Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
my harddrive performance is sucking
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8, 9  Next  
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
radonsg
Apprentice
Apprentice


Joined: 26 Jun 2004
Posts: 257
Location: Singapore

PostPosted: Wed Jun 30, 2004 7:08 pm    Post subject: Reply with quote

hdparm /dev/hde

/dev/hde:
multcount = 16 (on)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 20023/255/63, sectors = 321672960, start = 0

hdparm -i /dev/hde

/dev/hde:

Model=HDS722516VLSA80, FwRev=V34OA60A, SerialNo=VNRD3EC4CATKWM
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
BuffType=DualPortCache, BuffSize=7938kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=268435455
IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 udma2
AdvancedPM=yes: disabled (255) WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 3a:

* signifies the current active mode

hdparm -tT /dev/hde

/dev/hde:
Timing buffer-cache reads: 1952 MB in 2.00 seconds = 974.20 MB/sec
Timing buffered disk reads: 176 MB in 3.01 seconds = 58.48 MB/sec

Sata harddisk performance not that bad I feel.
_________________
Athlon64 3000+ on a MSI Neo2 Platinum (nVidia nForce3)
1 Sata Hitachi (HGST) 7K250
1 Sata Samsung SpinPoint
2 x 512MB TwinMos DDR400 Dual Channel
Geforce 5200
Back to top
View user's profile Send private message
CrAzYChIcK
n00b
n00b


Joined: 21 Aug 2004
Posts: 6
Location: NY

PostPosted: Thu Aug 26, 2004 6:17 am    Post subject: Reply with quote

this is a pretty good thread...
figured i'd post a solution to a problem i had while trying to enable udma mode 6 on my drive. could not get it to do anything over udma 33, turned out it was because i didn't have the Master hdd on the end of the ide cable. worked fine after i put it on the end. udma 133 now. (i feel dumb lol)

/dev/hda:
Timing buffer-cache reads: 1824 MB in 2.00 seconds = 911.23 MB/sec
Timing buffered disk reads: 144 MB in 3.04 seconds = 47.42 MB/sec

(Maxtor 6Y060L0 60GB 2MB cache ATA133 on nforce2 IDE controller)
Back to top
View user's profile Send private message
feld
Guru
Guru


Joined: 29 Aug 2004
Posts: 593
Location: WI, USA

PostPosted: Mon Sep 06, 2004 5:53 am    Post subject: Reply with quote

the drive i have my Linux on is slave......

so ur saying that if i swapped my drives and put it as root and redo my fstab and grub.conf to apply accordingly then it would run better?




-Feld
Back to top
View user's profile Send private message
dungara
n00b
n00b


Joined: 09 Oct 2004
Posts: 1

PostPosted: Sat Oct 09, 2004 10:47 am    Post subject: strange behaviour with CD-RW no difference due to dma Reply with quote

Hi,
I find strange behaviour with CD-RW( /dev/hdc) on ICH5 chip.
with:
Code:
unmaskirq    =  1 (on)
using_dma    =  0 (off)
keepsettings =  0 (off)
readonly     =  0 (off)
readahead    = 64 (on)

I get following speeds
Code:
 Timing cached reads:   1984 MB in  2.00 seconds = 991.16 MB/sec
 Timing buffered disk reads:   10 MB in  3.71 seconds =   2.70 MB/sec

now I enabled dma:
Code:
#hdparm -d 1 /dev/hdc
#hdparm /dev/hdc
 IO_support   =  1 (32-bit)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 keepsettings =  0 (off)
 readonly     =  0 (off)
 readahead    = 64 (on)
#hdparm -tT /dev/hdc
 Timing cached reads:   1952 MB in  2.00 seconds = 975.17 MB/sec
 Timing buffered disk reads:   10 MB in  3.70 seconds =   2.70 MB/sec

There is no difference in speed
Although my cd-rw supports udma2 mode.
Code:
# hdparm -I /dev/hdc
..............
ATAPI CD-ROM, with removable media
        Model Number:       SONY    CD-RW  CRX225E
..............
Standards:
        Used: ATAPI for CD-ROMs, SFF-8020i, r2.5
        Supported: CD-ROM ATAPI-2
...............
Capabilities:
        LBA, IORDY(cannot be disabled)
        DMA: mdma0 mdma1 mdma2 udma0 udma1 *udma2
             Cycle time: min=120ns recommended=120ns
        PIO: pio0 pio1 pio2 pio3 pio4
             Cycle time: no flow control=227ns  IORDY flow control=120ns


Also when I run cdrecord
Code:
#cdrecord -scanbus -dev=ATAPI:/dev/hdc
..........
Warning: The related Linux kernel interface code seems to be unmaintained.
Warning: There is absolutely NO DMA, operations thus are slow.
...........

Any idea why there is no difference after enabling dma?
Why cdrecord complains that there is no dma?
PS: system config
linux-2.6.9-rc2
configured with dma enabled
I result of harddisk(/dev/hda)
Code:
/dev/hda:
 Timing cached reads:   1940 MB in  2.00 seconds = 969.66 MB/sec
 Timing buffered disk reads:  170 MB in  3.01 seconds =  56.47 MB/sec
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Fri Apr 01, 2005 1:51 pm    Post subject: Reply with quote

when i do a hdparm -tT /dev/hda i get these strange readings:
Code:
Timing cached reads: 2056 MB in 2.00 seconds = 1027.13 MB/sec
Timing buffered disk reads: 4 MB in 3.07 seconds = 1.30 MB/sec

when i test my other drive i get 990Mb/s and 35MB/s.

i've checked the hdparm -i outputs and the only difference is the BuffType= option.
On the a drive it is unknown and on the b drive it is DualPortCache.
How can i fix this, and get both my drives up to speed???

[EDIT]
sorry didn't realize it was an old thread.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Zorglab
n00b
n00b


Joined: 05 Nov 2004
Posts: 11

PostPosted: Fri Apr 01, 2005 3:23 pm    Post subject: Reply with quote

squanto wrote:
Bloody Bastard: thank you for enlightening me. I did not know that. Now it makes sense that my 2 drives would have similar ratings for buffered cache reads. :P

Andrew

Can anyone beat my 45.98MB / sec? just curious. I run ReiserFS with notail on my WD special edition, on Epox 8KHA+ board. :?:


Code:
# hdparm -Tt /dev/hdc

/dev/hdc:
 Timing cached reads:   1768 MB in  2.00 seconds = 882.37 MB/sec
 Timing buffered disk reads:  164 MB in  3.00 seconds =  54.60 MB/sec


with -d1X69c1u1m16 on the 120GB 7200rpm Ultra ATA/100 HDD of my laptop. Quite impressive, isn't it ?
BTW I'm using reiser4, but I don't know if this affects the results, because I also have ntfs, fat32 and ext2 partitions on this disk.
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Fri Apr 01, 2005 4:31 pm    Post subject: Reply with quote

Sade wrote:
when i do a hdparm -tT /dev/hda i get these strange readings:
Code:
Timing cached reads: 2056 MB in 2.00 seconds = 1027.13 MB/sec
Timing buffered disk reads: 4 MB in 3.07 seconds = 1.30 MB/sec

when i test my other drive i get 990Mb/s and 35MB/s.

i've checked the hdparm -i outputs and the only difference is the BuffType= option.
On the a drive it is unknown and on the b drive it is DualPortCache.
How can i fix this, and get both my drives up to speed???

[EDIT]
sorry didn't realize it was an old thread.

Doesn't matter, its obviously a popular thread ;-)
Can you show us the full hdparm -i for both disks.
mvh
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
John5788
Advocate
Advocate


Joined: 06 Apr 2004
Posts: 2140
Location: 127.0.0.1

PostPosted: Fri Apr 01, 2005 9:14 pm    Post subject: Reply with quote

Code:
[/home/john5788 :: 01:11 PM :: 0]
[root@john5788.x5788.net] $ hdparm -tT /dev/hda

/dev/hda:
 Timing cached reads:   1900 MB in  2.00 seconds = 949.20 MB/sec
 Timing buffered disk reads:  166 MB in  3.03 seconds =  54.72 MB/sec

[/home/john5788 :: 01:12 PM :: 0]
[root@john5788.x5788.net] $ hdparm -i /dev/hda

/dev/hda:

 Model=HDS722580VLAT20, FwRev=V32OA60A, SerialNo=VNR21FC2RB0EZL
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=52
 BuffType=DualPortCache, BuffSize=1794kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=160836480
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=yes: disabled (255) WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 3a:

 * signifies the current active mode


7200rpm 80gb deskstar is /dev/hda


Code:
[/home/john5788 :: 01:13 PM :: 0]
[root@john5788.x5788.net] $ hdparm -tT /dev/hdb

/dev/hdb:
 Timing cached reads:   1844 MB in  2.00 seconds = 922.14 MB/sec
 Timing buffered disk reads:  152 MB in  3.03 seconds =  50.22 MB/sec

[/home/john5788 :: 01:13 PM :: 0]
[root@john5788.x5788.net] $ hdparm -i /dev/hdb

/dev/hdb:

 Model=WDC WD800EB-00DJF0, FwRev=77.07W77, SerialNo=WD-WCAHL5449667
 Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq }
 RawCHS=16383/16/63, TrkSize=57600, SectSize=600, ECCbytes=74
 BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=156301488
 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: device does not report version:

 * signifies the current active mode



/dev/hdb is a 5400rpm 80gb western digital

Code:
[/home/john5788 :: 01:14 PM :: 0]
[root@john5788.x5788.net] $ hdparm -tT /dev/hdd

/dev/hdd:
 Timing cached reads:   1848 MB in  2.00 seconds = 923.68 MB/sec
 Timing buffered disk reads:   60 MB in  3.07 seconds =  19.53 MB/sec

[/home/john5788 :: 01:14 PM :: 0]
[root@john5788.x5788.net] $ hdparm -i /dev/hdd

/dev/hdd:

 Model=SAMSUNG SV2042H, FwRev=PK100-16, SerialNo=0263J1BR141684
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
 RawCHS=16383/16/63, TrkSize=34902, SectSize=554, ECCbytes=4
 BuffType=DualPortCache, BuffSize=426kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=39865392
 IORDY=yes, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: ATA/ATAPI-4 T13 1153D revision 17:

 * signifies the current active mode


/dev/hdd is a 20gb samsung 5400rpm.
_________________
John5788
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Fri Apr 01, 2005 9:53 pm    Post subject: Reply with quote

Cintra wrote:
[Can you show us the full hdparm -i for both disks.

Ok here goes:
Code:
/dev/hda:

 Model=ST380011A, FwRev=3.06, SerialNo=3JV1HH9N
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=unknown, BuffSize=2048kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=156301488
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

/dev/hdb:

 Model=Maxtor 4G120J6, FwRev=GAK819K0, SerialNo=G60L3DRE
 Config={ Fixed }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57
 BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=240121728
 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 udma5 *udma6
 AdvancedPM=yes: disabled (255) WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 0:


hda is a seagate barracuda and it is also the drive performing badly atm.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sat Apr 02, 2005 4:49 am    Post subject: Reply with quote

Ok, by looking back some pages I see other Seagate drives have ' BuffType=unknown' so you can forget that..

There was at least one other Seagate with similar results to yours,
i.e. very good 'Timing cached reads' and very low 'buffered disk reads'.

So, do you have anything set up in /etc/conf.d/hdparm and do you start hdparm when booting?

I have the following in /etc/conf.d/hdparm:
Code:

hda_args="-m16 -u1 -c3 -d1"
hdb_args="-m16 -u1 -c3 -d1"

and 'hdparm' in default run level as shown with 'rc-status -a'

Just do the following after you have edited /etc/conf.d/hdparm, if you don't have it set yet
Code:
 # rc-update add hdparm default
 * hdparm added to runlevel default
 * rc-update complete.

and see if that helps
mvh

Edit: I just came across this thread where the poster used
Code:
hdparm -i -M -m 1 -d -a -c 3 -u 1 -t -T -X 69 /dev/hdx

on a Seagate like yours with following results
Code:
 Timing buffer-cache reads:   1940 MB in  2.00 seconds = 969.66 MB/sec
 Timing buffered disk reads:  160 MB in  3.02 seconds =  52.97 MB/sec

see https://forums.gentoo.org/viewtopic-t-167101-highlight-st380011a.html
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Sat Apr 02, 2005 11:23 am    Post subject: Reply with quote

Cintra wrote:
So, do you have anything set up in /etc/conf.d/hdparm and do you start hdparm when booting?

I use this, to config the driver at bootup (i did the rc-update):
Code:
# cat /etc/conf.d/hdparm
hda_args="-a256A1c1d1m16u1"
cdrom0_args="-d1c1u1

Cintra wrote:
Edit: I just came across this thread where the poster used
Code:
hdparm -i -M -m 1 -d -a -c 3 -u 1 -t -T -X 69 /dev/hdx

on a Seagate like yours with following results
Code:
 Timing buffer-cache reads:   1940 MB in  2.00 seconds = 969.66 MB/sec
 Timing buffered disk reads:  160 MB in  3.02 seconds =  52.97 MB/sec

i tried those settings with these results:
Code:
root@jpc j # hdparm -iMm1dac3u1tTX69 /dev/hda

/dev/hda:
 setting 32-bit IO_support flag to 3
 setting multcount to 1
 setting unmaskirq to 1 (on)
 setting xfermode to 69 (UltraDMA mode5)
 multcount    =  1 (on)
 IO_support   =  3 (32-bit w/sync)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 readahead    = 256 (on)

 Model=ST380011A, FwRev=3.06, SerialNo=3JV1HH9N
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=unknown, BuffSize=2048kB, MaxMultSect=16, MultSect=1
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=156301488
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

 * signifies the current active mode

 acoustic     =  0 (128=quiet ... 254=fast)
 Timing cached reads:   1984 MB in  2.00 seconds = 991.16 MB/sec
 Timing buffered disk reads:    8 MB in  3.08 seconds =   2.59 MB/sec

root@jpc j # hdparm -im1da254c3u1tTX69W /dev/hda

/dev/hda:
 setting fs readahead to 254
 setting 32-bit IO_support flag to 3
 setting multcount to 1
 setting unmaskirq to 1 (on)
 multcount    =  1 (on)
 IO_support   =  3 (32-bit w/sync)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 readahead    = 248 (on)

 Model=ST380011A, FwRev=3.06, SerialNo=3JV1HH9N
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=unknown, BuffSize=2048kB, MaxMultSect=16, MultSect=1
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=156301488
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

 * signifies the current active mode

 Timing cached reads:   1988 MB in  2.00 seconds = 993.16 MB/sec
 Timing buffered disk reads:   20 MB in  3.04 seconds =   6.59 MB/sec

The settings u posted didn't make a difference.
I tried setting the accoustic value manualy but that gave an error.
I also added the W value (not w) and that gave some improved results, but still much too slow.

P.S. notice how u write all the options against each other. also works in the /etc/conf.d/hdparm.

-------------------
EDIT
-------------------
I did a hdparm -I /dev/hda and this is what i found:
Code:
ATA device, with non-removable media
   Model Number:       ST380011A                               
   Serial Number:      3JV1HH9N           
   Firmware Revision:  3.06   
Standards:
   Used: ATA/ATAPI-6 T13 1410D revision 2
   Supported: 6 5 4 3
Configuration:
   Logical      max   current
   cylinders                   16383   65535
   heads      16   1
   sectors/track   63   63
   --
   CHS current addressable sectors:    4128705
   LBA    user addressable sectors:  156301488
   LBA48  user addressable sectors:  156301488
   device size with M = 1024*1024:       76319 MBytes
   device size with M = 1000*1000:       80026 MBytes (80 GB)
Capabilities:
   LBA, IORDY(can be disabled)
   bytes avail on r/w long: 4   Queue depth: 1
   Standby timer values: spec'd by Standard
   R/W multiple sector transfer: Max = 16   Current = 16
   Recommended acoustic management value: 128, current value: 0
   DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5
        Cycle time: min=120ns recommended=120ns
   PIO: pio0 pio1 pio2 pio3 pio4
        Cycle time: no flow control=240ns  IORDY flow control=120ns
Commands/features:
   Enabled   Supported:
      *   READ BUFFER cmd
      *   WRITE BUFFER cmd
      *   Host Protected Area feature set
      *   Look-ahead
      *   Write cache
      *   Power Management feature set
      Security Mode feature set
      *   SMART feature set
      *   FLUSH CACHE EXT command
      *   Mandatory FLUSH CACHE command
      *   Device Configuration Overlay feature set
      *   48-bit Address feature set
      SET MAX security extension
      *   DOWNLOAD MICROCODE cmd
      *   SMART self-test
      *   SMART error logging
Security:
      supported
   not   enabled
   not   locked
   not   frozen
   not   expired: security count
   not   supported: enhanced erase
HW reset results:
   CBLID- above Vih
   Device num = 0 determined by the jumper
Checksum: correct


and this is the part that worries me:
Code:
Configuration:
   Logical      max   current
   cylinders                   16383   65535
   heads      16   1
   sectors/track   63   63

as u can see, my drive has more cylinders than max and less heads.
I know to little about harddrives to tell if these values are hardware related or, software.
My guess is that the number of heads and cylinders are manufactured in.

P.P.S. sorry for the long post.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sat Apr 02, 2005 12:17 pm    Post subject: Reply with quote

This is a real puzzle..

Do you have both disks on the same IDE cable?
CD/DVDs on their own IDE cable?
And what chip have you configured in 'IDE chipset support'
Does dmesg show anything interesting about IDE after you boot up, like UDMA(100)?
In fact, why not post it.. and the IDE chipset part of your .config file

mvh

I have the same..
Code:
Configuration:
        Logical         max     current
        cylinders       16383   65535
        heads           16      1
        sectors/track   63      63

_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Sat Apr 02, 2005 4:46 pm    Post subject: Reply with quote

Cintra wrote:
Do you have both disks on the same IDE cable?
CD/DVDs on their own IDE cable?

Yes, the hdd's are on primary IDE and the cdd's are on the secondary.

Cintra wrote:
Does dmesg show anything interesting about IDE after you boot up, like UDMA(100)?

Yes it does:
Code:
jpc linux # dmesg | grep IDE
ACPI: PCI Interrupt Link [LIDE] (IRQs 3 4 5 6 7 9 10 11 12 14 15) *0, disabled.
Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
NFORCE2: IDE controller at PCI slot 0000:00:09.0
Probing IDE interface ide0...
Probing IDE interface ide1...
Probing IDE interface ide2...
Probing IDE interface ide3...
Probing IDE interface ide4...
Probing IDE interface ide5...

jpc linux # dmesg | grep UDMA
NFORCE2: 0000:00:09.0 (rev a2) UDMA133 controller
hda: 156301488 sectors (80026 MB) w/2048KiB Cache, CHS=16383/255/63, UDMA(100)
hdb: 240121728 sectors (122942 MB) w/2048KiB Cache, CHS=16383/255/63, UDMA(133)
hdc: ATAPI 40X DVD-ROM DVD-R CD-R/RW drive, 2000kB Cache, UDMA(33)
hdd: ATAPI 40X CD-ROM CD-R/RW drive, 4096kB Cache, UDMA(33)


Cintra wrote:
In fact, why not post it.. and the IDE chipset part of your .config file

# snip out .config
# ATA/ATAPI/MFM/RLL support
CONFIG_IDE=y
CONFIG_BLK_DEV_IDE=y
#
# Please see Documentation/ide.txt for help/info on IDE drives
#
# CONFIG_BLK_DEV_IDE_SATA is not set
# CONFIG_BLK_DEV_HD_IDE is not set
CONFIG_BLK_DEV_IDEDISK=y
CONFIG_IDEDISK_MULTI_MODE=y
CONFIG_BLK_DEV_IDECD=y
# CONFIG_BLK_DEV_IDETAPE is not set
# CONFIG_BLK_DEV_IDEFLOPPY is not set
# CONFIG_BLK_DEV_IDESCSI is not set
# CONFIG_IDE_TASK_IOCTL is not set

#
# IDE chipset support/bugfixes
#
CONFIG_IDE_GENERIC=y
CONFIG_BLK_DEV_CMD640=y
# CONFIG_BLK_DEV_CMD640_ENHANCED is not set
# CONFIG_BLK_DEV_IDEPNP is not set
CONFIG_BLK_DEV_IDEPCI=y
CONFIG_IDEPCI_SHARE_IRQ=y
# CONFIG_BLK_DEV_OFFBOARD is not set
CONFIG_BLK_DEV_GENERIC=y
# CONFIG_BLK_DEV_OPTI621 is not set
CONFIG_BLK_DEV_RZ1000=y
CONFIG_BLK_DEV_IDEDMA_PCI=y
# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
CONFIG_IDEDMA_PCI_AUTO=y
# CONFIG_IDEDMA_ONLYDISK is not set
# CONFIG_BLK_DEV_AEC62XX is not set
# CONFIG_BLK_DEV_ALI15X3 is not set
CONFIG_BLK_DEV_AMD74XX=y
# CONFIG_BLK_DEV_ATIIXP is not set
# CONFIG_BLK_DEV_CMD64X is not set
# CONFIG_BLK_DEV_TRIFLEX is not set
# CONFIG_BLK_DEV_CY82C693 is not set
# CONFIG_BLK_DEV_CS5520 is not set
# CONFIG_BLK_DEV_CS5530 is not set
# CONFIG_BLK_DEV_HPT34X is not set
# CONFIG_BLK_DEV_HPT366 is not set
# CONFIG_BLK_DEV_SC1200 is not set
CONFIG_BLK_DEV_PIIX=y
# CONFIG_BLK_DEV_NS87415 is not set
# CONFIG_BLK_DEV_PDC202XX_OLD is not set
# CONFIG_BLK_DEV_PDC202XX_NEW is not set
# CONFIG_BLK_DEV_SVWKS is not set
# CONFIG_BLK_DEV_SIIMAGE is not set
# CONFIG_BLK_DEV_SIS5513 is not set
# CONFIG_BLK_DEV_SLC90E66 is not set
# CONFIG_BLK_DEV_TRM290 is not set
CONFIG_BLK_DEV_VIA82CXXX=y
# CONFIG_IDE_ARM is not set
# CONFIG_IDE_CHIPSETS is not set
CONFIG_BLK_DEV_IDEDMA=y
# CONFIG_IDEDMA_IVB is not set
CONFIG_IDEDMA_AUTO=y
# CONFIG_BLK_DEV_HD is not set



Cintra wrote:
I have the same..
Code:
Configuration:
        Logical         max     current
        cylinders       16383   65535
        heads           16      1
        sectors/track   63      63
Ok thx, i thought i had partitioned my drive wrong or something, and i had to repartition.
I am releived.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sat Apr 02, 2005 7:25 pm    Post subject: Reply with quote

Ok thanks for the lists..
one thing I'm not sure of is which chipset your nforce uses. lspci will show it..
The other thing is that you have a lot of chipsets set as <y>.
It would be OK if unnecessary ones were <m>, but I don't think a number of <y> can be a good thing.
My asus p4pe uses the piix chipset.
When you do make any changes, you can set CONFIG_IDE_GENERIC=y to <>. Its for legacy ide and has been known to affect hdparm results & performance.
mvh

Edit: just found this tread https://forums.gentoo.org/viewtopic-t-183857-highlight-a7n8x+ide+chip.html with a bunch of your card's configs and they all had lots of <y> against various chipsets... ;-) ?
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Sat Apr 02, 2005 9:18 pm    Post subject: Reply with quote

i'm trying the settings posted on the link u supplied. I feel kinda stupid not finding that thread.
Here is a snip out if my lspci.
u say it tells me what driver to use, i don't know where to look/read.
Code:
jpc j # lspci -vvs 00:09.0
0000:00:09.0 IDE interface: nVidia Corporation nForce2 IDE (rev a2) (prog-if 8a [Master SecP PriP])
        Subsystem: ASUSTek Computer Inc.: Unknown device 0c11
        Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
        Status: Cap+ 6Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- >TAbort- <MAbort- >SERR- <PERR-
        Latency: 0 (750ns min, 250ns max)
        Region 4: I/O ports at f000 [size=16]
        Capabilities: [44] Power Management version 2
                  Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
                  Status: D0 PME-Enable- DSel=0 DScale=0 PME-
i guess the chipset is nforce2, but i don't see it in the kernel.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sun Apr 03, 2005 5:42 am    Post subject: Reply with quote

At last - if you run 'make xconfig' its easy to see a chipset called 'AMD and nvidia IDE support' or whatever, see https://forums.gentoo.org/viewtopic-t-311867-highlight-nforce2+ide+chipset.html

Thats the one where you particularly need a <y> and I assume its the one that shows up as
Code:
CONFIG_BLK_DEV_AMD74XX=y
which you have selected already.

The big question is whether the others, PIIX, VIA82CXXX, RZ1000, and CMD640 are affecting your performance...
I would suggest changing them to <m> first, and CONFIG_IDE_GENERIC=y to No.

If that doesn't help, it leaves the question of whether your drives are connected correctly, i.e. with the Master on the end of the cable?

Best of luck ;-)
mvh
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Sun Apr 03, 2005 10:01 am    Post subject: Reply with quote

Cintra wrote:
If that doesn't help, it leaves the question of whether your drives are connected correctly, i.e. with the Master on the end of the cable?
i always connect drives in that matter, but just in case i checked. and the drives are connected correctly.

i also removed all the drivers except the amd and nvidia one. (the one the thread mentioned) and i still get the same hdparm timings.

a few post back up i posted a snip ou of my dmesg witch said the drive was in UDMA(100). can that be the problem?
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sun Apr 03, 2005 10:59 am    Post subject: Reply with quote

Good, at least you are still in business ;-)
Ref
Quote:
a few post back up i posted a snip of my dmesg which said the drive was in UDMA(100). can that be the problem?

I don't think so. my Western Digitals are both UDMA100 and I have buffered disk reads on both around 43.30 MB/sec.
In my case adding PIIX=<y> made a huge difference, and changing my disk cable for one of the latest round ones pushed things up to where they are now. (thats why I'm interested in this post)

I'm afraid it looks as though you are stuck, unless someone else has a brainwave?
Alternatively you could consider investing in something like HighPoint's 'Rocket 133SB' pci adapter to run that disk on..

I'll keep my eyes open
mvh
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Sun Apr 03, 2005 4:56 pm    Post subject: Reply with quote

damm, that's a downer.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Sun Apr 03, 2005 6:03 pm    Post subject: Reply with quote

Haven't stopped thinking about your problem.. am currently doing a 'brainstorm' on it with freemind ;-)
Don't suppose you have a spare disk cable to try?
Oh, do the results stay exactly the same if you repeat the -t/T test a number of times in succession?
mvh
Edit What i/o scheduler are you using?
Code:
# cd /sys/block/hdb/queue/
p4pe queue # cat scheduler
noop anticipatory deadline [cfq]

_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Mon Apr 04, 2005 3:49 pm    Post subject: Reply with quote

Cintra wrote:
Don't suppose you have a spare disk cable to try?

I tried a brand new ATA133 cable, with same results
Cintra wrote:
Oh, do the results stay exactly the same if you repeat the -t/T test a number of times in succession?

No, the results vary between 1Mb/s and 10Mb/s most times below 3Mb/s
Cintra wrote:
Edit What i/o scheduler are you using?

yep like u, i use the "noop [anticipatory] deadline cfq"

P.S. i tried knoppix3.7, and it gave the same (poor) results.
P.S.S. i tried sisoft-sandra on Wxp, and it gave me a result of 45Mb/s, i don't know if it is comparable, but it clearly states the drive is capable of something more then 3Mb/s.
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Mon Apr 04, 2005 6:46 pm    Post subject: Reply with quote

Oh, the knoppix results are bad news.. however, there is one thing you could try which just might be affecting knoppix as well.
Just above the config option where you set your amd & nvidia chipset there's an option called
Code:
Force enable legacy 2.0.X HOSTS to use DMA

I have come across a number of 'fixes' where turning this off improved poor performance, and one case where turning it on (at the same time as making PIIX = <y>...) gave better performance - it was probably PIIX which improved things as it did for me. So if its <y> try turning it off, and vice versa.

EDIT: I see from your earlier post that it wasn't set then.. :-(

Another thing I wonder about is how full that disk is, i.e. % free space?
Also what filesystems are used on it's various partitions?
Its possible btw to do 'hdparm -tT /dev/hdax' for individual partitions (forget /boot tho').
Btw did you look in /var/log/messages for DMA messages?
And, finally (for now ;-) ) some guys improved results by stopping 'KDED media manager' in KDE Control center > KDE components > service manager. I've turned mine off too because it prints acres of messages if my packet-write disk is unmounted..
Hows that for some ideas :-)
Mvh
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
Sade
Guru
Guru


Joined: 22 Mar 2005
Posts: 406
Location: Netherlands - Eindhoven

PostPosted: Mon Apr 04, 2005 7:07 pm    Post subject: Reply with quote

Yeah nice list of suggestions, and....
I think we(read you) naled the problem, in fact there was no problem.

i did a:
Code:
hdparm -tT /dev/hda1 (boot-reiser3.6) and got the usual 1.99MB/s
hdparm -tT /dev/hda3 (root-reiser3.6) and got a wonderfull 52.43MB/s
hdparm -tT /dev/hda6 (home-ext3) and got a very nice 40.77MB/s


i can leap with joy, and i feel stupid at the same time, knowing that the whole time there was no problem!

P.S. Cintra, thx alot for your time and effort, although there was nothing to solve, i did learn a lot in the proces, and i thank u for that.
P.P.S if you'r still intrested in my %freespace, i'm more then willing to supply :).
_________________
| 1.6Ghz atom N270 | adopt an unanswerd post | a nice way to post a config file |
Back to top
View user's profile Send private message
Cintra
Advocate
Advocate


Joined: 03 Apr 2004
Posts: 2111
Location: Norway

PostPosted: Mon Apr 04, 2005 7:22 pm    Post subject: Reply with quote

Hey thats fantastic! :-D
Though I think its very strange your /dev/hda results got pulled down so
Another weird problem gets fixed by perseverance... thats the great thing with the tough ones, you learn a lot on the way.
And now I have the beginnings of a great mindmap for this kind of problem in the future
All the best Sade
regards
_________________
"I am not bound to please thee with my answers" W.S.
Back to top
View user's profile Send private message
EL CHAVO!
n00b
n00b


Joined: 24 May 2005
Posts: 2

PostPosted: Sun Aug 21, 2005 1:09 am    Post subject: Great Via chipset tip Reply with quote

Buchu wrote:
Pepoles who have an via chipset can enable in the kernel configuration HTP366 in ATA/IDE/... for user had and i don't need hdparm optimizations ( it doesn't change nothing for me)

sleepiiiiiiiiiiiiiiiiiiiiiiiiiiiiiing zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz


Thanks so much for this tip! Yeah, i know the post is old but I've gotta acknowledge when someone submits a good tip that was impossible to find on my own! My AZ11E mobo with via chipset would refuse to use dma, until I enabled the HPT36X in the kernel, who knew? EL CHAVO!
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8, 9  Next
Page 8 of 9

 
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