This series implements several fixes for intel-iommu. The first
is to make use of the device coherent_dma_mask when allocating
coherent DMA buffers, for which we move dma_generic_alloc_coherent
out of the x86 code so we can use it by both of the current users
of intel-iommu. Next, we add a little more to detecting when a
device can't support passthrough mode. Then we reinstate RMRRs
for devices that get kicked out of passthrough mode. And finally,
a trivial printk change to be less verbose on boot.
Ideally I'd like to entertain the idea of getting these into 2.6.32
because the coherent mapping issue is actually a regression since
2.6.31 and will cause some devices to fail to initialize in passthrough
mode (cciss for one). Thanks,
Alex
---
Alex Williamson (5):
intel-iommu: Quiet unnecessary output
intel-iommu: Reinstate RMRRs if a device is removed from passthrough domain
intel-iommu: Use max_pfn to determine whether a device can passthrough
intel-iommu: Use dma_generic_alloc_coherent() for passthrough mappings
dma: create dma_generic_alloc/free_coherent()
arch/x86/include/asm/dma-mapping.h | 3 --
arch/x86/kernel/pci-dma.c | 31 --------------------
arch/x86/kernel/pci-nommu.c | 10 +++++--
drivers/pci/intel-iommu.c | 55 +++++++++++++++++++++++++++++++++---
include/linux/dma-mapping.h | 44 +++++++++++++++++++++++++++++
5 files changed, 102 insertions(+), 41 deletions(-)
--
To unsubscribe from this list: send the line "unsubscribe linux-pci" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
