Introduce intel_iommu=forcedac commandline option.
This option is helpful to verify the pci device capability
of handling physical dma'able address greater than 4G.
Signed-off-by: Anil S Keshavamurthy <[email protected]>
---
Documentation/kernel-parameters.txt | 7 +++++++
drivers/pci/intel-iommu.c | 6 +++++-
2 files changed, 12 insertions(+), 1 deletion(-)
Index: linux-2.6.22-rc3/Documentation/kernel-parameters.txt
===================================================================
--- linux-2.6.22-rc3.orig/Documentation/kernel-parameters.txt 2007-06-04 12:40:29.000000000 -0700
+++ linux-2.6.22-rc3/Documentation/kernel-parameters.txt 2007-06-04 12:40:41.000000000 -0700
@@ -785,6 +785,13 @@
bypassed by not enabling DMAR with this option. In
this case, gfx device will use physical address for
DMA.
+ forcedac
+ With this option iommu will not optimize to look
+ for io virtual address below 32 bit forcing dual
+ address cycle on pci bus for cards supporting greater
+ than 32 bit addressing. The default is to look
+ for translation below 32 bit and if not available
+ then look in the higher range.
io7= [HW] IO7 for Marvel based alpha systems
See comment before marvel_specify_io7 in
Index: linux-2.6.22-rc3/drivers/pci/intel-iommu.c
===================================================================
--- linux-2.6.22-rc3.orig/drivers/pci/intel-iommu.c 2007-06-04 12:40:29.000000000 -0700
+++ linux-2.6.22-rc3/drivers/pci/intel-iommu.c 2007-06-04 12:40:41.000000000 -0700
@@ -53,6 +53,7 @@
static int dmar_disabled;
static int __initdata dmar_map_gfx = 1;
+static int dmar_forcedac = 0;
#define DUMMY_DEVICE_DOMAIN_INFO ((struct device_domain_info *)(-1))
static DEFINE_SPINLOCK(device_domain_lock);
@@ -69,6 +70,9 @@
} else if (!strncmp(str, "igfx_off", 8)) {
dmar_map_gfx = 0;
printk(KERN_INFO"Intel-IOMMU: disable GFX device mapping\n");
+ } else if (!strncmp(str, "forcedac", 8)) {
+ printk (KERN_INFO"Intel-IOMMU: Enabling DAC for PCI supporting > 32Bit DMA\n");
+ dmar_forcedac = 1;
}
str += strcspn(str, ",");
@@ -1500,7 +1504,7 @@
start_addr = IOVA_START_ADDR;
- if (pdev->dma_mask <= DMA_32BIT_MASK) {
+ if ((pdev->dma_mask <= DMA_32BIT_MASK) || (dmar_forcedac)) {
iova = iommu_alloc_iova(domain, addr, size, start_addr,
pdev->dma_mask);
} else {
--
On Wed, 06 Jun 2007 11:57:05 -0700
[email protected] wrote:
> --- linux-2.6.22-rc3.orig/Documentation/kernel-parameters.txt 2007-06-04 12:40:29.000000000 -0700
> +++ linux-2.6.22-rc3/Documentation/kernel-parameters.txt 2007-06-04 12:40:41.000000000 -0700
> @@ -785,6 +785,13 @@
> bypassed by not enabling DMAR with this option. In
> this case, gfx device will use physical address for
> DMA.
> + forcedac
You want
forcedac [X86-64]
> + With this option iommu will not optimize to look
> + for io virtual address below 32 bit forcing dual
> + address cycle on pci bus for cards supporting greater
> + than 32 bit addressing. The default is to look
> + for translation below 32 bit and if not available
> + then look in the higher range.