LKML Archive on
help / color / mirror / Atom feed
From: Robert Hancock <>
To: linux-kernel <>
Cc: Petr Vandrovec <>,
Subject: Re: [PATCH] Unbreak MSI on ATI devices
Date: Thu, 21 Dec 2006 08:36:39 -0600	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <fa.yZrxrHh1AWLcv/>

Petr Vandrovec wrote:
> Hello Jeff,
>   I'm using second patch below for couple of months to get MSI on all
> devices present on my notebook which are MSI capable (except IDE - notebook
> uses IDE in legacy mode and seems unhappy with transition to native MSI-based
> mode; maybe I could try do the job with libata now when I switched; and VGA, I 
> do not use dri...).  All worked nicely until last sync, when I noticed that my 
> USB devices suddenly stopped working (it took me few weeks as I do not use USB
> regulary). 
> After poking around I've found that problem is that at least ATI USB-HCDs
> apply INTX enable even for MSI, despite warning in the PCI specification that
> it should apply only to MSI (actually I have feeling that on these USB devices 
> disabling INTX in MSI mode drives their INTA# line active as when ohci1394 
> module got loaded kernel complained about interrupt being continuously 
> activated for no good reason (TI's 7421 is one of few MSI-incapable devices
> in my box).
> So my question is - what is real reason for disabling INTX when in MSI mode?
> According to PCI spec it should not be needed, and it hurts at least chips
> listed below:
> 00:13.0 0c03: 1002:4374 USB Controller: ATI Technologies Inc IXP SB400 USB Host Controller
> 00:13.1 0c03: 1002:4375 USB Controller: ATI Technologies Inc IXP SB400 USB Host Controller
> 00:13.2 0c03: 1002:4373 USB Controller: ATI Technologies Inc IXP SB400 USB2 Host Controller 
> I believe that these devices from same vendor accept disabling INTX while in MSI
> fine (I did not notice problems with them even with INTX disabling code in msi.c):
> 00:14.5 0401: 1002:4370 (rev 02) Multimedia audio controller: ATI Technologies Inc IXP SB400 AC'97 Audio Controller (rev 02)
> 00:14.6 0703: 1002:4378 (rev 02) Modem: ATI Technologies Inc ATI SB400 - AC'97 Modem Controller (rev 02)
> None of devices in the box assert INTX while in MSI even if INTX is enabled.
> So I'd like to see first patch below accepted.  If there are some devices which
> require INTX disabling, then apparently decision whether to disable it or no
> has to be moved to device drivers, or some blacklist/whitelist must be created...

Linus' original position on this issue was that any devices which broke 
from disabling INTX when going into MSI mode were just broken and we 
should blacklist MSI entirely for these devices. The reason this change 
went in is that some devices don't automatically disable INTX when MSI 
is turned on (somewhat contrary to the PCI spec apparently).

This whole issue might need to be reopened though..

Robert Hancock      Saskatoon, SK, Canada
To email, remove "nospam" from
Home Page:

       reply	other threads:[~2006-12-21 14:39 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <fa.yZrxrHh1AWLcv/>
2006-12-21 14:36 ` Robert Hancock [this message]
2006-12-21  7:55 Petr Vandrovec
2006-12-21  8:21 ` Jeff Garzik
2006-12-25  4:01 ` Daniel Barkalow
2007-01-04 21:59 ` Roland Dreier
2007-01-05  9:07   ` Petr Vandrovec
2007-01-06  0:54     ` Daniel Barkalow
2007-01-05 23:57   ` Daniel Barkalow

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:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \ \ \ \ \ \
    --subject='Re: [PATCH] Unbreak MSI on ATI devices' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

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).