From: Jeff Garzik <jgarzik@pobox.com>
To: Tejun Heo <htejun@gmail.com>
Cc: linux-ide@vger.kernel.org, albertcc@tw.ibm.com
Subject: Re: [PATCH 06/12] sata_sil: convert to new reset mechanism
Date: Sat, 28 Jan 2006 13:09:15 -0500 [thread overview]
Message-ID: <43DBB34B.1050302@pobox.com> (raw)
In-Reply-To: <1138089922913-git-send-email-htejun@gmail.com>
Tejun Heo wrote:
> Convert sata_sil to use new reset mechanism. sata_sil is fairly
> generic and can directly use std routine.
>
> Signed-off-by: Tejun Heo <htejun@gmail.com>
Major NAK.
The placement of every hardware register read/write has been very
carefully laid out, and your reset hardware re-orders a lot of that.
Code re-ordering is fine. Changing the order of I/O operations is not.
With this patch, sata_sil fails to wake up and initialize the SATA phy!
Proper code transformation is art. You're doing really well, but this
was a bit of a hiccup. You should have presented a series of patches
which changed over to your new ata_drive_probe_reset() stuff, while -not
changing the hardware operations at all-. Like a mathematical solution,
your code is then provably correct. If there is a regression, then you
know it is a driver bug, and not a problem with hardware acting weird.
If the driver can access the SATA PHY registers, the very first thing it
should usually do is the code in sata_std_hardreset()... everything
except the 0x301 value write. Or in other words, follow the code in
__sata_phy_reset() PRECISELY (presuming that ATA_FLAG_SATA_RESET is not
set).
sata_sil hardware initialization order and behavior should not change at
all [though certainly you will change it sometime in the future].
So... I will drop all other patches from you, let you resync with the
latest stuff in libata-dev.git#upstream (NOTE: renamed from
upstream-2.6.17 to upstream), and wait for your next barrage of patches ;-)
Jeff
next prev parent reply other threads:[~2006-01-28 18:09 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-01-24 8:05 [PATCHSET] libata: new reset mechanism, take#2 Tejun Heo
2006-01-24 8:05 ` [PATCH 01/12] libata: export ata_busy_sleep Tejun Heo
2006-01-27 4:18 ` Jeff Garzik
2006-01-24 8:05 ` [PATCH 04/12] libata: implement ata_drive_probe_reset() Tejun Heo
2006-01-28 1:58 ` Jeff Garzik
2006-01-24 8:05 ` [PATCH 03/12] libata: new ->probe_reset operation Tejun Heo
2006-01-24 8:05 ` [PATCH 09/12] ata_piix: convert pata to new reset mechanism Tejun Heo
2006-01-24 8:05 ` [PATCH 07/12] sata_sil24: convert " Tejun Heo
2006-01-24 8:05 ` [PATCH 12/12] ahci: add softreset Tejun Heo
2006-01-24 8:05 ` [PATCH 08/12] sata_sil24: add hardreset Tejun Heo
2006-01-24 8:05 ` [PATCH 05/12] libata: implement standard reset component operations and ->probe_reset Tejun Heo
2006-01-28 1:58 ` Jeff Garzik
2006-01-24 8:05 ` [PATCH 11/12] ahci: convert to new reset mechanism Tejun Heo
2006-01-24 8:05 ` [PATCH 10/12] ata_piix: convert sata " Tejun Heo
2006-01-24 8:05 ` [PATCH 02/12] libata: modify ata_dev_try_classify Tejun Heo
2006-01-27 4:20 ` Jeff Garzik
2006-01-24 8:05 ` [PATCH 06/12] sata_sil: convert to new reset mechanism Tejun Heo
2006-01-28 18:09 ` Jeff Garzik [this message]
2006-01-28 18:27 ` Jeff Garzik
2006-01-28 23:17 ` Tejun
2006-01-29 2:41 ` 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=43DBB34B.1050302@pobox.com \
--to=jgarzik@pobox.com \
--cc=albertcc@tw.ibm.com \
--cc=htejun@gmail.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 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.