OoO Vers la fin de l'après-midi du mardi 19 mars 2002, vers 16:33,
Jakob Østergaard <jakob@unthought.net> disait:
>> I have an Abit KR7A-RAID with an embedded HPT372 IDE controller. I
>> have put one disk on each controller. These two disks support UDMA
>> 5. I can transfer (dd of=/dev/null) at full speed (40 Mb) from the two
>> disks at the same time without any problem.
>>
>> However, when I build the RAID array (software raid 0 with md, not
>> with hptraid which works fine but seems to be too slow), I immediatly
>> get a DMA error. raidstop does exactly the same thing.
>>
>> I think there is some tricks in md driver which makes things weird
>> since I have absolutely no problem without md driver, even when I
>> stress the disks but I get an immediate error when I use md.
>>
>> Any clue ?
> Try stressing *both* drives without using MD.
Well, stressing both drives doesn't trigger a DMA error : both drives
can happily transfer to /dev/null at 40 MB/s at the same time.
> MD doesn't pull any special IDE tricks - it should not matter at all
> whether you do a "dd if=/dev/md0 of=/dev/null" or a
> "dd if=/dev/hda of=/dev/null & dd if=/dev/hdc of=/dev/null"
What could I do to "trace" into md driver to discover what triggers
this DMA error ?
--
panic("Detected a card I can't drive - whoops\n");
2.2.16 /usr/src/linux/drivers/net/daynaport.c
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
