WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-devel

Re: [Xen-devel] xen and pci

To: Adam Sulmicki <adam@xxxxxxxxxxxx>
Subject: Re: [Xen-devel] xen and pci
From: Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>
Date: Sat, 30 Oct 2004 08:43:15 +0100
Cc: Niraj Tolia <ntolia@xxxxxxxxx>, xen-devel@xxxxxxxxxxxxxxxxxxxxx
Delivery-date: Sat, 30 Oct 2004 08:45:05 +0100
Envelope-to: steven.hand@xxxxxxxxxxxx
In-reply-to: Your message of "Fri, 29 Oct 2004 19:07:33 EDT." <20041029190544.I55224-100000@xxxxxxxxxxxxxxxxxxxx>
List-archive: <http://sourceforge.net/mailarchive/forum.php?forum=xen-devel>
List-help: <mailto:xen-devel-request@lists.sourceforge.net?subject=help>
List-id: List for Xen developers <xen-devel.lists.sourceforge.net>
List-post: <mailto:xen-devel@lists.sourceforge.net>
List-subscribe: <https://lists.sourceforge.net/lists/listinfo/xen-devel>, <mailto:xen-devel-request@lists.sourceforge.net?subject=subscribe>
List-unsubscribe: <https://lists.sourceforge.net/lists/listinfo/xen-devel>, <mailto:xen-devel-request@lists.sourceforge.net?subject=unsubscribe>
Sender: xen-devel-admin@xxxxxxxxxxxxxxxxxxxxx
> On Fri, 29 Oct 2004, Niraj Tolia wrote:
> 
> > I believe that Xen hides bridge devices from dom0. However, it should
> > not really affect you.
> 
> well the thing is that I'm trying to run PMCIA ethernet card, and with
> those two lines missing
> 
>       0000:02:00.0 CardBus bridge: Texas Instruments PCI1420
>       0000:02:00.1 CardBus bridge: Texas Instruments PCI1420
> 
> it seems like the yenta driver in xen0 can't find the bridge.

Long term, Xen's PCI code needs some work -- currently it is very
similar to the Linux code and so Xen probes behind all bridges and
reports the results to dom0. It therefore hides the bridges themselves
as otherwise Linux would see some devices twice and get confused.
The solution I think is for Xen not to do any start-of-day probing and
leave that entirely up to Linux. Xen should just tell dom0 where the
root buses are, and let dom0 go from there. Then just need a little
code in Xen to set up IRQ routing when a PCI device is enabled --
there are suitable hooks for thsi already.

*However* I think that pcmcia/cardbus bridges are safe to tell dom0
about (Xen cannot probe behind them) and they have a special subtype
that we could detect. So we can rejig the test that hides the bridges
to be a little bit smarter. 

Please try the following patch, which you can apply by pasting into a
file dev.patch and then:
 'cd xen/common ; chmod 0644 physdev.c ; patch -p1 <path/to/dev.patch'

 -- Keir

--- common/physdev.c    2004-10-22 11:51:50.918741798 +0100
+++ common-new/physdev.c        2004-10-30 08:40:33.513342843 +0100
@@ -747,7 +747,8 @@
          *
          * In Linux2.6 we set pcibios_scan_all_fns().
          */
-        if ( dev->hdr_type != PCI_HEADER_TYPE_NORMAL )
+        if ( (dev->hdr_type != PCI_HEADER_TYPE_NORMAL) &&
+             (dev->hdr_type != PCI_HEADER_TYPE_CARDBUS) )
             continue;
         pdev = xmalloc(sizeof(phys_dev_t));
         pdev->dev = dev;


-------------------------------------------------------
This SF.Net email is sponsored by:
Sybase ASE Linux Express Edition - download now for FREE
LinuxWorld Reader's Choice Award Winner for best database on Linux.
http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel

<Prev in Thread] Current Thread [Next in Thread>