oct 14, 2007 10:02 am Post subject: |
|
|
are wrote: | Don't have any comparison, but i'm sure, linux supports my laptop much better than any other system now. |
I started with vista/linux dual boot (and got rid of vista due to personal dislike and a total vista crash), windows never reached more than 2:30 battery life, linux reaches up to 2:40. Powertop/Lesswats did a great job in improving linux, your battery will now last longer in linux. At least with the latest kernel.
So how about sending out some spam? "Enlarge your battery life! Install linux! Instand download availabel at ...." would be a good topic... _________________ root@localhost# whois POEM-RIPE55-SONG
root@localhost# : ( ) { : | : & } ; : |
|
Back to top |
|
|
pietrek n00b
Joined: 09 Oct 2007 Posts: 14
|
Posted: Sun Oct 14, 2007 10:38 am Post subject: |
|
|
What happened to fbsplash? I'm missing my big Gentoo logo on the framebuffer |
|
Back to top |
|
|
SlashBeast Retired Dev
Joined: 23 May 2006 Posts: 2922
|
Posted: Sun Oct 14, 2007 11:26 am Post subject: |
|
|
Thx! This patches solved my problem with emerge python when portage_tmpdir is on reiser4
Code: | # mount |grep dropzone2
/dev/md0 on /mnt/dropzone2 type reiser4 (rw)
/mnt/dropzone2 on /var/tmp/portage type none (rw,bind) |
But, www-servers still have problem and send '0 bytes' on static files, php under mod_fastcgi work. |
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
Posted: Sun Oct 14, 2007 1:10 pm Post subject: |
|
|
I'm changing from the old server to a new and faster server (Intel Xeon 1,86GHz with 4 processors, 4GB RAM, 2TB hard disk...).
And there's a new domain name available: http://waninkoko.es and http://kamikaze.waninkoko.es |
|
Back to top |
|
|
are Apprentice
Joined: 03 Jan 2006 Posts: 188
|
Posted: Sun Oct 14, 2007 1:24 pm Post subject: |
|
|
Dottout wrote: | only reiser4 users? anyone who run a working ext4 -o mballoc root?I can't, it looks broken. however, data=writeback + delalloc increased a lot my root's performances |
Please excuse my dumb question: What is "-o mballoc", can I use it with jfs too? (Never understand the hype for reiserfs, jfs is the second best performer in every single discipline meaning the best over all tests).
Thanks and best regards
are |
|
Back to top |
|
|
Dottout l33t
Joined: 07 Mar 2006 Posts: 882
|
Posted: Sun Oct 14, 2007 1:28 pm Post subject: |
|
|
no you cant. each fs has its own set of mount options. jfs flags are here: /usr/src/linux/Documentation/filesystems/jfs.txt |
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
Posted: Sun Oct 14, 2007 1:34 pm Post subject: |
|
|
My portage partition uses ext4 and I mount it with extents, delalloc and mballoc |
|
Back to top |
|
|
Dottout l33t
Joined: 07 Mar 2006 Posts: 882
|
Posted: Sun Oct 14, 2007 1:38 pm Post subject: |
|
|
which baselayout version are you using? I tried only with baselayout-2.0.0_rc5, I should make a second attempt with rc4
edit: tested also with older baselayout versions. it hangs at 'remounting r\w'
Last edited by Dottout on Sun Oct 14, 2007 1:54 pm; edited 1 time in total |
|
Back to top |
|
|
Waninkoko Guru
Joined: 13 May 2005 Posts: 549
|
|
Back to top |
|
|
|