Hi, i note a recent post regarding a similar error, however my
circumstances are different.
I dont get an error writting, only reading, its on a raid0 partition
under hpt366 controllers.
To test writing i did "time dd=/dev/zero of=test.out bs=4000
count=50000"
The result was
real 0m34.312s
user 0m0.780s
sys 0m32.640s
Which means i was writting at about 56MB/s, no errors, i was impressed
by this :)
To test reading i did
dd if=test.out of=/dev/null bs=4000
and also dd if=/dev/md0 of=/dev/null
Initially i get
hdk: timeout waiting for DMA
ide_dmaproc: chipset supported ide_dma_timeout func only:14
followed clossly by a re-occuring
ide_dmaproc: chipset supported ide_dma_lostirq func only: 13
hde: lost interrupt
hdg: lost interrupt
To bypass raid i did multiple
dd if=/dev/hde1 of=/dev/null and dd if=/dev/hdg1 of=/dev/null in
seperate windows and monitored performace using top in another windows.
each of the two dd precesses was using about 38% and kswapd 70%, 28%
idle (dual processor)
if i add another process, dd if=/dev/hdi1 of=/dev/null then kswapd jumps
to 88% and i start getting the ide_dma_lostirq funct only:13 errors, one
dd maintains its 30% loads, the other two disapear.
My guess is kswapd (pri 18) starved dd (pri 6) for cpu time leading to
the timeouts.
Should i expect kswapd to be using 88% cpu resources ?
I only have 32MB swap space, but none apears to be used.
Am i doing anything obviously wrong here ?
I dont understand how i get no error writing, but i do reading.
Im using a bp6 dual 433 celeron, 128MB ram, to provide me with two extra
udma channels i have an abit hotrod pci hpt366 controller card, giving
me 6 ide channels.
I have one 6.4GB quantum drive that i boot from that contains my distro.
I have a raid0 array comprised of 4 ide drives each on there own hpt366
channel.
I have two 20.5GB IBM-DPTA-372050 plugged into the hpt366 channels on my
motherboard, the hpt366 detects them as udma4
one 20.5GB Quantum KX that hpt366 detects as udma4
one 18.5GB Quantum KA that hpt366 detects as udma2
/proc/dma only conatins
4: cascade
/proc/interrupts contains
CPU0 CPU1
0: 5949 2760 IO-APIC-edge timer
1: 127 56 IO-APIC-edge keyboard
2: 0 0 XT-PIC cascade
8: 0 0 IO-APIC-edge rtc
12: 0 0 IO-APIC-edge PS/2 Mouse
13: 1 0 XT-PIC fpu
14: 95607 28123 IO-APIC-edge ide0
17: 0 2 IO-APIC-level eth0
18: 1999 10879 IO-APIC-level ide4, ide5
19: 4 6 IO-APIC-level ide2, ide3
NMI: 8638 8638
LOC: 8616 8608
ERR: 0
NOTE : It took me ages to get the onboard hpt366 and pci hpt366
controllers to use different interupts.
Thanks
Glenn McGrath
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Apr 23 2000 - 21:00:14 EST