From: Jeff Garzik <jgarzik@pobox.com>
To: Alan <alan@lxorguk.ukuu.org.uk>
Cc: Linus Torvalds <torvalds@osdl.org>,
Alessandro Suardi <alessandro.suardi@gmail.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] libata: fix combined mode (was Re: Happy New Year (and v2.6.20-rc3 released))
Date: Tue, 02 Jan 2007 16:29:00 -0500 [thread overview]
Message-ID: <459ACE9C.7020107@pobox.com> (raw)
In-Reply-To: <20070102212701.4b4535cf@localhost.localdomain>
Alan wrote:
>>> This is a silly complaint because the SFF layer in libata doesn't handle
>>> this case yet anyway.
>> Yes, it's "silly" people people use configurations you find inconvenient.
>>
>> At least one embedded x86 case cares, that I know of. They only needed
>> to make two minor changes to make it work.
>
> *It is not part of 2.6.20*
>
>> The code no long reserves resources for the "extra" PCI BAR that often
>> exists on PCI controllers regardless of legacy/native mode. Previously,
>> the code called pci_request_regions() to reserve ALL regions attached to
>> the PCI device.
>
> We use BAR5 on two devices in legacy mode. Both of those reserve all the
> other resources.
Translation: You want to hand-wave away an obvious regression that YOU
have created with your fix-to-a-fix.
> We can fix BAR5 in .21 when all the combined mode crap
> goes away.
Translation: Problems disappear in 2.6.21 because Jeff will revert the
code I touched to its previous state -- always calling
pci_request_regions() -- and all the problems I introduced by avoiding
pci_request_regions() will go away.
Why INTRODUCE these 2.6.20 Alan-isms, if they are going away in 2.6.21?
>> You have suddenly decided that it's OK to --not reserve at all-- these
>> additional regions.
>
> It's not ideal - but it is perfectly sufficient for 2.6.20
>
>> Proof: The AHCI PCI BAR (#5, zero-based) is clearly NOT reserved, even
>> though we talk to it, in piix_disable_ahci() of ata_piix.c.
>
> We always claim the other BARs so catch a collision.
Where? AFAICS, it is crystal clear the behavior:
* Prior to your patch, ata_piix in legacy mode calls
pci_request_regions() to intentionally reserve ALL regions on the PCI
device.
* After your patch, the code explicitly calls pci_request_region() for
BARs 0-4, but never for BAR5.
Another driver is now free to claim a PCI BAR, and start running the
hardware in AHCI mode, whee!
Jeff
next prev parent reply other threads:[~2007-01-02 21:29 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-01-01 1:19 Happy New Year (and v2.6.20-rc3 released) Linus Torvalds
2007-01-01 1:39 ` Gene Heskett
2007-01-01 2:40 ` Alexey Dobriyan
2007-01-01 7:22 ` Cyrill V. Gorcnov
[not found] ` <4598BC0F.6070003@vc.cvut.cz>
2007-01-01 9:41 ` Cyrill V. Gorcnov
2007-01-02 2:19 ` H. Peter Anvin
2007-01-01 11:29 ` Arkadiusz Miskiewicz
2007-01-01 13:28 ` Alessandro Suardi
2007-01-01 20:13 ` Linus Torvalds
2007-01-01 20:49 ` Jeff Garzik
2007-01-01 21:31 ` Alan
2007-01-01 23:34 ` Linus Torvalds
2007-01-02 2:32 ` Theodore Tso
2007-01-02 11:58 ` [PATCH] libata: fix combined mode (was Re: Happy New Year (and v2.6.20-rc3 released)) Alan
2007-01-02 12:07 ` Alessandro Suardi
2007-01-02 14:00 ` Alan
2007-01-02 16:12 ` Theodore Tso
2007-01-02 21:00 ` Jeff Garzik
2007-01-02 21:27 ` Alan
2007-01-02 21:29 ` Jeff Garzik [this message]
2007-01-02 21:32 ` Jeff Garzik
2007-01-02 22:48 ` Alan
2007-01-02 22:45 ` Alan
2007-01-02 23:01 ` Jeff Garzik
2007-01-02 23:27 ` Alan
2007-01-02 23:43 ` Jeff Garzik
2007-01-03 0:36 ` Alan
2007-01-03 1:02 ` Jeff Garzik
2007-01-02 23:54 ` Jeff Garzik
2007-01-03 0:39 ` Alan
2007-01-02 23:04 ` Jeff Garzik
2007-01-03 19:59 ` Steve Wise
2007-01-01 21:26 ` Happy New Year (and v2.6.20-rc3 released) Alan
2007-01-02 16:09 ` Linus Torvalds
2007-01-01 15:52 ` Olaf Hering
2007-01-02 19:16 ` 2.6.20-rc3: known unfixed regressions Adrian Bunk
2007-01-03 18:15 ` Steve Youngs
2007-01-03 20:03 ` Adrian Bunk
2007-01-02 19:24 ` 2.6.20-rc3: known regressions with patches available (part 1) Adrian Bunk
2007-01-02 19:26 ` Jens Axboe
2007-01-02 19:34 ` Adrian Bunk
2007-01-02 20:45 ` Rene Herman
2007-01-02 20:51 ` Jens Axboe
2007-01-02 19:30 ` 2.6.20-rc3: known regressions with patches available (part 2) Adrian Bunk
2007-01-03 19:06 ` [2.6.20-rc3] INFO: possible recursive locking detected (was: Happy New Year (and v2.6.20-rc3 released)) Tilman Schmidt
2007-01-03 20:59 ` 2.6.20-rc3: known unfixed regressions (v2) Adrian Bunk
2007-01-04 13:23 ` Horst H. von Brand
2007-01-03 21:04 ` 2.6.20-rc3: known regressions with patches (v2) Adrian Bunk
2007-01-03 21:30 ` Jiri Kosina
2007-01-03 21:32 ` Jiri Kosina
2007-01-04 17:46 ` 2.6.20-rc3: known unfixed regressions (v3) Adrian Bunk
2007-01-04 18:13 ` Matthew Wilcox
2007-01-05 0:25 ` Horst H. von Brand
2007-01-06 21:04 ` 2.6.20-rc3: known unfixed regressions (v4) Adrian Bunk
2007-01-06 22:06 ` Brice Goglin
2007-01-07 23:49 ` Adrian Bunk
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=459ACE9C.7020107@pobox.com \
--to=jgarzik@pobox.com \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=alessandro.suardi@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=torvalds@osdl.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