public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Jeff Garzik <jeff@garzik.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Arjan van de Ven <arjan@infradead.org>,
	linux-kernel@vger.kernel.org, gregkh@suse.de,
	linux-pci@atrey.karlin.mff.cuni.cz,
	Benjamin Herrenschmidt <benh@kernel.crashing.org>
Subject: Re: [patch] Make MMCONFIG space (extended PCI config space) a driver opt-in issue
Date: Sat, 22 Dec 2007 23:13:46 -0500	[thread overview]
Message-ID: <476DE07A.4000204@garzik.org> (raw)
In-Reply-To: <alpine.LFD.0.9999.0712221939570.21557@woody.linux-foundation.org>

Linus Torvalds wrote:
> On Sat, 22 Dec 2007, Jeff Garzik wrote:
>> Regardless of whether a driver is loaded or not, you may NEED to see extended
>> capabilities.  The system may NEED to see those capabilities just to parse
>> them for sane operation.
> 
> And that's just not true.
> 
> I don't know why you even claim so. There is absolutely *zero* need for 
> MMCONFIG for 100% of old hardware, and old hardware is exactly the 
> problem. The hardware that people will run next year isn't the issue.

I'm not claiming anything about old hardware.

It should be self-evident that mmconfig doesn't work on old hardware, is 
not needed on old hardware, should not be turned on for old hardware, 
and in general should never disturb old hardware.


> There may be an absolute need for MMCONFIG for future machines and future 
> drivers, but there is not a single machine in existence today that really 
> depends on it. You're just making things up. At worst, you lose some 
> not-so-important error reporting, no more.
> 
> So don't go spreading obvious untruths,


The facts as they exist today:

1) Existing 256-byte config space devices have been known put 
capabilities in the high end (>= 0xc8) of config space.

2) It is legal for PCI-Express to put capabilities anywhere in PCI 
config space, including extended config space.  (I hope our PCI cap 
walking code checks for overruns...)

3) Most new machines ship with PCI-Express devices with extended config 
space.

Therefore it is provable /possible/, and is indeed logical to conclude 
that capabilities in extended config space will follow the same pattern 
that existing hw designers have been following...  but only once the 
current OS's have stable extended-config-space support.

Maybe that day will never come, but it is nonetheless quite possible 
without today's PCI Express spec for this to happen.

We must handle this case, even if mmconfig is COMPLETELY DISABLED (i.e. 
normal mode today).

	Jeff




  parent reply	other threads:[~2007-12-23  4:13 UTC|newest]

Thread overview: 61+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-12-22 12:31 [patch] Make MMCONFIG space (extended PCI config space) a driver opt-in issue Arjan van de Ven
2007-12-22 12:35 ` [patch] opt the sky2 driver into using extended config space Arjan van de Ven
2007-12-22 20:28   ` Stephen Hemminger
2007-12-22 14:20 ` [patch] Make MMCONFIG space (extended PCI config space) a driver opt-in issue Jeff Garzik
2007-12-22 14:47   ` Arjan van de Ven
2007-12-22 15:54     ` Matthew Wilcox
2007-12-22 16:02       ` Arjan van de Ven
2007-12-22 21:10         ` Matthew Wilcox
2007-12-23 23:33           ` Benjamin Herrenschmidt
2007-12-22 18:06     ` Linus Torvalds
2007-12-22 19:30       ` Martin Mares
2007-12-22 20:15         ` Arjan van de Ven
2007-12-22 20:36           ` Martin Mares
2007-12-23  1:29           ` Jeff Garzik
2007-12-23  1:26       ` Jeff Garzik
2007-12-23  3:46         ` Linus Torvalds
2007-12-23  4:11           ` Jeff Garzik
2007-12-23  4:35             ` Linus Torvalds
2007-12-23  4:52               ` Jeff Garzik
2007-12-23  5:00                 ` Linus Torvalds
2007-12-23  5:09                   ` Jeff Garzik
2007-12-23 21:09                   ` Benjamin Herrenschmidt
2007-12-23 21:15                     ` Martin Mares
2007-12-23 22:32                       ` Ivan Kokshaysky
2007-12-23 23:06                       ` Benjamin Herrenschmidt
2007-12-23 23:19                         ` Benjamin Herrenschmidt
2007-12-23  5:27                 ` Loic Prylli
2007-12-23  5:44                   ` Jeff Garzik
2007-12-23  8:33                     ` Loic Prylli
2007-12-23 11:03                     ` Ivan Kokshaysky
2007-12-23 17:32                       ` Linus Torvalds
2007-12-24  7:31                         ` Jeff Garzik
2007-12-24 18:51                           ` Linus Torvalds
2007-12-24 21:22                             ` Matthew Wilcox
2007-12-27 11:46                             ` Jeff Garzik
2007-12-27 14:09                               ` Arjan van de Ven
2007-12-24  7:22                       ` Jeff Garzik
2007-12-24 15:47                         ` Ivan Kokshaysky
2007-12-23 21:13                     ` Benjamin Herrenschmidt
2007-12-23 10:33                 ` Arjan van de Ven
2007-12-24  7:04                   ` Jeff Garzik
2007-12-24 11:49                     ` Arjan van de Ven
2007-12-24 11:54                       ` Jeff Garzik
2007-12-24 12:00                         ` Arjan van de Ven
2007-12-25  9:22                           ` Martin Mares
2007-12-25  9:40                             ` Martin Mares
2007-12-23  4:13           ` Jeff Garzik [this message]
2007-12-23  4:18             ` Jeff Garzik
2007-12-23  4:44               ` Linus Torvalds
2007-12-23  5:04                 ` Jeff Garzik
2007-12-23  5:18                   ` Linus Torvalds
2007-12-23  5:22                     ` Jeff Garzik
2007-12-23  4:40             ` Linus Torvalds
2007-12-23 10:18               ` Martin Mares
2007-12-23  5:09             ` Loic Prylli
2007-12-23  5:57             ` H. Peter Anvin
2007-12-23  1:34     ` Jeff Garzik
2007-12-22 20:43 ` Benjamin Herrenschmidt
     [not found] <fa.pFsY/52FEkQYqrDPnPMxmcUOsRY@ifi.uio.no>
2007-12-22 16:22 ` Robert Hancock
2007-12-22 19:25   ` Greg KH
2007-12-23  0:56     ` 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=476DE07A.4000204@garzik.org \
    --to=jeff@garzik.org \
    --cc=arjan@infradead.org \
    --cc=benh@kernel.crashing.org \
    --cc=gregkh@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@atrey.karlin.mff.cuni.cz \
    --cc=torvalds@linux-foundation.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