On Tue, Mar 26, 2013 at 08:40:18PM +0100, Andreas Degert wrote:
> dmesg output appended. I was wrong, there are io page faults
> when I trigger dma transfer with the soundcard, like you expected.
> At the end of the dmesg output are the IO_PAGE_FAULT's from
> trying to do a playback.
Okay, thanks for trying this out. I think I know where the problem is,
can you try the attached patch (without my debug-patch) and report me
whether it fixes the issue?
Thanks,
Joerg
diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index c1c74e0..4647b50 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -2466,18 +2466,16 @@ static int device_change_notifier(struct notifier_block *nb,
/* allocate a protection domain if a device is added */
dma_domain = find_protection_domain(devid);
- if (dma_domain)
- goto out;
- dma_domain = dma_ops_domain_alloc();
- if (!dma_domain)
- goto out;
- dma_domain->target_dev = devid;
-
- spin_lock_irqsave(&iommu_pd_list_lock, flags);
- list_add_tail(&dma_domain->list, &iommu_pd_list);
- spin_unlock_irqrestore(&iommu_pd_list_lock, flags);
-
- dev_data = get_dev_data(dev);
+ if (!dma_domain) {
+ dma_domain = dma_ops_domain_alloc();
+ if (!dma_domain)
+ goto out;
+ dma_domain->target_dev = devid;
+
+ spin_lock_irqsave(&iommu_pd_list_lock, flags);
+ list_add_tail(&dma_domain->list, &iommu_pd_list);
+ spin_unlock_irqrestore(&iommu_pd_list_lock, flags);
+ }
dev->archdata.dma_ops = &amd_iommu_dma_ops;
_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu
