[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-devel] Possible bug with pass-through hot-plug?


  • To: xen-devel@xxxxxxxxxxxxxxxxxxx
  • From: Tom Rotenberg <tom.rotenberg@xxxxxxxxx>
  • Date: Wed, 7 Oct 2009 11:00:10 +0200
  • Delivery-date: Wed, 07 Oct 2009 02:00:38 -0700
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; b=tusX58QH9md+RnaO8+TJAY9NxBx/s4StID4RbSXhh6E+hOW+EkrKC3HdXHTDlDZ3iI YeSjhC+loCngKs0bpqd5g3yDCdCbQaLLCvGnPCV/z6+PujqaK6au4LPnd/6VAmadMUXO E4jNRZrvLeLmQJmsK6D1uH5bvss2YxALyvPFg=
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>

Hi,

I'm testing the pass-through code, and trying to hot-plug & hot-remove
a PCI device which is in PT, and i think there is a bug in the
hot-plug code.
If i put the pass-through device on a "low" slot number, such as:
00:02.0 - then the hot-plug & hot-remove works, but if i use a "high"
slot number, such as: 00:1b.0, the hot-plug & hot-remove doesn't work.

>From looking at the qemu logs, it looks like the gpe event is "stuck"
and isn't finishing. I suspect, that it's because the gpe eventrelated
procedure in the DSDT, is level triggered (_L03), and thus, if the OS
doesn't handle it fast enough, it keeps getting the interrupt, and is
in kind of  "gpe storm" (that's why a "low" slot number, is working,
because it has less lines of the _L03 function in the DSDT to
evaluate).

Does anyone knows something about it? knows how to fix it?

Tom

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.