Re: PATCH: DMA enabled dump for ATA please test

From: Christian Brueffer <chris_at_unixpages.org>
Date: Thu, 12 Jan 2006 20:43:25 +0100
On Thu, Jan 12, 2006 at 12:39:03AM +0100, Søren Schmidt wrote:
> 
> Evening!
> 
> Originally I wanted to finally add dump functionality to ata-raid, but 
> quickly discovered that it would be a pain todo the old way.
> So I decided to reuse strategy for dumping by implementing polled mode 
> in ATA, and have dump simply call strategy.
> (This could be generalised for all drivers, and would make driver 
> writing easier).
> Now this also gave me the chance to use DMA on dumps, and that does do 
> wonders for dump speed.
> There is a few chipsets (ICH6+7 in AHCI mode and Marvell) that wont do 
> dumps yet not even PIO but they didn't work before either.
> 
> Anyhow this needs a bit testing on various HW out there before it hits 
> the tree, so please give it a go and let me know...
> 

I have tested this on my Thinkpad T41p with the following hardware:

atapci0: <Intel ICH4 UDMA100 controller> port 0x1f0-0x1f7,0x3f6,0x170-0x177,0x376,0x1860-0x186f at device 31.1 on pci0
ata0: <ATA channel 0> on atapci0
ata1: <ATA channel 1> on atapci0
[...]
ad0: 57231MB <HTS726060M9AT00 MH4OA6DA> at ata0-master UDMA100
acd0: CDRW <UJDA755yDVD/CDRW/1.72> at ata1-master UDMA33

With the patched kernel it stops after printing the acd0 line and spews
out the following:

acd0: req=0xc350f8d0 TEST_UNIT_READY semaphore timeout  !!  DANGER Will Robinson  !!

Do you need boot -v output? (I'll have to hook up a serial or firewire
console, that's why I haven't provided it already).

- Christian

-- 
Christian Brueffer	chris_at_unixpages.org	brueffer_at_FreeBSD.org
GPG Key:	 http://people.freebsd.org/~brueffer/brueffer.key.asc
GPG Fingerprint: A5C8 2099 19FF AACA F41B  B29B 6C76 178C A0ED 982D

Received on Thu Jan 12 2006 - 18:43:28 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:50 UTC