The system normally runs 2.1.107. When initially built ckraid
said about 6MB/sec with a single drive, with two drives on one
controller it said about 3.5MB/sec, with two drives on different
controllers (the normal configuration) ckraid said about 4.5MB/s.
Since I got in earlier than everyone else this morning I decided
to bounce it to 2.1.115. Same hardware config, same kernel config,
just a different version. Ckraid said 1.5MB/sec. *1.5*. It never
peaked above 2MB/sec. Killed it. Rebooted 2.1.107. Back to 4.5MB/sec.
Hmmm... It just occurred to me that 2.1.107 might have been a
non-SMP build and 2.1.115 might be SMP (single processor PII/400
machine). Even so - a *third* the I/O throughput???
I don't think the eata_dma driver has changed between 107 and 115
and being bus master DMA the only real possibilities are the request
queue or the interrupt handler (with the DPT that's one interrupt
per request to signal completion - the hardware does the work).
_Something_ seems to have gone horribly wrong here. Anyone any
guesses what?
Mike
-- .----------------------------------------------------------------------. | Mike Jagdis | Internet: mailto:mike@roan.co.uk | | Roan Technology Ltd. | | | 54A Peach Street, Wokingham | Telephone: +44 118 989 0403 | | RG40 1XG, ENGLAND | Fax: +44 118 989 1195 | `----------------------------------------------------------------------'
- 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.altern.org/andrebalsa/doc/lkml-faq.html