ATA-100 disk under 2.2.18 -- horrible performance
I have an IBM ATA-100 30G harddrive, attached to a ASUS CUSL2-C (Intel 815EP-based) mobo. The kernel is not even using the drive in UDMA mode. The performance is *horrible*: every time I copy to the disk the CPU usage goes up all the way and I'm seeing things like load average: 5.47, 5.02, 3.27, which is only caused by disk access. ,[ dmesg ] | PCI_IDE: unknown IDE controller on PCI bus 00 device f9, VID=8086, DID=244b | PCI_IDE: not 100% native mode: will probe irqs later | ide0: BM-DMA at 0xa800-0xa807, BIOS settings: hda:DMA, hdb:pio | ide1: BM-DMA at 0xa808-0xa80f, BIOS settings: hdc:DMA, hdd:pio | hda: QUANTUM FIREBALLP LM10.2, ATA DISK drive | hdc: IBM-DTLA-307030, ATA DISK drive | ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 | ide1 at 0x170-0x177,0x376 on irq 15 | hda: QUANTUM FIREBALLP LM10.2, 9797MB w/1900kB Cache, CHS=19906/16/63, UDMA | hdc: IBM-DTLA-307030, 29314MB w/1916kB Cache, CHS=59560/16/63 ` As you can see, the other HD (which is an ATA-66 Quantum) is at least used in UDMA mode. What is the reason the kernel dislikes the drive so much? ;^) Do you suppose I've missed some option in kernel configuration? Any other possible reasons? Could the fact that the whole drive is one 30G ext2fs partition have anything to do with the slowness? Many thanks for any suggestions, -- Arcady Genkin Don't read everything you believe.
Re: ATA-100 disk under 2.2.18 -- horrible performance
Arcady Genkin wrote: I have an IBM ATA-100 30G harddrive, attached to a ASUS CUSL2-C (Intel 815EP-based) mobo. The kernel is not even using the drive in UDMA mode. The performance is *horrible*: every time I copy to the disk the CPU usage goes up all the way and I'm seeing things like load average: 5.47, 5.02, 3.27, which is only caused by disk access. ,[ dmesg ] | PCI_IDE: unknown IDE controller on PCI bus 00 device f9, VID=8086, DID=244b | PCI_IDE: not 100% native mode: will probe irqs later looks like the system doesn't know what kind of controller it is try patching from www.linux-ide.org nate -- ::: ICQ: 75132336 http://www.aphroland.org/ http://www.linuxpowered.net/ [EMAIL PROTECTED]
Re: ATA-100 disk under 2.2.18 -- horrible performance
On 07 Jan 2001 03:58:45 -0500, Arcady Genkin said: I have an IBM ATA-100 30G harddrive, attached to a ASUS CUSL2-C (Intel 815EP-based) mobo. The kernel is not even using the drive in UDMA mode. The performance is *horrible*: every time I copy to the disk the CPU usage goes up all the way and I'm seeing things like load average: 5.47, 5.02, 3.27, which is only caused by disk access. ,[ dmesg ] | PCI_IDE: unknown IDE controller on PCI bus 00 device f9, VID=8086, DID=244b | PCI_IDE: not 100% native mode: will probe irqs later | ide0: BM-DMA at 0xa800-0xa807, BIOS settings: hda:DMA, hdb:pio | ide1: BM-DMA at 0xa808-0xa80f, BIOS settings: hdc:DMA, hdd:pio | hda: QUANTUM FIREBALLP LM10.2, ATA DISK drive | hdc: IBM-DTLA-307030, ATA DISK drive | ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 | ide1 at 0x170-0x177,0x376 on irq 15 | hda: QUANTUM FIREBALLP LM10.2, 9797MB w/1900kB Cache, CHS=19906/16/63, UDMA | hdc: IBM-DTLA-307030, 29314MB w/1916kB Cache, CHS=59560/16/63 ` As you can see, the other HD (which is an ATA-66 Quantum) is at least used in UDMA mode. What is the reason the kernel dislikes the drive so much? ;^) Do you suppose I've missed some option in kernel configuration? Any other possible reasons? Could the fact that the whole drive is one 30G ext2fs partition have anything to do with the slowness? Well i am using the identical ibm drive on an abit mb without ATA100 support. In my kernel config i set IDEDMA_AUTO=y so the kernel automatically enables DMA. When i run hdparm -i it shows: UDMA modes: mode0 mode1 mode2 mode3 *mode4 mode5 And hdparm -t shows disk reads about 30MB/sec. Maybe you have to manually enable DMA for this drive.
Re: ATA-100 disk under 2.2.18 -- horrible performance
Nate Amsden [EMAIL PROTECTED] writes: ,[ dmesg ] | PCI_IDE: unknown IDE controller on PCI bus 00 device f9, VID=8086, DID=244b | PCI_IDE: not 100% native mode: will probe irqs later looks like the system doesn't know what kind of controller it is try patching from www.linux-ide.org Woohooo! ,[ dmesg ] | PIIX4: IDE controller on PCI bus 00 dev f9 | PIIX4: chipset revision 1 | ... | hdc: IBM-DTLA-307030, 29314MB w/1916kB Cache, CHS=59560/16/63, UDMA(100) ` Thanks!!! The only question is whether I used the right patch. There seem to be two different kinds of patches (judging from the filename format), but no documentation about the difference between them. Any idea what the difference would be between: ide.2.2.18.1221.patch.gz and ide.2.2.18-27.all.20001208.patch.gz besides that the first one is newer? I used the first patch. Many thanks, -- Arcady Genkin Don't read everything you believe.
Re: ATA-100 disk under 2.2.18 -- horrible performance
Arcady Genkin wrote: | PIIX4: IDE controller on PCI bus 00 dev f9 | PIIX4: chipset revision 1 | ... | hdc: IBM-DTLA-307030, 29314MB w/1916kB Cache, CHS=59560/16/63, UDMA(100) ` Thanks!!! cool! glad it works :) The only question is whether I used the right patch. There seem to be two different kinds of patches (judging from the filename format), but no documentation about the difference between them. Any idea what the difference would be between: ide.2.2.18.1221.patch.gz and ide.2.2.18-27.all.20001208.patch.gz besides that the first one is newer? I used the first patch. i'm not sure, i would of used the first one too i think. check the filesize differences ? but if it patched cleanly and seems to be working i'd say your set to go .. nate -- ::: ICQ: 75132336 http://www.aphroland.org/ http://www.linuxpowered.net/ [EMAIL PROTECTED]