Siyuan Wang (wangsiyuanbuaa@gmail.com) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/1652
-gerrit
commit 0420ca8bac447417e11ec766dc65ad85ede0c47a Author: Siyuan Wang wangsiyuanbuaa@gmail.com Date: Mon Oct 29 12:08:39 2012 +0800
tyan s8226: change lapic of lapic_cluster 0 to 0x30
amd family15 cpu has a lapic_cluster which is the same level with pci domain. Now, the lapic is 0x20 which is already taken by CPU. there are two CPUs on s8226 and each CPU has 8 cores. CPU 0 takes lapic from 0x10 to 0x17 and CPU 1 takes from 0x20 to 0x27.
Now, the dmesg complains as follow: do_IRQ: 1.55 No irq handler for vector (irq -1)
so I change lapic of lapic_cluster to 0x30. although I am not sure what's the exactly meaning of lapic_cluster, this is a workaround to solve this issue.
Change-Id: Ie17820e0f11d11c35e7e2351a454fa8e823bbf2c Signed-off-by: Siyuan Wang SiYuan.Wang@amd.com Signed-off-by: Siyuan Wang wangsiyuanbuaa@gmail.com --- src/mainboard/tyan/s8226/devicetree.cb | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/src/mainboard/tyan/s8226/devicetree.cb b/src/mainboard/tyan/s8226/devicetree.cb index 4459019..c7bd8c3 100644 --- a/src/mainboard/tyan/s8226/devicetree.cb +++ b/src/mainboard/tyan/s8226/devicetree.cb @@ -19,8 +19,7 @@ chip northbridge/amd/agesa/family15/root_complex device lapic_cluster 0 on chip cpu/amd/agesa/family15 - device lapic 0x20 on end #f15 - #device lapic 0x10 on end #f10 + device lapic 0x30 on end end end device pci_domain 0 on