On Wed, 2009-07-01 at 16:59 -0700, Linus Torvalds wrote:
> I also _really_ hate how you do
>
> (unsigned long)pte >> VTD_PAGE_SHIFT ==
> (unsigned long)first_pte >> VTD_PAGE_SHIFT
Kill this, in favour of just looking to see if the incremented pte
pointer has 'wrapped' onto the next page. Which means we have to check
it _after_ incrementing it, not before.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This would have found the bug in i386 pci_unmap_addr() a long time ago.
We shouldn't just silently return without doing anything.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Since we're using cmpxchg64() anyway (because that's the only way to do
an atomic 64-bit store on i386), we might as well ditch the extra
locking and just use cmpxchg64() to ensure that we don't add the page
twice.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
As with other functions, batch the CPU data cache flushes and don't keep
recalculating PTE addresses.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The loop condition was wrong -- we should free a PMD only if its
_entire_ range is within the range we're intending to clear. The
early-termination condition was right, but not the loop.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Instead of calling domain_pfn_mapping() repeatedly with single or
small numbers of pages, just pass the sglist in. It can optimise the
number of cache flushes like domain_pfn_mapping() does, and gives a huge
speedup for large scatterlists.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
There's no need for the separate iommu_alloc_iova() function, and
certainly not for it to be global. Remove the underscores while we're at
it.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
As with dma_pte_clear_range(), don't keep flushing a single PTE at a
time. And also micro-optimise the setting of PTE values rather than
using the helper functions to do all the masking.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
It's a bit silly to repeatedly call domain_flush_cache() for each PTE
individually, as we clear it. Instead, batch them up and flush a whole
range at a time. We might as well refrain from recalculating the PTE
address from scratch each time round the loop too.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This is fairly broken anyway -- it doesn't take hotplug into account.
We should probably be checking page_is_ram() instead.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Most of its callers are having to shift for themselves anyway, so we might
as well do it in iommu_flush_iotlb_psi().
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
... and use it in the trivial cases; the other callers want individual
(and bisectable) attention, since I screwed them up the first time...
Make the BUG_ON() happen on too-large virtual address rather than
physical address, too. That's the one we care about.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Use unaligned address for domain->max_addr. That algorithm isn't ideal
anyway -- we should probably just look at the last iova in the tree.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
With some cleanup of intel_unmap_page(), intel_unmap_sg() and
vm_domain_exit() to no longer play with 64-bit addresses.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Add some helpers for converting between VT-d and normal system pfns,
since system pages can be larger than VT-d pages.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
There's no need for the GFX workaround now we have 'iommu=pt' for the
cases where people really care about performance. There's no need to
have a special case for just one type of device.
This also speeds up the iommu=pt path and reduces memory usage by
setting up the si_domain _once_ and then using it for all devices,
rather than giving each device its own private page tables.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
In caching mode, domain ID 0 is reserved for non-present to present
mapping flush. Device IOTLB doesn't need to be flushed in this case.
Previously we were avoiding the flush for domain zero, even if the IOMMU
wasn't in caching mode and domain zero wasn't special.
Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Drop the e820 scanning and use existing function for finding valid
RAM regions to add to 1:1 mapping.
Signed-off-by: Chris Wright <chrisw@redhat.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Identity mapping for IOMMU defines a single domain to 1:1 map all PCI
devices to all usable memory.
This reduces map/unmap overhead in DMA API's and improve IOMMU
performance. On 10Gb network cards, Netperf shows no performance
degradation compared to non-IOMMU performance.
This method may lose some of DMA remapping benefits like isolation.
The patch sets up identity mapping for all PCI devices to all usable
memory. In the DMA API, there is no overhead to maintain page tables,
invalidate iotlb, flush cache etc.
32 bit DMA devices don't use identity mapping domain, in order to access
memory beyond 4GiB.
When kernel option iommu=pt, pass through is first tried. If pass
through succeeds, IOMMU goes to pass through. If pass through is not
supported in hw or fail for whatever reason, IOMMU goes to identity
mapping.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
* git://git.infradead.org/~dwmw2/iommu-2.6.31:
intel-iommu: Fix one last ia64 build problem in Pass Through Support
VT-d: support the device IOTLB
VT-d: cleanup iommu_flush_iotlb_psi and flush_unmaps
VT-d: add device IOTLB invalidation support
VT-d: parse ATSR in DMA Remapping Reporting Structure
PCI: handle Virtual Function ATS enabling
PCI: support the ATS capability
intel-iommu: dmar_set_interrupt return error value
intel-iommu: Tidy up iommu->gcmd handling
intel-iommu: Fix tiny theoretical race in write-buffer flush.
intel-iommu: Clean up handling of "caching mode" vs. IOTLB flushing.
intel-iommu: Clean up handling of "caching mode" vs. context flushing.
VT-d: fix invalid domain id for KVM context flush
Fix !CONFIG_DMAR build failure introduced by Intel IOMMU Pass Through Support
Intel IOMMU Pass Through Support
Fix up trivial conflicts in drivers/pci/{intel-iommu.c,intr_remapping.c}
Conflicts:
arch/mips/sibyte/bcm1480/irq.c
arch/mips/sibyte/sb1250/irq.c
Merge reason: we gathered a few conflicts plus update to latest upstream fixes.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Enable the device IOTLB (i.e. ATS) for both the bare metal and KVM
environments.
Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Make iommu_flush_iotlb_psi() and flush_unmaps() more readable.
Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
PAGE_MASK is 0xFFFFF000 on i386 -- even with PAE.
So it's not sufficient to ensure that you use phys_addr_t or uint64_t
everywhere you handle physical addresses -- you also have to avoid using
the construct 'addr & PAGE_MASK', because that will strip the high 32
bits of the address.
This patch avoids that problem by using PHYSICAL_PAGE_MASK instead of
PAGE_MASK where appropriate. It leaves '& PAGE_MASK' in a few instances
that don't matter -- where it's being used on the virtual bus addresses
we're dishing out, which are 32-bit anyway.
Since PHYSICAL_PAGE_MASK is not present on other architectures, we have
to define it (to PAGE_MASK) if it's not already defined.
Maybe it would be better just to fix PAGE_MASK for i386/PAE?
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In iommu_flush_write_buffer() we read iommu->gcmd before taking the
register_lock, and then we mask in the WBF bit and write it to the
register.
There is a tiny chance that something else could have _changed_
iommu->gcmd before we take the lock, but after we read it. So we could
be undoing that change.
Never actually going to have happened in practice, since nothing else
changes that register at runtime -- aside from the write-buffer flush
it's only ever touched at startup for enabling translation, etc.
But worth fixing anyway.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
As we just did for context cache flushing, clean up the logic around
whether we need to flush the iotlb or just the write-buffer, depending
on caching mode.
Fix the same bug in qi_flush_iotlb() that qi_flush_context() had -- it
isn't supposed to be returning an error; it's supposed to be returning a
flag which triggers a write-buffer flush.
Remove some superfluous conditional write-buffer flushes which could
never have happened because they weren't for non-present-to-present
mapping changes anyway.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
It really doesn't make a lot of sense to have some of the logic to
handle caching vs. non-caching mode duplicated in qi_flush_context() and
__iommu_flush_context(), while the return value indicates whether the
caller should take other action which depends on the same thing.
Especially since qi_flush_context() thought it was returning something
entirely different anyway.
This patch makes qi_flush_context() and __iommu_flush_context() both
return void, removes the 'non_present_entry_flush' argument and makes
the only call site which _set_ that argument to 1 do the right thing.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The domain->id is a sequence number associated with the KVM guest
and should not be used for the context flush. This patch replaces
the domain->id with a proper id value for both bare metal and KVM.
Signed-off-by: Yu Zhao <yu.zhao@intel.com>
Acked-by: Weidong Han <weidong.han@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This updated patch should fix the compiling errors and remove the extern
iommu_pass_through from drivers/pci/intel-iommu.c file.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The patch adds kernel parameter intel_iommu=pt to set up pass through
mode in context mapping entry. This disables DMAR in linux kernel; but
KVM still runs on VT-d and interrupt remapping still works.
In this mode, kernel uses swiotlb for DMA API functions but other VT-d
functionalities are enabled for KVM. KVM always uses multi level
translation page table in VT-d. By default, pass though mode is disabled
in kernel.
This is useful when people don't want to enable VT-d DMAR in kernel but
still want to use KVM and interrupt remapping for reasons like DMAR
performance concern or debug purpose.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Acked-by: Weidong Han <weidong@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
