View previous topic :: View next topic |
Author |
Message |
m0p Apprentice
Joined: 20 Jun 2005 Posts: 205 Location: en_GB
|
Posted: Mon Jul 25, 2005 11:20 am Post subject: |
|
|
Is there any plans to add some ext3 latency patches like they have been doing with morph-sources? As a die-hard ext3 user, some ext3 patches might be cool. |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Mon Jul 25, 2005 11:49 am Post subject: |
|
|
I dont know if i should stop arguing about people not reading announcements (ie. patchlistS) ,
or make my BIG FAT RED FONT even BIGGER AND FATTER ......
MM-patches IN THIS THREAD'S ANNOUNCEMENT!!!! wrote: |
(...)
ext3-reduce-allocate-with-reservation-lock-latencies.patch
(..)
|
In case anyone asks where did the other ones go: in latest mm-tree they got all consolidated into this one single patch.
PS:
im also a die-hard ext3 user, and my filesystem has probably all ext3 tweaks possible applied.... _________________ Retired gentoo user |
|
Back to top |
|
|
ktech Guru
Joined: 12 Apr 2003 Posts: 340
|
Posted: Mon Jul 25, 2005 2:57 pm Post subject: |
|
|
would be nice if some user benchmarked both ext3 and reiser3... |
|
Back to top |
|
|
fallow Bodhisattva
Joined: 08 Jan 2004 Posts: 2208 Location: Poland
|
Posted: Mon Jul 25, 2005 3:11 pm Post subject: |
|
|
ktech wrote: | would be nice if some user benchmarked both ext3 and reiser3... |
OMG, use search button and find tons of benchmarks like this.
Holly Jesus , ppl just use "search" or ASK google.
and about ppl not reading the RELEASE INFO / RELEASE ANNOUCEMENT and asking questions answered in rel.info
hehe Its very normal, so then just don't worry
example : https://forums.gentoo.org/viewtopic-p-2301717.html#2301717
cheers. _________________ "Time is a companion that goes with us on a journey. It reminds us to cherish each moment, because it will never come again. What we leave behind is not as important as how we have lived" J-L. Picard |
|
Back to top |
|
|
ktech Guru
Joined: 12 Apr 2003 Posts: 340
|
Posted: Mon Jul 25, 2005 3:19 pm Post subject: |
|
|
but I was talking about ext3 with that last set of patches applied... |
|
Back to top |
|
|
JetAce44 Apprentice
Joined: 02 Feb 2004 Posts: 152
|
Posted: Mon Jul 25, 2005 8:11 pm Post subject: |
|
|
Still have the problem mentioned before with a large (559gb) reiserfs 3.6 partition.
Deletion works perfectly under my dev-sources 2.6.5 kernel, however I'd like to stick to a more recent kernel.
Also, I'm still having trouble with ivtv:
Code: |
FATAL: Error inserting msp3400 (/lib/modules/2.6.12-nitro5/kernel/drivers/media/video/msp3400.ko): Unknown symbol in module, or unknown parameter (see dmesg)
FATAL: Module saa7115 not found.
FATAL: Error inserting tuner (/lib/modules/2.6.12-nitro5/kernel/drivers/media/video/tuner.ko): Unknown symbol in module, or unknown parameter (see dmesg)
FATAL: Module ivtv not found.
FATAL: Error running install command for ivtv
|
Dmesg:
Code: |
msp3400: Unknown parameter `simple'
tuner: Unknown parameter `type'
|
Any ideas? _________________ Tyan Tiger MPX 2466
2x AMD Athlon MP 1800+
1gig Crucial PC2100 ECC
NV 6600GT
1x 1000JB WD 100gig 8mb Cache IDE Drive
4x 2000JD WD 200gig 8mb Cache Sata Drive (Under RAID 5 on a 3ware 8506-4LP)
Registered Linux User #346685 |
|
Back to top |
|
|
mathfeel l33t
Joined: 03 Aug 2004 Posts: 700
|
Posted: Mon Jul 25, 2005 9:00 pm Post subject: |
|
|
IPW2200 not working...
I chose NO for kernel configure, then try to emerge ipw2200 from portage with error:
Code: | /var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_adhoc_create':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:5459: warning: implicit declaration of function `ieee80211_channel_to_index'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:5462: error: structure has no member named `passive_only'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_wx_set_encodeext':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:6670: error: structure has no member named `host_encrypt_msdu'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:6674: error: structure has no member named `host_encrypt_msdu'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:6679: warning: implicit declaration of function `ieee80211_wx_set_encodeext'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_wx_get_encodeext':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:6688: warning: implicit declaration of function `ieee80211_wx_get_encodeext'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_associate_network':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:7290: error: `WLAN_CAPABILITY_SHORT_SLOT' undeclared (first use in this function)
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:7290: error: (Each undeclared identifier is reported only once
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:7290: error: for each function it appears in.)
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_sw_reset':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:8048: error: structure has no member named `host_encrypt_msdu'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:8059: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:8071: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_wx_set_freq':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:8215: error: structure has no member named `passive_only'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_wx_set_txpow':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:8802: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `ipw_wx_set_wireless_mode':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:9216: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:9221: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:9230: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:9237: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:9243: error: structure has no member named `abg_true'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c: In function `shim__set_security':
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:10198: error: structure has no member named `encode_alg'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:10198: error: structure has no member named `encode_alg'
/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.c:10244: error: `SEC_ENCRYPT' undeclared (first use in this function)
make[2]: *** [/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6/ipw2200.o] Error 1
make[1]: *** [_module_/var/tmp/portage/ipw2200-1.0.6/work/ipw2200-1.0.6] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.12-nitro5'
make: *** [modules] Error 2
!!! ERROR: net-wireless/ipw2200-1.0.6 failed.
!!! Function linux-mod_src_compile, Line 491, Exitcode 2
!!! Unable to make KSRC=/usr/src/linux KSRC_OUTPUT=/usr/src/linux IEEE80211_INC=/usr/include all.
!!! If you need support, post the topmost build error, NOT this status message.
|
Which version is the ipw2200 in nitro-5? |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Mon Jul 25, 2005 9:29 pm Post subject: |
|
|
Err, am i retarded or are you people making fun of me?
Now, in the 12-th TEXTline of the announcement it reads:
Green text, 12 line of text, announcement post wrote: |
This kernel already contains the ipw2100 and ipw2200 drivers same version as portage !!!
|
Also, a few posts back i did a suggestion which might resolve the issue with the driver in kernel (don't try to compile the ebuild version, you wont make it... )
Tiger683 wrote: | Need4Speed:
try this maybe:
tagwar wrote: |
(..) I had the same problem... emerge unmerge ipw2100-firmware and emerge ipw2100-firmware fixed it... if I remember correctly..
(..)
|
|
_________________ Retired gentoo user |
|
Back to top |
|
|
Illissius Guru
Joined: 31 Jul 2004 Posts: 395 Location: Hungary
|
Posted: Tue Jul 26, 2005 12:27 pm Post subject: |
|
|
I have noticed something: mapped watermark is dumb. When you aren't using much memory, it's great, no swapping at all -- but once you *do* start using a lot of memory, it completely sucks, because what it does is set the amount of memory for file cache and 'normal' memory to fixed amounts -- so even though I have 512MB of memory, programs can't actually use more than 360-somethingMB of it... everything is getting completely swapped to hell and back at least ten times per second, I do free -m, and a good 160 or so MB of memory is being taken up by file cache. That, sucks.
I could set the knob to a higher value, meaning less space for file cache, but
(a) say if I set it to 99, would that mean I only ever have 5-10MB of file cache (even if everything else is only taking up 100MB), or is it at least smart enough to use any extra unused memory for file cache too? (given the above, I'm not sure whether to be confident);
and more importantly (b) setting it to anything but the default value is supposedly unstable (why provide a setting when you aren't supposed to touch it, I don't know) -- and I did get stuff occasionally killed from oom when I tried it, so no thanks.
So, getting to the point finally, would it be possible to provide a menuconfig option to choose between the old swap code (w/ the swappiness knob) and mapped watermark, or is it more complicated than that? If it is, I may have to switch to vanilla- or gentoo-sources, because I can hardly stand this anymore
(nitro is great and works flawlessly otherwise, so I wouldn't like that) _________________ Work is punishment for failing to procrastinate effectively.
last.fm |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Tue Jul 26, 2005 2:15 pm Post subject: |
|
|
Illissius:
If it only is possible by any means, i'll be more than pleased to do it....
So, lets hope, i might as well do another nitro, but the problem is:
i'd also be more than pleased to finally touch 2.6.13 code, especially because im
not very keen to break anything between nitro5 and nitro6...
if it will work, i might provide a nitro6 with something like it plus a few updates....
until then, let's get busy
T _________________ Retired gentoo user |
|
Back to top |
|
|
tagwar Tux's lil' helper
Joined: 11 Aug 2004 Posts: 147 Location: Karlsruhe
|
Posted: Tue Jul 26, 2005 4:57 pm Post subject: |
|
|
JetAce44 wrote: |
Also, I'm still having trouble with ivtv:
Any ideas? |
I'm running ivtv-0.3.6u nicely with 2.6.12-nitro3.. dont remember if I already tried it on nitro5, but I think so..
your error looks like misconfiguration in modprobe.conf.... sorry, dont have time now to dig it out, but for a quick hint try:
Code: | cd /etc;grep -R msp3400 mod*|grep simple; grep -R tuner mod*|grep type |
and remove all occurances of "simple" together with msp3400 and "type" together with tuner.
Also make sure you're using the tuner module from ivtv and not the one provided with the kernel... same goes for all other modules provided with ivtv... (who would have guessed... ) |
|
Back to top |
|
|
m0p Apprentice
Joined: 20 Jun 2005 Posts: 205 Location: en_GB
|
Posted: Tue Jul 26, 2005 5:19 pm Post subject: |
|
|
Hmm, is the CFQ scheduler any better than genetic anticipatory? And, does the genetic lib help with latency, or does it just increase latency? |
|
Back to top |
|
|
Illissius Guru
Joined: 31 Jul 2004 Posts: 395 Location: Hungary
|
Posted: Tue Jul 26, 2005 7:40 pm Post subject: |
|
|
Tiger683 wrote: | Illissius:
If it only is possible by any means, i'll be more than pleased to do it.... |
Thanks a lot
Tiger683 wrote: | So, lets hope, i might as well do another nitro, but the problem is:
i'd also be more than pleased to finally touch 2.6.13 code |
Well... my problem with rc kernels is they tend to be hit and miss (some randomly lock up, others don't); I'd have no problem with 2.6.13, as long as it's a hit . (but this may be hard to tell beforehand...) _________________ Work is punishment for failing to procrastinate effectively.
last.fm |
|
Back to top |
|
|
JetAce44 Apprentice
Joined: 02 Feb 2004 Posts: 152
|
Posted: Wed Jul 27, 2005 12:02 am Post subject: |
|
|
tagwar wrote: | JetAce44 wrote: |
Also, I'm still having trouble with ivtv:
Any ideas? |
I'm running ivtv-0.3.6u nicely with 2.6.12-nitro3.. dont remember if I already tried it on nitro5, but I think so..
your error looks like misconfiguration in modprobe.conf.... sorry, dont have time now to dig it out, but for a quick hint try:
Code: | cd /etc;grep -R msp3400 mod*|grep simple; grep -R tuner mod*|grep type |
and remove all occurances of "simple" together with msp3400 and "type" together with tuner.
Also make sure you're using the tuner module from ivtv and not the one provided with the kernel... same goes for all other modules provided with ivtv... (who would have guessed... ) |
Got it running finally, it just required a bit of tweaking with the ivtv-driver source file.
The one problem that still plauges this kernel is my problems with deleting and creating large files. I know it is something in the kernel, because my previous 2.6.5 kernel worked fine.
I checked the filesystem, and it is completely clean. My system disk, also reiserfs3.6 does not have the same problem. Ideas anyone? Could enabling reiser4 support but not using reiser4 cause this? _________________ Tyan Tiger MPX 2466
2x AMD Athlon MP 1800+
1gig Crucial PC2100 ECC
NV 6600GT
1x 1000JB WD 100gig 8mb Cache IDE Drive
4x 2000JD WD 200gig 8mb Cache Sata Drive (Under RAID 5 on a 3ware 8506-4LP)
Registered Linux User #346685 |
|
Back to top |
|
|
mathfeel l33t
Joined: 03 Aug 2004 Posts: 700
|
Posted: Wed Jul 27, 2005 12:31 am Post subject: |
|
|
Tiger683 wrote: wrote: |
Green text, 12 line of text, announcement post wrote: |
This kernel already contains the ipw2100 and ipw2200 drivers same version as portage !!!
|
|
Sorry, but I also notice that many lines below, there is an [EDIT] and it say "Reported as NOT WORKING"
Anything, it DOES work. One less thing to remerge when I recompile kernel. NICE JOB guys... |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Wed Jul 27, 2005 8:23 am Post subject: |
|
|
mathfeel: thanx, will remove that edit line then
cheers
T
PS: People, we'll be getting an official nitro sources fbsplash theme sometime soon,
if anyone feels like helping XmN with this, please do so, i already added the progressbar to his 1024x768 config,
look into this thread if interested (its still work in progress, if you really want to use it now, read the whole thread carefully) :
https://forums.gentoo.org/viewtopic-t-363372.html _________________ Retired gentoo user |
|
Back to top |
|
|
Need4Speed Guru
Joined: 06 Jun 2004 Posts: 497
|
Posted: Wed Jul 27, 2005 10:31 pm Post subject: |
|
|
hmm, I still can't get my 2200bg to work, i've reemerged everything but I'm still getting the error I posted. I guess I'll just try some older kernels and see if they work. |
|
Back to top |
|
|
bosyotech Tux's lil' helper
Joined: 29 Jan 2005 Posts: 79 Location: earth
|
Posted: Thu Jul 28, 2005 3:38 am Post subject: |
|
|
since the DOC (ck) has released a wonderful benchmarking tool for desktop usage maybe the kernel hackers can use this to finally create the ultimate kernel for desktops? http://members.optusnet.com.au/ckolivas/interbench/ _________________ of all the things i've lost, i miss my mind the most |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Thu Jul 28, 2005 10:37 am Post subject: |
|
|
Bosyotech:
baeksu wrote: | vrln wrote: | sorry this is a bit off topic, but seems like CK has released http://members.optusnet.com.au/ckolivas/interbench/ --> has anyone tried it yet and compared ck-based patchsets (like nitro) to vanilla 2.6? Would be interesting to see some statistics. | I did a bunch of tests with different recent patchsets (ck-based) and a vanilla kernel. I won't bother putting the results here, but the tests indicated three interesting points.
1. Differences between ck and vanilla kernels were fuzzy at best
Con K. wrote: | ...I have to admit you need to squint hard to be convinced that -ck is better overall. Both mainline and -ck perform better in different load settings (source) |
Though latency of "X" was generally lower with ck-patches, latencies were higher for "Burn" and "Compile".
2. Genetic anticipatory io-scheduler introduces higher read/write latencies that cfq
Yes, on all three ck-based patchsets that I tried, cfq had lower latencies in io than genetic anticipatory.
3. Staircase scheduler tunables decreases latency more than just ck-patches
Though there's obviously performance costs, and stability risks as well, test results were significantly better for lower values for the tunables. Personally, it seems that if you don't mind sacrificing performance (a.k.a. "speed") for interactivity, finding lower-but-still-stable values for the tunables is not a bad idea.
I've been running nitro5 for about 24 hours now, and I haven't had any stability issues yet. As for the test results, it just goes to show how fuzzy the concept of "interactivity" (and responsiveness) can be, and how difficult it is to measure and compare.
There's a definite (and obvious to most) trade-off between performance, interactivity/responsiveness, and stability. At least one of these has to be sacrificed to improve the others. |
Couldn't put it better myself _________________ Retired gentoo user |
|
Back to top |
|
|
kare n00b
Joined: 06 Jun 2004 Posts: 26 Location: Friedberg, Germany
|
Posted: Thu Jul 28, 2005 4:01 pm Post subject: |
|
|
Need4Speed wrote: | hmm, I still can't get my 2200bg to work, i've reemerged everything but I'm still getting the error I posted. I guess I'll just try some older kernels and see if they work. |
I had the same problem and solved it disabling ipw2200 and ieee80211 in the kernel and emerging ipw2200-1.0.4 and ipw2200-firmware-2.3.
I know it's not the best, but it works for the moment. |
|
Back to top |
|
|
Debentoo_Gao Tux's lil' helper
Joined: 21 Nov 2003 Posts: 85 Location: Shanghai,China
|
Posted: Sat Jul 30, 2005 5:59 pm Post subject: |
|
|
m0p wrote: | Debentoo_Gao wrote: | Could ati-drivers-8.4.13-r1 work with nitro-sources-2.6.12-r5 & gcc 4.0.1 ? I can't use the 2.6.12 patches & gcc 4.x patches together |
I found a patch that "supposedly" works for gcc4 and 2.6.12, but it wouldn't patch succesfuly. |
Could u give me the patches I'd like to try it _________________ More Love~~More Dream~~More Happiness
Free Your Mind~~Free Your Body~~Free Your Life
http://www.myblue.ws |
|
Back to top |
|
|
Phlogiston Veteran
Joined: 27 Jan 2004 Posts: 1925 Location: Europe, Swizerland
|
Posted: Mon Aug 01, 2005 6:35 pm Post subject: |
|
|
hi guys!
I was on holiday for 3 weeks and there have been 3 nitro releases in this time: Great work Tiger
So now I must try out latest nitro... I'll report soon how it works. Now I'm still running 2.6.12-nitro2
cu
/edit: hey tiger, did you remove lirc in kernel? No problem for me, but if its still there, I would give it a try at least _________________ Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor |
|
Back to top |
|
|
bakkus n00b
Joined: 17 Dec 2002 Posts: 33
|
Posted: Mon Aug 01, 2005 7:46 pm Post subject: MPPC-patch |
|
|
Would it be possible for one of you nitro-devs to post the file linux-2.6.12-mppe-mppc-1.3.patch.gz as a separate patch somewhere?
The official site for the patch seems down (not been able to contact it for the last two days) and as the wireless network at my college uses this thing I really need it
Thanks in advance |
|
Back to top |
|
|
Tiger683 Veteran
Joined: 08 Jan 2005 Posts: 1347 Location: Heffner's House
|
Posted: Tue Aug 02, 2005 8:37 am Post subject: |
|
|
Hi, guys, i'm currently on vacations (and for the next two months).
I was writing my exams recently so no time for nitro.
If anyone wants a new release, bug seppe, maybe he will be
able to do something about it
cheers,
T _________________ Retired gentoo user |
|
Back to top |
|
|
Phlogiston Veteran
Joined: 27 Jan 2004 Posts: 1925 Location: Europe, Swizerland
|
Posted: Tue Aug 02, 2005 8:42 am Post subject: |
|
|
Tiger683 wrote: | Hi, guys, i'm currently on vacations (and for the next two months).
cheers,
T |
Have a nice time.
PS: New nitro working like a charm here _________________ Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor |
|
Back to top |
|
|
|
|
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
|
|