Post #159,946
6/15/04 12:28:38 AM
|
Shouldn't kernel 2.6 be faster than 2.4?
I just installed it and this thing is dragging. Granted, I've got a low-end machine, but I thought this upgrade was supposed to boost speed.
===
Implicitly condoning stupidity since 2001.
|
Post #159,953
6/15/04 6:26:08 AM
|
Odd. Faster here...
Imric's Tips for Living
- Paranoia Is a Survival Trait
- Pessimists are never disappointed - but sometimes, if they are very lucky, they can be pleasantly surprised...
- Even though everyone is out to get you, it doesn't matter unless you let them win.
|
Nothing is as simple as it seems in the beginning, As hopeless as it seems in the middle, Or as finished as it seems in the end.
|
|
Post #159,990
6/15/04 10:16:56 AM
|
Depends.
By default the 2.6.x kernels are faster.
But, Debian is getting a bit whacked right now during the rush to release (yeah right... probably 3-5 months away) that some things are being done that may slow down you machine.
hdparm is a good thing to install for the 2.6.x kernel. Turning on DMA and 32-bit access and mult-sec transfers.
Also, look at what is being loaded as far as modules... sometimes things are getting loaded you don't want. Personally I am deleting all IPV6, OpenSoundSystem (OSS) and some AGP modules. I then usae alsa for sound (default in 2.6) Then I start looking at other things just running.
You might be a bit memory starved.
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #159,994
6/15/04 10:39:41 AM
|
Heh, you think?
===
Implicitly condoning stupidity since 2001.
|
Post #160,010
6/15/04 1:54:49 PM
|
Perhaps you need to kill all the *IDLE* processes?
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Blog]
|
Post #160,024
6/15/04 3:16:45 PM
|
And now, I shall SHUN you a second time
===
Implicitly condoning stupidity since 2001.
|
Post #160,193
6/16/04 10:42:53 PM
|
Just did lsmod ... *way* too much running
Not much listed in /etc/modules, so where do I go to start trimmimg?
===
Implicitly condoning stupidity since 2001.
|
Post #160,253
6/17/04 2:30:49 PM
|
Personally:
You will have lotsa modules now with 2.6.
But all depends on what is there. I nuke ipv6 modules (they are re-installable) and OSS (open sound system) and some agp stuff as my nvidia stuff dunnah like agpgart or via_agp
There is a blacklist of modules you never want loaded at startup. Hotplug and discover will load eveyrthing under the blooming sun right now. Blacklists for each one :)... Fun eh?
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,256
6/17/04 2:55:00 PM
6/17/04 2:56:08 PM
|
Dupe, ignore.
This isn't the dupe you're looking for.\r\n\r\nMove along, move along.
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Blog]
Edited by pwhysall
June 17, 2004, 02:56:08 PM EDT
|
Post #160,257
6/17/04 2:55:01 PM
|
feh
ccordelia:/home/peter# uname -a\r\nLinux cordelia 2.6.6 #1 SMP Sat Jun 5 20:15:08 BST 2004 i686 GNU/Linux\r\ncordelia:/home/peter# lsmod\r\nModule Size Used by\r\nfglrx 207116 10\r\ncordelia:/home/peter#\r\n :-)
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Blog]
|
Post #160,304
6/17/04 7:04:20 PM
|
Welcome to we-be-a-ninnie.
Peter, how many times have you Grouched at me for compiling a kernel?
Peter, are you telling me *YOU* compiled a NON-STOCK kernel then?
FOR SHAME.... FOR SHAME!
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,365
6/18/04 1:40:59 AM
|
Er, never. Why?
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Blog]
|
Post #160,388
6/18/04 8:56:36 AM
|
...
Everytime I compiled a kernel YOU said something on the order of: Nobody needs to compile a kernel for a "Home" system e-var, you stupid Colonial Twit. Those kinds of comments stick in ones memory a bit longer.
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,511
6/19/04 6:06:20 AM
|
You're thinking of someone else.
And it's "illiterate colonial baboon", you illiterate colonial baboon.
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Blog]
|
Post #160,551
6/19/04 6:01:04 PM
|
That explains it...
You "Stupid Empirical Idiot."
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,065
6/15/04 9:35:21 PM
|
Still slow as hell booting, and even worse opening apps
I know I'm low on memory, but no more than I was with the 2.4 kernel. I just booted into the old kernel to check and it's still noticeable faster.
===
Implicitly condoning stupidity since 2001.
|
Post #160,178
6/16/04 8:56:07 PM
|
Are you sure you are using hdparm yet?
Cause iffin you ain't you are seriously missing out. I know that 2.6 is much more usable when DMA, Mults, 32/wsync access and read-ahed enbaled. 2.4 also benefits from that, but not as much as 2.6 From my /etc/hdparm.conf: /dev/hda { io32_support = 3 read_ahead_sect = 256 dma = on read_only = on interrupt_unmask = on } /dev/hdc { io32_support = 3 read_ahead_sect = 256 dma = on interrupt_unmask = on } /dev/hde { mult_sect_io = 16 lookahead = on io32_support = 3 read_ahead_sect = 256 dma = on acoustic_management = 254 } /dev/hdg { mult_sect_io = 16 io32_support = 3 read_ahead_sect = 256 dma = on acoustic_management = 254 }
Be cautioned as the read_ahead_sect is LOTS lower for 2.4 (I think 8 or 16)
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,185
6/16/04 10:11:00 PM
|
Mine, the whole file is commented out
How do I know what's the right config for me?
===
Implicitly condoning stupidity since 2001.
|
Post #160,220
6/17/04 8:10:46 AM
|
Here is what I use...
The Specs from the HD or: duke:~# hdparm -acdgiIkmnru /dev/hde\n\n/dev/hde:\n multcount = 16 (on)\n IO_support = 3 (32-bit w/sync)\n unmaskirq = 0 (off)\n using_dma = 1 (on)\n keepsettings = 0 (off)\n nowerr = 0 (off)\n readonly = 0 (off)\n readahead = 256 (on)\n geometry = 65535/16/63, sectors = 117347328, start = 0\n\n Model=WDC WD600BB-23CAA1, FwRev=17.07W17, SerialNo=WD-WMA8F3496798\n Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq }\n RawCHS=16383/16/63, TrkSize=57600, SectSize=600, ECCbytes=40\n BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16\n CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=117347328\n IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}\n PIO modes: pio0 pio1 pio2 pio3 pio4\n DMA modes: mdma0 mdma1 mdma2\n UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5\n AdvancedPM=no WriteCache=enabled\n Drive conforms to: device does not report version:\n\n * signifies the current active mode\n\n\nATA device, with non-removable media\n Model Number: WDC WD600BB-23CAA1\n Serial Number: WD-WMA8F3496798\n Firmware Revision: 17.07W17\nStandards:\n Supported: 5 4 3 2\n Likely used: 6\nConfiguration:\n Logical max current\n cylinders 16383 16383\n heads 16 16\n sectors/track 63 63\n --\n CHS current addressable sectors: 16514064\n LBA user addressable sectors: 117347328\n device size with M = 1024*1024: 57298 MBytes\n device size with M = 1000*1000: 60081 MBytes (60 GB)\nCapabilities:\n LBA, IORDY(can be disabled)\n bytes avail on r/w long: 40 Queue depth: 1\n Standby timer values: spec'd by Standard, with device specific minimum\n R/W multiple sector transfer: Max = 16 Current = 16\n Recommended acoustic management value: 128, current value: 254\n DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5\n Cycle time: min=120ns recommended=120ns\n PIO: pio0 pio1 pio2 pio3 pio4\n Cycle time: no flow control=120ns IORDY flow control=120ns\nCommands/features:\n Enabled Supported:\n * READ BUFFER cmd\n * WRITE BUFFER cmd\n * Host Protected Area feature set\n * Look-ahead\n * Write cache\n * Power Management feature set\n Security Mode feature set\n * SMART feature set\n * Device Configuration Overlay feature set\n * Automatic Acoustic Management feature set\n SET MAX security extension\n * DOWNLOAD MICROCODE cmd\n * SMART self-test\n * SMART error logging\nSecurity:\n supported\n not enabled\n not locked\n not frozen\n not expired: security count\n not supported: enhanced erase\nHW reset results:\n CBLID- above Vih\n Device num = 0 determined by CSEL\nChecksum: correct\nduke:~# Pull the bits and pieces out of that.
-- [link|mailto:greg@gregfolkert.net|greg], [link|http://www.iwethey.org/ed_curry|REMEMBER ED CURRY!] @ iwethey
Heard near the SCOG employee entry/exit way:
Security: We got another Mass Exodus Doorway Jam.
|
Post #160,226
6/17/04 10:00:35 AM
|
So that command shows what the drive supports?
Then I just need to find the bits that match the hdparm config file? Okay, will try tonight.
===
Implicitly condoning stupidity since 2001.
|