All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jeff Garzik <jeff@garzik.org>
To: Tejun Heo <tj@kernel.org>, Mikael Pettersson <mikpe@it.uu.se>
Cc: Peter Favrholdt <linux-ide@how.dk>, linux-ide@vger.kernel.org
Subject: Re: [PATCH #upstream-fixes] sata_promise: request follow-up SRST
Date: Tue, 25 Nov 2008 12:27:40 -0500	[thread overview]
Message-ID: <492C358C.7030403@garzik.org> (raw)
In-Reply-To: <49263F80.2060607@kernel.org>

Tejun Heo wrote:
> sata_promise hardreset doesn't seem to be able to acquire the initial
> D2H Reg FIS after hardreset leading to hardreset timeouts.  Request
> follow-up SRST.
> 
>  http://article.gmane.org/gmane.linux.ide/36186
> 
> Signed-off-by: Tejun Heo <tj@kernel.org>
> ---
> Peter, can you please test this one too?  It's essentially the same
> code just slightly prettier.  Mikael, what do you think about this?
> 
> Jeff, please commit only after Peter's verification and Mikael's ACK.
> 
> Thanks.
> 
>  drivers/ata/sata_promise.c |    7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
> index ba9a257..917038a 100644
> --- a/drivers/ata/sata_promise.c
> +++ b/drivers/ata/sata_promise.c
> @@ -710,7 +710,12 @@ static int pdc_sata_hardreset(struct ata_link *link, unsigned int *class,
>  			      unsigned long deadline)
>  {
>  	pdc_reset_port(link->ap);
> -	return sata_sff_hardreset(link, class, deadline);
> +
> +	/* sata_promise can't reliably acquire the first D2H Reg FIS
> +	 * after hardreset.  Do non-waiting hardreset and request
> +	 * follow-up SRST.
> +	 */
> +	return sata_std_hardreset(link, class, deadline);

hrm....   at this point we have deviated massively from the standard 
Promise method of hard reset...

* set PMP port
* make hotplug irqs
* reset port# in PDC_GLOBAL_CTL
* pdc_reset_port()
* reset FPDMA -- probably a good idea even if not doing FPDMA
* clear SATA Error register (0xffffffff)
* clear errors-reported-from-link-layer register
* wait standard length of time for SATA connection
* clear hotplug bits
* set PMP port

The PDC_GLOBAL_CTL bitbang is the most notable missing element in the 
hard reset path, though we also miss clearing an apparently-important 
error register as well.  FPDMA reset would be a good idea I think, even 
if not in use, mainly for paranoia's sake and because that's what 
Promise's driver does.

	Jeff




  parent reply	other threads:[~2008-11-25 17:27 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-11-13 21:21 FYI: BUG in SATA Promise 300 TX4 (2.6.24 - 2.6.27-3) w/Linux Linda Walsh
2008-11-16  6:04 ` Tejun Heo
2008-11-16 11:08   ` Mikael Pettersson
2008-11-16 14:24     ` Tejun Heo
2008-11-16 16:48     ` Brad Campbell
2008-11-17  2:01       ` Tejun Heo
2008-11-16 17:34     ` Peter Favrholdt
2008-11-16 17:39       ` Peter Favrholdt
2008-11-17  2:01         ` Tejun Heo
2008-11-17 11:47           ` Peter Favrholdt
2008-11-18  1:11             ` Tejun Heo
2008-11-18 18:03               ` Peter Favrholdt
2008-11-19  1:55                 ` Tejun Heo
2008-11-20 10:22                   ` Peter Favrholdt
2008-11-20 11:10                     ` Mikael Pettersson
2008-11-21  4:42                       ` Tejun Heo
2008-11-21  4:56                       ` [PATCH #upstream-fixes] sata_promise: request follow-up SRST Tejun Heo
2008-11-22 16:30                         ` Mikael Pettersson
2008-11-23 22:38                         ` Peter Favrholdt
2008-11-25 13:00                         ` Peter Favrholdt
2008-11-26  2:46                           ` Tejun Heo
2008-11-26  8:12                             ` Peter Favrholdt
2008-11-26 23:07                               ` Peter Favrholdt
2008-11-25 17:27                         ` Jeff Garzik [this message]
2008-11-25 21:17                           ` Mikael Pettersson
2008-11-29 21:50                           ` Mikael Pettersson
2008-11-30 15:06                             ` Peter Favrholdt
2009-02-10  4:30                             ` Jeff Garzik
2009-02-10 17:28                               ` Mikael Pettersson
2009-02-10 21:13                                 ` Jeff Garzik
2009-02-23 12:17                                   ` [PATCH #upstream-fixes] sata_promise: request follow-up SRST - it works Peter Favrholdt
  -- strict thread matches above, loose matches on Subject: below --
2009-09-25  5:24 [PATCH #upstream-fixes] sata_promise: request follow-up SRST Walter Hutchins
2009-09-26  2:57 ` Robert Hancock
2009-09-26 18:42   ` Walter Hutchins

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=492C358C.7030403@garzik.org \
    --to=jeff@garzik.org \
    --cc=linux-ide@how.dk \
    --cc=linux-ide@vger.kernel.org \
    --cc=mikpe@it.uu.se \
    --cc=tj@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.