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 09/12] ata_piix: convert pata to new reset mechanism Tejun Heo
2006-01-24 8:05 ` [PATCH 12/12] ahci: add softreset 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 03/12] libata: new ->probe_reset operation Tejun Heo
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
2006-01-24 8:05 ` [PATCH 08/12] sata_sil24: add hardreset Tejun Heo
2006-01-24 8:05 ` [PATCH 07/12] sata_sil24: convert to new reset mechanism Tejun Heo
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 10/12] ata_piix: convert sata to new reset mechanism 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 11/12] ahci: convert to new reset mechanism Tejun Heo
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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).