Linux ATA/IDE development
 help / color / mirror / Atom feed
From: Pat LaVarre <p.lavarre@ieee.org>
To: Jeff Garzik <jgarzik@pobox.com>
Cc: linux-ide@vger.kernel.org
Subject: Re: [PATCH] libata atapi work #2.1
Date: 16 May 2004 09:39:52 -0600	[thread overview]
Message-ID: <1084721991.3085.6.camel@patibmrh9> (raw)

Jeff G:

> Subject: Re: SATA ATAPI work in progress
> Date: 2004-05-15 16:32:56
> ...
> Seems like we are pretty close.  Once I add
> the code to issue a REQUEST SENSE, I bet
> things start working.

I agree!  I'm delighted to report,

I now see SATA stream up past 1.5 GB/min via ioctl SG_IO read & write of
the disc blocks, after I apply together all four of the patches you
first posted as:

> Subject: [PATCH] libata update 20040515-1
> Subject: [PATCH] libata atapi work #2.1
> Subject: [PATCH] libata DMADIR support
> Subject: [PATCH] Re: SATA ATAPI work in progress

Heads up, I was myself slow to appreciate:

Yes still in -bk3 after patch.2.1 we need to apply that fourth patch
i.e. what I had termed patch.7 i.e. ata_host_intr += ATA_PROT_ATAPI_DMA.

> Once I add the code to issue a REQUEST SENSE, ...

Also, of course, to avoid dying for the lack of auto sense, I have to
duck all errors.

Once I died by mistyping an LBA, off by one nybble.  Repeatedly I die
except when I remember to clear the x 6 29 Reset and x 6 28 GoneReady
unit attentions, via an abuse which works here i.e. repeatedly
unsolicited "REQUEST SENSE".  I notice in particular that, with this
device, modprobe seemingly provokes x 6 29 Reset.

Pat LaVarre


$ # demo of 1.5 GB/min, in or out
$
$ plscsi /dev/sg0 -i x12 -x "03 00:00:00 12 00"
$ plscsi /dev/sg0 -i x12 -x "03 00:00:00 12 00"
$
$ time pldd of if=/dev/sg0 bs=64k sbs=2k skip=0 count=1600
 
real    0m4.134s
user    0m0.002s
sys     0m0.155s
$
$ dc -e '64 1024 * 1600 * 4.134 / p 60 * p'
25364683
1521880980
$
$ time pldd of=/dev/sg0 if bs=64k sbs=2k seek=0 count=1600
 
real    0m4.125s
user    0m0.001s
sys     0m0.150s
$
$ dc -e '64 1024 * 1600 * 4.125 / p 60 * p'
25420024
1525201440
$


$ # demo of reset implicit in modprobe
$
$ plscsi /dev/sg0 -i x12 -x "03 00:00:00 12 00" -v
x 00000000 03 00:00:00 12 00 .. .. .. .. .. .. .. .. .. .. "C@@@R@"
x 00000000 70:00:00:00 00:00:00:18 00:00:00:00 00:00:00:00 "p@@@@@@X@@@@@@@@"
x 00000010 00:00 .. .. .. .. .. .. .. .. .. .. .. .. .. .. "@@"
// 0 = plscsi.main exit int
$
$ sudo modprobe -r ata-piix
$ sudo modprobe ata-piix
$
$ plscsi /dev/sg0 -i x12 -x "03 00:00:00 12 00" -v
x 00000000 03 00:00:00 12 00 .. .. .. .. .. .. .. .. .. .. "C@@@R@"
x 00000000 70:00:06:00 00:00:00:18 00:00:00:00 29:00:00:00 "p@F@@@@X@@@@)@@@"
x 00000010 00:00 .. .. .. .. .. .. .. .. .. .. .. .. .. .. "@@"
// 0 = plscsi.main exit int
$



             reply	other threads:[~2004-05-16 15:40 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-05-16 15:39 Pat LaVarre [this message]
2004-05-16 23:20 ` [PATCH] libata atapi work #2.1 Jeff Garzik
2004-05-17 14:56   ` Pat LaVarre
2004-05-17 17:57     ` Jeff Garzik
2004-05-17 19:24       ` Pat LaVarre
2004-05-18  1:48     ` Pat LaVarre
2004-05-18 21:10       ` Pat LaVarre
  -- strict thread matches above, loose matches on Subject: below --
2004-05-15 22:12 Jeff Garzik

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1084721991.3085.6.camel@patibmrh9 \
    --to=p.lavarre@ieee.org \
    --cc=jgarzik@pobox.com \
    --cc=linux-ide@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox