CD and 2.6 - more data fewer answers
From: Bill Davidsen
Date: Thu Dec 16 2004 - 15:07:18 EST
I have a 2.6.7-rc1-mm1 test system with both IDE CD and USB CD-R. Both
work fine as I normally use them, the CD-R is my primary backup device.
With all the discussion of DMA or not DMA, I decided to try some
reads... I tried readcd using /dev/hdc and the user was 60%, sys 40%, no
idle. Using ATAPI:0,0,0 I saw user 60%, sys 2% idle 38%. Then using the
USB device I saw user 4%, sys 9%, idle 12%, wio 75%. The USB is much
faster, so this looked right, and it appeared that on data reads DMA was
being used. I have no idea where all the user CPU was going.
Then I turned on -c2scan to read 2352 byte sectors without the final
error correct. With /dev/hdc I saw use 42%, sys 68%. Using ATAPI:0,0,0 I
saw user 47%, sys 53%. And using the USB device the c2scan ended without
reading any data.
The difference between /dev/hdc and ATAPI:0,0,0 is small, although
consistent. The access by SCSI number failed with the USB drive, and
/dev/scd0 resulted in "readcd: Invalid argument. Cannot get SCSI I/O
buffer" when tried (yes, as root as well).
I make no claim that this sheds light on the question, just a sprinkle
of data to provide flavor. I'll bring in an audio CD tomorrow and see
what trying to rip that does, unless someone beats me to it.
-bill davidsen (davidsen@xxxxxxx)
"The secret to procrastination is to put things off until the
last possible moment - but no longer" -me
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/