During testing of a large customer UV configuration with a lot of
I/O devices, certain IOMMU functions were either consuming too much
time or not functioning correctly.
This patchset addresses those problems.
Signed-off-by: Mike Travis <[email protected]>
--
When the IOMMU is being used, each request for a DMA mapping requires
the intel_iommu code to look for some space in the DMA mapping table.
For most drivers this occurs for each transfer.
When there are many outstanding DMA mappings [as seems to be the case
with the 10GigE driver], the table grows large and the search for
space becomes increasingly time consuming. Performance for the
10GigE driver drops to about 10% of it's capacity on a UV system
when the CPU count is large.
The workaround is to specify the iommu=pt option which sets up a 1:1
identity map for those devices that support enough DMA address bits to
cover the physical system memory. This is the "pass through" option.
But this can only be accomplished by those devices that pass their
DMA data through the IOMMU (VTd). But Host Bridge Devices connected
to System Sockets do not pass their data through the VTd, thus the
following error occurs:
IOMMU: hardware identity mapping for device 1000:3e:00.0
Failed to setup IOMMU pass-through
BUG: unable to handle kernel NULL pointer dereference at 000000000000001c
This patch fixes that problem but removing Host Bridge devices from
being identity mapped, given that they do not generate DMA ops anyways.
Signed-off-by: Mike Travis <[email protected]>
Reviewed-by: Mike Habeck <[email protected]>
---
drivers/pci/intel-iommu.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
--- linux.orig/drivers/pci/intel-iommu.c
+++ linux/drivers/pci/intel-iommu.c
@@ -46,6 +46,7 @@
#define ROOT_SIZE VTD_PAGE_SIZE
#define CONTEXT_SIZE VTD_PAGE_SIZE
+#define IS_HOSTBRIDGE_DEVICE(pdev) ((pdev->class >> 8) == PCI_CLASS_BRIDGE_HOST)
#define IS_GFX_DEVICE(pdev) ((pdev->class >> 16) == PCI_BASE_CLASS_DISPLAY)
#define IS_ISA_DEVICE(pdev) ((pdev->class >> 8) == PCI_CLASS_BRIDGE_ISA)
#define IS_AZALIA(pdev) ((pdev)->vendor == 0x8086 && (pdev)->device == 0x3a3e)
@@ -2183,7 +2184,7 @@ static int iommu_should_identity_map(str
* take them out of the 1:1 domain later.
*/
if (!startup)
- return pdev->dma_mask > DMA_BIT_MASK(32);
+ return pdev->dma_mask == DMA_BIT_MASK(64);
return 1;
}
@@ -2198,6 +2199,9 @@ static int __init iommu_prepare_static_i
return -EFAULT;
for_each_pci_dev(pdev) {
+ /* Skip PCI Host Bridge devices */
+ if (IS_HOSTBRIDGE_DEVICE(pdev))
+ continue;
if (iommu_should_identity_map(pdev, 1)) {
printk(KERN_INFO "IOMMU: %s identity mapping for device %s\n",
hw ? "hardware" : "software", pci_name(pdev));
--
Subject: Intel iommu: Speed up processing of the identity_mapping function
When there are a large count of PCI devices, and the pass
through option for iommu is set, much time is spent in the
identity_mapping function hunting though the iommu domains to
check if a specific device is "identity mapped".
Speed up the function by checking the cached info to see if
it's mapped to the static identity domain.
Signed-off-by: Mike Travis <[email protected]>
---
drivers/pci/intel-iommu.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
--- linux-2.6.32.orig/drivers/pci/intel-iommu.c
+++ linux-2.6.32/drivers/pci/intel-iommu.c
@@ -2124,10 +2124,10 @@ static int identity_mapping(struct pci_d
if (likely(!iommu_identity_mapping))
return 0;
+ info = pdev->dev.archdata.iommu;
+ if (info && info != DUMMY_DEVICE_DOMAIN_INFO)
+ return (info->domain == si_domain);
- list_for_each_entry(info, &si_domain->devices, link)
- if (info->dev == pdev)
- return 1;
return 0;
}
The __intel_map_single function is not honoring the passed in
DMA mask. This results in not using the coherent DMA mask when
called from intel_alloc_coherent().
Signed-off-by: Mike Travis <[email protected]>
Reviewed-by: Mike Habeck <[email protected]>
Acked-by: Chris Wright <[email protected]>
---
drivers/pci/intel-iommu.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
--- linux.orig/drivers/pci/intel-iommu.c
+++ linux/drivers/pci/intel-iommu.c
@@ -2582,8 +2582,7 @@ static dma_addr_t __intel_map_single(str
iommu = domain_get_iommu(domain);
size = aligned_nrpages(paddr, size);
- iova = intel_alloc_iova(hwdev, domain, dma_to_mm_pfn(size),
- pdev->dma_mask);
+ iova = intel_alloc_iova(hwdev, domain, dma_to_mm_pfn(size), dma_mask);
if (!iova)
goto error;
--