On Wed, May 30, 2018, 15:55 Peter Grehan wrote:
> >> If bhyve has its own attachment id, it is trivial
> >> to add a special case on it quickly, too.
> >
> > vendor_id -- 0x1275
>
> Actually, this can also be AMD. A better check would be if the system
> is running virtualized.
>
I'm
If bhyve has its own attachment id, it is trivial
to add a special case on it quickly, too.
vendor_id -- 0x1275
Actually, this can also be AMD. A better check would be if the system
is running virtualized.
later,
Peter.
___
Hi Justin,
I can add one either tonight or tomorrow.
Thanks.
If bhyve has its own attachment id, it is trivial
to add a special case on it quickly, too.
vendor_id -- 0x1275
later,
Peter.
___
svn-src-head@freebsd.org mailing list
Hi Peter,
On Wed, May 30, 2018, 15:44 Peter Grehan wrote:
> Hi Nathan,
>
> > There are a ton of ARM boards that need this too. You can find one-off
> > hacks all through the tree and one of the nice things about this change
> > is that all of those can be consolidated/removed now. If we are
Hi Nathan,
There are a ton of ARM boards that need this too. You can find one-off
hacks all through the tree and one of the nice things about this change
is that all of those can be consolidated/removed now. If we are going to
have some #ifdef and special cases, it would be better to make
On 05/30/18 02:42, Peter Grehan wrote:
PCIe only permits 1 device on an endpoint, so some devices ignore
the device
part of B:D:F probing. Although ARI likely fixes this, not all
platforms
support ARI completely or correctly, so some devices end up
showing up 32
times on the
On Wed, May 30, 2018 at 4:42 AM, Peter Grehan wrote:
>>PCIe only permits 1 device on an endpoint, so some devices ignore the
>> device
>>part of B:D:F probing. Although ARI likely fixes this, not all
>> platforms
>>support ARI completely or correctly, so some devices end up showing
PCIe only permits 1 device on an endpoint, so some devices ignore the device
part of B:D:F probing. Although ARI likely fixes this, not all platforms
support ARI completely or correctly, so some devices end up showing up 32
times on the bus.
I think this might have broken bhyve -
Author: jhibbits
Date: Wed May 30 02:41:47 2018
New Revision: 334365
URL: https://svnweb.freebsd.org/changeset/base/334365
Log:
Restrict PCIe maxslots to 0, instead of PCI_SLOTMAX
Summary:
PCIe only permits 1 device on an endpoint, so some devices ignore the device
part of B:D:F