atom feed122 messages in com.xensource.lists.xen-develRe: [Xen-devel] Re: APIC rework
FromSent OnAttachments
49 earlier messages
Konrad Rzeszutek WilkOct 12, 2009 1:02 pm 
Pasi KärkkäinenOct 14, 2009 2:13 pm 
Konrad Rzeszutek WilkOct 15, 2009 1:03 pm 
Boris DerzhavetsOct 16, 2009 12:47 am.gz
Pasi KärkkäinenOct 16, 2009 2:01 am 
Konrad Rzeszutek WilkOct 20, 2009 9:57 am.patch
Pasi KärkkäinenOct 21, 2009 4:53 am 
Konrad Rzeszutek WilkOct 21, 2009 11:31 am 
Pasi KärkkäinenOct 21, 2009 11:51 am 
Jeremy FitzhardingeOct 21, 2009 12:49 pm 
Pasi KärkkäinenOct 21, 2009 1:21 pm 
Pasi KärkkäinenOct 27, 2009 8:46 am 
Konrad Rzeszutek WilkOct 27, 2009 9:59 am.makefile, .c
Pasi KärkkäinenOct 27, 2009 10:29 am 
Konrad Rzeszutek WilkOct 27, 2009 12:40 pm 
Pasi KärkkäinenOct 27, 2009 12:45 pm 
Konrad Rzeszutek WilkOct 27, 2009 1:12 pm 
Pasi KärkkäinenOct 27, 2009 1:17 pm 
Pasi KärkkäinenOct 27, 2009 1:23 pm 
Pasi KärkkäinenOct 27, 2009 1:35 pm 
Jeremy FitzhardingeNov 11, 2009 4:46 pm 
Jeremy FitzhardingeNov 11, 2009 4:59 pm 
Jeremy FitzhardingeNov 12, 2009 3:50 pm 
Zhang, XiantaoNov 12, 2009 9:26 pm 
Keir FraserNov 12, 2009 11:24 pm 
Jeremy FitzhardingeNov 13, 2009 3:56 pm 
Keir FraserNov 14, 2009 12:04 am 
Zhang, XiantaoNov 16, 2009 2:37 am.patch, .patch
Jeremy FitzhardingeNov 16, 2009 10:37 am 
Zhang, XiantaoNov 16, 2009 7:12 pm 
Keir FraserNov 16, 2009 7:44 pm 
Jeremy FitzhardingeNov 16, 2009 9:12 pm 
Jeremy FitzhardingeNov 16, 2009 9:19 pm 
Keir FraserNov 16, 2009 9:43 pm 
Zhang, XiantaoNov 17, 2009 4:45 am.patch
Keir FraserNov 17, 2009 5:04 am 
Zhang, XiantaoNov 17, 2009 6:16 am 
Jeremy FitzhardingeNov 17, 2009 10:50 am 
Keir FraserNov 17, 2009 11:49 am 
Jiang, YunhongNov 17, 2009 7:11 pm 
Zhang, XiantaoNov 17, 2009 7:24 pm 
Zhang, XiantaoNov 17, 2009 7:37 pm 
Keir FraserNov 18, 2009 1:36 am 
Konrad Rzeszutek WilkNov 18, 2009 6:14 am 
Konrad Rzeszutek WilkNov 18, 2009 6:29 am 
Zhang, XiantaoNov 19, 2009 5:45 pm 
Zhang, XiantaoNov 19, 2009 5:47 pm 
Zhang, XiantaoNov 24, 2009 2:04 am.patch, .patch
Jeremy FitzhardingeNov 24, 2009 11:25 am 
Konrad Rzeszutek WilkNov 24, 2009 11:43 am 
Jeremy FitzhardingeNov 24, 2009 3:34 pm 
Zhang, XiantaoNov 24, 2009 5:41 pm 
Zhang, XiantaoNov 24, 2009 6:43 pm 
Konrad Rzeszutek WilkNov 25, 2009 5:41 am 
Konrad Rzeszutek WilkNov 25, 2009 6:09 am 
Zhang, XiantaoNov 25, 2009 7:21 am 
Konrad Rzeszutek WilkNov 25, 2009 10:00 am 
Jeremy FitzhardingeNov 25, 2009 10:58 am 
Jeremy FitzhardingeNov 25, 2009 11:13 am 
Zhang, XiantaoNov 25, 2009 5:11 pm 
Zhang, XiantaoNov 26, 2009 3:52 am 
Konrad Rzeszutek WilkNov 30, 2009 6:26 am 
Konrad Rzeszutek WilkNov 30, 2009 6:33 am 
Zhang, XiantaoDec 2, 2009 6:13 pm 
Konrad Rzeszutek WilkDec 3, 2009 6:37 am 
Stefan KuhneDec 4, 2009 8:07 am 
Pasi KärkkäinenDec 4, 2009 10:57 am 
Jeremy FitzhardingeDec 4, 2009 11:26 am 
Pasi KärkkäinenJan 1, 2010 9:20 am 
Konrad Rzeszutek WilkJan 4, 2010 5:37 am 
Pasi KärkkäinenJan 4, 2010 11:41 am 
Konrad Rzeszutek WilkJan 14, 2010 12:05 pm 
Pasi KärkkäinenJan 14, 2010 11:18 pm 
Subject:Re: [Xen-devel] Re: APIC rework
From:Konrad Rzeszutek Wilk (konr@oracle.com)
Date:Nov 24, 2009 11:43:38 am
List:com.xensource.lists.xen-devel

At least dom0 parses this info from DSDT, so we can't have the assuption whether
it is used or not, I think. And I also agree to add a new physdev_op to handle
this case, and it should be better way to go. Based on this idea, I worked out the patch, attached! In this patch, we
introduced a new physdev_op PHYSDEVOP_setup_gsi for each GSI setup, and each
domain can require to map each GSI in this case. In addition, I believe it is very safe to port the hypervisor patch to xen-3.4-x
tree and keeps pv_ops dom0 running on it, since no logic is changed. BTW, I
also tested apic and non-apic cases, they works fine after applying the patches.

But I don't think you tested PCI front and PCI back.

Mainly these lines worry me (can you inline the patch next time too, please):

+ map_irq.domid = DOMID_SELF; + map_irq.type = MAP_PIRQ_TYPE_GSI; + map_irq.index = gsi; + map_irq.pirq = irq; + rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);

For PCI passthrough to work, the domid needs to be for the guest domain, while in this case it is set to Dom0.

There is already a method of extracting the domain id for PCI devices passed to the guest. Look in the 'xen_create_msi_irq' function.